CADTH Reimbursement Review

Belimumab (Benlysta)

Sponsor: GlaxoSmithKline Inc.

Therapeutic area: Lupus nephritis

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Lupus is an autoimmune disease characterized by inflammatory processes that can occur in various tissues and organs of the body.¹ A common form of lupus is systemic lupus erythematosus (SLE)² with an estimated prevalence of about 1 in 2,000 individuals in Canada.^3,4 The age of onset is primarily between 16 and 55 years,⁵ with females more commonly affected than males (9:1).⁴ The median ages at diagnosis for white females range from 37 to 50 years, in white males from 50 to 59, in Black females from 15 to 44, and in Black males from 45 to 64.⁵ Kidney injury is common in SLE, with lupus nephritis (LN) occurring in about 50% of patients with SLE,⁶ usually within 5 years of SLE diagnosis.⁷ Kidney involvement can remain silent or asymptomatic for a significant period of time;⁸ however, patients may experience fatigue, joint and muscle pain, edema, rash, and a variety of other symptoms.⁹ The disease is associated with substantial morbidity⁶ as serious complications include progression to end-stage renal disease (ESRD),¹⁰ in which patients require dialysis or kidney transplant.⁸

Initial treatment options for induction of class III, IV, and/or V active LN include a high-dose corticosteroid taper as well as immunosuppressive drugs such as mycophenolate mofetil (or mycophenolic acid) or cyclophosphamide.^8,11 The clinical expert consulted by CADTH for this review noted that other treatments in addition to standard of care for patients with an inadequate response to first-line induction therapy may include off-label use of rituximab, cyclosporin, or tacrolimus. The clinical expert for this review stated that in all cases of class III, IV, and/or V active LN, use of antimalarials (i.e., hydroxychloroquine), bone protection (vitamin D, calcium, possibly antiresorptive drugs), immunizations with nonlive vaccines, and adjunct treatment with renin-angiotensin blockade and statins should be considered.

Belimumab inhibits the B lymphocyte stimulator protein and thus reduces B-cell activity.¹² The IV administration is 10 mg/kg, administered over an hour, at 2-week intervals for the first 3 doses and at 4-week intervals thereafter in addition to standard of care therapy for patients with active LN. The recommended dose for subcutaneous injection, in addition to standard of care therapy, is 400 mg (two 200 mg injections) once weekly for 4 doses, then 200 mg once weekly thereafter for the treatment of adult patients with active LN.

The objective of this report is to perform a systematic review of the beneficial and harmful effects of belimumab IV infusion or subcutaneous injection in addition to standard of care therapy for the treatment of active LN in adults.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups who responded to CADTH’s call for patient input and from the clinical expert consulted by CADTH for the purpose of this review.

Patient Input

Four responses to CADTH’s call for patient input for the belimumab submission were received. These consisted of submissions from Arthritis Consumer Experts, Lupus Ontario, a joint submission from the Kidney Foundation of Canada and Lupus Canada, and a cooperative submission from the Canadian Arthritis Patient Alliance, the Arthritis Society, the Canadian Skin Patient Alliance, and CreakyJoints. Patient input was gathered from surveys, video interviews, and focus group discussions among lupus patients across Canada: 34 respondents (88% female) from Arthritis Consumer Experts, 10 respondents (90% female) with SLE from Lupus Ontario, and 38 respondents (73% with LN and approximately 15% caregivers) from the Kidney Foundation of Canada and Lupus Canada. The cooperative submission conducted a focus group of 3 patients with LN as well as a video interview with 1 patient. The submission from Arthritis Consumer Experts also included an in-depth interview with 1 patient. Seventeen patients (6 from a previous survey) in the included submissions had experience with the treatment under review.

Patients reported managing SLE was difficult given the severity of the physical symptoms, such as debilitating fatigue, joint pain, flares, skin rashes, nausea, loss of appetite, bruising, back pain, brain fog, mobility issues, and mental health issues. Respondents reported that currently available treatments are difficult to tolerate because of the many side effects. While describing their experiences with the drug under review, patients reported both positive and negative outcomes. Some patients described experiencing side effects such as severe allergic reaction, extreme nausea, sleep deprivation, frequent urinary tract infections (UTIs), depression, and psychosis. Other patients reported an overall decrease in their disease symptoms and improvement in physical ability, leading to improvement in their health-related quality of life (HRQoL).

The key outcomes patients would like addressed by a new therapy are reduction of side effects and number of medications used; reduction in fatigue, flares, pain, and rash and skin irritations; increased mobility and participation in physical activities; overall improvement in HRQoL; improved engagement in social activities; and better affordability of medication.

Clinician Input

Input From the Clinical Expert Consulted by CADTH

According to the clinical expert consulted by CADTH for this review, response to current standard of care induction therapy is suboptimal (only 20% to 35% of patients achieve a complete renal response [CRR] within 6 months to 12 months after initiation of induction therapy; of those who do respond, 20% to 35% relapse within 3 years to 5 years), indicating a major unmet need as up to 40% of patients with LN can develop chronic kidney disease and progress to ESRD, requiring dialysis or transplant. Other unmet needs include lack of adherence, side effects (e.g., with prednisone), and recurrent flares that cause progressive organ damage, and only a few treatments are safe in pregnancy in a disease that largely affects those who can become pregnant. Currently, no treatments provide a long-term cure or long-term medication-free survival. According to the clinical expert, the current place in therapy for belimumab would be as add-on therapy to existing standard of care (i.e., corticosteroids and mycophenolate mofetil/mycophenolic acid or cyclophosphamide) in patients with class III or IV (with or without class V) or pure class V active LN who have not attained an adequate renal response after 2 months to 3 months of induction therapy; however, the clinical expert noted that the time to initiate belimumab may vary from start of induction therapy to 3 months to 6 months after initiation of induction therapy, dependent on disease severity, patient response, and expert physician judgment. Other factors that may identify active patients most likely to respond to belimumab include those with previous episodes of class III or IV (with or without class V) or class V LN in whom another flare may cause a serious decline in renal function; patients with active class III or IV (with or without class V) or class V LN with chronically impaired renal function or who still have prednisone use greater than 7.5 mg/day after 3 to 6 months of induction; or who have extrarenal manifestations in addition to LN. The clinical expert identified those least likely to benefit from belimumab as including patients with active LN who are not currently receiving standard of care induction therapy, patients for whom induction with both mycophenolate mofetil/mycophenolic acid and cyclophosphamide drugs has failed, and patients with an estimated glomerular filtration rate (eGFR) less than or equal to 30 mL/min/1.73 m². (It is unknown if belimumab would be efficacious in these patients as they were excluded from the BLISS-LN trial.) Further, patients who have attained a CRR after 6 months to 12 months of induction therapy will probably derive little incremental benefit from belimumab as add-on to standard of care and should not be considered as candidates for belimumab.

In the opinion of the clinical expert, a clinically meaningful response to belimumab would include, sequentially, at least a 25% reduction in proteinuria (as defined by a urine protein-creatinine ratio [uPCR]) after 2 to 3 months of induction therapy, at least a 50% reduction in proteinuria after 6 months of therapy, reduction in corticosteroids to less than or equal to 7.5 mg/day after 6 months to 12 months of therapy, proteinuria no greater than 0.5 g per 24 hours to 0.7 g per 24 hours after 12 months of therapy (the response time can be delayed to 18 months to 24 months if baseline proteinuria is in the nephrotic range [i.e., > 3.5 g per 24 hours]), and an eGFR no worse than 10% to 20% of pre-flare value and greater than or equal to 60 mL/min/1.73 m² after 12 months of therapy.

Clinician Group Input

The Canadian Network for Improved Outcomes for Systemic Lupus Erythematosus and 31 associated physicians provided input for this review.

The clinician group and physicians agreed that there are some treatment gaps and unmet needs in the current LN therapeutics. These unmet needs include inability to achieve complete remission from existing treatment options (e.g., mycophenolate mofetil and cyclophosphamide); increased risk of multiple complications from moderate or high doses of corticosteroids; subsequent ESRD and renal replacement therapy associated with disease flares; and difficulty in maintaining adherence.

The clinician group and physicians indicated that a clinically meaningful response to treatment would include any of the following: complete remission (proteinuria less than 0.5 g per 24 hours) within 12 months of starting treatment, reduction in daily prednisone dose to levels lower than 7.5 mg, or reduction in the frequency and intensity of flares. According to the clinician group, sufficient time (at least 12 months) should be allowed for these outcomes to be observed and treatment should be discontinued after 12 months in cases where no response can be demonstrated. The input stated that belimumab is expected to cause a shift in the current treatment paradigm for LN by addressing the disease mechanism. Its ability to modulate the maturation and functional differentiation of B cells, which produce autoantibodies that are central in SLE pathogenesis and tissue damage, renders it most suitable for patients whose LN has not reached at least partial remission, patients experiencing early and frequent flares, patients with steroid-dependent disease, and patients with adherence issues.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The drug plans identified implementation issues related to considerations for initiation, renewal, discontinuation, and prescription of therapy, as well as generalizability and system and economic issues. The clinical expert consulted by CADTH weighed evidence from the BLISS-LN trial and other clinical considerations to provide responses to the drug programs’ implementation questions. Refer to Table 4 for more details.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

One double-blind, placebo-controlled, randomized controlled trial (BLISS-LN) was included in this review (107 sites in 21 countries; N = 448) evaluating the efficacy, safety, and tolerability of an IV treatment regimen of belimumab 10 mg/kg in adult patients with class III or IV (with or without the presence of class V) or pure class V active LN while receiving standard of care. The primary objective was to evaluate the effect of belimumab 10 mg/kg compared to placebo in renal response as measured by the difference in the proportion of patients who achieve a primary efficacy renal response (PERR) at week 104. The key secondary objectives were to evaluate the effect of belimumab 10 mg/kg compared to placebo on CRR at week 104, PERR at week 52, time to renal-related event or death, and ordinal renal response (ORR) at week 104. Patients were defined as having treatment failure if they did not follow the corticosteroid rules (i.e., failed to taper corticosteroids to ≤ 10 mg/day by week 24 and to not exceed this dose of 10 mg/day through week 104); if they received additional immunosuppressive drugs (except topical drugs) beyond the induction and maintenance regimens; if the use of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or antimalarial drugs was initiated after week 24; or if the patient’s standard therapy (cyclophosphamide followed by azathioprine or mycophenolate mofetil) exceeded permitted doses. Baseline patient characteristics, including age, sex, and race, were balanced between groups. Patients were predominantly female (88.3% in the belimumab group and 87.9% in the placebo group) and predominantly Asian (51.1% in the belimumab group and 48.9% in the placebo group), and 84% of patients in both the belimumab and placebo groups were categorized as being in renal biopsy class III or IV (with or without class V) according to the local reader. Disease characteristics such as SLE and LN disease duration, mean score for the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) with modified scoring for proteinuria (SLEDAI-S2K), mean eGFR, and mean uPCR were balanced between treatment groups at baseline. There were some baseline differences in use of concomitant medications between groups, including of antimalarials (74.4% and 69.1%) and of prednisone (steroids were converted to prednisone equivalent), with a mean dose of 66.5 (standard deviation [SD] = 99.6) mg/day and 72.5 (SD = 133.2) mg/day for the belimumab and placebo groups, respectively. Patients also used immunosuppressants (88.8% in the belimumab group and 86.1% in the placebo group) and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (65.9% in the belimumab group and 67.3% in the placebo group) at baseline.

Efficacy Results

In BLISS-LN, the primary end point, PERR at week 104, was statistically significant in favour of the belimumab group (43.0% versus 32.3%), with an adjusted between-group difference of 10.66% (95% confidence interval [CI], 1.89 to 19.42; P = 0.0311). In addition, statistically significant differences in favour of the belimumab group were reported for all key secondary end points of CRR at week 104 (30.0% versus 19.7%), with an adjusted between-group difference of 10.27% (95% CI, 2.40 to 18.14; P = 0.0167); PERR at week 52 (46.6% versus 35.4%), with an adjusted between-group difference of 11.12% (95% CI, 2.25 to 19.99; P = 0.0245); ORR at week 104 (30.0% belimumab versus 19.7% placebo achieving a complete response), with an adjusted between-group difference of 10.27% (95% CI, 2.40 to 18.14; P = 0.0167); and time to renal-related event or death (15.7% belimumab versus 28.3% placebo experiencing an event), with a hazard ratio (HR) of 0.51 (95% CI, 0.34 to 0.77; P = 0.0014).

Subgroup analyses based on baseline renal biopsy class (class III or class IV, versus class III and V or class IV and V, versus class V) and the induction regimen (cyclophosphamide versus mycophenolate mofetil) for the primary and the key secondary end points were generally consistent with the overall results for all subgroups except for the baseline renal biopsy class V subgroup and the cyclophosphamide followed by azathioprine subgroup, the results of which were not statistically significantly different between treatment groups. However, the study was not designed or powered to evaluate efficacy in subgroups, and the small number of patients in the class V and the cyclophosphamide followed by azathioprine subgroups might have led to the lack of statistical significance between treatment groups.

In terms of secondary outcomes, a higher proportion of patients in the belimumab group compared with the placebo group received an average daily prednisone dose of less than or equal to 7.5 mg at week 104 since the previous 4-week visit (40.8% versus 29.6%), with an odds ratio (OR) of 1.65 (95% CI, 1.11 to 2.45). In terms of disease activity, the least squares (LS) mean change from baseline in SLEDAI-S2K at week 104 was –7.7 (standard error [SE] = 0.46) in the belimumab group and –6.1 (SE = 0.47) in the placebo group, with an LS mean difference of –1.5 (95% CI, –2.4 to –0.6). The proportion of patients with a SLEDAI-S2K score less than 4 at week 104 was higher in the belimumab group compared to the placebo group (27.8% versus 18.4%), with an OR of 1.76 (95% CI, 1.11 to 2.78). A higher proportion of patients experienced a severe flare postbaseline through to week 104 in the placebo group (31.4%) than in the belimumab group (18.8%), and the risk of experiencing a severe flare at any time, based on the Safety of Estrogens in Lupus Erythematosus – National Assessment (SELENA) SLEDAI Flare Index (SFI), was lower in patients in the belimumab group compared with patients in the placebo group (HR = 0.57; 95% CI, 0.39 to 0.84).

No HRQoL data were assessed in the BLISS-LN trial.

Harms Results

The key harms reported in the BLISS-LN trial are summarized in Table 2.

Table 2: Summary of Key Results From the BLISS-LN Trial

CI = confidence interval; CMH = Cochran-Mantel-Haenszel; CRR = complete renal response; eGFR = estimated glomerular filtration rate; GFR = glomerular filtration rate; HR = hazard ratio; LS = least squares; NMSC = nonmelanoma skin cancer; OR = odds ratio; ORR = ordinal renal response; PERR = primary efficacy renal response; SAE = serious adverse event; SD = standard deviation; SE = standard error; SFI = SELENA SLEDAI Flare Index; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria; uPCR = urine protein-creatinine ratio.

^aCMH estimates are adjusted for induction regimen (cyclophosphamide vs. mycophenolate mofetil) and race (Black vs. non-Black).

^bOR (95% CI) and P value are from a logistic regression model for comparison between belimumab and placebo, with covariates treatment group, induction regimen (cyclophosphamide vs. mycophenolate mofetil), race (Black vs. non-Black), baseline uPCR, and baseline eGFR.

^cStudy withdrawal, treatment failures, and investigational product discontinuation imputed as “no response.” Investigational product discontinuation and treatment failure not related to renal disease or study withdrawal were censored in the time-to-event analysis. A CRR: uPCR was < 0.5 g/g; eGFR no more than 10% below pre-flare GFR or within normal range; no treatment failure. A partial renal response: ≥ 50% decrease from baseline in uPCR and 1 of the following: uPCR value < 1 g/g if baseline value was ≤ 3 g/g, or uPCR value < 3 g/g if the baseline value was > 3 g/g; eGFR no more than 10% below baseline GFR or within normal range; and no treatment failure.

^dEvents are defined as the first event experienced among the following: death, progression to end-stage renal disease, doubling of serum creatinine from baseline, renal worsening, or renal-related treatment failure. Patients who discontinue randomized treatment, withdraw from the study, or are lost to follow-up are censored on that date. Patients who completed the 104-week treatment period are censored at the week 104 visit. Time to event is defined as (event date – treatment start date + 1).

^eFrom a Cox proportional hazards model for comparison between belimumab and placebo, adjusting for induction regimen (cyclophosphamide vs. mycophenolate mofetil), race (Black vs. non-Black), baseline uPCR, and baseline eGFR.

^fWeek 104 statistics are from an analysis of covariance model comparing belimumab and placebo, with covariates for treatment group, baseline SLEDAI-S2K score, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^gP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

^hOR (95% CI) and P value are from a logistic regression model for comparison between belimumab and placebo, with covariates treatment group, baseline SLEDAI-S2K score, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

ⁱSteroids were converted to prednisone equivalent. The prednisone average daily dose since previous visit was calculated at every 4-week visit after baseline. All prednisone doses since the visit 4 weeks prior, up to and including the current visit, were summed and divided by the number of days in the period. Days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation for average daily prednisone dose.

^jOR (95% CI) and P value are from a logistic regression model for comparison between belimumab and placebo, with covariates treatment group, baseline prednisone dose, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^kAnalysis excludes severe flares that were triggered only by an increase in SELENA SLEDAI score to more than 12. Treatment failure is considered an event. Patients who discontinue randomized treatment, withdraw from study, are lost to follow-up, die, or complete week 104 are censored at the last flare assessment date, death date, or the week 104 study visit. Time to first severe flare is defined as (event date – treatment start date + 1) .

^lOnly includes postbaseline severe flares.

^mFrom a Cox proportional hazards model for comparison between belimumab and placebo, adjusting for induction regimen (cyclophosphamide vs. mycophenolate mofetil) and race (Black vs. non-Black).

ⁿIncludes all deaths that occurred during the double-blind phase including off treatment.

Source: Clinical Study Report for BLISS-LN.¹³

Rates of adverse events (AEs) were similar across treatment groups (95.5% belimumab versus 94.25% placebo). Frequent AEs that occurred more commonly in the belimumab group than in the placebo group were UTI (19.2% versus 15.6%), cough (12.5% versus 8.5%), and upper abdominal pain (6.3% versus 2.7%). The proportion of patients experiencing at least 1 serious AE (SAE) was similar between treatment groups (25.9% belimumab versus 29.9% placebo). The most common SAEs were pneumonia (4.0% versus 3.1%), herpes zoster (1.8% versus 0.9%), gastroenteritis (0% versus 2.2%), lung infection (0.9% versus 1.3%), LN (0.4% versus 1.8%), and UTI (1.3% versus 0.9%). A similar number of withdrawals due to AEs occurred in the belimumab group and the placebo group (12.9% versus 12.9%), with the most common reason for withdrawal in both groups being pneumonia (4.0% versus 3.1%).

Eleven deaths occurred during the double-blind phase of the BLISS-LN trial, mainly due to infections, with 6 deaths in the belimumab group (2.7%) and 5 deaths the placebo group (2.2%).

Common notable harms in the BLISS-LN trial included postinfusion-related systemic reactions (11.6% belimumab versus 12.9% placebo); serious infections of herpes zoster (2.2% versus 0.9%), active tuberculosis (0.9% versus 0.4%), and sepsis (0% versus 0.4%); malignancies (including nonmelanoma skin cancer) (1.3% versus 0%), and serious suicidal behaviour (0.4% versus 0%).

Critical Appraisal

In terms of limitations, a greater proportion of patients discontinued from the placebo group than the belimumab group, which may have led to bias in the results in favour of belimumab. However, the sensitivity analyses that assessed the impact of missing data generally showed results supportive of the primary analysis. Regarding calculations of patients’ average daily prednisone dose in the BLISS-LN trial, days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation, which would likely have underestimated the average dose of prednisone used in the study and may also have led to bias, although the direction of bias is unknown. Improvements in HRQoL were identified as important outcomes by the patient groups providing input for this review. However, no HRQoL data were collected in the BLISS-LN trial; hence, it is unknown what impact belimumab would have on HRQoL.

The product monograph for belimumab authorized both IV and subcutaneous formulations for LN.¹² However, the approval of subcutaneous formulations was based on extrapolated data, and there is no clinical evidence regarding the subcutaneous formulation for patients with active LN.

Overall, the clinical expert consulted felt the characteristics of the patient population enrolled in the trials was a good representation of the target population and did not identify any issues with the use of concurrent treatments or conduct of the trial that could substantially affect the generalizability of the findings.

Indirect Comparisons

No indirect evidence was available.

Other Relevant Evidence

Data from 1 open-label extension (OLE) study were summarized in this report.

Description of Studies

The OLE study provided supplemental safety and efficacy data for patients who received IV belimumab 10 mg/kg plus standard of care for up to 28 weeks (N = 254) among eligible patients who completed all assessments at week 104 in the BLISS-LN trial. Patients received the first dose at week 104 of the double-blind phase of the BLISS-LN trial (marked as day 0 for the open-label phase). There were 2 groups in the extension phase: the placebo-to-belimumab group (patients switching from placebo to belimumab) and the belimumab-to-belimumab group (patients remaining on belimumab). Criteria for the open-label phase allowed for the use of concomitant medications including immunosuppressants and corticosteroids, which were prohibited in the BLISS-LN trial. Also, PERR and CRR were evaluated based on observed data at open-label weeks 12, 24, and 28, and criteria were required to be met at a single time point only, meaning criteria did not have to be met on consecutive visits as was required for the double-blind phase of the BLISS-LN trial.

Efficacy Results

In the OLE study, results found that the proportion of patients achieving the PERR increased from baseline to week 28 in both the belimumab-to-belimumab group (from 71% to 75%) and the placebo-to-belimumab group (from 60% to 67%) when using the open-label phase criteria. Post hoc analyses found that when using the pivotal trial-defined criteria for a PERR, the proportion of patients achieving a PERR from baseline to week 28 decreased in the belimumab-to-belimumab group (from 66% to 52%) and remained stable in the placebo-to-belimumab group (from 54% to 53%). Similar results were found for the proportion of patients who had a CRR. Reductions in PERR and CRR rates at open-label week 28 in the belimumab-to-belimumab group were mainly due to discontinuations (n = 8) or intake of concomitant medications (n = 9) allowed during the OLE phase but counted as treatment failures for the post hoc statistical analysis. Median uPCR and eGFR remained similar at baseline and at week 28 in both groups. There were no marked changes in the proportion of patients with SLEDAI-S2K scores less than 4 or in the proportion of patients receiving an average daily prednisone-equivalent dose of less than or equal to 7.5 mg in either group from baseline to week 28.

Harms Results

The proportion of patients experiencing at least 1 AE in the open-label phase was higher in the belimumab-to-belimumab group (70%) than in the placebo-to-belimumab group (62%). The most common AEs occurring in at least 5% of patients in either group included infections and infestations (49% versus 42%), musculoskeletal and connective tissue disorders (12% versus 13%), skin and subcutaneous tissue disorders (13% versus 8%), gastrointestinal disorders (10% versus 9%), and respiratory, thoracic, and mediastinal disorders (11% versus 4%) in the belimumab-to-belimumab versus placebo-to-belimumab groups, respectively. The number of patients experiencing at least 1 SAE during the open-label phase was low (8% in the belimumab-to-belimumab group and 4% in the placebo-to-belimumab group). The number of withdrawals due to AEs was also very low in both groups (3% versus 0.8%). Common notable harms included postinfusion systemic reactions (4% versus 3%) and infections of special interest (opportunistic infections, herpes zoster, tuberculosis, sepsis) (5% versus 2%) in the belimumab-to-belimumab versus placebo-to-belimumab groups, respectively. Two serious infections of special interest were reported in the belimumab-to-belimumab group, 1 for serious tuberculosis and another for serious disseminated herpes zoster. One case of suicidal behaviour occurred in a patient diagnosed with an adjustment disorder. This patient recovered and completed the treatment throughout the open-label phase. One death, deemed SLE related, occurred during the open-label phase in the placebo-to-belimumab group.

Critical Appraisal

The extension study allowed for the investigation of the long-term efficacy and harms associated with belimumab for an additional 28 weeks for eligible patients who completed the BLISS-LN trial. As there was no active comparator and all outcomes were descriptive in nature, it is difficult to make any inferences regarding the results. Furthermore, extension studies are often limited by selection bias, as only patients who are tolerant to treatment and complete the parent studies are eligible to enrol in the OLE study. An additional limitation is the open-label nature of treatment, which can bias the reporting of subjective end points (i.e., harms). Lastly, the relatively short duration of the OLE study is insufficient to observe appreciable benefit among those who had transitioned from placebo to belimumab.

Conclusions

In adult patients with class III, IV, and/or V active LN, treatment with belimumab 10 mg/kg in addition to standard of care statistically significantly improved renal response as measured by the primary outcome PERR, relative to placebo, based on 104-week data from the BLISS-LN trial, the results of which were deemed to be clinically meaningful by the clinical expert consulted for this review. All key secondary outcomes, including CRR at week 104, PERR at week 52, ORR at week 104, and time to renal-related event or death, showed statistically significant differences in favour of belimumab. The trial showed that a greater proportion of patients in the belimumab group than in the placebo group had reductions in average daily prednisone use to less than 7.5 mg since the previous 4-week visit and showed improvements in mean SLEDAI-S2K score. Also, fewer patients in the belimumab group experienced severe flares than those in the placebo group. HRQoL was not assessed in the trial, and therefore the impact of belimumab on HRQoL in patients with LN is unknown. In addition, the long-term efficacy of belimumab on reducing flare rates is unknown. AEs, including infections, occurred with similar frequency in the 2 treatment groups. Data from the pivotal trial and the OLE study do not suggest issues of tolerability or safety, although the extension study was limited by lack of a control group.

Introduction

Disease Background

Lupus is an autoimmune disease characterized by inflammatory processes that can occur in various tissues and organs of the body.¹ A common form of lupus is SLE,² with an estimated prevalence of about 1 in 2,000 individuals in Canada.^3,4 The age of onset is primarily between 16 years and 55 years, with females more commonly affected than males (9:1).⁵ Kidney involvement is common in SLE, and LN occurs in about 50% of lupus patients,⁶ usually within 5 years of SLE diagnosis.⁷ Kidney involvement can remain silent or asymptomatic for a significant period of time;⁸ however, patients may experience fatigue, joint and muscle pain, edema, rash, and a variety of other symptoms.⁹ The typical disease course of LN is characterized by episodes of flares in between periods of disease inactivity.¹⁴ LN flares and uncontrolled disease activity between flares contribute to accumulating nephron loss and progressive decline in renal function.¹⁵ It is estimated that about 5% to 20% of patients with LN will progress to ESRD within 10 years of an initial SLE diagnosis,¹⁰ eventually requiring dialysis or kidney transplant.⁸ Those with proliferative LN have a significantly higher rate of mortality than those without LN, with about 5% to 25% of patients dying due to kidney disease within 5 years of onset.⁶ Patients with SLE who develop LN usually present at a younger age than those without LN,⁶ and evidence suggests that there is a slightly higher prevalence of LN in males with SLE than females with SLE (1.1:1 to 1.7:1).¹⁰ Various cohort studies, including those with Canadian data,¹⁶ suggest that the prevalence of LN is higher in Black, Hispanic, and Asian individuals than in white individuals¹⁰ and that Hispanic and Black patients are more likely to progress to kidney failure than white patients.⁶ Guidelines state the importance of early therapy to prevent progressive kidney damage while reducing medication-associated toxicities.⁸ According to the clinical expert consulted for this review, diagnosis, treatment, and monitoring of patients with LN should be by a specialist physician, such as a rheumatologist or nephrologist.

Standards of Therapy

According to the clinical expert consulted for this review, the International Society of Nephrology/Renal Pathology Society classification represents the gold standard for histological classification of a renal biopsy¹⁷ and is widely used throughout Canada. Immunosuppressive treatment is recommended in active class III (focal proliferative), IV (diffuse proliferative), and V (membranous) disease or mixed class III plus V or class IV plus V disease. According to the expert consulted, the current treatment paradigm for active class III to V LN consists of 2 phases: induction therapy with aggressive immunosuppression, usually lasting 3 months to 12 months; long-term maintenance with less intensive immunosuppression for at least 3 years to 5 years to prevent renal relapses. Guidelines from the European League Against Rheumatism and the European Renal Association–European Dialysis and Transplant Association (EULAR/ERA-EDTA) and Kidney Disease: Improving Global Outcomes (KDIGO) recommended that first-line induction treatment options include a high-dose corticosteroid taper and either mycophenolate mofetil or mycophenolic acid or cyclophosphamide,^8,11 which are used off label in Canada. After a satisfactory response has been achieved with induction therapy, patients are switched to a subsequent maintenance treatment, which — according to the clinical expert — most often consists of mycophenolate mofetil or mycophenolic acid at a lower dose than is used in the induction phase, with azathioprine (off label) generally reserved for when the patient is considering pregnancy.⁸ The use of corticosteroids should be gradually tapered in the maintenance phase⁸ due to their substantial long-term toxicity and, according to the clinical expert, if continued, should not exceed a dose of 5 mg/day to 7.5 mg/day after 6 months to 12 months of therapy. The clinical expert suggested that multitarget therapy can be considered in patients with an inadequate response to first-line induction therapy (i.e., refractory cases), which may include off-label use of rituximab, cyclosporin, or tacrolimus in addition to standard induction therapy. As the single randomized trial comparing rituximab as add-on therapy to standard of care versus standard of care alone in LN (the LUNAR trial) did not achieve its primary outcome¹⁸ and as supportive data for the use of calcineurin inhibitors (i.e., cyclosporin or tacrolimus) combined with mycophenolate mofetil (or other forms of mycophenolate) in patients of race or ethnicity other than Asian are scarce,^19,20 these approaches are not recommended as standard of care by the most recent versions of either the EULAR/ERA-EDTA or KDIGO guidelines.^8,11 The clinical expert suggested that in all cases of active class III to V LN, the use of antimalarials (i.e., hydroxychloroquine), bone protection (vitamin D, calcium, possibly antiresorptive drugs), immunizations with nonlive vaccines, and lifestyle modifications to reduce cardiovascular risk are recommended, and adjunct treatment with renin-angiotensin blockade and statins should be considered.

According to the clinical expert, important treatment goals include the optimization of kidney function (i.e., preservation or improvement in eGFR) accompanied by reduction in proteinuria, reduction of glucocorticoid use, prevention of relapse of LN, achievement of SLE remission or low disease activity, minimization of treatment-related AEs and accumulation of damage in the kidney and other organs, and decreased mortality. Secondary treatment goals can include improvement in urinary sediment, clinical parameters (serum hemoglobin, serum albumin), serological markers (i.e., anti–double-stranded DNA [anti-dsDNA], complement C3 and C4), HRQoL, and ability to maintain employment.

Drug

B cells are believed to play an important role in the pathophysiology of LN, which may lead to damage affecting multiple organ systems.¹² Belimumab reduces the survivability of B cells by limiting the activity of B lymphocyte stimulator protein. Belimumab is indicated in addition to standard therapy for reducing disease activity in adult patients with active LN. The Health Canada–recommended dose for IV administration is 10 mg/kg, administered over an hour, at 2-week intervals for the first 3 doses and at 4-week intervals thereafter. The product monograph indicates that the infusion rate may be slowed or interrupted if the patient develops an infusion reaction. In the event of a serious infusion-related or hypersensitivity reaction (e.g., anaphylaxis), treatment should be discontinued immediately and appropriate therapy should be administered. The Health Canada–recommended dose for subcutaneous injection is 400 mg (two 200 mg injections) once weekly for 4 doses, then 200 mg once weekly thereafter. For adult patients with LN transitioning from belimumab IV therapy to subcutaneous therapy, the first subcutaneous dose is to be administered 1 week to 2 weeks after the last IV dose. This transition can occur anytime after the patient completes the first 2 IV doses. The product monograph further states that belimumab should be used in combination with corticosteroids and mycophenolate or cyclophosphamide for induction, or mycophenolate or azathioprine for maintenance, and that the patient’s condition should be evaluated continuously.¹²

The sponsor-requested reimbursement indication of belimumab differs from the Health Canada–approved indication. The sponsor reimbursement request is for belimumab in addition to standard therapy for treatment of active LN in adult patients who have class III, class IV, and/or class V active LN, and if no improvements in disease activity or symptoms are observed after 6 months, use should be discontinued.

Belimumab has been approved for use in Europe and the US for the treatment of adult patients with active LN who are receiving standard therapy. The subcutaneous formulation of belimumab was previously reviewed by CADTH in 2020 for SLE.

Key characteristics of biologic drugs used in the treatment of LN are presented in Table 3.

Table 3: Key Characteristics of Belimumab, Rituximab, Antimalarials, Corticosteroids, and Immune Suppressants

LN = lupus nephritis; PML = progressive multifocal leukoencephalopathy; SLE = systemic lupus erythematosus.

^aHealth Canada–approved indication.

Sources: Product monographs for Benlysta,¹² Rituxan,²¹ hydroxychloroquine,²² prednisone,²³ Procytox,²⁴ and CellCept.²⁵

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups. The full original patient input(s) received by CADTH have been included in the stakeholder section at the end of this report.

Four responses to CADTH’s call for patient input for the belimumab submission were received. These consisted of submissions from Arthritis Consumer Experts, Lupus Ontario, a joint submission from the Kidney Foundation of Canada and Lupus Canada, and a cooperative submission from the Canadian Arthritis Patient Alliance, the Arthritis Society, the Canadian Skin Patient Alliance, and CreakyJoints. Patient input was gathered from surveys, video interviews, and focus group discussions among lupus patients across Canada: 34 respondents (88% female) from Arthritis Consumer Experts, 10 respondents (90% female) with SLE from Lupus Ontario, and 38 respondents (73% with LN and approximately 15% caregivers) from the Kidney Foundation of Canada and Lupus Canada. The cooperative submission conducted a focus group of 3 patients with LN as well as a video interview with 1 patient. The submission from Arthritis Consumer Experts also included an in-depth interview with 1 patient. Seventeen patients (6 from a previous survey) in the included submissions had experience with the treatment under review.

Patients reported managing SLE was difficult given the severity of the physical symptoms, such as debilitating fatigue, joint pain, flares, skin rashes, nausea, loss of appetite, bruising, back pain, brain fog, mobility issues, and mental health issues. Respondents reported that currently available treatments are difficult to tolerate because of the many side effects, such as headaches, brain fog, additional fatigue, frequent infections, eye issues, osteoporosis, upset stomach, gastric issues, insomnia, hair loss, weight gain or loss, mood swings, changes in kidney function, skin changes, heart disease, acne, nausea, vomiting, decrease in white blood count, bone marrow toxicity, liver toxicity, and bladder-related problems. While describing their experiences with the drug under review, patients reported both positive and negative outcomes. Some patients described experiencing side effects, such as severe allergic reaction, extreme nausea, sleep deprivation, frequent UTIs, depression, and psychosis. Other patients reported an overall decrease in their disease symptoms and improvement in physical ability, leading to improvement in their HRQoL.

The key outcomes patients would like addressed by a new therapy are reduction of side effects and number of medications used; reduction in fatigue, flares, pain, and rash and skin irritations; increased mobility and participation in physical activities; overall improvement in HRQoL; improved engagement in social activities; and better affordability of medication. Patients identified financial barriers and difficulty in receiving reimbursement from private insurance and provinces while accessing belimumab a major concern, as costs to patients without drug coverage can be upward of $2,500 per month.

Clinician Input

Input From Clinical Expert Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol; assisting in the critical appraisal of clinical evidence; interpreting the clinical relevance of the results; and providing guidance on the potential place in therapy). The following input was provided by 1 clinical specialist with expertise in the diagnosis and management of LN.

Unmet Needs

According to the clinical expert consulted, response to current standard of care induction therapy is suboptimal, with only 20% to 35% of patients achieving a CRR within 6 months to 12 months from onset of LN, indicating a major unmet need. Of those that do achieve adequate disease control, 20% to 35% relapse within 3 years to 5 years. At least 20% of patients with LN develop chronic kidney disease within 10 years of LN onset, and up to 40% (in the most aggressive forms of proliferative LN) develop ESRD after 15 years and require dialysis or transplant. The expert noted that current treatments cause substantial short- and long-term side effects (i.e., mycophenolate mofetil mofetil or other forms of mycophenolate, cyclophosphamide, and azathioprine are associated with gastrointestinal intolerance and increased risk of infection; cyclophosphamide is associated with premature ovarian failure, infertility, and hematologic and bladder cancer; azathioprine is associated with hepatotoxicity; cyclosporin and tacrolimus require frequent monitoring of blood levels and can cause long-term reduction in kidney function; rituximab is associated with increased risk of infection; and corticosteroids are associated with premature cardiovascular and cerebrovascular disease, osteoporosis, avascular necrosis, cataracts, and diabetes). The expert also noted that there is a high rate of noncompliance with oral medications and only a few treatments are safe in pregnancy, which include hydroxychloroquine, corticosteroids, azathioprine, and tacrolimus.

Place in Therapy

The clinical expert noted that belimumab would be implemented as an add-on therapy to existing standard of care (i.e., corticosteroids and mycophenolate mofetil or other forms of mycophenolate or cyclophosphamide) for active LN class III or class IV (with or without class V) or pure class V if there is an inadequate response after 2 months to 3 months of induction therapy. In cases where there is no response whatsoever or a worsening of symptoms after 2 weeks to 4 weeks of induction therapy, consideration could be given to initiating belimumab as add-on therapy at an earlier stage of induction. In the clinical expert’s opinion, the efficacy and safety of initiating belimumab anytime within the 60-day window after initiation of standard of care induction therapy has been demonstrated in the BLISS-LN trial. In the trial, in 82.5% of patients, belimumab was initiated within 4 weeks after initiation of induction therapy and in 57.6% of patients, belimumab was initiated within 2 weeks after initiation of induction therapy.²⁶ However, given that 20% to 35% of patients will respond adequately to standard of care induction therapy alone and given the significant medication burden associated with standard of care induction therapy, the clinical expert noted that it would be prudent in patients who are gradually improving to wait for a maximum of 2 months to 3 months to assess response to standard of care before initiating belimumab. This should be sufficient time to allow for a response to be seen but should not prolong an ineffective treatment and risk increasing renal damage. If the response is deemed inadequate, belimumab could be added to standard of care. However, should the patient worsen or demonstrate no improvement whatsoever within 2 weeks to 4 weeks after initiation of induction therapy, the clinical expert believed it would be appropriate to initiate belimumab earlier in the induction course, either at the time of initiation of standard of care induction or within 1 month after initiation of standard of care induction (as was the case for most patients in BLISS-LN).

The clinical expert noted that BLISS-LN did not assess the efficacy of administering belimumab later in the induction phase (i.e., after > 60 days of induction therapy). The expert noted that for patients who show some initial response within the first 2 months to 3 months of induction therapy and then plateau or deteriorate or are unable to decrease steroids between months 3 to 6, there may be benefit in adding belimumab to standard of care at 3 months to 6 months after induction therapy (although there are no data from the trial addressing this use of belimumab). Hence, the clinical expert noted that it may be appropriate to provide the prescriber the flexibility to initiate belimumab anytime from start of standard of care induction to within the first 3 months to 6 months after initiation of standard of care induction therapy, based on clinician expert judgment.

Addition of belimumab to standard of care would represent a shift in the LN treatment paradigm. Currently, induction therapy is sequential, with a single immunosuppressive drug followed by maintenance, again with a single drug. Integration of belimumab into the induction regimen would represent a multitarget or combination approach.

Patient Population

According to the clinical expert, the patients best suited for treatment with belimumab in addition to standard of care are those with active class III or IV (with or without class V) or active class V LN. Other patients who may be most likely to respond to belimumab include patients with active class III or IV (with or without class V) or class V LN who are experiencing chronically impaired renal function, or patients for whom prednisone is unable to be decreased to less than or equal to 7.5 mg/day after 3 months to 6 months of induction, or patients who have extrarenal manifestations in addition to LN. Also, patients with previous episodes of class III or IV (with or without class V) or class V LN in whom another flare may cause a serious decline in renal function may respond to belimumab. The patients least suited for treatment with belimumab include those with active LN who are not currently receiving standard of care induction therapy; patients for whom induction with both mycophenolate mofetil (or other forms of mycophenolate) and cyclophosphamide has failed, as the efficacy of belimumab has only been demonstrated when added to induction with either of these drugs; patients with an eGFR less than or equal to 30 mL/min/1.73 m² as these patients were excluded from the BLISS-LN trial; and patients who have attained a CRR after 6 months to 12 months of induction therapy, as these patients are less likely to derive additional benefit from belimumab. According to the clinical expert, LN should be suspected in any patient with SLE with an unexplained decline in renal function or a urine protein exceeding 500 mg/day. Such patients should undergo a prompt renal biopsy (providing there are no contraindications, such as increased risk of bleeding), which is the gold standard for diagnosis of LN. Based on renal biopsy findings, a decision would be made regarding the need to initiate induction therapy for LN. Several other laboratory parameters, (i.e., serum hemoglobin, serum albumin, urinalysis, anti-dsDNA, anti-C1q, complement C3 and C4) are also measured, and expert physician clinical examination is performed before confirming a diagnosis of LN and initiating induction therapy. The clinical expert states that kidney biopsy is indispensable and cannot be replaced by laboratory variables.

Assessing Response to Treatment

The clinical expert noted that a clinically meaningful response to standard of care would include, sequentially:

• at least a 25% reduction in proteinuria (as defined by a uPCR) after 3 months of therapy

• at least a 50% reduction in proteinuria after 6 months of therapy

• reduction in corticosteroids to less than or equal to 7.5 mg/day after 6 months to 12 months of therapy

• proteinuria no greater than 0.5 g per 24 hours to 0.7 g per 24 hours after 12 months of therapy (the response time can be delayed to 18 months to 24 months if baseline proteinuria is in the nephrotic range [i.e., > 3.5 g per 24 hours])

• an eGFR no worse than 10% to 20% of pre-flare value and greater than or equal to 60 mL/min/1.73 m² after 12 months of therapy.

Key secondary indicators of renal response according to the clinical expert would include increased time to renal-related event or death, prevention of relapse of LN, decreased disease activity in extrarenal domains, minimization of treatment-related AEs and accumulation of damage in the kidney and other organs, and decreased mortality. Other treatment goals can include improvement in clinical parameters (serum hemoglobin, serum albumin), urinalysis, serological markers (i.e., anti-dsDNA, complement C3 and C4), HRQoL, and ability to maintain employment.

Discontinuing Treatment

The clinical expert recommended discontinuing treatment with belimumab immediately if the patient experienced a severe adverse reaction to belimumab (e.g., anaphylaxis) or became pregnant (the Health Canada product monograph indicates that belimumab should not be used during pregnancy unless the potential benefit justifies the potential risk to the fetus). According to the clinical expert, it would be reasonable to discontinue treatment with belimumab after 6 months to 12 months of therapy if there was no improvement or a worsening in renal function or proteinuria. The clinical expert noted that in BLISS-LN, the percentage of patients achieving the primary outcome (i.e., PERR) over time was identical through to week 20 after randomization and started to diverge at week 24 through to week 104 (with a greater percentage achieving PERR in the belimumab group). Hence, it would be appropriate to wait at least 6 months to assess response to belimumab, and in patients with higher levels of proteinuria the clinical expert believed it would be appropriate to wait up to 12 months to determine if the response is sufficient to warrant continuation. Hence, the clinical expert stated that it would be optimal to provide physicians the flexibility to prescribe belimumab for up to 12 months before assessing response. The expert would also discontinue treatment if there was a lack of steroid-sparing effect within 6 months to 12 months of therapy. If a CRR was achieved (as defined by stabilization or normalization of renal function and improvement in proteinuria to less than 700 mg/day) and maintained for at least 3 years to 5 years, it may be reasonable to discontinue therapy with belimumab with the expectation that remission will be maintained. However, there are currently no data to support duration of maintenance therapy, and EULAR/ERA-EDTA guidelines recommend at least 3 years to 5 years with current standard of care maintenance therapy with mycophenolate mofetil (or other forms of mycophenolate) or azathioprine before attempting gradual tapering.

Prescribing Conditions

According to the clinical expert, IV belimumab should be administered in a hospital outpatient or community infusion centre. Diagnosis, treatment, and monitoring of patients with LN who might receive belimumab should be by a specialist, typically a rheumatologist or nephrologist experienced in the management of LN.

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by clinician groups. The full original clinician group input(s) received by CADTH have been included in the stakeholder section at the end of this report.

The Canadian Network for Improved Outcomes for Systemic Lupus Erythematosus and 31 associated physicians provided input for this review.

The clinician group and physicians agreed that there are some treatment gaps and unmet needs in the current LN therapeutics. These unmet needs include inability to achieve complete remission from existing treatment options (e.g., mycophenolate mofetil (or other forms of mycophenolate) and cyclophosphamide); increased risk of multiple complications from moderate or high doses of corticosteroids; subsequent ESRD and renal replacement therapy associated with disease flares; and difficulty in maintaining adherence.

The clinician group and physicians indicated that a clinically meaningful response to treatment would include any of the following: complete remission (proteinuria < 0.5 g per 24 hours) within 12 months of starting treatment, reduction in daily prednisone dose to levels less than or equal to 7.5 mg, or reduction in the frequency and intensity of flares. According to the clinician group, sufficient time (at least 12 months) should be allowed for these outcomes to be observed and treatment should be discontinued after 12 months in cases where no response can be demonstrated. The input stated that belimumab is expected to cause a shift in the current treatment paradigm for LN by addressing the disease mechanism. Its ability to modulate the maturation and functional differentiation of B cells, which produce autoantibodies that are central in SLE pathogenesis and tissue damage, renders it most suitable for patients not achieving at least partial remission, patients experiencing early and frequent flares, patients with steroid-dependent disease, and patients with adherence issues.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 4.

Table 4: Summary of Drug Plan Input and Clinical Expert Response

CDEC = CADTH Canadian Drug Expert Committee; eGFR = estimated glomerular filtration rate; ISN = International Society of Nephrology; LN = lupus nephritis; pCPA = pan-Canadian Pharmaceutical Alliance; PERR = primary efficacy renal response; RPS = Renal Pathology Society; uPCR = urine protein-creatinine ratio.

Clinical Evidence

The clinical evidence included in the review of belimumab is presented in 3 sections. The first section, the Systematic Review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence selected from the literature that met the selection criteria specified in the review. The third section includes a sponsor-submitted long-term extension study that was considered to address important gaps in the evidence included in the systematic review.

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of belimumab IV infusion or subcutaneous injection in addition to standard of care therapy for the treatment of active LN in adults.

Methods

Studies selected for inclusion in the systematic review will include pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 5. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Table 5: Inclusion Criteria for the Systematic Review

AE = adverse event; eGFR = estimated glomerular filtration rate; FACIT-F = Functional Assessment of Chronic Illness Therapy – Fatigue; HRQoL = health-related quality of life; LN = lupus nephritis; Lupus QoL = Lupus Quality of Life questionnaire; RCT = randomized controlled trial; SAE = serious adverse event; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SF-36 = Short Form (36) Health Survey; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria; UTI = urinary tract infection; WDAE = withdrawal due to adverse event.

^aClasses according to the International Society of Nephrology/Renal Pathology Society 2003 classification of LN: class III = focal LN; class IV = diffuse LN; class V = membranous LN.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.²⁷ Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) via Ovid and Embase (1974–) via Ovid. All Ovid searches were run simultaneously as a multifile search. Duplicates were removed using Ovid deduplication for multifile searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were belimumab and LN. Clinical trial registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

No filters were applied to limit retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. Refer to Appendix 1 for the detailed search strategies.

The initial search was completed on August 4, 2022. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on November 23, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool for Searching Health-Related Grey Literature checklist.²⁸ Included in this search were the websites of regulatory agencies (FDA and the European Medicines Agency). Google was used to search for additional internet-based materials. Refer to Appendix 1 for more information on the grey literature search strategy.

In addition, the manufacturer of the drug was contacted for information regarding unpublished studies.

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

Findings From the Literature

A total of 299 studies were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 6. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 6: Details of Included Studies

AE = adverse event; AESI = adverse event of special interest; anti-dsDNA = anti-double-stranded DNA; CRR = complete renal response; C-SSRS = Columbia – Suicide Severity Rating Scale; DB = double-blind; ELISA = enzyme-linked immunosorbent assay; ISN = International Society of Nephrology; LN = lupus nephritis; OLE = open-label extension; ORR = ordinal renal response; PERR = primary efficacy renal response; RBC = red blood cell; RCT = randomized controlled trial; RPS = Renal Pathology Society; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SFI = SELENA SLEDAI Flare Index; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria; uPCR = urine protein-creatinine ratio; WBC = white blood cell.

Source: Clinical Study Report for BLISS-LN;¹³ Furie et al. (2020);²⁹ Rovin et al. (2021).²⁶

Description of Studies

One double-blind, placebo-controlled randomized controlled trial met the inclusion criteria for the systematic review (BLISS-LN).

The objective of the BLISS-LN study was to determine the efficacy, safety, and tolerability of belimumab in adult patients with active LN (classes III, IV, V, or V in combination with III or IV). The trial took place in 107 sites across 21 countries including China, the US, and the Philippines, with 1 site in Canada. A total of 448 patients who met the eligibility criteria during screening were randomized to 1 of 2 treatment groups — 10 mg/kg IV belimumab plus standard of care or placebo plus standard of care — in a 1:1 ratio (Figure 2). In addition to the investigational product (IP), all patients received standard of care regimens chosen by the investigators, which were initiated within 60 days before receiving the IP and which included high-dose corticosteroids and either IV cyclophosphamide for induction therapy followed by azathioprine for maintenance therapy or oral mycophenolate mofetil for induction and maintenance therapy. The randomization of all eligible patients was stratified by their induction regimen (high-dose corticosteroids plus cyclophosphamide versus high-dose corticosteroids plus mycophenolate mofetil) and race (Black versus non-Black). The study recommended that the high-dose corticosteroid regimen included 0 to 3 IV pulses of methylprednisolone (500 mg/pulse to 1,000 mg/pulse) followed by an oral prednisone dose of up to 60 mg/day that had to be tapered to less than or equal to 10 mg/day by week 24. The recommended taper regimen suggested reducing corticosteroids by approximately 5 mg/week. In addition to standard of care, patients received IV belimumab or placebo on days 1 (baseline), 15, and 29 and every 28 days thereafter to week 100, with final assessments at week 104. Hydroxychloroquine as well as angiotensin-converting enzyme inhibitors or angiotensin receptor blockers were encouraged for all patients. Patients who received treatment with the IP through week 100 and completed week 104 assessments in the double-blind period could enter into a 6-month OLE study.

Patients were defined as having treatment failure if they did not follow the corticosteroid rules (i.e., failed to taper corticosteroids to ≤ 10 mg/day by week 24 and not exceed this dose through week 104); if they received additional immunosuppressive drugs (except topical drugs) beyond the induction and maintenance regimens; if the use of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or antimalarial drugs was initiated after week 24; or if the patient’s standard therapy (i.e., cyclophosphamide, azathioprine, or mycophenolate mofetil) exceeded permitted doses. Patients were to be withdrawn from the study treatment if they missed 3 or more consecutive IP doses, took prohibited medication or a prohibited dose of concurrent medication, experienced unacceptable toxicity, became pregnant, withdrew consent, or tested positive for hepatitis B at screening.

Figure 2: Flow Chart of BLISS-LN Study Design

Source: Clinical Study Report for BLISS-LN.¹³

Populations

Inclusion and Exclusion Criteria

The BLISS-LN trial enrolled adult patients aged 18 years and older with active, biopsy-proven LN class III or IV, with or without class V, or pure class V membranous LN. In addition, patients had to have an antinuclear antibody titre greater than or equal to 1:80 and/or anti-dsDNA greater than or equal to 30 IU/mL and active renal disease requiring induction therapy with high-dose corticosteroids with either cyclophosphamide or mycophenolate mofetil or oral forms of mycophenolate. Patients were excluded if both cyclophosphamide and mycophenolate mofetil induction therapy had previously failed, if they had received cyclophosphamide induction therapy within 3 months of the trial, if they had an eGFR less than 30 mL/min/1.73 m², if they were pregnant or breastfeeding, or if they had been on dialysis, been treated with belimumab, or received any B cell–targeted therapy (e.g., rituximab) in the past year.

Baseline Characteristics

Baseline characteristics were generally well balanced between treatment groups in the BLISS-LN trial (Table 7). The mean (SD) age of patients enrolled in the BLISS-LN trial was 33.7 (10.74) years and 33.4 (10.6) years in the belimumab and placebo groups, respectively. Patients were predominantly female (88.3% and 87.9%) and predominantly Asian (51.1% and 48.9%), and 84% of patients in both the belimumab and placebo groups had renal biopsy class III or IV (with or without class V) according to the local reader. The mean of LN disease duration was 2.3 years (SD = 4.3 years) and 2.4 years (SD = 4.1 years) in the belimumab and placebo groups, respectively. The mean uPCR levels were 3.2 g/g (SD = 2.7 g/g) and 3.5 g/g (SD = 3.6 g/g), and the mean eGFR levels were 100.0 mL/min/1.73 m² (SD = 37.7 mL/min/1.73 m²) and 101.0 mL/min/1.73 m² (SD = 42.7 mL/min/1.73 m²) in the belimumab and placebo groups, respectively. A total of 73.5% of patients in both groups received mycophenolate mofetil as induction therapy. Patients in both the belimumab and placebo groups had previous cyclophosphamide therapy (19.3% and 21.1%) and previous mycophenolate mofetil therapy (13.9% and 15.2%).

There were some baseline differences in use of concomitant medications between groups (Table 8), with slightly more patients using antimalarials in the belimumab group than in the placebo group (74.4% and 69.1%). A total of 96.9% and 94.2% of patients used steroids at baseline in the belimumab and placebo groups, respectively. The mean dose of prednisone at baseline (steroids were converted to prednisone equivalent) was slightly lower in the belimumab group, at 66.5 mg/day (SD = 99.6 mg/day), than in the placebo group, at 72.5 mg/day (SD = 133.2 mg/day). Patients also used immunosuppressants (88.8% and 86.1%) and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (65.9% and 67.3%) at baseline in the belimumab and placebo groups, respectively, the proportions of which were comparable between groups.

Table 7: Summary of Baseline Characteristics of Patients in the BLISS-LN Trial

anti-dsDNA = anti–double-stranded DNA; BLyS = B lymphocyte stimulator; BMI = body mass index; eGFR = estimated glomerular filtration rate; LN = lupus nephritis; PGA = physician global assessment; SD = standard deviation; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SLE = systemic lupus erythematosus; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria; uPCR = urine protein-creatinine ratio.

^aPatients are only counted in 1 category.

^bMycophenolate mofetil or any other form of mycophenolate.

^cBased on the presence of class III, IV, and V results, although other classes may be present also.

^dNot all patients had an assessment by the central reader.

^ePre-flare is defined as the most recent values obtained before screening and before first manifestations of the current renal flare, as reported by investigator if available.

^fDerived from pre-flare serum creatinine.

^gTwo different laboratory assays were used for BLyS protein: 366 samples were analyzed with 1 assay, and 79 samples were analyzed with the other.

Source: Clinical Study Report for BLISS-LN.¹³

Table 8: Summary of Systemic Lupus Erythematosus and Lupus Nephritis Treatments Patients Receiving at Baseline

SD = standard deviation.

^aSteroids were converted to prednisone equivalent.

Source: Clinical Study Report for BLISS-LN.¹³

Interventions

In the BLISS-LN trial, eligible patients were randomized in a 1:1 ratio to receive IV belimumab 10 mg/kg or matching placebo administered over at least 1 hour. Patients were dosed on days 0 (baseline), 14, and 28 and then every 28 days thereafter through 100 weeks, with a final evaluation for the double-blind treatment period at 104 weeks. Patients who received treatment with the IP through week 100 and completed week 104 assessments in the double-blind period could enrol in a 28-week OLE study. Belimumab and placebo were supplied as open-label vials, and the IP was reconstituted and diluted by an unblinded qualified person (i.e., site pharmacist or designee), independent of the study. The infusion rate of the IP could be slowed or interrupted if a patient developed an infusion reaction, but the dose could not be altered. If a clinically significant AE possibly related to the IP was suspected, doses could be delayed by up to 2 weeks or 1 dose could be withheld at the investigators’ discretion.

In addition to the IP, all patients received standard of care, which included high-dose corticosteroids and either:

• cyclophosphamide for induction therapy (500 mg by IV infusion every 2 weeks for 6 infusions) followed by azathioprine for maintenance therapy (target dose of 2 mg/kg/day)

• mycophenolate mofetil for induction and maintenance therapy with recommended oral dosing of 0.5 g twice daily for the first week, increasing to 1 g twice daily for the second week, and then 1.5 g twice daily for the third and subsequent weeks (mycophenolate sodium was permitted in lieu of mycophenolate mofetil for induction and/or maintenance therapy with a recommended oral dose from 720 mg/day to 2,160 mg/day).

The induction therapy was chosen by the investigator and initiated within 60 days before patients received the IP. Patients were able to switch maintenance regimens if certain tolerability or toxicity issues occurred. Patients were to be withdrawn from the study treatment if they missed 3 or more consecutive IP doses, took prohibited medication or a prohibited dose of concurrent medication, experienced unacceptable toxicity, became pregnant, withdrew consent, or tested positive for hepatitis B at screening.

High-Dose Corticosteroids

All patients were on a daily corticosteroid regimen as part of the induction therapy. The study-recommended corticosteroid regimen included 0 to 3 IV pulses of methylprednisolone followed by oral prednisone for a total daily dose up to 60 mg. Corticosteroids had to be tapered over time to less than or equal to 10 mg/day by week 24. Short-term rescue treatment was permitted between weeks 24 and 76 for reasons other than LN. No corticosteroid rescue treatment was allowed from week 76 to week 104.

Concomitant Medications

Concomitant medications, including LN-related treatments, were recommended and used by most patients in the BLISS-LN trial as shown in Table 8. Most patients in both the belimumab and placebo groups received antimalarials (74.4% versus 69.1%) and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (65.9% versus 67.3%). Other immunosuppressives (e.g., methotrexate) were allowed provided these were started before baseline and met eligibility criteria. Starting a new angiotensin-converting enzyme inhibitor, angiotensin receptor blocker, or antimalarial treatment after the week 24 visit would be considered treatment failure.

Prohibited medications included new immunosuppressant drugs (other than as part of the induction and maintenance regimens), corticosteroid use outside of the limits, other investigational drugs (biologic or nonbiologic), anti–tumour necrosis factor therapy (e.g., adalimumab, etanercept, infliximab), other biologics (e.g., rituximab, abatacept, IV immunoglobulin), or plasmapheresis.

Outcomes

A list of efficacy outcomes identified in the CADTH review protocol and assessed in the clinical trials included in this review is provided in Table 9. These outcomes are further summarized below. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 4.

Table 9: Summary of Outcomes of Interest Identified in the CADTH Review Protocol

anti-dsDNA = anti–double-stranded DNA; CRR = complete renal response; eGFR = estimated glomerular filtration rate; HRQoL = health-related quality of life; NR = not reported; ORR = ordinal renal response; PERR = primary efficacy renal response; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SFI = SELENA SLEDAI Flare Index; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria.

Source: Clinical Study Report for BLISS-LN.¹³

Primary Efficacy Renal Response

PERR at week 104 was the primary outcome and PERR at week 52 was a key secondary outcome in the BLISS-LN trial. PERR was a dichotomous composite outcome (“responder” versus “nonresponder” [from original source, referred to as “response” versus “no response” hereafter) that was considered achieved when all 3 of the following components were met: uPCR was less than or equal to 0.7 g/g, eGFR was no more than 20% below pre-flare value or at least 60 mL/min/1.73 m², and there was no treatment failure (i.e., patients did not take a protocol-prohibited or -restricted medication or dose). “Response” was defined as a response at the week 48 or week 100 visit that was confirmed by a repeat measure at the week 52 or week 104 visit, respectively. Patients who met treatment failure criteria were to continue in the study.

Complete Renal Response

CRR at week 104 was a key secondary outcome in the pivotal trial. CRR was a composite outcome that was considered achieved when all 3 of the following components were met: uPCR was less than or equal to 0.5, eGFR was no more than 10% below pre-flare value or within the normal range of at least 90 mL/min/1.73 m², and there was no treatment failure (i.e., patients did not take a protocol-prohibited or -restricted medication or dose). A response was defined as a response at the week 100 visit confirmed by a repeat measurement at the week 104 visit.

Ordinal Renal Response

ORR at week 104 was a key secondary outcome in the BLISS-LN trial in which patients achieved a CRR, partial renal response, or no response. A CRR was achieved if a patient had all of the following:

• uPCR less than 0.5 g/g

• eGFR no more than 10% below pre-flare glomerular filtration rate (GFR) or within normal range

• no treatment failure (i.e., did not take a protocol-prohibited or -restricted medication or dose).

A partial renal response was achieved if a patient had all of the following:

• 50% or greater decrease from baseline in uPCR and either:

  • uPCR value less than 1 g/g if baseline value was less than or equal to 3 g/g

  • uPCR value less than 3 g/g if the baseline value was greater than 3 g/g

• eGFR no more than 10% below baseline GFR or within normal range

• no treatment failure (i.e., did not take a protocol-prohibited or -restricted medication or dose).

The complete and partial renal responses required a response at the week 100 visit that was confirmed by a repeat measurement at the week 104 visit. Patients not meeting criteria for either a complete or partial renal response were classified as ”no response.”

Reduction in Proteinuria

Secondary outcomes in the BLISS-LN trial included the proportion of patients with a uPCR less than or equal to 0.7 g/g or a uPCR less than 0.5 g/g by visit, as well as the absolute and percent change in proteinuria over time as measured by the uPCR change from baseline by visit while on treatment.

Increase or Stabilization of eGFR

Secondary outcomes in the BLISS-LN trial included the proportion of patients with an eGFR no more than 20% below pre-flare value or at least 60 mL/min/1.73 m² by visit and the proportion of patients with an eGFR no more than 10% below the pre-flare value or within the normal range (≥ 90 mL/min/1.73 m²) by visit. Post hoc analyses of the BLISS-LN trial included outcomes of time to 30% and 40% decline in eGFR from baseline using either only on-treatment data or all on-study data. Post hoc analyses also included sustained 30% and 40% decline in eGFR, which was defined as a 30% or 40% decrease in eGFR from baseline and was confirmed by the last 2 eGFR values in the trial.

Reduction in Corticosteroid Use

In the BLISS-LN trial, corticosteroid use was converted to a prednisone-equivalent dose. Secondary outcomes include the proportion of patients who received an average daily prednisone dose of less than or equal to 5 mg or less than or equal to 7.5 mg since the previous visit. The prednisone average daily dose since previous visit was calculated at every 4-week visit after baseline. All prednisone doses since the previous 4-week visit were summed and divided by the number of days in the period. Days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation for average prednisone dose.

Time to Renal-Related Event or Death

Time to renal-related event or death was a key secondary outcome in the trial, defined as the first event occurring after day 1 among the following: death by any cause; ESRD (defined as the need for chronic dialysis or renal transplant); doubling of serum creatinine (compared with baseline, confirmed with a second measurement at least 3 weeks later); renal worsening as evidenced by increased proteinuria and/or impaired renal function; or renal disease–related treatment failure (based on adjudication of treatment failures). Renal worsening was defined in the trial as a reproducible increase in 24-hour urine protein levels (as measured in uPCR) to more than 1 g if the baseline value was less than 0.2 g, or more than 2 g if the baseline value was between 0.2 g and 1 g, or more than twice the value at baseline if the baseline value was more than 1 g. Impaired renal function was defined as a reproducible decrease in GFR of more than 20% accompanied by at least 1 of the following: proteinuria (> 1), red blood cell casts, or white blood cell casts.

Reduction in Chronic Kidney Disease Stage

This outcome was not reported in the BLISS-LN trial.

Disease Activity: SLEDAI-S2K

In the trial, disease activity was assessed with the SELENA SLEDAI, a measure of disease activity consisting of 24 items across 9 organ systems that are scored based on the time of visit or preceding 10 days (Appendix 4).³⁰ The items are answered yes or no (presence or absence), and answers are weighted to arrive at a total score (range, 0 to 105), with higher scores indicating greater disease activity. The items include the following: seizure, psychosis, organic brain syndrome, visual disturbance, cranial nerve disorder, lupus headache, cerebrovascular attack, vasculitis, arthritis, myositis, urinary casts, hematuria, proteinuria, pyuria, rash, alopecia, mucosal ulcers, pleurisy, pericarditis, low complement, increased DNA binding, fever, thrombocytopenia, and leukopenia.³⁰ As noted in Appendix 4, a minimal important difference (MID) was not found for the SLEDAI-2K in patients with LN; however, the SLE literature suggests a minimal clinically meaningful increase of 3 or 4 points in the SLEDAI-2K for prediction of worsening and suggests a minimal clinically meaningful decrease in score of 1 to 2 points for improvement.³¹ In the BLISS-LN trial, the proteinuria component of the SELENA SLEDAI was modified with the SLEDAI-2K, which considers new as well as persistent proteinuria of more than 0.5 g per 24 hours, to create the SLEDAI-S2K. Assessed end points included SLEDAI-S2K change from baseline by visit and SLEDAI-S2K score less than 4 by visit.

Organ Damage: Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index

In the pivotal trial, organ damage was assessed using the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI) as a secondary outcome. Assessed end points included SDI change from baseline by visit and the percent of patients with any SDI worsening (change > 0) compared with baseline by visit. The SDI was developed to assess irreversible damage in patients with SLE independently of its cause (SLE activity, therapy, comorbidities) but occurring after disease onset. Damage is usually defined as a clinical feature that must be continuously present for at least 6 months to score. The SDI consists of 42 items in 12 domains, with a maximum score of 46 (higher scores denote more damage). The SDI is defined for 12 organ systems (possible scores): peripheral vascular (0 to 5), ocular (0 to 2), neuropsychiatric (0 to 6), renal (0 to 3), pulmonary (0 to 5), cardiovascular (0 to 6), gastrointestinal (0 to 6), musculoskeletal (0 to 7), skin (0 to 3), endocrine (diabetes) (0 to 1), gonadal (0 to 1), and malignancies (0 to 2). The SDI global score is the sum of the damage scores for all 12 organ systems. An SDI greater than or equal to 1 indicates worsening.³² The SDI has been found to be a predictor of mortality, and SDI scores have been shown to increase with disease duration.³³ No formal MID has been estimated for patients with SLE or LN. An SDI greater than or equal to 1 indicates damage, which may remain stable or increase over time.³²

Disease Flare Frequency and Severity: SFI Flares

In the pivotal trial, a modified version of SFI was used to assess flares. The SFI is a disease-specific composite measure that classifies flares as mild/moderate or severe, based on criteria of clinical activity, need for additional treatment, or physician global assessment (PGA) score.³³ In the pivotal trials, mild or moderate flares and severe flares were defined according to the following criteria:

• Mild or moderate flare (any of the following):

  • Change in SLEDAI-2K score of at least 3 points but no more than 12 points compared to previous visit

  • New or worse discoid, photosensitivity, profundus, cutaneous vasculitis, bullous lupus, nasopharyngeal ulcers, pleuritis, pericarditis, arthritis, or SLE fever

  • Increase in prednisone, but not exceeding 0.5 mg/kg/day

  • Addition of nonsteroidal anti-inflammatory drug or hydroxychloroquine for SLE activity

  • A 1.0 or greater increase in PGA score, but not to more than 2.5.

• Severe flare (any of the following):

  • Change in SLEDAI-2K score of more than 12 points compared to previous visit

  • New or worse central nervous system SLE, vasculitis, nephritis, myositis, hemolytic anemia (hemoglobin < 70 g/L or decrease in hemoglobin > 30 g/L) requiring doubling of prednisone or prednisone increase to more than 0.5 mg/kg/day or hospitalization

  • Increase in prednisone to more than 0.5 mg/kg/day

  • New cyclophosphamide, azathioprine, methotrexate, or mycophenolate for SLE activity

  • Hospitalization for SLE

  • Increase in PGA score to more than 2.5.

In the BLISS-LN trial, flares originally marked severe were downgraded if the only reason they had been marked as such was a change in SELENA SLEDAI score to greater than 12, because this may indicate only a modest increase in disease activity.

Health-Related Quality of Life

This outcome was not reported in the BLISS-LN trial.

Reduction in Symptoms

This outcome was not reported in the BLISS-LN trial.

Serological Outcomes: Decrease in Anti-dsDNA Antibodies

Anti-dsDNA serum levels usually increase as the clinical activity of the disease increases, most often before the clinical deterioration of kidney function.¹⁰ In the BLISS-LN trial, the mean change from baseline in anti-dsDNA through to week 104 was calculated, as was the proportion of patients who shifted from a positive anti-dsDNA (≥ 30 IU/mL) at baseline to a negative anti-dsDNA (< 30 IU/mL) at week 104.

Serological Outcomes: Increase in Complement C3 and C4 Levels

Complement C3 and C4 typically decrease as the clinical activity of the disease increases, most often before the clinical deterioration of kidney function.¹⁰ Decrease in circulating complement levels are associated with LN and may also be diagnostic markers in this disease.¹⁰ In the BLISS-LN trial, the mean change from baseline in complement C3 and C4 levels through to week 104 was calculated, as was the proportion of patients who shifted from a low level in each complement (C3 < 90 mg/dL; C4 < 10 mg/dL) at baseline to a normal or high level (C3 ≥ 90 mg/dL; C4 ≥ 10 mg/dL) at week 104.

Safety Assessments

AEs were any untoward medical occurrence that may or may not have been related to the study drug. The criteria for SAEs included death; events that were life threatening, required hospitalization or prolongation of hospitalization, or resulted in persistent or significant disability or incapacity; congenital anomalies or birth defects; events of possible drug-induced liver injury with hyperbilirubinemia; or events that may have required medical or surgical intervention to prevent 1 of the other outcomes listed.

In addition, the Columbia – Suicide Severity Rating Scale (C-SSRS) was utilized as a safety assessment for suicidality. The C-SSRS is an assessment tool that evaluates suicidal ideation and behaviour and was completed at every visit during the double-blind phase in the BLISS-LN trial. It is made up of 10 categories, all of which maintain binary responses (yes or no) to indicate a presence or absence of behaviours that are significantly predictive of completed suicide.³⁴ The outcome of the C‐SSRS is a numerical score obtained from the aforementioned categories.

Statistical Analysis

Sample Size Determination and Power Calculation

In the BLISS-LN trial, the original sample size calculations using ORR as the primary outcome resulted in a target sample size of N = 464 patients (232 per arm) to achieve at least 85% power to detect a treatment difference for the initial primary outcome. The primary outcome was revised to PERR with a final sample size of N = 448, with 80% power (2-sided test; significance level of 0.05) to detect a 13.6% between-group difference and a minimum detectable difference of 9.7%, assuming response rates of 40% and 53.6% for placebo and belimumab, respectively.

Efficacy outcomes were analyzed in the modified intention-to-treat (mITT) population, which included all the patients who underwent randomization and received at least 1 dose of belimumab or placebo.

Statistical Test or Model

The main components of the statistical test and model for the BLISS-LN trial are discussed in Table 10. The primary outcome, PERR at week 104, as well as 2 of the key secondary outcomes — CRR at week 104 and PERR at week 52 — were compared between belimumab and placebo using logistic regression. All statistical tests were 2-sided and performed at an overall significance level of alpha = 0.05. The key secondary outcome of time to renal-related event or death was assessed using a Cox proportional hazards model, and ORR was analyzed using rank analysis of covariance (ANCOVA). The primary and all key secondary outcomes controlled for induction regimen (cyclophosphamide versus mycophenolate mofetil), race (Black versus non-Black), baseline proteinuria, and baseline eGFR.

Proportion outcomes as defined in Table 10 were compared between belimumab and placebo using logistic regression. For continuous outcomes, ANCOVA was used to evaluate change from baseline in SLEDAI-S2K, and rank ANCOVA was used to evaluate change from baseline in proteinuria and SDI. All analyses for these proportion outcomes and continuous outcomes controlled for baseline value, induction regimen (cyclophosphamide versus mycophenolate mofetil), and race (Black versus non-Black).

Other time-to-event outcomes, including time to first PERR, time to first CRR that is maintained through week 52 or week 104, and time to first severe SFI flare, were assessed using a Cox proportional hazards model controlling for induction regimen (cyclophosphamide versus mycophenolate mofetil) and race (Black versus non-Black).

Post hoc analyses of the BLISS-LN trial by Rovin et al. (2022)²⁶ included end points of time to 30% and 40% decline in eGFR from baseline and sustained 30% and 40% decline in eGFR, which were analyzed with a Cox proportional hazards model and logistic regression models, respectively. Analyses were adjusted for induction regimen, race, baseline uPCR, and baseline eGFR.

Data Imputation Methods

Patients who discontinued treatment or withdrew from the study were imputed as no response from the visit after the first missed IP dose for the primary and major secondary efficacy outcomes other than time to renal-related event or death; the latter had data censored from the first missed treatment dose.

For the primary outcome (PERR at week 104), both the week 100 and week 104 visits were used to derive the outcome. If a patient missed the week 100 or week 104 visit, the week 96 renal response was used in place of the missing visit. If the patient was missing both week 100 and week 104 visits or had only 1 nonmissing visit among weeks 96, 100, and 104, the patient was imputed as no response. Imputation for missing visits was limited to the current visit and 2 previous visits, thereby limiting reproducibility to 3 visits. This 3-visit rule was used for determining renal response at all visits in which renal response is derived. The same rules were applied if either lab components (eGFR or uPCR) were missing for the week 100 or week 104 visits.

For end points for renal-specific measures, including SLEDAI-S2K organ improvement or worsening, serum creatinine doubling, or progression to ESRD, a last observation carried forward strategy was used in which the last observed value on or before the intercurrent event was carried forward for all subsequent time points through week 104. The secondary outcome of SDI used a worst observation carried forward imputation method and was presented for data for both on-treatment and on-study analyses.

Censoring Rules for Time-to-Event Analyses

For the key secondary outcome of time to renal-related event or death, intercurrent events of IP discontinuation, treatment failure not related to renal event, loss to follow-up, and study withdrawal resulted in censoring unless they occurred after a renal-related event. Death and treatment failure related to renal event were counted as an event if they occurred on or before the events that resulted in censoring. For the secondary outcome of time to first maintained PERR and CRR, IP discontinuation, treatment failure, or withdrawal were imputed as no response. For the secondary outcome of time to severe SFI flares, a treatment failure was imputed as an event, while IP discontinuation and study withdrawal were censoring events.

Subgroup Analyses

Of the subgroups of interest to this review, the BLISS-LN study conducted preplanned analyses based on baseline renal biopsy class (class III or class IV, versus class III + V or class IV + V, versus class V) and induction regimen (cyclophosphamide versus mycophenolate mofetil).

Sensitivity Analyses

In the BLISS-LN trial, various preplanned sensitivity analyses were conducted to investigate the impact of imputing intercurrent events as no response for the primary and key secondary outcomes. Prespecified tipping point analyses were conducted on the outcomes of PERR at week 104, CRR at week 104, and time to renal-related event or death to evaluate the robustness of the results of the conclusions when the intercurrent event assumptions are changed. Additional sensitivity analyses were conducted as shown in Table 10.

Table 10: Statistical Analysis of Efficacy Outcomes in the BLISS-LN Trial

ANCOVA = analysis of covariance; CRR = complete renal response; eGFR = estimated glomerular filtration rate; ESRD = end-stage renal disease; GFR = glomerular filtration rate; IP = investigational product; NR = not reported; ORR = ordinal renal response; PERR = primary efficacy renal response; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SFI = SELENA SLEDAI Flare Index; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria; uPCR = urine protein-creatinine ratio.

^aTo be performed only if ≥ 15% of patients were excluded from the modified intention-to-treat population.

^bPerformed if PERR at week 52 is statistically significant per the testing hierarchy.

Sources: Clinical Study Report for BLISS-LN;¹³ Rovin et al. (2022).²⁶

Multiplicity Testing

The primary and the 4 major secondary efficacy outcomes were evaluated based on a prespecified stepdown sequential testing procedure to control the overall type I error rate. If the prior end point was statistically significant (P < 0.05), then testing of the next sequential end point would proceed. The primary and 4 major secondary outcomes were evaluated for statistical significance in this prespecified approach, as follows:

• PERR at week 104

• CRR at week 104

• PERR at week 52

• Time to renal-related event or death

• ORR at week 104

Analyses of other efficacy outcomes, including the individual components of these composite outcomes, were not subject to any multiplicity adjustment.

Results

Patient Disposition

A summary of the patient disposition in the pivotal trial is available in Table 11. In BLISS-LN, 797 patients were screened for eligibility into the trial, 43.8% of whom either did not meet the trial’s inclusion criteria or did meet the exclusion criteria. A total of 224 patients were randomized into each of the belimumab and placebo groups. There were fewer study discontinuations in the belimumab group (16.6%) than in the placebo group (24.2%). The major reason for study discontinuation was withdrawal by patient (8.5% belimumab and 11.7% placebo), followed by AEs (3.1% and 4.5%) and investigator discretion (2.2% and 4.9%). A total of 34.5% of patients in the belimumab group and 40.8% in the placebo group prematurely discontinued the IP, with the main reasons including an AE (13.5% belimumab and 13.5% placebo), lack of efficacy (8.1% and 9.0%), or patient meeting protocol-defined stopping criteria (5.4% and 9.9%).

Table 11: Patient Disposition in the BLISS-LN Trial

IP = investigational product; mITT = modified intention to treat; PP = per protocol.

Source: Clinical Study Report for BLISS-LN.¹³

Table 12: Duration of Treatment Exposure Through Week 104 (mITT)

mITT = modified intention to treat; SD = standard deviation.

^aDuration of exposure (days) = (last infusion date – first infusion date + 28). Only complete dates were used when calculating duration of exposure. First and last infusion dates were used, regardless of any missed doses.

^bBased on the visit week of the last infusion received.

Source: Clinical Study Report for BLISS-LN.¹³

Exposure to Study Treatments

Exposure data from the BLISS-LN trial are summarized in Table 12. The mean (SD) treatment duration was 577 (243.5) days in the belimumab group and 546 (251.3) days in the placebo group. Most patients in both treatment groups had a duration of exposure > 76 weeks (69.5% belimumab versus 63.7% placebo), and most received 19 to 27 infusions (71.7% versus 65.5%).

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported below. Results of the subgroup analyses were available for certain outcomes and will be presented under each respective efficacy outcome. Detailed efficacy data are available in Appendix 3.

Renal Response Activity

Primary Efficacy Renal Response

Key results of the primary outcome, PERR at week 104, are summarized in Table 13. There was a statistically significant difference in the number of patients who achieved a PERR in the belimumab group versus the placebo group (43.0% versus 32.3%), with an adjusted between-group difference of 10.66% (95% CI, 1.89 to 19.42; P = 0.0311). A larger proportion of patients in the treatment group met each component of the composite outcome than in the placebo group; however, the difference was not statistically significant for the eGFR component, with an adjusted between-group difference of 7.10%, (95% CI, –1.84 to 16.04; P = 0.1599). A greater percentage of patients in the belimumab group achieved a PERR as of week 24, and this difference was maintained through week 104 (Figure 3). A total of 43.0% of patients in the belimumab group and 32.3% in the placebo group achieved a PERR that was maintained through to week 104. The HR for time to first PERR that was maintained through week 104 was 1.46 (95% CI, 1.07 to 1.98).

Table 13: Summary of Key Response Variables in the BLISS-LN Trial (mITT)

CI = confidence interval; CMH = Cochran-Mantel-Haenszel; CRR = complete renal response; eGFR = estimated glomerular filtration rate; HR = hazard ratio; mITT = modified intention to treat; OR = odds ratio; ORR = ordinal renal response; PERR = primary efficacy renal response; uPCR = urine protein-creatinine ratio.

^aCMH estimates are adjusted for induction regimen (cyclophosphamide vs. mycophenolate mofetil) and race (Black vs. non-Black).

^bOR (95% CI) and P value are from a logistic regression model for comparison between belimumab and placebo, with covariates treatment group, induction regimen (cyclophosphamide vs. mycophenolate mofetil), race (Black vs. non-Black), baseline uPCR, and baseline eGFR.

^cFrom Cox proportional hazards model for comparison between belimumab and placebo, adjusting for induction regimen (cyclophosphamide vs. mycophenolate mofetil), race (Black vs. non-Black), baseline uPCR, and baseline eGFR.

^dP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

^ePatients did not take a protocol-prohibited or -restricted medication or dose. The same treatment failure rules apply for PERR and CRR.

^fStudy withdrawal, treatment failures, and investigational product discontinuation imputed as “no response.” Investigational product discontinuation or treatment failure not related to renal disease or study withdrawal are censored in the time-to-event analysis.

^gEvents are defined as the first event experienced among the following: death, progression to end-stage renal disease, doubling of serum creatinine from baseline, renal worsening, or renal-related treatment failure. Patients who discontinue randomized treatment, withdraw from the study, or are lost to follow-up are censored on the date. Patients who complete the 104-week treatment period are censored at the week 104 visit. Time to event is defined as (event date – treatment start date + 1).

Source: Clinical Study Report for BLISS-LN.¹³

In BLISS-LN, a statistically significant difference was observed in the belimumab group in the proportion of patients who achieved the key secondary outcome of PERR at week 52 (46.6% versus 35.4%), with an adjusted between-group difference of 11.12% (95% CI, 2.25 to 19.99; P = 0.0245).

Various sensitivity analyses were conducted on the primary outcome, all of which were consistent with the result of the primary efficacy analysis. For the primary outcome, the tipping point analysis varying the no response assumption for all withdrawn patients regardless of treatment found that 21.2% of those in the placebo group considered no response needed to be altered to response to tip the conclusion from statistical significance to nonsignificance, assuming no additional patients on belimumab were considered response; however, this was deemed highly implausible.

Subgroup analysis: A summary of PERR at week 104 stratified by baseline renal biopsy class and induction regimen subgroups is presented in Appendix 3. Results numerically favoured the belimumab group for all subgroups except for the baseline renal biopsy class V and the cyclophosphamide followed by azathioprine subgroups, the results of which were not statistically significantly different between treatment groups. However, the study was not designed or powered to evaluate efficacy in subgroups, and the small number of patients in the class V and the cyclophosphamide followed by azathioprine subgroups might have led to the lack of statistical significance between treatment groups.

Complete Renal Response

In BLISS-LN, as shown in Table 13, the key secondary outcome of CRR at week 104 was statistically significant in favour of belimumab (30.0% versus 19.7%; adjusted between-group difference of 10.27% [95% CI, 2.40 to 18.14; P = 0.0167]). A total of 30.0% of patients in the belimumab group and 19.7% in the placebo group had achieved a CRR that was maintained through to week 104. The HR for time to first CRR that was maintained through week 104 was 1.58 (95% CI, 1.08 to 2.31).

As shown in Figure 5, a greater proportion of patients in the belimumab group than in the placebo group achieved a CRR at each visit; however, CIs overlapped until approximately week 72, after which a separation was maintained through week 104.

Various sensitivity analyses and tipping point analyses were conducted on the outcome of CRR at week 104, all of which were consistent with the result of the primary efficacy analysis.

Subgroup analysis: A summary of CRR at week 104 stratified by baseline renal biopsy class and induction regimen subgroups is presented in Appendix 3. Results numerically favoured the belimumab group for all subgroups except for the baseline renal biopsy class V and the cyclophosphamide followed by azathioprine subgroups, the results of which were not statistically significantly different between treatment groups. However, the study was not designed or powered to evaluate efficacy in subgroups, and the small number of patients in the class V and the cyclophosphamide followed by azathioprine subgroups might have led to the lack of statistical significance between treatment groups.

Ordinal Renal Response

For the key secondary outcome of ORR at week 104, there was a statistically significant treatment difference between the belimumab and placebo groups in the composite outcome of patients with complete response, partial response, and no response (P = 0.0096), as shown in Table 13. The proportion of patients with complete response was higher in the belimumab group than in the placebo group (30.0% versus 19.7%), and the proportion of patients with partial response was similar between the belimumab group and the placebo group (17.5% versus 17.0%). A higher proportion of patients in the placebo group were had no response (63.2%) than in the belimumab group (52.5%).

The results of sensitivity analyses of ORR including urinary sediment at week 104 and ORR including urinary sediment using calculated GFR at week 104 were supportive of the primary analysis.

Reduction in Proteinuria

A secondary outcome in the BLISS-LN trial was the proportion of patients achieving uPCR less than or equal to 0.7 g/g at week 104, the results of which, as shown in Table 13, were higher for patients in the belimumab group compared with the placebo group (44.4% versus 33.6%), with an OR of 1.54 (95% CI, 1.04 to 2.29). Also, a higher proportion of patients in the belimumab group than in the placebo group achieved a uPCR less than 0.5 g/g at week 104 (39.5% versus 28.7%), with an OR of 1.58 (95% CI, 1.05 to 2.38).

As shown in Table 15, the mean absolute change from baseline to week 104 in uPCR was –2.33 (SD = 2.88) for the belimumab group (n = 138) and –2.43 (SD = 2.65) in the placebo group (n = 128). The mean percent change from baseline to week 104 in uPCR was –70.66% (SD = 43.98%) and –62.36% (SD = 49.43%) for the belimumab and placebo groups, respectively.

Increase or Stabilization of eGFR

As shown in Table 13, at week 104 in the BLISS-LN trial, a higher proportion of patients in the belimumab group than in in the placebo group achieved an eGFR no more than 20% below pre-flare value or at least 60 mL/min/1.73 m² (57.4% versus 50.2%); however, no statistically significant difference was detected between groups, with an adjusted difference of 7.10 (95% CI, –1.84 to 16.04; P = 0.1599). Similarly, a higher proportion of patients in the belimumab group than in the placebo group achieved an eGFR at week 104 that was no more than 10% below pre-flare value or within the normal range (≥ 90 mL/min/1.73 m²) (46.6% versus 39.9%); however, no statistically significant difference was detected between groups, with an adjusted difference of 6.68 (95% CI, –2.31 to 15.67; P = 0.1539).

As shown in Table 15, there was a numerical increase in the mean (SD) observed GFR values from baseline to week 104 for the belimumab group, which increased from 100.2 (37.68) mL/min/1.73 m² to 111.3 (35.75) mL/min/1.73 m² and decreased in the placebo group from 101.0 (42.61) mL/min/1.73 m² to 100.8 (29.18) mL/min/1.73 m².

In a post hoc analysis of on-study patients in the BLISS-LN trial, fewer patients in the belimumab group than in the placebo group experienced a 30% decline in eGFR (8.5% versus 17.0%) between baseline and week 104, with an HR of 0.47 (95% CI, 0.27 to 0.83), and fewer patients in the belimumab group than in the placebo group experienced a 40% decrease in eGFR (4.5% versus 11.7%) between baseline and week 104, with an HR of 0.35 (95% CI, 0.17 to 0.74).

In a post hoc analysis, a sustained 30% decline in eGFR by the end of the study (as confirmed by the last 2 observed eGFR measurements) was reported by fewer patients in the belimumab group than in the placebo group (3.6% versus 11.2%), with a OR of 0.29 (95% CI, 0.13 to 0.68). A sustained 40% decline in eGFR was reported by 1.8% and 6.7% of patients in the belimumab and placebo groups, respectively, with an OR of 0.25 (95% CI, 0.08 to 0.78).

Reduction in Corticosteroid Use

In BLISS-LN, corticosteroid use was converted to a prednisone-equivalent dose. As shown in Table 15, a higher proportion of patients in the belimumab group compared with the placebo group received an average daily prednisone dose of less than or equal to 7.5 mg since the previous visit (40.8% versus 29.6%), with an OR of 1.65 (95% CI, 1.11 to 2.45). A total of 82 patients (36.8%) in the belimumab group received an average daily prednisone dose less than or equal to 5 mg since the previous visit, compared to 62 (27.8%) in the placebo group, with an OR of 1.51 (95% CI, 1.01 to 2.27) as of week 104.

Time to Renal-Related Event or Death

As shown in Table 13, there was a statistically significant difference in the key secondary outcome of risk of a renal-related event or death through to week 104 between patients in the belimumab and placebo groups (15.7% versus 28.3%), with an HR of 0.51 (95% CI, 0.34 to 0.77; P = 0.0014). As shown in Table 14, most events that contributed to this composite end point were renal worsening (7.6% belimumab versus 17.5% placebo) and renal-related treatment failures (7.2% belimumab versus 9.0% placebo).

Table 14: Number of Renal-Related Events or Deaths by Group in the BLISS-LN Trial (mITT)

ESRD = end-stage renal disease; mITT = modified intention to treat.

^aRepresents first event for each patient with an event. A patient who died following a renal-related event will be included in this analysis as a renal-related event, not a death.

^bProgression to ESRD is defined as the need for chronic dialysis or renal transplant.

^cDoubling of serum creatinine compared with baseline, confirmed with a second measurement at least 3 weeks later.

^dRenal worsening is defined by increased proteinuria (a reproducible increase in urine protein-creatinine ratio to > 1 g if the baseline value was < 0.2 g, to > 2 g if the baseline value was between 0.2 g and 1 g, or more than twice the value at baseline if the baseline value was > 1 g) or impaired renal function (a reproducible decrease in glomerular filtration rate of > 20%, accompanied by proteinuria [> 1] and/or cellular [red blood cell and/or white blood cell] casts).

^eBased on adjudication of treatment failures.

Source: Clinical Study Report for BLISS-LN.¹³

Reduction in Chronic Kidney Disease Staging

This outcome was not reported in the BLISS-LN trial.

Disease Activity

Systemic Lupus Erythematosus Disease Activity Index 2000 With Modified Scoring for Proteinuria

As shown in Table 15, the LS mean change from baseline in SLEDAI-S2K at week 104 was –7.7 (SE = 0.46) in the belimumab group and –6.1 (SE = 0.47) in the placebo group, with an LS mean difference of –1.5 (95% CI, –2.4 to –0.6). The proportion of patients with a SLEDAI-S2K score less than 4 at week 104 was higher in the belimumab group compared with the placebo group (27.8% versus 18.4%), with an OR of 1.76 (95% CI, 1.11 to 2.78).

Table 15: Summary of Other Secondary Response Variables in the BLISS-LN Trial (mITT)

ANCOVA = analysis of covariance; BSA = body surface area; CI = confidence interval; GFR = glomerular filtration rate; HR = hazard ratio; LS = least squares; mITT = modified intention to treat; NR = not reported; OR = odds ratio; SD = standard deviation; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SE = standard error; SFI = SELENA SLEDAI Flare Index; SLEDAI-S2K = Systemic Lupus Erythematosus Disease Activity Index 2000 with modified scoring for proteinuria; uPCR = urine protein-creatinine ratio.

^aWeek 104 statistics are from an ANCOVA model comparing belimumab and placebo, with covariates for treatment group, baseline SLEDAI-S2K score, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^bP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

^cOR (95% CI) and P value are from a logistic regression model for comparison between belimumab and placebo, with covariates treatment group, baseline SLEDAI-S2K score, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^dWeek 104 P value is from a rank ANCOVA model comparing belimumab and placebo, with covariates for treatment group, baseline SDI score, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^eP value is from a rank ANCOVA model comparing belimumab and placebo, with covariates for treatment group, baseline uPCR value, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^fSteroids were converted to prednisone equivalent. The prednisone average daily dose since previous visit was calculated at every 4-week visit after baseline. All prednisone doses since the visit 4 weeks prior, up to and including the current visit, were summed and divided by the number of days in the period. Days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation for average daily prednisone dose.

^gOR (95% CI) and P value are from a logistic regression model for comparison between belimumab and placebo, with covariates treatment group, baseline prednisone dose, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^hAnalysis excludes severe flares triggered only by an increase in SELENA SLEDAI score to > 12. Treatment failure is considered an event. Patients who discontinue randomized treatment, withdraw from study, are lost to follow-up, die, or complete week 104 are censored at the last flare assessment date, the death date, or the week 104 study visit. Time to first severe flare is defined as (event date – treatment start date + 1).

ⁱOnly includes postbaseline severe flares.

^jFrom Cox proportional hazards model for the comparison between belimumab and placebo adjusting for induction regimen (cyclophosphamide vs. mycophenolate mofetil) and race (Black vs. non-Black).

Source: Clinical Study Report for BLISS-LN.¹³

Organ Damage

Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index

In the pivotal trial, a similar proportion of patients in the belimumab group (n = 138) and the placebo group (n = 129) experienced SDI worsening (change > 0) at week 104 compared with baseline (7.8% versus 6.5%), with an OR of 0.85 (95% CI, 0.33 to 2.18) (Table 15).

Disease Flare Frequency and Severity

SELENA SLEDAI Flare Index As shown in Table 15, in the BLISS-LN trial, the percentage of patients who experienced a severe SFI flare postbaseline through to week 104 was 18.8% in the belimumab group and 31.4% in the placebo group. The risk of experiencing a severe flare at any time, based on the SFI, was lower in patients in the belimumab group than in the placebo group, with an HR of 0.57 (95% CI, 0.39 to 0.84). Of the patients experiencing a severe flare, the median study day to first severe flare was 204 for the belimumab group and 262.5 for the placebo group.

Health-Related Quality of Life

This outcome was not reported in the BLISS-LN trial.

Reduction in Symptoms

This outcome was not reported in the BLISS-LN trial.

Serological Outcomes

Decrease in Anti-dsDNA Antibodies

As shown in Table 16, in the BLISS-LN trial, the mean change in anti-dsDNA antibody levels from baseline to week 104 was greater for the belimumab group than the placebo group, with an LS mean change from baseline of –188.6 IU/mL (SD = 629.3 IU/mL) in the belimumab group and –30.6 IU/mL (SD = 682.0 IU/mL) in the placebo group. Among the patients who were anti-dsDNA antibody positive (≥ 30 IU/mL) at baseline (n = 38 belimumab; n = 14 placebo), a greater proportion of patients in the belimumab group shifted to negative (< 30 IU/L) at week 104 than in the placebo group (35.5% versus 14.4%).

Table 16: Summary of Serological Outcomes of Interest in BLISS-LN (mITT)

anti-dsDNA = anti–double-stranded DNA; CI = confidence interval; LS = least squares; mITT = modified intention to treat; SD = standard deviation; SE = standard error.

^aOnly on-treatment data are displayed. Week 104 statistics are from an analysis of covariance model comparing belimumab and placebo, with covariates for treatment group, baseline complement value, induction regimen (cyclophosphamide vs. mycophenolate mofetil), and race (Black vs. non-Black).

^bP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

^cAnti-dsDNA positive (≥ 30 IU/mL); anti-dsDNA negative (< 30 IU/mL).

^dBaseline status is defined as low (C3 < 90 mg/dL; C4 < 10 mg/dL) or normal or high (C3 ≥ 90 mg/dL; C4 ≥ 10 mg/dL).

Source: Clinical Study Report for BLISS-LN.¹³

Increase in Complement C3 and C4 Levels

As shown in Table 16, for both complement C3 and C4, the mean change from baseline to week 104 was greater for the belimumab group than the placebo group, with an LS mean difference of 6.1 (95% CI, 0.2 to 12.0) in complement C3 and 3.0 (95% CI, 1.3 to 4.7) in complement C4. Among patients with low C3 at baseline (n = 82 belimumab; n = 69 placebo), the percentage of patients who shifted to normal or high at week 104 was 58.5% in the belimumab group and 46.4% in the placebo group. Among patients with a low C4 at baseline (n = 42 belimumab; n = 34 placebo), the percentage of patients who shifted to normal or high at week 104 was 81.0% in the belimumab group and 61.8% in the placebo group.

Harms

Only those harms identified in the review protocol are reported below. Refer to Table 17 for detailed harms data.

Table 17: Summary of Harms (Safety Population) in the BLISS-LN Trial

NMSC = nonmelanoma skin cancer; SAE = serious adverse event; UTI = urinary tract infection.

^aFrequency > 5%.

^bFrequency ≥ 2 patients.

^cIncludes all deaths that occurred during the double-blind phase, including off treatment.

^dDeveloped fatal SAEs while on study treatment; death may have occurred anytime thereafter.

^eFatal SAEs occurred after the on-treatment period.

Source: Clinical Study Report for BLISS-LN.¹³

Adverse Events

The proportion of patients experiencing at least 1 AE in the BLISS-LN trial was similar between the belimumab group (95.5%) and the placebo group (94.2%). The most common AEs were upper respiratory tract infection (32.1% belimumab versus 31.3% placebo), diarrhea (18.8% versus 20.1%), UTI (19.2% versus 15.6%), headache (13.4% versus 15.6%), nasopharyngitis (13.8% versus 12.9%), and arthralgia (10.3% versus 14.7%).

Frequent AEs that occurred more commonly in the belimumab group than in the placebo group were UTI (19.2% versus 15.6%), cough (12.5% versus 8.5%), and upper abdominal pain (6.3% versus 2.7%).

Serious AEs

The proportion of patients experiencing at least 1 SAE was similar between treatment groups (25.9% belimumab versus 29.9% placebo). The most common SAEs were pneumonia (4.0% versus 3.1%), herpes zoster (1.8% versus 0.9%), gastroenteritis (0% versus 2.2%), lung infection (0.9% versus 1.3%), LN (0.4% versus 1.8%), and UTI (1.3% versus 0.9%).

IP Discontinuations Due to AEs

There was a similar proportion of IP discontinuation due to AEs in the belimumab group and the placebo group (12.9% versus 12.9%), with the most common reason for withdrawal for both groups being pneumonia (2.2% versus 0.4%) and lupus nephritis (1.8% versus 1.3%),

Mortality

Eleven deaths occurred during the double-blind phase of the BLISS-LN trial, mainly due to infections, with 6 deaths (2.7% of patients) in the belimumab group and 5 deaths (2.2% of patients) in the placebo group (Table 17). In 7 patients (4 belimumab; 3 placebo), the deaths occurred while on study treatment, 5 of which were deemed related to the IP by the investigator. In the remaining 4 patients (2 belimumab; 2 placebo), the deaths occurred posttreatment,1 of which was deemed related to the IP.

Notable Harms

Common notable harms (as outlined in the CADTH protocol) included postinfusion-related systemic reactions (11.6% belimumab versus 12.9% placebo); serious infections of herpes zoster (2.2% versus 0.9%), active tuberculosis (0.9% versus 0.4%), and sepsis (0% versus 0.4%); malignancies (including nonmelanoma skin cancer) (1.3% versus 0%); and serious suicidal behaviour (0.4% versus 0%).

Critical Appraisal

Internal Validity

In the BLISS-LN trial, patients were allocated to treatment groups with randomization stratified by relevant prognostic factors of patient induction and maintenance regimen (high-dose corticosteroids plus cyclophosphamide followed by azathioprine versus high-dose corticosteroids plus mycophenolate mofetil followed by mycophenolate mofetil) and race (Black versus non-Black) to reduce the risk of confounding. The primary and key secondary end points were controlled for multiplicity, but all other end points — as well as all post hoc analyses — were not. As a result, outcomes assessed other than the primary and key secondary end points may be at risk of a type I error and should be viewed as supportive evidence for the overall effects of belimumab. The primary outcome for BLISS-LN was PERR at week 104, which was changed after initiating patient recruitment from the original primary end point of ORR (complete, partial, no response) at week 104, where renal response was determined by changes in urinary sediment, proteinuria, and renal function. The changes to the primary end point were made before unblinding of the study results; therefore, the risk of operational bias is low. An mITT population was used that included all patients who received at least 1 dose of the IP. Given that the vast majority of patients were included in the mITT population (99.6%), this analysis population would be unlikely to introduce bias into the study.

Overall, baseline demographics (e.g., sex, race, age, and gender) and disease characteristics (e.g., mean eGFR and uPCR) were generally similar and balanced between groups in the BLISS-LN trial. There was, however, a greater percentage of patients in the belimumab group than in the placebo group taking antimalarials (74.4% versus 69.1%) and steroids (96.9% versus 94.2%) at baseline. The belimumab group had a slightly lower mean dosage of prednisone than the placebo group (mean = 66.5 mg/day [SD = 99.6 mg/day] mg/day versus mean = 72.5 mg/day [SD = 133.2 mg/day]). The clinical expert consulted for this review stated that these differences were minimal and not likely to bias results. Overall, it is unclear how this may have affected the results as the differences may have underestimated or overestimated the effects of belimumab relative to placebo.

The BLISS-LN trial reported power calculations for the primary end point and detected a statistically significant difference between the belimumab group and the placebo group for the primary outcome and key secondary outcomes, while adhering to their statistical testing hierarchy for the multiplicity adjustment. The primary efficacy outcome was a composite outcome, which was considered to be appropriate and clinically relevant by the clinical expert consulted. Other outcomes of interest to this review (e.g., increase or stabilization of eGFR and change in SLEDAI-S2K scores from baseline) were not controlled for multiplicity. Moreover, analyses for time to 30% and 40% decline in eGFR from baseline and sustained 30% and 40% decline in eGFR were conducted post hoc. Furthermore, as discussed in Appendix 4, the SLEDAI-S2K, as it uses a single weighted score to summarize disease activity, on the one hand standardizes the judgment of disease activities, while on the other it could mask the underlying importance of organ systems that are contributing to the total score (i.e., the same score could represent multiple mild disease in many organs or severe disease in a single organ, or an unchanged score may occur despite worsening in 1 organ system if there is also improvement in another system).

A relatively high percentage of patients withdrew from the BLISS-LN study in each of the belimumab (16.6%) and placebo (24.2%) groups, with the primary reasons for study discontinuation being withdrawal of consent, AE, and investigator discretion. A slightly higher proportion of patients in the placebo group than in the belimumab group discontinued due to withdrawal of consent (11.7% versus 8.5%) and due to investigator discretion (4.9% versus 2.2%). Patients who discontinued the study were classified as no response, and when more patients discontinued in the placebo group, this may have biased the results in favour of belimumab as these patients were considered no response whether they were responding at the time of discontinuation or not. However, sensitivity and tipping point analyses performed by the sponsor, which examined the extent to which results were affected by response assumptions for intercurrent events, were generally supportive of the findings of the primary analyses.

Patient HRQoL was identified as an important outcome by the patient groups providing input for this review, specifically symptoms such reduction in fatigue, flares, pain, rash, and skin irritations. However, HRQoL was not assessed in the BLISS-LN trial; hence, it is unknown what impact belimumab would have on HRQoL.

Regarding calculations of patients’ average daily prednisone dose in the BLISS-LN trial, days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation. This assumption may have led to lower estimates of average prednisone doses in both the belimumab and placebo groups. It is unclear in which direction this may have biased results as the number of missing doses in each treatment group is unknown.

For the subgroups identified as important to this review (i.e., baseline renal biopsy class and induction regimen), the analyses were exploratory; the study was not designed or powered to evaluate efficacy in subgroups, and results were limited by the small number of patients in these subgroups. With these limitations in mind, overall subgroup analyses should be viewed as supportive evidence for the overall effects of belimumab.

External Validity

Although there was only 1 site in Canada in the BLISS-LN trial, the clinical expert consulted by CADTH agreed that the baseline patient characteristics in regard to age, gender, race, and disease activity in the BLISS-LN trial were reflective of patients they see in Canadian clinical practice for the present indication. At baseline, a greater proportion of patients were assigned to mycophenolate mofetil induction therapy (73.5%) than to cyclophosphamide as per the investigator’s choice. The clinical expert consulted for this review stated that this is reflective of Canadian clinical practice and many patients with LN may be considering pregnancy and cyclophosphamide is associated with premature ovarian failure and infertility. Overall, the clinical expert consulted felt the characteristics of the patient population enrolled in the trials were a good representation of the target population and did not identify any issues with the use of concurrent treatments or conduct of the trials that could substantially affect the generalizability of the findings.

The product monograph for belimumab authorized both IV and subcutaneous formulations for LN. However, this recommendation was based on extrapolated data, and there is no clinical evidence regarding the subcutaneous formulation for patients with active LN.

The BLISS-LN trial did not include rituximab as a comparator; therefore, the efficacy and harms of belimumab in addition to standard of care compared to the addition of rituximab to standard of care in the treatment of LN is unknown. However, rituximab is used as a salvage therapy for refractory cases of LN, therefore it may not be appropriate to include it as a comparator in addition to standard of care.

Indirect Evidence

Objectives and Methods for the Summary of Indirect Evidence

No direct comparative data for the use of belimumab in the adult LN population were identified in the systematic review; thus, a search for indirect evidence was conducted.

A focused literature search for indirect treatment comparisons (ITCs) dealing with LN was run in MEDLINE All (1946–) on August 4, 2022. No limits were applied to the search. The search results were screened by 1 researcher to identify ITCs that met the patient, intervention, comparator, and outcome criteria in the review protocol (Table 5). No ITC was submitted by the sponsor; however, the indirect comparison feasibility assessment included in the sponsor’s submission was reviewed.

No relevant ITCs were found in the literature search. CADTH’s review of the sponsor’s feasibility assessment is provided subsequently.

Appraisal of the Feasibility Assessment

The sponsor conducted a feasibility assessment to determine if the clinical trials for belimumab and rituximab in patients with LN were sufficiently similar to permit valid comparison in an ITC. Four studies were evaluated: the BLISS-LN¹³ trial (belimumab plus mycophenolate mofetil or cyclophosphamide versus mycophenolate mofetil or cyclophosphamide), an open-label pilot study by Li et al. (2009)³⁵ (rituximab versus rituximab plus cyclophosphamide), the LUNAR trial¹⁸ (rituximab plus mycophenolate mofetil versus mycophenolate mofetil), and a study by Zhang et al. (2015)³⁶ (rituximab plus cyclophosphamide versus cyclophosphamide).

The authors of the feasibility assessment determined that it was possible to construct 2 stratified evidence networks, which essentially splits BLISS-LN into 2 studies to conduct separate network meta-analyses for cyclophosphamide and for mycophenolate mofetil while preserving randomization. However, the feasibility assessment identified several important differences in baseline characteristics and potential effect modifiers — including dosing regimens, gender, ethnicity, geography, duration of LN, mean complement C3, mean complement C4, and LN class — between studies and concluded it was not possible to generate robust estimates of the comparative treatment effects due to between-study heterogeneity.

Based on the information presented in the sponsor’s feasibility assessment, the CADTH reviewer agrees that the heterogeneity between the studies is significant; thus, any ITC is unlikely to produce robust estimates of comparative efficacy or safety. There were differences in enrolment criteria related to baseline renal biopsy class as Zhang et al. only enrolled patients with severe refractory LN. The mean age was generally consistent across studies: between 29 years and 40 years. However, sex and racial distributions were not consistent across trials. The mycophenolate mofetil doses were consistent across trials. However, cyclophosphamide doses varied between 500 mg and 800 mg across trials, and cyclophosphamide dosing regimens varied. In Zhang et al., the cyclophosphamide dosing regimen was 800 mg at weeks 1 and 3 in the treatment group and 800 mg monthly for 48 weeks in the comparator group, as opposed to 500 mg biweekly for 12 weeks total in both treatment groups in the BLISS-LN trial. Furthermore, trials had inconsistent definitions for the end point of CRR as only the BLISS-LN trial used eGFR in its criteria for CRR, while only the LUNAR trial incorporated serum creatinine levels, and only the Zhang et al. trial included serum albumin levels in its criteria. Other key end points such as uPCR, eGFR, and harms-related data were not reported in the Zhang et al. study,³⁶ which would disconnect the cyclophosphamide network, preventing the conduct of network meta-analysis. Furthermore, the sample size of the study population in Li et al.³⁷ (n = 16) was limited and unlikely to be a representative population of patients with LN, which may limit the ability to use methods such as unanchored matching-adjusted indirect comparison with Zhang et al.

Other Relevant Evidence

This section includes submitted long-term extension studies and additional relevant studies included in the sponsor’s submission to CADTH that were considered to address important gaps in the evidence included in the systematic review.

Long-Term Extension Studies

Description of Study

The OLE phase of the BLISS-LN trial was a 28-week study that took place between February 2015 and March 2000. During the extension phase, all eligible patients received 10 mg/kg IV belimumab plus standard of care every 28 days, continued until week 24. A final evaluation was done at week 28, 4 weeks after receiving the last dose. Patients receiving belimumab or placebo through week 100 and completing all assessments of week 104 of the double-blind phase were considered eligible for participation in the extension study. Patients received the first dose at week 104 of the double-blind phase (marked as day 0 for the OLE phase) after completing all assessments of week 104. There were 2 groups in the extension phase: the placebo-to-belimumab group (patients switching from placebo to belimumab) (n = 122) and the belimumab-to-belimumab group (patients remaining on belimumab) (n = 132). Prohibited medications applicable during the double-blind phase (e.g., immunosuppressants and corticosteroids) were not applicable in the OLE phase (exceptions were live vaccines, biologics, and other investigational drugs).

Populations

Baseline characteristics showed a similar trend in both the placebo-to-belimumab and belimumab-to-belimumab groups. The characteristics were also similar to the baseline patient characteristics during the double-blind phase. The mean (SD) age of patients in the open-label phase was 36 (10) years in both the placebo-to-belimumab and belimumab-to-belimumab groups. Patients were predominantly female (90% and 89%) and predominantly Asian (55% and 54%) in the placebo-to-belimumab and belimumab-to-belimumab groups, respectively. Moreover, 82% and 86% of patients in the placebo-to-belimumab and belimumab-to-belimumab groups, respectively, had LN class III or IV (with or without class V).

At OLE baseline, patients in the belimumab-to-belimumab group compared with the placebo-to-belimumab group had lower median uPCR values (0.3 g/g versus 0.4 g/g), a higher proportion of patients with SLEDAI-S2K scores less than 8 (n = 103 [78%] versus n = 84 [69%]), a lower average daily prednisone-equivalent dose (median = 5.0 mg/day versus median = 7.5 mg/day), a lower proportion of patients with positive biomarkers (anti-dsDNA: n = 64 [49%] versus n = 85 [70%]; anti-C1q: n = 60 [55%] versus n = 60 [68%]), and fewer patients with low C3 or C4 levels (C3: n = 37 [28%] versus n = 45 [37%]; C4: n = 12 [9%] versus n = 18 [15%]).

Statistical Analysis

For the open-label phase of the BLISS-LN study, no hypothesis testing of treatment differences was conducted. The last available value before the first dose of treatment on day 0 was defined as the open-label baseline value. Descriptive statistics were used to summarize safety and efficacy end points, assessed only for the duration of the open-label phase, unless otherwise specified. Analyses on end points were conducted based on observed data with no imputation for withdrawal. All enrolled patients receiving 1 or more doses of open-label belimumab treatment were defined as the open-label safety population. While safety end points for the open-label phase were assessed in the safety population, 1 patient was excluded from efficacy analysis for issues related to good clinical practice nonadherence. The efficacy population was defined as the mITT population for the open-label phase. For the post hoc analyses of PERR and CRR using the double-blind criteria, withdrawal from the trial, treatment failure, and discontinuation of study treatment were imputed as nonresponse.

Patient Disposition

A total of 257 of 448 patients (57.4%) from the double-blind phase enrolled in the open-label phase. Of these, 255 were included in the safety population and 254 in the open-label mITT population, receiving 1 or more doses of 10 mg/kg IV belimumab. In total, 122 and 132 patients were placed in the placebo-to-belimumab group and the belimumab-to-belimumab group, respectively. Of these 254 patients, 9 withdrew from the study, leaving 245 (96.5%) completing the open-label phase through week 28. Among the patients who withdrew, more than half did so due to AEs (55.6%), following withdrawal of consent (22.2%), protocol deviation (11.1%), and loss to follow-up (11.1%).

Exposure to Study Treatments

The median (range) duration of exposure to belimumab 10 mg/kg during the OLE study was 196 (56 to 214) days in the belimumab-to-belimumab group and 196 (85 to 207) days in the placebo-to-belimumab group (following approximately 2 years of blinded exposure to belimumab).

Efficacy

The efficacy end points (assessed as secondary end points after safety assessments) for the OLE study were to evaluate the proportions of patients with PERR and the proportions of patients with CRR at week 28. During the open-label phase, PERR was defined as follows: uPCR less than or equal to 0.7, eGFR no more than 20% below the open-label baseline eGFR or greater than or equal to 60 mL/min/1.73 m², and no prohibited medications. CRR was defined as follows: uPCR less than 0.5, eGFR no more than 10% below the open-label baseline eGFR or greater than or equal to 90 mL/min/1.73 m², and no prohibited medications. PERR and CRR were evaluated based on observed data at open-label weeks 12, 24, and 28, and criteria were required to be met at a single time point only, meaning criteria did not have to be met on consecutive visits as was required for the double-blind phase.

The post hoc analyses of PERR and CRR at open-label week 28 were performed according to the double-blind phase criteria to assess the maintenance of response over time; in other words, the eGFR component was relative to the pre-flare value, and the PERR and CRR criteria were required to be met on 2 consecutive visits (at week 100 and week 104 in the double-blind phase and at week 24 and week 28 in the open-label phase).

Using the OLE criteria, the number of patients with PERR and CRR increased from baseline to week 28 in both groups (PERR: 93 [71%] to 91 [75%] in the belimumab-to-belimumab group and 73 [60%] to 79 [67%] in the placebo-to-belimumab group; CRR: 63 [48%] to 76 [62%] in the belimumab-to-belimumab group and 44 [36%] to 57 [48%] in the placebo-to-belimumab group). In the post hoc analyses, the number of patients in the belimumab-to-belimumab group with double-blind criteria of PERR and CRR at baseline was 87 (66%) and 60 (46%), respectively, and 69 (52%) and 54 (41%), respectively, at week 28, showing a reduction from the baseline. In the placebo-to-belimumab group, the number of patients with double-blind criteria of PERR and CRR at baseline was 66 (54%) and 41 (34%), respectively; the number of patients was maintained through to week 28, with 64 (53%) and 43 (35%) achieving PERR and CRR, respectively.

In addition to PERR and CRR, other efficacy end points assessed during the open-label phase were proportions of patients with a uPCR less than 0.5 g/g from double-blind week 4 through open-label week 28 by visit; mean eGFR levels from double-blind baseline to open-label week 28 by visit; proportions of patients with SLEDAI-S2K score less than 4 at open-label week 28; proportions of patients with average daily prednisone-equivalent dose less than or equal to 5 mg, less than or equal to 7.5 mg, or less than or equal to 10 mg at open-label week 28; time to a sustained PERR and CRR from double-blind baseline through to open-label week 28 (using double-blind phase criteria; post hoc analyses); proportion of patients with a doubling of serum creatinine (compared with open-label baseline) and/or progression to kidney failure (defined as the need for chronic dialysis or renal transplant); and percentage changes from open-label baseline in biomarker levels (anti-dsDNA, complement C3 andC4) to open-label week 28 among patients who were anti-dsDNA positive and or had low C3 or C4 at open-label baseline.

In terms of efficacy end points, the median uPCR values remained stable both in the belimumab-to-belimumab group, from 0.3 g/g (interquartile range [IQR], 0.1 to 0.7 g/g) at OLE baseline to 0.2 g/g (IQR, 0.1 to 0.5 g/g) at week 28, and in the placebo-to-belimumab group, from 0.4 g/g (0.1 to 1.0 g/g) at OLE baseline to 0.4 g/g (0.1 to 0.9 g/g) at week 28. The number of patients with a uPCR less than 0.5 g/g increased in the belimumab-to-belimumab group from 89 (67%) at OLE baseline to 91 (75%) at week 28 and remained stable in the placebo-to-belimumab group, from 71 of 122 patients (58%) at OLE baseline to 68 of 118 patients (58%) at week 28. The median (interquartile range) eGFR values remained stable in the belimumab-to-belimumab group, from 108 mL/min/1.73 m² (IQR, 88 to 130 mL/min/1.73 m²) at OLE baseline to 106 mL/min/1.73 m² (88 to 127 mL/min/1.73 m²) at week 28, and remained stable in the placebo-to-belimumab group, from 105 mL/min/1.73 m² (IQR, 84 to 123 mL/min/1.73 m²) at OLE baseline to 104 mL/min/1.73 m² (81 to 124 mL/min/1.73 m²) at week 28. The number of patients who received an average daily prednisone dose of less than or equal to 7.5 mg remained stable in the belimumab-to-belimumab group, from 85 (64%) at OLE baseline to 83 (65%) at week 28 and remained stable in the placebo-to-belimumab group, from 62 (51%) at OLE baseline to 66 (55%) at week 28. There were no marked changes in the proportions of patients with SLEDAI-S2K scores less than 4 in either group, which ranged from 64 of 132 patients (49%) at OLE baseline to 64 of 122 patients (53%) at week 28 for the belimumab-to-belimumab group and from 44 of 122 patients (36%) at OLE baseline to 40 of 120 patients (33%) at week 28 for the placebo-to-belimumab group.

Harms

The primary objective of this OLE study was to evaluate the safety of belimumab for another 28 weeks. The proportion of patients experiencing at least 1 AE in the open-label phase was slightly higher in the belimumab-to-belimumab group (70%) than in the placebo-to-belimumab group (62%). The most common AEs by system organ class occurring in at least 5% of patients in either group included infections and infestations (49% versus 42%), musculoskeletal and connective tissue disorders (12% versus 13%), skin and subcutaneous tissue disorders (13% versus 8%), gastrointestinal disorders (10% versus 9%), and respiratory, thoracic, and mediastinal disorders (11% versus 4%) in the belimumab-to-belimumab and placebo-to-belimumab groups, respectively.

The proportion of patients with at least 1 treatment-related AE during the open-label phase was similar between the groups: 18% and 20% in the belimumab-to-belimumab and placebo-to-belimumab groups, respectively. The proportion of patients experiencing at least 1 SAE during the open-label phase was low (8% in the belimumab-to-belimumab group and 4% in the placebo-to-belimumab group) compared with the double-blind phase (25.9% belimumab versus 29.9% placebo). The proportion of withdrawals due to AEs was also very low in both groups (3% versus 0.8%).

Common notable harms included postinfusion systemic reactions (4% versus 3%) and infections of special interest (opportunistic infections, herpes zoster, tuberculosis, sepsis) (5% versus 2%) in the belimumab-to-belimumab versus placebo-to-belimumab groups, respectively. Two serious infections of special interest were reported in the belimumab-to-belimumab group, 1 for serious tuberculosis and another for serious disseminated herpes zoster. Among the 4 depression events reported in the belimumab-to-belimumab group, 1 case of suicidal behaviour occurred in a patient diagnosed with an adjustment disorder who took an overdose of lorazepam. This patient recovered and completed the treatment throughout the open-label phase. One death occurred during the open-label phase in the placebo-to-belimumab group. The death was reported to be associated with fatal serious AEs of multiple organ dysfunction syndrome, sepsis secondary to nosocomial pneumonia, and chronic kidney disease.

Critical Appraisal

Internal Validity

The baseline characteristics in the open-label patients were similar and balanced between the placebo-to-belimumab and belimumab-to-belimumab groups. Nonobjective end points like harms and SLEDAI-S2K can be biased because of the open-label nature of the trial. Moreover, patients who had required treatment with prohibited medications due to their disease activity were withdrawn from treatment during the double-blind phase and were not enrolled in the open-label phase, which may have led to selection bias favouring belimumab, as patients who completed the double-blind phase were those who had better tolerance with belimumab or placebo with background immunosuppressants and were more willing to participate in the open-label phase and potentially more likely to respond well to belimumab over time. The same validity, reliability, and responsiveness concerns for the tools used to measure the outcomes described for the randomized phase also apply to the extension period.

Since no hypothesis testing of treatment differences was conducted for the active treatment period, it is not possible to draw causal conclusions based on the descriptive results. The lack of a comparator group also poses risk of confounding. Although between-group differences were observed in uPCR, proportions of patients with SLEDAI-S2K scores less than 8, average prednisone-equivalent doses, and proportions of autoantibody-positive patients and patients with low C3 or C4 levels, favouring the belimumab-to-belimumab group, it is not possible to draw causal conclusions (i.e., changes following the randomized phase cannot be attributed to the treatment with any certainty) without a randomized comparator. The safety results were consistent with those observed during the double-blind randomized phase. Since the OLE phase was a crossover study and not controlled, differences in safety profile between the belimumab-to-belimumab and placebo-to-belimumab groups should be viewed in light of this limitation. Since the treatment failure criterion regarding prohibited medications (e.g., immunosuppressants and corticosteroids) was less restrictive during the open-label phase than during the double-blind phase, these differences made it difficult to evaluate the rates of maintenance of PERR and CRR across the 2 phases of the study. Lastly, the relatively short duration of the open-label phase is not enough to observe appreciable benefit among those who had transitioned from placebo to belimumab.

External Validity

The external validity of the extension phase is equivalent to that reported for the randomized phase. As with the randomized phase, per the clinical expert there are no major concerns for the generalizability of the findings to Canadian clinical practice.

Discussion

Summary of Available Evidence

The systematic review included a double-blind randomized controlled trial that evaluated the efficacy and safety of IV belimumab 10 mg/kg plus standard of care compared to placebo plus standard of care in adult patients with active LN (histological classes III, IV, V, or V in combination with III or IV LN). The primary outcome in the BLISS-LN study was PERR at week 104, a dichotomous end point (response versus no response); “response” was defined by a uPCR less than or equal to 0.7 g/g, an eGFR no more than 20% below pre-flare value or greater than or equal to 60 mL/min/1.73 m², and no use of prohibited therapy for treatment failure.

One OLE study was also summarized in this report and provided supplemental safety and efficacy data for patients who received IV belimumab 10 mg/kg plus standard of care for up to 28 weeks (N = 254) among eligible patients who completed the BLISS-LN study. No indirect evidence comparing belimumab to other treatments for LN was submitted by the sponsor, and none was found in the literature.

Interpretation of Results

Efficacy

The clinical expert who consulted on this review suggested that the outcomes used in clinical practice align with those used in the BLISS-LN study, with the most recent KDIGO and EULAR/ERA-EDTA clinical practice guidelines for LN suggesting that a complete response to therapy should aim for proteinuria of less than 0.5 g per 24 hours to less than 0.7 g per 24 hours within 12 months of initiating therapy (but may require an additional 12 months in those with nephrotic range proteinuria) alongside stabilization or improvement in eGFR within 10% to 15% of baseline (i.e., pre-flare).^8,11 The BLISS-LN study met its primary end point of PERR at week 104 as well as key secondary end points including CRR at week 104, PERR at week 52, ORR at week 104, and time to renal-related event or death. About 11% more patients in the belimumab group than in the placebo group achieved the primary outcome of PERR at week 104 and the key secondary outcome of CRR at week 104. Patients who received belimumab had a significantly lower risk of a renal-related event or death than those who received placebo. The clinical expert consulted deemed these primary and key secondary results to be clinically meaningful in favour of belimumab 10 mg/kg in a disease with limited therapeutic options and substantial morbidity and mortality.

The subgroup analyses based on baseline renal biopsy class and induction regimen for the primary and key secondary outcomes were generally consistent with the overall results for all subgroups except for the baseline renal biopsy class V and cyclophosphamide followed by azathioprine subgroups, the results of which were not statistically significantly different between treatment groups. However, the study was not designed or powered to evaluate efficacy in subgroups, and the small number of patients in the class V and the cyclophosphamide followed by azathioprine subgroups might have led to the lack of statistical significance between treatment groups.

Secondary analyses in the BLISS-LN trial were generally supportive of the results of the primary outcome. However, as some secondary outcomes were not included in the hierarchy, they were not controlled for the type I error rate; therefore, any results should be viewed as overall supportive evidence of the effects of belimumab relative to placebo. The clinical expert consulted on this review noted the importance of reducing the use of corticosteroids in the management of LN due to their substantial long-term toxicity. Results of the BLISS-LN trial found that at week 104 a greater proportion of patients in the belimumab group than in the placebo group had an average daily dose of prednisone less than or equal to 7.5 mg since their previous 4-week visit, with a between-group difference of 11.21%, which the clinical expert found to be clinically meaningful.

Regarding disease activity, the SLEDAI-S2K, which measures general disease activity, was modified in the BLISS-LN study to consider new as well as persistent proteinuria of more than 0.5 g per 24 hours. Secondary analyses found a greater proportion of patients in the belimumab group than in the placebo group had a SLEDAI-S2K score less than 4 by week 104. As discussed in Appendix 4, the SLEDAI-S2K uses a single weighted score to summarize disease activity, and it could have the potential to mask the underlying importance of organ systems that are contributing to the total score (i.e., the same score could represent multiple mild disease in many organs or severe disease in a single organ, or an unchanged score may occur despite worsening in 1 organ system if there is also improvement in another system). The clinical expert noted that it is clinically meaningful that a greater proportion of patients in the belimumab group had a SLEDAI-S2K score less than 4 by week 104, but as only the global SLEDAI-S2K score was presented, it is not known whether the improvements were in renal versus extrarenal systems; it would have been of interest to present the SLEDAI-S2K results stratified by organ system.

The clinician and patient group input received for this review indicated that patients would like to experience fewer flares. In the BLISS-LN study, patients in the belimumab group experienced fewer severe SFI flares and had a lower risk of experiencing a severe SFI flare than those in the placebo group. According to the clinician group and clinical expert input, flares have been associated with worse outcomes, increasing the risk of progression to ESRD and dialysis among patients with LN; hence, these results were found to be clinically meaningful to the clinical expert consulted.

The clinician group input noted that HRQoL is significantly impaired in patients with LN. The patient group input received for this CADTH review indicated that patients would like new therapies that provide increased mobility, increase ability to participate in physical and social activities, and improve overall HRQoL. As the BLISS-LN study did not assess HRQoL, the impact of belimumab in addition to standard of care on HRQoL is unclear. The clinical expert stated that, while HRQoL outcomes would have been of interest, the renal-focused primary and key secondary outcomes included in BLISS-LN trial are the outcomes of greatest importance in an LN trial.

Other limitations include the fact that a greater proportion of patients discontinued from the placebo group than the belimumab group, which may have led to bias in favour of belimumab. However, the sensitivity analyses that assessed the impact of missing data generally showed results that were supportive of the primary analysis. Regarding calculations of patients’ average daily prednisone dose in the BLISS-LN trial, days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation, which may also have led to bias, although the direction of bias is unknown.

The product monograph for belimumab authorized both IV and subcutaneous formulations for LN.¹² However, this recommendation was based on extrapolated data, and there is no clinical evidence regarding the subcutaneous formulation for patients with active LN. While the product monograph states that belimumab is indicated for adult patients in addition to standard of care for the treatment of active LN, the sponsor reimbursement request is for adult patients with class III, IV, and/or V active LN in addition to standard of care, and if no improvements in disease activity and/or symptoms are observed after 6 months, use of belimumab should be discontinued. Patients with class III, IV, and/or V active LN were enrolled and assessed for the end points in the BLISS-LN trial, and according to the clinical expert consulted, patients with class I, II, or VI LN should not be candidates for immunotherapy such as belimumab. The clinical expert and clinician group input noted that treatment with belimumab should be discontinued after 6 months to 12 months of therapy if there was no improvement in renal functioning or proteinuria, suggesting the reimbursement request to discontinue treatment after 6 months may have been too conservative. In the opinion of the clinical expert, in the BLISS-LN trial, the percentage of patients achieving the primary outcome over time were identical through to week 20 after randomization and started to diverge at week 24 through to week 104 (with a greater percentage achieving PERR in the belimumab group). Hence, it would be appropriate to wait at least 6 months to assess response to belimumab, and in patients with higher levels of proteinuria the clinical expert believed it would be appropriate to wait up to 12 months to determine if the response is sufficient to warrant continuation.

There were no head-to-head studies comparing the efficacy and safety of belimumab relative to rituximab, and the clinical expert believed it would be inappropriate to include a comparator arm in BLISS-LN consisting of rituximab added to standard of care. Although rituximab is used as add-on therapy to mycophenolate in refractory cases of LN, neither the EULAR/ERA-EDTA nor KDIGO guidelines recognize it as standard of care. Both guidelines recommend either mycophenolate or cyclophosphamide as standard of care and advise that rituximab is reserved as salvage therapy for refractory cases. Hence, it would have been inappropriate to include a comparator arm consisting of rituximab added to standard of care. The sponsor did not conduct an ITC as the heterogeneity between studies is significant and any ITC is therefore unlikely to produce robust estimates of comparative efficacy or safety. Thus, the comparative efficacy and safety of belimumab relative to rituximab is unknown.

Harms

Based on its mechanism of action, targeting B cells, infection would be 1 of the notable harms that should be monitored with belimumab. The clinical expert consulted by CADTH felt that the safety profile of belimumab was in line with other treatments. There has been no indication from the pivotal trial that there is an increased risk of mortality due to AE while on belimumab. Eleven deaths occurred during the double-blind phase of the BLISS-LN trial (2.7% of patients in the belimumab group versus 2.2% in the placebo group), mainly due to infections, and 1 death (0.4% of patients) occurred in the OLE, which was deemed SLE related. Most patients reported 1 or more AEs during the BLISS-LN trial, with UTI, cough, and upper abdominal pain reported more frequently among patients who received belimumab versus placebo. SAEs were comparable across treatment groups (25.9% of patients in the belimumab group versus 29.9% in the placebo group). The OLE study confirmed these findings (N = 254; duration up to 28 weeks), although the conclusions that can be drawn are limited by the lack of control group. Concerns over infection risk with belimumab also need to be weighed against the risk of infection of current standard of care medications, including immunosuppressants and corticosteroids, which are known for their increased infection risk.

Psychiatric AEs were a notable harm of this review, and there was no indication of an increased risk of these events occurring with belimumab compared to placebo. Patients judged recently to be at high risk of suicide were excluded from BLISS-LN; thus, the safety of belimumab in these patients is unknown.

Conclusions

In adult patients with class III, IV, and/or V active LN, treatment with belimumab 10 mg/kg in addition to standard of care statistically significantly improved renal response as measured by the primary outcome PERR, relative to placebo, based on 104-week data from the BLISS-LN trial, the results of which were deemed to be clinically meaningful by the clinical expert consulted for this review. All key secondary outcomes, including CRR at week 104, PERR at week 52, ORR at week 104, and time to renal-related event or death, showed statistically significant differences in favour of belimumab. The trial showed that a greater proportion of patients in the belimumab group than in the placebo group had reductions in average daily prednisone use to less than 7.5 mg since the previous 4-week visit and showed improvements in mean SLEDAI-S2K score. Also, fewer patients in the belimumab group experienced severe flares than those in the placebo group. HRQoL was not assessed in the trial, and therefore the impact of belimumab on HRQoL in patients with LN is unknown. In addition, the long-term efficacy of belimumab on reducing flare rates is unknown. AEs, including infections, occurred with similar frequency in the 2 treatment groups. Data from the pivotal trial and the OLE study do not suggest issues of tolerability or safety, although the extension study was limited by a lack of a control group.

References

1.Daniel J Wallace UpToDate: Overview of the management and prognosis of systemiclupus erythematosus in adults. In: Pisetsky DS, Schur PH, Curtis MR, eds2021.

2.Asif S, Bargman J, Auguste B. A review of the AURORA and BLISS trials: will it revolutionize the treatment of lupus nephritis? Curr Opin Nephrol Hypertens. 2022;31(3):278-282. PubMed

3.Bernatsky S, Joseph L, Pineau C, Tamblyn R, Feldman D, Clarke A. A population-based assessment of systemic lupus erythematosus incidence and prevalence—results and implications of using administrative data for epidemiological studies. Rheumatology. 2007;46(12):1814-1818. PubMed

4.Carter EE, Barr SG, Clarke AE. The global burden of SLE: prevalence, health disparities and socioeconomic impact. Nature reviews rheumatology. 2016;12(10):605-620. PubMed

5.Peter H Schur BHH. UpToDate: Epidemiology and pathogenesis of systemic lupuserythematosus. In: David S Pisetsky, Curtis MR, eds2021.

6.Parikh SV, Almaani S, Brodsky S, Rovin BH. Update on lupus nephritis: core curriculum 2020. Am J Kidney Dis. 2020;76(2):265-281. PubMed

7.Mahajan A, Amelio J, Gairy K, et al. Systemic lupus erythematosus, lupus nephritis and end-stage renal disease: a pragmatic review mapping disease severity and progression. Lupus. 2020;29(9):1011-1020. PubMed

8.Rovin BH, Adler SG, Barratt J, et al. KDIGO 2021 clinical practice guideline for the management of glomerular diseases. Kidney Int. 2021;100(4):S1-S276. PubMed

9.Diseases NIoDaDaK. Lupus and Kidney Disease (Lupus Nephritis). 2017: https://www.niddk.nih.gov/health-information/kidney-disease/lupus-nephritis. Accessed 2022 Sept 16.

10.Anders H-J, Saxena R, Zhao M-h, Parodis I, Salmon JE, Mohan C. Lupus nephritis. Nature reviews Disease primers. 2020;6(1):1-25. PubMed

11.Fanouriakis A, Kostopoulou M, Cheema K, et al. 2019 Update of the Joint European League Against Rheumatism and European Renal Association-European Dialysis and Transplant Association (EULAR/ERA-EDTA) recommendations for the management of lupus nephritis. Annals of the rheumatic diseases. 2020;79(6):713-723. PubMed

12.Benlysta (belimumab): lyophilized powder for intravenous infusion 120 mg in 5 mL vial, 400 mg in 20 mL vial, solution for subcutaneous injection 200 mg in 1 mL [product monograph]. Mississauga (ON): GlaxoSmithKline Inc.; 2021 Jul 29: https://pdf.hres.ca/dpd_pm/00042499.PDF. Accessed 2022 Sept 09.

13.Clinical Study Report: BEL114054. A phase 3, randomized, double-blind, placebo-controlled study to evaluate the efficacy and safety of belimumab plus standard of care versus placebo plus standard of care in adult subjects with active lupus nephritis [internal sponsor's report]. City (PROV): GlaxoSmithKline; 2020 Mar 27.

14.Yap DY, Yung S, Chan TM. Lupus nephritis: an update on treatments and pathogenesis. Nephrology. 2018;23:80-83. PubMed

15.Anders H-J, Rovin B. A pathophysiology-based approach to the diagnosis and treatment of lupus nephritis. Kidney Int. 2016;90(3):493-501. PubMed

16.Hanly JG, O’Keeffe AG, Su L, et al. The frequency and outcome of lupus nephritis: results from an international inception cohort study. Rheumatology. 2016;55(2):252-262. PubMed

17.Weening JJ, D'Agati VD, Schwartz MM, et al. The classification of glomerulonephritis in systemic lupus erythematosus revisited. J Am Soc Nephrol. 2004;15(2):241-250. PubMed

18.Rovin BH, Furie R, Latinis K, et al. Efficacy and safety of rituximab in patients with active proliferative lupus nephritis: the Lupus Nephritis Assessment with Rituximab study. Arthritis Rheum. 2012;64(4):1215-1226. PubMed

19.Kasitanon N, Boripatkosol P, Louthrenoo W. Response to combination of mycophenolate mofetil, cyclosporin A and corticosteroid treatment in lupus nephritis patients with persistent proteinuria. International journal of rheumatic diseases. 2018;21(1):200-207. PubMed

20.Chen W, Liu Q, Chen W, et al. Outcomes of maintenance therapy with tacrolimus versus azathioprine for active lupus nephritis: a multicenter randomized clinical trial. Lupus. 2012;21(9):944-952. PubMed

21.Rituxan (rituximab): intravenous infusion 10mg/mL [product monograph]. Mississauga (ON): Hoffmann-La Roche Ltd.; 2019 Oct 10: https://pdf.hres.ca/dpd_pm/00053554.PDF. Accessed 2022 Sept 09.

22.Hydroxychloroquine: Tablets, 200 mg, Oral [product monograph]. Brampton (ON): Sanis Health Inc.; 2022 Feb 18: https://pdf.hres.ca/dpd_pm/00064783.PDF. Accessed 2022 Set 09.

23.Prednisone: Tablets 5 mg, and 50 mg [product monograph]. Toronto (ON): Teva Canada Limited; 2020 Dec 7 https://pdf.hres.ca/dpd_pm/00059252.PDF. Accessed 2022 Sept 09.

24.Procytox (Cyclophosphamide): Tablets 25 mg, 50 mg, powder for injection 200 mg, 500 mg, 1000 mg, 2000 mg per vial [product monograph]. Mississauga (ON): Baxter Corporation 2012 Sept 07: https://pdf.hres.ca/dpd_pm/00017604.PDF. Accessed 2022 Sept 09.

25.CellCept (mycophenolate mofetil}: Capsules 250 mg, film-coated tablets 500 mg, powder for oral suspension 200 mg/mL (when reconstituted), for injection (as hydrochloride) 500 mg/vial [product monograph]. Mississauga (ON): Hoffmann-La Roche Ltd.; 2021 Jul 08: https://pdf.hres.ca/dpd_pm/00063434.PDF. Accessed 2022 Sept 09.

26.Rovin BH, Furie R, Teng YKO, et al. A secondary analysis of the Belimumab International Study in Lupus Nephritis trial examined effects of belimumab on kidney outcomes and preservation of kidney function in patients with lupus nephritis. Kidney Int. 2022;101(2):403-413. PubMed

27.McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer Review of Electronic Search Strategies: 2015 guideline statement. J Clin Epidemiol. 2016;75:40-46. PubMed

28.Grey matters: a practical tool for searching health-related grey literature. Ottawa (ON): CADTH; 2019: https://www.cadth.ca/grey-matters. Accessed 1800 Jan 1.

29.Furie R, Rovin BH, Houssiau F, et al. Two-Year, Randomized, Controlled Trial of Belimumab in Lupus Nephritis. N Engl J Med. 2020;383(12):1117-1128. PubMed

30.Mikdashi J, Nived O. Measuring disease activity in adults with systemic lupus erythematosus: the challenges of administrative burden and responsiveness to patient concerns in clinical research. Arthritis Res Ther. 2015;17:183. PubMed

31.Yee C-S, Farewell VT, Isenberg DA, et al. The use of Systemic Lupus Erythematosus Disease Activity Index-2000 to define active disease and minimal clinically meaningful change based on data from a large cohort of systemic lupus erythematosus patients. Rheumatology. 2011;50(5):982-988. PubMed

32.Castrejon I, Tani C, Jolly M, Huang A, Mosca M. Indices to assess patients with systemic lupus erythematosus in clinical trials, long-term observational studies, and clinical care. Clin Exp Rheumatol. 2014;32(5 Suppl 85):S-85-95. PubMed

33.Stoll T, Seifert B, Isenberg DA. SLICC/ACR Damage Index is valid, and renal and pulmonary organ scores are predictors of severe outcome in patients with systemic lupus erythematosus. Br J Rheumatol. 1996;35(3):248-254. PubMed

34.Posner K, Oquendo MA, Gould M, Stanley B, Davies M. Columbia Classification Algorithm of Suicide Assessment (C-CASA): classification of suicidal events in the FDA’s pediatric suicidal risk analysis of antidepressants. Am J Psychiatry. 2007;164(7):1035-1043. PubMed

35.Li EK, Tam L-S, Zhu TY, et al. Is combination rituximab with cyclophosphamide better than rituximab alone in the treatment of lupus nephritis? Rheumatology. 2009;48(8):892-898. PubMed

36.Zhang J, Zhao Z, Hu X. Effect of rituximab on serum levels of anti-C1q and antineutrophil cytoplasmic autoantibodies in refractory severe lupus nephritis. Cell Biochem Biophys. 2015;72(1):197-201. PubMed

37.Li K, Yu Y, Gao Y, Zhao F, Liang Z, Gao J. Comparative Effectiveness of Rituximab and Common Induction Therapies for Lupus Nephritis: A Systematic Review and Network Meta-Analysis. Front Immunol. 2022;13:859380. PubMed

38.Gladman DD, Ibanez D, Urowitz MB. Systemic lupus erythematosus disease activity index 2000. The Journal of rheumatology. 2002;29(2):288-291. PubMed

39.Touma Z, Urowitz MB, Gladman DD. Systemic lupus erythematosus disease activity index 2000 responder index-50 website. The Journal of rheumatology. 2013;40(5):733-733. PubMed

40.Jesus D, Rodrigues M, Matos A, Henriques C, Pereira da Silva JA, Ines LS. Performance of SLEDAI-2K to detect a clinically meaningful change in SLE disease activity: a 36-month prospective cohort study of 334 patients. Lupus. 2019;28(5):607-612. PubMed

41.Yee CS, Isenberg D, Prabu A, et al. BILAG-2004 index captures systemic lupus erythematosus disease activity better than SLEDAI-2000. Ann Rheum Dis. 2008;67(6):873-876. PubMed

42.Ward MM, Marx AS, Barry NN. Comparison of the validity and sensitivity to change of 5 activity indices in systemic lupus erythematosus. The Journal of rheumatology. 2000;27(3):664-670. PubMed

43.Gladman DD, Urowitz MB, Kagal A, Hallett D. Accurately describing changes in disease activity in Systemic Lupus Erythematosus. The Journal of rheumatology. 2000;27(2):377-379. PubMed

44.Gladman D, Ginzler E, Goldsmith C, et al. The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for systemic lupus erythematosus. Arthritis Rheum. 1996;39(3):363-369. PubMed

45.Romero-Diaz J, Isenberg D, Ramsey-Goldman R. Measures of adult systemic lupus erythematosus: updated version of British Isles Lupus Assessment Group (BILAG 2004), European Consensus Lupus Activity Measurements (ECLAM), Systemic Lupus Activity Measure, Revised (SLAM-R), Systemic Lupus Activity Questionnaire for Population Studies (SLAQ), Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K). Arthritis Care Res. 2011;63(Suppl 11):S37-S46. PubMed

46.Jolly M, Annapureddy N, Arnaud L, Devilliers H. Changes in quality of life in relation to disease activity in systemic lupus erythematosus: post-hoc analysis of the BLISS-52 Trial. Lupus. 2019;28(14):1628-1639. PubMed

47.Isenberg D, Sturgess J, Allen E, et al. Study of flare assessment in systemic lupus erythematosus based on paper patients. Arthritis Care Res (Hoboken). 2018;70(1):98-103. PubMed

48.Clinical Study Report: D3461C00004. TULIP 2: A multicentre, randomised, double-blind, placebo-controlled, phase 3 study evaluating the efficacy and safety of anifrolumab in adult subjects with active systemic lupus erythematosus [internal sponsor's report]. Gaithersburg (MD): AstraZeneca PLC; 2019 Dec 16.

49.Buyon JP, Petri MA, Kim MY, et al. The effect of combined estrogen and progesterone hormone replacement therapy on disease activity in systemic lupus erythematosus: a randomized trial. Ann Intern Med. 2005;142(12 Pt 1):953-962. PubMed

50.Petri M, Kim MY, Kalunian KC, et al. Combined oral contraceptives in women with systemic lupus erythematosus. N Engl J Med. 2005;353(24):2550-2558. PubMed

51.Uribe AG, Vilá LM, McGwin G, Sanchez ML, Reveille JD, Alarcón GS. The Systemic Lupus Activity Measure-revised, the Mexican Systemic Lupus Erythematosus Disease Activity Index (SLEDAI), and a modified SLEDAI-2K are adequate instruments to measure disease activity in systemic lupus erythematosus. The Journal of rheumatology. 2004;31(10):1934-1940. PubMed

52.Shariati-Sarabi Z, Monzavi SM, Ranjbar A, Esmaily H, Etemadrezaie H. High disease activity is associated with high disease damage in an Iranian inception cohort of patients with lupus nephritis. Clin Exp Rheumatol. 2013;31(1):69-75. PubMed

53.Castrejon I, Rua-Figueroa I, Rosario MP, Carmona L. Clinical composite measures of disease activity and damage used to evaluate patients with systemic lupus erythematosus: a systematic literature review. Reumatol Clin. 2014;10(5):309-320. PubMed

54.Feld J, Isenberg D. Why and how should we measure disease activity and damage in lupus? Presse Med. 2014;43(6P2):e151-e156.

55.Furie RA, Petri MA, Wallace DJ, et al. Novel evidence-based systemic lupus erythematosus responder index. Arthritis Rheum. 2009;61(9):1143-1151. PubMed

56.Petri M, Buyon J, Kim M. Classification and definition of major flares in SLE clinical trials. Lupus. 1999;8(8):685-691. PubMed

57.Isenberg DA, Allen E, Farewell V, et al. An assessment of disease flare in patients with systemic lupus erythematosus: a comparison of BILAG 2004 and the flare version of SELENA. Ann Rheum Dis. 2011;70(1):54-59. PubMed

58.Coresh J, Turin TC, Matsushita K, et al. Decline in estimated glomerular filtration rate and subsequent risk of end-stage renal disease and mortality. JAMA. 2014;311(24):2518-2531. PubMed

59.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. PubMed

60.Matsushita K, Mahmoodi BK, Woodward M, et al. Comparison of risk prediction using the CKD-EPI equation and the MDRD study equation for estimated glomerular filtration rate. JAMA. 2012;307(18):1941-1951. PubMed

Appendix 1: Literature Search Strategy

Note that this appendix has not been copy-edited.

Clinical Literature Search

Overview

Interface: Ovid

Databases:

• MEDLINE All (1946 to present)

• Embase (1974 to present)

• Note: Subject headings and search fields have been customized for each database. Duplicates between databases were removed in Ovid.

Date of search: August 4, 2022

Alerts: Biweekly search updates until project completion

Search filters applied: No filters were applied to limit the retrieval by study type.

Limits:

• Conference abstracts: excluded

Table 18: Syntax Guide

Multi-Database Strategy

1. (benlysta* or belimumab* or benlista* or benlystia* or LimphoStat-B or LimphoStatB or LymphoStat-B or LymphoStatB or 73B0K5S26A or L04AA26 or HGS-1006 or HGS1006 or BEL-114333 or BEL114333 or GSK-1550188 or GSK1550188).ti,ab,kf,ot,hw,rn,nm.

2. Lupus Nephritis/

3. ((lupus or scleros* or lupoid* or SLE) adj5 (neph* or kidney or mesangial* or renal)).ti,ab,kf.

4. (glomeruloneph* or glomerulosclerosis).ti,ab,kf.

5. or/2-4

6. 1 and 5

7. 6 use medall

8. *belimumab/

9. (benlysta* or belimumab* or benlista* or benlystia* or LimphoStat-B or LimphoStatB or LymphoStat-B or LymphoStatB or L04AA26 or HGS-1006 or HGS1006 or BEL-114333 or BEL114333 or GSK-1550188 or GSK1550188).ti,ab,kf,dq.

10. 8 or 9

11. lupus erythematosus nephritis/

12. ((lupus or scleros* or lupoid* or SLE) adj5 (neph* or kidney or mesangial* or renal)).ti,ab,kf,dq.

13. (glomeruloneph* or glomerulosclerosis).ti,ab,kf,dq.

14. or/11-13

15. 10 and 14

16. 15 use oemezd

17. (conference abstract or conference review).pt.

18. 16 not 17

19. 7 or 18

20. remove duplicates from 19

Clinical Trials Registries

ClinicalTrials.gov

Produced by the US National Library of Medicine. Targeted search used to capture registered clinical trials.

[Search -- Studies with results | Benlysta or belimumab AND lupus nephritis]

WHO ICTRP

International Clinical Trials Registry Platform, produced by the World Health Organization. Targeted search used to capture registered clinical trials.

[Search terms -- Benlysta or belimumab AND lupus nephritis]

Health Canada’s Clinical Trials Database

Produced by Health Canada. Targeted search used to capture registered clinical trials.

[Search terms -- Benlysta or belimumab AND lupus nephritis]

EU Clinical Trials Register

European Union Clinical Trials Register, produced by the European Union. Targeted search used to capture registered clinical trials.

[Search terms -- Benlysta or belimumab AND lupus nephritis]

Grey Literature

Search dates: July 24 to 27, 2022

Keywords: Benlysta, belimumab, lupus nephritis

Limits: none

Relevant websites from the following sections of the CADTH grey literature checklist Grey Matters: A Practical Tool for Searching Health-Related Grey Literature were searched:

• Health Technology Assessment Agencies

• Health Economics

• Clinical Practice Guidelines

• Drug and Device Regulatory Approvals

• Advisories and Warnings

• Drug Class Reviews

• Clinical Trials Registries

• Databases (free)

• Health Statistics

• Internet Search

• Open Access Journals

Appendix 2: Excluded Studies

Note this appendix has not been copy-edited.

Table 19: Excluded Studies

Appendix 3: Detailed Outcome Data

Note this appendix has not been copy-edited.

Figure 3: Proportion of Patients With PERR by Visit From Week 8 to Week 104 in BLISS-LN (mITT)

mITT = modified intention to treat.

Source: Clinical Study Report for BLISS-LN.¹³

Table 20: PERR Response by Subgroup — Baseline Renal Biopsy Class and Induction Regimen (mITT)

CI = confidence interval; mITT = modified intention to treat; OR = odds ratio.

^aOdds ratio (OR), 95% CI, and p value are from a logistic regression model run within the subgroup level for the comparison between belimumab and placebo with covariates treatment group, induction regimen (cyclophosphamide versus mycophenolate mofetil), race (Black versus Non-Black), baseline uPCR, and baseline eGFR.

^bP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

Source: Clinical Study Report for BLISS-LN.¹³

Table 21: CRR Response by Subgroup — Baseline Renal Biopsy Class and Induction Regimen (mITT)

CI = confidence interval; mITT = modified intention to treat; OR = odds ratio.

^aOdds ratio (OR), 95% CI, and p value are from a logistic regression model run within the subgroup level for the comparison between belimumab and placebo with covariates treatment group, induction regimen (cyclophosphamide versus mycophenolate mofetil), race (Black versus Non-Black), baseline uPCR, and baseline eGFR.

^bP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

Source: Clinical Study Report for BLISS-LN.¹³

Figure 4: Time to PERR That Is Maintained Through Week 104 (mITT)

mITT = modified intention to treat.

Note: The at-risk numbers are the number of patients who have the potential to experience the event at that time point. Patients without PERR at week 104 are censored at the last available visit up through week 104. Patients with investigational product discontinuation, treatment failure, study withdrawal, lost to follow-up, or death are censored. Time to event is defined as (event date – treatment start date + 1).

Source: Clinical Study Report for BLISS-LN.¹³

Figure 5: Proportion of Patients With CRR by Visit From Week 8 to Week 104 in BLISS-LN (mITT)

mITT = modified intention to treat.

Source: Clinical Study Report for BLISS-LN.¹³

Figure 6: Time to CRR That Is Maintained Through Week 104 (mITT)

CRR = complete renal response; mITT = modified intention to treat.

Note: The at-risk numbers are the number of patients who have the potential to experience the event at that time point. Figure is truncated at day 742. Patients without CRR at week 104 are censored at the last available visit up through week 104. Patients with investigational product discontinuation, treatment failure, study withdrawal, lost to follow-up, or death are censored. Time to event is defined as (event date - treatment start date + 1).

Source: Clinical Study Report for BLISS-LN.¹³

Figure 7: Time to Renal-Related Event or Death Through Week 104 (mITT)

mITT = modified intention to treat.

Note: The at-risk numbers are the number of subjects who have the potential to experience the event at that time point. Figure is truncated at day 742.

Source: Clinical Study Report for BLISS-LN.¹³

Figure 8: Proportion of Patients in Response Group With an Average Daily Dose of Prednisone of 7.5 mg or Less Since Prior 4-Week Visit (mITT)

mITT= modified

Note: All prednisone doses since the previous 4-week visit were summed and divided by the number of days in the period. Days where a patient did not have a prednisone dose recorded were considered as 0 mg for the day in the calculation for average prednisone dose.

Source: Clinical Study Report for BLISS-LN.¹³

Figure 9: Mean Observed GFR^a Values by Visit in BLISS-LN (Safety Population)

BSA = body surface area; GFR = glomerular filtration rate.

^a From creatinine adjusted for BSA (mL/min/1.73 m^2.).

Source: Clinical Study Report for BLISS-LN.¹³

Appendix 4: Description and Appraisal of Outcome Measures

Note this appendix has not been copy-edited.

Aim

To describe the following outcome measures and review their measurement properties (validity, reliability, responsiveness to change, and MID):

• SLEDAI-2K

• SDI

• SFI

• C-SSRS

Findings

Table 22: Summary of Outcome Measures and Their Measurement Properties

BILAG=British Isles Lupus Assessment Group; C-SSRS = Columbia – Suicide Severity Rating Scale; ESR = erythrocyte sedimentation rate; FACIT-F = Functional Assessment of Chronic Illness Therapy – Fatigue; LN = lupus nephritis; MID = minimal important difference; NA = not applicable; PGA = physician global assessment; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; SELENA SLEDAI = Safety of Estrogens in Lupus National Assessment-Systemic Lupus Erythematosus Disease Activity Index; SF-36 = 36-Item Short Form Survey; SFI = SELENA SLEDAI Flare Index; SLE= systemic lupus erythematosus; SLEDAI-2K = Systemic Lupus Erythematosus Disease Activity Index 2000; v2= version 2.

SLEDAI-S2K

Description and Scoring

The SELENA SLEDAI is a measure of disease activity that was derived by consensus among experts in rheumatology, followed by regression models to assign relative weights to each parameter.³⁰ The SELENA SLEDAI assessment is based on the presence of 24 individual items in 9 organ systems which are given a weighted score and summed if present at the time of the visit or in the preceding 10 days and incudes the use of signs and symptoms, laboratory tests, and physician’s assessment for this purpose. Each descriptor has a weighted score and the sum of all 24 descriptor scores falls between 0 and 105, with higher scores representing higher disease activity, and 0 representing inactive disease. However, not many patients could achieve a score of > 45.^49,50 In the BLISS-LN trial, the SLEDAI-S2K is a modified version of the original SLEDAI. The proteinuria descriptor was modified with the use of the SLEDAI-2000 (SLEDAI-2K) version that captures new as well as persistent proteinuria of more than 0.5 g per 24 hours.

Validity

In a study of 334 patients with SLE in Portugal, a strong Spearman rank correlation (0.824) was observed between the SLEDAI-2K and the PGA at the 36-month follow-up, supporting the construct validity of the SLEDAI-2K in patients with SLE.^.40 In another study of 92 patients with SLE, a good correlation coefficient of 0.677 between the SLEDAI-2K and PGA was identified, indicating construct validity.⁵¹

Reliability

The reliability of the SLEDAI-2K was demonstrated using interrater reliability between 2 raters in a study of 93 patients with SLE.⁴¹ Results found agreement between the raters for each of the items ranging between 81.7% and 100%.⁴¹

Responsiveness

In terms of responsiveness, in one study, the SLEDAI-2K was unable to detect a clinically meaningful improvement or worsening in SLE disease activity; as it failed to identify more than 60% of cases with a worsening or improvement, which was defined as a change of 0.3 points in the patient global assessment PGA.⁴² The BILAG-2004 has been found to be more responsive to change in disease activity than the SLEDAI-2K.⁴¹ Using a summary score to describe disease activity as in the SLEDAI-2K can mask the underlying organ systems that are contributing to the score (i.e., the same score could indicate mild disease in multiple organs or severe disease in 1 organ; or an unchanged score may occur despite worsening in 1 organ system if there is also improvement in another system).³²

The validity, reliability, and responsiveness were not assessed for patients with LN.

MID

One study identified a minimal clinically meaningful increase of 3 or 4 points for prediction of increase in therapy (worsening) and suggest a minimal clinically meaningful decrease in score of 1 to 2 points for improvement.³¹ Another study found that the SLEDAI-2K score increased by > 3 points when the clinician assessed that the patient was experiencing a flare.⁴³

MID not assessed for patients with LN.

Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI)

Description and Scoring

The SDI was developed by the international collaboration, Systemic Lupus International Collaborating Clinics.³² The purpose of the assessment is to score irreversible damage, regardless of cause. Damage is defined as irreversible change in an organ system that has occurred since the onset of SLE, and is present for at least 6 months.³² The tool is completed by a physician and consists of 42 items in 12 domains (peripheral vascular, ocular, neuropsychiatric, renal, pulmonary, cardiovascular, gastrointestinal, musculoskeletal, skin, endocrine (diabetes), gonadal, and malignancies) with a maximum score of 47 points (higher scores denote more damage).^32,45 The items are rated as present or absent and, in the case of recurring events, such as a stroke, there is a possibility of providing a rating of 2 or 3 points to an item.³² At diagnosis of SLE, the SDI score is 0 by definition.⁴⁵ Damage is considered if the SDI score is ≥ 1 and damage can remain stable or increase over time, however points should not decrease.⁴⁵

Validity

To assess the validity of the SDI, centres who treated patients with SLE submitted 2 assessments, 5 years apart, on 2 patients with active disease (one patient with increase in damage over the 5 years and one patient with stable damage) and 2 patients with inactive disease (one patient with increase in damage and one patient with stable damage).⁴⁴ The cases (14 cases in 3 separate packages) were written up in a uniform format and sent back out, in mixed order, to the centres where the SDI was completed by 20 physicians (2 assessments per patient at time 1 and time 2). The SDI scores of patients with damage after 5 years were increased by a greater degree compared with patients with stable disease (2.08 points versus 0.24 points).⁴⁴ The SDI scores of patients with active disease also increased more compared with patients with inactive disease (1.48 points versus 0.83 points).⁴⁴ A study of 71 patients found that the SDI was associated with SLEDAI-S2K (r=0.742) and the European Consensus Lupus Activity Measurement (ECLAM) (r=0.699).⁵² The SDI and BILAG have been found to have weak correlation (Spearman correlation coefficient 0.19).⁵³

Reliability

Among 20 SLICC members who completed the SDI on 42 SLE cases, there was moderate agreement between raters (ICC = 0.553).⁴⁵ Similarly, when the SDI was completed by another physician based on retrospective review of patient cases, interobserver reliability was moderate (kappa 0.47, 95% CI 0.28 to 0.66).⁵³

Responsiveness

The SDI is a statistically significant predictor of clinically important outcomes. In a 10-year retrospective study of 80 patients with SLE, the mean SDI renal damage score at one year after diagnosis was a significant predictor of end-stage renal failure (at 1 year: renal failure versus no renal failure, SDI renal damage score 0.33 versus 0.03; at 5 years: SDI renal damage score 1.33 versus 0.14; at 10 years: SDI renal damage score 2.80 versus 0.35).³³ The total SDI score was also associated with end-stage renal failure at 5 and 10 years.³³ The SDI pulmonary damage score at 1 year after diagnosis was a significant predictor of death within 10 years, however total SDI score was not associated with death.³³ More recent studies with larger cohorts of patients have shown that the SDI is a predictor of mortality. Patients with SLE (N=1,297) were identified within 2 years of a first clinical visit from 8 centres, and followed for 2, 5-10, and >10 years.⁴⁵ The SDI increased over time and was found to be higher among patients who died.⁴⁵ In the University of Toronto Lupus Clinic, 263 patients were followed for 10 years.⁵⁴ Within 10 years, 25% of patients who exhibited damage at the first SDI assessment (i.e., one year after diagnosis) died, compared with 7.3% of patients who had no early signs of damage.⁵⁴

The validity, reliability, and responsiveness were not assessed for patients with LN.

MID

No formal MID has been assessed for patients with SLE or LN. An SDI ≥ 1 indicates damage which can remain stable or increase over time.³²

SELENA SLEDAI Flare Index (SFI)

Description and Scoring

The SFI is used to identify and classify flares as mild/moderate or severe, based on clinical activity, need for additional treatment, or PGA score.⁵⁵ The original definitions of mild/moderate and severe flares were reached by consensus of the investigators of the SELENA trials.⁵⁶ In the BLISS-LN trial,^13,28 a modified version of the SFI was used, using the SLEDAI-2K instead of the SELENA SLEDAI. In this trial, mild/moderate flare and severe flare were defined according to the following criteria:

• Mild or moderate flare:

  • change in SLEDAI-2K score of ≥ 3 points but not > 12 points, or

  • new or worse discoid, photosensitive, profundus, cutaneous vasculitis, or bullous lupus, or

  • nasopharyngeal ulcers, pleuritis, pericarditis, arthritis, or SLE fever, or

  • increase in prednisone, but not > 0.5 mg/kg/day

  • added NSAID or hydroxychloroquine for SLE

  • ≥ 1.0 increase in PGA score (but not > 2.5).

• Severe flare:

  • change in SLEDAI-2K score > 12 points compared to previous visit, or

  • new or worse CNS-SLE, vasculitis, nephritis, myositis, PIt < 60,000, hemolytic anemia (Hb < 70 g/L or decrease in Hb > 30 g/L); requiring double prednisone, or prednisone increase to > 0.5 mg/kg/day, or hospitalization

  • increase in prednisone increase to > 0.5 mg/kg/day

  • new cyclophosphamide, azathioprine, methotrexate, or mycophenolate for SLE

  • hospitalization for SLE, or

  • increase in PGA score to > 2.5.^13,28

Validity

In a post hoc analysis of BLISS-52 trial data with 867 patients with SLE, the occurrence of a new SFI flare using the SELENA SLEDAI was associated with a significant change in the FACIT-F and all domains of the SF-36v2 except role emotional scores, indicating convergent validity.⁴⁶ In a small study of 16 patients who were each evaluated by 4 physicians, there was 52% agreement between the SFI and BILAG-2004 flare index in classifying patients as having no flare, or mild, moderate or severe flare.⁵⁷ It was unclear, however, if this study used the SFI, or the modified SFI. The agreement among raters on the SFI was fair (ICC 0.21, 95% CI, 0.08 to 0.48), and lower than the BILAG 2004 assessment of flares.⁵⁷

Reliability

A study evaluated the modified SFI using paper-based cases of patients with SLE.⁴⁷ Initially, 988 cases were assessed by 3 physicians for degree of flare or presence of disease activity and rated as severe, moderate, or mild flare, or persistent/ongoing disease. For those cases where there was agreement by the 3 physicians (N=451 cases), they were moved on the second part of the study and assessed by 18 pairs of physicians with 3 instruments, BILAG-2004 flare index, SFI, and modified SFI. The assessments based on these instruments were compared with the assessments conducted initially in the first stage of the study by the 3 physicians. For the modified SFI, assessments matched the conclusions of the thee physicians in 70% of cases (weighted kappa 0.74).⁴⁷ The discrepancies were concentrated in classifying moderate flares as severe flares, and identifying persistent activity as a flare.⁴⁷ There was also an issue of over-scoring due to classifying treatment change as a flare, even when there were no new or worsening clinical features.⁴⁷ The authors of this study indicate that “the problem of capturing lupus flare accurately” is not completely solved.⁴⁷

The validity and reliability were not assessed for patients with LN. No literature was identified regarding the responsiveness of the instrument in patients with SLE or LN.

C-SSRS

Assesses the lethality of attempts and other features of ideation (frequency, duration, controllability, reasons for ideation, and deterrents), all of which are significantly predictive of completed suicide.³⁴ In the pivotal trials,^28,48 2 different versions of the questionnaire were used: one assessing the last 12 months prior to the assessment and another assessing the time since last visit. Suicidal ideation was defined as a “yes” answer at any time in the respective study period to any one of the 5 (re-ordered) suicidal ideation questions ranging from category 1 “wish to be dead” to category 5 “active suicidal ideation with specific plan and intent” on the C-SSRS. Suicidal behaviour was defined as a “yes” answer at any time in the respective study period, to any one of the 5 (re-ordered) suicidal behaviour questions ranging from category 6: preparatory acts or behaviour” to category 10: “completed suicide” on the C-SSRS. Evidence related to the validity, reliability, responsiveness or MID of the instrument among patients with SLE or LN was not identified.

Appendix 5: Summary of Other Studies

Note this appendix has not been copy-edited.

Aim

The aim of this section was to summarize and appraise evidence from Coresh et al., 2014 that was used to inform the pharmacoeconomic model.

Findings

In a meta-analysis conducted by Coresh et al. (2014),⁵⁸ the associations of decline in eGFR with subsequent progression to ESRD as well as all-cause mortality risk have been characterized to evaluate lesser declines in eGFR as potential alternative end points for CKD progression. The prognostic contribution of change in eGFR over 1, 2 and 3 years to subsequent ESRD and mortality had been assessed in cohorts taken from the Chronic Kidney Disease Prognosis Consortium (CKD-PC), which consisted of up to 1.7 million participants with 12,344 ESRD events and 223,944 deaths. Among the included cohorts (from separate datasets), 22 cohorts presented with a repeated measure of serum creatinine over 1 to 3 years to assess the relationship of change in eGFR on subsequent ESRD, whereas 35 cohorts included mortality outcome data. Patient data transfer and a random-effects meta-analysis were conducted between July 2012 and September 2013, with baseline assessments done between 1975-2012.⁵⁸

In this study, the CKD-EPI 2009 creatinine equation was used to calculate eGFR.^59,60 Since the doubling of serum creatinine corresponds to a change in eGFR of 57% or greater reduction with the CKD-EPI equation (for serum creatinine ≥ 0.9 mg/dL in males and ≥ 0.7 mg/dL in females), the primary data presentation used in this study was based on percent change in eGFR. ESRD was defined as an initiation of renal replacement therapy or death due to kidney disease other than acute kidney injury in this study. While the primary outcome of interest for this study was to measure ESRD cases after the baseline period, an analysis had been conducted for all-cause mortality as well as cardiovascular mortality and non-cardiovascular mortality, given that majority of CKD patients die before reaching ESRD. The covariates assessed in this study included – age, sex, race/ethnicity (Black versus non-Black), systolic blood pressure, total cholesterol, history of diabetes or CVD, and first eGFR.⁵⁸

For statistical analyses, a 2-stage analytic approach had been applied, where in the first step, each study had been analyzed individually, followed by a random-effects meta-analysis in the second step. The adjusted HRs of ESRD and mortality after the end of the baseline period had been modelled as a spline function of percent change in eGFR including the covariates. Using the weighted average baseline risk, the meta-analyzed adjusted HRs for percent change in eGFR had been translated to absolute risk of ESRD and mortality at 1, 3, 5, and 10 years after the baseline period. One year baseline risk had been calculated for the following covariates: no change in eGFR, a first eGFR of 50 mL/min/1.73m², age 60 years, male, non-Black, a systolic blood pressure of 130 mm Hg, a total cholesterol of 5 mmol/L, no history of diabetes or CVD. Using a weighted average, risk had been scaled for longer follow-up and pooled across cohorts in the analyses.⁵⁸

From the analyses, the adjusted HRs for ESRD were found to be 32.1 (95% CI: 22.3 to 46.3) and 5.4 (95% CI: 4.5 to 6.4) for −57% and −30% eGFR changes, respectively, for patients with a baseline eGFR <60 mL/min/1.73m², while corresponding adjusted HRs for mortality were 3.7 (95% CI: 3.2 to 4.4) and 1.8 (95% CI: 1.6 to 1.9) for –57% and −30% eGFR changes, respectively, showing exponentially higher adjusted HRs for ESRD and mortality with larger eGFR declines. Results compiled from the whole consortium indicated that changes of −30% or greater eGFR (6.9%, 95% CI: 6.4 to 7.4%) were more common than changes of −57% eGFR (0.79%, 95% CI: 0.52 to 1.06%), This association had been found to be strong and consistent across length of baseline (1 or 3 years), baseline eGFR, age, diabetes status, or albuminuria.⁵⁸

The average adjusted 10-year risks of ESRD for eGFR changes of–57%, –40%, –30% and 0% were 99% (95% CI: 95-100%), 83% (95% CI: 71 to 93%), 64% (95% CI: 52 to 77%), versus 18% (95% CI: 1 to –22%) respectively, for a baseline eGFR of 35 mL/min/1.73m², adjusted for aforementioned covariates and competing mortality risk. The corresponding absolute all-cause mortality risks were 77% (95% CI: 71 to 82%), 60% (95% CI: 56 to 63%), 50% (95% CI: 47 to 52%), versus 32% (95% CI: 31 to 33%), showing a similar but weaker pattern. Results from this study indicated that eGFR decline starting at severely reduced eGFR was associated with very high rates of ESRD during the subsequent 1-5 years, whereas eGFR decline starting at moderately reduced or normal eGFR was associated with a lower risk with ESRD occurring after 10 or more years. Moreover, a consistently higher absolute all-cause mortality risk had been observed with larger eGFR declines for all levels of baseline eGFR and across different subsequent follow-up time. Results from this study demonstrated that the average absolute risk of ESRD was very strongly related to the first eGFR, the length of follow-up and the change in eGFR.⁵⁸

This study provided some evidence in support of using lesser declines in eGFR as an alternative end point for CKD progression, rather than using the established CKD progression end points like ESRD or doubling of serum creatinine. Since these events occur at much later stage in CKD, and a 30% decline in eGFR was found to be approximately 10 times more common than a doubling of serum creatinine as well as associated with an approximately 5-fold increased risk of ESRD after adjusting for covariates including the first eGFR, consideration of lesser declines in eGFR as an alternative end point addresses the limiting feasibility issue of clinical nephrology trials. The study had excluded ESRD cases before baseline period from the relevant analyses, which may potentially lead to bias. Moreover, the authors had acknowledged that standardization of serum creatinine values might not be consistent across time and studies, as well as heterogeneity issues related to variation in design across cohorts.⁵⁸

Pharmacoeconomic Review

Executive Summary

The executive summary comprises 2 tables (Table 1 and Table 2) and a conclusion.

Table 1: Submitted for Review

NOC = Notice of Compliance.

Table 2: Summary of Economic Evaluation

AZA = azathioprine; CKD = chronic kidney disease; CYC = cyclophosphamide; eGFR = estimated glomerular filtration rate; ICER = incremental cost-effectiveness ratio; LN = lupus nephritis; LY = life-year; MMF = mycophenolate mofetil; QALY = quality-adjusted life-year; ST = standard therapy.

Conclusions

Evidence from the BLISS-LN trial suggested that compared to placebo, belimumab plus standard therapy (ST) significantly improved renal response as measured by the primary outcome — primary efficacy renal response — for adult patients with active class III, IV, and/or V lupus nephritis (LN). Time to renal-related events or death also showed statistically significant differences in favour of belimumab, with a greater proportion of patients in the belimumab group experiencing reductions in prednisone use and fewer severe flares than in the placebo group.

As a result of the inappropriate model structure and limitations or uncertainty in the clinical data, CADTH was unable to derive a base-case analysis. Instead, CADTH undertook exploratory reanalyses to address limitations related to flare data, where an alternative extrapolation was selected to inform the time to first renal flare event (and subsequently the annual probability of a renal flare) to better align with clinical expert feedback sought by CADTH.

Based on the CADTH exploratory reanalysis, belimumab plus mycophenolate mofetil (MMF) was associated with an incremental cost-effectiveness ratio (ICER) of $352,880 per quality-adjusted life-year (QALY) gained (incremental costs = $201,083; incremental QALYs = 0.57) versus MMF alone. Belimumab plus cyclophosphamide (CYC) followed by azathioprine (AZA) was dominated (more costly, less effective) by belimumab plus MMF. The CADTH exploratory reanalysis results were similar to those submitted by the sponsor, in that belimumab was not cost-effective at commonly used willingness-to-pay thresholds. At a willingness-to-pay threshold of $50,000 per QALY gained, belimumab plus CYC followed by AZA would require a price reduction of 73% and belimumab plus MMF would require a price reduction of 58% to be considered cost-effective versus CYC followed by AZA alone. However, given the uncertainty around the economic model, further price reductions may be necessary to ensure the cost-effectiveness of belimumab plus ST.

There remains a significant degree of uncertainty in the cost-effectiveness of belimumab plus ST (CYC followed by AZA or MMF monotherapy) compared to ST alone for adult patients with active LN. CADTH was unable to adjust for major limitations, including inability to model the reimbursement population, uncertainties in the modelled disease progression, model structure that failed to adequately reflect the management of active LN in clinical practice, and utility values for patients with LN. As such, the cost-effectiveness of belimumab plus ST should be interpreted with caution.

Stakeholder Input Relevant to the Economic Review

This section is a summary of the feedback received from the patient groups, registered clinicians, and drug plans that participated in the CADTH review process.

CADTH received patient input from Arthritis Consumer Experts, Lupus Ontario, the Kidney Foundation of Canada and Lupus Canada, as well as a jointly by Canadian Arthritis Patient Alliance, Arthritis Society, Canadian Skin Patient Alliance, and CreakyJoints. The joint group gathered patient input by recruiting 1 patient for a video interview and 3 patients for a focus group. Responses from a previously conducted 2019 survey for the use of belimumab for systemic lupus erythematosus (SLE) were also used to inform greater context around living with SLE (6 participants). Patients noted that SLE impacts all aspects of their lives, including difficulties in contributing and participating at school or work due to fatigue, pain, and other symptoms. Patients mentioned that there are many challenges in finding the right combination of drugs to help manage their SLE and LN as responses to medications can vary significantly. Many patients expressed a desire to reduce their use of steroids due to concerns of increased bone density loss. All 3 patients in the focus group and all 6 previous survey responders had experience with belimumab for treating their SLE symptoms. Broadly, 1 patient had negative side effects, another did not have any changes in symptoms or side effects, and all other remaining patients reported an overall decrease in their disease symptoms and increased ability to participate in activities of daily living.

Registered clinical input was received from the Canadian Network for Improved Outcomes in Systemic Lupus Erythematosus and associated physicians (rheumatologists and nephrologists). Clinical input noted that the current treatment strategies for LN are aimed at suppressing or modulating the autoimmune response and preserving renal function, without increasing the risk of adverse events, while minimizing the use of more harmful drugs. Clinician input recognized that belimumab is the first approved drug to address the disease mechanism of modulating the maturation and functional differentiation of the B cells and, therefore, is expected to cause a shift in the current treatment paradigm for LN. Specifically, clinical input believed that belimumab would be most beneficial to patients who do not achieve at least partial remission in a reasonable time period after commencing treatment with currently available therapies, patients who experience early flares and cannot reduce their daily prednisone dose, patients who have frequent flares, and patients for whom adherence is a major factor for treatment failure.

Feedback from drug plans noted uncertainty around the appropriateness of placebo as the comparator in the BLISS-LN trial, given available off-label therapies (e.g., calcineurin inhibitors or rituximab) currently used in practice. Additionally, the plans highlighted the waxing and waning nature of the condition and how it could negatively impact reporting response over time. Lastly, the drug plans highlighted the different modes of administration of belimumab (IV or subcutaneous [SC]) and how efficacy could be impacted.

Several of these concerns were addressed in the sponsor’s model:

• The impact of LN on patients’ quality of life was captured via utility values.

• Adverse events associated with belimumab and renal flares were included within the analysis.

CADTH was unable to address the following concerns raised from stakeholder input:

• cost-effectiveness of belimumab plus ST against available off-label therapies such as rituximab

• cost-effectiveness of SC belimumab.

Economic Review

The current review is for belimumab (Benlysta) for patients with active LN.

Economic Evaluation

Summary of Sponsor’s Economic Evaluation

Overview

The sponsor submitted a cost-utility analysis to assess the cost-effectiveness of belimumab plus ST (defined as MMF monotherapy or CYC followed by AZA for induction and/or maintenance) against ST alone as a treatment for adult patients with active LN. This model population aligned with the BLISS-LN trial population and the Health Canada indication.¹ The reimbursement request had the following additional criteria: adult patients aged 18 years and older who are receiving ST and who have class III, class IV, and/or class V active LN, and if no improvement in disease activity and/or symptoms are observed after 6 months, use should be discontinued.²

Belimumab is a recombinant human immunoglobulin G1 lambda (IgG1λ) monoclonal antibody that is to be used alongside ST. Belimumab is available for both IV and SC administration, where IV belimumab is available in 120 mg and 400 mg strengths and SC belimumab is available in a 200 mg per unit strength, distributed in packs of 4.¹ The recommended IV dose for patients with LN is 10 mg/kg at weeks 0, 2, and 4, and every 4 weeks thereafter, whereas the SC regimen includes a 400 mg dose once weekly for the first 4 doses, then 200 mg once weekly thereafter.¹ The submitted price for belimumab is $305.7100 for the 120 mg per 5 mL vial lyophilized powder for IV infusion, $1,019.0100 for the 400 mg in 20 mL vial lyophilized powder for IV infusion, and $1,581.5900 for the 200 mg in 1 mL for SC injection (only available in packages of 4).³ The comparators for this analysis were ST regimens of MMF monotherapy or CYC followed by AZA for induction and/or maintenance.²

Outcomes of the model included QALYs and life-years over a lifetime horizon of 70 years. Discounting (1.5% per annum) was applied for both costs and outcomes, and a cycle length of 1 year was used.

Model Structure

The sponsor submitted a Markov model consisting of 7 health states: estimated glomerular filtration rate (eGFR) greater than 60 mL/min, eGFR 30 mL/min to 59 mL/min, eGFR 15 mL/min to 29 mL/min, dialysis dependent, renal transplant, posttransplant dialysis dependent, and death (Figure 1).² Patients entered the model based on their baseline eGFR and the percent decline in eGFR at the end of the 2-year BLISS-LN trial period (i.e., stable eGFR or those with a 1% to 20%, 21% to 25%, 26% to 30%, 31% to 40%, and 41% to 57% decline). The proportion of patients experiencing varying degrees of eGFR decline during the BLISS-LN study was also used to inform the rate of progression to end-stage kidney disease (hereby referred to as end-stage renal disease) or death.² Once a patient entered an ESRD health state (i.e., dialysis dependent, renal transplant, posttransplant dialysis dependent), they could no longer transition backward.² Patients could transition to the death health state from any health state in the model.²

It was assumed that patients would transition through the model at a rate relative to their baseline eGFR and percentage decline from baseline eGFR during the 2-year trial period versus having a treatment-specific efficacy.² Patients discontinued treatment at a rate relative to renal worsening, defined by BLISS-LN, for the first 2 years and at a rate reflective of renal flares thereafter.²

Model Inputs

The target population was informed by the BLISS-LN trial, which enrolled patients with LN. The mean age was 33.4 years, and the proportion of females was 88.1%.⁴

Model transition probabilities were sourced from the BLISS-LN trial and published literature. Transition probabilities between eGFR greater than 60 mL/min, 30 mL/min to 59 mL/min, 15 mL/min to 29 mL/min, and less than 15 mL/min health states for the first 2 years (i.e., cycles) of each treatment arm were informed by BLISS-LN, whereas the remaining transition probabilities between the eGFR health states were informed by an international study examining 700 patients with LN over a mean follow-up of 5.2 years.⁵ The latter transition probabilities were adjusted so that the proportion experiencing ESRD and death at years 3, 5, and 10 aligned with the estimates reported by Coresh et al. (2014) per baseline eGFR and percent decline category.^2,6 Transition probabilities between ESRD health states (i.e., dialysis dependent, renal transplant, posttransplant dialysis dependent) were sourced from Nuijten et al. (2015).⁷

Discontinuation for the first 2 cycles in the model was based on the reported rate of renal worsening in the BLISS-LN trial, defined as increased proteinuria and/or impaired renal function.² From cycle 3 onward, it was assumed patients would discontinue at a rate equal to the probability of renal flares,² where time to first flare data were taken from the BLISS-LN trial and used to derive the annual flare probabilities fitted with a lognormal parametric function.

The annual probabilities of adverse events for both belimumab plus ST and ST alone were informed using data from the BLISS-LN trial and applied over the time horizon of the model as costs and disutilities.²

Health state utility values in the model were informed by 3-Level EQ-5D estimates for chronic kidney disease (CKD) health states, as reported by Jesky et al. (2016).⁸ Estimates were reported by CKD stages, including G1/G2, G3a, G3b, G4, and G5, which corresponded to greater than or equal to 60 mL/min, 45 mL/min to 59 mL/min, 30 mL/min to 44 mL/min, 15 mL/min to 29 mL/min, and less than 15 mL/min eGFR health states, respectively.² It was assumed that the renal transplant health state had a utility value equivalent to the eGFR greater than or equal to 60 health state, and the posttransplant dialysis-dependent health state had the same utility value as the dialysis-dependent health state.² Renal flare and adverse event–related disutility values were informed by published literature.^9-12 In the sponsor’s base case, a steroid sparing utility was also applied as reported by Bexelius et al. (2013).¹³

The drug acquisition cost of belimumab was provided by the sponsor, while the cost of all other therapies (i.e., AZA, CYC, MMF, prednisone) were informed by the Ontario Drug Benefit formulary or DeltaPA.^14,15 Treatment adherence was assumed to be 100%.² Drug administration, disease management, renal flare, and end-of-life costs were sourced from the Ontario Nurses’ Association¹⁶ and published literature.^17-21 All costs were expressed as 2021 Canadian dollars.²

Summary of Sponsor’s Economic Evaluation Results

All analyses were run probabilistically (1,000 iterations). Submitted deterministic analyses were aligned with the probabilistic results. The probabilistic findings are presented in the Base-Case Results section.

Base-Case Results

Based on the sponsor’s probabilistic sequential analysis, belimumab plus CYC followed by AZA was dominated (i.e., more costly and less effective) by belimumab plus MMF. The least costly nondominated therapy was CYC followed by AZA, with the second and third most expensive therapies being MMF and belimumab plus MMF. Belimumab plus MMF was associated with an additional 0.57 QALYs at an additional cost of $196,902 compared to MMF, resulting in an ICER of $345,269 per QALY gained. A disaggregated summary of the incremental analysis is presented in Appendix 3.

The results of the sponsor’s probabilistic pairwise base-case analysis can be found in Appendix 3. It demonstrated that belimumab plus CYC followed by AZA was associated with an ICER of $515,277 per QALY gained versus CYC followed by AZA alone. For the comparison between belimumab plus MMF and MMF alone, the ICER was $345,269 per QALY gained.

Based on the deterministic results, the majority (94%) of the incremental QALYs for belimumab versus CYC followed by AZA or MMF were found to be accrued during the extrapolation period (i.e., after the approximately 2 years observed from BLISS-LN).

Table 3: Summary of the Sponsor’s Economic Evaluation Results (Sequential Analysis)

AZA = azathioprine; CYC = cyclophosphamide; ICER = incremental cost-effectiveness ratio; MMF = mycophenolate mofetil; QALY = quality-adjusted life-year.

Source: Sponsor’s pharmacoeconomic submission.²

Sensitivity and Scenario Analysis Results

The sponsor conducted several scenario analyses pertaining to different discounting, maximum treatment duration, utility sources, inclusion of treatment waning, inclusion of proteinuria disutilities, exclusion of administration costs, reduced renal flare costs and disutilities, and perspective. No scenarios had an important effect on the ICER, with the following exceptions: when a 5-year maximum treatment duration was included, belimumab plus MMF had an ICER of $29,858 per QALY versus CYC followed by AZA; when waning was included, belimumab plus MMF had an ICER of $315,434 per QALY versus CYC followed by AZA, and belimumab plus CYC followed by AZA was extendedly dominated through CYC followed by AZA and belimumab plus MMF.

CADTH Appraisal of the Sponsor’s Economic Evaluation

CADTH identified several key limitations to the sponsor’s analysis that have notable implications for the economic analysis:

• The reimbursement request population cannot be modelled. The modelled population in the sponsor’s base-case analysis explores the cost-effectiveness of belimumab plus ST (i.e., either MMF alone or CYC followed by AZA) in the full Health Canada indication, which is not aligned with the reimbursement request, which includes the following criteria: adult patients aged 18 years and older who are receiving ST and who have class III, class IV, and/or class V active LN, and if no improvements in disease activity and/or symptoms are observed after 6 months, use should be discontinued. The sponsor’s model was not sufficiently flexible to report the cost-effectiveness of belimumab plus ST in the reimbursement request population, incorporating the criteria that patients who do not respond to treatment within 6 months discontinue. It was noted by the sponsor that a half-cycle correction implemented within the sponsor’s economic model may account for patients discontinuing at 6 months due to increased proteinuria and/or impaired renal function. However, the underlying data from BLISS-LN did not capture the sponsor’s criteria; therefore, the sponsor’s modelling practice may represent some form of discontinuation but is neither representative nor a direct substitute for the reimbursement request population criteria of patients who do not respond within 6 months.

  • CADTH was unable to address this limitation within the model. The cost-effectiveness of belimumab plus ST in the reimbursement request population is uncertain because the criteria specifying discontinuation in patients who did not respond to treatment within 6 months was not incorporated into the model’s structure or clinical efficacy inputs.

• Uncertainties in modelled disease progression. In the sponsor-submitted model, the distribution of patients across health states for the first 2 years of the model was based on the baseline eGFR and percentage decline data from the BLISS-LN trial for both the CYC followed by AZA and the MMF regimens. Results from the sponsor’s base-case analysis found a large difference between the life-years gained between ESRD health states (Appendix 3); this does not align with the clinical expert feedback received by CADTH, which stated that CYC followed by AZA or MMF alone were likely to have similar efficacy. Discrepancies between the ST regimen results is likely due to uncertainties from the CYC followed by AZA subgroup. As noted in the CADTH clinical report, subgroup data regarding CYC followed by AZA were insignificant and imprecise, as analyses were limited by the small number of patients. As a result, uncertainty is likely propagated into the reported transition probabilities between eGFR health states from the trial period, which has significant downstream implications as renal outcomes are primarily driven by eGFR improvements from the first 2 cycles (i.e., during the BLISS-LN trial period).

  For transitions between eGFR health states (excluding ESRD health states) for year 2 onward, inputs are informed by values from Hanly et al. (2016) and adjusted to reflect the risk of ESRD and death for each baseline eGFR as a percentage decline category, as reported by Coresh et al. (2014).^5,6 According to the CADTH clinical report, Coresh et al. (2014) provided some evidence in support of using lesser declines in eGFR as an alternative end point for CKD progression, and clinical expert feedback further confirmed that transition probabilities from Hanly et al. (2016) would likely reflect Canadian clinical practice and that declining eGFR is predictive of development of ESRD; however, it was noted that this was likely an oversimplification as other factors not captured in the model, such as comorbidities and proteinuria levels, would also impact disease progression.

  • CADTH was unable to address this limitation within the model due to the model structure and data limitations. The direction and magnitude of the impact on the cost-effectiveness results for belimumab plus ST are unknown.

• Model structure does not adequately reflect the management of active LN in clinical practice. In the sponsor-submitted economic model, subsequent therapies were not considered, and it was assumed that once patients discontinued belimumab plus ST, they would continue to receive ST alone. However, clinical expert feedback received by CADTH noted that this modelled treatment pattern is not reflective of clinical practice, as patients who discontinue belimumab treatment and/or have inadequate response to first-line induction therapy (e.g., those who do not respond to CYC followed by AZA within 4 months to 6 months) would instead be initiated on additional or different therapies as their disease would still need to be treated (i.e., reintroducing high-dose corticosteroids, modifying ST, and/or putting patients on an alternative therapy such as calcineurin inhibitors [i.e., tacrolimus, cyclosporine] or rituximab in addition to ST). Costs and outcomes associated with subsequent therapies after belimumab discontinuation were not modelled.

  Additionally, clinical expert feedback noted that patients in remission for 3 years to 5 years would likely discontinue immunosuppressive therapy; however, long-term data informing how discontinuation would differ between treatments and patients’ responses are limited. Although the sponsor incorporated functionality to discontinue a proportion of patients at a certain time (i.e., 5 years) in the model, this would not accurately capture patients in remission.

  • CADTH was unable to address this limitation due to the model structure and limited data informing comparative efficacy between belimumab plus ST versus other treatments for LN, such as calcineurin inhibitors and rituximab. The direction and magnitude of the impact on the cost-effectiveness results for belimumab plus ST are unknown.

• Effect of belimumab on long-term flare rate is unknown. In the sponsor’s base-case analysis, the annual rate of renal flares (i.e., 5.8% and 13.0% for belimumab plus ST and ST alone, respectively) was determined using the probability of a flare observed at 2 years based on a lognormal curve fitted to the time to first flare Kaplan-Meier data from BLISS-LN. The lognormal curve was selected as it was associated with the lowest Akaike information criterion and Bayesian information criterion values.² The Akaike information criterion and Bayesian information criterion reflect statistical fit in the observed trial period (i.e., interpolation); selection based on Akaike information criterion or Bayesian information criterion alone may not accurately predict long-term flare rates for treatments (i.e., extrapolation). Clinical expert feedback received by CADTH noted that the sponsor’s use of lognormal curves may be conservative, as 20% to 35% of patients who experience adequate disease control are expected to show relapse within 5 years and 27% to 66% of patients with SLE in remission eventually have subsequent flares.^22-24 However, given the absence of evidence, the long-term efficacy of belimumab on reducing flare rates is unknown.

  • CADTH was unable to fully address this limitation due to limitations in data availability regarding the effect of belimumab on long-term flare rate. Owing to this and other limitations that could not be addressed, CADTH was unable to derive a base-case analysis. Instead, an exploratory reanalysis using more appropriate assumptions was conducted. In the CADTH exploratory reanalysis, a generalized gamma curve was used to inform the time to first renal flare, and subsequently the annual probability of a renal flare, where an extrapolated 29.0% and 13.8% of patients in the ST alone arm and the belimumab plus ST arms, respectively, were expected to experience a renal flare by year 5.

  • To examine the impact of the extrapolated benefit of flares, CADTH conducted a scenario analysis assuming no difference in annual renal flare probabilities between belimumab plus ST and ST alone for year 2 onward.

• Utilities informed by CKD patients. In the sponsor’s base-case analysis, health state utility values were informed by a UK study examining the health-related quality of life associated with CKD.⁸ Clinical expert feedback sought by CADTH noted that patients with active LN are likely to have worse quality of life than their CKD counterparts, and therefore Jesky et al. (2016) values may not accurately reflect the quality of life of patients with active LN.

  • Due to limitations in available data, CADTH was unable to address this limitation. The direction and magnitude of the impact of the cost-effectiveness results are unknown.

• Cost-effectiveness model is overly complex. The submitted model was found to be lacking transparency and to be highly inefficient. Coding of the model was spread over multiple sheets, including more than 126 sheets for individual Markov trace calculations.

  • CADTH was unable to address this limitation.

Additional limitations were identified but were not considered to be key limitations. These limitations are outlined subsequently:

• Model ICERs were associated with variability. The sponsor’s probabilistic analyses were conducted using 1,000 iterations as it was estimated that ICER stabilization would occur at approximately 750 iterations onward. However, CADTH conducted multiple probabilistic analysis runs at 1,000 iterations and noted that the ICER was associated with some variability.

  • CADTH conducted the probabilistic exploratory reanalyses using 5,000 iterations.

Additionally, the following key assumptions were made by the sponsor and have been appraised by CADTH (refer to Table 4).

Table 4: Key Assumptions of the Submitted Economic Evaluation (Not Noted as Limitations to the Submission)

eGFR = estimated glomerular filtration rate; ESRD = end-stage renal disease; LN = lupus nephritis; ST = standard therapy.

CADTH Reanalyses of the Economic Evaluation

Base-Case Results

Due to limitations with the model design and efficacy data, CADTH was unable to determine a base-case estimate of the cost-effectiveness of belimumab plus ST compared to ST alone (CYC followed by AZA or MMF alone) in the modelled population. As such, the changes in Table 6 reflect a CADTH exploratory reanalysis, rather than a base-case estimate of cost-effectiveness. The CADTH exploratory reanalysis was derived by making changes in model parameter values and assumptions, in consultation with feedback from clinical experts.

The results of the CADTH exploratory reanalysis were consistent with those submitted by the sponsor. The exploratory reanalysis conducted by CADTH demonstrated that belimumab plus MMF compared to MMF was associated with 0.57 additional QALYs at an incremental cost of $201,083, resulting in an ICER of $352,880 per QALY gained. Belimumab plus CYC followed by AZA was dominated (more costly, less effective) by belimumab plus MMF. At a willingness-to-pay threshold of $50,000 per QALY, there was a 0.7% and 0.4% probability of, respectively, belimumab plus CYC followed by AZA and belimumab plus MMF being cost-effective compared to their ST alone. A summary of the CADTH exploratory reanalysis is presented in Table 6.

Table 5: CADTH Revisions to the Submitted Economic Evaluation