CADTH Reimbursement Review

Dalbavancin (Xydalba)

Sponsor: Paladin Labs Inc.

Therapeutic area: Acute bacterial skin and skin structure infections

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Clinical Review

Executive Summary

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

Introduction

Acute bacterial skin and skin structure infections (ABSSSIs) are bacterial infections of skin and associated tissues. They are caused by the penetration of bacterial skin flora into soft tissues. ABSSSIs include cellulitis, erysipelas, wound infections, and major cutaneous abscesses.¹ ABSSSIs may be acquired in either the community or hospital setting and are a frequent cause of morbidity.^2-4 Worldwide, the most common cause of ABSSSIs is Staphylococcus aureus, including methicillin-resistant Staphylococcus aureus (MRSA).^2,5 In Canada, the estimated prevalence of ABSSSIs requiring IV antibiotics is 0.39%,^6,7 and First Nations communities are disproportionately affected.⁷ Severity ranges from nonsevere infections to severe, life-threatening infections that affect deeper tissue layers.^2,8,9 In North America, the mortality rate among patients hospitalized for skin and soft tissue infections (SSTIs) is approximately 0.5%.¹⁰ Compared with non-MRSA infections, ABSSSIs caused by MRSA are associated with increased risk of mortality, longer hospital stays, and higher health care costs.¹¹

The microbiologic diagnosis of ABSSSIs can be difficult, especially in cases of closed infection, where a microbiological specimen cannot be obtained without invasive procedures, such as biopsy. Hence, in many cases, the microbiological cause of ABSSSIs is unknown. Among outpatients, lack of compliance may contribute to failure of initial therapy, which may result in hospitalization.^12-14 ABSSSIs can lead to missed work or school, the inconvenience of inpatient treatment, emotional distress, and embarrassment, all of which can affect a patient’s quality of life.¹⁵

The current standard of therapy for ABSSSIs in Canada is a cephalosporin for methicillin-sensitive Staphylococcus aureus (MSSA) infections (i.e., IV cefazolin, IV ceftriaxone, or oral cephalexin), and IV vancomycin, IV daptomycin, or oral or IV linezolid for MRSA infections (daptomycin and linezolid are rarely used). For severe ABSSSI, treatment includes IV piperacillin plus tazobactam or IV ertapenem. According to the clinical expert consulted by CADTH, an oral route of administration is well suited for the treatment of ABSSSI, as oral antibiotics are inexpensive, easy to use, and can be administered at home. However, the potential lack of adherence to oral antibiotics is a concern. The goals of therapy are prevention of mortality, improvements in accessibility and compliance, and reducing the risk of IV-related complications, hospital visits, and health care costs.

Dalbavancin is a semisynthetic bactericidal lipoglycopeptide active against susceptible strains of gram-positive bacteria.¹⁶

Dalbavancin is indicated for the treatment of adults with ABSSSI caused by susceptible isolates of the following gram-positive micro-organisms: Staphylococcus aureus (including methicillin-susceptible and methicillin-resistant strains), Streptococcus pyogenes, Streptococcus agalactiae, Streptococcus dysgalactiae, Streptococcus anginosus group (including Streptococcus anginosus, Streptococcus intermedius, and Streptococcus constellatus), and Enterococcus faecalis (vancomycin-susceptible strains).¹⁶ The Health Canada recommended dosage is 1,500 mg administered either as a single dose or as 1,000 mg followed 1 week later by a 500 mg IV infusion administered over 30 minutes. The sponsor’s requested reimbursement criteria for dalbavancin aligns with the Health Canada indication. Dalbavancin underwent a priority review process by Health Canada and received a Notice of Compliance on May 5, 2021. Dalbavancin has no other Health Canada–approved indication and has not previously been reviewed by CADTH.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical expert noted many gaps in the treatment of ABSSSIs, including the fact that current IV treatment is expensive and complicated (potentially requiring nursing effort, a peripherally inserted central catheter [PICC] line, repeated visits, and automated IV pumps), there is a risk of IV-line complications, patient adherence to treatment can be poor, obtaining a microbiological can be difficult, and the current standard of care can have toxic effects (e.g., IV vancomycin, which requires therapeutic monitoring). Sometimes, given the difficulty in obtaining a swab to identify the pathogen, the diagnosis may be incorrect and delay the treatment process. In addition, patients may be misdiagnosed with ABSSSIs and given unnecessary treatment. In terms of access, people in remote areas or Indigenous communities may have poor or limited access to IV treatment and may face geographic barriers accessing treatment after discharge.

According to the clinical expert, dalbavancin could change the current treatment paradigm, given its unique pharmacokinetics and its ease of use compared to other treatments. Dalbavancin would be considered a first-line treatment and would replace current IV antibiotic treatments for ABSSSIs caused by gram-positive bacteria (unless oral antibiotic treatment was available). In the opinion of the clinical expert, there would be no reason to try other treatments before dalbavancin. Given that dalbavancin does not require repeat IV infusion, PICC line placement, or therapeutic drug monitoring, it is much easier to access than other standards of therapy and may lead to better patient satisfaction. In addition, dalbavancin could prevent repeat visits, hospital admission, and poor adherence, which is observed with other therapies. According to the clinical expert, patients best suited for treatment with dalbavancin include those with MRSA infection, those who inject drugs, those identified by physician assessment at presentation, and those who are clinically stable and candidates for home IV, who have mild-to-moderate infection, are competent, and have access to a nearby hospital. Patients least suited for dalbavancin include those with infections that require surgical debridement and those with polymicrobial infections. According to the clinical expert, response in a trial setting is best assessed using FDA outcome criteria at 48 to 72 hours: alive, a 20% reduction in the affected area, and no rescue treatment. In clinical practice, there are typically daily assessments, which may not be necessary; also, assessment may be highly variable between physicians (e.g., determining time to resolution of redness). Patients can be re-treated, typically around day 8, if there is no response. However, physicians should investigate the reasons for nonresponse and adjust the treatment plan accordingly.

The clinical expert indicated that dalbavancin could be administered in any setting in which there is IV access (e.g., emergency room [ER], long-term care facility, hospital, clinic, or at a patient’s home). Dalbavancin does not require a specialist for prescription, although its use should be regulated carefully by an antimicrobial stewardship committee. Treatment should not be used off-label without trial evidence. The clinical expert added that trials should be done in patients with osteomyelitis, endocarditis, abscess, prosthetic joint infection, pneumonia, and meningitis.

The clinical expert also noted that dalbavancin has some potential harms. If an allergic reaction occurred, there would be no way to withdraw the drug. Also, the convenience of dalbavancin dosing could increase the unnecessary treatment of wrong diagnoses, lead to failure to change treatment after culture results are received, lead to increased IV treatment when oral treatment would be adequate, or increase unnecessary prophylaxis. According to the clinical expert, there may be concerns related to antimicrobial resistance with the use of dalbavancin, if used inappropriately; with dalbavancin, there is prolonged duration of selection toward antimicrobial resistance in human gut flora. Antimicrobial resistance is associated with enormous cost to the Canadian health system and mortality among Canadians; therefore, antimicrobial use must be minimized, both in the spectrum of action and the duration of treatment, to reduce antimicrobial resistance.

Drug Program Input

Input was obtained from the drug programs that participate in the CADTH reimbursement review process. The following were identified as key factors that could potentially affect the implementation of a CADTH recommendation for dalbavancin: considerations related to the initiation of therapy, discontinuation of therapy, prescribing, generalizability, and care provisions. The clinical expert consulted by CADTH provided advice on the potential implementation issues raised by the drug programs.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

Four randomized controlled trials (RCTs) were included in this review: the DISCOVER 1, DISCOVER 2, DUR001-303, and VER001-9 trials. The first 3 RCTs were considered pivotal trials by Health Canada. The DISCOVER 1, DISCOVER 2, and DUR001-303 studies were sponsored by Durata Therapeutics, and the VER001-9 study was sponsored by Vicuron Pharmaceuticals Inc.

The DISCOVER 1 (N = 573) and DISCOVER 2 (N = 739) studies were phase III, multicentre, 1:1 randomized, double-blind (DB), noninferiority studies comparing the efficacy and safety of dalbavancin to vancomycin (with a possible switch to oral linezolid) in patients with a known or suspected gram-positive ABSSSI. The primary objective in both trials was to compare clinical efficacy 48 to 72 hours after study-drug initiation between dalbavancin and a vancomycin and linezolid regimen. Clinical response was defined as no increase in lesion size, absence of pyrexia at 48 to 72 hours in the intention-to-treat (ITT) population, and no new systemic antibacterial treatment for gram-positive ABSSSIs. In both trials, the key secondary objectives included the following: clinical response 48 to 72 hours after study-drug initiation, based on measurements of ABSSSI lesion size (≥ 20% reduction in lesion area); clinical efficacy at day 14 or 15 after study-drug initiation (end of treatment [EOT] visit), based on lesion size, local signs, temperature, and receipt of nonstudy antibiotics; and clinical efficacy at the day 28 short-term follow-up (SFU) visit, based on lesion size, local signs, temperature, and receipt of nonstudy antibiotics. Patients assigned to dalbavancin received a 1,000 mg dose on day 1 followed by a 500 mg dose on day 8 with a possible switch to oral placebo (if switching criteria were met), for a total duration of 14 days. Patients assigned to IV vancomycin received at least 3 days of therapy with the option to switch to oral linezolid to complete 10 to 14 days of therapy. For both treatment arms, the total course of therapy (IV and oral) was 14 days. Treatment was initiated on day 1 and efficacy and safety assessments took place on days 2, 3, 4, and 8, as well as at the EOT visit (day 14). Following the EOT visit, patients were to return for SFU at day 28 and long-term follow-up (LFU) at day 70 (2 months after the EOT visit).

The DUR001-303 study (N = 698) was a phase III, multicentre, 1:1 randomized, DB, noninferiority study designed to compare single-dose versus 2-dose IV dalbavancin regimens in patients with known or suspected gram-positive ABSSSIs. The primary objective of this study was to compare the efficacy of treatment with a single dose of dalbavancin 1,500 mg to treatment with a 2-dose regimen of dalbavancin (1,000 mg on day 1 followed by 500 mg on day 8) 48 to 72 hours after the initiation of treatment. Clinical response was defined as the patient being alive, not receiving rescue therapy for ABSSSI, and having at least a 20% decrease in lesion area. The secondary objectives of this study were clinical status at day 14 or 15 (EOT visit) and day 28 (± 2 days) after study-drug initiation, and safety. Other objectives looked at health care resource use, including length of stay (LOS) in the hospital. Patients in the single-dose group received a single dose of IV dalbavancin on day 1 and a dalbavancin-matching placebo on day 8. Patients randomly assigned to the 2-dose dalbavancin group received the first dose of dalbavancin on day 1 and the second dose of dalbavancin on day 8. Treatment was initiated on day 1, and efficacy and safety assessments took place on day 3 to 4, day 8, day 14 or 15 (EOT visit), and day 28 (final visit).

The VER001-9 study (N = 854) was a phase III, multicentre, 2:1 randomized, DB, noninferiority study that aimed to determine whether dalbavancin is noninferior to IV linezolid (with a possible switch to oral linezolid) in adults with complicated skin and skin structure infections (cSSSIs) due to gram-positive pathogens, based on clinical response, defined as survival status, temperature, and no rescue therapy. The primary objective of the VER001-9 study was to compare the clinical efficacy and safety of dalbavancin (2-dose regimen) with that of a linezolid regimen in the treatment of adults with cSSSIs due to gram-positive pathogens. A clinical response of success was defined as sufficient resolution of the local and systemic signs and symptoms of skin and skin structure infections (SSSIs) such that the patient did not receive new systemic antibacterial treatment for SSSI. Additional objectives included hospital use and LOS. Treatment was initiated on day 1, and efficacy and safety assessments took place on day 4, day 8, in the 3 days following treatment completion on day 14 (EOT visit), day 28 (test of cure [TOC] visit), and on day 39 (LFU).

In the DISCOVER 1, DISCOVER 2, and DUR001-303 studies, patients were 18 to 85 years of age with a known or suspected ABSSSI (major cutaneous abscess, surgical-site or traumatic-wound infection, or cellulitis), accompanied by at least 75 cm² of erythema, at least 2 signs of ABSSSI (purulent drainage or discharge, fluctuance, heat or localized warmth, tenderness on palpation, and swelling or induration), at least 1 systemic sign of infection, and infection severity requiring a minimum of 3 days of IV therapy. The most common infection type was cellulitis (47.3% to 54.2% of patients), followed by major cutaneous abscess (25.0% to 30.2%) and wound infection (18.2% to 26.4%). Of patients with a known pathogen, 16.7% to 28.8% had a MRSA infection and 41.8% to 58.0% had an MSSA infection.

In the VER001-9 study, inclusion and exclusion criteria were similar to the other 3 trials, although no threshold was set for the size of erythema and the threshold for increased white blood cells (WBCs) was > 10,000/mm³. Major cutaneous abscess was the most common infection type (30.4% to 33.3%), followed by cellulitis (27.5% to 29.7%) and wound infection (19.2% to 21.2%), and 62.5% to 64.3% of patients were hospitalized at study entry. Of patients with a known pathogen, 88.8% to 90.6% had a Staphylococcus aureus infection and 51% had a MRSA infection.

Efficacy Results

The key outcomes from the RCTs are summarized in Table 2. Overall, the 2-dose regimen of dalbavancin was considered noninferior to the comparator regimens, and single-dose dalbavancin was noninferior to the 2-dose regimen in terms of clinical response at 48 to 72 hours across the pivotal trials and at day 28 in the VER001-9 study.

Clinical Response or Success

In the DISCOVER 1 study, the clinical response rate at 48 to 72 hours was 83.3% in the dalbavancin group and 81.8% in the vancomycin group (treatment difference = 1.5%; 95% confidence interval [CI], –4.6% to 7.9%). The lower limit of the 95% CI for the treatment difference was higher than the pre-specified noninferiority margin of –10%. Therefore, the noninferiority of dalbavancin to vancomycin was concluded. The proportion of patients in the ITT population with clinical success at the EOT visit (day 14) was similar between groups, with 81.9% in the dalbavancin group and 86.7% in the comparator group, for a between-group difference of –4.8% (95% CI, –10.7% to 1.3%). The proportion of patients with clinical success at day 28 was 83.7% for dalbavancin and 88.1% for the comparator (treatment difference = –4.4%; 95% CI, –10.1% to 1.4%).

In the DISCOVER 2 study, clinical response rate at 48 to 72 hours was 76.8% in the dalbavancin group versus 78.3% in the vancomycin group (treatment difference = –1.5%; 95% CI, –7.4% to 4.6%). The lower limit of the 95% CI for the treatment difference was higher than the pre-specified noninferiority margin of –10%. Therefore, the noninferiority of dalbavancin to vancomycin was concluded. In addition, clinical response at the EOT visit (day 14 or 15) in the clinically evaluable (CE) population was consistent with the ITT population, with a treatment difference of 2.8% (95% CI, –6.7% to 0.7%) in favour of the comparator. At the EOT visit (day 14), 88.7% in the dalbavancin group and 85.6% in the comparator group had achieved clinical success, for a between-group difference of 3.1% (95% CI, –1.8% to 8.0%). At day 28, 88.1% in the dalbavancin group and 84.5% in the comparator group had achieved clinical success, for a between-group difference of 3.6% (95% CI, –1.1% to 8.9%).

Sensitivity analyses conducted in both the DISCOVER 1 and DISCOVER 2 studies were consistent with the results of the primary analysis.

In the DUR001-303 study, the clinical response rate at 48 to 72 hours was 84.2% in the 2-dose arm of dalbavancin and 81.4% in the single-dose arm, with a between-group difference of 2.9% (95% CI, –8.5% to 2.8%). The lower limit of the 95% CI for the treatment difference was higher than the pre-specified noninferiority margin of –10%. Therefore, the noninferiority of 1-dose dalbavancin to 2-dose dalbavancin was concluded. At the EOT visit (day 14) and at day 28, the treatment difference for the single-dose versus the 2-dose regimens was –0.9% (95% CI, –6.3% to 4.6%) and 0.6% (95% CI, –6.0% to 4.8%), respectively.

Clinical response at 48 to 72 hours across subgroups (MRSA versus MSSA; infection type) showed similar clinical response rates across all trials and groups.

In the VER001-9 study, clinical response was assessed at the EOT (day 14) and TOC (day 28) visits in the ITT population of the study. The clinical response rate at the TOC visit, the primary end point, was 76.5% in the dalbavancin arm and 82.7% in the linezolid arm. The treatment difference was –6.2% (95% CI, –12.03% to –0.27%). The lower limit of the 95% CI for the treatment difference of –6.5% remained above the pre-specified noninferiority margin of –12.5%. Therefore, the noninferiority of dalbavancin to linezolid was claimed. In terms of clinical response at the EOT visit, results were consistent with those observed at 48 to 72 hours, with 80.6% of patients treated with dalbavancin and 86.9% of patients treated with linezolid deemed to have achieved clinical success in the ITT analysis. Clinical response was also evaluated in the CE population at the TOC visit (day 28), and results were consistent with that of the ITT population (treatment difference = –2.21%; 95% CI, –7.28% to 2.86%).

Clinical Failure at EOT

In the DISCOVER 1 study, 38 patients (13.2%) in the dalbavancin group and 29 (10.2%) patients in the vancomycin group in the ITT population were clinical failures at the EOT visit (day 14). The most commonly reported reasons for clinical failure were that local signs of fluctuance and localized heat or warmth had not resolved (84.2% of patients on dalbavancin and 79.3% of patients on vancomycin), local signs of tenderness on palpation and swelling or induration were worse than mild (23.7% and 34.5%, respectively), and the patient received a new nonstudy systemic antibacterial treatment (34.2% and 13.8%, respectively).

In the DISCOVER 2 study, 32 (8.6%) patients in the dalbavancin group and 33 (9.0%) patients in the vancomycin group in the ITT population were clinical failures at the EOT visit. The most commonly reported reasons for clinical failure at the EOT visit in both treatment groups were the same as in the DISCOVER 1 study: local signs of fluctuance and localized heat or warmth had not resolved (53.1% of those on dalbavancin and 60.6% of those on vancomycin); local signs of tenderness on palpation and swelling or induration were worse than mild (34.4% and 48.5%, respectively); and the patient received a new nonstudy systemic antibacterial treatment (28.1% and 42.4%, respectively).

In the DUR001-303 study, 42 (12.0%) patients on single-dose dalbavancin and 36 (10.3%) patients on 2-dose dalbavancin in the ITT population were clinical failures at the EOT visit. The most common reasons for clinical failure were lesion size that did not decrease from baseline (73.8% on single-dose dalbavancin and 66.7% on 2-dose dalbavancin), local signs of tenderness on palpation and swelling or induration that were worse than mild (21.4% and 19.4%, respectively), and the patient received a new nonstudy systemic antibacterial treatment (21.4% and 16.7%, respectively).

Harms Results

A summary of the key harms reported in the RCTs are summarized in Table 2.

In the DISCOVER 1 and DISCOVER 2 studies, the incidence of treatment-emergent adverse effects (TEAEs) was lower in the dalbavancin group than in the vancomycin group (34.9% versus 39.4% in the DISCOVER 1 study and 31.3% versus 36.8% in the DISCOVER 2 study). Across the pivotal trials, the number and type of TEAEs were similar between groups, with the most commonly reported TEAEs being headache, nausea, hypertension, and rash. The TEAEs in the DISCOVER 2 study were similar to those in the DISCOVER 1 study. In the DUR001-303 study, the incidence of TEAEs was similar with the single-dose and 2-dose dalbavancin regimens (20.1% versus 19.9%). The most common (≥ 2%) TEAEs in the single-dose and 2-dose treatment groups, respectively, were nausea (3.4% versus 2.0%), headache (1.7% versus 1.2%), and vomiting (1.7% versus 0.9%). In the VER001-9 study, the most commonly (≥ 2%) reported TEAEs throughout the entire study period for the dalbavancin and linezolid groups, respectively, were nausea (3.2% versus 5.3%), diarrhea (2.5% versus 5.7%), and headache (1.9% versus 1.8%).

In the DISCOVER 1 study, treatment-emergent serious adverse events (SAEs) were less commonly reported in the dalbavancin group than in the comparator group (1.8% versus 4.2%, respectively). In the DISCOVER 2 study, the number of treatment-emergent SAEs was similar in the dalbavancin and comparator groups (3.3% versus 3.8%, respectively). In the DUR001-303 study, the percentage of patients with SAEs was similar in the single-dose and 2-dose treatment groups (2.0% versus 1.4%). In the VER001-9 study, the rate of SAEs was similar in the dalbavancin and linezolid groups (7.5% versus 8.4%).

The proportion of patients who discontinued treatment due to adverse events (AEs) was similar in the 2 treatment groups across the RCTs (1.8% or less in all groups).

The number of deaths was similar between groups in all trials (0.7% or less of each group), except in the DISCOVER 1 study, where there were deaths in 1.8% of the vancomycin group and none in the dalbavancin group. The most common notable harms were infusion-related reactions (1.8% or less of each group), rash or hypersensitivity reactions (0.8% to 2.1% of each group in the pivotal trials), and hepatic AEs (0.8% to 2.1% in the DISCOVER 1 and DISCOVER 2 studies).

Table 2: Summary of Key Results From Pivotal and Protocol-Selected Studies

AE = adverse event; CI = confidence interval; DAL = dalbavancin; EOT = end of treatment; ITT = intention-to-treat; LZD = linezolid; NR = not reported OR = odds ratio; SAE = serious adverse event; VAN = vancomycin; WDAE = withdrawal due to adverse event.

^aThis was the primary end point. The 95% CI was calculated using the Miettinen and Nurminen method, without adjustment in the DUR001-303 study and with adjustment for the randomization stratification factor of presence or absence of fever at baseline in the DISCOVER 1 and DISCOVER 2 studies.

^bPrimary end point.

Note: Linezolid in refers to oral linezolid in the pivotal trials, and linezolid in the VER001-9 study refers to IV linezolid (with a possible switch to oral linezolid). The noninferiority margin was –10% for the pivotal trials and –12.5% for the VER001-9 trial, and there was no adjustment for multiple outcomes. In the DISCOVER 1 and DISCOVER 2 studies, the OR was stratified for the presence or absence of fever at baseline, and the 95% CI for the OR was calculated using the Cochran-Mantel-Haenszel method.

Source: Clinical Study Reports.^7-20

Critical Appraisal

Internal Validity

The noninferiority design was adequately powered, and the threshold used (–10% to –12.5%) was justified. According to the clinical expert consulted by CADTH, the 10% noninferiority margin is somewhat wide, but a loss of approximately 5% in clinical benefit would not be too concerning for the treatment of ABSSSI. Also, the expert indicated that although vancomycin is expected to be efficacious in this population (especially in those with MRSA), linezolid is a bacteriostatic rather than bactericidal drug and may not be as efficacious as other available options. Although the dosing of vancomycin aligned with the recommended dose for this population, it may not have been the most appropriate comparator, given that majority of patients with Staphylococcus aureus infections were MSSA and not MRSA. According to the clinical expert, more appropriate drugs for patients with MSSA include cephalosporins, such as IV ceftriaxone or oral cephalexin.

The primary outcome in the DUR001-303 study corresponds to the primary efficacy end point recommended in the FDA draft guidance for the development of drugs for ABSSSIs,¹ and the primary outcomes in the DISCOVER 1 and DISCOVER 2 studies were also similar. The primary outcomes for the DISCOVER 1 and DISCOVER 2 studies also considered resolution of fever and did not require a decrease in lesion size for clinical response, which is more aligned with the way patients are assessed in clinical practice, as the requirement for a 20% reduction in lesion area (part of the FDA-recommended outcome) is arbitrary, and fever, WBC count, and pain are typically assessed. The clinical expert noted that the outcomes for the VER001-9 study were less well defined. Overall, the expert considered the efficacy outcomes and follow-up in the 4 studies to be appropriate for guidance on the use of dalbavancin in patients with ABSSSIs.

Treatment compliance was higher with dalbavancin than with other IV therapies and with the single-dose than with the 2-dose regimen. The use of concomitant antibiotic therapy was similar among groups across all trials; hence, bias toward treatment outcome is low. Subgroup analyses did not include statistical testing and were limited in number, so results should be considered exploratory. Overall, the subgroup results across all trials were similar to those in the primary analysis.

External Validity

The clinical expert consulted by CADTH agreed that the baseline patient characteristics in the pivotal trials and the VER001-9 study were reflective of the patients with ABSSSIs they encounter in Canadian clinical practice. Although the majority of patients in each study were enrolled at trial sites in the US and Europe, the population enrolled in the trial was consistent with the population expected to be treated in Canadian clinical practice. In general, lesion size was quite large across all trials, (e.g., approximately 300 cm² up to approximately 750 cm²), which is indicative of more severe ABSSSI; this may limit generalizability to populations with smaller lesion sizes. Furthermore, the subgroup analyses (e.g., bacteremia, infection type, MRSA status) underwent no statistical comparisons and had smaller sample sizes, which limit the generalizability to a broader population. According to the clinical expert, the concomitant medications used in the trial were reflective of those encountered in Canadian clinical practice.

Indirect Comparisons

Description of Studies

The published indirect treatment comparison (ITC) included in the sponsor’s submission included a systematic review, and the authors used a Bayesian network meta-analysis (NMA) to evaluate the relative clinical and safety efficacy of IV antibiotics for the treatment of ABSSSI in adults. The efficacy outcomes of interest were clinical treatment success and microbiological success, and the safety outcomes were discontinuation due to AEs or SAEs, any AE or any SAE experienced by the patients, and all-cause mortality.

Efficacy Results

The NMA shows no evidence of a difference in clinical success between dalbavancin and vancomycin (odds ratio [OR] = 0.99; 95% credible interval [CrI], 0.68 to 1.51), linezolid (OR = 0.69; 95% CrI, 0.41 to 1.00), or daptomycin (OR = 1.05; 95% CrI, 0.61 to 2.10).

Safety Results

The ITC reported that dalbavancin is associated with a lower likelihood of AEs than linezolid, a lower likelihood of SAEs than vancomycin and daptomycin, and a lower risk of all-cause mortality than vancomycin, linezolid, and tigecycline. However, the CrIs were wide, particularly for SAEs and mortality.

Critical Appraisal

The ITC had several limitations, including high clinical heterogeneity in study design (e.g., patient selection criteria, definitions of response, and timing of assessment) and in patient characteristics. High statistical heterogeneity was also present in the pairwise meta-analyses. The structure of the network was star-shaped with 1 closed loop, with some contrasts represented by 1 or 2 trials. Because of the small number of trials included in the NMA, the ability to estimate between-trial variance was limited. As a result, analyses with uncommon events, such discontinuations due to AEs or SAEs, would likely produce imprecise estimates. Furthermore, there were limitations related to the lack of reporting of certain items that would have better informed the certainty of the indirect evidence; the authors did not report risk of bias when pooling the studies, and did not adequately report sensitivity and subgroup analyses to investigate the root of heterogeneity or conduct a meta-regression that would adjust for effect modifiers that could influence the results. Overall, there was substantial uncertainty around the ITC because of imprecision and risk of bias.

Other Relevant Evidence

Description of Studies

The ENHANCE²¹ and ADVANCE²² pre-post pragmatic studies provided further evidence on the efficacy (hospital LOS and hospital admission rate) and safety of dalbavancin in patients with ABSSSI.

The ENHANCE study²¹ was conducted to estimate the difference in infection-related total hospital admission days during initial care (the period between enrolment and 10 to 14 days after treatment initiation) and follow-up (the period between the end of initial care to 30 days after treatment initiation) between patients with ABSSSIs who received care before (pre-period) and after (post-period) the implementation of a critical pathway that was developed for the management of patients with ABSSSIs who were admitted to the hospital. The intervention was the critical pathway, which comprised the identification of patients who met criteria that were developed based on guidelines on the management of ABSSSIs in the hospital setting and outpatient parenteral antibiotic therapy; and the administration of dalbavancin to patients who met the criteria and were subsequently discharged to an outpatient setting at the discretion of the treating physician. During the pre-period, only the first component of the critical pathway was implemented; patients received usual care for ABSSSIs, which was defined as the antibiotic with coverage for the known or suspected gram-positive infection selected by the treating physician or site. During the post-period, both components of the critical pathway were implemented, with the initiation of 1,500 mg dalbavancin as a single IV dose over 30 minutes in a new set of patients enrolled based on the same guideline-based criteria used in the pre-period. For both the pre-period and the post-period, patients were assessed at baseline (date of enrolment), 48 to 72 hours after the date of enrolment or discharge, 10 to 14 days after the date of enrolment, and 44 to 51 days after date of enrolment.

The ADVANCE study²² was conducted to estimate the difference in hospital admission rates at initial care between patients with ABSSSIs who received care before (pre-period) and after (post-period) implementation of the critical pathway. The design of the ADVANCE study was similar to that of the ENHANCE study,²¹ with a few differences. First, the ENHANCE study recruited patients with ABSSSIs who were admitted to the ER, whereas the ADVANCE study recruited patients with ABSSSIs who presented to the ER. Therefore, patients enrolled in the ADVANCE study received dalbavancin at the point of care in the ER and were subsequently sent home at the discretion of the treating physician. Finally, patients enrolled in the ADVANCE study had an additional assessment 24 hours after enrolment in the post-period, whereas patients in the pre-period were not assessed 48 to 72 hours after enrolment.

Efficacy Results

The mean difference between the pre-period and post-period in total infection-related LOS during the entire ENHANCE study was 1.6 days (95% CI, 0.6 to 2.6 days; P = 0.003), in favour of the post-period. The results of the secondary analysis, after adjustment for age and immunocompromised status, were generally consistent with those of the primary analysis. The results from the analysis that included time spent by patients in prolonged (greater than 1 day) observation and patients who completed the study were generally consistent with the observed difference in the primary analysis for inpatients only for the entire duration of the study.

The ABSSSI-related hospital admission rate at initial care in the ADVANCE study was 38.5% and 17.6% in the pre-period and post-period groups, respectively. The difference between the pre-period and post-period groups in the ABSSSI-related hospital admission rate at initial care was in favour of the post-period group (P < 0.001). The results of the secondary analysis, after adjustment for age, race, insurance type, prior resource use, and systemic inflammatory response syndrome (SIRS) score, were generally consistent with the primary analysis.

All secondary outcomes were exploratory, as the studies were not powered for secondary outcomes, and no adjustments for multiple comparisons were made. In the ADVANCE study, the 12-Item Short Form Survey mental and physical health component scores at day 14, relative to baseline, were used to assess health-related quality of life (HRQoL). The trial found no difference between the 2 assessment groups.

Harms Results

In the ENHANCE study, 3 (6.2%) and 20 (47.6%) patients in the pre-period and post-period groups, respectively, reported at least 1 AE. The most common (> 5%) AE was pyrexia, which occurred in 3 patients (7.1%) in the post-period group. No patients discontinued the study due to any AE, and no deaths were reported in the study. A total of 1 (2.1%) and 3 (7.1%) patients in the pre-period and post-period groups, respectively, reported at least 1 SAE.

In the ADVANCE study, 22 patients (14.1%) and 68 (44.4%) patients in the pre-period and post-period groups, respectively, reported at least 1 AE. The most common (> 5%) AEs were cellulitis and diarrhea, both of which occurred in 8 (5.2%) patients in the post-period group. In the post-period group, no patients discontinued the study due to an AE, and no deaths were reported; in the pre-period group, 2 deaths were reported, 1 due to congestive cardiac failure and 1 due to a road traffic accident. A total of 11 (7.0%) and 16 (10.5%) patients in the pre-period and post-period groups, respectively, reported at least 1 SAE; most common SAEs were not reported.

Critical Appraisal

Internal Validity

The interpretation of the efficacy and safety results from both the ENHANCE and ADVANCE studies may be limited due to the pre-post pragmatic (nonrandomized and open-label) study design. Each site enrolled patients consecutively into both the pre-period and post-period groups and, as a result, the study lacked a concurrent control and patients were not randomized to a treatment. The study design may have also introduced the risk for time-related confounders, such as changes between the pre-period and the post-period in local antimicrobial resistance patterns and in the site-specific and physician-specific approach to usual care for the treatment of ABSSSI. In comparison to the post-period, in which site staff were trained on the critical pathway and the post-period protocol, staff were only trained on the pre-period protocol for the pre-period. As a result, it is uncertain how much of the treatment effect observed in the post-period group was due to the efficacy of dalbavancin, rather than the effectiveness of the entire critical pathway.

Of note, treating physicians were not trained on the protocol (e.g., study objectives, exclusion and inclusion criteria, intervention, and comparator treatment, and outcomes of interest) in the pre-period to limit performance bias when selecting usual care for the treatment of ABSSSIs. This was particularly important because patients were discharged or sent home from the ER at the discretion of the treating physician during the pre-period and the post-period, which may have had an impact on the LOS and hospital admission rate.

The interpretation of the results may be further limited by missing data. Because the analyses were conducted using the observed-cases approach and because of a relatively high and unbalanced study withdrawal rate due to patients lost to follow-up — 4 (8.3%) patients in the pre-period and 9 (20.9%) patients in the post-period in the ENHANCE study — the direction of attrition bias is uncertain. The interpretation of the results in the ADVANCE study may be further limited by the 104 (68.0%) patients in the post-period group having received concomitant therapy with other antibiotics at initial care, thereby removing the difference in the treatment received between the pre-period and the post-period.

External Validity

To optimize the selection of patients with ABSSSIs that best represented real-world outpatients, both the ENHANCE and ADVANCE studies used minimal inclusion and exclusion criteria. Although this would support generalizability to clinical practice, there may still be potential differences between study sites and Canadian practice in the approach to usual care for the treatment of ABSSSI, according to guidelines on the management of patients with ABSSSIs, in the acute-care setting and between the recommended treatment options or regimens and local antimicrobial resistance patterns. Further, the generalizability of the results to the population of patients with ABSSSIs in Canada may be limited, given that all patients were sourced from 1 hospital in the ENHANCE study.

Conclusions

In patients with ABSSSIs, dalbavancin demonstrated noninferiority to vancomycin (with a possible switch to oral linezolid) and IV linezolid (also with a possible switch to oral linezolid), with similar rates of clinical response and clinical success. The dalbavancin single-dose treatment regimen was also shown to be noninferior to the 2-dose treatment regimen. A key limitation from these trials was that the comparators did not fully reflect the standard of care typically used in Canadian practice for patients with ABSSSIs, as identified by the clinical expert, as there were no comparators specific to the treatment of MSSA infections, which may reduce the generalizability of the findings. The ITC compared dalbavancin with vancomycin, linezolid, and daptomycin, but the efficacy and safety results were uncertain because of the heterogeneity across trials and the sparse safety data. The pragmatic studies showed an improvement in total infection-related LOS (the ENHANCE study) and the hospital admission rate at initial care (the ADVANCE study) after the introduction of a dalbavancin care pathway; however, the pre-post study design and use of a pathway intervention limit interpretation. The AEs reported in the RCTs did not give rise to any safety concerns (in either the single-dose or 2-dose regimens), and the safety profile of dalbavancin was similar and possibly better than that of vancomycin and linezolid for patients with ABSSSIs.

Introduction

Disease Background

ABSSSIs are bacterial infections of skin and associated tissues, also referred to as SSTIs. They are caused by the penetration of bacterial skin flora into soft tissues. ABSSSIs include cellulitis, erysipelas, wound infection, and major cutaneous abscesses.¹ ABSSSIs are defined by lesion size²³ (at least 75 cm²), as measured by the area of redness, edema, or induration.² ABSSSIs may be acquired in either the community or hospital setting, and are a frequent cause of morbidity.^2-4 Staphylococcus aureus and streptococci are responsible for most simple community-acquired infections. Worldwide, the most common cause of ABSSSIs is Staphylococcus aureus, including MRSA.^2,5 ABSSSIs are a frequent cause of morbidity in both community and hospital settings.^3,4 In Canada, the estimated prevalence of ABSSSIs requiring IV antibiotics is 0.39%,^6,7 and First Nations communities are disproportionately affected.⁷ Severity ranges from nonsevere infections to severe, life-threatening infections that affect deeper tissue layers.^2,8,9 In North America, the mortality rate among patients hospitalized for SSTIs is approximately 0.5%.¹⁰ ABSSSIs caused by MRSA are associated with increased risk of mortality, longer hospital stays, and higher health care costs than non-MRSA infections.¹¹

ABSSSIs may be closed (cellulitis) or open (wound infection). The microbiologic diagnosis can be difficult, especially in cases of closed infection, where a microbiological finding cannot be obtained without invasive procedures, such as biopsy. Hence, in many cases, the microbiological cause of ABSSSIs is unknown. Microbiologic diagnosis is associated with delay because of culture growth and antimicrobial susceptibility testing; in many cases, treatment is completed without microbiologic diagnosis, and appropriateness of treatment is unknown. Among outpatients, lack of compliance with antibiotic treatment also contributes to the failure of initial therapy.¹⁴ ABSSSIs affect patient quality of life due to missed work or school, the inconvenience of inpatient treatment, emotional distress, and embarrassment.¹⁵ In Canada, ABSSSIs represent a major burden to patients; for example, it is estimated that patients hospitalized for cellulitis have a mean hospital stay of 7.1 days.^11,24

Standards of Therapy

According to the clinical expert consulted by CADTH, the current standard of therapy for ABSSSIs in Canada is a cephalosporin for MSSA infections (i.e., IV cefazolin, IV ceftriaxone, or oral cephalexin) and IV vancomycin, IV daptomycin, or oral or IV linezolid for MRSA infections (daptomycin and linezolid are rarely used). For severe ABSSSIs, treatment would include IV piperacillin plus tazobactam or IV ertapenem. According to the clinical expert consulted by CADTH, an oral route of administration is well suited for the treatment of ABSSSIs, as oral antibiotics are inexpensive, easy to use, and can be administered at home. However, the potential lack of adherence to oral antibiotics is still a concern. Treatment begins in a step-wise manner, beginning with microbiologic diagnosis (if possible), surgery (if needed), admission to hospital (if needed), IV or oral antibiotics, and follow-up with inpatient or outpatient care to monitor response to treatment. The physician will decide on the duration of treatment and will monitor for toxicities and AEs. The microbiologic diagnosis of ABSSSIs guides the course of treatment (e.g., treatment of MRSA is different than treatment for MSSA). However, treatment of ABSSSIs is difficult because the pathogen type is not always known. This may also contribute to antimicrobial resistance, given that treatment is often initiated before microbiologic diagnosis. Antimicrobial resistance affects the prognosis and duration of treatment. Therefore, a goal of therapy is to minimize antimicrobial resistance, both in spectrum of action and duration of treatment. The prognosis of ABSSSIs is generally cure, but some patients may have recurrent ABSSSIs, which is more likely in patients with lymphedema or other skin diseases. According to the clinical expert, the goals of therapy are prevention of mortality, improvement in accessibility and compliance, and reduction in the risk of IV-related complications, hospital visits, and health care costs.

According to the clinical expert, treatment is generally initiated in ERs, and patients are referred to outpatient infusion clinics where available. However, some patients receive care in the ER only. Given the lack of randomized trials, many factors are unknown about the treatment of ABSSSIs, including when to switch from an IV to oral antibiotic, the ideal duration of treatment, and the efficacy of various treatments, including penicillin G.

The wider impacts of antimicrobial resistance are outlined in a report by the Council of Canadian Academies Expert Panel on the Potential Socio-Economic Impacts of Antimicrobial Resistance in Canada.²⁵ The panel estimated that in 2018, there were 5,400 deaths directly attributable to antimicrobial resistance, an additional $1.6 billion in health care system costs from expenses related to antimicrobial resistance, and a reduction in gross national product by $2.8 billion related to the impact on productivity.²⁵

Drug

Dalbavancin is a semisynthetic bactericidal lipoglycopeptide antibiotic used against susceptible strains of gram-positive bacteria.¹⁶ Its mechanism of action is the interruption of cell-wall synthesis; the drug binds to the terminal D-alanyl-D-alanine of the stem peptide in nascent cell-wall peptidoglycan, which prevents cross-linking of disaccharide subunits, resulting in bacterial cell death.¹⁶

Dalbavancin is indicated for the treatment of adults with ABSSSI caused by susceptible isolates of the following gram-positive micro-organisms: Staphylococcus aureus (including methicillin-susceptible and methicillin-resistant strains), Streptococcus pyogenes, Streptococcus agalactiae, Streptococcus dysgalactiae, the Streptococcus anginosus group (including Streptococcus anginosus, Streptococcus intermedius, and Streptococcus constellatus), and E. faecalis (vancomycin-susceptible strains).¹⁶ The Health Canada recommended dosage is 1,500 mg, administered either as a single dose or as a 1,000 mg dose followed 1 week later by a 500 mg dose, all administered by IV infusion over 30 minutes. The sponsor’s requested reimbursement criteria for dalbavancin aligns with the Health Canada indication. Dalbavancin underwent a priority review by Health Canada and received a Notice of Compliance on May 5, 2021. Dalbavancin has no other Health Canada–approved indication and has not previously been reviewed by CADTH.

Key characteristics of treatments used for ABSSSIs are presented in Table 3.

Stakeholder Perspectives

Patient Group Input

No patient group input was received for this review.

Clinician Input

Input From Clinical Experts Consulted by CADTH

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

Unmet Needs

According to the clinical expert, current IV treatment is expensive and complicated, requiring nursing effort and PICC line insertion, and increasing the risk of IV-line complications, repeated visits, automated IV pumps, and/or home IV visits. IV vancomycin is toxic and requires therapeutic drug monitoring. Sometimes, given the difficulty in obtaining a culture specimen to identify the pathogen, the diagnosis may be incorrect and delay the treatment process. There is no accurate diagnostic biomarker or imaging study to assist in diagnosis. In addition, many patients have red skin on the legs caused by other diseases, so ABSSSI treatment is often administered unnecessarily. Response to treatment is often slow, which leads to prescribers often treating for durations longer than necessary. The longer treatment duration is further perpetuated in the ER settings, where a different doctor assesses the skin each day without comparison to previous appearances. Compliance is also an issue with oral antibiotics, especially in people who inject drugs who may fail to return or fail to comply with oral antibiotics; patients who inject drugs are at higher risk of ABSSSI due to injection. In terms of access, people in remote areas or Indigenous communities may have poor or limited access to IV treatment and may face geographic barriers accessing treatment after discharge.

Place in Therapy

According to the clinical expert, dalbavancin could change the current treatment paradigm, given its unique pharmacokinetics and ease of use compared to other treatments. Dalbavancin would be considered a first-line treatment (unless oral antibiotics were available and suitable for use by the patient) and would replace current IV antibiotic treatments for ABSSSIs caused by gram-positive bacteria. In the opinion of the clinical expert, there would be no reason to try other IV treatments before dalbavancin. Given that dalbavancin does not require repeat IV infusion, PICC line placement, or therapeutic drug monitoring, it is much easier to access than other treatments and may lead to higher patient satisfaction. In addition, dalbavancin could prevent the repeat visits, hospital admission, and poor compliance observed with other therapies.

Patient Population

According to the clinical expert, patients best suited for treatment with dalbavancin include those with MRSA infection, those who inject drugs, those identified by physician assessment at presentation, and candidates for home IV who are clinically stable, have mild-to-moderate infection, are competent, and have access to a nearby hospital if they worsen after dalbavancin injection. Patients least suited for dalbavancin include those with infections that require surgical debridement.

Assessing Response to Treatment

According to the clinical expert, response in a trial setting is best assessed using the FDA outcome criteria at 48 to 72 hours: alive, a 20% reduction in affected area, and no rescue treatment. In clinical practice, there are typically daily assessments, which may not be necessary; also, assessment may be highly variable between physicians (e.g., determining time to resolution of redness). Patients can be re-treated, typically around day 8, if there is no response. However, physicians should investigate reasons for nonresponse and adjust the treatment plan accordingly.

Discontinuing Treatment

Dalbavancin is a single-dose or 2-dose regimen, so criteria for discontinuation are not applicable in most cases, according to the expert.

Prescribing Conditions

Dalbavancin could be given in any setting in which there is IV access (e.g., ER, long-term care, hospital, clinic, or home visit). Dalbavancin does not require a specialist for prescription, although it may be beneficial to have antimicrobial stewardship involved to ensure appropriate use. Treatment should not be used off-label without trial evidence. The clinical expert added that trials should be conducted in patients with osteomyelitis, endocarditis, abscess, prosthetic joint infection, pneumonia, and meningitis.

Additional Considerations

Antimicrobial treatment is associated with selection toward antimicrobial resistance in human flora. This antimicrobial resistance is transmitted among humans, animals, and the environment, and is associated with enormous cost to the Canadian health system and mortality. Therefore, antimicrobial use must be minimized, both in spectrum of action and duration of treatment, to reduce antimicrobial resistance. Furthermore, there has been little new drug development in antimicrobials.

Dalbavancin has potential harms, including an increased risk of antimicrobial resistance due to prolonged duration of selection toward antimicrobial resistance among gut flora (with fecal excretion for up to 70 days) and the inability to withdraw the drug if allergic reaction occurs. Because of the associated convenience, the availability of dalbavancin could increase unnecessary treatment by increasing treatment of the wrong diagnosis, increasing IV treatment when oral treatment would be adequate, and increasing unnecessary antimicrobial prophylaxis. Discontinuation of close patient follow-up due to prolonged dosing intervals could lead to a lack of recognition of worsening; patients could be discharged on dalbavancin pending culture results, which the provider might fail to review because the patient is discharged. If the culture reveals gram-negative infection, dalbavancin would not be a suitable treatment.

Clinician Group Input

No clinician group input was received for this review.

Drug Program Input

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

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

ABSSSI = acute bacterial skin and skin structure infection; CDEC = CADTH Canadian Drug Expert Committee; ER = emergency room; SAE = serious adverse event.

Clinical Evidence

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

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of dalbavancin injection (1,500 mg, administered as a single dose or as 1 dose of 1,000 mg followed by a second dose of 500 mg 1 week later) for the treatment of adults with ABSSSI caused by susceptible isolates of the following gram-positive micro-organisms: Staphylococcus aureus (including methicillin-susceptible and methicillin-resistant strains), Streptococcus pyogenes, Streptococcus agalactiae, Streptococcus dysgalactiae, Streptococcus anginosus group (including Streptococcus anginosus, Streptococcus intermedius, Streptococcus constellatus), and E. faecalis (vancomycin-susceptible strains).

Methods

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

Table 5: Inclusion Criteria for the Systematic Review

ABSSSI = acute bacterial skin and skin structure infection; AE = adverse event; HRQoL = health-related quality of life; MRSA = methicillin-resistant Staphylococcus aureus; MSSA = methicillin-susceptible Staphylococcus aureus; RCT = randomized controlled trial; SAE = serious adverse event; SIRS = systemic inflammatory response system; vs. = versus; WDAE = withdrawal due to adverse event.

^aSIRS criteria include tachycardia (heart rate > 90 beats/min), tachypnea (respiratory rate > 20 breaths/min), fever or hypothermia (temperature > 38°C or < 36°C), and leukocytosis, leukopenia, or bandemia (WBCs > 1,200/mm³, < 4,000/mm³, or bands ≥ 10%).

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the Peer Review of Electronic Search Strategies (PRESS) checklist.³¹

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946—) via Ovid and Embase (1974—) via Ovid. All Ovid searches were run simultaneously as a multifile search. Duplicates were removed using Ovid deduplication for multifile searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were Xydalba (dalbavancin) and ABSSSIs. The following clinical trials registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

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

The initial search was completed on May 13, 2022. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on September 28, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.³² Included in this search were the websites of regulatory agencies (FDA and European Medicines Agency). Google was used to search for additional internet-based materials. Refer to Appendix 1 for more information on the grey literature search strategy.

These searches were supplemented by reviewing bibliographies of key papers and through contacts with appropriate experts. In addition, the sponsor was contacted for information regarding unpublished studies.

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

A focused literature search for ITCs dealing with ABSSSIs was run in MEDLINE All (1946–) on May 12, 2022. No limits were applied to the search.

Findings From the Literature

A total of 4 studies were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 6. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 6: Details of Included Studies

ABSSSI = acute bacterial skin and skin structure infection; AE = adverse event; CD4 = cluster of differentiation 4; DB = double-blind; EOT = end of treatment; FV = first visit; ITT = intention-to-treat; LFU = long-term follow-up; ME = microbiologically evaluable; microITT = microbiological intention-to-treat; MRSA = methicillin-resistant Staphylococcus aureus; MSSA = methicillin-sensitive Staphylococcus aureus; RCT = randomized controlled trial; SFU = short-term follow-up; SSSI = skin and skin structure infection; TOC = test of cure; WBC = white blood cell.

Note: Day 14 after the initial dose of treatment was considered the EOT visit in all trials. Day 28 is the SFU for the DISCOVER 1 and DISCOVER 2 studies, the final visit for the DUR001-303 study, and the TOC visit for the VER001-9 study.

^aResource use categories included any additional visits (including urgent care), any additional procedures, any additional tests, any home visits or nursing care, and any ER visits. The percentage of patients in each category is reported.

Source: Clinical Study Reports.^17-20

Description of Studies

Four sponsor-submitted trials were included in this review: the DISCOVER 1, DISCOVER 2, DUR001-303, and VER001-9 trials. The DISCOVER 1, DISCOVER 2, and DUR001-303 studies were sponsored by Durata Therapeutics, and the VER001-9 study was sponsored by Vicuron Pharmaceuticals Inc.

The DISCOVER 1 (N = 573) and DISCOVER 2 (N = 739) studies were phase III, multicentre, 1:1 randomized, DB, noninferiority studies comparing the efficacy and safety of dalbavancin to a vancomycin and linezolid regimen in patients with known or suspected gram-positive ABSSSI. The DISCOVER 1 study had 54 sites in 7 countries (including 1 in Canada), and the DISCOVER 2 study had 86 sites in 14 countries (with no sites in Canada). The primary objective in both trials was to compare clinical efficacy 48 to 72 hours after study-drug initiation between the dalbavancin and vancomycin regimens. Clinical response was defined as no increase in lesion size and absence of pyrexia at 48 to 72 hours in the ITT population. In both trials, the key secondary objectives included the following: clinical response 48 to 72 hours after study-drug initiation, based on measurements of ABSSSI lesion size (≥ 20% reduction in lesion area); clinical efficacy day 14 or 15 after study-drug initiation (EOT visit) based on lesion size, local signs, temperature, and receipt of nonstudy antibiotics; and clinical efficacy at the day 28 SFU visit based on lesion size, local signs, temperature, and receipt of nonstudy antibiotics. Patients assigned to dalbavancin received a 1,000 mg dose on day 1 followed by a 500 mg dose on day 8, with a possible switch to oral placebo (if switching criteria were met), for a total duration of 14 days. Patients assigned to IV vancomycin received at least 3 days of therapy with the option to switch to oral linezolid to complete 10 to 14 days of therapy. Patients were stratified by the presence or absence of fever at baseline (a minimum of 25% of patients had to have fever at baseline), geographic region, and infection type (cellulitis, major abscess [to a maximum of 30% of the total study population], and traumatic-wound or surgical-site infection), using block randomization (block size of 4) with an interactive voice-activated randomization system (IVRS). For both treatment arms, the total course of therapy (IV and oral) was 14 days. Treatment was initiated on day 1, and efficacy and safety assessments took place on days 2, 3, 4, and 8 as well as at the EOT visit (day 14). Following the EOT visit, patients were to return for SFU at day 28 and LFU at day 70 (2 months after the EOT visit). Baseline assessments were performed in the 24 hours before the first dose.

The DUR001-303 study (N = 698) was a phase III, multicentre, 1:1 randomized, DB, noninferiority study designed to compare single-dose and 2-dose IV dalbavancin regimens in patients with known or suspected gram-positive ABSSSI. The DUR001-303 study had 60 sites in 11 countries, with no sites in Canada. The primary objective of this study was to compare the efficacy of a single dose of dalbavancin 1,500 mg to a 2-dose regimen of dalbavancin (1,000 mg on day 1 followed by 500 mg on day 8) 48 to 72 hours after the initiation of treatment. The secondary objectives of this study were clinical status at day 14 or 15 (EOT visit) and day 28 (± 2 days) after study-drug initiation and safety. Other objectives looked at health care resource use, including hospital LOS. Patients in the single-dose group received a single dose of IV dalbavancin on day 1, and a dalbavancin-matching placebo on day 8. Patients randomly assigned to the 2-dose dalbavancin group received the first dose on day 1 and the second dose on day 8. Treatment was initiated on day 1, and efficacy and safety assessments took place on day 3 or 4, day 8, day 14 or 15 (EOT visit), and day 28 (final visit). Baseline assessments were performed in the 24 hours before the first dose.

The VER001-9 study (N = 854) was a phase III, multicentre, 2:1 randomized, DB, noninferiority study that aimed to determine whether dalbavancin is noninferior to IV or oral linezolid treatment in adults with cSSSIs due to gram-positive pathogens, based on clinical response, defined by survival status, temperature, and use of rescue therapy. The VER001-9 study included 65 sites, including 21 in Canada, in 7 countries. Eligible patients in the VER001-9 study were randomized in a 2:1 ratio, with an IVRS, to receive treatment with IV dalbavancin or IV linezolid (possibly followed by oral linezolid). Patients were stratified by geographic location and site. The primary objective of the VER001-9 study was to compare the clinical efficacy and safety of dalbavancin with that of a linezolid regimen for the treatment of adults with cSSSIs due to gram-positive pathogens. The secondary objective was to compare the microbiological efficacy between treatment arms and to obtain dalbavancin pharmacokinetic and pharmacokinetic and/or pharmacodynamic data in patients with this disease entity. Additional objectives included hospital use and LOS. Treatment was initiated on day 1 and efficacy and safety assessments took place on day 4, day 8, in the 3 days after treatment completion (the day 14 [EOT], day 28 [TOC], and day 39 [LFU] visits). Baseline assessments were performed in the 24 hours before the first dose.

Populations

Inclusion and Exclusion Criteria

In the DISCOVER 1, DISCOVER 2, and DUR001-303 studies, patients were 18 to 85 years of age with a known or suspected ABSSSI (major cutaneous abscess, surgical-site or traumatic-wound infection, or cellulitis) accompanied by at least 75 cm² of erythema, at least 2 signs of ABSSSI (purulent drainage or discharge, fluctuance, heat or localized warmth, tenderness on palpation, and swelling or induration), at least 1 systemic sign of infection (a body temperature of 38°C or greater; WBC count above 12,000 cells/mm³; a manually performed WBC differential count with at least 10% band forms, regardless of peripheral WBC count); and infection severity requiring a minimum of 3 days of IV therapy.

In the VER001-9 study, the inclusion and exclusion criteria were similar to those in the other 3 trials, although no threshold was set for the size of erythema and the threshold for an elevated WBC count was greater than 10,000/mm³.

Baseline Characteristics

Baseline characteristics of all included studies are presented in Table 7. Overall, baseline characteristics were well balanced in the groups and studies. In the DISCOVER 1 study, most patients in the dalbavancin and vancomycin treatment groups were male (59.0% versus 60.7%) and white (91.7% versus 90.9%), and median age was 50 years in both treatment groups. Cause, type, and presentation of ABSSSIs were similar in the dalbavancin and vancomycin treatment groups; the most common ABSSSI was cellulitis (48.6% versus 49.1%). Median body mass index (BMI) was elevated in both treatment groups (27.80 kg/m² and 27.60 kg/m², respectively). The median (minimum to maximum) area of ABSSSI erythema at baseline was 333.00 cm² (25.6 to 3,400.0 cm²) and 367.75 cm² (77.6 to 3,675.0 cm²) in the dalbavancin and vancomycin treatment groups, respectively, which was larger than the minimum (75 cm²) required. SIRS criteria were met by 61.6% of patients in each treatment group. More than 80% of patients in the dalbavancin and vancomycin treatment groups had fever at baseline (81.9% versus 82.2%). Approximately 1-third of patients in each treatment group had a temperature of at least 38°C and either a WBC count above 12,000 cells/mm³ or bands of at least 10%. Staphylococcus aureus was the most common organism isolated in the dalbavancin and vancomycin treatment groups (79.7% versus 82.6%); fewer than 30% of Staphylococcus aureus isolates were MRSA (28.8% versus 25.2%) and approximately half were MSSA (51.0% versus 56.8%).

In the DISCOVER 2 study, the majority of patients in the dalbavancin and vancomycin treatment groups were male (60.1% versus 54.6%) and white (88.4% versus 87.0%), and the median age was approximately 50 years (49.0 versus 51.0 years). Median BMI was elevated in both treatment groups (27.40 kg/m² versus 27.60 kg/m²). Cause, type, and presentation of ABSSSIs were similar in the dalbavancin and vancomycin treatment groups, with cellulitis being the most common infection, accounting for a mean of 54.1% of infections in the 2 groups (53.4% versus 54.9%), followed by major abscesses, accounting for a mean of 24.0% of infections in the 2 groups (24.3% versus 23.6%), and traumatic-wound or surgical-site infections, accounting for a mean of 21.8% of infections in the 2 groups (22.1% versus 21.5%). The planned minimum area of ABSSSI erythema required for a patient to participate in the study was 50 cm² on the face or 75 cm² on other anatomic sites; however, actual ABSSSI erythema area measurements recorded in the studies were generally much larger than the minimum required. The median (minimum to maximum) area of ABSSSI erythema at baseline was 313.50 cm² (85.1 to 5,100.0 cm²) and 362.40 cm² (72.0 to 3,922.0 cm²) in the dalbavancin and vancomycin treatment groups, respectively. SIRS criteria were met by 42.7% of patients in the dalbavancin group. More than 80% of patients in the dalbavancin and vancomycin groups had fever at baseline (81.9% versus 82.2%). Approximately 1-third of patients in each treatment group had a temperature of at least 38°C and either a WBC count above 12,000 cells/mm³ or bands of at least 10%. Staphylococcus aureus was the most common organism isolated in the dalbavancin and vancomycin groups (73.4% versus 73.6%); no more than 25% of Staphylococcus aureus isolates were MRSA (25.0% versus 16.1%) and approximately half were MSSA (48.4% versus 58.0%).

In the DUR001-303 study, the single-dose and 2-dose dalbavancin groups were well matched with respect to demographic and baseline characteristics. The majority of patients in the 1-dose and 2-dose treatment groups were male (58.5% versus 58.2%) and White (89.4% versus 89.1%). Median age was 49 years in the 1-dose group and 50 years in the 2-dose group. Cause, type, and presentation of ABSSSI were similar in the 2 treatment groups. Cellulitis constituted a mean of 47.4% of infections in the 1-dose and 2-dose treatment groups (47.3% versus 47.6%), and median BMI was elevated in both groups (26.90 kg/m² versus 27.80 kg/m²). The actual ABSSSI erythema area measurements recorded in this study were much larger than the minimum required. The median (minimum to maximum) area of ABSSSI erythema at baseline was 296.05 cm² (56.0 to 4,325.0 cm²) and 293.25 cm² (76.5 to 2,668.0 cm²) in the 1-dose and 2-dose dalbavancin groups, respectively. The largest lesions were observed in patients with cellulitis. SIRS criteria were met by 42.4% of patients in the single-dose group and by 44.4% of the 2-dose group. The majority of patients in the ITT population had fever at baseline (83.1% of patients in the 1-dose group and 81.6% of patients in the 2-dose group). The majority of patients in the 1-dose and 2-dose treatment groups in the microbiological intention-to-treat (microITT) population had monomicrobial infections (72.4% versus 73.6%). Staphylococcus aureus was the most common organism isolated in the 1-dose and 2-dose groups (65.2% versus 65.9%); approximately 1-third of the Staphylococcus aureus isolates were MRSA and approximately 2-thirds were MSSA.

In the VER001-9 study, groups were well matched with respect to demographic and baseline characteristics. The majority of patients were male (61.5% of the study population) and white (68.4%), while 18.0% of the patients identified as Hispanic or Latino and 10.7% were Black. Mean age was similar in the dalbavancin and linezolid treatment arms (47.1 years versus 46.0 years). Study-specific medical history was also similar in the 2 treatment arms, except a significantly higher percentage of patients in the dalbavancin arm than in the linezolid arm had vascular disease at study entry. Centres in North America enrolled and dosed 86.5% of the patient population. Cause, type, and presentation of SSSIs were similar in the treatment arms. Infections included predominantly major abscesses (32.3%), cellulitis (28.2%), and other deep soft tissue infections (16.6%). Hallmark signs and symptoms of SSSI reported in more than 95% of patients at baseline included erythema, heat or localized warmth, pain or tenderness on palpation, and swelling or induration. Drainage or discharge (69.9% of patients) and fluctuance (57.3% of patients) were also common.

Table 7: Summary of Baseline Characteristics (ITT Population)

ABSSSI = acute bacterial skin and skin structure infection; BMI = body mass index; DAL = dalbavancin; hs-CRP = high sensitivity C-reactive protein; ITT = intent-to-treat; IVRS = Interactive Voice Randomization System; LZD = linezolid; N = number of patients in the ITT population; MRSA = methicillin-resistant Staphylococcus aureus; MSSA = methicillin-sensitive Staphylococcus aureus; NR = not reported; SD = standard deviation; SIRS = systemic inflammatory response syndrome; SSTI = skin structure and tissue infection; VAN = vancomycin; WBC = white blood cell count.

Note: The percentages were based on nonmissing data, unless otherwise stated. Meeting SIRS criteria was defined as having 2 or more of the following: temperature < 36°C or > 38°C, heart rate > 90 beats per minute, respiratory rate > 20 breaths per minute, or WBC count < 4,000 cells/mm³ or > 12,000 cells/mm³ or > 10% bands. The values reported by the local laboratory were used to determine SIRS; wound infections include both surgical-site infection and traumatic-wound infection.

^aArea was defined as the longest length × the widest perpendicular width.

^bPatients can have more than 1 complicating factor and/or signs of systemic infection.

^cDenominator n1 is the number of patients with nonmissing values.

^dThe percentages were calculated as shown in the row labels. Specimen types identified as other include purulent fluid taken by aseptic aspiration; specimens taken during curettage, debridement, or incision and drainage; scrapings from the wound base; superficial wound smears; and various types of swabs.

Source: Clinical Study Reports.^17-20

Interventions

In the DISCOVER 1 and DISCOVER 2 studies, patients were randomly assigned to 1 of 2 treatment groups on day 1 no more than 4 hours before their first dose of the study drug. Patients assigned to dalbavancin received a 1,000 mg dose on day 1 followed by a 500 mg dose on day 8, with a possible switch to oral placebo if criteria were met, for a total duration of 14 days. Reduced dalbavancin doses of 750 mg on day 1 and 375 mg on day 8 were given to patients with creatinine clearance (CrCl) values of less than 30 mL/min who were not receiving regular hemodialysis or peritoneal dialysis. An IV placebo infusion was given every 12 hours for 3 to 14 days to correspond to the vancomycin dosage regimen, and the patient was switched to an oral placebo regimen to match oral linezolid if they met the criteria for a switch. Vancomycin-matched placebo was prepared using the same diluent as the active vancomycin for IV infusion at each study site. All patients received a single IV dose of either dalbavancin (dalbavancin treatment arm) or its placebo (vancomycin arm) on day 8.

Patients assigned to the vancomycin treatment group received an IV infusion of vancomycin every 12 hours for 3 to 14 days. Patients with normal renal function received vancomycin doses of 1,000 mg or 15 mg/kg. Patients with impaired renal function had their dosages and intervals of vancomycin treatment adjusted by an unblinded pharmacist, as necessary, based on local standard of care, renal function, and vancomycin levels. An IV placebo infusion was also given on days 1 and 8 to match the dalbavancin dosing regimen. Dalbavancin-matched placebo was prepared using 250 mL of 5% dextrose. All unit doses were prepared and administered at the study site by an unblinded pharmacist or by a designated member of the clinical pharmacy staff. A total of 20 to 28 doses of vancomycin or linezolid (or matching placebo) were to be administered.

In the DISCOVER 1 and DISCOVER 2 studies, after a minimum of 72 hours, if the blinded investigator determined that a patient met the predefined criteria for a switch to oral therapy, the IV placebo or IV vancomycin treatment was stopped and the patient was given oral therapy (e.g., oral linezolid 600 mg every 12 hours for patients in the vancomycin treatment group or matching placebo for patients in the dalbavancin treatment group). A switch to oral therapy occurred if both of the following conditions were met:

• In the previous 24 hours, the patient had 4 temperature measurements, each separated by approximately 6 hours, in which all 4 measurements were 37.6°C or lower.

• There was unequivocal improvement in some or all of the clinical signs of the ABSSSI under study; if some signs had not improved, none should have worsened (i.e., stability for that sign was observed).

In the DUR001-303 study, patients in the single-dose dalbavancin group received a single dose of dalbavancin on day 1 and a dalbavancin-matched IV placebo on day 8. The dalbavancin dose on day 1 was 1,500 mg for patients with CrCl of 30 mL/min or greater and for patients with CrCl of less than 30 mL/min who were receiving regular hemodialysis or peritoneal dialysis. In patients with CrCl of less than 30 mL/min who were not receiving regular hemodialysis or peritoneal dialysis, the dalbavancin dose on day 1 was 1,000 mg. Patients randomly assigned to the 2-dose regimen of dalbavancin followed the same dosing schedule as that described for the 2 DISCOVER studies. Dalbavancin-matched IV placebo was prepared using 250 mL of 5% dextrose. All unit doses were prepared and administered at the study site by an unblinded pharmacist or by a designated member of the clinical pharmacy staff. Patients in the DUR001-303 study received the first dose of the study drug no more than 4 hours after randomization.

In the VER001-9 study, patients were randomly assigned in a 2:1 ratio to treatment with dalbavancin or linezolid. Patients in the IV dalbavancin group received a 1,000 mg dose on day 1 followed by a 500 mg dose on day 8. In addition, IV placebo was administered every 12 hours afters the initial dalbavancin dose, with a possible switch to oral placebo every 12 hours. IV placebo was supplied by the sites. IV linezolid was administered in a dose of 600 mg every 12 hours, with possible switch to oral linezolid if switching criteria were met (i.e., improvement in fever or clinical improvement at the SSSI site after at least 24 hours of parenteral therapy). IV linezolid and IV placebo were supplied by the sites.

Concomitant Medications

Across the pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303), any medication taken by the patient other than the study drug was considered concomitant medication. Nonantibacterial medications were limited to those essential to the patient. Concomitant and adjunctive systemic and topical antibacterials were prohibited during the study, up to day 14, with the following exceptions:

• Oral vancomycin in doses of 125 mg or 250 mg every 6 hours could be used in both treatment groups for the treatment of C. difficile infections and could be continued as required for the duration of the study. The sponsor did not provide oral vancomycin.

• Metronidazole, IV or oral, in a dose of 500 mg every 8 hours could be used in both treatment groups for the treatment of C. difficile infections and could be continued as required for the duration of the study. The sponsor did not provide metronidazole. As an adjunctive therapy, it could be used in both treatment groups for suspected anaerobic pathogens and could be continued as required for the duration of the study.

• Other antibacterials that did not achieve therapeutic levels in the serum (e.g., nitrofurantoin, fosfomycin, norfloxacin) or at the site of the ABSSSI could be considered.

• Aztreonam could be used for the treatment of the ABSSSI caused by gram-negative bacteria, but only systemic aztreonam could be administered empirically at randomization for a presumed gram-negative contribution to the ABSSSI. Empirical use of aztreonam after randomization was not permitted. Use of aztreonam to treat a culture-confirmed infection at any time during the study was acceptable.

The use of any other investigational drug was prohibited, and patients could not participate in any other studies that involved marketed products during the study. Nonsteroidal anti-inflammatory drugs, such as acetaminophen, were administered if patients had a temperature of 38°C or higher. Patients in this study who were given the oral study drug were prohibited from the following medications: inhibitors of monoamine oxidase A or B (e.g., phenelzine, isocarboxazid, methylene blue), direct-acting and indirect-acting sympathomimetic drugs (e.g., pseudoephedrine), serotonin 5-HT₁-receptor agonists (triptans), meperidine, and buspirone. Drugs such as dopamine and epinephrine were to be reduced and titrated to achieve the desired response. Patients were not to be switched to an oral study drug if they had received any of the following medications: any serotonergic psychiatric medication, any other psychiatric medication, or any related medication in the previous 2 weeks; fluoxetine in the previous 5 weeks; or both. In addition, patients treated with an oral study drug were not to receive any of these contraindicated medications until at least 24 hours after their final dose of the oral study drug.

In the VER001-9 study, all medications required by the patient to manage underlying illnesses were permitted. Systemic antimicrobials were prohibited during the study, up to the TOC visit, with the following exceptions:

• metronidazole for the treatment of gram-negative anaerobes

• fluconazole for the treatment of fungal infections

• aztreonam for the treatment of gram-negative bacteria.

Topical antibacterial therapies (e.g., mupirocin), although permitted before study entry, were prohibited during the study. Nonabsorbable oral antimicrobial therapies for digestive tract decontamination (e.g., oral vancomycin or neomycin) were permitted.

The following nondrug adjunctive therapies were permitted for the treatment of ABSSSIs in all trials: debridement at the bedside; topical solutions, including antiseptic drugs such as iodopovidone (Betadine); and local bedside wound care per hospital protocol.

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 8. These end points are summarized below.

Table 8: Summary of Outcomes of Interest Identified in the CADTH Review Protocol

HRQoL = health-related quality of life; LOS = length of stay; NR = not reported. TOC = test of cure.

Analysis Populations

Five analysis sets were defined: the ITT population, the safety population, the microITT population, the CE population, and the microbiologically evaluable population. For the purpose of this review, only the ITT, safety, CE, and microITT populations will be discussed.

ITT Population

For all pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303), the ITT population consisted of all randomly assigned patients, regardless of whether or not they received the study drug. A patient was considered randomly assigned when the pharmacist or pharmacist’s designee received the treatment assignment from the IVRS.

Safety or Modified ITT Population

In the DISCOVER 1 and DISCOVER 2 studies, all patients in the ITT population who received at least 1 dose of the dalbavancin or vancomycin (active) study drug were included in the safety population. In the DUR001-303 study, all patients in the ITT population who received at least 1 (active) dose of dalbavancin were included in the microITT population, which also served as the safety population. In the VER001-9 study, all patients who received any dose of the study medication were included in the safety analyses (ITT population). Descriptive statistics were compiled for AEs, clinical laboratory test results, and vital signs.

CE Population

All trials involved a CE population. In the pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303), 2 different CE populations were defined, based on the timing of the outcome assessment evaluated: CE-EOT and CE-final visit (the term CE population refers collectively to both of these populations). Any patient who met all of the following criteria was included in the CE population for the respective visit:

• fulfilled patient selection criteria

• received dalbavancin as randomized, specifically in the DUR001-303 study, which means 1 dose for patients in the single-dose treatment group and 2 doses for patients in the 2-dose treatment group

• received a new, nonstudy systemic antibacterial drug (other than aztreonam or metronidazole) for the treatment of ABSSSI or received more than 1 dose of a concomitant systemic antibacterial therapy (except systemic aztreonam and oral or IV metronidazole) for a non-ABSSSI indication; these patients were considered to be evaluable failures and therefore included in the CE population for the respective visit

• received appropriate adjunctive antibacterial coverage if the patient had a culture-documented mixed ABSSSI (1 or more gram-positive pathogens with 1 or more gram-negative aerobic or anaerobic organisms)

• had an outcome assessment from which clinical status could be evaluated in the CE populations

  • for the CE-EOT population, that included all patients with an assessment in the time window from study day 12 to day 18, plus those who had assessments in the 3 days before premature discontinuation from the study drug and those who were considered clinical failures and received a concomitant antibiotic for worsening of the primary ABSSSI lesion before study day 12

  • for the CE-final visit population, that included all patients with an assessment in the time window from study day 26 through study day 30, plus those who were considered clinical failures at the EOT visit and received a concomitant antibiotic for worsening of the primary ABSSSI lesion.

In the VER001-9 study, patients in the CE population met all of the following conditions:

• fulfilled inclusion and exclusion criteria such that a clinical response could be evaluated, unless an exemption was granted before randomization

• received at least 3 days worth of study medication based on start and stop dates (i.e., stop date minus start date, plus 1 day)

• had not received more than 24 hours of another systemic or topical antibacterial therapy with documented activity against the causative organism between the 7 days before the first dose of the study medication and the TOC assessment, unless the indication for the new antibiotic was lack of efficacy, in which case the outcome was considered a failure and the patient remained evaluable

• did not have a clinical response considered indeterminate at the time point being analyzed

• received adjunctive antibacterial coverage with aztreonam or metronidazole if the patient had a culture-documented mixed SSSI (i.e., concomitant gram-negative aerobic and/or anaerobic pathogens), unless the organism was not considered by the investigator to be causative

• had not received unanticipated surgical intervention during the study, unless the response was considered a clinical failure.

Patients who were CE at the EOT visit (day 14) and did not return for a TOC visit (day 28) and/or violated the protocol in a manner that affected evaluability before the TOC visit were considered CE at the EOT visit, but not at the TOC visit. Similarly, patients who did not attend the EOT visit (and thus had a clinical response considered indeterminate for EOT and were not CE at the EOT visit) were CE for TOC if they returned for the TOC visit and had not otherwise violated the protocol before the TOC visit. Therefore, there were 2 groups in the CE population: CE at the EOT visit (day 14), and CE at the TOC visit (day 28).

Primary End Points

Clinical Response

All pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303) used clinical response at 48 to 72 hours as the primary end point. In the VER001-9 study, the primary end point was clinical response at day 28.

In the DISCOVER 1 and DISCOVER 2 studies, clinical response was based on lesion size and temperature. A patient was defined as a clinical responder if the following 2 criteria were met:

• The patient had a temperature no higher than 37.6°C (regardless of method of measurement) 48 to 72 hours after the first dose of the study therapy followed by 2 additional temperature measurements no higher than 37.6°C (regardless of method of measurement) separated by at least 3 hours and no more than 9 hours, with no intervening temperature above 37.6°C.

• The patient had no increase in lesion area 48 to 72 hours after the first dose of study-drug therapy from the baseline measurement. Lesion area was defined as length × width, and lesion size was defined as length or width. The baseline lesion measurement was defined as the measurement taken closest to but before the first dose of the study drug. If multiple lesion measurements were taken in the 48 to 72 hours after the first dose of the study drug, the most recent lesion measurement was used.

Patients was met at least 1 of the following criteria were defined as a clinical nonresponder:

• The patient had an increase over baseline in lesion size 48 to 72 hours after initiation of the study drug.

• The patient did not have a temperature of 37.6°C or lower (regardless of method of measurement) 48 to 72 hours after the first dose of the study therapy followed by 2 additional temperature measurements below 37.6°C, each separated by approximately 6 hours, or had an intervening temperature above 37.6°C.

• The patient died from any cause in the 72 hours after the first dose of the study drug.

• The patient initiated a new, systemic antibacterial with gram-positive activity for the ABSSSI under study in the 72 hours after the first dose of the study drug.

• The patient had missing data at 48 to 72 hours for lesion size or temperature, such that a clinical outcome could not be defined.

In the DISCOVER 1 and DISCOVER 2 studies, IV treatment was initiated and the patient’s temperature was recorded on day 1. Efficacy assessments were made on days 2, 3, 4, 8, and 14 or 15 of the treatment period. An EOT assessment took place on days 14 or 15, or in the 3 days after premature discontinuation of treatment. A SFU was planned for day 28 and a final LFU at day 70.

In the DUR001-303 study, patients were defined as clinical responders if they met the following 3 criteria:

• alive

• received no rescue therapy for ABSSSI before the infection-site assessment at 48 to 72 hours

• examination of the ABSSSI lesion demonstrated a decrease of at least 20% in lesion area (calculated as the longest length multiplied by the longest perpendicular width) relative to the baseline measurement. If an antibiotic was given for a reason other than rescue therapy, the patient was not considered a nonresponder.

Nonresponders were those who did not meet the aforementioned criteria. Baseline assessments were performed in the 24 hours before the first dose of the study drug. On day 2, patients being treated on an outpatient basis were contacted by the investigator to check for worsening of the presenting ABSSSI lesion. Efficacy assessments were performed on day 3 to 4; on day 8, at which time patients were administered the second dose of the study drug; on day 14 or 15, which was defined as the EOT visit; and on day 28, which was defined as the final visit.

In all pivotal trials, patients were considered to have an indeterminate outcome if any data needed to determine whether the outcome was a success or failure were missing. For example, if the assessment of local signs was not completed at the EOT visit, the patient was considered to have an indeterminate response for the analysis at the EOT visit. Patients with an indeterminate response were included in the denominator for analyses of the ITT population and were considered treatment failures.

In the VER001-9 study, the primary efficacy end point was the clinical response rate in the CE population at day 28, which was defined as the TOC visit. Clinical response (success, indeterminate, failure) was assessed by the investigator, as follows:

• Success meant sufficient resolution of the local and systemic signs and symptoms of SSSI, such that the patient did not receive new, systemic antibacterial treatment for SSSI.

• Failure meant persistence of 1 or more local or systemic signs and symptoms of the SSSI, such that new, systemic antibacterial treatment was given for the SSSI, or the patient died during the study period and the SSSI was the sole cause or was thought to be contributory toward death. All EOT failures were carried forward as failures to the TOC visit.

• Indeterminate meant there were no post-baseline local or systemic signs or symptoms data available to assess clinical response (e.g., patient lost to follow-up); the patient received the study medication for fewer than 72 hours; or another reason (must be specified). Indeterminate responses were grouped with failures for the ITT analysis.

Clinical status was assessed by the investigator at baseline; during treatment (day 4 + 1 day), day 8 (± 1 day), day 14 (defined as EOT); in the 3 days after completion of therapy; and at the TOC visit (14 ± 2 days after completion of therapy). The investigator assigned a clinical response at the EOT and TOC visits. Patients with a successful clinical response at the TOC visit were contacted for LFU assessment.

Secondary End Points

Clinical Response

In the VER001-9 study, clinical response was additionally evaluated in the CE population at day 14 (EOT) and, as a secondary end point, in the ITT population at the EOT and day 28 (TOC) visits. Clinical response and failure were defined as they were for the primary efficacy variable.

Clinical Status of Success

For all pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303), a secondary outcome measure was clinical status at the EOT (day 14 or 15) and day 28 visits. To accommodate a specific request from the European Medicines Agency directed to the sponsor as part of scientific advice, the clinical status at day 14 or 15 in the CE population was selected as a clinical end point of special interest in the DISCOVER 1 and DISCOVER 2 studies.

In the DISCOVER 1 and DISCOVER 2 studies, a response was considered a clinical success if it met following 5 criteria:

• The patient’s lesion size, as defined by erythema, was decreased from baseline.

• The patient’s temperature was no higher than 37.6°C (by any measurement method).

• Local signs of fluctuance and localized heat or warmth were absent.

• Local signs of tenderness on palpation and swelling or induration were no worse than mild.

• For patients with a wound infection, the severity of purulent drainage was improved and no worse than mild, relative to baseline.

In the DUR001-303 study, a response was defined as a clinical success if it met the following 5 criteria:

• lesion area must be decreased by at least 80% from baseline and, at the final visit, must be decreased by at least 90% from baseline

• temperature is no higher than 37.6°C

• local signs of tenderness on palpation and swelling or induration are no worse than mild

• at the EOT visit, local signs of fluctuance and localized heat or warmth must be improved from baseline and no worse than mild, and at the final visit, local signs of fluctuance and localized heat or warmth must be absent; for patients with a wound infection, the severity of purulent drainage must be improved and no worse than mild relative to baseline

• for patients with a wound infection the severity of purulent drainage is improved and no worse than mild relative to baseline.

For all trials, a patient was defined as having indeterminate clinical status if any of the data needed to determine clinical success or clinical failure, as defined above, were missing.

Clinical Status of Failure

In the DISCOVER 1 and DISCOVER 2 studies, a response was defined as a clinical failure if at least 1 of the following criteria was met:

• the patient’s lesion size, defined by erythema, was not decreased from baseline

• local signs of fluctuance and localized heat or warmth had not resolved

• local signs of tenderness on palpation and swelling or induration were worse than mild

• for patients with a wound infection, the severity of the purulent drainage was the same as it was at baseline, or worse, or was worse than mild

• the patient had a temperature higher than 37.6°C (by any measurement method) at the visit

• the patient received a new, nonstudy, systemic antibacterial treatment for the ABSSSI at any time from the first dose of the study drug through the visit

• the patient died before the visit

• unless preplanned as part of the nondrug therapy for the ABSSSI, the patient required surgical intervention more than 72 hours after the start of therapy for the treatment of the ABSSSI under study

• the patient received study therapy for the ABSSSI under study beyond the protocol treatment period as a result of the investigator’s assessment that additional drug therapy was needed for treatment of the underlying skin infection.

In the DUR001-303 study, clinical failure was defined as the opposite of success, death before the visit, or the administration of the study therapy for ABSSSIs beyond the protocol treatment period.

Systemic Signs and Symptoms

In the DISCOVER 1 and DISCOVER 2 studies, time to resolution of fever was based on the patient’s first temperature measurement that indicated the patient was afebrile, although resolution required 3 sequential temperatures no higher than 37.6°C. Patients who were lost to follow-up while on the study drug and continued to have fevers were censored at their last visit. Patients who continued to have fevers were censored on the date they last received the study drug.

In the DUR001-303 study, time to resolution of fever was based on the patient’s first temperature measurement of ≤ 37.6°C, regardless of method of measurement. Patients who continued to have fever were censored on the last date they received the study drug or on the last visit date, whichever was later.

Need for Substitution With Another Antibiotic

The need for substitution with another antibiotic was considered a reason for failure in the assessments of clinical response and clinical status (EOT).

Health-Related Quality of Life

HRQoL was not assessed in the trials.

Hospitalization Rate and Hospital LOS

In the DUR001-303 and VER001-9 studies, an efficacy end point was resource use. In the DUR001-303 study, resource use was based on days in the hospital, outpatient visits, and ER visits. Data were collected on day 14 or 15 and on day 28 (± 2 days).

Safety

Safety assessments were made at every visit. For all trials, the safety parameters included AEs, clinical laboratory evaluations, vital signs, and physical examination assessments. AEs were coded using the Medical Dictionary of Regulatory Activities (MedDRA), version 14.0 (DISCOVER 1 and DISCOVER 2 studies) or version 17.0 (DUR001-303 study) or higher of the System Organ Class and Preferred Term levels. In the VER001-9 study, AEs included any adverse experience, whether or not it was considered drug-related, that occurred during a patient’s study participation. This included any side effect, injury, toxicity, sensitivity reaction, intercurrent illness, or sudden death. A pre-existing condition was 1 that was present on physical examination at the start of the study and was reported as part of the subject’s medical history. Any change from the previous physical examination during study treatment (e.g., worsening in frequency, intensity, or character of the pre-existing condition) was considered an AE.

Safety evaluations included TEAEs, which were defined as AEs that appeared, increased in frequency, or worsened in severity during the period that began with the first dose of the study drug (including events that occurred during study-drug administration) and ended at the final visit.

Statistical Analysis

Primary Outcome of the Studies

Sample Size Determination

In the DISCOVER 1 and DISCOVER 2 studies, the point estimate used for sample size determination was based on a retrospective analysis of the VER001-9 study, and included 428 patients with cellulitis, major abscess, or surgical- or traumatic-wound infection with a baseline lesion area of at least 75 cm². In the VER001-9 study, clinical response was 70% (95% CI, 65% to 74%) for the outcome measures of cessation of spread and absence of fever approximately 48 to 72 hours after the first dose. The sponsor noted that there are study design and analysis issues (e.g., measurement error, missing data, different data-collection times) that could have biased the point estimate. Thus, because the VER001-9 study may have provides a reduced estimate of the true clinical response rate, a point estimate of 85%, which corresponds to the upper limit of the CI for the analysis, excluding missing data, was used to determine the sample size. In addition to a 1-sided alpha of 0.025, a noninferiority margin of 10% and a 90% power were assumed. Based on the method of Farrington and Manning,³⁶ the sample size for the DISCOVER 1 and DISCOVER 2 studies was 556 randomized patients (ITT population). There is also sufficient power for the secondary outcome of clinical status at the EOT visit, assuming a 10% noninferiority margin. Data from the VER001-9 study indicate an 81% clinical success rate at the EOT visit (in the ITT population). The outcome rate for the programmatic determination of clinical status at the EOT visit was expected to be similar to that of the investigator’s assessment in the VER001-9 study. Assuming an outcome rate of 80% in the ITT population and 90% in the CE population (and an 85% evaluability rate), there is an 83% and 94% power, respectively, for the outcome of clinical status at the EOT visit.

In the DUR001-303 study, the proposed sample size was 205 patients per arm, based on the method of Farrington and Manning.³⁶ This assumed a noninferiority margin of 10%, a power of 90%, a 1-sided alpha level of 0.025, and a 90% treatment response, which is consistent with FDA guidance. The expected treatment response was estimated from the DISCOVER 1 and DISCOVER 2 trials, given the similar patient population. In those studies, the response rate for the primary outcome measure was 89.9% in the DISCOVER 1 study and 87.6% in the DISCOVER 2 study. Therefore, the overall response rate was expected to be approximately 90% in the DUR001-303 study. The aggregate blinded response rate was assessed when approximately 60% of patients were enrolled. If the aggregate response rate was less than 90%, a sample size adjustment could be made. To ensure that the point estimate of greater than a 20% reduction used in the estimation of sample size was valid for this study, an interim analysis for sample size re-estimation was performed when early clinical response data, at 48 to 72 hours, were available for 60% of the patients. The interim analysis involved a sample size re-estimation to either confirm that the initial sample size estimate was adequate or to increase the sample size (number of randomly assigned patients) to ensure that the study had adequate power to determine whether a single 1,500 mg dose of dalbavancin is noninferior to the same total dose given as 1,000 mg on day 1 and 500 mg on day 8 for the primary outcome measure. The sample size re-estimation was based on the blinded overall (not by treatment group) clinical response rate and was conducted by an independent, blinded statistician.

In the VER001-9 study, the sample size calculation assumed that if the 2 treatment groups (dalbavancin and linezolid) were equally effective and the clinical success rate was 85% at the day 28 visit, then 258 patients in the dalbavancin group and 129 patients in the linezolid group (CE population) would be required to ensure, with 90% power, that the lower bound of a 1-sided 97.5% CI for the true difference in efficacy between the groups would not exceed –12.5%. Assuming an evaluability rate of at least 70%, approximately 555 patients will need to be enrolled to obtain 386 CE patients. An additional 150 patients were enrolled in the study in a 2:1 ratio (dalbavancin:linezolid), resulting in a total patient enrolment of approximately 705 to augment the safety database.

Statistical Test

All trials used a noninferiority analysis for the primary efficacy end point.

In the DISCOVER 1 and DISCOVER 2 studies, the primary efficacy analysis was performed in the ITT population. The noninferiority test was a 1-sided hypothesis test performed at the 2.5% level of significance and was based on the lower limit of the 2-sided 95% CI. The primary efficacy analysis was adjusted for the randomization stratification factor of the presence or absence of fever at baseline. The number and percentage of patients in each treatment group defined as clinical responders and nonresponders were tabulated. To test the null hypothesis, a 2-sided 95% CI for the observed difference in primary outcome rates (dalbavancin treatment group minus vancomycin treatment group) was calculated. If the lower limit of the 95% CI for the treatment difference in the ITT population exceeded –10%, then the null hypothesis was rejected and the noninferiority of dalbavancin to vancomycin was concluded. The 2-sided 95% CI for noninferiority testing based on the difference in clinical response rates at 48 to 72 hours was computed using the method proposed with stratification by Miettinen and Nurminen (1985).³⁷

In the DUR001-303 study, the primary efficacy analysis used the same noninferiority test as the DISCOVER 1 and DISCOVER 2 studies for the observed difference in the primary outcome measure (single-dose dalbavancin group minus 2-dose dalbavancin group). If the lower limit of the 95% CI for the difference in ITT population exceeds –10%, then the null hypothesis will be rejected and the noninferiority of 1-dose dalbavancin to 2-dose dalbavancin will be declared. In the primary efficacy analysis, the 95% CI was computed using Cochran-Mantel-Haenszel stratum weights, without adjustments for the stratification factors employed in randomization (i.e., geographic region, infection type, and prior use of antibiotics).

In the VER001-9 study, the primary analysis was a 2-step comparison of the clinical success rates (success versus failure) between treatment arms (Hwang 1999). A 1-sided 97.5% CI was calculated for the true difference in efficacy (dalbavancin minus linezolid) using the normal approximation method for CIs. In the first step, noninferiority was concluded if the lower limit of the 1-sided CI did not exceed –12.5%. In the second step, if the lower limit of the 1-sided CI was greater than 0, dalbavancin was considered strict-sense superior to linezolid. In practice, the 2-sided 95% CI of the treatment difference is displayed. The lower bound of this interval is identical to the lower bound of the planned 1-sided 97.5% CI. The primary end point (clinical response rate at the TOC visit in the CE population) was the sole end point for confirmatory statistical testing. All other end points, populations, and time points were considered secondary and supportive. Therefore, no adjustments for multiple end points were considered necessary by the sponsor.

Data Imputation Methods

For the primary outcome measure (clinical response at 48 to 72 hours), patients were considered to have missing data if there was no lesion measurement at baseline and/or in the 48- to 72-hour time period (after the first dose of the study drug). Patients were considered to have missing data if there were 3 missing temperature measurements in the 48- to 72-hour time period taken 6 hours (± 3 hours) apart. Patients with missing data were defined as a nonresponders for the primary analysis (ITT analysis).

The DUR001-303 study handled missing data as described for the DISCOVER 1 and DISCOVER 2 studies. In addition, the DUR001-303 study used multiple imputation, with a Markov chain Monte Carlo method, for the secondary outcomes. Two models were run; the first used the type of infection as a predictive variable and the second used the clinical response at 48 to 72 hours as a predictive variable. For both analyses, 50 imputation datasets were created with imputed data for the missing values and existing values for complete cases.

For all pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303), patient responses were considered indeterminate if data were not available for the evaluation of efficacy for any reason. By definition, patients with an indeterminate response were included in the denominator for analyses in the ITT and microITT populations and are considered failures.

In the VER001-9 study, there were no plans to impute missing data in the study. Responses for patients with missing data related to the clinical response were considered indeterminate.

Subgroup Analyses

In the pivotal trials and the VER001-9 study, the subgroup analyses were planned a priori in the statistical analysis plan (SAP) for specific groups of patients. For each subgroup, a forest plot and the respective outcome’s 95% CI was provided. In the DISCOVER 1 and DISCOVER 2 studies, subgroup analyses were conducted for the primary outcome, clinical response at 48 to 72 hours, and clinical status at the EOT and day 28 visits, based on the following groups:

• infection type

• sex (female versus male)

• geographic region (North America versus the rest of the world)

• prior surgical intervention

• presence or absence of fever

• presence or absence of bacteremia

• nonsteroidal anti-inflammatory drug use before administration of the study drug

• antipyretic use before administration of the study drug, excluding patients with an infection-related major abscess

• SIRS criteria met.

In the DUR001-303 study, subgroup analyses were conducted for the primary outcome, clinical response at 48 to 72 hours, and clinical status at the EOT and day 28 visits, based on the following groups:

• age (< 65 or ≥ 65 years)

• sex

• race

• ethnicity

• randomization before or after interim analysis

• geographic region

• infection type

• prior use of antibiotics

• presence or absence of fever

• presence or absence of bacteremia

• SIRS criteria met

• inpatient or outpatient treatment.

In the VER001-9 study, subgroup analyses were conducted for the primary outcome, clinical response at the TOC visit (day 28) in the CE population, and clinical response at the TOC visit (day 28) in the ITT population, based on the following categories: age, sex, ethnicity, presence of diabetes, presence of vascular disease, infection category, geographic area, and surgical intervention.

The following subgroup, planned a priori in the statistical analyses plan, aligned with the subgroup pre-specified in the protocol for this CADTH review: presence or absence of bacteremia, MRSA versus MSSA infection, and infection type (major abscess, cellulitis, and wound). Only the subgroup identified in the CADTH review protocol is reported in the efficacy section.

Sensitivity Analyses

In the DISCOVER 1 and DISCOVER 2 studies, 13 separate pre-specified sensitivity analyses of the primary efficacy analysis were performed to determine how missing data and different measurement criteria would affect the results.

In the DUR001 to 3 study, sensitivity analysis of the primary efficacy analysis was conducted using an expanded window of 36 to 75 hours. An additional analysis of clinical status at the EOT visit (day 14 or 15) and on day 28 was conducted in which the definition of clinical success was switched from “local signs of fluctuance and localized heat/warmth are absent” to “local signs of fluctuance and localized heat/warmth are improved from baseline and no worse than mild.” In addition, the primary efficacy outcome was adjusted for the randomization stratification factors of prior use of antibiotics, infection type, and geographic region as captured in the IVRS.

In the VER001-9 study, a post hoc sensitivity analysis was conducted that excluded patients with an indeterminate clinical response in both treatment arms for clinical response at the TOC visit.

Secondary Outcomes of the Studies

Secondary efficacy end points were for descriptive purposes and no conclusions of noninferiority were made.

In the DISCOVER 1 and DISCOVER 2 studies, the observed difference in percentage of patients with a clinical success (dalbavancin group minus the vancomycin group) was determined, and a 2-sided 95% CI for the observed difference was computed using the method of Miettinen and Nurminen (1985),³⁷ with stratification for the presence or absence of fever at baseline. Cochran-Mantel-Haenszel weights were used for the stratum weights in the calculation of the CI. The reasons (a patient may have more than 1 reason for clinical failure) for clinical failure were summarized by treatment group, by the presence or absence of fever at baseline, and by infection type. The reasons for an indeterminate response were also summarized by treatment group. Two-sided 95% CIs for the observed differences in clinical success rates were calculated for descriptive purposes. Exploratory subgroup analyses of clinical status at the EOT visit were also conducted.

In the DUR001-303 study, noninferiority of the secondary end points was not formally tested; however, from previous phase III studies, it was expected that with a 1-sided hypothesis test performed at the 2.5% level of significance, the lower limit of the 95% CI for the difference in response rates in the ITT population would be greater than –15%.

In the VER001-9 study, clinical response end points were analyzed using the same method as for the primary end point. For the additional end points of time from IV to oral switch and discontinuation of IV access following IV to oral switch, summary statistics were produced.

Sensitivity Analyses

In the DISCOVER 1 and DISCOVER 2 studies, a sensitivity analysis of the secondary outcome measure of clinical status was conducted, which emphasized improvement, rather than absence, in fluctuance and warmth of the lesion. In this analysis, the criteria for success, “local signs of fluctuance and localized heat/warmth are absent,” were replaced with “if present at baseline, local signs of fluctuance and localized heat/warmth must be improved from baseline.” The number and percentage of patients with a clinical status of success and failure at the EOT and SFU visits, based on the revised definition, is determined in each treatment group in the ITT and CE populations. The observed difference in the percentage of patients with a clinical success (dalbavancin group minus the vancomycin group) was determined, and a 2-sided 95% CI for the observed difference was computed using the method of Miettinen and Nurminen (1985),³⁷ with stratification for the presence or absence of fever at baseline. Cochran-Mantel-Haenszel weights are used for the stratum weights in the calculation of the CI. Multiple imputation methods, using a Markov chain Monte Carlo full-data imputation, will be used to define missing data (i.e., patients with an indeterminate outcome). Exploratory subgroup analyses (patients with bacteremia) of clinical status at the EOT visit were also conducted.

In the DUR001-303 study, sensitivity and additional analyses for the secondary efficacy end point were performed in the same manner as described for the primary efficacy analysis, except that the patient’s lesion size had to be decreased by at least 80% from baseline at the EOT visit and at least 90% from baseline at the final visit. A further sensitivity analysis was conducted in which 95% CIs were calculated in which clinical failures at the EOT visit carried forward to the final visit and clinical status was adjusted by stratification factors.

The VER001-9 study included post hoc sensitivity analyses that will not be considered by CADTH.

|||||||| ||||||||||||| |||||||| || ||||||||| | || ||| ||||| |||||||| ||||||| ||||| ||||||||| || ||| |||||||| ||||||||| ||||| ||||||| |||||||| |||||||| ||| ||||||||| |||||||| || ||||||| |||||||| |||| |||||||||| |||||||| ||||||||| | ||| ||| ||||| |||||||| ||||||| ||||| ||||||||| || ||||||||| | ||||||||| ||||||| || ||| ||||||||| |||||||| || ||||||| |||||||| |||| ||||| || ||||||||| ||| ||||||||||| || |||| | |||| || ||||||| |||| |||||||| ||||||||||||||||| ||| ||||||| ||| |||||| ||| |||||||||| || || || ||||||| || ||||||| |||| || ||||| |||| ||| |||||||||| |||||||| || ||| ||||| |||||||| ||||| ||| || |||||||||| || || ||| ||||||||| | ||| ||||| ||||| |||||||| || |||||| || ||| ||||||||||||| |||||||| || ||||| || |||| |||| ||| || |||||||| |||||||| |||||||| || |||| | ||||| ||||||| || ||||| |||||||||| ||||||||| ||||||| ||||||||||| ||||| ||| |||||||| || ||||||||| | || ||| ||||| |||||||| ||||||| ||||| ||||||||| || ||| |||||||| ||||||||| ||| |||| ||||||| || ||||||||| ||| ||||||||| |||| |||| || |||||||| || || ||||||||| ||| |||||| |||| ||||||||||||| ||| ||||||| || |||||||| ||| ||||||||| || |||||| |||| |||| ||| || ||||||| ||||||||||||| |||||||| |||| ||||||| || ||| ||||| |||||| || |||||||| || | ||||||||| || |||||||||||||| ||| ||||||| || ||||| ||| || |||||||| |||||||| |||||||| || |||| ||||| || ||||||||| ||||||||| | ||| ||||| |||||| |||||||| ||||||| || ||| ||||||||||||| |||||||| || ||||| || |||| |||| ||| || |||||||| |||||||| |||||||| || |||| | ||||| ||||||| || ||||| |||||||||| ||||||||| ||||||| ||||||||||| ||||| ||||||||| | ||| |||||| ||||| |||||||| ||||||||||| ||||||| || ||| ||||||| |||||| |||| ||||| || ||| ||||||| ||||||| ||||||||| |||||||||| ||| |||||| || |||||||| ||||||| |||| ||| || ||||||| ||||||||||| | |||||||||| |||| |||| || ||| ||||||||| || ||||||||| | ||||||||| ||| ||||| ||| |||||||| |||||| |||| ||| ||||||| |||| ||| |||||||| || ||| ||||| || ||| |||||||| |||||||| |||| || ||| ||||||| ||||||||| ||||||| || ||| ||||||| |||||||| || |||||||||| |||||||| || ||| |||| ||||||| ||| |||||||||| |||| ||||| ||||||||| ||||||||| || ||| |||||||||| || | |||| || ||| || |||| ||||| ||||||||| || |||||||| || |||||||| ||||||||| |||||||||| || ||||| |||||||| ||||||||| ||||| |||| | |||||||||| || ||| ||||||||| ||||||||| | || ||| ||||| |||||||| |||||||||| || |||||||||| |||||||| || ||||||| ||| |||||| ||||||||| ||| ||| |||||||||| || ||||||||||||| |||||||| |||||||| || ||||||| || ||||||| |||||||| ||| |||||||| | || ||||| || ||||| ||||||||||| ||| |||| ||||||| || ||| | |||||||| |||| ||||||||| ||| || ||||| |||||||||| || |||||||||| |||||||| || || | ||||| || |||| || ||||||||| |||||||| ||| ||| || |||||||||| |||| ||||||||| || |||||||| |||| |||||||| ||| |||||||| ||| ||||| || ||||| ||||||||||| ||||||||| |||||||| ||||| || ||||||||| |||||||||||| || ||||||||| ||||

Results

Patient Disposition

A summary of patient disposition in the pivotal trials (DISCOVER 1, DISCOVER 2, and DUR001-303) and the VER001-9 study is available in Table 10. The most common reasons for failing screening (≥ 10% of patients) were receipt of a systemic antibiotic with gram-positive activity for more than 24 hours before screening, the potential presence of osteomyelitis or septic arthritis, and refusal to provide informed consent. The rate of study discontinuation was similar between treatment groups for all trials. In all trials, the major reason for discontinuation was loss to follow-up (4.0% to 5.2% in the pivotal trials and 7.7% to 7.8% in the VER001-9 study).

Table 10: Patient Disposition

CE-EOT = clinically evaluable population at end of treatment; DAL = dalbavancin; ITT = intention-to-treat; LZD = linezolid; MicroITT = microbiological intention-to-treat; VAN = vancomycin.

Source: Clinical Study Reports.^17-20

Exposure to Study Treatments

DISCOVER 1 and DISCOVER 2 Studies

In the DISCOVER 1 and DISCOVER 2 studies, at least 98% of patients in the ITT population had a compliance of at least 80% in each treatment group for both placebo and the active study drug, and the majority of patients in both treatment groups had 100% compliance with treatment.

DUR001-303 Study

In the ITT population, 93.0% of patients in both treatment groups completed treatment with both doses of the randomly assigned study drug, per-protocol (active and placebo considered together); more patients were compliant with the first dose (99.6% of patients) than with both doses (93.0% of patients).

VER001-9 Study

All patients received at least 1 dose of IV study medication in this study. Each IV dose was considered to represent 7 days of therapy because of the long half-life. Therefore, patients in the dalbavancin arm were exposed to treatment for either 7 or 14 days, depending on whether the day 8 dalbavancin infusion was administered. Exposure for patients who received linezolid depended on the number of IV and oral doses received. Oral linezolid was taken by 265 (93.6%) patients. Most patients in both treatment arms were exposed to treatment for at least 14 days (87.2% in the dalbavancin group versus 85.2% in the linezolid group). Average exposure was 3.8 days (range, 1 to 16 days) for IV linezolid and 11.4 days (range, 1 to 20 days) for oral linezolid.

|||||||| |||||||||| || |||| || ||||||||| |||||||| |||||||||| ||| |||||||| |||||||||| ||| ||||||||| |||||| |||||| ||| ||||||||| || ||||| || ||| ||||| ||| ||||||||| |||||||| |||||||||| ||| |||||||| ||||||||||| |||| ||||| ||||||||| ||| |||||||| || |||||| ||| ||||||| |||||||| |||||||| |||||||||| |||| ||||||| ||||||| |||||| |||||| |||| ||||||

Table 11: Redacted

||| | |||||||||||| ||| | |||||||||||||||| ||| | |||||||||| ||| | ||||||| |||||| ||| | ||||||| |||||| | | |||||| || |||||||| |||| || |||||||||||| || | ||| ||||||||| || | |||||||| |||||||||| ||| | ||||||||||||| |||| |||||||||| ||| ||||||| || ||| |||||| || ||||| |||||||| ||||||| || ||| |||||| || |||||||| |||||| ||||| |||||||| ||||| ||| ||||| || ||| |||||| || |||| |||| ||||||| ||||| ||||||| ||| ||||||||| ||||| |||| || ||||| |||| ||| |||||||| || ||| ||||||| || |||||||| ||| ||||||| ||| || ||||| ||||| ||||||| |||||||||| |||||||| |||||||||||| ||||||||||| |||||||||| ||| ||| ||||||||||||| |||||||||||||||| |||||||| ||||| ||||||||||||

Table 12: Redacted

||| | |||||||||||| || | ||||||||||||||| |||||||| ||| | |||||||||||||||| ||| | |||||||||| | | |||||| || |||||||| |||| || |||||||||||| | | |||||| || ||| |||||||| || ||| ||| ||||||||||| || | ||| ||||||||| ||| | ||||||| ||||||| |||||| ||| | ||||||||||||||||| ||| ||||||||||| |||| |||||||||| || ||||||||| || ||| ||| ||||||| | ||||||| ||| |||| |||| |||| | |||| || |||||||||| |||| ||||| |||||||||| ||| ||| |||||||| || ||| ||||||||

Table 13: Redacted

||| | |||||||||||| ||| | |||||||||||||||| ||| | |||||||||||||||| ||| ||||||||||| |||| |||||||||| || ||||||||| || ||| ||| ||||||| | ||||||| ||| |||| |||| |||| | |||| || |||||||||| |||| ||||| |||||||||| ||| ||| |||||||| || ||| ||||||||

Concomitant Medication

Overall, in the DISCOVER 1 and DISCOVER 2 studies, the numbers of patients who used antibacterial and/or nonantibacterial medications either before the first dose of study drug or concomitantly during the study were similar across treatment groups. In the ITT population of the DISCOVER 1 study, the number of patients who took any systemic or topical antibacterial medications in addition to the study drugs from the start of the study drug through to the EOT visit was similar in the dalbavancin and vancomycin treatment groups (26 [9.0%] versus 25 [8.8%]); all medications were systemic. The medication most commonly taken in both treatment groups was aztreonam. The most common nonantibacterial medications were anesthetics, antihistamines, antiseptics and disinfectants, and antithrombotic medications. Medications taken in the dalbavancin treatment group were metronidazole and aztreonam and in the vancomycin treatment group were amoxicillin plus clavulanic acid, penicillin, and aztreonam. Medications taken in both treatment groups were most commonly in the Anatomic Therapeutic Chemical classes, including anti-inflammatory and antirheumatic products, other analgesics and antipyretics, and opioids. In the ITT population, approximately 1-fifth of patients had incision and drainage of their ABSSSI before the start of the study drug (18.8% of patients and 20.7% patients in the dalbavancin and vancomycin treatment groups, respectively). Other commonly performed nondrug interventions included aspiration, application or change of surface dressings, and wound packing.

In the DUR001-303 study, overall, the number of patients who used antibacterial and/or nonantibacterial medications either before the first dose of the study drug or concomitantly during the study was similar across treatment groups. For antibacterial medications, patients who used any systemically administered antibiotic with a gram-positive spectrum that achieves therapeutic concentrations in the serum or at the ABSSSI site were excluded from the study. However, an exception to this exclusion criterion was allowed for patients who received, during the 14 days before the first dose of the study drug, a single dose of an antibacterial drug with a half-life no longer than 12 hours. Twenty-two patients (6.3%) in the single-dose group and 19 patients (5.4%) in the 2-dose group received a systemic antibacterial drug in the 14 days before study-drug assignment. The number of patients who used any systemic antibacterial medication during the period from the first dose of the study drug through to the EOT visit was similar across treatment groups: 39 patients (11.2%) in the single-dose group and 54 patients (15.5%) in the 2-dose group. Per protocol, any patient who used a systemic antibacterial medication for ABSSSIs during this period was to be considered a treatment failure. The number of patients who used any systemic antibacterial medication during the period from the first dose of the study drug through to the final visit (day 28) was also similar between groups.

In the VER001-9 study, the majority of patients in each treatment arm received at least 1 concomitant medication. Analgesics were the most common class of concomitant medications, taken by 72.2% of patients overall, and by a similar percentage of patients in each treatment arm. Vicodin and paracetamol were the most prescribed analgesics. Systemic antibacterials other than study medication, taken by 52.5% of patients overall, were the second most common class of concomitant medication. Twelve percent of patients in each treatment arm received antibacterial therapy in violation of the protocol. In addition to protocol violators, the overall use of concomitant antibiotics included protocol-defined use of systemic antibacterials, such as antibiotics administered to patients who failed treatment, aztreonam and/or metronidazole given for gram-negative aerobe and/or anaerobe coverage in patients with mixed infection, no more than 24 hours of antibacterial therapy (either before or concomitant), and antibiotics with a documented lack of in vitro activity against baseline gram-positive pathogens. Aztreonam and metronidazole were administered concomitantly to a similar proportion of patients in the 2 treatment arms (for aztreonam, 15.6% and 14.8% of patients in the dalbavancin and linezolid groups, respectively, and for metronidazole, 12.6% and 11.7% of patients, respectively).

Efficacy

Only efficacy outcomes and analyses of subgroups identified in the review protocol are reported here.

Clinical Response

Clinical response results for all trials are presented in Table 14. In the DISCOVER 1 study, clinical response at 48 to 72 hours was reported in 83.3% of the dalbavancin group and 81.8% of the vancomycin group, with a between-group difference of 1.5% (95% CI, –4.6% to 7.9%). The lower limit of the 95% CI for the treatment difference (1.5%) in the ITT population was –4.6%, which was higher than the pre-specified lower limit of –10%. Therefore, noninferiority of dalbavancin to vancomycin was concluded. The results of clinical response at the EOT visit by investigator assessment also showed similar results between treatment groups (90.3% with dalbavancin and 91.9% with comparator), with a 1.6% treatment difference (95% CI, –6.4% to 3.1%).

In the DISCOVER 2 study, clinical response at 48 to 72 hours was also similar between groups, with a greater response in the comparator arm (78.3%) than in the dalbavancin (76.8%), and a –1.5% treatment difference (95% CI, –7.4% to 4.6%). The lower limit of the 95% CI for the treatment difference of –1.5% in the ITT population was –7.4, which was higher than the pre-specified lower limit of –10%. Therefore, noninferiority of dalbavancin to vancomycin was concluded. In addition, clinical response at the EOT visit in the CE population was consistent with the ITT population, with a treatment difference of 2.8% (95% CI, –6.7% to 0.7%).

In the DISCOVER 1 and DISCOVER 2 studies, multiple sensitivity analyses were conducted for clinical response at 48 to 72 hours in the ITT population (refer to Appendix 3). Results of all of the sensitivity analyses were consistent with results of the primary efficacy analysis, with the treatment difference ranging from 0.2% to 3.3% in the DISCOVER 1 study and from 1.7% to 2.8% in the DISCOVER 2 study; the lower limit of the 95% CI for the treatment difference was always higher than –10%. This included the sensitivity analysis that excluded temperature from the primary outcome measure and defined success as at least a 20% decrease from baseline in lesion area.

In the DUR001-303 study, clinical response at 48 to 72 hours was observed in 84.2% of patients in the 2-dose dalbavancin arm and in 81.4% of the single-dose dalbavancin arm, with a between-group difference of –2.9% (95% CI, –8.5% to 2.8%). The lower limit of the 95% CI for the treatment difference of –2.9% in the ITT population was –8.5%, which was higher than the pre-specified lower limit of –10%. Therefore, noninferiority of 1-dose to 2-dose dalbavancin was concluded. In addition, clinical response at the EOT visit was almost identical between groups (treatment difference = –0.20%; 95% CI, –4.2% to 3.8%).

In the VER001-9 study, clinical response was assessed at day 14 (EOT) and day 28 (TOC) in the ITT population. The percentage of patients with clinical response at the TOC visit was 82.7% in the linezolid arm and 76.5% in the dalbavancin arm. The treatment difference was –6.15% (95% CI, –12.03% to –0.27%). The lower limit of the 95% CI for the treatment difference of –12.03% remained above the pre-specified lower limit of –12.5%. Therefore, noninferiority of dalbavancin to linezolid was claimed. However, because the upper bound of the 95% CI was just slightly below zero (–0.27%), the study authors felt that although dalbavancin may be considered statistically inferior to linezolid, the difference was not clinically meaningful. In terms of clinical response at the EOT visit, results were consistent with those observed at 48 to 72 hours, with 80.6% of patients in the dalbavancin arm and 86.9% of patients in the linezolid arm deemed clinical successes in the ITT analysis. Clinical response was also evaluated in the CE population at the TOC visit, and results were consistent with those of the ITT population (treatment difference = –2.21% in favour of linezolid; 95% CI, –7.28% to 2.86%).

Table 14: Clinical Response at 48 to 72 Hours Across Pivotal Trials and at the TOC Visit in the VER001-9 Trial (ITT Population)

CI = confidence interval; DAL = dalbavancin; ITT = intention-to-treat; LZD = linezolid; NR = not reported; OR = odds ratio; VAN = vancomycin.

Note: In the DISCOVER 1 and DISCOVER 2 studies, the OR of clinical response in the dalbavancin treatment group relative to the vancomycin treatment group stratified for the presence or absence of fever at baseline and the 95% CI for the OR were calculated using the Cochran-Mantel-Haenszel method.

^a95% CI was calculated using the Miettinen and Nurminen method without adjustment in the DUR001-303 study and with adjustment for the randomization stratification factor of presence or absence of fever at baseline in the DISCOVER 1 and DISCOVER 2 studies.

^bThe time point for assessment in the VER001-9 study is TOC (e.g., day 28).

^cBased on investigator assessment.

Source: Clinical Study Reports.^17-20

Reasons for clinical nonresponse are presented in Table 15. In the DISCOVER 1 and DISCOVER 2 studies, reasons for clinical nonresponse at 48 to 72 hours were similar in the 2 treatment arms. In the DISCOVER 1 study, the most common reasons for clinical nonresponse were fever only (approximately 29% of patients in each group), increase in lesion size only (27.1% in the dalbavancin group and 25.0% in the vancomycin group), and no evidence of fever but temperature criteria not met, as described in the SAP (25.0% of patients in each group). Most patients in both treatment groups were nonresponders, either because of fever or not meeting the criteria for temperature measurement (54.2% in the dalbavancin group versus 53.8% in the vancomycin group). In the DISCOVER 1 study, missing data were the reason for nonresponse in approximately 20% of patients with nonresponse status (22.9% of patients in the dalbavancin group and 19.2% in the vancomycin group).

In the DISCOVER 2 study, the most common reasons for nonresponse in the dalbavancin and vancomycin groups were the same as in the DISCOVER 1 study: fever only (31.4% versus 35.0%), increase in lesion size only (29.1% versus 22.5%), and no evidence of fever but temperature criteria not met, as described in SAP (24.4% versus 22.5%). The majority of patients in both treatment groups were nonresponders, either because of fever or not meeting the criteria for temperature measurement (55.8% in the dalbavancin group and 57.5% in the vancomycin group). In the DISCOVER 2 study, missing data accounted for a relatively small proportion of clinical nonresponse in each treatment group.

In the DUR001-303 study, the proportion of patients determined to be nonresponders for each of these reasons was similar across treatment groups in the ITT population. The most common reason for clinical nonresponse was lesion area not decreased by at least 20% relative to baseline (63.1% of patients in the single-dose group and 61.8% in the 2-dose group). Other reasons for clinical nonresponse included missing lesion data at any point (33.8% of patients in the single-dose group versus 32.7% in the 2-dose group), and lesion area collected outside of the 48- to 72-hour window (23.1% in the single-dose group versus 20.0% in the 2-dose group). A sensitivity analysis was conducted using a 36- to 75-hour time window instead of a 48- to 72-hour window, and 9 of the nonresponder patients in the single-dose group and 4 in the 2-dose group were considered responders. Deaths and use of rescue medications accounted for only a small proportion of the patients in each treatment group who were nonresponders (≤ 6.2% in the single-dose group and ≤ 7.3% in the 2-dose group).

In the VER001-9 study, there were 48 (8.4%) and 20 (7.1%) clinical failures in the dalbavancin and linezolid groups, respectively, in the CE population at the TOC visit (day 28). The most common reasons for clinical nonresponse were not reported.

Table 15: Reasons for Clinical Nonresponse at 48 to 72 Hours and on Day 28 Across Pivotal Trials (ITT Population)

ABSSSI = acute bacterial skin and skin structure infection; CE = clinically evaluable; DAL = dalbavancin; ITT = intention-to-treat; LZD = linezolid; NA = not applicable; TOC = test of cure; VAN = vancomycin.

Note: Patients could have had more than 1 reason for clinical failure. Lesion size was defined as lesion area only.

^aThe patient had at least 1 temperature measurement, but did not have 3 temperature measurements 3 to 9 hours apart in the 48- to 72-hour window and had no temperature > 37.6°C after the 48-hour time point.

^bThe patient initiated a new, systemic antibacterial with gram-positive activity for the ABSSSI under study in the first 72 hours.

^cAll temperature data in the 48- to 72-hour window are missing for the patient.

Source: Clinical Study Reports.^17-19

Clinical Status at EOT (Day 14) and Day 28

Summary results of clinical status at the EOT and day 28 visits in the ITT population and CE population are presented in Table 16. Clinical status of success or failure was not evaluated as defined in the pivotal trials in the VER001-9 study, and hence is not reported.

Clinical Status at EOT (Day 14)

In the DISCOVER 1 study, after adjustment for baseline infection type or fever status, the proportion of patients in the ITT population who achieved clinical success at the EOT visit (day 14) was similar in the comparator and dalbavancin groups (86.7 versus 81.9%), with a between-group difference of –4.8% (95% CI, –10.7% to 1.3%). Results of clinical success at the EOT visit were similar in the CE and ITT populations. In the DISCOVER 2 study, results were similar in the dalbavancin and comparator groups in the ITT population (88.7% versus 85.6%), with a between-group difference of 3.1% (95% CI, –1.8% to 8.0%). In the DISCOVER 2 study, clinical success at the EOT visit was higher in the CE population than in the ITT population for both treatment groups, and there was little difference between treatment arms (treatment difference = 0.8%; 95% CI, –3.3% to 5.0%). In the DUR001-303 study, there was little to no difference in the proportion of patients achieving clinical success at day 28 between the single-dose and 2-dose regimens of dalbavancin in the ITT population (treatment difference = –0.8%; 95% CI, –6.3% to 4.6%). This result was consistent with that in the CE population.

Clinical Status at Day 28

In all trials, the proportion of patients achieving clinical success were similar with dalbavancin and its comparator. In the DISCOVER 1 study, after adjustment for baseline infection type or fever status, the proportion of patients classified as a clinical success at day 28 was similar in treatment groups in the ITT populations (83.7% of patients in the dalbavancin group versus 88.1% in the comparator group; treatment difference = –4.4; 95% CI, –10.1% to 1.4%). In the DISCOVER 2 study, the proportion of patients achieving clinical success at SFU was similar in the 2 treatment groups in the ITT population (88.1% versus 84.5%; treatment difference = 3.6%; 95% CI, –1.1% to 8.9%). In the DUR001-303 study, there was little to no difference in the proportion of patients achieving clinical success at day 28 in the single-dose and 2-dose dalbavancin groups in the ITT population (treatment difference = –0.6%; 95% CI, –6.0% to 4.8%). Results for clinical success in the CE population at day 28 were consistent with those for the ITT population for all trials.

Table 16: Redacted