Reimbursement Review

Burosumab (Crysvita)

Sponsor: Kyowa Kirin Canada, Inc.

Therapeutic area: Treatment of X-linked hypophosphatemia

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Clinical Review

Abbreviations

Executive Summary

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

Table 1: Background Information on Application Submitted for Review

NOC = Notice of Compliance; XLH = X-linked hypophosphatemia.

Introduction

X-linked hypophosphatemia (XLH) is a rare, chronically debilitating genetic disorder.¹ It is characterized by renal phosphate wasting and consequent defective bone mineralization caused by inactivating mutations in the PHEX gene.^2-4 Patients with XLH produce excess of the protein FGF23, leading to the impaired conservation of phosphate and consequent hypophosphatemia,⁵ suppressing the production of 1,25-dihydroxyvitamin D — or 1,25(OH)₂D in short — and resulting in a decrease in the intestinal absorption of calcium and phosphate.^1,6

XLH in children is characterized by vitamin D–resistant rickets.¹ Adults with XLH can display manifestations such as osteomalacia, fractures and pseudofractures, early-onset osteoarthritis, and enthesopathies.^7-12 These abnormalities in adults with XLH result in musculoskeletal pain and stiffness, impaired mobility and physical function, fatigue, and reduced health-related quality of life (HRQoL).^7,13

Published information about the incidence and prevalence of XLH is limited. The estimated prevalence of XLH in Norway is 1 case per 100,000 children.¹⁴ The estimated prevalence of hypophosphatemic rickets in southern Denmark is 4.8 cases per 100,000 people (children and adults)¹⁵ and 2.03 cases per 100,000 people in Colombia.¹⁶ A recent population-based cohort study using a large primary care database in the UK estimated adult XLH prevalence as being 1.57 cases per 100,000 people.¹⁷ There are no known reported prevalence estimates for Canada.

In adults, primary treatment generally consists of the continued use of oral phosphate and active vitamin D analogues as well as pain management and orthopedic interventions.^1,6

Burosumab is a recombinant human immunoglobulin G subclass 1 monoclonal antibody that inhibits the biological activity of FGF23 and thereby increases both renal phosphate reabsorption and the serum concentration of 1,25(OH)₂D.¹⁸ It is indicated for the treatment of XLH in adult and pediatric patients aged 6 months and older.¹⁸ The burosumab dosing regimen for the treatment of XLH in adult patients is 1 mg/kg rounded to the nearest 10 mg up to a maximum dose of 90 mg, administered by subcutaneous (SC) injection every 4 weeks.¹⁸

Burosumab was previously reviewed by CADTH and a recommendation to reimburse with conditions was issued in May 2020 for the treatment of pediatric patients with XLH; a recommendation not to reimburse was issued for adults.¹⁹ The sponsor has submitted additional data and requested reassessment of the reimbursement request for the treatment of adult patients with XLH, as this population is included in the indication approved by Health Canada. The sponsor is proposing the following criteria for adult patients with XLH, which the sponsor indicated align with international consensus guidelines.⁶

Burosumab treatment can be initiated in adult patients (≥ aged 18 years) who have:

• a clinical presentation consistent with XLH, including —

  • fasting hypophosphatemia, and

  • normal renal function (defined as fasting serum creatinine below the age-adjusted upper limit of normal [ULN])

• a confirmed PHEX gene variant in either the patient or in a directly related family member with appropriate X-linked inheritance and —

  • persistent bone and/or joint pain due to XLH, and/or

  • osteomalacia that limits daily activities, and/or

  • pseudofractures or osteomalacia-related fractures

• an insufficient response or refractoriness to conventional therapy or if patients experience complications related to conventional therapy.

The sponsor also proposed that patients should be assessed on an annual basis for continued benefit; treatment with burosumab can be renewed as long as the patient does not meet any of the discontinuation criteria. It was also proposed that in adult patients, burosumab should be discontinued if any of the following occur: hyperparathyroidism, nephrocalcinosis, evidence of fracture or pseudofracture based on radiographic assessment, or intolerable adverse events (AEs) (e.g., 1 patient discontinued the UX023-CL303 study [hereafter known as the CL303 trial] due to restless leg syndrome worsening from baseline).

Perspectives of Patients, Clinicians, and Drug Programs

The information in this section is a summary of input provided by the patient and clinician groups that responded to the call by Canada’s Drug Agency (CDA-AMC) for input and from the clinical expert consulted by CDA-AMC for the purpose of this review.

Patient Input

Input was submitted for this review by the Canadian XLH Network, a national, not-for-profit, patient support organization for people living and dealing with XLH. Information for this input was gathered through an online survey of XLH adult patients, family members, and caregivers from December 2 to 15, 2023.

Survey respondents indicated that symptoms of XLH during adulthood differed from childhood symptoms. When asked about adult symptoms, 44% of patients reported severe pain, 28% reported a loss of mobility, 21% reported a lack of energy, 21% had an increase in dental issues, and 26% had developed arthritis and/or spinal stenosis. All of these symptoms were reported to significantly impact patients’ quality of life as well as their social and psychological well-being.

Survey respondents indicated that with conventional treatment (a combination of phosphate and calcitriol), patients need to take large doses of phosphate up to 5 times daily and calcitriol 1 to 2 times daily; this addresses the issue of low phosphate but does not address pain and other serious symptoms of XLH. In addition, conventional treatment has serious side effects, such as nephrocalcinosis, kidney disease, calcium deposits, and parathyroid issues, all while allowing XLH to continue progressing. Furthermore, phosphate is very expensive and hard to access due to supply chain issues.

Respondents indicated that there is a need for treatment options that are accessible, affordable, and easier to take and that can boost energy levels and muscle function, reduce pain, and improve bone health and overall quality of life, with fewer side effects.

Clinician Input

Input From Clinical Experts Consulted by CDA-AMC

The clinical expert consulted by CDA-AMC noted that the goals of treatment in adults are to reduce osteomalacia and pseudofractures to alleviate generalized bone pain, enhance mobility that may be reduced, and cure any non-union fractures. Unmet needs pertained to the fact that while current treatment reduces downstream effects of the elevated FGF23 levels, while attempting to normalize serum phosphorus and 1,25(OH)₂D, may further elevate FGF23 levels to cause a feedback loop that limits the efficacy of conventional treatment. The clinical expert also noted that there is a side effect burden to conventional therapy, including gastrointestinal upset due to oral phosphate, and hypercalciuria and nephrocalcinosis due to 1,25(OH)₂D treatment, which can reduce kidney function and cause secondary hyperparathyroidism. In addition, the clinical expert stated that the majority of patients (> 70%) continue to have symptoms of pain, mobility issues, or complications despite treatment. Furthermore, since active vitamin D may need to be administered twice daily and oral phosphate is usually administered several times per day, adherence may not be optimal.

Per the clinical expert, burosumab would represent a shift in the current treatment paradigm as it addresses the underlying disease at an upstream level rather than a downstream level. They noted that treatment with burosumab is likely to be lifelong as the cause of the disease is a genetic mutation, which results in consequences that persist throughout life.

Per the clinical expert, symptomatic patients with bone pain due to bone disease (i.e., due to osteomalacia, pseudofractures, and nonunion fractures) are best suited for treatment. However, they also noted there may be benefit in treating adults with limited symptomatology to increase activity levels and a sense of well-being.

In the clinical expert’s practice, they would consider a reduction in bone pain, a reduction in fractures, and the healing of fractures to be clinically meaningful responses to therapy. Laboratory evidence of the normalization of serum phosphorus and biomarkers of bone metabolism (e.g., alkaline phosphatase) and the absence of elevations in serum creatinine or parathyroid hormone (PTH) as well as the absence of the development or acceleration of nephrocalcinosis would also be considered clinically meaningful responses.

The clinical expert noted that patients who are experiencing a sustained decline in serum phosphorus levels despite adherence to therapy (suggesting that burosumab treatment is not working) or who develop a severe allergic reaction to burosumab should discontinue therapy. Therapy should be continued if initiated during childhood as long as the patient does not meet any of the discontinuation criteria, since the consequences of elevated FGF23 can also be seen in adults. Specialist attention would likely be required to diagnose, treat, and monitor patients receiving burosumab (i.e., either an endocrinologist or rheumatologist with knowledge of the disorder).

Clinician Group Input

No input was received by clinician groups by the deadline of the call for input.

Drug Program Input

Input was obtained from the drug programs that participate in the CDA-AMC reimbursement review process. The following were identified as key factors that could potentially impact the implementation of a CDA-AMC recommendation for burosumab:

• considerations for the initiation of therapy

• considerations for the discontinuation of therapy

• care provision issues.

The clinical expert consulted by CDA-AMC provided advice on the potential implementation issues raised by the drug programs. Refer to Table 4 for more details.

Clinical Evidence

Systematic Review

Description of Studies

The major focus for the reassessment of this indication was additional data analysis results for the 48-week and 96-week mark of the CL303 clinical trial, as well as an ad hoc week 48 analysis of the placebo-emergent arm (placebo treatment during the first 24 weeks, switching to burosumab after 24 weeks).^20,21 The CL303 trial, which was included in the original submission, was a phase III, double-blind, placebo-controlled, randomized controlled trial (RCT) consisting of a 24-week placebo-controlled period and 2 open-label extensions providing 96 weeks of total follow-up. Patients in this study had to be aged 18 years to 65 years, with a diagnosis of XLH supported by classic clinical features of adult XLH (such as short stature or bowed legs) and either a documented PHEX mutation (in either the patient or in a directly related family member with appropriate X-linked inheritance) or a serum intact FGF23 level of greater than 30 pg/mL by Kainos assay; biochemical findings consistent with XLH — namely, serum phosphorus of less than 0.81 mmol/L and a ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate (TmP/GFR) of less than 2.5 mg/dL; an estimated glomerular filtration rate of 60 mL per minute or more (using the Chronic Kidney Disease Epidemiology Collaboration equation); or an estimated glomerular filtration rate of 45 mL per minute to less than 60 mL per minute at the second screening visit, with confirmation that the renal insufficiency was not due to nephrocalcinosis; as well as the presence of skeletal pain attributed to XLH and/or osteomalacia based on a Brief Pain Inventory (BPI) worst pain score of 4 or more at the first screening visit.

The proportion of patients attaining serum phosphorus levels above the lower limit of normal (LLN) (0.81 mmol/L) at the midpoint of the dosing cycle from baseline to week 24 was the primary outcome of the study. Key secondary end points were also measured at 24 weeks and included change in the following patient-reported outcome (PRO) measures: the BPI worst pain score, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) stiffness score, and the WOMAC physical function score. Other secondary end points included domains of the BPI, WOMAC, and Brief Fatigue Inventory (BFI) measured at week 24, week 48, and week 96. WOMAC is a self-administered questionnaire assessing pain, stiffness, and physical functioning in patients with hip and knee osteoarthritis, comprising pain, physical function, and stiffness domains; a higher score indicates worse pain, stiffness, and functional limitations. The BPI is a self-reported questionnaire designed to provide information about pain intensity (the sensory dimension) and the degree to which pain interferes with daily living (the reactive dimension); a high score represents a high level of pain intensity or pain interference. The BFI is a self-reported questionnaire that assesses the severity of fatigue and the impact of fatigue on daily functioning, measuring fatigue and the interference of fatigue on daily life; the items are measured on a 0 to 10 numeric rating scale and a score of 7 to 10 is considered severe fatigue.²²

The proportion of patients attaining serum phosphorus levels above the LLN at the end of their dosing cycle (i.e., 4 weeks after dosing) was also a secondary end point measured at week 48, as were measures of bone metabolism (bone-specific alkaline phosphatase [BALP]), 1,25(OH)₂D, and phosphorus homeostasis (TmP/GFR and tubular reabsorption of phosphate [TRP]), measured at week 24, week 48, and week 96. Exploratory end points were active pseudofractures and/or fractures, as well as the 6-minute walk test (6MWT), a supervised test that measures the distance a patient can walk on a hard, flat surface over a 6-minute period. Both were measured at week 24 and week 48 (neither exploratory outcome was measured at week 96).

Baseline characteristics were generally balanced between the 2 treatment arms. In terms of medical history, a numerically higher proportion of patients in the burosumab arm had osteoarthritis (69.1% of patients versus 57.6% of patients in the placebo arm). A numerically higher proportion of patients in the burosumab arm were classified as having a BPI average pain score of more than 6 (32.4% of patients in the burosumab arm and 25.6% of patients in the placebo arm); similarly, a numerically higher proportion of patients in the burosumab arm were classified as having a BPI worst pain score of more than 6 (77.9% of patients in the burosumab arm and 65.2% of patients in the placebo arm). A numerically higher proportion of patients in the burosumab arm had nephrocalcinosis than in the placebo arm (16.2% versus 7.6% of patients, respectively). The majority of patients in the burosumab and placebo arms (86.8% and 93.9% of patients, respectively) had received both vitamin D analogues and phosphate before the trial. There were no notable imbalances in baseline laboratory characteristics. A higher proportion of patients in the placebo arm had active pseudofractures at baseline (51.5%) than patients in the burosumab arm (42.6%). The majority of patients in both arms had had previous orthopedic surgery (66.2% of patients in the burosumab arm and 71.2% of patients in the placebo arm) or were taking nonopioid pain medications at baseline (65.2% of patients in the placebo arm and 69.1% of patients in the burosumab arm).

Efficacy Results

Proportion of Patients With Serum Phosphorus Levels Greater Than LLN

Following crossover to burosumab after week 24, the additional data from the reassessment reported that the proportion of patients in the placebo-emergent arm with midpoint serum phosphorus levels greater than LLN was 89.4% (95% confidence interval [CI], 79.7% to 94.8%) at week 48 and 68.2% (95% CI, 56.2% to 78.2%) at week 96. The proportion of patients with serum phosphorus levels greater than LLN in the burosumab-emergent arm (burosumab treatment during the first 24 weeks with continued burosumab after 24 weeks) was 83.8% (95% CI, 73.3% to 90.7%) at week 48 and 82.4% (95% CI, 71.6% to 89.6%) at week 96. There was no information on the patients with end point serum phosphorus levels greater than LLN for week 48 and week 96.

Brief Pain Inventory

Additional information submitted for the BPI worst pain scores reported that at week 48, the least squares mean (LSM) change from baseline in the placebo-emergent arm was −1.53 (95% CI, −1.98 to −1.09) and in the burosumab-emergent arm was −1.09 (95% CI, −1.51 to −0.66). At week 96, the LSM changes from baseline in the placebo-emergent arm was −0.99 (95% CI, −1.51 to −0.47) and in the burosumab-emergent arm was −1.48 (95% CI, −2.07 to −0.90).

BPI pain interference results at week 48 reported an LSM change from baseline of −1.27 (95% CI, −1.77 to −0.78) in the placebo-emergent arm and −1.04 (95% CI, −1.51 to −0.56) in the burosumab-emergent arm. At week 96, the LSM change from baseline was −1.08 (95% CI, −1.59 to −0.57) in the placebo-emergent arm and −1.43 (95% CI, −1.89 to −0.97) in the burosumab-emergent arm.

BPI pain severity results reported that at week 48, the LSM change from baseline in the 2 study arms was −1.20 (95% CI, −1.58 to −0.81) in the placebo-emergent group and −0.85 (95% CI, −1.16 to −0.54) in the burosumab-emergent group. At week 96, the LSM change from baseline was −1.18 (95% CI, −1.57 to −0.80) in the placebo-emergent arm and −1.42 (95% CI, −1.87 to −0.97) in the burosumab-emergent arm.

Western Ontario and McMaster Universities Osteoarthritis Index

For WOMAC physical function, at week 48, the LSM change from baseline was –6.35 (95% CI, –11.94 to –0.76) in the placebo-emergent arm and –7.76 (95% CI, –11.97 to –3.55) in the burosumab-emergent arm. At week 96, the LSM change from baseline was –8.41 (95% CI, –13.80 to –3.01) in the placebo-emergent arm and –9.02 (95% CI, –13.47 to –4.57) in the burosumab-emergent arm.

WOMAC stiffness scores reported that at week 48, the LSM change from baseline was –15.29 (95% CI, –22.23 to –8.35) for the placebo-emergent arm and –16.03 (95% CI, –22.53 to –9.53) in the burosumab-emergent arm. At week 96, the LSM change from baseline was –17.67 (95% CI, –24.99 to –10.34) in the placebo-emergent arm and –15.32 (95% CI, –22.33 to –8.31) in the burosumab-emergent arm.

WOMAC pain scores were not analyzed, but a trend toward numerically increasing reductions was reported between week 48 and week 96, for both the placebo-emergent and burosumab-emergent treatment arms.

6-Minute Walk Test

At week 48, the mean total distance walked at baseline was 367.28 m (standard deviation [SD] = 104.22) in the placebo-emergent arm and 365.66 m (SD = 125.44) in the burosumab-emergent arm. The LSM change from baseline in total distance walked was –5.71 (95% CI, –21.70 to 10.28) in the placebo-emergent arm and 5.92 (95% CI, –15.00 to 26.84) in the burosumab-emergent arm. This outcome was not measured at week 96.

Brief Fatigue Inventory

At week 48, the LSM change from baseline in BFI worst fatigue was –1.23 (95% CI, –1.84 to –0.62) in the placebo-emergent arm and –1.01 (95% CI, –1.57 to –0.45) in the burosumab-emergent arm. At week 96, the LSM change from baseline was –0.82 (95% CI, –1.53 to –0.11) in the placebo-emergent arm and –0.75 (95% CI, –1.35 to –0.26) in the burosumab-emergent arm.

At week 48, the LSM change from baseline in BFI global fatigue was –0.73 (95% CI, –1.34 to –0.12) in the placebo-emergent arm and –0.46 (95% CI, –1.01 to 0.09) in the burosumab-emergent arm. At week 96, the LSM change from baseline was –0.86 (95% CI, –1.43 to –0.29) in the placebo-emergent arm and –0.80 (95% CI, –1.36 to –0.25) in the burosumab-emergent arm.

Fractures and Pseudofractures

The reassessment submission’s additional 24-week analyses reported a higher probability of a fully healed fracture at 24 weeks in the burosumab arm: 0.458 in the burosumab arm versus 0.048 in the placebo arm (odds ratio [OR] = 16.76 [95% CI, 4.93 to 56.95]).

At 48 weeks, 46.2% of patients in the placebo arm and 57.1% of patients in the burosumab arm reported healed active fractures. In addition, 33.3% of patients in the placebo-emergent arm and 64.7% of patients in the burosumab-emergent arm reported healed pseudofractures. The probability of a fully healed fracture was 0.725 (95% CI, 0.516 to 0.933) in the burosumab-emergent arm and 0.386 (95% CI, 0.718 to 0.594) in the placebo-emergent arm. Fracture outcomes were not measured at 96 weeks.

Key Serum Biomarkers

At week 48, the LSM change from baseline for the levels of serum 1,25(OH)₂D was 10.50 (95% CI, 5.76 to 15.24) in the placebo-emergent arm and 7.24 (95% CI, 2.44 to 12.04) in the burosumab-emergent arm. At week 96, the serum 1,25(OH)₂D was 3.43 (95% CI, –1.17 to 8.03) in the placebo-emergent arm and 1.95 (95% CI, –2.66 to 6.57) in the burosumab-emergent arm.

At week 48, the LSM change from baseline in TmP/GFR in the placebo-emergent arm was 0.55 (95% CI, 0.38 to 0.72) and 0.48 (95% CI, 0.30 to 0.65) in the burosumab-emergent arm. At week 96, the LSM change was 0.29 (95% CI, 0.12 to 0.46) in the placebo-emergent arm and 0.46 (95% CI, 0.29 to 0.62) in the burosumab-emergent arm.

At week 48, the LSM change from baseline in TRP was 0.02 (95% CI, 0.00 to 0.05) for the placebo-emergent arm and 0.03 (95% CI, 0.02 to 0.05) in the burosumab-emergent arm. At week 96, LSM changes from baseline in the placebo-emergent group was –0.01 (95% CI, –0.04 to 0.02), while the burosumab-emergent group was 0.03 (95% CI, 0.01 to 0.05).

At week 48, the LSM change from baseline in BALP in the placebo-emergent arm was 6.69 (95% CI, 2.91 to 10.47) and in the burosumab-emergent arm was 0.23 (95% CI, –3.36 to 3.81). At week 96, the LSM change in the placebo-emergent arm was –2.49 (95% CI, –6.19 to 1.21) and –2.76 (95% CI, –5.98 to 0.45) in the burosumab-emergent arm.

Harms Results

Overall, 97% of patients in the placebo-emergent arm and 100% of patients in the burosumab-emergent arm experienced a treatment-emergent adverse event (TEAE). There were differences between the proportions of patients experiencing some TEAEs between the burosumab-emergent arm during the trial and the placebo-emergent arm after initiating burosumab. Specifically, there were differences in the proportion of patients in the burosumab-emergent arm and placebo-emergent arm reporting tooth abscesses (28% and 8%, respectively), vitamin D deficiency (22% and 11%, respectively), injection site reactions (12% and 25%, respectively), diarrhea (19% and 8%, respectively), upper respiratory tract infection (18% and 3%, respectively), nausea and dizziness (both 16% and 8% in each arm, respectively), depression (13% and 5%, respectively), hypoesthesia (10% and 5%, respectively), migraine (10% and 3%, respectively), oropharyngeal pain (6% and 12%, respectively), injection site pruritus (4% and 12%, respectively), and ectopic mineralization (0% and 11%, respectively).

During the placebo-controlled period, a serious adverse event (SAE) was reported in 1 patient in the placebo-emergent arm and in 2 patients in the burosumab-emergent arm. In the placebo-emergent arm during burosumab treatment, 10 patients overall reported SAEs. The burosumab-emergent arm reported SAEs in 12 patients during the whole trial. There were no withdrawals due to AEs and 1 death due to a traffic accident in the burosumab-emergent arm (judged not related to treatment).

AEs of special interest included injection site reactions, hypersensitivity, hyperphosphatemia, ectopic mineralization, and restless leg syndrome. A total of 16 (24%) patients in the placebo-emergent arm reported injection site reactions after initiating burosumab and 8 (12%) patients reported injection site reactions before initiating burosumab. In addition, 7 (11%) patients in the placebo-emergent arm experienced ectopic mineralization, which was not reported in any of the other treatment arms.

Noting the higher proportions of patients in the burosumab-emergent arm experiencing TEAEs and serious TEAEs, the reassessment included an exposure-adjusted analysis reporting incidence rates in each arm; it reported generally similar incidence rates in the placebo-emergent and burosumab-emergent arms.

Critical Appraisal

Internal Validity

There are some limitations impacting the internal validity of the CL303 trial. First, there were some concerns with imbalances in certain medical characteristics, which could impact the outcomes. A numerically higher proportion of patients in the burosumab arm had osteoarthritis (69.1% versus 57.6% of patients) and nephrocalcinosis (16.2% versus 7.6% of patients) relative to the placebo arm, and a numerically higher proportion of patients in the placebo arm (51.5% versus 41.6%) had active pseudofractures at baseline. These could bias the assessment of efficacy on outcomes pertaining to pain, active pseudofracture healing, and physical function. In addition, the sample size was powered only for the primary end point to demonstrate a statistically significant difference at week 24 for the primary outcome of serum phosphorus, and therefore there could be a lack of power for key secondary but clinically important outcomes, such as the PROs; this adds uncertainty regarding the exact magnitude of benefit for these outcomes. Furthermore, a total of 6 patients had discontinued from the placebo-emergent arm and 8 patients had discontinued from the burosumab-emergent arm as of week 96. Given the relatively small sample size, this could represent a notable loss to follow-up. In addition, by virtue of the study design, at week 48 all patients crossed over to receive burosumab. The lack of control group makes it difficult to attribute the changes in efficacy outcomes and harms to burosumab alone during the open-label phases, which may be of particular importance for the harms results as a possible cumulative side effect burden was observed. Furthermore, the reporting of PRO results, as subjective measures, could be impacted by the open-label design. There were some missing data for serum biomarkers and PRO scores (data from approximately 59 patients were available at the 96-week mark), which were handled by exclusion from the analysis, but the potential impacts of this choice were not explored by sensitivity analysis relative to other missing data methods. For fracture outcomes, only targeted radiography was performed to check the progress of fractures after the initial scan at baseline, and these scans did not appear to be identifying new fractures. This could impact the detection of new fractures in particular as the development or absence of fractures in non–X-rayed sites may be missed. Lastly, patients in the burosumab-emergent arm had higher rates of pain medication usage; these differences may bias the efficacy results, particularly those of the WOMAC pain and BPI measures.

Table 2: Summary of Key Results From Pivotal Studies and RCT Evidence

1,25(OH)₂D = 1,25-dihydroxyvitamin D; 6MWT = 6-minute walk test; BALP = bone-specific alkaline phosphatase; BFI = Brief Fatigue Inventory; BPI = Brief Pain Inventory; CI = confidence interval; GEE = generalized estimating equation; LLN = lower limit of normal; LSM = least squares mean; NR = not reported; OR = odds ratio; SAE = serious adverse event; SD = standard deviation; TEAE = treatment-emergent adverse event; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate; TRP = tubular reabsorption of phosphate; vs. = versus; WDAE = withdrawal due to adverse event; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

^aThe 95% CI for the proportion of patients who attain mean serum phosphorus levels above the LLN was calculated using the Wilson score method.

^bThe P value was not adjusted for multiplicity.

^cEstimates of LSMs and P values for change from week 24 to week 48 in the placebo-emergent arm are from an ad hoc GEE model, similar to those for the planned analysis.

^dThe 95% CI and P value corresponds to the probability of a fracture being graded as fully healed for the week 48 analysis.

^eHarms in the placebo-emergent arm are those that occurred in the placebo arm after initiating burosumab (i.e., the treatment continuation period [week 24 to week 48], extension period I [week 48 to week 96], or extension period II [through week 149]). Harms in the burosumab-emergent group are those that occurred during any treatment period.

Sources: Sponsor’s Summary of Clinical Evidence,²³ CL303 Clinical Study Report,²⁰ and additional information provided by the sponsor.^21,24

External Validity

There are some limitations impacting the external validity of the study. The study focused on adult patients with XLH but did not specifically select patients diagnosed with XLH as adults; furthermore, the patient population included were majority white and majority female, which could have under-represented patient populations, including Indigenous Peoples and males. The majority of the patient population reported receiving vitamin D and phosphate before treatment with burosumab, which does not provide any information on the effectiveness of burosumab for adults with XLH who have received no prior treatments. The frequent visits and dose adjustment protocols used in the trial setting may not exactly reflect daily clinical practice in Canada, and the optimized efficacy and safety profile during the trial may not be extrapolatable to the general patient population. Moreover, patients were prohibited from using certain concomitant medications during the trial. This may not represent prescribing patterns in routine practice and may impact the generalizability of the findings from these additional data analyses. For fracture healing, which may take longer to capture, the duration of the study may not have been a long enough time to fully determine the impact of burosumab on these outcomes. Furthermore, the PRO measures used in the study were noted by the clinical expert to not be routinely used in clinical practice, suggesting that the impact of treatment on subjective measures such as pain, fatigue, and stiffness in the clinical trial may not be easily translated into these settings. Lastly, while the sponsor provided minimal clinically important differences (MCIDs) for the WOMAC, BPI, and BFI domain scores, it is important to note that apart from 2 domains of WOMAC, where a general population sample was cited by the sponsor, these MCIDs were generated from cross-sectional data (the UX023-CL001 study),²⁵ phase II clinical trial data (the UX023-CL203 study),²⁵ and/or CL303 (pivotal trial) data,²⁴ all collected by the sponsor. The UX023-CL001 and UX023-CL203 studies were cross-sectional data and had a small sample size (N = 20), respectively, and were considered exploratory. The CL303 trial data are both the data source for the MCIDs and the data analyzed in the trial. Therefore, there is no external reference data in a population with XLH to use as a comparison for meaningful clinical change, and there remains a lack of confirmatory data on the meaningfulness of these score changes in the general adult XLH population.

Long-Term Extension Studies

Description of Studies

Study BUR02^26,27 was an open-label, phase III study evaluating the long-term efficacy and safety of burosumab in adult patients with XLH. It was undertaken using patient populations that had completed the CL303 study²⁰ (a phase III RCT that evaluated measures of phosphate metabolism, PROs, and fractures and/or pseudofractures in adults with XLH) or the UX023-CL304 study, hereafter known as the CL304 trial (this was a phase III, single-arm study that evaluated measures of osteomalacia in patients with XLH who received burosumab treatment, and was not appraised in the current submission).²⁸ Patients completing the CL303 study were eligible to transition to the BUR02 study; however, there was an interval between the CL303 and BUR02 trials (a mean of 9 months; range, 6 months to 16 months) where interim burosumab treatment was provided via an early access program only to the patients for whom the drug supply was accessible.

Efficacy Results

Serum Phosphate Levels Above the LLN

At the baseline of the BUR02 trial, 34.3% of patients had serum phosphate levels above the LLN. The proportion increased to 55.9% at week 12 and remained mostly within a range between 55% and 75% in subsequent visits. At the end of the study, 66.7% of the patients had serum phosphate levels above the LLN.

Key Serum Biomarkers

At the CL303 trial baseline, the mean TmP/GFR was 0.55 (SD = –0.15) mmol/L and it increased to 0.70 (SD = 0.26) mmol/L at week 12a; this level was sustained through both the CL303 and BUR02 studies. At the final analysis, the mean (SD) TmP/GFR was 0.62 (SD = 0.22) mmol/L and it increased to 0.69 (SD = 0.14) mmol/L at week 48b; these levels were sustained over time.

At the interim analysis, the mean (SD) serum 1,25(OH)₂D was 79.95 (SD = 29.77) pmol/L at the CL303 trial baseline, 98.56 (SD = 30.27) pmol/L at week 48a, and 83.36 (SD = 32.97) pmol/L at week 72a. At the baseline of the BUR02 trial, the mean (SD) serum 1,25(OH)₂D was 78.43 (SD = 41.49) pmol/L and it increased to 92.85 (SD = 36.06) pmol/L at week 12b, remaining consistent to week 48b of the BUR02 study.

According to the final analysis, at baseline, the mean (SD) serum concentration of 1,25(OH)₂D was 32.67 (SD = 16.35) pg/mL. At week 12, the 1,25(OH)₂D concentration increased to 39.86 (SD = 15.57) pg/mL. At week 24, week 48, week 72, and week 96, the mean (SD) serum 1,25(OH)₂D levels were 36.34 (SD = 9.80) pg/mL, 37.04 (SD = 7.83) pg/mL, 38.16 (SD = 11.30) pg/mL, and 41.01 (SD = 12.80) pg/mL, respectively. At the end of the study, the mean (SD) serum 1,25(OH)₂D was 38.53 (SD = 12.70) pg/mL.

Patient-Reported Outcomes

Based on the interim analyses in the CL303 study, the LSM (standard error [SE]) of WOMAC stiffness scores was –14.77 points (SE = 4.03 points) at week 36a; this reduction was sustained at all subsequent time points in the 2 studies. Similar results were reported for the WOMAC physical function score.

In the final analysis of the BUR02 study, the mean (SD) stiffness score was 55.15 (18.75) at baseline, and the mean (SD) change was –3.13 (17.68) at week 12. The mean stiffness scores were maintained at lower than baseline throughout subsequent visits. The mean (SD) changes in stiffness score from baseline to week 24, week 48, and week 96 were –9.19 (SD = 22.89) points, –8.62 (SD = 18.63) points, and –9.09 (SD = 20.48) points, respectively. At the end of the BUR02 study, the mean score decreased by –14.52 (22.61) points. Similar decreases were observed for the WOMAC pain score and the physical function score.

Based on the interim analyses in the CL303 study, the LSM change from baseline in the BPI average worst pain scores at week 12a was −0.88 (SE = 0.281) and it decreased from baseline at all subsequent time points in the 2 studies except for week 24a. The BPI pain interference scores had also decreased from baseline with an LSM change from baseline of −1.22 (SE = 0.309) at week 12a and at all subsequent time points in both studies except week 24a.

Similarly, according to the final analysis from the BUR02 trial, the mean (SD) BPI worst pain score was 5.78 (SD = 1.725) points at baseline. The mean change in BPI worst pain score from baseline to week 12 was −0.51 (SD = 1.698) points, and this level was maintained at lower than baseline at week 24, week 36, week 48, week 72, and week 96.

The mean BPI pain severity score was 4.52 (SD = 1.657) points at baseline (N = 32) and the mean change in BPI worst pain score from baseline was −0.40 (SD = 1.416) points at week 12 (N = 12). These values were maintained throughout subsequent visits. Similar decreases were observed for the BPI pain interference score.

Based on the interim analyses, the LSM of the BPI average worst fatigue scores decreased from baseline results and were consistent at all subsequent time points. Similar trends were observed for the BFI global fatigue score and fatigue interference score. The BFI fatigue severity scores had decreased from baseline with an LSM of −1.45 (SE = 0.45) at week 12a and at all time points through to the end of the BUR02 trial.

According to the final analysis, at baseline of the BUR02 study, the mean BFI worst fatigue score was 5.91 (SD = 1.75) points. The mean change in worst fatigue score from baseline to week 24, week 48, week 72, and week 96 were −0.49 (SD = 1.78) points, −0.46 (SD = 2.00) points, −0.34 (SD = 2.24) points, and −0.64 (SD = 1.73) points, respectively. Similar trends were observed for the BFI global fatigue score and fatigue interference score.

6-Minute Walk Test

At the interim analysis, the 6MWT actual distance walked increased from the CL303 trial baseline at week 24a to week 48b. At the final analysis, at the baseline of the BUR02 study, the mean actual distance walked was 393.3 (SD = 93.25) m. After BUR02 study entry and continuation with burosumab treatment, the mean changes in actual walking distance increased from baseline to week 12, and all subsequent visits.

Harms Results

Safety data were not evaluated as part of the interim analysis. At the final analysis, all patients had received all scheduled doses and no patients had skipped doses. Almost all patients (n = 34) experienced 1 or more TEAE but most events were mild to moderate in severity. Among the patients who experienced a TEAE, the most common TEAEs were vitamin D deficiency (55.9%), arthralgia (38.2%), and hypophosphatemia (26.5%).

Six patients experienced SAEs (17.1%); these events occurred in single patients from each subgroup. No patients experienced related treatment-emergent SAEs. No deaths or TEAEs leading to death were reported during the CL303 study. No patient had a TEAE that led to withdrawal of the study drug or study discontinuation. There was no notable difference in the overall incidence of AEs between the 2 subgroups.

Critical Appraisal

Internal Validity

The open-label designs of the BUR02 study could bias the magnitude of the efficacy of subjective PRO results due to unblinded exposure to the study medication during the treatment period. In addition, the absence of control arm in the BUR02 study and the lack of data beyond week 96 make interpretation of the long-term sustainability of treatment effect challenging.

The interim analysis showed that the clinical effect of burosumab decreased when treatment was interrupted and returned after patients resumed the medication, but analysis based on the doses received by the patients was not performed and it cannot be confirmed whether those who received 1 dose versus 6 doses of burosumab would have different outcomes.

Furthermore, treatment history and concomitant medications during the gap between the pivotal studies and the BUR02 trial were not assessed, limiting the ability to interpret the outcomes efficiently.

External Validity

As the BUR02 study consisted of patients who took part in the CL303 and CL304 parent studies, it is reasonable to expect that the same strengths and limitations related to generalizability apply to the extension studies.

The patient population of those studies may not be reflective of the wider, more heterogeneous clinical population in terms of demographic and clinical characteristics; therefore, the results presented may differ from those observed in a real-world clinical setting. The study population was not reflective of the Canadian population and therefore the patients enrolled may not reflect the gender, racial, or ethnic diversity of the Canadian population, which may reduce the generalizability of results.

Indirect Comparisons

No indirect comparisons were submitted as part of this review.

Studies Addressing Gaps in the Evidence From the Systematic Review

Description of Studies

The Disease Monitoring Program (DMP) is a cohort study intended to last for up to 10 years per patient and enrol at least 500 adult and pediatric patients with XLH at up to 39 sites in the US, Canada, and Latin America.²⁹ Patients receiving burosumab in a real-world setting (i.e., outside of clinical trials), those enrolled in the DMP after receiving burosumab in a clinical trial setting, and those who were not receiving burosumab at all (i.e., receiving conventional therapy or no treatment) were included. An analysis of the year 1 data was submitted, consisting of data collected from 2 matched patient cohorts: patients who were reported to be receiving conventional therapy at baseline (July 16, 2018) and who never received burosumab during the DMP, and patients who reported receiving burosumab in a real-world setting and who initiated burosumab at any point after DMP initiation. Patients provided information on demographics, family history, diagnostic history, medical and surgical history, growth history, disease-specific clinical symptoms and progression, concomitant medications and therapies, disability, and quality of life.

Outcomes

The outcomes of interest were serum phosphate levels and WOMAC pain, WOMAC stiffness, and WOMAC physical function scores at the year 1 mark. Information on outcomes was collected at the baseline visit and again at the approximate year 1 visit.

Statistical Analysis

The 2 patient cohorts were balanced on baseline characteristics using propensity score (PS) matching algorithms that included the following: demographics (age, race, and gender), clinical characteristics (weight, height, body mass index, serum phosphate level, WOMAC pain score, WOMAC stiffness score, and WOMAC physical function score), disease and/or medical characteristics (PHEX mutation positivity, age at XLH diagnosis, number of historical fractures, osteoarthritis, and enthesopathy and/or bone spurs and/or osteophytes).

Mean changes to outcome variables between the baseline visit and the year 1 visit were calculated for the cohorts; changes in outcomes were only calculated for those patients who had a baseline and year 1 measure for that outcome. For continuous baseline variables, the F-test was performed to check for the equality of variance between the 2 cohorts, and equal or unequal variance student t test was used. For categorical baseline variables, a chi-square test was performed with a P value of 0.05 or less being considered statistically significant.

Efficacy Results

The matching procedure balanced cohorts with respect to race, weight at baseline, height at baseline, WOMAC pain score, and WOMAC stiffness score. A total of 44% of patients in the burosumab cohort reported receiving conventional therapy at baseline, and 56% of patients reported receiving no treatment. All patients in the conventional therapy cohort reported receiving conventional therapy. There was a mean delay of 245.8 (SD = 275.2) days in initiating burosumab in the burosumab cohort, and the year 1 visit for patients occurred an average of 408.8 (SD = 94.0) days after the baseline visit in the burosumab cohort and 431.3 (SD = 89.3) days in the conventional therapy cohort.

The proportion of patients in the burosumab cohort with serum phosphate levels greater than LLN was 20.0% at baseline and 58.3% at the year 1 visit; this attained statistical significance relative to the conventional therapy cohort (28.6% of patients had serum phosphate levels > LLN at year 1; P value = 0.0013). There was no significant difference between the 2 cohorts in terms of the change in WOMAC physical function, WOMAC pain, or WOMAC stiffness scores at the year 1 visit.

Harms Results

Information on harms was not provided for the DMP study.

Critical Appraisal

The design of the DMP study is subject to some notable limitations due to missing key information. It is unclear when the initiation of burosumab occurred in the burosumab cohort whereas the analysis appeared to consider the time between baseline and burosumab initiation as time spent on burosumab treatment. The treatment patterns of the cohort after baseline but before burosumab initiation are also not known. The dosing of all therapies during the study, conventional or burosumab, is largely unknown. While transparently discussed in the submission, this remains an important consideration as potential variations in real-world practice or differences in the degrees of patient adherence to therapy are unaccounted for in the assessment. There was no information provided on the recruitment methods of sites or patients; therefore, the study settings are largely unknown. There was also no information on which point in the dosing cycle (e.g., midpoint, end point) the serum phosphate results were measured. Since the pivotal trial demonstrated that there are notable variations in the proportion of patients with serum phosphorus levels greater than LLN at the end point versus the midpoint of the dosing cycle, this could greatly impact the definition of the interventions and render inference very uncertain. The results must also be interpreted in the context of no harms data having been reported, which is an important consideration as this leaves a considerable knowledge gap in understanding the full impact of burosumab treatment. Furthermore, the patients in the burosumab cohort comprised both patients who had been receiving conventional therapy at baseline and those who hadn’t been receiving any therapy — the magnitude of benefit due to burosumab treatment may vary within subgroups of patients based on their previous treatment patterns, which is not explored in sensitivity analyses in the cohort study. There is also no discussion of the methods used to identify the variables included in the PS matching. The matching itself did not attain balance on fractures (38.0% in the burosumab cohort versus 49.3% in the conventional therapy cohort) or the country variable, and as such any country-level differences in practice would not be controlled for in this analysis. There is also the possibility of selection bias as approximately half the patients entering the burosumab cohort had no treatment at baseline, and without treatment history it isn’t known whether these patients were refractory to conventional therapy or their disease activity levels were such that it was not needed.

There are also limitations on the generalizability of this cohort study. Less than a quarter of the participants were from Canada, and therefore results may not translate directly to the characteristics of this clinical population. In addition, with an average of 245.8 days until the first burosumab exposure and a mean duration between visits of 408.8 days, the burosumab cohort was treated for less time than was covered in the CL303 pivotal trial and BUR02 long-term extension (LTE), limiting the applicability of these results to longer time periods. Furthermore, similar to the CL303 trial, the cohort study used the same MCIDs and therefore the same limitations apply regarding the lack of an externally validated measure of clinical meaningfulness. Overall, the potential biases that may or may not be imparted by the presence of missing information greatly complicates the definition of the intervention and comparator, as well as any causal inference linking burosumab treatment to the observed results, rendering it difficult to draw conclusions regarding the relationship between burosumab treatment and patient outcomes in a real-world setting.

Conclusions

The major areas of the reassessment addressed the lack of clinically meaningful results in the domains of pain, physical function, and fatigue in adults with XLH, as well as a lack of active comparator data against conventional therapy for XLH. Additional data from the CL303 trial broadly showed normalization of serum phosphorus in a notable majority of patients that persisted in many patients over time, although a waning in the proportion of patients with serum phosphorus levels greater than LLN was observed at 96 weeks after enrolment. A trend toward increased healing in fractures or pseudofractures was also noted along with a statistically significant OR of full healing relative to no healing at all at 24 weeks, although longer-term data remained lacking. While potentially notable reductions in WOMAC scores, particularly stiffness scores, were reported and reductions maintained over longer time periods, there was a lack of notable impact noted in the BPI pain and fatigue scores, with reductions of at most 2 points from baseline. The meaningfulness of these changes remains unknown due to the fact that the MCIDs provided in the submission were derived from the same dataset as that of the pivotal trial and are thus hampered by a lack of external validity. Data from the safety assessment of burosumab noted no serious safety signals but a potentially cumulative impact of TEAEs, which was identified through an analysis adjusting for the duration of burosumab exposure; this is a potentially important consideration as treatment with burosumab will be lifelong, per the clinical expert consulted by CDA-AMC. The LTE study also reported an increase in vitamin D deficiency and hypophosphatemia at later time points, although the clinical impact of these results is unclear. The reassessment was not able to conclude anything about comparative evidence due to limitations in the real-world evidence portion and there remains an information gap on the safety of burosumab relative to conventional therapy.

Introduction

Disease Background

Content in this section has been informed by materials submitted by the sponsor and clinical expert input. The following has been summarized and validated by the CDA-AMC review team.

XLH is a rare, chronically debilitating genetic disorder.¹ It is characterized by renal phosphate wasting and consequent defective bone mineralization caused by inactivating mutations in PHEX.^2-4 In the absence of functional PHEX, patients with XLH produce excess FGF23, leading to the impaired conservation of phosphate and consequent hypophosphatemia.⁵ Excess FGF23 also suppresses 1,25(OH)₂D production, resulting in decreased intestinal absorption of calcium and phosphate.^1,6

Phosphorus plays a critical role in several essential biological processes, bone formation, and metabolism, and is an essential component of cell membranes and nucleic acids.^2,3 Patients usually develop clinical symptoms during the first or second year of life.⁶ XLH in children is characterized by vitamin D–resistant rickets, and results in variable degrees of delayed walking, a waddling gait, leg bowing, enlarged cartilages, bone pain, craniosynostosis, dental abscesses, and impaired growth.¹ Adults with XLH can have manifestations such as fractures and pseudofractures, and early-onset osteoarthritis and enthesopathies causing musculoskeletal pain, stiffness, and fatigue.^7-12 These musculoskeletal abnormalities in adults with XLH also result in impaired mobility and physical function, and reduced HRQoL.^7,13

As is the case with many rare diseases, published information about the incidence and prevalence of XLH is limited. The estimated prevalence of XLH in Norway is 1 case per 100,000 children.¹⁴ The estimated prevalence of hypophosphatemic rickets in southern Denmark is 4.8 cases per 100,000 people (children and adults),¹⁵ and 2.03 cases per 100,000 people in Colombia.¹⁶ A recent population-based cohort study using a large primary care database in the UK estimated adult XLH prevalence as being 1.57 cases per 100,000 people.¹⁷ There are no known reported prevalence estimates for Canada.

Standards of Therapy

Content in this section has been informed by materials submitted by the sponsor and clinical expert input. The following has been summarized and validated by the CDA-AMC review team.

Due to the multivariate manifestations of XLH, particularly in adulthood, care involves pharmacological and non-pharmacological management, and can include the involvement of many specialists (e.g., endocrinologists, rheumatologists, geneticists, orthopedic surgeons, dentists, nephrologists, pain specialists). The goals of therapy for XLH in adults are to normalize phosphate concentrations as well as to decrease the extent of osteomalacia and pseudofractures to reduce bone pain, enhance mobility, cure nonunion fractures, and/or improve fracture healing or surgical recovery. Reduction of these events increases patients’ ability to participate in life activities and improves HRQoL.

In Canada, conventional therapy for adults with XLH involves oral phosphate supplements and active vitamin D analogues (calcitriol or alfacalcidol).³⁰ Active vitamin D analogues are publicly funded for XLH, while phosphate supplementation is accessible as an over-the-counter product. The clinical expert consulted by CDA-AMC noted that current treatment generally reduces hypophosphatemia and low 1,25(OH)₂D levels, which are downstream effects of the elevated FGF23 levels; however, they do not reverse the course of disease. Furthermore, conventional therapy has a notable side effect burden. Per the clinical expert, frequent phosphate administration may produce gastrointestinal upset and secondary or tertiary hyperparathyroidism, and 1,25(OH)₂D treatment may produce hypercalciuria and nephrocalcinosis that may potentially lead to renal failure; these complications may require additional therapy. Furthermore, the clinical expert noted that although not all patients respond to available treatments, even in those who respond with the normalization of serum phosphate and 1,25(OH)₂D, this may further elevate FGF23 levels and set up an unwanted feedback loop that limits the efficacy of conventional treatment.

Drug Under Review

Burosumab is a recombinant human immunoglobulin G subclass 1 monoclonal antibody that binds to the N-terminal domain of FGF23. This inhibits the biological activity of FGF23, thereby increasing both renal phosphate reabsorption and the serum concentration of 1,25(OH)₂D.¹⁸ It is indicated for the treatment of XLH in adult and pediatric patients aged 6 months and older.¹⁸ The burosumab dosage regimen for the treatment of XLH in adult patients is 1 mg/kg rounded to the nearest 10 mg up to a maximum dose of 90 mg, administered by SC injection every 4 weeks.¹⁸

Burosumab was previously reviewed by CADTH and a recommendation to reimburse with conditions was issued in May 2020 for treatment of pediatric patients with XLH.¹⁹ The sponsor has submitted additional data and requested a reassessment of the reimbursement request for the treatment of adult patients with XLH, as this population is included within the indication approved by Health Canada. The sponsor is proposing the following criteria for adult patients with XLH, which they indicated align with international consensus guidelines.⁶

Burosumab treatment is proposed by the sponsor to be initiated in adult patients (≥ aged 18 years) who have:

• a clinical presentation consistent with XLH, including —

  • fasting hypophosphatemia, and

  • normal renal function (defined as fasting serum creatinine below the age-adjusted ULN)

• a confirmed PHEX gene variant in either the patient or in a directly related family member with appropriate X-linked inheritance and —

  • persistent bone and/or joint pain due to XLH, and/or

  • osteomalacia that limits daily activities, and/or

  • pseudofractures or osteomalacia-related fractures

• insufficient response or are refractory to conventional therapy or if patients experience complications related to conventional therapy.

Key characteristics of burosumab are summarized in Table 3 with other treatments available for XLH.

Table 3: Key Characteristics of Burosumab, Alfacalcidol, Calcitriol, Sodium Phosphate, and Cinacalcet Hydrochloride

HPT = hyperparathyroidism; PTH = parathyroid hormone; SC = subcutaneous; XLH = X-linked hypophosphatemia.

^aThe brand name of the tablets is Jamp-sodium phosphate effervescent.

^bHealth Canada–approved indication.

Sources: Product monographs of burosumab,¹⁸ alfacalcidol,³¹ calcitriol,³² sodium phosphates,³³ phosphate effervescent,^34,35 and cinacalcet hydrochloride.³⁶

Perspectives of Patients, Clinicians, and Drug Programs

Patient Group Input

This section was prepared by the CDA-AMC review team based on the input provided by patient groups. The full original patient input that CDA-AMC received has been included in the Perspectives of Patients, Clinicians, and Drug Programs section of this report.

Input was submitted for this review by the Canadian XLH Network, a national, not-for-profit, patient support organization for people living and dealing with XLH. Information for this input was gathered through an online survey of XLH adult patients and family and caregivers from December 2 to 15, 2023.

Of the 57 respondents to the survey, 46% live in Canada, 88% identify as female, and 88% are aged between 30 years and 59 years. The majority of the respondents were diagnosed as children, and 12% were diagnosed as adults. Respondents indicated that symptoms of XLH during adulthood differed from childhood symptoms. When asked about adult symptoms, 44% of patients reported severe pain, 28% reported a loss of mobility, 21% reported a lack of energy, 21% had an increase in dental issues, and 26% had developed arthritis and/or spinal stenosis. All of these symptoms were reported to significantly impact patients’ quality of life, and their social and psychological well-being.

Survey respondents indicated that while conventional treatment (a combination of phosphate and calcitriol) was available to adult patients in Canada, patients need to take large doses of phosphate up to 5 times daily and calcitriol 1 to 2 times daily, a treatment that addresses the issue of low phosphate but does not address pain and other serious symptoms of XLH. In addition, conventional treatment has serious side effects, such as nephrocalcinosis, kidney disease, calcium deposits, and parathyroid issues, all while allowing XLH to continue progressing. Furthermore, phosphate is very expensive and hard to access due to supply chain issues. To manage their pain, all patients surveyed use over-the-counter or prescription pain medication, and most of the respondents (88%) reported having had major surgeries and anticipating more in the future.

Respondents indicated that there is a need for treatment options that are accessible and affordable, are easier to take, can boost energy levels and muscle function, reduce pain, and improve bone health and overall quality of life, with fewer side effects.

Clinician Input

Input From Clinical Experts Consulted by CDA-AMC

All of the CDA-AMC review teams include at least 1 clinical specialist with expertise regarding 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, 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 XLH in adults.

Unmet Needs

Per the clinical expert, the goals of treatment in adults are to reduce osteomalacia and pseudofractures, cure any nonunion fractures to alleviate generalized bone pain, and enhance mobility that may have been reduced. Reduction of these events can increase the ability to maintain employment and improve HRQoL. It is often recommended that only symptomatic adults (i.e., those displaying symptoms of disease along with indicative laboratory markers and imaging studies) be treated.

The clinical expert noted that current treatment improves hypophosphatemia and low 1,25(OH)₂D levels, which are downstream effects of the elevated FGF23 levels that occur in this disease. However, the treatment, by normalizing serum phosphate and 1,25(OH)₂D, may further elevate FGF23 levels to cause a feedback loop that limits the efficacy of conventional treatment. There is also a side effect burden to conventional therapy, including gastrointestinal upset and secondary hyperparathyroidism due to oral phosphate, as well as hypercalciuria and nephrocalcinosis leading to decreased renal function due to 1,25(OH)₂D treatment. Thiazide diuretics and cinacalcet may be added to conventional phosphate and active vitamin D therapy to treat complications of conventional therapy, such as hypercalciuria and hyperparathyroidism, respectively. In addition, the clinical expert noted that the majority of patients (> 70%) continue to have symptoms of pain, mobility issues, or complications despite treatment. Since active vitamin D may need to be administered twice daily and oral phosphate is usually administered several times per day, adherence may not be optimal.

Place in Therapy

Per the clinical expert consulted by CDA-AMC, burosumab would represent a shift in the current treatment paradigm as it addresses the underlying disease at an upstream level rather than a downstream level. Since burosumab inhibits the action of FGF23, renal TRP would be increased and serum phosphorus levels elevated into the normal range; therefore, oral phosphate therapy should no longer be required. With the reduction in FGF23 action, 1,25(OH)₂D synthesis can increase and supplementation should also no longer be required.

The clinical expert noted that mechanistically, burosumab should be used as a first-line treatment by itself and not in combination with other treatments — namely, sodium phosphate and 1,25(OH)₂D. For pragmatic reasons, a trial with conventional therapy could be initiated and then burosumab substituted if side effects persist. The expert noted that treatment with burosumab is likely to be lifelong as the cause of the disease remains. It is possible for burosumab to stop working, but this would most likely be due to adherence issues rather than the therapy itself.

Patient Population

The clinical expert noted that symptomatic patients with bone pain due to bone disease (i.e., due to osteomalacia, pseudofractures, and nonunion fractures) are best suited for treatment. Likewise, treatment should likely not be considered in adults who are asymptomatic unless they have asymptomatic fractures detected by radiography. In the expert’s experience, most patients they have met with tended to be symptomatic, although as a rare disease, the overall group of patients is small.

However, the clinical expert also noted that there may be benefit in treating adults with limited symptomatology to increase activity levels and a sense of well-being. In addition, patients with XLH scheduled for an elective orthopedic surgical procedure such as joint replacement might benefit from a course of therapy (either burosumab or phosphate and calcitriol) for 6 months before the procedure to ensure optimal healing of the bone and the secure placement of hardware.

Assessing the Response to Treatment

In clinical practice, a reduction in bone pain, a reduction in fractures, and the healing of fractures would be considered by the clinical expert consulted by CDA-AMC to be clinically meaningful responses to therapy, particularly if they were accompanied by improved motor ability, reduced stiffness, and the improved ability to perform activities of daily living. Laboratory evidence of the normalization of serum phosphate and biomarkers of bone metabolism (e.g., alkaline phosphatase) and the absence of elevations in serum creatinine or PTH as well as the absence of the development or acceleration of nephrocalcinosis would also be important.

Per the clinical expert, several outcomes measured in clinical trials are not routinely used in clinical practice — specifically, standardized measures of osteoarthritis symptoms to determine symptoms and the 6MWT.

Discontinuing Treatment

Per the clinical expert, patients who are experiencing a sustained decline in serum phosphorus levels despite adherence to therapy (suggesting that burosumab treatment is not working) or who develop a severe allergic reaction to burosumab should discontinue therapy. Therapy should be continued if initiated during childhood since the consequences of elevated FGF23 can also be seen in adults. The expert noted that the discontinuation criteria provided in the CADTH recommendation for the pediatric indication (hyperparathyroidism, nephrocalcinosis, or evidence of fracture or pseudofracture based on radiographic assessment)¹⁹ would also be reasonable in adults.

Prescribing Considerations

The expert noted that specialist care would be required to diagnose, treat, and monitor patients receiving burosumab — either an endocrinologist or rheumatologist with knowledge of the disorder. The clinical expert acknowledged that patients could also likely be taught to self-administer burosumab.

Clinician Group Input

This section was prepared by the CDA-AMC review team based on the input provided by clinician groups. The full original clinician group input that CDA-AMC received has been included in the Perspectives of Patients, Clinicians, and Drug Programs section of this report.

No input was received by clinician groups by the deadline of the call for input.

Drug Program Input

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

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

BPI = Brief Pain Inventory; LLN = lower limit of normal; RSS = Rickets Severity Scale; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate; XLH = X-linked hypophosphatemia.

Clinical Evidence

The objective of the CDA-AMC Clinical Review Report is to review and critically appraise the clinical evidence submitted by the sponsor on the beneficial and harmful effects of burosumab 10 mg/mL, 20 mg/mL, and 30 mg/mL for SC administration in the treatment of XLH in adults. The focus has been placed on comparing burosumab to relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the reassessment of burosumab for adults with XLH is presented in 3 sections with the CDA-AMC critical appraisal of the evidence included at the end of each section. The first section, the systematic review, includes pivotal studies and RCTs that were selected according to the sponsor’s systematic review protocol. The second section includes a sponsor-submitted LTE study. The third section includes an additional study that was considered by the sponsor to address important gaps in the systematic review evidence.

Included Studies

Clinical evidence from the following are included in the CDA-AMC review and appraised in this document:

• One pivotal RCT (the CL303 study) identified in the systematic review

• One LTE study

• One additional study addressing gaps in the evidence.

Systematic Review

Content in this section has been informed by materials submitted by the sponsor. The following has been summarized and validated by the CDA-AMC review team.

Description of Studies

Characteristics of the included studies are summarized in Table 5.

Table 5: Details of Studies Included in the Systematic Review

1,25(OH)₂D = 1,25-dihydroxyvitamin D; 6MWT = 6-minute walk test; BALP = bone-specific alkaline phosphatase; BFI = Brief Fatigue Inventory; BPI = Brief Pain Inventory; GFR = glomerular filtration rate; LLN = lower limit of normal; LTE = long-term extension; PTH = parathyroid hormone; RCT = randomized controlled trial; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate; TRP = tubular reabsorption of phosphate; ULN = upper limit of normal; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; XLH = X-linked hypophosphatemia.

Sources: Sponsor’s Summary of Clinical Evidence²³ and the CL303 Clinical Study Report.²⁰

The CL303 study²⁰ was a randomized, double-blind, placebo-controlled, multicentre phase III study evaluating the safety and efficacy of burosumab in adult patients with XLH. A total of 163 patients were screened; 134 patients were enrolled in the study and randomized in a 1:1 ratio to receive burosumab or placebo. The most common reason for not advancing past screening was that patients did not meet the minimum pain level in the inclusion criteria (N = 10 patients). Randomization was stratified based on region (North America, the European Union, Japan, and South Korea) and was also planned to be stratified on pain intensity using the BPI worst pain scores (less than or equal to 6 or greater than 6 at the first screening visit). Due to stratification misclassification, 97 patients and 37 patients were randomized according to their BPI average pain scores of less than or equal to 6 and greater than 6, respectively. A total of 44 patients and 90 patients were randomized per the planned stratification of BPI worst pain less than or equal to 6 and greater than 6, respectively. The submission investigated the consequences of the misclassification and due to the high correlation between worst pain and average pain scores, it judged the impact to be minimal; analyses were done using the actual randomization strata. There were no study sites in Canada.

Patients randomized to burosumab received 1 mg/kg (rounded to the nearest 10 mg) or placebo administered subcutaneously every 4 weeks (28 days) for 24 weeks. After completing the week 24 visit, patients randomized to placebo crossed over into a treatment continuation phase to receive open-label burosumab 1 mg/kg (rounded to the nearest 10 mg) subcutaneously every 4 weeks (28 days) until week 48. During the placebo-controlled period, patients had clinic and/or home health visits at 2-week intervals beginning at baseline, plus visits at week 1 and week 21; the window for visits was plus or minus 3 days. During the treatment continuation and extension periods, patients had clinic and/or home visits at 2-week or 4-week intervals, plus or minus 5 days.

All patients and investigators remained blinded to the original double-blind treatment assignments until week 48 analyses were completed. All patients continued burosumab treatment on the same dosing regimen in the treatment continuation and extension phases. Analyses up until week 96 were included in the CL303 study as part of the systematic review; results from the CL303 study participants who continued, plus participants from another study, CL304, who continued up until treatment extension II were reported as part of the LTE study section. Full details of the overall study design are in Figure 1.

Figure 1: Study Design for CL303 Trial

BPI = Brief Pain Inventory; FU = follow-up; P = phosphorus; TC = telephone call; XLH = X-linked hypophosphatemia.

Source: Sponsor’s Summary of Clinical Evidence.²³

Populations

Inclusion and Exclusion Criteria

The study population consisted of adult patients aged between 18 years and 65 years with a minimum threshold of bone or joint pain at baseline and a diagnosis of XLH, defined as clinical and biochemical features consistent with XLH and/or a confirmed PHEX mutation (self or a family member consistent with X-linked inheritance). If patients were receiving chronic pain medications, they had to have been on a stable regimen for at least 21 days before screening and be willing to maintain medications at the same stable dose (a maximum of 60 mg oral morphine equivalents per day). There were no specific inclusion criteria for previous treatment patterns before study enrolment, but patients had to discontinue any conventional therapies between the screening visit and randomization (a 2-week washout period). Per the exclusion criteria, patients would have had to have an absence of, or control over, calcium and parathyroid-related complications, as well as no recent treatments that could impact bone metabolism and no planned surgeries.

Interventions

Both placebo and burosumab solutions had the same composition, apart from the presence of the active ingredient, and were supplied as a sterile, clear, colourless, preservative-free solution in single-use, 5 mL vials. The burosumab vials contained 1 mL of burosumab at a concentration of 30 mg/mL, and the placebo vials contained 1 mL of placebo. Both interventions were prepared and administered in the same manner; trained personnel at study sites administered the study drug by SC injection every 28 days into the abdomen, upper arms, or thighs, rotating injection sites with each administration. A maximum of 1.5 mL was administered in any 1 site, and if a larger volume were required, it was administered in multiple sites.

For patients randomized to burosumab, the dose was calculated based on baseline body weight up to a maximum dose of 90 mg, and remained fixed for the duration of the study provided that serum phosphorus levels did not exceed 5.0 mg/dL at any time or 4.5 mL/dL on 2 occasions, and body weight did not change by more than 20%. If serum phosphorus increased to more than 5.0 mg/dL at any time, the patient treatment assignment was unblinded to the investigator (if this occurred during the placebo-controlled treatment period) and the dose of burosumab was decreased by half. If serum phosphorus increased above the ULN (4.5 mg/dL) but did not exceed 5.0 mg/dL, the patient treatment assignment was unblinded (if this occurred during the placebo-controlled treatment period) and the dose of burosumab decreased by half only if a second phosphorus result exceeded the ULN. Following a downward dose adjustment, the investigator, together with the sponsor’s medical monitor, determined if, when, and how to titrate the dose upward.

If at any time during the study a patient did not receive a dose within 21 days of a scheduled dose, that dose was skipped and the next dose administered at the next scheduled dosing visit. Doses were administered no fewer than 14 days apart.

Throughout the study, a patient's diet or medication schedule was not to change significantly unless medically indicated. Investigators could prescribe any concomitant medications or treatments deemed necessary to provide adequate supportive care except for those specified as prohibited medications: pharmacologic vitamin D metabolites or analogues, oral phosphate, aluminum hydroxide antacids, acetazolamides, thiazides, bisphosphonate therapy, denosumab therapy, teriparatide therapy, chronic use of systemic corticosteroids, PTH suppressors, and any other monoclonal antibody therapy. Oral supplementation of 1,25(OH)₂D could be provided if serum levels decreased to less than 20 ng/mL. Nonsteroidal anti-inflammatory drugs, opioids, or other narcotic pain medications were permitted, being limited to 60 mg oral morphine equivalents per day. Pain medication use (prescription and over the counter) was recorded by patients in a pain medication diary for 7 consecutive days before the baseline visit and visits at week 12, week 24, week 36, and week 48. Concomitant therapies and medications were reviewed and recorded in the patient electronic Clinical Report Form at each study visit and during home health visits, as applicable. Patients were required to maintain any chronic pain medication(s) at a stable dose(s) and schedule through the placebo-controlled period (up to week 24); this requirement did not apply for subsequent treatment periods.

Outcomes

A list of efficacy end points assessed in this Clinical Review Report is provided in Table 6, followed by descriptions of the outcome measures. Summarized end points are based on outcomes included in the sponsor’s Summary of Clinical Evidence²³ as well as any outcomes identified as important to this review according to the clinical expert consulted by CDA-AMC and input from patient and clinician groups and public drug plans. Using the same considerations, the CDA-AMC review team select end points that were deemed to be most relevant to inform the CDA-AMC expert committee deliberations and finalized this list of end points in consultation with members of the expert committee.

The proportion of patients attaining serum phosphorus levels above the LLN (0.81 mmol/L) at the midpoint of the dosing cycle (i.e., when serum levels of burosumab are at their peak) was the primary outcome of the study. The clinical expert consulted by CDA-AMC confirmed that this outcome is a surrogate for clinical outcomes in XLH and an efficacy end point as burosumab’s mechanism of action equalizes serum phosphorus.

Key secondary end points and secondary end points were PRO measures of pain, fatigue, stiffness, and physical function. The key secondary end points were measured at 24 weeks and included change in the BPI worst pain score, and WOMAC stiffness and WOMAC physical function scores. Other secondary end points included domains of the BPI, WOMAC, and BFI. The clinical expert consulted by CDA-AMC noted that these PRO measures are not commonly used in routine clinical practice. PROs at each visit were summarized, analyzed, and reported at week 24, week 48, and week 96 as secondary outcomes.

The proportion of patients attaining serum phosphorus levels greater than the LLN at the end of their dosing cycle (i.e., 4 weeks after dosing) was also a secondary end point as were measures of bone metabolism (BALP), serum 1,25(OH)₂D, and phosphorus homeostasis (TmP/GFR and TRP). The clinical expert consulted by CDA-AMC noted that these serum biomarkers are important outcomes to measure in XLH management.

Exploratory end points were active pseudofractures and/or fractures, as well as the 6MWT. The clinical expert noted that pseudofractures and fractures were important efficacy outcomes, but that the 6MWT was not routinely used in clinical practice.

Table 6: Efficacy Outcomes Summarized From CL303 Study

1,25(OH)₂D = 1,25-dihydroxyvitamin D; 6MWT = 6-minute walk test; BALP = bone-specific alkaline phosphatase; BFI = Brief Fatigue Inventory; BPI = Brief Pain Inventory; LLN = lower limit of normal; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate; TRP = tubular reabsorption of phosphate; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

^aStatistical testing for these end points was adjusted for multiple comparisons (e.g., hierarchical testing) at the 24-week mark of the trial.

Source: Sponsor’s Summary of Clinical Evidence.²³

Detailed Outcome Descriptions

A summary of the PROs measured in the study can be found in Table 7, followed by detailed descriptions of all outcomes. Briefly, the submission supplied MCIDs for the BPI and WOMAC scores, which were validated by the sponsor using data from 3 studies in patients with XLH as a source: the UX023-CL001 study (a cross-sectional survey of adults with XLH [N = 201 patients]),²⁵ the UX023-CL203 study (a phase II trial in adults with XLH [N = 20 patients]),²⁵ and the CL303 study (the pivotal trial in this submission [N = 134 patients]).²⁰ MCIDs for the BFI were also provided by the sponsor using the CL303 trial alone as a data source.²⁵ The MCID for the 6MWT came from assessment in a hypophosphatasia population and was not provided by the sponsor.

Table 7: Summary of PRO Measures and Their Measurement Properties

6MWT = 6-minute walk test; BFI = Brief Fatigue Inventory; BPI = Brief Pain Inventory; MCID = minimal clinically important difference; PRO = patient-reported outcome; VAS = visual analogue scale; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

Proportion of Patients With Serum Phosphorus Levels Greater Than LLN

The level of serum phosphorus was selected as an outcome based on the mechanism of action of burosumab inhibiting FGF23, which plays a role in regulating serum phosphorus. Per the clinical expert, the proportion of patients attaining serum phosphorus levels greater than LLN at the midpoint of the dosing cycle determines the efficacy when the burosumab dose effect is maximal, and the proportion of patients attaining serum phosphorus levels greater than LLN at the end of the dosing cycle determines efficacy when the burosumab dose effect is at its minimum level due to its biological half-life.

Serum phosphorus was analyzed by the local laboratory at screening visit 1 and screening visit 2 for eligibility, and by the central laboratory at all other indicated time points.

Brief Pain Inventory

The BPI is a self-reported questionnaire designed to provide information about pain intensity (the sensory dimension, 4 items) and the degree to which pain interferes with daily living (the reactive dimension, 7 items). It is recommended by the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials as a core outcome measure of pain.⁴² Four items assess a patient’s pain intensity — at its worst intensity in the last 24 hours, at its least intensity in the last 24 hours, average pain, and pain right now — using a 0 to 10 numeric rating scale, with 0 representing “no pain” and 10 representing “pain as bad as you can imagine.” For the 7 items assessing pain interference with functioning, patients are asked to rate how their pain interferes with 7 life domains, including general activity, mood, walking ability, normal work, relations with others, sleep, and enjoyment of life, on a similar numeric rating scale. The anchor points in each item of the interference scale are 0 (no interference) and 10 (complete interference). The scores for the 2 BPI subscales (pain intensity and pain interference) range from 0 to 10 and are calculated using the means of their corresponding items’ scores.

The total BPI score is the mean of the 2 subscale scores. A high score represents a high pain intensity or pain interference. Although originally developed for the evaluation of breast, prostate, colon, rectum, or gynecologic cancer pain, it has also been shown to be a reliable and valid instrument for the evaluation of nonmalignant chronic pain (e.g., low back pain, osteoarthritis, rheumatoid arthritis, multiple sclerosis) across various languages. It is also commonly used for nonmalignant pain.^43,44

Patients recorded responses to the 4 questions of the BPI pain severity dimension (worst, least, average, and now) in a paper diary for the week preceding the baseline and week 12, week 24, week 36, and week 48 visits. The modified short-form BPI was administered at baseline and at week 12, week 24, week 36, week 48, week 72, and week 96.

Western Ontario and McMaster Universities Osteoarthritis Index

WOMAC is a self-administered questionnaire assessing pain, stiffness, and physical functioning in patients with hip and knee osteoarthritis. It is a valid, reliable, and responsive measure of outcome in knee osteoarthritis^46,50,51 and has been widely used in other painful musculoskeletal disorders, such as lower back pain, rheumatoid arthritis, systemic lupus erythematosus, and fibromyalgia.¹² WOMAC consists of 24 items divided into 3 subscales:⁴⁷

• pain (5 items) — during walking, using stairs, in bed, sitting or lying, and standing upright

• stiffness (2 items) — after first waking and later in the day

• physical function (17 items) — using stairs, rising from sitting, standing, bending, walking, getting in and out of a car, shopping, putting on and taking off socks, rising from bed, lying in bed, getting into and out of a bath, sitting, getting on and off a toilet, heavy domestic duties, and light domestic duties.

There are 2 scale formats for WOMAC: a 10 cm visual analogue scale and a 5-point Likert scale. The 2 formats were found to be highly correlated and to yield similar precision for discriminating treatments in patients with osteoarthritis.⁴⁸ The Likert version of WOMAC was used in the CL303 study. It is rated on an ordinal scale of 0 to 4, where 0 means the lowest level of symptoms or physical disability. Each subscale is summated to a maximum score of 20, 8, and 68 for pain, stiffness, and physical function, respectively, providing a maximum global score of 96 (the sum of the 3 subscales).⁵² Higher scores on WOMAC indicate worse pain, stiffness, and functional limitations.

One study suggests that the WOMAC tool has acceptable face and content validity in the adult XLH population.¹² Another study that enrolled adult patients with XLH also tested the scale discriminant validity and convergent-divergent validity and supported the use of WOMAC in this study population.⁵²

The WOMAC tool was administered at baseline and week 12, week 24, week 36, week 48, week 72, and week 96.

Brief Fatigue Inventory

The BFI is a self-reported questionnaire to assess the severity of fatigue and the impact of fatigue on daily functioning. Two dimensions are measured in this 9-item instrument: fatigue (3 items) and the interference of fatigue on daily life (6 items pertaining to general activity, mood, walking ability, normal work, relations with others, and enjoyment of life). The items are measured on a 0 to 10 numeric rating scale. For the dimension of severity of fatigue, 0 represents “no fatigue” and 10 represents “fatigue as bad as you can imagine.” For the dimension of interference from fatigue, 0 represents “does not interfere” and 10 represents “completely interferes.” A score of 7 to 10 is considered severe fatigue.²² A global fatigue score can be obtained by averaging all the items on the BFI.⁵⁴

The BFI has been validated and used in patients with various conditions, including cancer, osteoarthritis, and rheumatoid arthritis.^20,22 The construct validity, concurrent validity, and discriminant validity of BFI have been demonstrated in cancer patients. Reliability of the BFI was assessed based on 1 study that included 305 adult patients with cancer (coefficient alphas were 0.95 to 0.96).²²

Patients recorded responses to the 3 questions of the BFI fatigue dimension (now, as usual, and at its worst) in a paper diary for the week preceding the baseline visit and the week 12, week 24, week 36, and week 48 visits. The complete BFI was administered at baseline and at week 12, week 24, week 36, week 48, week 72, and week 96.

Change in 1,25(OH)₂D, Serum Phosphorus, Urinary Phosphorus, TmP/GFR, and TRP

Samples for these serum and urinary markers were generally collected following a minimum overnight fasting time of 8 hours and before study drug administration (if applicable); 2-hour fasting urine samples were used for select measurements. Local laboratory analyses were performed at screening visit 1 and screening visit 2, and all other analyses were performed at a central laboratory. Two-hour fasting urine collection was required to calculate TmP/GFR and TRP using simultaneously measured urine and blood creatinine and phosphorus concentrations. Both 2-hour and 24-hour urine were used for measurements of urinary phosphorus, creatinine, and calcium.

Change in BALP

Samples for serum and urinary markers were generally collected following a minimum overnight fasting time of 8 hours and before study drug administration (if applicable). Local laboratory analyses were performed at screening visit 1 and screening visit 2, and all other analyses were performed at a central laboratory.

6-Minute Walk Test

The 6MWT is a supervised test that measures the distance a patient can walk on a hard, flat surface over a 6-minute period.⁵⁶ A specific protocol outlining training, level of support provided to the patient, and standardization of distance available for the patient to walk (30 m) is provided by the American Thoracic Society.⁵⁶

The 6MWT has been used and validated in multiple adult patient populations with cardiopulmonary conditions (e.g., heart failure, chronic obstructive pulmonary disease, pulmonary hypertension).⁵⁶ Multiple studies have also established a proposed MCID in these populations. Reported distances associated with a noticeable functional improvement range from 54 m in patients with stable chronic obstructive pulmonary disease and 43 m in patients with heart failure.⁵⁶ Initial improvements in 6MWT results should be interpreted with caution, given that there has been a well-documented learning effect in patients previously unfamiliar with the test.⁵⁸ Motivation, encouragement, and co-operation can have a significant positive impact on the results, and the magnitude of these effects could be comparable to the effect of interventions.^59,60 This could be of special concern in situations where blinding is not present or is compromised, particularly when no comparator arm is available. Age, height, and weight of children can impact the results, although this may be less of a concern for adult patients who are not actively growing.

The reliability and validity of the 6MWT were evaluated in 24 patients with hypophosphatasia. The test-retest reliability was high for children, adolescents, and adults (i.e., Pearson correlation coefficients ranged from 0.81 to 0.95).⁵⁷ MCIDs for patients with hypophosphatasia were estimated using distribution-based methods (31 m for children and adults and 43 m for adolescents).⁵⁷

The 6MWT was administered at study visit 1 for practice purposes to minimize training effects. For patients unable to complete the 6MWT at screening, the test was not performed during the study. The 6MWT was also administered at baseline and at week 12, week 24, week 36, week 48, week 72, and, if applicable, every 24 weeks during the treatment extension period, and at the end of the LTE phase. Assistive devices could be used, with any use noted on the clinical report form.

Active Pseudofractures and/or Fractures

A radiographic skeletal survey was conducted at the baseline visit to determine the number and subsequent healing or resolution of current pseudofractures and fractures, as well as progression of enthesopathy. Standard radiographs were obtained of the chest, lateral spine, right and left hand and wrist, right and left humerus, right and left radius and ulna, right and left femur and pelvis, right and left tibia and fibula, and right and left foot. Targeted radiography at locations predetermined by the skeletal survey as areas with active pseudofractures or fractures was performed starting at week 12, week 24, week 36, and week 48 to monitor the frequency and healing of fractures and/or pseudofractures. During the treatment extension periods (week 48 to week 96, and the LTE phase), targeted X-rays were only performed at clinic visits if indicated to follow healing of any newly diagnosed fractures.

Postbaseline radiographs were compared with baseline radiographs using a predefined list of abnormalities by a trained central reader who was blinded to treatment assignment. Existing and new pseudofractures and fractures were graded as either healed, partially healed, or fully healed.

Statistical Analysis

A summary of the analysis methods, sensitivity analyses, and missing data methodology for the outcomes in the CL303 study can be found in Table 8, followed by a detailed description of the statistical analysis.

Table 8: Statistical Analysis of Efficacy End Points From CL303 Study

6MWT = 6-minute walk test; BOCF = baseline observation carried forward; BPI = Brief Pain Inventory; CMH = Cochran-Mantel-Haenszel; GEE = generalized estimating equation; LLN = lower limit of normal; LOCF = last observation carried forward; mBOCF = modified baseline observation carried forward; NR = not reported; vs. = versus; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

^aGEE estimates are from a GEE model that includes the change from baseline for the 6MWT end point as the dependent variable and region, visit, treatment, actual randomization stratification, and treatment-by-visit as fixed factors, the baseline of each 6MWT as a covariate, and compound symmetry covariance structure.

Sources: Sponsor’s Summary of Clinical Evidence²³ and the CL303 Clinical Study Report.²⁰

Statistical Tests and Models

The primary outcome was analyzed using a Cochran-Mantel-Haenszel test to compare the percentages of patients in the burosumab and placebo groups who attained a serum phosphorus level above the LLN (0.81 mmol/L) at the midpoint of the dosing interval, as averaged across dose cycles between baseline and week 24, controlling for BPI average pain score and geographic region. The primary end point was tested at the 2-sided alpha level of 0.05. Descriptive analyses were performed for the changes to week 48 and week 96.

Key secondary outcomes and secondary outcomes were analyzed at a 2-sided alpha level of 0.05 using generalized estimating equation (GEE) linear regression models. They included treatment, BPI average pain randomization stratification factors (except for the BPI worst pain outcome), region, visit, and interaction of treatment-by-visit as fixed factors, and was adjusted for baseline measurement. The models employed a compound symmetry covariance structure, which specified constant variance for assessments and constant covariance between assessments over time. In general, the P value for testing the statistical significance between the 2 treatment groups was provided for the change from baseline to week 24. Type III tests for the LSMs were used for the statistical comparison and 95% CIs also were reported.

To control the familywise error rate at the 0.05 level, a hierarchical testing procedure was used for the primary and key secondary efficacy end points. First, if the primary efficacy end point was statistically significant (i.e., the burosumab group was superior to the placebo group [P value < 0.05 by a 2-sided test]), then the 3 key secondary end points were tested as a group at the 0.05 level. The Hochberg adjustment was applied for multiple testing for the 3 end points, ordering the nominal P values for the end points from largest to smallest to determine the significance level at which they were tested. The largest P value was tested at a significance level of 0.05, the second-largest P value was tested at a significance level of 0.025, and the smallest P value was tested at a significance level of 0.0167. The full description of the testing procedure is shown in Figure 2.

Descriptive analyses were performed for the changes to week 48 and week 96. At week 48, in addition to the planned analysis, an ad hoc analysis using a GEE model similar to that of the key secondary and secondary outcomes was performed for the change from week 24 to week 48 in BPI worst pain, WOMAC stiffness, and WOMAC physical function scores in the placebo to burosumab group, with the LSM difference, 95% CI, and P value reported.

Figure 2: Flow Chart for the Multiplicity Adjustment in CL303 Study

Alpha = alpha; BPI = Brief Pain Inventory; N = no; signif. = significant; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; Y = yes.

Source: Sponsor’s Summary of Clinical Evidence.²³

The exploratory outcomes were summarized via descriptive statistics using methods similar to the descriptive statistical methods applied for the primary, key secondary, and secondary outcomes after week 24. In addition to the planned analysis, fracture and pseudofracture grades were analyzed using a hierarchical generalized linear mixed proportional odds model for binary responses, with the dependent variable being healed, partially healed, unchanged, or worse. The model included treatment, visit, treatment-by-visit and fracture type as fixed factors as well as random intercepts for patients and fractures nested within patients. Estimates of the probabilities for fracture grade occurrence, ORs for healed fractures at each visit, and 95% CIs were reported.

All efficacy analyses were conducted per protocol.

Missing Data and Sensitivity Analyses

For all outcomes, there was no imputation or accounting for missingness in the data in the primary analyses. Sensitivity analyses were carried out for the key secondary outcomes and consisted broadly of carrying forward select observations to the corresponding end points. In the last observation carried forward (LOCF) analysis, the last nonmissing postbaseline observation before discontinuation was carried forward to the corresponding end point for evaluation. In the baseline observation carried forward analysis, the nonmissing baseline observation was carried forward to the corresponding end point for evaluation. The modified baseline observation carried forward analysis proceeded as follows: for discontinuations due to AEs or death, the worst of baseline observation carried forward and LOCF was used; otherwise, LOCF was used. The results of sensitivity analyses were not appraised in the submission reassessment as they applied only to key secondary outcomes.

Sample Size and Power Calculation

A sample size of 60 patients per group (total sample size of 120 patients) was planned to provide 95% statistical power to detect a 50% difference between the burosumab and placebo groups in the percentages of patients attaining mean serum phosphorus levels above the LLN at the midpoint and end-of-dose intervals between baseline and week 24, at the 2-sided significance level of 0.05. The determination of the sample size for this study was based on the assumption that the percentage of patients who attained mean serum phosphorus levels in the normal range at the midpoint of the dose interval from baseline to week 24 would be 60% and 10% in the burosumab and placebo groups, respectively. The submission did not specify any literature or clinical source as the basis for using a 50% difference between treatment arms and did not specify any expected SD or losses to follow-up.

With a total sample size of 120 patients, the CL303 study design also had greater than or equal to 80% power to detect a mean difference of 1.0 in change from baseline between the burosumab and placebo groups in BPI worst pain, assuming a mean change from baseline of 2.0 in the burosumab group and of 1.0 in the placebo group, a common SD of 1.8, and a 10% dropout rate. The submission did not specify any literature or clinical source for the difference between treatment arms.

Subgroup Analyses

Subgroup analyses by baseline BPI worst pain (> 6 or ≤ 6), an actual stratification factor based on BPI average pain (> 6 or ≤ 6), and region (North America, the European Union, Japan, and South Korea) were performed separately for select efficacy end points. Select AE summaries were presented by region as well. Subgroup analyses by sex and race were also conducted separately for select efficacy end points. If the number of patients within a particular race category was too small, alternative categorizations of race were considered (e.g., white versus non-white [from original source]). Subgroup results were not appraised by CDA-AMC.

Analysis Populations

The efficacy analyses were conducted on the primary analysis set, which included all randomized patients who received at least 1 dose of the study drug during the placebo-controlled period; patients were analyzed per protocol.

The safety analysis set consisted of all randomized patients who received at least 1 dose of the study drug. Patients were analyzed according to the actual treatment received.

The pharmacokinetic analysis set consisted of all patients in the safety analysis set who had 1 or more evaluable burosumab concentration.

Table 9: Analysis Populations of CL303 Study

CDA-AMC = Canada’s Drug Agency.

Source: Sponsor’s Summary of Clinical Evidence.²³

Results

Patient Disposition

A total of 163 patients were screened for inclusion in the trial; 29 patients did not advance past screening and 134 patients were randomized into the treatment arms (66 patients in the placebo arm and 68 patients in the burosumab arm) — more than the minimum numbers required by the power calculations. The most common reason for not advancing past screening was the inclusion criteria of worst pain score of 4 or more. After randomization, discontinuations were similar between the study arms. Full details of patient disposition are in Table 10.

Table 10: Summary of Patient Disposition From CL303 Study

BPI = Brief Pain Inventory; PAS = primary analysis set; PTH = parathyroid hormone; ULN = upper limit of normal.

Source: Sponsor’s Summary of Clinical Evidence.²³

Of note, an error in the stratification plan for the randomization led to a number of patients being stratified according to their BPI average pain domain scores as opposed to their BPI worst pain domain scores. This resulted in 34.8% of patients in the placebo arm being stratified into the “6 or less” strata as opposed to the “greater than 6” strata, which they would have been if stratified as planned. In the burosumab arm, 45.6% of patients were stratified into the “6 or less” strata as opposed to the “greater than 6” strata, which they would have been if stratified as planned.

Table 11: Randomization Stratification Categories (PAS)

BPI = Brief Pain Inventory; PAS = primary analysis set.

^aPer the study protocol, stratification was planned according to BPI worst pain. However, a programming issue resulted in randomization actually being stratified according to BPI average pain. The baseline values for both pain scores were defined as the mean of the values from the baseline visit and the previous 7 days of diary scores.

Sources: Sponsor’s Summary of Clinical Evidence²³ and the CL303 Clinical Study Report.²⁰

Baseline Characteristics

Overall, the patient population encompassed by these baseline characteristics was majority female (65.2% female and 34.8% male in the placebo arm and 64.7% female and 35.3% male in the burosumab arm) and majority white (80.3% white, 13.6% of Asian descent, 4.5% Black or African American, and 1.5% other in the placebo arm; 80.9% white, 17.6% of Asian descent, and 1.5% other in the burosumab arm). Baseline characteristics were generally balanced between the 2 treatment arms.

In terms of medical history, a numerically higher proportion of patients in the burosumab arm had osteoarthritis (69.1% versus 57.6% of patients in the placebo arm). Similarly, a numerically higher proportion of patients in the burosumab arm were classified as having a BPI average pain score of more than 6 (32.4% versus 25.8% of patients in the placebo arm); this was similar for the BPI worst pain score where the proportion of patients with a score of more than 6 was 77.9% in the burosumab arm and 65.2% in the placebo arm. A numerically higher proportion of patients in the burosumab arm had nephrocalcinosis than in the placebo arm (16.2% versus 7.6% of patients, respectively). The majority of patients in the burosumab and placebo arms (86.8% and 93.9% of patients, respectively) had received both vitamin D analogues and phosphate before the trial. There were no notable imbalances in baseline laboratory characteristics. A higher proportion of patients in the placebo arm had active pseudofractures at baseline (51.5%) than patients in the burosumab arm (42.6%). The majority of patients in both arms had had previous orthopedic surgery (71.2% of patients in the placebo arm and 66.2% of patients in the burosumab arm) or were taking nonopioid pain medications at baseline (66.7% of patients in the placebo arm and 69.1% of patients in the burosumab arm).

Table 12: Summary of Baseline Characteristics From Studies Included in the Systematic Review (PAS)

1,25(OH)₂D = 1,25-dihydroxyvitamin D; BPI = Brief Pain Inventory; PAS = primary analysis set; SD = standard deviation; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate; XLH = X-linked hypophosphatemia.

^aMean of the baseline visit score and diary scores from the 7 days before the baseline visit.

^bNonsteroidal anti-inflammatory medications and acetaminophen or paracetamol.

Source: Sponsor’s Summary of Clinical Evidence.²³

Exposure to Study Treatments

Full details of study treatment exposure are in Table 13. Briefly, as the CL303 trial was a crossover study design, patients originally randomized to the placebo arm had approximately 30 patient-years of less time exposed to burosumab. The mean duration of exposure in days was 477.0 days in the placebo arm and 634.0 days in the burosumab arm. No patients were reported to have discontinued treatment due to nonadherence; further information on adherence was not available.

Table 13: Summary of Patient Exposure From Studies Included in the Systematic Review

IQR = interquartile range; SD = standard deviation.

^aTotal patient-years of exposure = the sum of duration of exposure to burosumab (for all patients in each treatment group) ÷ 365.25.

Source: Sponsor’s Summary of Clinical Evidence.²³

Concomitant Medications and Cointerventions

Most patients (98.5%) received 1 or more concomitant medication during the study. The most common classes (reported in ≥ 25% of patients overall) were analgesics (83.6%), anti-inflammatory and antirheumatic products (80.6%), vitamins (64.2%), systemic antibacterials (54.5%), all other nontherapeutic products (53.7%), systemic antihistamines (44.0%), cough and cold preparations (36.6%), drugs for acid-related disorders (34.3%), corticosteroids for systemic use (32.8%), nasal preparations (30.6%), and psychoanaleptics (26.9%).

Based on patients’ diaries, overall pain medication use at the beginning of the placebo-controlled period was 69.1% in the burosumab arm and 66.7% in the placebo arm, and at the end of the placebo-controlled period was 66.2% in the burosumab arm and 60.6% in the placebo arm. Opioid use at the beginning of the placebo-controlled period was 25.0% in the burosumab arm and 19.7% in the placebo arm; at the end of the placebo-controlled period, 23.5% of patients in the burosumab arm and 21.2% of patients in the placebo arm reported opioid use. At week 48 (the last week of data collection), pain medication use was reported by 63.5% of patients in the burosumab-emergent arm and by 54.0% of patients in the placebo-emergent arm. Opioid use at week 48 was reported by 20.6% of patients in the burosumab-emergent arm and by 17.5% of patients in the placebo-emergent arm.

Efficacy

The proportion of patients who attained serum phosphorus concentrations greater than the LLN at both midpoints and end points of the dosing cycle is shown in Table 14. Changes from baseline to week 24, week 48, and week 96 in the key secondary and secondary PRO measures are summarized in Table 15. Key pseudofracture and fracture outcomes at week 24 and week 48 are summarized in Table 16, and key serum biomarker results are found in Table 17. A discussion of these outcomes follows each table.

Table 14: Summary of Key Serum Phosphorus Efficacy Results From CL303 Study (PAS)

CI = confidence interval; LLN = lower limit of normal; NR = not reported; PAS = primary analysis set.

^aThe 95% CI for the proportion of patients who attain mean serum phosphorus levels above the LLN was calculated using the Wilson score method.

^bThe P value was adjusted for multiplicity in the 24-week analysis.

Sources: Sponsor’s Summary of Clinical Evidence,²³ the CL303 Clinical Study Report,²⁰ and additional information provided by the sponsor.²¹

Proportion of Patients with Serum Phosphorus Levels Greater Than LLN

As noted in the previous CADTH reimbursement review, the proportion of patients with midpoint serum phosphorus levels greater than LLN was statistically significantly greater in the burosumab arm compared to the placebo arm.³⁰ The additional information on serum phosphorus levels following crossover to burosumab reported that the proportion of patients with midpoint serum phosphorus levels greater than LLN was 89.4% (95% CI, 79.7% to 94.8%) in the placebo-emergent arm and 83.8% (95% CI, 73.3% to 90.7%) in the burosumab-emergent arm at week 48. The proportion of patients with serum phosphorus levels greater than LLN at week 96 was 68.2% (95% CI, 56.2% to 78.2%) in the placebo-emergent arm and 82.4% (95% CI, 71.6% to 89.6%) in the burosumab-emergent arm. There was no information on the patients with serum phosphorus levels greater than LLN at the ends of dosing cycles for week 48 and week 96.

Brief Pain Inventory

As noted in the previous CADTH reimbursement review, the LSM differences in the placebo and burosumab arms for worst pain, pain interference, and pain severity at 24 weeks did not attain statistical significance.³⁰ In addition, the LSM change in BPI worst pain and BPI pain interference did not reach the MCID threshold specified by the current submission (≥ –1.72 points for worst pain and ≥ –1.00 points for pain interference) in either treatment arm.

BPI Worst Pain

At week 48, the LSM change from baseline in the placebo-emergent arm was −1.53 (95% CI, −1.98 to −1.09) and −1.09 (95% CI, −1.51 to −0.66) in the burosumab-emergent arm; the point estimates did not attain the MCID provided by the submission. The ad hoc analysis for the placebo-emergent group showed the change in BPI worst pain between week 24 and week 48 was −1.18 (95% CI, −1.61 to −0.76; P value < 0.0001). At week 96, the LSM changes from baseline at week 48 were −0.99 (95% CI, −1.51 to −0.47) in the placebo-emergent arm and −1.48 (95% CI, −2.07 to −0.90) in the burosumab-emergent arm. The point estimates did not surpass the MCID provided in the submission.

BPI Pain Interference

At week 48, the LSM change from baseline was −1.27 (95% CI, −1.77 to −0.78) in the placebo-emergent arm and −1.04 (95% CI, −1.51 to −0.56) in the burosumab-emergent arm. The point estimates in both arms surpassed the MCIDs provided in the submission. At week 96, the LSM change from baseline in the placebo-emergent group was −1.08 (95% CI, −1.59 to −0.57) in the placebo-emergent arm and −1.43 (95% CI, −1.89 to −0.97) in the burosumab-emergent arm. The point estimates in both arms surpassed the MCIDs provided in the submission.

Table 15: Summary of Key PRO Efficacy Results From CL303 Study (PAS)

6MWT = 6-minute walk test; BFI = Brief Fatigue Inventory; BPI = Brief Pain Inventory; CI = confidence interval; GEE = generalized estimating equation; LSM = least squares mean; NR = not reported; PAS = primary analysis set; PRO = patient-reported outcome; SD = standard deviation; SE = standard error; vs. = versus; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

^aThe P value was adjusted for multiplicity for the week 24 analysis.

^bEstimates of LSMs and P values for change from week 24 to week 48 in the placebo-emergent group are from an ad hoc GEE model, similar to those for the planned analysis.

Sources: Sponsor’s Summary of Clinical Evidence,²³ the CL303 Clinical Study Report,²⁰ and additional information provided by the sponsor.²¹

BPI Pain Severity

At week 48, the LSM change from baseline was –1.20 (95% CI, –1.58 to –0.81) in the placebo-emergent arm and –0.85 (95% CI, –1.16 to –0.54) in the burosumab-emergent arm. At week 96, the LSM change from baseline was –1.18 (95% CI, –1.57 to –0.80) in the placebo-emergent arm and –1.42 (95% CI, –1.87 to –0.97) in the burosumab-emergent arm. There were no MCIDs provided for this outcome.

Western Ontario and McMaster Universities Osteoarthritis Index

As noted in the previous CADTH reimbursement review, the LSM difference for change in WOMAC physical function score did not attain statistical significance at week 24 following the Hochberg adjustment for multiplicity, and the WOMAC stiffness scores did attain statistical significance after the Hochberg adjustment.³⁰ In addition to this, the WOMAC physical function and WOMAC stiffness score point estimates did not attain the MCIDs provided in the current submission (≥ –8.00 points for physical function and ≥ –10.00 points for stiffness).

WOMAC Physical Function

At week 48, the LSM change from baseline was –6.35 (95% CI, –11.94 to –0.76) in the placebo-emergent arm and –7.76 (95% CI, –11.97 to –3.55) in the burosumab-emergent arm. These point estimates did not surpass the MCID provided by the submission. At week 96, the LSM change from baseline was –8.41 (95% CI, –13.80 to –3.01) in the placebo-emergent arm and –9.02 (95% CI, –13.47 to –4.57) in the burosumab-emergent arm. The point estimates in both arms surpassed the MCID provided in the submission.

WOMAC Stiffness

At week 48, the LSM change from baseline was –15.29 (95% CI, –22.23 to –8.35) in the placebo-emergent arm and –16.03 (95% CI, –22.53 to –9.53) in the burosumab-emergent arm. The point estimates surpassed the MCID in both arms. The ad hoc analysis for the change in scores between week 24 and week 48 in the placebo-emergent arm showed a statistically significant reduction in scores following burosumab initiation at –15.82 (95% CI, –21.30 to –10.34; P value < 0.0001). At week 96, the LSM change from baseline was –17.67 (95% CI, –24.99 to –10.34) in the placebo-emergent arm and –15.32 (95% CI, –22.33 to –8.31) in the burosumab-emergent arm. Both point estimates surpassed the MCID.

WOMAC Pain

LSM changes from baseline were not reported in the submission. A trend toward numerically increasing reductions in observed WOMAC pain scores was reported in the submission between week 48 and week 96, for both the placebo-emergent and burosumab-emergent treatment arms.

6-Minute Walk Test

At week 48, the mean total distance walked at baseline was 367.28 (SD = 104.22) m in the placebo-emergent arm and 365.66 (SD = 125.44) m in the burosumab-emergent arm. The LSM change in total distance walked was −5.71 (95% CI, −21.70 to 10.28) in the placebo-emergent arm and 5.92 (95% CI, −15.00 to 26.84) in the burosumab-emergent arm. These point estimates did not surpass the MCIDs provided. This outcome was not measured at week 96.

Brief Fatigue Inventory

As noted in the previous CADTH reimbursement request, the change from baseline to week 24 in BFI worst fatigue and global fatigue scores did not attain statistical significance.³⁰ In addition to this, the point estimates and 95% CIs did not surpass the MCIDs provided by the sponsor in the current submission for either score (≥ –1.5 points for worst fatigue and ≥ –1.2 points for global fatigue).

BFI Worst Fatigue

At week 48, the LSM change from baseline was –1.23 (95% CI, –1.84 to –0.62) in the placebo-emergent arm and –1.01 (95% CI, –1.57 to –0.45) in the burosumab-emergent arm. The point estimates did not surpass the MCID provided by the submission. At week 96, the LSM change from baseline was –0.82 (95% CI, –1.53 to –0.11) in the placebo-emergent arm and –0.75 (95% CI, –1.35 to –0.26) in the burosumab-emergent arm. The point estimates did not surpass the MCID provided by the submission.

BFI Global Fatigue

At week 48, the LSM change from baseline was –0.73 (95% CI, –1.34 to –0.12) in the placebo-emergent arm and –0.46 (95% CI, –1.01 to 0.09) in the burosumab-emergent arm. The point estimates did not surpass the MCID provided in the submission. At 96 weeks, the LSM change from baseline was –0.86 (95% CI, –1.43 to –0.29) in the placebo-emergent arm and –0.80 (95% CI, –1.36 to –0.25) in the burosumab-emergent arm. The point estimates did not surpass the MCID provided in the submission.

Active Fractures and Pseudofractures

As noted in the previous CADTH reimbursement review, 50.0% of patients in the burosumab arm had active fractures graded as healed and 0.0% of patients in the placebo arm had active fractures graded as healed.³⁰ A total of 41.2% of patients in the burosumab arm and 9.0% of patients in the placebo arm had active pseudofractures graded as healed. The reassessment submission’s additional 24-week analyses reported a probability of a fully healed fracture at 24 weeks in the burosumab arm of 0.458 and a probability of 0.048 in the placebo arm. The OR of a fully healed fracture was 16.76 (95% CI, 4.93 to 56.95) for patients treated with burosumab (P < 0.0001).

At 48 weeks, 46.2% of patients in the placebo-emergent arm and 57.1% of patients in the burosumab arm had a fracture status of healed. A total of 33.3% of patients in the placebo-emergent arm and 64.7% of patients in the burosumab-emergent arm reported healed pseudofractures. The probability of a fully healed fracture was 0.725 (95% CI, 0.516 to 0.933) in the burosumab-emergent arm and 0.386 (95% CI, 0.718 to 0.594) in the placebo-emergent arm.

Table 16: Summary of Key Fracture-Related Efficacy Results From CL303 Study (PAS)

CI = confidence interval; NR = not reported; OR = odds ratio; PAS = primary analysis set.

^aThe 95% CI and P value correspond to the probability of a fracture being graded as fully healed for the week 48 analysis, vs. 0 probability of healing.

Sources: Sponsor’s Summary of Clinical Evidence²³ and the CL303 Clinical Study Report.²⁰

Table 17: Summary of Key Serum Biomarker Efficacy Results From CL303 Study (PAS)

1,25(OH)₂D = 1,25-dihydroxyvitamin D; BALP = bone-specific alkaline phosphatase; CI = confidence interval; LSM = least squares mean; NR = not reported; PAS = primary analysis set; SD = standard deviation; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate; TRP = tubular reabsorption of phosphate.

^aWeek 22 values were used for the week 24 analysis as this biomarker was not measured at week 24.

Sources: Sponsor’s Summary of Clinical Evidence,²³ the CL303 Clinical Study Report,²⁰ and additional information provided by the sponsor.²¹

Key Serum Biomarkers

Serum 1,25(OH)₂D

The previous CADTH reimbursement request reported that the LSM difference in serum 1,25(OH)₂D concentrations was statistically significantly larger in the burosumab arm relative to the placebo arm.³⁰ At week 48, the LSM change in the levels of serum 1,25(OH)₂D was 10.50 (95% CI, 5.76 to 15.24) in the placebo-emergent arm and 7.24 (95% CI, 2.44 to 12.04) in the burosumab-emergent arm. At week 96, the LSM change from baseline was 3.43 (95% CI, –1.17 to 8.03) in the placebo-emergent arm and 1.95 (95% CI, –2.66 to 6.57) in the burosumab-emergent arm.

TmP/GFR and TRP

As noted in the previous CADTH reimbursement review, the LSM difference showed a statistically significant increase in TmP/GFR for the burosumab arm relative to the placebo arm at 24 weeks.³⁰ At week 48, the LSM change from baseline in TmP/GFR was 0.55 (95% CI, 0.38 to 0.72) in the placebo-emergent arm and 0.48 (95% CI, 0.30 to 0.65) in the burosumab-emergent arm. At week 96, the LSM change was 0.29 (95% CI, 0.12 to 0.46) in the placebo-emergent arm and 0.46 (95% CI, 0.29 to 0.62) in the burosumab-emergent arm.

In the previous CADTH reimbursement review, the change in TRP from baseline at 24 weeks was statistically significantly greater in the burosumab arm.³⁰ At week 48, the LSM change from baseline was 0.02 (95% CI, 0.00 to 0.05) for the placebo-emergent arm and 0.03 (95% CI, 0.02 to 0.05) in the burosumab-emergent arm. At week 96, LSM change from baseline in the placebo-emergent group was –0.01 (95% CI, –0.04 to 0.02), while the LSM change in the burosumab-emergent group was 0.03 (95% CI, 0.01 to 0.05).

Bone-Specific Alkaline Phosphatase

The previous CADTH reimbursement review noted statistically nonsignificant changes in BALP at 24 weeks.³⁰ At week 48, the LSM change from baseline in BALP in the placebo-emergent arm was 6.69 (95% CI, 2.91 to 10.47) and the LSM change from baseline in the burosumab-emergent arm was 0.23 (95% CI, –3.36 to 3.81). At week 96, the LSM change was –2.49 (95% CI, –6.19 to 1.21) in the placebo-emergent arm and –2.76 (95% CI, –5.98 to 0.45) in the burosumab-emergent arm.

Harms

Overall, 97% of patients in the placebo-emergent arm and 100% of patients in the burosumab-emergent arm experienced a TEAE. The most common TEAEs in the burosumab-emergent arm over the entire duration of the trial were arthralgia (41%), nasopharyngitis (41%), headache (32%), tooth abscess (28%), and back pain and pain in extremity (27% each). The most common TEAEs in the placebo-emergent arm during the placebo-controlled period (first 24 weeks) were arthralgia (24%), pain in extremity (15%), fatigue (11%), and oropharyngeal pain (11%). After initiating burosumab, the most common TEAEs in the placebo-emergent arm were arthralgia (36%), nasopharyngitis (36%), headache (27%), and back pain (26%). Table 18 contains harms data for the most commonly reported TEAEs (reported in > 10% of patients in either arm). The data for the burosumab arm were presented together for the placebo-controlled period and the open-label period, and the data for the placebo arm were presented separately for the placebo-controlled period and the open-label period.

Adverse Events

Overall, there were differences between the burosumab-emergent arm during the entire trial and the placebo-emergent arm after initiating burosumab for some TEAEs. Specifically, there were differences between the burosumab-emergent arm and the placebo-emergent arm in the proportion of patients reporting tooth abscesses (28% and 8%, respectively), vitamin D deficiency (22% and 11%, respectively), injection site reactions (12% and 25%, respectively), diarrhea (19% and 8%, respectively), upper respiratory tract infection (18% and 3%, respectively), nausea and dizziness (both 16% and 8% in each arm, respectively), depression (13% and 5%, respectively), hypoesthesia (10% and 5%, respectively), migraine (10% and 3%, respectively), oropharyngeal pain (6% and 12%, respectively), injection site pruritus (4% and 12%, respectively), and ectopic mineralization (0% and 11%, respectively).

Serious Adverse Events

During the placebo-controlled period, an SAE was reported in 1 patient (invasive ductal breast carcinoma) in the placebo-emergent arm and 2 patients (irritable bowel syndrome and back pain) in the burosumab-emergent arm.

In the placebo-emergent arm, 10 patients overall reported SAEs. A total of 6 patients reported SAEs between week 24 and week 48 (presyncope, palpitations, cervical spinal stenosis, periodontal disease, pseudarthrosis, and subdural hematoma) and 4 patients reported SAEs between week 48 and week 96 (partial seizures, arthralgia [separate events for each knee], joint range of motion decreased, and benign parathyroid tumour).

The burosumab-emergent arm overall reported SAEs in 12 patients. A total of 4 patients reported SAEs between week 24 and week 48 (colitis, procedural nausea and/or vomiting, spinal column stenosis and myelopathy, and musculoskeletal pain) and 4 patients reported SAEs between week 48 and week 96 (knee deformity, cholelithiasis and gastroenteritis, spondylolisthesis, and duodenal ulcer).

Withdrawals Due to AEs

No patients stopped treatment due to AEs.

Mortality

There was 1 death reported in the burosumab-emergent cohort, which was due to a road traffic accident. Per the submission, this was judged not related to the study treatment.

Notable Harms

AEs of special interest included injection site reactions, hypersensitivity, hyperphosphatemia, ectopic mineralization, and restless leg syndrome. A total of 16 (24%) patients in the placebo-emergent arm reported injection site reactions after initiating burosumab and 8 (12%) patients reported injection site reactions before initiation. In addition, 7 (11%) patients in the placebo-emergent arm experienced ectopic mineralization, which was not reported in any of the other treatment arms. Lastly, there were 0 reports of hyperphosphatemia in the placebo-emergent arm during the placebo-controlled period; hyperphosphatemia was reported in 6% of patients in the placebo-emergent arm after initiating burosumab and in 6% of patients in the burosumab-emergent arm throughout the trial.

Noting the higher proportions of patients in the burosumab-emergent arm experiencing TEAEs and serious TEAEs, the submission included an exposure-adjusted analysis reporting incidence rates in each arm, which revealed generally similar incidence rates in the placebo-emergent and burosumab-emergent arms. Full results of this analysis were not available in the submission.

Table 18: Summary of Harms Results From CL303 Study (SAS)

AE = adverse event; SAE = serious adverse event; SAS = safety analysis set; TEAE = treatment-emergent adverse event.

^aThis includes the treatment continuation period (week 24 to week 48), extension period I (week 48 to week 96), or extension period II (through week 149).

^bThese were the most common TEAEs that occurred in 10% or more of patients in either study arm.

Source: Sponsor’s Summary of Clinical Evidence.²³

Critical Appraisal

Internal Validity

The CL303 study was a double-blind, randomized trial with a 24-week placebo-controlled period, followed by an open-label treatment continuation phase (24 additional weeks, all patients receiving burosumab) and open-label treatment extension I (48 additional weeks) for adult patients with XLH. The additional data analyses for this reassessment of reimbursement focused on longer-term efficacy by comparing the within-group changes in the placebo-emergent group and the burosumab-emergent group at week 48 and week 96. Of note, the submission claimed an error in the stratification randomization process such that patients were randomized on the basis of the BPI average pain score rather than the BPI worst pain score, which affected the majority of patients. Analyses were conducted on the basis of actual randomization. Overall, there appeared to be a low risk of bias from the randomization process itself, as it was stratified and centrally conducted; treatment allocation concealment and blinding processes were not applicable to the data considered in the reassessment as the study design was open-label at week 48 and week 96.

Most baseline characteristics were balanced between study arms at randomization; however, there were some concerns with imbalances in certain medical characteristics, which may impact the outcomes. A numerically higher proportion of patients in the burosumab arm relative to the placebo arm had osteoarthritis (69.1% versus 57.6% of patients, respectively) and nephrocalcinosis (16.2% versus 7.6% of patients, respectively), and a numerically higher proportion of patients in the placebo arm relative to the burosumab arm had active pseudofractures at baseline (51.5% versus 41.6%, respectively). Osteoarthritis and pseudofractures both may cause pain and could bias the assessment of efficacy on outcomes pertaining to pain, and the imbalance in pseudofractures could impact the active pseudofracture healing outcome. Both of these outcomes may also impact the physical function outcome. This limitation would apply to the results at all time points (24 weeks, 48 weeks, and 96 weeks).

There are some major limitations pertaining to both the open-label study design and the statistical analyses, which may cast a degree of inconclusiveness on the sustainability of treatment effect at later time points. First, the sample size was powered only for the primary end point to demonstrate a statistically significant difference at week 24 for the primary outcome of serum phosphorus, and therefore there could be a lack of power for key secondary but clinically important outcomes, such as the PROs. This adds uncertainty to the exact magnitude of benefit for these outcomes. Of note, there were fluctuations in the primary outcome results at week 48 and week 96 — specifically, in the placebo-emergent group, the percentage of patients attaining serum phosphorus levels above the LLN was 89% at week 48 and 68% at week 96; in the burosumab-emergent group, this percentage remained nearly unchanged (83% at week 48 and 82% at week 96). The explanation for this is unclear. The concerns around statistical power apply particularly to the week 48 ad hoc analysis. First, since this analysis was not preplanned, it should be considered exploratory as the timing of this analysis relative to the rest of the week 48 analyses is not clear. Second, a total of 6 patients had discontinued from the placebo-emergent arm and 8 patients had discontinued from the burosumab-emergent arm as of week 96; given the relatively small sample size, this could represent a notable loss to follow-up. Third, by study design, at week 48 all patients crossed over to receive burosumab. The lack of control group impacts causal inference and it is therefore difficult to attribute the changes in efficacy outcomes and harms to burosumab alone during the open-label phases. This may be of particular importance for the harms results as a possible cumulative side effect burden was observed.

There are some limitations that impact the interpretation of the results. While there was likely a low risk of bias for the measurement of laboratory outcomes (serum phosphorus, 1,25(OH)₂D, TmP/GFR, TRP, and BALP) as they were analyzed at a central laboratory, the reporting of PRO results, as subjective measures, could be impacted by the open-label design. In addition, information on unblinding for safety reasons or the measure of adherence to burosumab treatment was unavailable. Furthermore, there are some missing data for serum biomarkers and PRO scores (data from 59 patients in the placebo-emergent arm and from 58 patients in the burosumab-emergent arm were available at the 96-week mark), which were handled by exclusion from the analysis, but the potential impacts of this choice are not explored by sensitivity analysis relative to other missing data methods. For fracture outcomes, only targeted radiography was performed to check the progress of fractures after the initial scan at baseline, and these scans did not appear to be identifying new fractures, but tracking the progress of old ones. This could have impacted the detection of new fractures in particular, as the development or absence of fractures in non–X-rayed sites may be missed, which may bias the numbers of fractures in favour of healed ones. In addition, at week 48, patients in the burosumab-emergent arm had numerically higher rates of pain medication usage (63.5% of patients in the burosumab-emergent arm and 54.0% of patients in the placebo-emergent arm) and opioid use (20.6% of patients in the burosumab-emergent arm and 17.5% of patients in the placebo-emergent arm). These differences in the use of pain medications may bias the efficacy results, particularly of the WOMAC pain and BPI measures. Lastly, TEAE results were not reported separately for the placebo-controlled period and the open-label period in the burosumab-emergent cohort, and this could limit the comparability of the 2 treatment arms in terms of reporting the timing of TEAEs.

External Validity

Per the clinical expert consulted by CDA-AMC, the inclusion and exclusion criteria of the CL303 study were applicable and specific to patients with XLH, though they noted there remains a lack of data on what to do with a patient who is receiving burosumab and subsequently becomes pregnant. The study also focused on adult patients with XLH but did not specifically select patients diagnosed with XLH as adults; although this is likely a small number, this might impact the generalizability to this patient population. Furthermore, the patient population included was majority white and majority female, which could have under-represented patient populations that include Indigenous Peoples and males. Lastly, the majority of the patient population reported receiving vitamin D and phosphate before treatment with burosumab, which does not provide any information on the effectiveness of burosumab for treatment-naive adults with XLH.

There are limitations that affect the generalizability of these results to real-life situations. The frequent visits and dose adjustment protocols used in the trial setting may not exactly reflect daily clinical practice in Canada and therefore the optimized efficacy and safety profile during the trial may not be extrapolatable to the general patient population. Moreover, patients were prohibited from using certain concomitant medications such as pharmacologic vitamin D metabolites or analogues, oral phosphate, aluminum hydroxide antacids, acetazolamides, thiazides, bisphosphonate therapy, denosumab therapy, teriparatide therapy, or chronic use of systemic corticosteroids during the trial. This may not represent prescribing patterns in routine practice and may impact the generalizability of the findings from these additional data analyses for reassessment. For fracture healing, which may take longer to capture, the duration of the study may not have been a long enough time to fully determine the impact of burosumab on these outcomes. Furthermore, the PRO measures used in the study were noted by the clinical expert to not be routinely used in clinical practice, suggesting that the impact of treatment on subjective measures such as pain, fatigue, and stiffness in the clinical trial may not be easily translated into these settings. Lastly, while the sponsor provided MCIDs for the WOMAC, BPI, and BFI domain scores, it is important to note that apart from 2 domains of WOMAC, where a general population sample was cited by the sponsor, these MCIDs were derived from cross-sectional data (the UX023-CL001 study), phase II trial data (the UX023-CL203 study), and the CL303 pivotal trial data, all collected by the sponsor.²⁵ The CL303 study is limited by being the same data used in the efficacy analyses, and the UX023-CL001 and UX023-CL203 studies are also limited by being cross-sectional data and having a small sample size (N = 20), respectively. There is therefore no external data in a reference population with XLH to use as a comparison for meaningful clinical change. Furthermore, the MCIDs were obtained through post hoc analysis and not from studies designed for this purpose, and should be considered exploratory thresholds. Therefore, there remains a lack of confirmatory data on the meaningfulness of these score changes in the general adult XLH population.

LTE Study

Content in this section has been informed by materials submitted by the sponsor. The following has been summarized and validated by the CDA-AMC review team.

Description of Study

The BUR02 study^26,27 was an open-label, phase III study evaluating the long-term efficacy and safety of burosumab in adult patients with XLH. It was undertaken using patient populations that had completed the CL303²⁰ or CL304 study.²⁸ The CL303 trial²⁰ was a phase III RCT that evaluated measures of phosphate metabolism, PROs, and fractures and pseudofractures in adults with XLH and is appraised in the Systematic Review section. The CL304 study was a phase III, single-arm study that evaluated measures of osteomalacia in patients with XLH who received burosumab treatment; it was not appraised in the current submission.²⁸ Patients completing the CL303 trial were eligible to transition to the BUR02 study. However, there was an interval between the CL303 and BUR02 trials (a mean of 9 months; range, 6 months to 16 months) where interim burosumab treatment was provided via an early access program only to the patients for whom the drug supply was accessible. Therefore, some participants received compassionate burosumab treatment during this period whereas others did not. An exploratory analysis (not reported in this review) evaluated the effects of burosumab treatment continuation versus interruption. The time to implement compassionate use in each country varied and treatment during the interim period was not captured.^40,61

The primary objective of the BUR02 trial was to evaluate the long-term impact of burosumab treatment on maintaining serum phosphate levels within the normal range. Secondary objectives were to assess the effects of burosumab on PROs, clinical laboratory tests, and ambulatory function.²⁶

The planned duration of treatment in this study was until December 2021 or until burosumab became commercially available within each participating country. Due to the impact of COVID-19 and at the investigator’s discretion, patients were allowed to withdraw from the study sooner than December 2021.²⁶ The end-of-study visit occurred 4 weeks (± 7 days) after the patient received their last dose of burosumab on the study. A safety follow-up telephone call occurred 12 weeks (± 7 days) after the patient’s final study site visit to collect information on any ongoing or new AEs, SAEs, or concomitant medications. The end of the study was defined as the last day that protocol-specified assessments were conducted for the last patient. The maximum planned study duration was approximately 140 weeks.²⁶

An interim analysis was conducted and 2 data cut analyses were performed after the completion of week 12 and week 48 of the BUR02 trial to support a publication regarding the changes in outcomes between the CL303 and BUR02 trials, among the 31 patients from the CL303 study only, who enrolled in the BUR02 study.⁶¹ This section reports results from both the interim and final analysis provided by the submission.^26,61

Populations

The inclusion and exclusion criteria for the BUR02 study were consistent with those of the CL303 and CL304 parent trials. Briefly, adult participants (aged 18 years to 65 years) with a confirmed diagnosis of XLH, biochemical findings consistent with XLH, and a BPI worst pain score of 4 or more and who had completed the CL303 or CL304 study — including the final study visit — were eligible for the BUR02 trial regardless of their response to primary or secondary end points in the CL303 and CL304 trials. Major exclusion criteria for both the BUR02 trial and the interim analysis were clinically significant hypocalcemia, hypercalcemia, or any other concurrent condition that would interfere with participation or affect safety.

Patients who did not complete the CL303 or CL304 trial were included on a case-by-case basis. The interim analysis evaluated a subset of the BUR02 study population that had previously enrolled in the CL303 study. It was noted in the submission that patients from the CL304 study could not be included in the interim analysis because their inclusion would violate the principles of the GEE model that was used in the interim analysis.

Interventions

Participants in the BUR02 study continued on the burosumab regimen they received at the end of the parent trials: SC burosumab every 4 weeks. Doses were rounded to the nearest 10 mg, and the dose was to remain fixed unless serum phosphorus levels increased to more than 5.0 mg/dL, at which point the dose would be reduced by half, or the patient’s body weight changed by more than 20% from baseline, at which point the dose would be recalculated to account for the new body weight.

Outcomes

The primary outcome in the BUR02 study was the proportion of patients attaining a mean trough serum phosphate level (i.e., the serum level just before the administration of the next burosumab dose) above the LLN. Secondary outcomes included the effects of continued burosumab treatment on laboratory markers — namely, serum 1,25(OH)₂D, serum phosphate, urinary phosphate and renal phosphate reabsorption (TmP/GFR) — on PROs such as BPI, BFI, WOMAC, and the Short Form (36) Health Survey version 2, and on ambulatory function using the 6MWT. Harms outcomes were evaluated only in the final analysis by the incidence, frequency, and severity of AEs, TEAEs, and SAEs, including death.

Outcomes in the BUR02 study were generally defined and evaluated consistently with the methods of the CL303 study; the CL304 trial evaluated a smaller subset of PROs compared to the CL303 and BUR02 studies. The submission noted a couple of key differences between the CL303 and CL304 studies and the BUR02 trial. First, study samples were analyzed at different central laboratories for the BUR02 trial and the CL303 and CL304 studies, which used slightly different values for the LLN for each biochemical end point. Furthermore, only end point serum phosphate values were analyzed in the BUR02 trial while the CL303 study analyzed serum phosphorus levels at the midpoint and end point of the dose interval.

Statistical Analysis

No formal statistical testing was planned as part of the BUR02 study and all statistical analyses were descriptive and exploratory. No adjustment was made to account for multiplicity. Subgroups were predefined as patients who had taken placebo during the double-blind period from the CL303 study and patients who had received burosumab during the double-blind period of the CL303 study or during the CL304 study.

The final BUR02 trial analysis was based on the full analysis set that consisted of all enrolled patients who completed the CL303 and CL304 studies and received at least 1 burosumab dose during the BUR02 study. All changes were calculated from the BUR02 trial baseline; ad hoc analyses were also conducted to understand how outcomes had changed from the CL303 and CL304 studies baseline. The ad hoc analyses could not consider all PROs that were evaluated in the CL303 and BUR02 studies because the CL304 study evaluated a smaller subset of PROs. For the final BUR02 study analysis, study visits were described as starting at baseline (week 0b) and week 12b to week 144b.

The interim analysis considered patients from the CL303 trial and changes were calculated from the CL303 trial baseline using the same methods as those from the CL303 study. The change from the CL303 study baseline to each assessment time point in the CL303 and BUR02 trials was analyzed using a GEE repeated measures analysis to align with the methods used in the CL303 trial. The model included treatment, the actual randomization stratification factor based on BPI average pain (except the model for BPI worst pain), region, visit, and interaction of treatment-by-visit as fixed factors, adjusted for phase III baseline measurements. Compound symmetry was used as the covariance structure for the model, which specified constant variances for the assessments and constant covariances between the assessments over time. Nominal P values were calculated and these analyses were not adjusted to account for multiplicity. To differentiate the time points, study visits in the CL303 trial have been described as baseline and week 12a to week 96a, while study visits in the BUR02 trial have been described as week 0b to week 48b (where 0b corresponds to the BUR02 study entry or baseline).

Results

Patient Disposition

A total of 35 patients were enrolled at 10 centres: 34 patients from the CL303 study and 1 patient from the CL304 study. At the data cut-off in January 2021, 31 participants had received up to 48 weeks of further burosumab treatment. Patient disposition from both interim and final analyses is summarized in Table 19.

At the interim analysis, 31 (91.1%) patients from the CL303 trial had completed the 48-week treatment period in the BUR02 trial and there were no study or treatment discontinuations due to TEAEs or death. At the final analysis, all 35 patients from the CL303 and CL304 trials had received burosumab and 25 (71.4%) patients had completed the study. Ten (25%) patients prematurely terminated the study; common reasons for discontinuation were withdrawal of consent (2 patients) and “other reasons” (not specified, 8 patients). No patients discontinued due to AEs.

Table 19: Patient Disposition for BUR02 Study

NA = not applicable.

Sources: Sponsor’s interim analysis Clinical Study Report⁶¹ and BUR02 trial final Clinical Study Report.²⁷

Baseline Characteristics

Demographics and baseline disease characteristics are presented in Table 20. At the interim analysis, data were available up to week 48b from 31 patients who had previously participated in the CL303 trial. The study population had a mean age of 40.1 (SD = 12.2) years, with 67.7% of patients being female. Most of the patients had pathogenic PHEX variants (87.1%) and histories of pain medication, orthopedic surgery, and osteoarthritis.

Table 20: Summary of Baseline Characteristics for BUR02 Study

BMI = body mass index; SD = standard deviation.

Source: Sponsor’s ad hoc analyses comparing the 2 patient populations.⁶²

Concomitant Medications and Cointerventions

Concomitant vitamin D and analogues were common (88.6%) among the 35 patients enrolled in the BUR02 study. Based on the interim analysis, the number of patients using pain medication was 25 (81%) patients at the phase III baseline,16 (52%) patients at the open-label extension baseline, and 15 (48%) patients at the end of the open-label extension. Concomitant pain medications included anilides (80.0%), propionic acid derivatives (54.3%), natural opium alkaloids (22.9%), acetic acid derivatives and related substances (20.0%), opioids in combination with nonopioid analgesics (17.1%), and other opioids (17.1%). Glucocorticoids were used by 25.7% of patients and corticosteroids were used by 8.6% of patients.

At the end of the BUR02 study, 34 of the 35 (97.1%) patients continued to receive burosumab via the sponsor’s post-trial access program.

Exposure to Study Treatments

At the interim analysis, the mean burosumab dose at the CL303 trial baseline was 1.0 (SD = 0.06) mg/kg (range, 0.9 mg/kg to 1.1 mg/kg) and remained constant through week 48b of the BUR02 study. A slightly broader dose range was observed up to week 48b of the BUR02 study (0.8 mg/kg to 1.5 mg/kg).⁶¹

At the final analysis, all 35 patients had received all their scheduled doses; no doses were skipped. Among all 35 patients, the mean time from the first burosumab dose in the BUR02 trial to study completion or early withdrawal was 116.2 (SD = 30.7) weeks.²⁷

Efficacy

Full results for the serum biomarker interim analysis are contained in Table 21, and results from the final analysis are in Table 22. Highlights of these results are summarized as follows.

Proportion of Patients With Serum Phosphate Levels Greater Than LLN

At baseline, when the patients had completed their scheduled burosumab treatment in their previous study (the CL303 or CL304 trial), 34.3% of patients had serum phosphate levels above the LLN. The proportion increased to 55.9% at week 12 and remained primarily within a range of 55% to 75% in subsequent visits. At the end of the study, 66.7% of the patients reported serum phosphate levels above the LLN (Table 22).

Key Serum Biomarkers

Ratio of Renal Tubular Maximum Phosphate Reabsorption Rate to Glomerular Filtration Rate

At the CL303 study baseline, the mean TmP/GFR was 0.55 (SD = 0.15) mmol/L and increased to 0.70 (SD = 0.26) mmol/L at week 12a, with that level sustained through both studies.

At the final analysis, the mean TmP/GFR was 0.62 (SD = 0.22) mmol/L and it increased to 0.69 (SD = 0.14) mmol/L at week 48b, with these levels sustained over time. At the end of the study, the mean TmP/GFR was 0.71 (SD = 0.20) mmol/L.

1,25(OH)₂D Concentrations

At the interim analysis, the mean serum 1,25(OH)₂D was 79.95 (SD = 29.77) pmol/L at the CL303 study baseline, 98.56 (SD = 30.27) pmol/L at week 48a, and 83.36 (SD = 32.97) pmol/L at week 72a. At the BUR02 study baseline, the mean serum 1,25(OH)₂D was 78.43 (SD = 41.49) pmol/L and increased to 92.85 (SD = 36.06) pmol/L at week 12b, remaining consistent through to week 48b of the BUR02 study.

According to the final analysis, at baseline, the mean serum concentration of 1,25(OH)₂D was 32.67 (SD = 16.35) pg/mL. At week 12, the 1,25(OH)₂D concentration increased to 39.86 (SD = 15.57) pg/mL. At week 24, week 48, week 72, and week 96, the mean serum 1,25(OH)₂D was 36.34 (SD = 9.80) pg/mL, 37.04 (SD = 7.83) pg/mL, 38.16 (SD = 11.30) pg/mL, and 41.01 (SD = 12.80) pg/mL, respectively. At the end of the study, the mean serum 1,25(OH)₂D was 38.53 (SD = 12.70) pg/mL.

Table 21: Laboratory Tests of Efficacy (Interim Analysis)

1,25(OH)₂D = 1,25-dihydroxyvitamin D; NR = not reported; SD = standard deviation; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate.

Sources: Sponsor’s BUR02 Clinical Study Report²⁷ and additional information provided by the sponsor.⁶¹

Table 22: Laboratory Tests of Efficacy (Final Analysis)

1,25(OH)₂D = 1,25-dihydroxyvitamin D; LLN = lower limit of normal; NE = not estimable; SD = standard deviation; TmP/GFR = ratio of renal tubular maximum phosphate reabsorption rate to glomerular filtration rate.

Note: % = n/N assessed × 100 unless the denominator for the percentage calculation is specified.

Source: Sponsor’s BUR02 trial final Clinical Study Report.²⁷

Patient-Reported Outcomes

WOMAC

The full results for relevant WOMAC domains are in Table 23 (interim analysis) and Table 24 (final analysis). In summary, based on the interim analyses in the CL303 study, the LSM of WOMAC stiffness scores was –14.77 (SE = 4.03) at week 36a and this reduction was sustained at all subsequent time points in the 2 studies. Similar results were reported for the WOMAC physical function score.

In the final analysis of the BUR02 study, the mean stiffness score was 55.15 (SD = 18.75) points at baseline, and the mean change was –3.13 (SD = 17.68) points at week 12. The mean stiffness scores were maintained at lower than baseline throughout subsequent visits. The mean changes in stiffness score from baseline to week 24, week 48, and week 96 were –9.19 (SD = 22.89) points, –8.62 (SD = 18.63) points, and –9.09 (SD = 20.48) points, respectively. At the end of the BUR02 study, the mean score decreased by –14.52 (SD = 22.61) points. Similar decreases were observed for the WOMAC pain score and physical function score.

Table 23: Patient-Reported WOMAC Outcomes From CL303 Study and BUR02 Study (Interim Analysis)

CfB = change from baseline; SD = standard deviation; SE = standard error; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

Note: Consistent with the CL303 study methods, the least squares mean CfB was calculated at the interim analysis using a generalized estimating equation relative to the CL303 study baseline.

Source: Sponsor’s interim analysis data tables (Table 2.2.2).⁶¹

Table 24: Patient-Reported WOMAC Outcomes From CL303 Study and BUR02 Study (Final Analysis)

CfB = change from baseline; NE = not estimable; SD = standard deviation; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

Note: The arithmetic mean change from the BUR02 study baseline was calculated at the final analysis.

Source: Sponsor’s BUR02 trial final Clinical Study Report (Table 14.2.2.6.1 and Table 14.2.2.6.2).²⁷

Brief Pain Inventory

The interim analysis results for the BPI are available in Table 25, and the final analysis results are in Table 26. Briefly, based on the interim analyses in the CL303 study, the LSM change from baseline in the BPI average worst pain scores at week 12a was −0.88 (SE = 0.281) and decreased from baseline at all subsequent time points in the 2 studies except for week 24a. The BPI pain interference scores had also decreased from baseline with an LSM change from baseline of −1.22 (SE = 0.309) at week 12a and at all subsequent time points in both studies except week 24a.

Similarly, according to the final analysis from the BUR02 trial, the mean BPI worst pain score was 5.78 (SD = 1.725) points at baseline. The mean changes in BPI worst pain score from baseline to week 12 was −0.51 (SD = 1.698) points and these levels were maintained at lower than baseline at week 24, week 36, week 48, week 72, and week 96.

The mean BPI pain severity score was 4.52 (SD = 1.657) points at baseline (N = 32), and the mean change in BPI worst pain score from baseline was −0.40 (SD = 1.416) points at week 12 (N = 12). These values were maintained throughout subsequent visits. The mean BPI pain interference score was 3.48 (SD = 2.135) points at baseline, and the mean change in the BPI pain interference score from baseline was −0.37 (SD = 1.862) points at week 24. The lower values were maintained throughout subsequent visits in the BUR02 trial.

Table 25: Patient-Reported BPI Outcomes From CL303 Study and BUR02 Study (Interim Analysis)