CADTH Reimbursement Review

Palovarotene (Sohonos)

Sponsor: Ipsen Biopharmaceuticals Canada, Inc.

Therapeutic area: Fibrodysplasia (myositis) ossificans progressiva

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Ethics Review

Stakeholder Input

Clinical Review

Executive Summary

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

Table 1: Submitted for Review

NOC = Notice of Compliance.

Introduction

Fibrodysplasia (myositis) ossificans progressiva (FOP) is an ultra-rare congenital disease in which uncontrolled, progressive, abnormal growth of bone in nonskeletal tissues (e.g., muscles, tendons, and ligament) takes place through the process of heterotopic ossification (HO). FOP is caused by a recurrent, heterozygous, activating mutation of ACVR1, which is a member of the protein family bone morphogenetic protein (BMP) type I receptors.^1,2 The mutation occurs as a random event during the formation of reproductive cells (eggs or sperm) in the patient’s biologic parent or during early embryonic development.³ Although FOP is a congenital condition, ossification does not occur before birth. HO occurs in infancy and progresses throughout life. It may occur without warning or following a flare-up induced by trauma (e.g., intramuscular childhood immunizations, falls, surgeries, biopsies, and so on) or various viral illnesses.⁴ In the affected areas, ossification eventually leads to stiffness and limited movement of joints. As the disease progresses, patients with FOP experience increasingly limited mobility, affecting their ability to balance, walk, and sit, and/or severely restricting their movement. The development of bone at multiple soft tissue sites eventually leads to ankylosis (fusion) of the affected joints, including the spine and thoracic cage. Patients whose jaws are affected have difficulty eating and/or speaking. Eventually, FOP may result in complete immobilization. The permanent and cumulative effects of HO result in severe functional limitations in joint mobility and progressive disability; most patients with FOP require a wheelchair by their third decade of life.^5,6 As mobility begins to deteriorate due to HO, patients with FOP are at increased risk for a multitude of health morbidities, including fractures, severe restrictive lung disease, right-sided congestive heart failure, scoliosis, pressure ulcers, severe weight loss (due to jaw ankylosis), gastrointestinal issues, and acute and chronic pain.^7-9 Hearing impairment occurs in approximately half of all patients with FOP.¹⁰ As disability progresses, health-related quality of life (HRQoL) decreases.^7-9 In addition to being extremely debilitating, FOP is associated with shortened lifespan;⁶ the estimated median lifespan is 56 years.¹¹ Death among patients with FOP is mainly due to complications of restrictive chest wall disease.⁶ The worldwide prevalence of FOP is estimated to be 1 person in 2 million.¹² However, due to globally high rates of misdiagnosis, the number of those with FOP may be closer to 1 person in 1 million.^6,13 The prevalence of FOP does not differ across sex, race, ethnicity, or geography.^14,15 Currently, there are approximately 900 confirmed cases worldwide.¹⁶ Based on registry data, the estimated prevalence of FOP in Canada is 0.559 per million persons.¹⁷ Based on clinical expert input, there are approximately 20 known patients with FOP in Canada.

Palovarotene is a selective agonist of retinoic acid receptor gamma expressed in chondrogenic cells and chondrocytes that acts as a transcriptional repressor.¹⁸ By binding to retinoic acid receptor gamma, palovarotene decreases BMP signalling and subsequently inhibits the BMP Smad 1/5/8 signalling pathway. Palovarotene interference with these pathways inhibits chondrogenesis. This allows for normal muscle tissue repair to occur, which reduces damage to muscle tissue.¹⁸

Palovarotene has been approved by Health Canada to reduce the formation of HO in adults and in children aged 8 years and above for females and 10 years and above for males with fibrodysplasia (myositis) ossificans progressiva. It is the first drug to be approved by Health Canada for this indication. Palovarotene underwent a priority review at Health Canada. The sponsor’s reimbursement request is aligned with the Health Canada indication. Health Canada recommends that palovarotene be administered at a dose of 5 mg once daily for chronic treatment. At the onset of the first symptom indicative of a FOP flare-up (or following a substantial, high-risk, traumatic event likely to lead to a flare-up), Health Canada recommends, under the guidance of a health care professional, a flare-up regimen of 20 mg once daily for 4 weeks followed by 10 mg once daily for 8 weeks, for a total of 12 weeks (i.e., a flare-up regimen of 20 mg to 10 mg), even if symptoms resolve more quickly. Chronic treatment with palovarotene should be reinitiated after the completion of the flare-up treatment. Weight-based dosage is required in children aged 14 years or younger.

The objective of this review is to evaluate the beneficial and harmful effects of palovarotene to reduce the formation of HO in adults and in children aged 8 years and above for females and 10 years and above for males with FOP.

Stakeholder Perspectives

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

Patient Input

Patient input was provided from the Canadian FOP Network (CFOPN) and the Canadian Organization for Rare Disorders (CORD) in 1 joint submission. Together, these patient groups collected input from 3 patients with FOP who were currently receiving palovarotene, and from 1 caregiver of a patient in Canada with FOP. Patient input was collected through telephone interviews conducted from October 31, 2022, to November 9, 2022. In addition, a summary of patient and caregiver public testimonies given before the FDA Endocrinologic and Metabolic Drugs Advisory Committee (October 31, 2022) was included.

Patients and caregivers reported experiencing stress from the unrelenting vigilance required to avoid activities that could result in injury and trigger flare-ups (with onset that is often unpredictable or without a precipitating event), resulting in permanent bone growth and contributing to the progressive loss of physical movement and independence. The disease deprives patients of normal activities and life experiences, contributing to stigmatization, isolation, and despair. The extreme mental and physical toll of FOP is underscored by the following quote from a caregiver:

“…My child is well aware of [her] body slowly failing…outward deformations, the inability to take full breaths…limited abilities to participate in social events and no…physical events. …Each day brings with it additional agonizing truths…not knowing what tomorrow will bring or the real potential for a much-shortened life span.”

Due to the absence of approved, effective therapies for FOP, patients are left to restrict their lifestyles to protect against potentially injurious flare-ups and control disease progression. However, giving up all potentially injurious activities deprives patients of enjoyable experiences and meaning while offering no assurance that flare-ups or disease progression will be prevented.

Patients with FOP expressed a desire for access to treatments that reduce symptoms and prevent disease progression. They listed the following treatment outcomes as important: maintenance of mobility or increased mobility; reduced frequency and severity of flare-ups; reduced pain; and reduced or halted new bone growth. For some patients, simply halting disease progression such that they could adapt to their existing disease state and continue to enjoy activities that provide meaningful experiences would bring them satisfaction. Of those patients who have had experience with palovarotene, all expressed a desire to continue its use for as long as possible. Treatment with palovarotene has allowed these patients to maintain or even increase their mobility and has provided them with hope for improved quality of life and independence with continued use.

Clinician Input

Input From Clinical Experts Consulted by CADTH

According to the clinical experts consulted by CADTH for the purpose of this review, an ideal treatment for patients with FOP would be one that could be administered as early as possible — even in utero — and that could prevent and/or reduce HO while prolonging life lived with good quality of life. The clinical experts agreed that palovarotene should be used as a first-line therapy, with supportive pharmacological and non-pharmacological measures in place. However, the clinical experts cautioned that the optimal time to start treatment with palovarotene for pediatric patients is uncertain. The clinical experts agreed that palovarotene would be the main treatment for both adults and children, and that for patients aged 16 years and older, palovarotene may be combined with corticosteroids during flare-ups. Bisphosphonates may also be used concurrently with palovarotene, regardless of age. The clinical experts highlighted the practice of minimizing polypharmacy in patients aged younger than 16 years. Accordingly, consideration should be given to initiating palovarotene as the sole treatment in patients aged younger than 16 years; corticosteroids or other drugs may be added when it is deemed clinically that adjuvant therapy is needed.

Regarding gene mutations associated with FOP, the clinical experts noted that because palovarotene works downstream of the ACVR1 receptor, patients harbouring any gene mutations of the receptor resulting in HO formation through the Smad 1/5/8 pathway may benefit from palovarotene treatment. Accordingly, any mutation that works on the Smad pathway and upregulates it should be amenable to treatment with palovarotene. The clinical experts noted that assessing patients’ or caregivers’ likelihood of compliance with the treatment regimen is important to determine suitability before prescribing palovarotene. The clinical experts agreed that patients who do not have ankylosis of the whole body would benefit from treatment with palovarotene to prevent progressive disability. The clinical experts highlighted that patients with more advanced disease may benefit less from palovarotene because the extent to which HO can be reversed is unclear; however, these patients may still derive benefits from palovarotene through the prevention of new bone formation or from HO reversal to preserve jaw function and lung capacity. Of note, the clinical experts stressed that the use of palovarotene in young patients whose growth plates have not yet fused should be undertaken only after careful consideration and consultation with patients and their families. The clinical experts opined that when considering treatment with palovarotene in patients with open epiphyses, there is still uncertainty as to whether the potential benefits of treatment outweigh the potential harms.

The clinical experts noted that preservation of function is the most important marker of treatment response. The clinical experts advocated the use of clinical parameters (e.g., the Cumulative Analogue Joint Involvement Scale [CAJIS]) to assess response to treatment. In addition, the clinical experts noted that response to treatment with palovarotene should include an assessment of the prevention of comorbidities, such as the inability to walk, the need for a wheelchair, the inability to work, the inability to continue performing activities of daily living, decreased respiratory function, hearing loss, and — for patients with advanced disease — decreased jaw and lung function.

The clinical experts anticipate that treatment with palovarotene would continue indefinitely. Accordingly, they advocated for continual assessment of the risk and benefits of its use. The clinical experts suggested that the efficacy and safety of palovarotene be assessed annually in adults, and assessed in children every 6 months for efficacy and every 3 months for safety. Given the progressive nature of FOP, the clinical experts suggested that treatment response should be monitored for at least 2 years before a decision to discontinue treatment is made, unless the decision to stop treatment is by patient choice, the patient is noncompliant with treatment, or the patient is experiencing intolerable AEs. The clinical experts noted that increasing HO load as detected by whole-body CT (WBCT) (excluding the head) and deteriorating CAJIS score may suggest lack of treatment response.

The clinical experts agreed that, due to the rarity of the condition, all patients with FOP should be diagnosed and managed by a specialist physician who is either experienced in the management of FOP or has academic expertise in FOP. In addition, palovarotene should be prescribed by such a specialist. The clinical experts suggested that all patients should be evaluated in person by their specialist for baseline assessment with the CAJIS before being prescribed palovarotene and for periodic reassessments. The clinical experts agreed that a family doctor could collaborate with the expert physician in the continued monitoring and treatment of palovarotene. In children, the palovarotene prescription would typically be in the hands of a bone disease specialist. The clinical experts agreed that, using a team approach led by a specialist, patients experiencing mucocutaneous adverse events (AEs) may visit their family doctor for follow-up care. The clinical experts agreed that oral treatment with palovarotene can be taken at home or in any outpatient setting. In the event of a flare-up, patients should be instructed to take pictures of the flare-up site and inform their physician. The clinical experts added that all patients should be provided with up to a 3-day supply of flare-up dosing in case a flare-up occurs when pharmacies are closed or the physician is unavailable, such as over a weekend. Given the issues associated with treatment assessment in this patient population (i.e., lack of validated instruments, heterogeneity among patients, heterogeneity over time within each patient, and flare-ups), the clinical experts suggested the creation of a pan-Canadian expert panel accessible to each jurisdiction to adjudicate both the initiation and renewal of palovarotene treatment to facilitate a national, consensus-based approach to access.

Clinician Group Input

Clinical group input was provided by 5 clinicians with experience treating patients with FOP who attended the Canadian Endocrine Update. Some of the clinicians providing input have participated in clinical trials for the drug under review, with 1 clinician reporting on a patient with experience using palovarotene. The main unmet need of patients with FOP identified by the clinician group was for treatment(s) that alter the natural course of the disease. The clinician group anticipates that palovarotene would be used as a single drug (with or without corticosteroids) administered daily, with the potential for short-term dose increases during flare-ups. Further, the clinical group noted that palovarotene has the potential to be used in combination with other investigational drugs with different mechanisms of action in the future. The clinical group suggests that all patients who meet the approved Health Canada indication for palovarotene should be considered for treatment. Accordingly, consideration for treatment initiation, as listed by the clinical group, suggested lifetime patient monitoring by a multidisciplinary care team that includes surgeons, rheumatologists, and specialists in pediatric and adult orthopedics. The clinician group suggested that treatment with palovarotene should be discontinued in the event of notable adverse effects (e.g., premature epiphyseal fusion in children) or in the event of, or intention for, pregnancy in individuals of child-bearing capabilities. According to input by the clinical group, there are currently no tools for measuring outcomes in patients with FOP. The clinical group indicated that the following outcomes should be used to determine response to treatment: annualized change in HO volume; maintenance of mobility and reduction in the rate of flare-ups; disease stability; and maintenance of HRQoL. The clinical group stressed that the decision to initiate and discontinue treatment with palovarotene should include careful patient counselling and shared decision-making.

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 impact the implementation of a CADTH recommendation for palovarotene:

• considerations for the initiation of therapy

• considerations for the continuation or renewal of therapy

• considerations for the discontinuation of therapy

• considerations for prescribing therapy

• care provision issues

• system and economic issues.

Clinical Evidence

Pivotal and Protocol-Selected Studies

Description of Studies

One sponsor-conducted study that met the CADTH review protocol criteria was included in this systematic review. The MOVE trial¹⁹ is an ongoing, multicentre, nonrandomized, open-label, phase III study evaluating the efficacy of palovarotene in decreasing new HO volume in adult patients and pediatric patients aged 4 years and older with FOP compared to that observed in untreated patients who participated in the sponsor-conducted FOP Natural History Study (NHS) (Study PVO-1A-001]).²⁰ The MOVE trial was conducted in 2 parts. In part A, eligible patients received chronic dosing with palovarotene for up to 24 months and underwent flare-up–based treatment if they experienced a flare-up (as defined a priori) or a traumatic event likely to lead to flare-up, as confirmed by the investigator. In part B, all patients were provided the chronic and flare-up palovarotene dosing regimen for an additional 24 months, until palovarotene was commercially available, to obtain longer-term safety data. No new patients were enrolled in part B of the MOVE trial. The primary efficacy end point for the MOVE trial was annualized change in new HO volume, with the key secondary outcome being the proportion of patients with new HO. Secondary outcomes included: number of body regions with new HO; proportion of patients reporting flare-ups; and flare-up rate per patient-month exposure. Exploratory outcomes included: change in range of motion (ROM) as measured by the CAJIS for FOP; change in physical function as measured by the FOP Physical Function Questionnaire (FOP-PFQ); change in physical and mental function for patients aged 15 years or older; change in mental function for patients aged younger than 15 years (using the Patient-Reported Outcome Measurement Information System [PROMIS]); and incidence and volume of catastrophic HO.

After the discovery of a high rate of premature epiphyseal fusion in growing children enrolled in the MOVE trial, and interruption of the study due to futility at the time of interim analysis 2, the target population was amended to include only adults and children aged 8 years and older (for females) and aged 10 years and older (for males). All analyses related to the target population were based on post hoc analyses.

Overall, patients in the target population were predominately white (MOVE = 74.7%; NHS = 75.0%) and approximately half were male (MOVE = 54.4%; NHS = 51.1%). Patients enrolled in the MOVE trial were, on average, younger than those enrolled in the NHS (14.4 years versus 20.4 years). A greater proportion of patients were aged 8 years to 14 years (females) or 10 years to 14 years (males) in the MOVE study versus the NHS (43.0% versus 23.9%). Moreover, patients in the MOVE study were, on average, younger than those in the NHS at the time of FOP diagnosis (6.5 years versus 7.5 years). Other notable imbalances in baseline characteristics between patients in the MOVE study and those in the NHS included reported hearing loss (MOVE = 45.6%; NHS = 35.2%); pain (MOVE = 73.4%; NHS = 85.2%); lethargy (MOVE = 7.6%; NHS = 26.1%); change in mood and behaviour during last flare-up (MOVE = 11.4%; NHS = 39.8%); unknown (MOVE = 73.4%; NHS = 44.3%) or other (MOVE = 8.9%; NHS = 23.9%) reported cause of last flare-up; bone formation as a result of last flare-up (MOVE = 51.9%; NHS = 0%); and slightly worsened loss of movement as a result of last flare-up (MOVE = 2.5%; NHS = 20.5%).

Efficacy Results

Annualized New HO

The Bayesian analysis with no square-root transformation and negatives set to 0 by body region fit a 25% reduction (ratio of mean change = 0.75; 95% credible interval [CrI], 0.51 to 1.11) in the volume of annualized new HO among patients treated with palovarotene in the MOVE trial compared to untreated patients in the NHS.

The post hoc analysis of annualized new HO with no square-root transformation and negative values included estimated an annualized new HO volume in patients treated with palovarotene in the MOVE trial and untreated patients in the NHS of 11,419 mm³ (standard error = 3,782) and 25,796mm³ (standard error = 6,066), respectively. Compared to patients in the NHS, a 55.7% reduction in mean annualized new HO volume was observed among patients treated with palovarotene in the MOVE trial. Based on the weighted linear mixed effect (wLME) model, patients treated with palovarotene in the MOVE trial had an estimated reduction in new HO volume of 10,443 mm³ per year (95% confidence interval [CI], –23,538 to 26,534 mm³ per year; P = 0.1124) compared to untreated patients in the NHS when controlling for baseline HO divided by age. Accordingly, the wLME model estimated a 49% reduction in mean annualized new HO volume in patients treated with palovarotene in the MOVE trial compared to untreated patients in the NHS. The Wilcoxon rank sum test reported P value was 0.0107.

Proportion of Patients With Any New HO

The proportions of patients with any new HO among those treated with palovarotene in the MOVE trial and those untreated in the NHS at month 12 were 62.2% and 57.4%, respectively.

Body Region With New HO

The proportions of patients with 0 body regions with new HO at month 12 were 37.8% and 42.6% in the MOVE trial and the NHS, respectively. The proportions of patients with 1 body region with new HO at month 12 were 31.1% and 22.1% in the MOVE trial and the NHS, respectively. No clear, consistent trends were observed differentiating patients in the MOVE trial from those in the NHS.

Catastrophic HO

The proportions of patients with catastrophic new HO volumes (i.e., exceeding 100,000 mm³, 50,000 mm³, or 30,000 mm³) at month 12 in the MOVE trial were 1.3%, 9.1%, and 11.7%, respectively. Among patients in the NHS, the proportions with catastrophic new HO volumes (i.e., exceeding 100,000 mm³, 50,000 mm³, or 30,000 mm³) at month 12 were 3.8%, 11.4%, and 13.9%, respectively.

The proportions of patients with catastrophic annualized new HO volumes (i.e., exceeding 100,000 mm³, 50,000 mm³, or 30,000 mm³) at the last time point in the MOVE trial were 1.3%, 6.5%, and 16.9%, respectively. At the last time point in the NHS, the proportions of patients with catastrophic new HO volumes (i.e., exceeding 100,000 mm³, 50,000 mm³, or 30,000 mm³) at month 12 were 6.3%, 15.2%, and 24.1%, respectively.

Reported Flare-Ups

Among patients in the target population, 67.1% and 63.9% of patients in the MOVE trial and the NHS reported flare-ups. Based on post hoc analysis of the principal safety set, the rates of flare-up per patient-month were 0.11 (95% CI, 0.08 to 0.16) in the MOVE trial and 0.06 (95% CI, 0.05 to 0.08) in the NHS.

HO at Flare-Up Sites

Overall, the mean volumes of new HO at flare-up sites were 19,610 mm³ (95% CI, 11,135 mm³ to 28,084 mm³) among patients in the MOVE trial and 40,157 mm³ (95% CI, 9,189 mm³ to 71,124 mm³) among patients in the NHS. The mean volumes of new HO away from flare-up sites (following a flare-up) were 7,626 mm³ (95% CI, 3,845 mm³ to 11,407 mm³) among patients in the MOVE trial and 26,399 mm³ (95% CI, 8,539 mm³ to 44,259 mm³) among patients in the NHS.

Range of Motion

ROM was assessed for 12 joints (i.e., shoulder, elbow, wrist, hip, knee, and ankle on both the right and left sides) and 3 body regions (i.e., jaw, cervical spine [neck], and thoracic and lumbar spine) using the CAJIS. The CAJIS is a clinician-administered analogue scale of gross mobility restriction.²¹ Total CAJIS scores can range from 0 points to 30 points, with a higher score indicating greater impairment. At baseline, mean baseline CAJIS scores were similar between patients treated with palovarotene in the MOVE trial (10.8; standard deviation [SD] = 6.4) and untreated patients in the NHS (12.6; SD = 7.0). At month 12, both patients treated with palovarotene in the MOVE trial and untreated patients in the NHS experienced an increase in mean CAJIS scores from baseline (indicating deterioration) of 0.6 (SD = 2.1) and 0.6 (SD = 2.4), respectively.

Physical Function

Physical function was assessed using age-appropriate forms of the FOP-PFQ. The FOP-PFQ is a disease-specific, patient-reported outcome measure that assesses physical function.²² Lower FOP-PFQ scores indicate more difficulty and, as a result, greater functional impairment. At baseline, the mean percentages of worse score on the FOP-PFQ were similar between patients treated with palovarotene in the MOVE trial (mean = 46.1; SD = 27.6) and untreated patients in the NHS (mean = 47.6; SD = 28.0). At month 12, both patients treated with palovarotene in the MOVE trial and untreated patients in the NHS experienced increases from baseline (indicating deterioration) in the mean percentage of worse score on the FOP-PFQ of 2.94 (SD = 8.09) and 4.70 (SD = 9.02), respectively.

Health-Related Quality of Life

HRQoL was assessed using age-appropriate forms of the PROMIS Global Health scale (short form). The PROMIS Global Health scale is a set of person-centred measures that evaluates and monitors physical, mental, and social health in adults and children in the general population or living with chronic conditions.²³ PROMIS scores were converted to T-scores for analysis. A T-score of 50 is normal, with an increment of 10 representing 1 SD away from the norm. A T-score of less than 50 is indicative of worse health, while a T-score of greater than 50 is indicative of better health. In patients aged 15 years and older, mean baseline PROMIS scores were similar between the MOVE trial and the NHS for Global Physical Health T-scores (MOVE = 43.15 [SD = 7.93]; NHS = 43.35 [SD = 8.66]) and Global Mental Health T-scores (MOVE = 52.17 [SD = 7.95]; NHS = 52.70 [SD = 9.40]). The mean changes from baseline on the Global Physical Health scale in patients in the MOVE trial at month 6, month 12, and month 18 were –0.15 (SD = 3.92), –0.20 (SD = 5.16), and –1.91 (SD = 6.28), respectively. Among patients in the NHS, the mean changes from baseline on the Global Physical Health scale at month 6, month 12, and month 18 were –0.37 (SD = 6.79), –1.19 (SD = 6.62), and –0.66 (SD = 5.57), respectively.

In patients aged younger than 15 years, baseline scores on the PROMIS Global Health scales were similar among patients in the MOVE trial and the NHS, at 43.2 (SD = 7.9) and 43.4 (SD = 8.7), respectively. Mean T-score changes from baseline among patients treated with palovarotene in the MOVE trial at month 6, month 12, and month 18 were –0.15 (SD = 3.92), 0.20 (SD = 5.16), and –1.91 (SD = 6.28), respectively. Among untreated patients in the NHS, the mean T-score changes from baseline at month 6, month 12, and month 18 were –0.37 (SD = 6.79), –1.19 (SD = 6.62), and –0.66 (SD = 5.57), respectively.

Harms Results

Adverse Events

In the MOVE trial, at least 1 AE was reported by 96% and 94.3% of patients receiving chronic and flare-up dosing regimens, respectively. The most commonly reported AEs were related to mucocutaneous issues (83.3%), including dry skin (52.5%) and rashes (19.2%); gastrointestinal issues (63.6%), including dry lips (34.3%); infections and infestations (58.6%), including upper respiratory infection (20.2%); and musculoskeletal and connective tissue disorders, including arthralgia (24.2%) and pain in an extremity (18.2%). Overall, the reporting of AEs was similar during the chronic and flare-up regimens, with the exception of AEs related to gastrointestinal issues, which tended to be reported more often with chronic dosing than during flare-up dosing (63.6% versus 47.1%).

Serious Adverse Events

At least 1 serious adverse event (SAE) was reported by 19.2% and 17.1% of patients during the chronic and flare-up dosing regimens, respectively. The most commonly reported SAE was epiphyses premature fusion, which was observed in 11.1% of patients during chronic dosing and in 10% of patients during flare-up dosing.

Dose Modification to AEs

More patients required dose modification of palovarotene due to AEs during flare-up treatment (40%) than during chronic treatment (11.1%). The most common reasons for dose modification were drug eruption (chronic dose = 3.0%; flare-up dose = 12.9%), generalized pruritis (flare-up dose = 8.6%), erythema (flare-up dose = 4.3%), and pruritis (flare-up dose = 4.3%).

Treatment Interruption Due to AEs

Treatment interruptions due to AEs occurred among 16.2% and 15.7% of patients during the chronic and flare-up dosing regimens, respectively. The most common AE leading to treatment interruption was epiphyses premature fusion, which occurred in 6.1% of patients during the chronic dosing regimen.

Withdrawals Due to AEs

Withdrawals from the MOVE trial due to AEs occurred in 6.1% and 5.7% of patients during the chronic and flare-up regimens, respectively. Withdrawal due to epiphyses premature fusion occurred in 1 patient in each of the dosing regimen phases.

Mortality

There were no deaths due to AEs during the study period.

Notable Harms

Of the notable harms of interest, dry skin and dry lips occurred in 52.5% and 34.3% of patients, respectively, during chronic treatment and in 45.7% and 20.0% of patients, respectively, during flare-up treatment. Epiphyses premature fusion was observed in 11.1% of patients during chronic dosing and in 10% of patients during flare-up dosing. Hearing loss, pneumonia, suicidal ideation, and fractures occurred in less than 5% of patients who received treatment in the MOVE trial. There were no reported cases of osteoporosis, low bone density, decreased bone density, or onycholysis.

Harms Related to Growth

The mean changes in linear height z score in female children aged 8 years to younger than 14 years, or male children aged 10 years to younger than 14 years, were –0.36 (SD = 0.43) and –0.20 (SD = 0.34) in those who received treatment with palovarotene in the MOVE trial and in untreated patients in the NHS, respectively. Among patients aged 14 years to younger than 18 years, the mean change in linear height among those who received treatment with palovarotene in the MOVE trial (–0.02; SD = 1.54) was less than it was for those who were untreated in the NHS (–0.55; SD = 1.61). Compared to untreated patients in the NHS, a greater proportion of patients in the MOVE trial aged 8 years to younger than 14 years (female), or 10 years to younger than 14 years (male) (61.3% versus 41.2%), and aged 14 years to younger than 18 years (92.3% versus 88.9%) were documented with a pathological growth velocity rate of less than 4 cm per year. A similar trend was observed for other growth measures among female patients aged 8 years to younger than 14 years or male patients aged 10 years to younger than 14 years, in which a greater proportion of patients treated with palovarotene in the MOVE trial were documented with a pathological growth velocity rate of less than 2 cm per year for knee height (61.3% versus 52.9%) and a pathological growth velocity rate of less than 1.5 cm per year for tibial length (60.7% versus 50.0%). Among patients aged younger than 18 years, the proportion of patients with any epiphyseal growth plate abnormalities documented at month 12 were similar in the MOVE trial and the NHS (both 45.8%).

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

AE = adverse event; CAJIS = Cumulative Analogue Joint Involvement Scale; CI = confidence interval; FAS = full analysis set: FOP-PFQ = Fibrodysplasia Ossificans Progressiva Physical Function Questionnaire; HO = heterotopic ossification; NA = not applicable; NHS = Natural History Study; OR = odds ratio; ROM = range of motion; SAE = serious adverse event; SD = standard deviation; SE = standard error; WDAE = withdrawal due to adverse event; vs. = versus; wLME = weighted linear mixed effect.

Notes: The target population consisted of patients aged at least 8 years (females) or at least 10 years (males). The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020 for patients aged 14 years or older.

^aNegative new HO volumes were set to 0 by body region.

^bThe wLME least squares mean and standard error of mean were estimated from a mixed model with dependent variable annualized new HO and independent variables included fixed effects of treatment and baseline total HO divided by baseline age and a random patient effect.

^cP value is based on post hoc analysis and cannot be used to draw inferences for this end point.

^dUnadjusted comparison of annualized new HO volumes between the MOVE trial and the NHS using the sum of ranks.

^eHarms data are not specific to the chronic or flare-up regimen.

Source: Clinical Study Report for the MOVE study.¹⁹

Critical Appraisal

The open-label, nonrandomized study design of the MOVE trial makes interpretation of the efficacy and safety of palovarotene challenging. Imbalances in baseline characteristics between the MOVE trial and the NHS were observed for hearing loss; symptoms of pain, lethargy, and change in mood and behaviour at last flare-up; unknown or reported cause of last flare-up; method of reporting general medical history; common medication conditions; and common new-onset medication. While baseline total HO volume divided by age at baseline was adjusted for in the comparative analysis using a wLME model, none of the mentioned characteristics were accounted for in either the prespecified or post hoc analyses. The presence of these imbalances suggests that the groups may also be imbalanced in terms of unknown or unmeasurable prognostic factors; as a result, confounding cannot be ruled out. Of note, age at FOP diagnosis differed between patients in the MOVE trial and the NHS, with a greater proportion of patients in the NHS aged 18 years or older compared to the MOVE trial (51.5% versus 30.4%). Based on input from the clinical experts, older age may be a marker of more severe disease (given that HO tends to increase with age). However, the clinical experts added that it is difficult to ascertain the prognostic implication of age at diagnosis or the effect that age may have on treatment. The MOVE trial and the NHS were also imbalanced in terms of the proportions of patients who discontinued. Overall, a greater proportion of patients in the NHS versus the MOVE trial discontinued the study (71.1% versus 17.8%). It is uncertain whether those patients who discontinued from the NHS differed from those who remained. Accordingly, attrition bias cannot be ruled out.

To mitigate bias associated with open-label studies, the sponsor employed an objective, consistent, blinded outcome assessment of annualized change in new HO volume as assessed by a low-dose, WBCT. However, evidence for the validity of the measure is limited, and no evidence was found for the reliability or responsiveness of the measure. For more subjective outcomes, such as HRQoL and certain harms, there was the potential for performance and observer bias due to the open-label design.

Of note, there is uncertainty in what are considered clinically meaningful increases and decreases in new HO volume, as well as in what duration is required to observe a clinically meaningful treatment effect on HO. The clinical experts noted that it is uncertain if overall HO volume is a definitive surrogate for disease progression because the location of HO is an important consideration when evaluating the impact of changes in HO volume on functional outcomes. According to the clinical experts, patient-centred outcomes, such as those related to the preservation of function and HRQoL, are more meaningful markers of treatment response than HO volume alone in patients with FOP. Moreover, the MOVE trial did not assess a variety of other clinically relevant end points, such as pain associated with flare-ups; use of mobility and nonmobility aids; use of personal care tools and bathroom aids; hearing loss; and survival. Of particular concern was the absence of information on the treatment effect of palovarotene on respiratory function, given that most patients with FOP die because of respiratory issues related to ossification in their chest and lungs.

Regarding the statistical analysis of the treatment effect, the initial analysis of efficacy results followed a defined statistical analysis plan. However, unanticipated negative HO values led to deviation from the statistical analysis plan. At the time of the second interim analysis, statistical futility was determined based on the prespecified analysis, and dosing for patients in the FOP palovarotene program was paused. In response, the Data Monitoring Committee and sponsor were unblinded to all study data. After unblinding and review of the post hoc efficacy analysis, the study continued, and the data analysis was changed to accommodate negative HO values. Analyses of the target population were based on a post hoc analysis using a modified statistical analysis plan. Post hoc efficacy analyses are generally not recommended for drawing conclusions based on statistical measures of inference. Instead, assessment should focus on the estimated measures of treatment effect, the measure of uncertainty associated with those estimates, and the limitations of the analysis from which the estimates were generated. Several limitations across the multiple analytical approaches employed in the MOVE trial — including strong modelling assumptions that were not evaluated, informative missing data issues, and inappropriate or inadequate adjustment for key confounding variables — restricted the CADTH review team’s ability to draw conclusions from this evidence.

While the clinical experts noted that FOP is variable and heterogenous in its presentation and progression, they felt that the study population was reflective of what is observed in the clinical setting.

Other Relevant Evidence

Description of Studies

Two sponsor-conducted, phase II studies were described to provide additional context to the safety and efficacy of palovarotene for the treatment of FOP: Study 201 and Study 202. Study 201 (N = 80) was a phase II, randomized, double-blind, placebo-controlled trial evaluating the ability of different doses of palovarotene to prevent HO at the flare-up site in patients with FOP compared to placebo. Results are not reported for Study 201 in this CADTH clinical report because the study population and intervention were not aligned with the Health Canada indication and recommended dosage.

Study 202 was a phase II, multicentre, open-label study evaluating the safety and efficacy of different palovarotene dosing regimens in patients with FOP.²⁴ The study was conducted in 3 parts. The CADTH clinical review of Study 202 focused on parts B and C, in which the intervention was aligned with the Health Canada–recommended dosage. In part B, 54 pediatric and adult patients received chronic and flare-up treatment with palovarotene for up to 24 months. In part C, 48 patients from part B were followed for an additional 48 months. The primary end points for parts B and C were the proportions of flare-ups with no new HO at week 12 and the annualized change in new HO volume, respectively. Key secondary end points for parts B and C were change in ROM as assessed by CAJIS and change in physical function as assessed by the FOP-PFQ and PROMIS Global Physical Health T-score. The patient population in part B of Study 202 was mostly white (75.9%) and mostly female (57.4%), with a median age of 19.0 years (range, 7 years to 54 years).

Efficacy Results

In part B, 72.5% of patients had flare-ups with no new HO at week 12. In part C, the changes in mean volume of new HO from baseline were 9,332 mm³, 62,836 mm³, and 89,487 mm³ at months 12, 24, and 36, respectively.

The changes in CAJIS scores from baseline in part B were 1.1 and 15.0 at months 12 and 24, respectively; in part C, these were 0.5 and 2.7 at months 12 and 24, respectively. The changes in FOP-PFQ total scores from baseline in part B were 3.34 and 0.89 at months 12 and 24, respectively; in part C, these were 2.79 and 8.68 at months 12 and 24, respectively. The changes in adult PROMIS Global Physical Health T-scores from baseline in part B were –0.3 and 2.5 at months 12 and 24, respectively; in part C, these were 0.33 and –1.78 at months 12 and 24, respectively.

Harms Results

In Study 202, all patients in part B reported at least 1 AE, while 93.0% and 94.4% of patients receiving chronic and flare-up treatment in part C reported at least 1 AE, respectively. The most frequently reported AEs (reported by more than 30% of patients) for parts B and C were dry skin, skin exfoliation, arthralgia, and pain in an extremity. Serious AEs were reported in 11.4% of patients in the chronic dosing group and 8.6% of patients in the flare-up dosing group for part B, and in 11.6% of patients in the chronic dosing group and 22.2% of patients in the flare-up dosing group for part C. Dose modifications due to AEs occurred in 1 patient (2.3%) in the chronic dosing group and in 9 patients (25.7%) in the flare-up dosing group for part B, and in 14 patients (38.9%) in the flare-up dosing group of part C. Dose interruptions due to AEs occurred in 7 patients (15.59%) in the chronic dosing group and 4 patients (11.4%) in the flare-up dosing group for part B, and in 4 patients (9.3%) in the chronic dosing group and 6 patients (16.7%) in the flare-up dosing group for part C. Drug discontinuation due to AEs occurred in 2 patients (5.7%) in the flare-up dosing group for part B, and in 1 patient (2.3%) in the chronic dosing group and 1 patient (2.8%) in the flare-up dosing group for part C. No deaths were observed at any point in Study 202.

The proportions of patients experiencing an AE, SAE, dose interruption due to AE, drug discontinuation due to AE, or death during Study 202 were comparable to those observed in the MOVE trial. The proportion of patients in Study 202 requiring dose modifications due to AEs was slightly lower than in the MOVE trial.

Critical Appraisal

Parts B and C of Study 202 were limited by the noncomparative study design and lack of statistical testing, which precludes causal conclusions. In the absence of a comparative arm, conclusions about the comparative efficacy and safety of palovarotene versus the current standard of care cannot be drawn.

Conclusions

Based on the MOVE trial, treatment with palovarotene may have resulted in less annualized new HO volume in adults and children aged at least 8 years (females) or at least 10 years (males) with FOP. The strong potential for bias due to imbalances between the MOVE and NHS groups and unmeasured confounding, as well as attrition, introduced uncertainty into the magnitude of the treatment effect. Interpretation of the results for the primary end point must also take into consideration the limitations of the reported analyses due to the modelling approaches employed and the post hoc nature of the analyses. Direct conclusions based on measures of statistical inference alone are not recommended. Rather, assessment should focus on the estimated effects, with consideration for the associated uncertainty and noted limitations. The clinical relevance of the results is unclear because HO volume is neither a patient-centred outcome nor used in clinical practice, and there are no available minimally important difference (MID) estimates. No reductions in the number of reported flare-ups, ROM, physical function, or HRQoL were observed, and clinically important outcomes, such as respiratory function (including the need for ventilation and survival) were not assessed in the MOVE trial. Treatment with palovarotene carries an increased risk of premature epiphyseal fusion and may negatively affect linear growth, highlighting the need for ongoing assessment of its risk-benefit profile in growing children.

Introduction

Disease Background

FOP is an ultra-rare congenital disease of uncontrolled, progressive, abnormal growth of bone in nonskeletal tissues (e.g., muscles, tendons, and ligament) through the process of HO. This process causes bone to form outside the skeleton and is often associated with painful, recurrent episodes of soft tissue swelling (flare-ups). FOP is caused by a recurrent heterozygous activating mutation of ACVR1, which is a member of the protein family of BMP type I receptors.^1,2 The mutation occurs as a random event during the formation of reproductive cells (eggs or sperm) in the patient’s biologic parent or during early embryonic development.³ The ACVR1 protein is normally activated at appropriate times by ligand molecules.³ Activation occurs when ligands, such as BMPs or activin A, bind to the receptor to transmit signals through BMP Smad signalling pathways.² The protein FKBP12 can inhibit ACVR1 by binding to the receptor and preventing inappropriate activation in the absence of ligands.³ The most common FOP variant substitutes the protein building block (amino acid) with histidine for the amino acid arginine at position 206 of the ACVR1 protein (R206H).³ The R206H variant changes the shape of the ACVR1 protein, which disrupts the binding of the inhibitor protein FKBP12. As a result, the receptor is always “on,” even in the absence of ligands,³ causing the overgrowth of bone and cartilage associated with the signs and symptoms of FOP.³ The worldwide prevalence of FOP is estimated to be 1 in 2 million.¹² However, due to globally high rates of misdiagnosis, some believe the number of those with FOP may be closer to 1 in 1 million.^14,25 The prevalence of FOP does not differ across sex, race, ethnicity, or geography.^14,15 There are approximately 900 confirmed cases of FOP worldwide.¹⁶ Based on registry data, the estimated prevalence of FOP in Canada is 0.559 per million persons.¹⁷ Based on clinical expert input, there are approximately 20 known patients with FOP in Canada.

Characteristics and Disease Progression of FOP

At birth, FOP is most often characterized by a shortened great toe with a malformed distal first metatarsal and a missing or abnormal first phalanx and/or interphalangeal joint.^26,27 Other congenital signs of FOP include inward turning of the great toe toward the other toes (hallux valgus), abnormally short fingers and toes (macrodactyly), and/or permanent fixation of the fifth finger in the bent position (clinodactyly), as well as proximal medial tibial osteochondromas, malformation of the cervical vertebrae, and an abnormally short broad neck of the femur.^28-30

Although FOP is a congenital condition, ossification does not occur before birth. Progressive bone formation in the connective tissues (i.e., HO) occurs during infancy and early childhood and progresses throughout life. HO may occur without warning or following a flare-up induced by trauma (e.g., intramuscular childhood immunizations, falls, surgeries, biopsies, and so on) or various viral illnesses.⁴ The first signs of HO are the appearance of firm, tender swellings in certain parts of the body,³¹ which may be accompanied by redness or a feeling a warmth.⁴ The neck, back, chest, arms, and legs are usually the first areas affected, followed by the hips, ankles, wrists, elbows, shoulders, jaw, and abdominal wall. These soft tissue swellings mature through a cartilage-to-bone (endochondral) pathway to form mature heterotopic bone.

In the affected areas, ossification eventually leads to stiffness and limited movement of joints. As the disease progresses, patients with FOP experience increasingly limited mobility, affecting balance, walking, and sitting; they may also have severely restricted movement. The development of bone at multiple soft tissue sites eventually leads to ankylosis (fusion) of the affected joints. In addition, deformity of the spine, including side-to-side (scoliosis) and front-to-back (kyphosis) curvature, may occur. Patients whose jaws are affected will have difficulty eating and/or speaking. Eventually, FOP may result in complete immobilization. Patients often experience progressive pain and stiffness in the affected areas caused by abnormal bone growth that compresses the nerves in these areas (entrapment neuropathies). Of note, progression of ossification is not uniform across patients; while progression may be rapid in some patients, it may be gradual in others. Indeed, twin studies have shown that while disease phenotype is influenced by genetic determinants during prenatal development, environmental factors influence the postnatal progression of HO.³²

Morbidity and Mortality

The permanent and cumulative effects of HO result in severe functional limitations in joint mobility and progressive disability, such that most patients require a wheelchair by their third decade of life.^5,6 As mobility begins to deteriorate due to HO, patients with FOP are at increased risk for a multitude of health morbidities, including fractures, severe restrictive lung disease, right-sided congestive heart failure, scoliosis, pressure ulcers, severe weight loss (due to jaw ankylosis), gastrointestinal issues, and acute and chronic pain.^7-9 Hearing impairment occurs in approximately half of all patients with FOP.¹⁰

As disability progresses, the way FOP is managed changes, with a greater focus on daily care and support and increased reliance on living adaptations and assistance from caregivers.¹¹ In turn, FOP is associated with negative financial impacts due to the costs associated with living adaptations (e.g., home renovations, lifestyle changes, assistive devices), changes to career plans, and missed days of work for both patients and primary caregivers.¹¹ Consequently, HRQoL decreases as FOP progresses. Patients assessed as having the most severe physical limitations report an HRQoL that is close to the equivalent of death (i.e., a mean score of 0.05 on the 5-Level EQ-5D).¹¹

In addition to being extremely debilitating, FOP is associated with a shortened lifespan.⁶ The median lifespan for patients with FOP is 56 years (95% CI, 51 to 60).⁶ Death among patients with FOP is due mainly to complications of restrictive chest wall disease;^6,13 the most common causes are cardiorespiratory failure from thoracic insufficiency syndrome (with median age of 42 years at death) and pneumonia (with median age of 40 years at death).⁶ According to the clinical experts consulted by CADTH for the purpose of this review, the oldest known living person with FOP to date is aged 52 years.

Diagnosis

FOP is diagnosed clinically based on medical history and clinical examination. An accurate diagnosis can be made based on characteristic malformation of the patient’s big toe in addition to rapidly changing bony swellings on the head, neck, limbs, or back.²⁶ The clinical diagnosis of FOP is confirmed by genetic testing.²⁶ The clinical experts acknowledged that while genetic testing is readily available, there is a lack of awareness of FOP in Canadian practice, which can lead to diagnostic delays. Excess bone formation may be detected on X-rays, CT scans, and MRI, but imaging is not required for diagnosis.

Due to the rarity of the condition, FOP is commonly misdiagnosed as conditions such as cancer, aggressive juvenile fibromatosis, or fibrous dysplasia.^16,33-35 The rate of misdiagnosis of FOP is estimated to be 90% worldwide.³⁵ Unfortunately, during the course of investigating other possible causes for a patient’s signs and symptoms, catastrophic and irreversible damage often results from invasive procedures that are contraindicated for patients with FOP, such as intramuscular injections, biopsies and surgeries.^36,37

Standards of Therapy

The main treatment goals of FOP are to prevent HO and its associated disabilities and to prolong life. Accordingly, the current management is supportive and directed at preventing flare-ups (which trigger HO) and managing symptoms. According to the clinical experts consulted by CADTH for the purpose of this review, a critical component in the management of FOP is the prevention of trauma that may induce flare-ups and HO formation, with a focus on fall prevention. However, flare-ups can occur because of viral illnesses, stress, or for no obvious reasons, and are not necessarily related to trauma. Physiotherapy with gentle exercise, including ROM exercises and breathing exercises to maintain lung capacity, may also be used.

In the event of flare-ups, patients are prescribed a short, high-dose course of an oral glucocorticoid to lower pain and inflammation. Nonsteroidal anti-inflammatories and COX-2 inhibitors can also be used to treat the inflammation. However, while these measures may reduce pain, there is no evidence that these prevent HO. According to the clinical experts, the following drugs have been tried in this patient population: antihistamines, muscle relaxants, mast cell inhibitors, selective tyrosine kinase inhibitors, interlueukin-1 inhibitors, and leukotriene receptor antagonists. However, these medications are usually reserved for flare-ups that are refractory to other medications or for compassionate use, and their use is not supported by robust clinical evidence. The clinical experts noted that based on anecdotal evidence, bisphosphonates may be used for patients who are co-diagnosed with osteoporosis because these decrease bone remodelling and have an angiogenic effect. The clinical experts highlighted the use of nerve pain medications to manage nerve entrapment issues that may result from FOP. Finally, the clinical experts added that standard of care for patients with FOP may also involve regular assessment by a physiotherapist or an occupational therapist trained to alleviate disability using appropriate mobility aids and tools that can assist patients with self-care and hygiene to maintain their quality of life.

Currently, palovarotene is the only treatment approved for the treatment of FOP.

Drug

Palovarotene is a selective agonist of retinoic acid receptor gamma, a receptor expressed in chondrogenic cells and chondrocytes that acts as a transcriptional repressor.¹⁸ By binding to retinoic acid receptor gamma, palovarotene decreases BMP signalling and subsequently inhibits the Smad 1/5/8 signalling pathway. Palovarotene interference with these pathways inhibits chondrogenesis, thereby allowing for normal muscle tissue repair to occur. This reduces damage to muscle tissue.¹⁸

Palovarotene has been approved by Health Canada to reduce the formation of HO in adults and children aged 8 years and above for females and 10 years and above for males with fibrodysplasia (myositis) ossificans progressiva. Palovarotene underwent a priority review at Health Canada. The sponsor’s reimbursement request is aligned with the Health Canada indication. Health Canada recommends that palovarotene be administered at a dose of 5 mg once daily for chronic treatment. At the onset of the first symptom indicative of an FOP fare-up, or of a substantial high-risk traumatic event likely to lead to a flare-up, Health Canada recommends, under the guidance of a health care professional, a flare-up regimen of 20 mg once daily for 4 weeks followed by 10 mg once daily for 8 weeks, for a total of 12 weeks of treatment (known as a 20 mg to 10 mg flare-up regimen), even if symptoms resolve before 12 weeks. In the presence of persistent flare-up symptoms, the flare-up regimen may be extended in 4-week intervals with 10 mg palovarotene. Chronic treatment with palovarotene should be reinitiated after completion of the flare-up treatment. A weight-based dosage is required in children who are under 14 years of age. In the event of intolerable AEs during either the chronic or flare-up treatment, Health Canada recommends reducing the daily dose of palovarotene in a stepwise way at the discretion of the treating physician. Discontinuation of palovarotene should be considered if chronic dose reduction is not tolerated. Palovarotene may be used only for treating flare-ups.

The product monograph for palovarotene documents the following common side effects: dry skin, dry lips, pruritus, alopecia, rash, erythema, skin exfoliation, dry eyes, skin reaction, chapped lips, drug eruption, headache, paronychia, arthralgia, dry mouth, epistaxis, skin irritation, cheilitis, and nausea.¹⁸ The product monograph also contains warnings for serious side effects — including cellulitis and premature epiphyseal fusion in growing children — and for teratogenic risk.¹⁸

Stakeholder Perspectives

Patient Group Input

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

Patient input was provided by CFOPN and CORD in 1 joint submission. CFOPN is a nonprofit charitable organization staffed by volunteers that provides support and education and raises funds for research to support children and adults with FOP. CORD is a national network that works with government bodies, researchers, clinicians, and industry to advocate for health policy and health care systems, research, diagnosis, treatment, and services for patients with rare disorders in Canada. Together, the 2 groups provided input from 3 patients with FOP who are currently receiving palovarotene and from 1 caregiver of a patient in Canada with FOP. Patient input was collected through telephone interviews between October 31, 2022, and November 9, 2022. A summary of public testimonies from patients and caregivers, given before the FDA Endocrinologic and Metabolic Drugs Advisory Committee (October 31, 2022), is also included.

Patients and caregivers reported experiencing stress from the “unrelenting vigilance” required to avoid activities that may result in injury and trigger flare-ups (with onset that is often unpredictable or without a precipitating event), given that these can result in permanent bone growth that contributes to the progressive loss of physical movement and independence. The disease deprives patients of normal activities and life experiences, contributing to stigmatization, isolation, and despair. The extreme mental and physical toll of FOP is underscored by the following quote from a caregiver:

“…My child is well aware of [her] body slowly failing…outward deformations, the inability to take full breaths…limited abilities to participate in social events and no…physical events. …Each day brings with it additional agonizing truths…not knowing what tomorrow will bring or the real potential for a much-shortened life span.”

Due to absence of approved, effective therapies for FOP, patients are left with restrict their lifestyles to protect against potentially injurious flare-ups and to control disease progression. However, giving up all potentially injurious activities deprives patients of enjoyable experiences and meaning while providing no assurance of flare-up prevention or disease progression.

Patients with FOP expressed a desire for access to treatments that reduce symptoms and prevent disease progression. Patients listed the following treatment outcomes as important: maintenance of or increased mobility, reduced frequency and severity of flare-ups, reduced pain, and reduced or halted new bone growth. For some patients, simply halting disease progression, such that they could adapt to their existing disease state and continue to enjoy activities that provide meaningful experiences, would bring them satisfaction. Of those patients who have had experience with palovarotene, all expressed a desire to continue its use for as long as possible. Treatment with palovarotene has allowed these patients to maintain or even increase their mobility and has provided them with hope for improved quality of life and independence with continued use.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol; assisting in the critical appraisal of clinical evidence; interpreting the clinical relevance of the results; and providing guidance on the potential place in therapy). In addition, as part of the palovarotene review, a panel of 4 clinical experts from across Canada was convened to characterize unmet therapeutic needs, assist in identifying and communicating situations where there are gaps in the evidence that could be addressed through the collection of additional data, promote the early identification of potential implementation challenges, gain further insight into the clinical management of patients living with a condition, and explore the potential place in therapy of the drug (e.g., potential reimbursement conditions). A summary of this panel discussion is presented in the following section.

Unmet Needs

Disease progression in patients with FOP is characterized by progressive HO restricting joint mobility and lung function. As noted by the clinical experts, HO can occur in the absence of flare-up events. Based on input from the clinical experts, formation of HO in the jaw restricts eating and increases the risk of aspiration, while HO formation in the rib cage restricts lung capacity, increases the chance of fatal pneumonia, and causes thoracic insufficiency syndrome. Consequently, the outcomes associated with FOP are devastating, leading to severe disability, poor quality of life, and shortened life span. Currently, there are no medical interventions that can prevent HO or reverse its formation. According to the clinical experts consulted by CADTH for the purpose of this review, an ideal treatment for patients with FOP would be 1 that could be administered as early as possible — even in utero — and that could prevent and/or reduce HO while prolonging life lived with good quality of life. However, treatment during adulthood to prevent HO and progressive disability and preserve as much independence as possible is also very important.

Place in Therapy

To date, palovarotene is the only approved drug that addresses HO in patients with FOP. The clinical experts agreed that palovarotene should be used as first-line therapy, with supportive non-pharmacological measures in place. However, the clinical experts cautioned that the optimal time to start treatment with palovarotene for pediatric patients is uncertain.

The clinical experts agreed that palovarotene would be the main treatment for both adults and children. The clinical experts agreed that for patients aged 16 years and older, palovarotene may be combined with corticosteroids during flare-ups. Bisphosphonates may also be used concurrently with palovarotene, regardless of age. The clinical experts highlighted the practice of minimizing polypharmacy in the pediatric population under the age of 16. Accordingly, consideration should be given to initiating palovarotene as the sole treatment in patients under the age of 16; corticosteroids or other drugs may be added when it is deemed clinically that adjuvant therapy is needed.

The clinical experts noted that patients receiving treatment with palovarotene should be instructed to use sunscreen and other supportive measures to lessen the burdens associated with retinoid-based medications (e.g., sun sensitivity and dry skin). In addition, due to the teratogenic properties of palovarotene, the clinical experts stressed that patients of child-bearing potential should be directed to use effective birth control.

Patient Population

Regarding gene mutations associated with FOP, the clinical experts noted that because palovarotene works downstream of the activating mutation of ACVR1, patients harbouring any gene mutations of the receptor resulting in HO formation through the Smad 1/5/8 pathway may benefit from palovarotene treatment. Accordingly, any mutation that works on the Smad pathway and upregulates it should be amenable to treatment with palovarotene. The clinical experts expected that this would apply to all patients diagnosed with FOP, regardless of the associated mutation, given that it is unlikely that all variants are known.

The clinical experts noted that it is important to assess patient and/or caregiver likelihood of compliance with the treatment regimen to determine suitability before prescribing palovarotene. The clinical experts agreed that patients who do not have ankylosis of the whole body would benefit from treatment with palovarotene to prevent progressive disability. The clinical experts highlighted that patients with more advanced disease may benefit less from palovarotene because the extent to which HO can be reversed is unclear. However, the clinical experts added that patients with advanced disease may still derive benefits from palovarotene through the prevention of new bone formations or the reversal of HO, and that it is important to preserve jaw function and lung capacity in patients with advanced disease.

Of note, the clinical experts stressed that the use of palovarotene in young patients whose growth plates have not fused should be undertaken only after careful consideration and consultation with patients and their families. The main concern about premature epiphyseal fusion is that it leads to smaller stature. In addition, the clinical experts explained that when plates fuse prematurely, there is a theoretical risk that they may do so asymmetrically, leading to uneven limb length. The clinical experts noted that, given the severity of FOP, patients may be willing to accept significant side effects that are considered largely cosmetic in nature. The clinical experts were of the opinion that, when considering treatment with palovarotene in patients with open epiphyses, there is still uncertainty with regards to whether the potential benefits of treatment with palovarotene outweigh the potential harms.

Assessing Response to Treatment

The clinical experts noted that the frequency of flare-ups is not consistent over time; flare-ups are expected to occur more frequently at younger ages (i.e., in teenagers and those in their twenties). Accordingly, this frequency on its own may not be a reliable marker of response to treatment because it would be difficult to ascertain whether any decrease was a result of treatment or of patients getting older. The clinical experts agreed that, although total HO volume is a suitable outcome for proof of concept, it is not the best marker for treatment efficacy because the location of HO (denoting the functional area affected) may be more important than total HO for some patients. In addition, some patients may not wholly fit into a CT scanner, depending on how their joints have been immobilized by HO.

The clinical experts noted that preservation of function is the most important marker of treatment response. They advocated the use of clinical parameters to assess response to treatment. The CAJIS, which assesses gross mobility restriction, is an instrument used in clinical practice. Although the CAJIS is not sensitive to subtle changes and does not capture respiratory function, the clinical experts noted that it is a patient-centred tool that captures changes in function over long periods of time. The clinical experts noted that a CAJIS score of 27 points or greater is associated with death. For patients with advanced disease, measurement of jaw and lung function may be more appropriate to assess suitability for treatment with palovarotene. In addition, the clinical experts noted that response to treatment with palovarotene should include assessing for the prevention of comorbidities, such as the inability to walk, the need for a wheelchair, the inability to work, the inability to continue performing activities of daily living, respiratory function, and hearing loss.

The clinical experts anticipate that treatment with palovarotene would continue indefinitely. Accordingly, they advocated for continual assessment of the risk and benefits of its use. They also suggested that the efficacy and safety of palovarotene be assessed annually in adults, and assessed in children every 6 months for efficacy and every 3 months for safety. With time, safety assessments every 3 months in children may be extended to every 6 months if the patient is doing well and the impact of premature epiphyseal fusion lessens because the patient is approaching adult height. Given the issues associated with treatment assessment in this patient population (i.e., lack of validated instruments, heterogeneity among patients, heterogeneity over time within each patient, and flare-ups), the clinical experts suggested the creation of a pan-Canadian expert panel that would be accessible to each jurisdiction to adjudicate both the initiation and renewal of palovarotene treatment. This panel would be similar to the one currently in place to assess patients’ eligibility to be reimbursed for asfotase alfa for the treatment of hypophosphatasia, facilitating a national, consensus-based approach to access.

Discontinuing Treatment

Given the progressive nature of FOP, the clinical experts suggested that treatment response should be monitored for at least 2 years before a decision to discontinue treatment is made, unless the decision to stop treatment is by patient choice, the patient is noncompliant with treatment, or the patient is experiencing intolerable AEs. The clinical experts noted that AEs, including dry skin, rash, pruritus, sun sensitivity, hair loss, and pancreatitis, are important factors in deciding whether to discontinue treatment. The clinical experts noted that mucocutaneous side effects in children should be assessed because it may be more difficult for children to advocate for themselves to stay hydrated. The clinical experts noted that an increasing HO load (as detected by WBCT, excluding the head) and a deteriorating CAJIS score may suggest lack of treatment response; however, even a decrease in the rate of disease progression (compared with the pre-treatment rate) would still be beneficial. In patients who are unable to undergo CT scans, the determination of response to treatment may include measurement of lung function.

Prescribing Conditions

The clinical experts agreed that all patients with FOP should be diagnosed and managed by a specialist physician who is either experienced in the management of FOP or has academic expertise in FOP, due to the rarity of the condition. In addition, palovarotene should be prescribed by such a specialist. The clinical experts suggested that all patients should be evaluated in person by their specialist for baseline assessment with the CAJIS before being prescribed palovarotene and for periodic reassessments afterward. The clinical experts agreed that a family doctor could collaborate with the expert physician on the continued monitoring of the patient and their treatment with palovarotene. In children, the palovarotene prescription would typically be in the hands of a bone disease specialist. The clinical experts agreed that, based on a team approach led by a specialist, a patient experiencing mucocutaneous AEs could visit their family doctor for follow-up care. The clinical experts explained that many patients do not live in the same province as their specialist; therefore, telehealth and a pediatrician or adult physician who is local and accessible after-hours should be available to patients. Moreover, the clinical experts suggested that telehealth or other virtual platforms could be used to help patients access step-up treatment for flare-ups.

The clinical experts agreed that oral treatment with palovarotene can be taken at home or in any outpatient setting. The clinical experts suggested that, in the event of a flare-up, patients could be instructed to take pictures of the flare-up site and inform their physician. The clinical experts added that all patients should be provided with up to a 3-day supply of flare-up treatment in the event that a flare-up occurs when pharmacies are closed or physicians are not available (such as over a weekend). The clinical experts noted that patients and caregivers tend to recognize flare-ups as these occur, and that to date, the clinical experts have not observed any false positives when parents or caregivers report a flare-up. However, patients are sometimes unsure if they are experiencing a flare-up and may wait a day or 2 to call their physician.

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by 5 clinicians from 1 clinician group. The full original clinician group input received by CADTH has been included in the stakeholder section at the end of this report.

Clinical group input was provided by 5 clinicians with experience treating patients with FOP who attended the Canadian Endocrine Update, an annual international meeting on advances in endocrinology and metabolism. Some of the clinicians providing input have participated in clinical trials for the drug under review, with 1 clinician reporting on a patient with experience using palovarotene. The main unmet need of patients with FOP identified by the clinician group was for a treatment that could alter the natural course of the disease. The clinician group anticipates that palovarotene would be used as a single drug (with or without corticosteroids) administered daily, with the potential for short-term dose increases during flare-ups. Further, the clinical group noted that palovarotene has the potential to be used in combination with other investigational drugs with different mechanisms of action in the future. The clinician group suggests that all patients who meet the approved Health Canada indication for palovarotene should be considered for treatment. Accordingly, consideration for treatment initiation as listed by the clinical group suggests lifetime patient monitoring by a multidisciplinary care team that includes specialists in pediatric and adult orthopedics as well as surgeons and rheumatologists. The clinician group suggests that treatment with palovarotene should be discontinued in the event of notable adverse effects (e.g., premature epiphyseal fusion in children) or in the event of, or intention for, pregnancy in individuals of child-bearing capabilities. According to input by the clinical group, there are currently no tools for measuring outcomes in patients with FOP. The clinical group indicated that the following outcomes should be used to determine response to treatment: annualized change in HO volume, maintenance of mobility and reduction in rate of flare-ups, disease stability, and maintenance of HRQoL. The clinical group stressed that the decision to initiate or discontinue treatment with palovarotene should be made with careful patient counselling and shared decision-making.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may affect their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 3.

Table 3: Summary of Drug Plan Input and Clinical Expert Response

ADL = activity of daily living; AE = adverse event; CAJIS = Cumulative Analogue Joint Involvement Scale; CDEC = CADTH Canadian Drug Expert Committee; FOP = fibrodysplasia ossificans progressive; HO = heterotopic ossification.

Clinical Evidence

The clinical evidence included in the review of palovarotene is presented in 2 sections. The first section, the Systematic Review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada as well as those studies that were selected according to an a priori protocol. The second includes sponsor-submitted, long-term extension studies and additional relevant studies that were considered to address important gaps in the evidence included in the systematic review. The clinical evidence does not include indirect evidence.

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of palovarotene 1 mg, 1.5 mg, 2.5 mg, 5 mg, and 10 mg oral capsules for reducing the formation of HO in adults and children aged 8 years and above for females and 10 years and above for males with FOP.

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 4. Outcomes included in the CADTH review protocol reflect those considered to be important to patients, clinicians, and drug plans.

Table 4: Inclusion Criteria for the Systematic Review

AE = adverse event; FOP = fibrodysplasia ossifications progressiva; HO = heterotopic ossification; HRQoL = quality of life; RCT = randomized controlled trial; ROM = range of motion; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

^aMay include a background of best supportive care.

^bMay include hearing aids, personal care aids or tools, and bathroom aids.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.³⁸

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) through Ovid and Embase (1974–) through 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 Sohonos and palovarotene. 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 the 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 November 16, 2022. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on February 22, 2023.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the CADTH checklist, Grey Matters: A Practical Tool For Searching Health-Related Grey Literature.³⁹ The websites of regulatory agencies (FDA and the European Medicines Agency) were included in the search. 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 with 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.

Findings From the Literature

A total of 2 studies were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 5. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 5: Details of the MOVE Study and the NHS

AE = adverse event; AST = aspartate aminotransferase; ALT = alanine aminotransferase; CAJIS = Cumulative Analogue Joint Involvement Scale; C-SSRS = Columbia Suicide Severity Rating Scale; ECG = electrocardiogram; FOP = fibrodysplasia ossificans progressiva; FOP-PFQ = fibrodysplasia ossificans progressiva Physician Function Questionnaire; FPE = first patient enrolled; GHS = Global Health Score; GMS = Global Mental Score; HO = heterotopic ossification; LPC = last patient completed; NA = not applicable; NR = not reported; PCS = potentially clinically significant; PROMIS = Patient-Reported Outcome Measurement Information System; q.d. = once daily; ROM = range of motion; SAE = serious adverse event; ULN = upper limit of normal.

Note: One additional report was included.⁴³

^aThe final analysis was originally planned to occur after 24 months of follow-up; however, due to an interruption in the administration of the study drug, interim analysis 3 served as the final analysis.

Source: Clinical Study Reports for the MOVE study¹⁹ and the NHS.⁴⁰

Description of Studies

One sponsor-conducted study that met the CADTH review protocol criteria was included in the systematic review: Study PVO-1A-301 (MOVE).^19,41,42 Although the sponsor-conducted NHS of FOP (Study PVO-1A-001) did not meet the CADTH review protocol criteria,^20,40 the study served as the control group for the MOVE trial. Therefore, it is summarized.

The MOVE trial is an ongoing, multicentre, nonrandomized, open-label, phase III study evaluating the efficacy of palovarotene in decreasing new HO in adult and pediatric patients with FOP. The MOVE trial was conducted in 2 parts. In part A, eligible patients received chronic dosing with palovarotene for up to 24 months and underwent flare-up treatment if they experienced a flare-up. In part B, to obtain longer-term safety data, all patients were provided the chronic and/or flare-up palovarotene dosing regimens for an additional 24 months, until palovarotene was commercially available. No new patients were enrolled in part B of the MOVE trial.

A total of 107 patients were enrolled across 16 sites in 11 countries (Argentina, Australia, Brazil, Canada, France, Italy, Japan, Spain, Sweden, the UK, and the US). Of these, 5 patients (4.7%) were from Canada. All patients underwent low-dose WBCT, excluding the head, to assess HO at screening and during onsite study visits at months 6, 12, 18, and 24. Remote visits occurred at week 6 and at months 3, 9, 15, and 21. Patients who continued into part B of the MOVE study followed the same procedures and assessments as described for part A, except that WBCT imaging was performed annually at months 36 and 48. In the event of early termination or withdrawal of a patient, efforts were made by the study staff to have the patient complete end-of-treatment and end-of-study assessments onsite.

The primary efficacy end point for the MOVE trial was annualized change in new HO volume. The key secondary outcome was the proportion of patients with new HO. Secondary outcomes included the number of body regions with new HO; the proportion of patients reporting flare-ups; and the flare-up rate per patient-month exposure. Exploratory outcomes included change in ROM as measured by the CAJIS for FOP; change in physical function as measured by the FOP-PFQ; change in physical and mental function for patients aged greater than 15 years and change in mental function for patients aged younger than 15 years, both measured using the PROMIS; and the incidence and volume of catastrophic HO.

After the discovery of a high rate of premature epiphyseal fusion in growing children enrolled in the MOVE trial, a partial clinical hold on patients under the age of 14 was instituted. This was later amended to a permanent hold on girls aged younger than 8 years and boys aged younger than 10 years because of an unfavourable benefit-risk assessment due to the high rate of epiphyseal fusion. The Health Canada approval of palovarotene was based on the revised age thresholds. Consequently, this review focuses on the clinical evidence pertaining to patients aged 8 years and older for females and 10 years and older for males.

The NHS was a multicentre, natural history, noninterventional, longitudinal, 1-part study of patients with FOP caused by the R206H mutation of the ACVR1 gene. A total of 114 patients aged 4 years to 56 years were enrolled across 7 sites in 6 countries. No patients from Canada were included in the NHS. Part A of the NHS aimed to determine the optimal imaging method for assessing total body HO. In part B, 114 patients aged 65 years of age or younger were enrolled and completed baseline imaging using low-dose WBCT scans (excluding the head). Patients were followed for 36 months and assessed on a variety of end points related to disease characteristics, flare-ups, disease progression, and safety. At baseline, patients underwent a thorough examination to establish the current disease status in each patient. This was repeated annually at months 12, 24, and 36. Patients who were unable to undergo a procedure because of a safety concern (e.g., risk of flare-up) or physical limitation (e.g., pain or locked position) did not undergo that procedure. Patients were contacted by telephone to record any occurrences of new flare-ups since last contact, the use of concomitant medications, and experience of any AEs at months 8, 18, and 30. Patient-reported outcomes, including age-appropriate FOP-FPQ, age-appropriate PROMIS Global Health scale, and the FOP Assisted Devices and Adaptations Questionnaire, were also assessed at months 6, 18, and 30 by telephone contact. Patients were directed to call the study site immediately if they believed they were experiencing a flare-up between study visits. On confirmation of flare-up by the investigator, weekly telephone contact was made until the flare-up was resolved. If the flare-up could not be confirmed by the investigator, patients were contacted by telephone weekly until the symptoms of the suspected flare-up were resolved. Protocol amendment 5 pertaining to the NHS resulted in increasing telephone contact to every 3 months from every 6 months, except when annual clinical visits were scheduled (i.e., months 3, 6, 9, 15, 18, 21, 27, 30, and 33).

Protocol Amendment

The MOVE study protocol underwent 4 amendments before the database cut-off of February 28, 2020. Notable amendments included the following:

• Patients were included if they had FOP due to mutations other than R206H or were in previous phase II trials and could not receive the combination chronic and flare-up regimen due to country of residence or long travel distance (amendment 1).

• The primary efficacy analysis was restricted to patients with R206H mutation who had not previously been treated with palovarotene (amendment 1).

• Analysis methods for the primary and secondary efficacy end points were revised from the frequentist wLME model without square-root transformation to a Bayesian compound Poisson model with square-root transformation. Accordingly, the sample size determination was changed to coincide with the revised analysis methods (amendment 1).

• The timing of the second and third interim analyses was revised to be performed after all patients in the principal enrolled population had completed 12 months and then 18 months of follow-up, respectively (amendment 2).

• A 24-month extension period, designated part B, was added following the main study, designated part A (amendment 3).

• The MOVE study was to continue despite crossing the futility boundary (amendment 4).

A fifth protocol amendment was made after the database cut-off date. These amendments were related to the implementation of safety measures based on the Data Monitoring Committee recommendation after the risk of epiphyseal fusion was identified.

Populations

Inclusion and Exclusion Criteria

Key inclusion and exclusion criteria for the MOVE trial are summarized in Table 5. Briefly, patients eligible for inclusion in the MOVE trial were patients aged at least 4 years who had previously participated in the NHS,⁴⁰ had been clinically diagnosed with FOP with R206H ACVR1 mutation or other FOP variants associated with progressive HO, or had participated in Study PVO-1A-202²⁴ or Study PVO-1A-204 and could not receive the combination chronic and flare-up palovarotene regimen due to country of residence or long travel distance to participate in the phase II trials. In addition, patients were required to have a CAJIS score of 6 points to 16 points and must not have experienced flare-up symptoms within the past 4 weeks at the time of enrolment. Patients were required to be accessible for treatment and follow-up assessment and able to undergo all study procedures. Patients who were receiving flare-up treatment in Study PVO-1A-202 or Study PVO-1A-204 were unable to enrol in the MOVE trial until the flare-up treatment was completed and at least 4 weeks had passed since the last flare-up symptom. Patients were excluded from the trial if they weighed less than 10 kg; were pregnant or breastfeeding; had an intercurrent, nonhealed, fractured, or immobilized flare-up; and had recent retinoid or tetracycline therapy, vitamin A, beta carotene, or herbal products with fish oil, or cytochrome (CYP) 450 3A4 inhibitors or inducers.

Baseline Characteristics

The demographic and baseline disease characteristics of female patients aged at least 8 years and male patients aged at least 10 years in the MOVE trial and the NHS (principal safety set) are summarized in Table 6.

Approximately half of the patients were male (MOVE = 54.4%; NHS = 51.1%) and most were white (MOVE = 74.7%; NHS = 75.0%). Patients enrolled in the MOVE trial were, on average, younger than those enrolled in the NHS (14.4 years versus 20.4 years, respectively). A greater proportion of patients were aged 8 years to 14 years for females and 10 years to 14 years for males in the MOVE study than in the NHS (43.0% versus 23.9%). Moreover, patients in the MOVE study were, on average, younger than those in the NHS at the time of FOP diagnosis (6.5 years versus 7.5 years). Overall, clinical features and osteochondromas were comparable between the patients in the MOVE trial and the NHS. Almost all patients had been born with malformation of the great toes (99.4%), and approximately half had thumb malformation (51.5%). Osteochondromas of the tibia presented in 6.6% of all patients. Of note, a greater proportion of patients in the NHS were initially misdiagnosed compared to those in the MOVE trial (62.5% versus 44.3%). A history of flare-up was reported by 100% and 97.7% of patients in the MOVE trial and the NHS, respectively. The mean time since the last reported flare-up was longer among patients in the MOVE trial than among patients in the NHS (27.0 months versus 21.3 months). The most common location of the last flare-up was the hip (MOVE = 16.5%; NHS = 14.8%). The most common symptoms of the last flare-up were pain (MOVE = 73.4%; NHS = 85.2%) and swelling (MOVE = 70.9; NHS = 77.3%).

Notable imbalances in baseline characteristics between patients in the MOVE trial and the NHS included reported hearing loss (MOVE = 45.6%; NHS = 35.2%); pain symptoms (MOVE = 73.4%; NHS = 85.2%); lethargy (MOVE = 7.6%; NHS = 26.1%); change in mood and behaviour (MOVE = 11.4%; NHS = 39.8%) during last flare-up; unknown (MOVE = 73.4%; NHS = 44.3%) or other (MOVE = 8.9%; NHS = 23.9%) reported cause of last flare-up; bone formation as a result of last flare-up (MOVE = 51.9%; NHS = 0%); and slightly worse loss of movement as a result of last flare-up (MOVE = 2.5%; NHS = 20.5%).

Table 6: Demographic and Baseline Characteristics in the MOVE Trial and NHS in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Principal Safety Set)

FOP = fibrodysplasia ossificans progressiva; LOM = loss of movement; NHS = Natural History Study; ROM = range of motion; SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged < 14 years and January 24, 2020, for patients aged ≥ 14 years.

^aMultiple symptoms could be reported by patients.

Source: Clinical Study Report for the MOVE study.¹⁹

Interventions

Investigational Product

Chronic Treatment

All patients in the MOVE study self-administered palovarotene 5 mg oral capsules once daily for chronic dosing for up to 24 months. Patients aged younger than 18 years with less than 90% skeletal maturity on hand-wrist radiography at screening received a weight-adjusted equivalent chronic regimen.

Patients were directed to take palovarotene orally following a full meal at approximately the same time each day. Patients were instructed to avoid foods that are known to induce or inhibit activity of the CYP 3A4 enzyme (e.g., grapefruit, pomelo, or juices containing these fruits). For patients who have difficulty swallowing intact capsules or tablets due to ankylosis of the jaw, patients or their caregivers were permitted to sprinkle the contents of the capsule onto specific foods, as specified in the dosing instructions. Patients and caregivers were also directed to wear protective gloves when handling palovarotene capsule due to the potential for dermal absorption.

Flare-Up Treatment

For patients in the MOVE study, flare-based dosing was considered at the time of flare-up symptoms being reported by patients and/or legal guardians or caregivers to site personnel. Flare-up symptoms included, but were not limited to, pain, swelling, redness, decreased ROM, stiffness, and warmth. Only 1 symptom was required to define a flare-up. If the reported symptoms were consistent with previous flare-ups, included a patient-reported onset date, and were confirmed by the investigator as being associated with a flare-up, patients immediately began flare-up dosing. Flare-up dosing could also be initiated based on an investigator-confirmed, high-risk traumatic event likely to lead to a flare-up.

The flare-up dosing regimen, taken upon flare-up confirmation by the investigator, was palovarotene 20 mg once daily (or weight-based equivalent) for 4 weeks followed by palovarotene 10 mg once daily (or weight-based equivalent) for 8 weeks, for a total flare-up treatment duration of 12 weeks.

The flare-up dosing regimen could be extended in 4-week intervals while on treatment with palovarotene 10 mg and could continue until the flare-up resolved and the 4-week extension treatment was completed at the investigator’s discretion based on clinical signs and symptoms. In the event of an intercurrent flare-up (i.e., a new flare-up at a new location or marked worsening of an original flare-up) — or a confirmed high-risk traumatic event likely to lead to a new flare-up at any time during the flare-up treatment — the 12-week dosing regimen restarted upon the confirmation of a new intercurrent flare-up by the investigator.

Accordingly, a flare-up cycle included the first flare-up or traumatic event and any subsequent intercurrent flare-ups or traumatic events during the same dosing period. Once all flare-ups or traumatic events in a cycle had been resolved and flare-up-based treatment with palovarotene completed, patients resumed chronic treatment with palovarotene 5 mg (or weight-adjusted equivalent) once daily.

Weight-Based Dosing

All dosing was weight-adjusted in patients aged younger than 18 years with less than 90% skeletal maturity on hand-wrist radiography at screening. Additional radiographic assessments were performed at months 3, 9, 15, or 21 in patients who had received a flare-up dosing regimen since their last radiographic assessment. Weight-based dosing ceased when patients achieved at least 90% skeletal maturity on hand-wrist radiography. Weight-based dosing was incorporated for 3 weight categories: 20 kg to less than 40 kg; 40 kg to less than 60 kg; and greater than or equal to 60 kg. Weight-adjusted palovarotene doses are summarized in Table 7.

Dose Modification

The palovarotene dose could be reduced to the next lower dose, as shown in Table 7, if a patient experienced intolerable side effects. If the patient was already receiving the lowest possible dose, then treatment with palovarotene was discontinued. Among patients who required dose de-escalation as a result of an intolerable side effect for a flare-up, the flare-up dosing for subsequent flare-ups was determined by the investigator and medical monitor. In the event of partial or complete premature growth plate closure, with or without growth deceleration, treatment with palovarotene could be continued, interrupted, de-escalated, or discontinued based on clinical and radiographic information and following discussion of the risks and benefits with the patient and/or guardian(s).

Table 7: Weight-Adjusted Palovarotene Doses for Skeletally Immature Patients and Dose De-Escalation for All Patients in the MOVE Trial

Note: The 15 mg, 7.5 mg, and 2.5 mg equivalents are provided in the event of dose de-escalation from 20 mg, 10 mg, or 5 mg equivalents, respectively.

Source: Clinical Study Report for MOVE.¹⁹

Treatment Interruption

In the event of a medical condition requiring treatment with a prohibited concomitant medication (i.e., tetracycline and/or doxycycline; CYP450 3A4 inhibitor), treatment with palovarotene was discontinued for the duration of treatment. An appropriate 3-day washout period was considered before restarting treatment with palovarotene after the completion of tetracycline treatment. An appropriate washout period of 5 half-lives was considered before restarting treatment with palovarotene after the completion of treatments considered to be strong inhibitors of CYP 450 3A4.

Compliance With Treatment

Compliance with treatment was calculated as the duration of treatment over the expected duration of treatment.

The expected duration of treatment for chronic treatment was calculated as: (last chronic dose date minus first chronic dose date) minus the sum of all flare-up cycle dosing periods plus 1.

The expected duration of treatment for flare-up treatment was calculated as total expected flare-up dosing periods across all flare-up cycles. Each expected flare-up cycle dosing period was defined as the last flare-up cycle dose date minus the first flare-up cycle dose date plus 1.

Compliance was categorized as either less than 80% or 80% and greater.

Concomitant Medications

All patients received treatment per the standard of care set out in the 2019 FOP treatment guidelines, which could include corticosteroids (e.g., prednisone at 2 mg/kg orally to a maximum dose of 100 mg daily) for 4 days for acute flare-ups.⁹

The following medications were not permitted during chronic or flare-up treatment:

• vitamin A or beta carotene, multivitamins containing vitamin A or beta carotene, herbal preparations, or fish oil from the day before the start of treatment until the day of treatment

• synthetic oral retinoids other than palovarotene in the 30 days before treatment until the last day of treatment with palovarotene

• tetracycline or tetracycline derivatives

• medications identified as strong inhibitors or inducers of cytochrome CYP 450 3A4

• kinase inhibitors, such as imatinib and other drugs used off-label as potential treatments for FOP (i.e., rapamycin; a washout period of 5 half-lives was required before enrolment in the MOVE trial).

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 further summarized here. A detailed discussion and critical appraisal of the outcome measures are provided in Appendix 4.

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

HO = heterotopic ossification; HRQoL = health-related quality of life; ROM = range of motion.

Heterotopic Ossification

Annualized change in HO volume as measured by low-dose WBCT was the primary efficacy end point in the MOVE trial. Based on part A of the NHS, measurement of HO volume through low-dose WBCT (excluding the head) was determined by an imaging committee to be optimal over dual-energy X-ray absorptiometry in its ability to assess HO balanced against the estimated radiation exposure.⁴⁴ Compared to dual-energy X-ray absorptiometry, low-dose WBCT scans detected HO in more body regions; in addition, the inability to evaluate HO presence due to overlapping body regions (positional ambiguity) occurred less frequently during WBCT scans (mean number of nonevaluable regions per scan = 1.2 [SD = 1.5] versus 2.4 [SD = 1.4]).⁴⁴

The annualized change in new HO volume was compared between treated and untreated patients in the MOVE trial and the NHS, respectively. It was calculated by summing the increase in HO volume across all body regions for which new HO had occurred, where the increase in HO volume per region was defined as the square-root of the volumetric increase in that region. Scans were interpreted at an independent, central imaging laboratory. All 5 reviewers were blinded to whether the scan originated from the MOVE study or the NHS. The demarcation of the 9 anatomic regions was standardized to guide readers (Figure 2). At each scan, the presence of HO was first determined qualitatively in each of the 9 anatomic regions. If the presence of HO was detected qualitatively, then the total HO volume within that region was measured quantitatively and documented. The standardization measures included the use of anatomic diagrams; functionalities of the Mendelian Inheritance in Man workflow system; reader training and testing before initiation of reading of the study scans; quality control checks throughout the study; and an adjudication process whereby the same readers were assigned to the role of primary or adjudication reader for a given patient over time.

Figure 2: Diagram Depicting the 9 Anatomic Regions of CT Scans to Detect HO

Note: 1 = right shoulder, chest, upper back, neck (coverage includes right chest, neck [base of skull], and shoulder through the midhumerus); 2 = left shoulder, chest, upper back, neck (coverage includes left chest, neck [base of skull], and shoulder through midhumerus); 3 = midtorso (coverage includes torso from the 10th thoracic vertebra to above the iliac crests); 4 = right arm (coverage includes midhumerus through hand); 5 = left arm (coverage includes midhumerus through hand); 6 = right hip (coverage includes right iliac crest, hip through midfemur); 7 = left hip (coverage includes left iliac crest, hip through midfemur); 8 = right lower leg (coverage includes right midfemur through foot); 9 = left lower leg (coverage includes left midfemur through foot).

Source: Integrated Summary of Efficacy.⁴⁵

Additional outcomes related to HO that were assessed included:

• the proportion of patients with any new HO at month 12 as a key secondary efficacy end point

• the number of body regions with new HO at month 12 as a secondary efficacy end point

• a post hoc analysis of the incidence and volume of catastrophic HO per year (i.e., new HO volume greater than 50,000 mm³ and new HO volume greater than 30,000 mm³) as an exploratory efficacy outcome at month 12 and month 24.

Flare-Ups

A flare-up cycle included the first flare-up or traumatic event and any subsequent, intercurrent flare-ups or traumatic events during the same dosing period. The MOVE study defined a flare-up as an event with 1 or more flare-up signs, while the NHS defined a flare-up as the occurrence of 2 or more flare-up signs. The analysis of the proportion of patients reporting flare-ups in the MOVE trial was restricted to match the NHS definition. Accordingly, a flare-up was defined as follows:

• the presence of 2 of the 6 most common flare-up symptoms: pain, soft tissue swelling, redness, warmth, stiffness, and decreased ROM

• consistency with the patient’s previous flare-ups

• confirmation by the investigator.

Range of Motion

ROM was assessed for 12 joints (shoulder, elbow, wrist, hip, knee, and ankle on both the right and left sides) and 3 body regions (jaw, cervical spine [neck], and thoracic and lumbar spine) using the CAJIS.

The CAJIS is a clinician-administered analogue scale of gross mobility restriction at 15 anatomic locations.²¹ Locations of assessments are organized into 3 axial sites (neck, jaw, and thoraco-lumbar spine), 6 upper limb sites (right and left shoulders, elbows, wrists), and 6 lower limb sites (right and left hips, knees, and ankles).²¹ Each joint or region is scored on a scale of 0 points (uninvolved) to 2 points (ankylosed or completely involved). Total scores are calculated as the sum of the scores of all joints or regions, and range from 0 points (no involvement) to 30 points (maximally involved). Higher scores are indicative of poorer ROM. CAJIS subscores for upper extremities and mobility (lower extremities) may also be calculated. The CAJIS upper extremities subscore is the sum from 6 upper-body joints (shoulder, elbow, and wrists on both the right and left sides) and 1 region (cervical spine [neck]). These subscores range from 0 points (no involvement) to 12 points (maximally involved). The CAJIS mobility subscore is the sum from 6 lower-body joints (hip, knee, and ankle on both the right and left sides) and range from 0 (no involvement) to 12 (maximally involved). The CAJIS also records the presence (yes) or absence (no) of disability for each of the following items: walks, uses a wheelchair, needs some help with activities of daily living, needs complete assistance with activities of daily living.²¹ An estimated MID for the CAJIS was not identified by the CADTH review team for patients with FOP.

Physical Function

Physical function was assessed using age-appropriate forms of the FOP-PFQ.

The FOP-PFQ is a disease-specific, patient-reported outcome measure that assesses physical function.²² The adult version of the FOP-PFQ consists of 28 questions, while the pediatric version consists of 26 questions. Questions are scored on a scale from 1 (not able to do) to 5 (able to do without trouble, help, or an assistive device), with lower scores indicative of more difficulty and greater functional impairment. Total scores are the sum of each question. Scores range from 28 to 140 for the adult version and from 26 to 130 for the pediatric version. Scores may also be calculated for the upper extremities (score range = 15 to 140 for adults and 18 to 90 for pediatrics) and mobility (score range = 13 to 65 for adults and 8 to 40 for pediatric patients). Where a patient does not complete some (but not more than 20%) of the form questions, the total scores are calculated as the average of the observed scores multiplied by the number of expected question scores. An estimated MID for the FOP-PFQ was not identified by the CADTH review team for patients with FOP.

In the MOVE trial, patients aged 15 years or older were administered the adult form of the FOP-PFQ. For patients aged 8 years to 14 years, both the pediatric self-completed and the pediatric proxy-completed forms were administered. The proxy-completed forms were used for analyses unless only the self-completed forms were available. Analysis of the FOP-PFQ was performed for all patients (i.e., adults and children). Because the number of contributing questions differs between the adult and pediatric versions, the scores were transformed to reflect a percentage of the worst score. The percentage of worse scores ranged from 0% (best possible function) to 100% (worst possible function).

Health-Related Quality of Life

HRQoL was assessed using age-appropriate forms of the PROMIS Global Health scale short form.

The PROMIS Global Health scale is a set of person-centred measures that evaluates and monitors physical, mental, and social health in adults and children in the general population or who are living with chronic conditions.²³ The PROMIS Global Health scale short form (for patients aged 15 years and older) consists of 10 items on varying scales.⁴⁶ Both the Global Physical Health and Global Mental scores may be calculated for adults. The Global Physical Health score is the sum 4 questions and may range from 4 (worse health) to 20 (better health). Likewise, the Global Mental Health score is calculated as the sum of 4 questions and ranges from 4 (worse health) to 20 (better health). If a patient is missing any of the contributing scores, the Global Physical Health or Global Mental Health scores are not calculated. Meanwhile, the PROMIS Pediatric Global Health Scale (for patients aged 17 years and younger) consists of 9 questions on varying scales.⁴⁷ Only the total scores are calculated for the pediatric form. The pediatric total score is the sum of the first 7 questions and ranges from 7 (worse health) to 35 (better health). In the event that a patient is missing some (but not more than 3) of the contributing scores, the total score is calculated as the average of the observed score multiplied by the number of expected scores. An estimated MID for the PROMIS was not identified by the CADTH review team for patients with FOP.

In the MOVE trial, patients aged 15 years or older were administered the adult form of the PROMIS Global Scale. For patients aged 8 years to 14 years, both the pediatric self-completed and the pediatric proxy-completed forms were administered. The proxy-completed forms were used for analyses unless only the self-completed form was available. The Global Physical Health and Global Mental Health scores and total scores were converted to T-scores for analysis. A T-score of 50 is normal, with an SD of 10. A T-score of less than 50 is indicative of worse health, while a T-score of greater than 50 is indicative of better health.

Safety

Safety end points included AEs and SAEs. Treatment-emergent adverse events (TEAEs) were defined as AEs occurring on or after the first day of palovarotene treatment and on or before 7 days following the last dose of palovarotene. Safety end points were reported separately for chronic and flare-up treatment cycles. AEs were assessed at each onsite and remote visit during chronic and flare-up-based treatment.

A chronic TEAE was a TEAE with an onset date during the chronic treatment; a flare-up TEAE was a TEAE with an onset date during flare-up treatment.

For the efficacy interim and final analyses, TEAEs were obtained from patients’ dosing diaries. For all other predatabase lock analyses, TEAEs were ascertained using patients’ planned dosing logs, if insufficient diary data were available using the date of data extraction as the end date for TEAE assignment purposes. Any AE with a missing start date was considered to be a TEAE. For TEAEs with a missing date, if the month was the same month as when the flare-up treatment was provided, then the TEAE was considered to be associated with flare-up dosing.

Of note, patients could have multiple independent flare-up treatment cycles, each associated with AEs.

Statistical Analysis

Patients treated with palovarotene in the MOVE trial were compared to untreated patients in the NHS to assess the efficacy of palovarotene. Where applicable, data from untreated patients who were in the NHS for more than 36 months were used for comparative purposes.

Sample Size Determination

Sample size determination was based on simulations of available HO volumes as measured via WBCT scans from the NHS and the observed efficacy of palovarotene treatment in phase II studies. For the purpose of sample size determination, it was assumed that 45 patients from the NHS would have heterotopic ossification volume as measured by WBCT scans documented at baseline, year 1, and year 1.5, and that another 45 patients would have HO volume as measured by WBCT scans documented at baseline, year 1, and year 2. It was also assumed that of these 90 patients, 50% would enter the MOVE trial to begin treatment with palovarotene.

The MOVE trial was originally set to enrol 80 patients. The primary analysis comparing the annualized change in new HO volume between patients in the MOVE trial treated with palovarotene and untreated patients from the NHS was determined to have an overall probability of 0.92 for study success if palovarotene treatment reduced the number of regions with new HO by 30% and reduced the volume of new HO, conditional on new HO per body region, by 50%. The probabilities of declaring study success were determined to be 0.51, 0.79, and 0.89 at the first, second, and third interim analyses, respectively.

The MOVE study was amended (through a protocol amendment on February 19, 2019) to increase the sample size to a maximum of 110 patients; up to 99 of these had the R206H mutation and no previous exposure to palovarotene, and up to 11 of these had other mutations and previous participation in phase II studies. It was determined that the increased sample size did not substantially change the power of the trial or have a substantial impact on the primary and secondary analyses.

Planned Analysis

Interim and Futility Analysis

A total of 3 interim efficacy analyses were planned for part A of the MOVE trial using group sequential methods. The first interim analysis was planned when 35 patients had completed 1 year of follow-up. The second and third interim analyses were conducted when all patients among the principal enrolled population had completed 12 months and 18 months of follow-up, respectively. The percentage of patient follow-up for each analysis is summarized in Table 9.

On December 4, 2019, a “partial clinical hold” was issued by the FDA for patients aged younger than 14 years following an identified risk of premature epiphyseal fusion in growing children in the FOP palovarotene program.

A futility analysis was conducted at the second interim analysis to determine whether the MOVE trial should be stopped due to insufficient evidence of efficacy. Assuming a 1-sided, overall type I error rate of 2.5%, the Lan-DeMets alpha-spending function with O’Brien-Fleming parameterization was used to determine stopping boundaries. Futility was initially declared by the Data Monitoring Committee on January 15, 2020, for the MOVE trial. This declaration required the sponsor to pause dosing for all patients in the FOP palovarotene program. The Data Monitoring Committee and sponsor became unblinded to all study data. After unblinding and a review of the post hoc efficacy analysis, the Data Monitoring Committee recommended that “palovarotene be continued in skeletally mature children ≥ 14 years,” due to concerns about premature epiphyseal fusion.

The third interim analysis on May 26, 2020, included a total follow-up period that was 24% longer for patients treated with palovarotene than at the second interim analysis.

The final analysis was originally planned to occur after 24 months of follow-up; however, due to the interruption in the administration of the study drug following the declaration of futility, interim analysis 3 served as the final analysis. Accordingly, the primary data analysis included assessments collected on or before December 4, 2019, for patients aged younger than 14 years and on or before January 24, 2020, for patients aged 14 years or older.

Control of Type I Error

Based on the percentages summarized in Table 9, the alpha level threshold used to determine the significance of treatment effect at each analysis was determined using the Lan-DeMets alpha-spending function with O’Brien-Fleming parameterization and assuming a 1-sided, overall type I error rate of 2.5%. The 1-sided significance thresholds were 0.0058, 0.0103, 0.0156, and 0.0190 for the first, second, third interim, and final analyses, respectively.

The 1-sided significance threshold for the Lan-DeMets alpha-spending function with O’Brien-Fleming parameterization was recalculated at the time of each interim analysis based on the estimate of the information fraction. However, because the futility boundary was crossed at the second interim analysis, all subsequent analyses of efficacy data were supplied without applying thresholds for success or futility, given that these thresholds were no longer relevant.

Post hoc analyses related to the target population were conducted without controlling for multiplicity.

Table 9: Observed Follow-Up at Each Planned Analysis for the MOVE Trial

^aApproximate information based on expected accrual rates and total planned accrual in the principal enrolled population.

Source: Statistical Analysis Plan for the MOVE Trial.⁴⁸

Statistical Methods

Annualized Change in New HO volume

The analysis for the annualized change in new HO volume used a Bayesian compound Poisson (hereafter referred to as Bayesian) model that described the change in HO volume as the number of body regions with new HO volume (volume of new HO > 0 mm³) and the new HO volume (mm³) per region.

The model incorporated a square-root transformation of new HO volume per region and required that new HO volume be nonnegative. Some patients were observed to have a reduction in HO volume (e.g., negative HO volume) relative to baseline. For the purpose of this analysis, HO volumes for these patients were imputed with 0. Of note, the Bayesian model incorporates random effects to account for correlation between repeated measures within patients. Any missing data for a particular patient were considered missing at random (MAR) and were not imputed. The MAR assumption is based on the random-effects structures built into the Bayesian model, which assume that the missing value for a given patient can be explained by other observed values from that patient. The prior distributions for the variables in the Bayesian model were included on all parameters. However, justifications for the priors were not provided by the sponsor. Finally, estimates for the posterior distribution were generated using Gibbs sampling. Details regarding the estimation procedures or assessment of model convergence were not available to the CADTH clinical review team.

Analysis of the number of body regions with new HO was adjusted for each patient’s sex and age (< 18 years or ≥ 18 years) at the time of the WBCT scan. Analysis was repeated without the square-root transformation in a post hoc analysis to better understand the potential treatment effect of palovarotene. Conclusions about efficacy in the overall population were based on the repeated analysis without the square-root transformation.

After the study was stopped due to futility at the second interim analysis, the method of analyzing the primary end point for patients aged at least 8 years in females and at least 10 years for males was changed to Bayesian analysis without square-root transformation or negative values zeroed out by region. The same statistical plan added analysis through a wLME model without square-root transformation and with negative values included and a Wilcoxon rank sum test. For the wLME model analyses, HO volumes from the observations associated with the longest follow-ups with weights were included to account for the different lengths of follow-up in the MOVE trial and the NHS. The wLME model incorporated random effects to account for repeated measures for those patients who were in both the MOVE trial and the NHS. Missing data were not imputed because the model assumed these were MAR. The model was adjusted for baseline HO volume divided by age at the time of scan. Comparisons using the Wilcoxon rank sum test were performed unadjusted. All analyses related to the target population were post hoc.

Proportion of Patients With Any New HO and Number of Body Regions With New HO

The proportion of patients with any new HO and number of body regions with any new HO at month 12 was compared between the MOVE trial and the National History Study using the Bayesian compound Poisson model, as described for the primary analysis.

Flare-Ups

Flare-ups were reported as the flare-up rate per patient-month exposure and the proportion of patients reporting flare-ups:

• The flare-up rate per patient-month exposure was compared using a negative binomial regression on the number of flare-ups. An offset variable equal to the log of the number of years of follow-up was included to convert the estimate to an annualized rate.

• The proportion of untreated and treated patients reporting flare-ups was compared using the exact test of the difference in proportions.

Flare-up data were summarized descriptively as the number of flare-ups, the rate of flare-ups, and the number and percentage of patients with any flare-ups.

Exploratory End Points

Exploratory end points were summarized descriptively for change from baseline in CAJIS total score, CAJIS upper extremities subscore, and CAJIS mobility subscore; FOP-PFQ total score; PROMIS Global Physical Health and Global Mental Health T-scores; and catastrophic HO.

Subgroup Analysis

The statistical analyses used in the subgroup analyses included a Bayesian compound Poisson analysis (without square-root transformation and with negative new HO values zeroed out by body region) and a wLME model (without square-root transformation and with negative included) for annualized new HO volume by age, and age by sex. Proportions were compared using Fisher’s exact test, and counts were compared using negative binominal regression.

Analysis Populations

The patient populations for the purpose of data analysis are defined in Table 11.

Table 10: Statistical Analysis of Efficacy End Points in the MOVE Trial

CAJIS = Cumulative Analogue Joint Involvement Scale; HO = heterotopic ossification; FOP-FFQ = fibrodysplasia ossificans progressive Physician Function Questionnaire; NA = not applicable; PROMIS = Patient-Reported Outcome Measurement Information System; ROM = range of motion; wLME = weighted linear mixed effect.

Table 11: Summary of Analysis Populations

EP = enrolled population; FAS = full analysis set; HO = heterotopic ossification; NHS = Natural History Study; PPS = per-protocol set; PS = pharmacokinetic set; SS = safety set; WBCT = whole-body CT.

^aThe supplementary EP, supplementary FAS, supplementary PPS, supplementary SS, and supplementary PS were defined analogously as the principal sets described in the table for patients with the R206H ACVR1 mutation.

Results

Patient Disposition

Data on patient disposition for the target population (i.e., patients aged 8 years and older for females and patients aged 10 years and older for males) were not available.

Data on patient disposition for the overall population enrolled in the MOVE trial and the NHS are summarized in Table 12. A total of 107 patients and 114 patients with FOP were enrolled in the MOVE trial and the NHS, respectively. Of those enrolled in the MOVE trial, 88 patients (82.2%) had ongoing participation in the study as of interim analysis 3. Overall, 17.8% of patients enrolled in the MOVE trial discontinued from the study. The most common reasons for discontinuation were self-withdrawal (10.3%) and AEs (5.6%). Among patients in the NHS, 31 patients (27.2%) completed the study, while 39 patients (36.4%) crossed over into the MOVE trial. Overall, 81 patients who were enrolled in the NHS (71.1%) discontinued from the study. The most common reason for study discontinuation was enrolment into an intervention study (45.6%).

Table 12: Patient Disposition

Edc = electronic data capture; EP = enrolled population; FAS = full analysis set; NA = not applicable; NHS = Natural History Study; PP = per protocol.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

^aOther reasons included not wanting to travel (n = 1); enrolled in an interventional study not listed in the Edc (n = 1); enrolled in another interventional study (n = 3); enrolled in another non-Clementia interventional study (n = 1); worsening of clinical condition and impossibility to move to the centre (n = 1).

Source: Clinical Study Report for the MOVE study.¹⁹

Protocol Deviations

Protocol deviations for the target population were not available.

Protocol deviations for the overall population are summarized in Table 13. A total of 25 patients (23.4%) in the MOVE trial experienced at least 1 significant protocol deviation, namely related to receiving the wrong treatment or incorrect dose (n = 21; 19.6%). No patient discontinued the MOVE trial due to a protocol deviation.

Table 13: Significant Protocol Deviations (Per-Protocol Set)

^aWrong treatment or incorrect dose.

Source: Clinical Study Report for the MOVE study.¹⁹

Exposure to Study Treatments

Exposure to Investigational Product

Exposure to palovarotene in the overall study population is presented in Table 34, Appendix 3.

Exposure to palovarotene in the target population is presented in Table 14. Total mean exposures to the chronic and flare-up dosages of palovarotene were 1,718.6 mg (SD = 857.3 mg) and 2,372.9 mg (SD = 1,695.8 mg), respectively. The mean overall exposure to palovarotene in the MOVE trial was 3,346.5 mg (SD = 1,656.1 mg), with a mean duration of exposure of 75.4 weeks (SD = 21.8 weeks). Interruption and discontinuation of chronic palovarotene therapy occurred in 7 patients (8.1%) and 6 patients (7.0%), respectively. In total, 59 patients (68.6%) received flare-up treatment with palovarotene for a mean duration of 27.9 weeks.

Table 14: Exposure to Palovarotene in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older in the MOVE Study (Principal Safety Set)

SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Compliance With Investigational Product

In the overall study population, a total of 88 patients (89%) were at least 80% compliant with chronic and flare-up treatment with palovarotene. Among patients in the target population, 95.3% and 88.1% of patients treated with palovarotene were at least 80% compliant with chronic treatment and flare-up treatment, respectively (Table 15).

Table 15: Compliance With Palovarotene Treatment in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older in the MOVE Study (Principal Safety Set)

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

^aFor chronic treatment, time in days from first non–flare-up dose to last non–flare-up dose.

Source: Clinical Study Report for the MOVE study.¹⁹

Concomitant Medications

Prior medications reported at baseline and new-onset medications started during the study period for the overall population are presented in Table 35 and Table 36, respectively (refer to Appendix 3).

The use of concomitant medications in the target population during the MOVE trial and the NHS is summarized in Table 16. In patients in the MOVE trial, the 5 most commonly used concomitant medications during the study period were emollients and protectives (89.5%), nonsteroidal anti-inflammatory and antirheumatic drugs (65.1%), corticosteroids (55.8%), antihistamines for systemic use (48.8%), and drugs for peptic ulcers and gastro-esophageal reflux disease (37.2%). The 5 most commonly used concomitant medications among patients in the NHS were corticosteroids (72.1%), nonsteroidal anti-inflammatory and antirheumatic drugs (51.4%), analgesics and antipyretics (29.7%), drugs for peptic and gastro-esophageal reflux disease (17.1%), and beta-lactam antibacterials or penicillin (17.1%).

Patients in the MOVE trial and the NHS differed in their use of the following concomitant medications: nonsteroidal anti-inflammatory and antirheumatic drugs (MOVE = 65.1%; NHS = 51.4%); corticosteroids (MOVE = 55.8%; NHS = 72.1%); analgesics and antipyretics (MOVE = 50.0%; NHS = 29.7%); drugs for peptic and gastro-esophageal reflux disease (MOVE = 37.2%; NHS = 17.1%); dermatological preparations (MOVE = 34.9%; NHS = 0.9%); beta-lactam antibacterials (MOVE = 32.6%; NHS = 17.1%); herbal and/or traditional medicines (MOVE = 31.4%; NHS = 0.9%); antibiotics for topical use (MOVE = 25.6%; NHS = 0.9%); and opioids (MOVE = 22.1%; NHS = 2.7%).

Table 16: Concomitant Medication Use in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older in the MOVE Trial and the NHS

GORD = gastro-esophageal reflux disease; NHS = Natural History Study; nos = not otherwise specified; NR = not reported.

Note: The list excluded glucocorticoids and corticosteroids for systemic plan use.

^aReported use in greater than or equal to 10% of patients overall.

Source: Clinical Study Report for the MOVE study.¹⁹

Efficacy

Only those efficacy outcomes identified in the review protocol are reported here. Refer to Appendix 3 for detailed efficacy data.

Heterotopic Ossification

Annualized New HO

The observed volumes of annualized new HO for individual patients treated with palovarotene and untreated patients in the target population are illustrated in Figure 3. On visual inspection of a swimmers’ plot illustrating the observed volume of annualized new HO for individual patients treated with palovarotene and untreated patients in the target population, there were larger amounts of annualized new HO volumes in the untreated patients in the NHS than in the treated patients in the MOVE trial. More patients with negative new HO volumes were observed in the patients treated with palovarotene in the MOVE trial compared to the untreated patients in the NHS. On visual inspection, the results in the target population were consistent with the results for the overall study population (Figure 5, Appendix 3).

The posterior distribution of the ratio gamma (annual new HO volume under treatment versus no treatment) for the Bayesian analysis for the target population in patients treated with palovarotene in the MOVE trial versus untreated patients in the NHS is illustrated in Figure 4 in this section. The Bayesian analysis with no square-root transformation and negatives set to 0 by body regions fit a 25% reduction (ratio of mean change = 0.75; 95% CrI, 0.51 to 1.11) in the volume of annualized new HO among patients treated with palovarotene in the MOVE trial compared to untreated patients in the NHS.

The post hoc analysis of annualized new HO with no square-root transformation and with negative values included for the target population is summarized in Table 17. The mean annualized new HO volumes in patients treated with palovarotene in the MOVE trial and untreated patients in the NHS were 11,419 mm³ (standard error = 3,782 mm³) and 25,796mm³ (standard error = 6,066 mm³), respectively. Compared to patients in the NHS, a 55.7% reduction in mean annualized new HO volume was observed among patients treated with palovarotene in the MOVE trial. Based on the wLME model, palovarotene-treated patients in the MOVE trial had an estimated reduction in new HO volume of 10,443 mm³ per year (95% CI, –23,538 to 26,534; P = 0.1124) compared to untreated patients in the NHS when controlling for baseline HO divided by age. Accordingly, the wLME model estimated a 49% reduction in mean annualized new HO volume in patients treated with palovarotene in the MOVE trial compared to untreated patients in the NHS. The Wilcoxon rank sum test reported P value was 0.0107.

Of note, 39 patients transitioned from the NHS to the MOVE trial. Among these patients, a 61% reduction in mean annualized new HO volume was observed after they transitioned to receive palovarotene. Based on the wLME model, these patients had an estimated reduction in mean HO volume of –10,015 mm³ per year (95% CI, –19,737 to –292; P = 0.0438) compared to their untreated period in the NHS. The Wilcoxon rank sum test reported P value was 0.005.

Figure 3: Post Hoc Individual Patient Annualized New HO volume Distribution in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Post Hoc; Principal Full Analysis Set)

CI = confidence interval; HO = heterotopic ossification; ID = identification; NHS = Natural History Study; PVO = palovarotene.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Figure 4: Post Hoc Posterior Distribution Gamma — Bayesian Compound Poisson Model in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Principal Full Analysis Set)

HO = heterotopic ossification; pr = probability; SD = standard deviation.

Notes: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older. The area to the left of the red line indicates the probability of a greater than 30% reduction in annualized new HO in treated patients in the MOVE study versus untreated patients in the NHS. The area to the left of the green line indicates the probability of any reduction in annualized new HO in patients in the MOVE study versus those in the NHS. Values are non–square-root transformed; negative values are set to 0 by body region.

Source: Clinical Study Report for the MOVE study.¹⁹

Table 17: Post Hoc Analysis of Annualized New HO in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Principal Full Analysis Set)

CI = confidence interval; HO = heterotopic ossification; LS = least squares; NHS = Natural History Study; SE = standard error; vs. = versus; wLME = weighted linear mixed effect.

Notes: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older. The annualized change in HO volume was calculated by summing the increase in HO volume across all body regions for which new HO had occurred, where the increase in HO volume per region is defined as the square-root of the volumetric increase in that region.

^awLME LS mean and SE are estimated from a mixed model for the end point of annualized new HO, adjusting for the fixed effects of treatment and baseline total HO divided by baseline age and a random patient effect.

^bThe P value is based on post hoc analysis and cannot be used to draw inferences for this end point.

^cUnadjusted comparison of annualized new HO volumes between the MOVE study and the NHS using the sum of ranks.

Source: Clinical Study Report for the MOVE study.¹⁹

Proportion of Patients With Any New HO

The proportion of patients in the target population with any new HO at month 12 is summarized in Table 18. The proportions of patients with any new HO among patients treated with palovarotene in the MOVE trial and untreated patients in the NHS at month 12 were 62.2% and 57.4%, respectively.

Body Regions With New HO

The proportion of patients in the target population with any new HO at month 12 by number of body regions is summarized in Table 19. The proportions of patients with 0 body regions with new HO at month 12 were 37.8% and 42.6% in the MOVE study and the NHS, respectively. The proportions of patients with 1 body region with new HO at month 12 were 31.1% and 22.1% in the MOVE study and the NHS, respectively. No clear, consistent trends were observed differentiating patients in the MOVE study from those in the NHS.

Table 18: Proportion of Patients With Any New HO at Months 12 in Females Aged 8 Years or Older and Males Aged 10 Years or Older (Post Hoc; Principal Full Analysis Set)

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged greater than or equal to 14 years.

HO = heterotopic ossification.

^aNew HO volume greater than 00 mm³.

^bThe P value is based on a post hoc analysis and cannot be used to draw inferences for this end point.

Source: Clinical Study Report for the MOVE study.¹⁹

Table 19: Number of Body Regions With New HO in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Post Hoc; Principal Full Analysis Set)

HO = heterotopic ossification; NHS = natural history study.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

^aNew HO volume greater than 0 mm³.

^bThe P value is based on a post hoc analysis and cannot be used to draw inferences for this end point.

Source: Clinical Study Report for the MOVE study.¹⁹

Catastrophic HO

Results for catastrophic HO volume for the target population among patients treated with palovarotene from the MOVE trial and untreated patients in the NHS are summarized in Table 20. The proportions of patients with catastrophic new HO volumes greater than 100,000 mm³, greater than 50,000 mm³, and greater than 30,000 mm³ at month 12 in the MOVE trial were 1.3%, 9.1%, and 11.7%, respectively. Among patients in the NHS, the proportions of patients with catastrophic new HO volumes greater than 100,000 mm³, greater than 50,000 mm³ or greater than 30,000 mm³ at month 12 were 3.8%, 11.4%, and 13.9%, respectively.

The proportions of patients with catastrophic annualized new HO volumes exceeding greater than 100,000 mm³, greater than 50,000 mm³, and greater than 30,000 mm³ at the last time point in the MOVE trial were 1.3%, 6.5%, and 16.9%, respectively. At the last time point in the NHS, the proportions of patients with catastrophic new HO volumes greater than 100,000 mm³, greater than 50,000 mm³, and greater than 30,000 mm³ at month 12 were 6.3%, 15.2%, and 24.1%, respectively. Generally, at each time point, the mean catastrophic HO volume was numerically lower in patients treated with palovarotene in the MOVE trial than in untreated patients in the NHS.

Table 20: Catastrophic HO in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Post Hoc; Principal Full Analysis Set)

HO = heterotopic ossification; NE = not estimable; SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Flare-Ups

Reported Flare-Ups

Among patients in the target population, 67.1% and 63.9% of patients in the MOVE study and the NHS, respectively, reported flare-ups. Based on a post hoc analysis of the principal safety set, the flare-up rates per patient-month were 0.11 (95% CI, 0.08 to 0.16) in the MOVE trial and 0.06 (95% CI, 0.05 to 0.08) in the NHS.

HO at Flare-Up Sites

The volumes of new HO at month 12 at and away from flare-up sites in the overall population are summarized in Table 21. Overall, the mean volumes of new HO at flare-up sites among patients in the MOVE trial and the NHS were 19,610 mm³ (95% CI, 11,135 mm³ to 28,084 mm³) and 40,157 mm³ (95% CI, 9,189mm³ to 71,124 mm³), respectively. The mean volumes of new HO away from flare-up sites (following a flare-up) were 7,626 mm³ (95% CI, 3,845 mm³ to 11,407 mm³) among patients in the MOVE trial and 26,398 mm³ (95% CI, 8,539 mm³ to 44,259 mm³) in the NHS.

Range of Motion

Cumulative Analogue Joint Involvement Scale

Table 22 summarizes the CAJIS scores for patients in the target population. At baseline, mean CAJIS scores were similar between patients treated with palovarotene in the MOVE study (10.8; SD = 6.4) and untreated patients in the NHS (12.6; SD = 7.0). At month 12, patients treated with palovarotene in the MOVE study and untreated patients in the NHS experienced increases (deterioration) in mean CAJIS scores from baseline of 0.6 (SD = 2.1) and 0.6 (SD = 2.4), respectively.

Physical Function

FOP Physical Function Questionnaire

Table 23 summarizes the FOP-PFQ scores for patients in the target population. At baseline, the mean percentages of worse scores on the FOP-PFQ were similar between patients treated with palovarotene in the MOVE study (mean = 46.1; SD = 27.6) and untreated patients in the NHS (mean = 47.6; SD = 28.0). At month 12, both patients treated with palovarotene in the MOVE study and untreated patients in the NHS experienced an increase (deterioration) in the mean percentage of worse score on the FOP-PFQ from baselines of 2.94 (SD = 8.09) and 4.70 (SD = 9.02), respectively.

Table 21: Volume of New HO at and Away From Flare-Up Sites for All Patients (Principal Full Analysis Set)

CI = confidence interval; HO = heterotopic ossification; NHS = Natural History Study.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Table 22: Summary of CAJIS Total Scores in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Principal Safety Set)

CAJIS = Cumulative Analogue Joint Involvement Scale; NHS = Natural History Study; SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Table 23: Summary of FOP-PFQ Percentages of Worse Scores in Female Patients Aged 8 Years or Older and Male Patients Aged 10 Years or Older (Principal Safety Set)

FOP-PFQ = fibrodysplasia ossificans progressiva Physical Function Questionnaire; SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Health-Related Quality of Life

The PROMIS Global Health T-scores for patients aged at least 15 years of age and patients aged younger than 15 years are summarized in Table 24 and Table 25, respectively. Small changes from baseline were observed in patients treated with palovarotene in the MOVE trial and among untreated patients in the NHS at all postbaseline time points.

In patients aged 15 years and older, mean baseline scores were similar between those in the MOVE study and the NHS on the PROMIS Global Physical Health T-score (MOVE = 43.15 [SD = 7.93]; NHS = 43.35 [SD = 8.66]) and Global Mental Health T-score (MOVE = 52.17 [SD = 7.95]); NHS = 52.70 [SD = 9.40]). Mean changes from baseline on the Global Physical Health scale in patients in the MOVE trial at month 6, month 12, and month 18 were –0.15 (SD = 3.92), –0.20 (SD = 5.16), and –1.91 (SD = 6.28), respectively. Among patients in the NHS, the mean changes from baseline on the Global Physician Health scale at month 6, month 12, and month 18 were –0.37 (SD = 6.79), –1.19 (SD = 6.62), and –0.66 (SD = 5.57), respectively.

In patients under the age of 15 years, baseline PROMIS Global Health scores were similar among patients in the MOVE study and the NHS, at 43.2 (SD = 7.9) and 43.4 (SD = 8.7), respectively. Mean T-score changes from baseline among patients treated with palovarotene in the MOVE trial at month 6, month 12, and month 18 were –0.15 (SD = 3.92), 0.20 (SD = 5.16), and –1.91 (SD = 6.28), respectively. Among untreated patients in the NHS, mean T-score changes from baseline at month 6, month 12, and month 18 were –0.37 (SD = 6.79), –1.19 (SD = 6.62), and –0.66 (SD = 5.57), respectively.

Table 24: Summary of Adult PROMIS Global Health T-scores in Patients Aged At Least 15 Years (Principal Safety Set)

NHS = Natural History Study; PROMIS = Patient-Reported Outcomes Measurement Information System; SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Table 25: Summary of PROMIS Pediatric Global Health T-scores in Patients Aged Younger Than 15 Years of Age (Principal Safety Set)

NHS = Natural History Study; PROMIS = Patient-Reported Outcomes Measurement Information System; SD = standard deviation.

Note: The data cut-off date was February 28, 2020.

Source: Clinical Study Report for the MOVE study.¹⁹

Other Outcomes

The following outcomes were not assessed in the MOVE trial:

• pain and swelling associated with flare-ups

• respiratory function, including the need for ventilation

• use of mobility aids

• use of nonmobility aids, such as hearing aids, personal care aids or tool, or bathroom aids

• hearing loss

• soft tissue swelling and cartilage swelling at flare-up site

• survival.

Harms

Only those harms identified in the review protocol are reported here. A summary of harms data for the overall study population in the MOVE trial is shown in Table 26.

Adverse Events

At least 1 AE was reported by 96% and 94.3% of patients during the chronic dosing and flare-up dosing regimens in the MOVE trial, respectively. The most commonly reported AEs during chronic dosing were related to mucocutaneous issues (83.8%), including dry skin (52.5%) and rashes (19.2%); gastrointestinal issues (63.6%), including dry lips (34.3%); infections and infestations (58.6%), including upper respiratory infection (20.2%); and musculoskeletal and connective tissue disorders, including arthralgia (24.2%) and pain in an extremity (18.2%). Overall, the reporting of AEs was similar during the chronic dose and flare-up dosing regimens, with the exception of AEs related to gastrointestinal issues, which tended to be reported more commonly with chronic dosing than during flare-up dosing (63.6% versus 47.1%).

Serious Adverse Events

At least 1 SAE was reported by 19.2% and 17.1% of patients during the chronic dosing and flare-up dosing regimens, respectively. The most commonly reported SAE was epiphyses premature fusion, which was observed in 11.1% of patients during chronic dosing and in 10% of patients during flare-up dosing.

Dose Modification Due to Adverse Events

More patients required dose modification of palovarotene due to AE during flare-up treatment (40%) than during chronic treatment (11.1%). The most common reasons for dose modification were drug eruption (chronic dose = 3.0%; flare-up dose = 12.9%), generalized pruritis (flare-up dose = 8.6%), erythema (flare-up dose = 4.3%), and pruritis (flare-up dose = 4.3%).

Treatment Interruption Due to Adverse Events

Treatment interruptions due to AEs occurred among 16.2% and 15.7% of patients during the chronic and flare-up dosing regimens, respectively. The most common AE leading to treatment interruption was epiphyses premature fusion; treatment interruption for this reason occurred in 6.1% of patients on the chronic dosing regimen.

Withdrawals Due to Adverse Events

Withdrawal from the MOVE trial due to AE occurred in 6.1% and 5.7% of patients during the chronic and flare-up dosing regimens, respectively. Withdrawal due to epiphyses premature fusion occurred in 1 patient during each of the dosing regimen phases.

Mortality

There were no deaths to due AEs during the study period.

Notable Harms

Of the notable harms of interest, dry skin and dry lips occurred overall in 52.5% and 34.3% of patients, respectively, during chronic treatment; these occurred in 45.7% and 20.0% of patients, respectively, during flare-up treatment with palovarotene. Epiphyses premature fusion was observed in 11.1% of patients during chronic dosing and in 10% of patients during flare-up dosing. Hearing loss, pneumonia, suicidal ideation, and fractures occurred in less than 5% of patients who received treatment in the MOVE trial. There were no reported cases of osteoporosis, low bone density, decreased bone density, or onycholysis.

Harms Related to Growth

Growth measurements related to linear height, knee height, femur length, tibial length, and bilateral leg length in patients under the age of 18 years are summarized in Table 27 and Table 28. These measurements are categorized into 2 age groups: female patients aged 8 years to younger than 14 years and male patients aged 10 years to younger than 14 years; and patients of either sex aged 14 years to younger than 18 years. The mean changes in linear height z score in female patients aged 8 years to 14 years and male patients aged 10 years to 14 years were –0.36 (SD = 0.43) and –0.20 (SD = 0.34) in those who received treatment with palovarotene in the MOVE trial and who were untreated in the NHS, respectively. Among patients aged 14 years to younger than 18 years, the mean change in linear height among those who received treatment with palovarotene in the MOVE trial (–0.02; SD = 1.54) was less than among those who were untreated in the NHS (–0.55; SD = 1.61). Compared to untreated patients in the NHS, greater proportions of patients in the MOVE study who were aged 8 years to younger than 14 years (females) or 10 years to younger than 14 years (males) (61.3% versus 41.2%) and aged 14 years to younger than 18 years (either sex) (92.3% versus 88.9%) were documented with a pathological growth velocity rate of less than 4 cm per year. A similar trend was observed for other growth measures among patients aged 8 years to 14 years (females) or 10 years to younger than 14 years (males): a greater proportion of patients treated with palovarotene in the MOVE trial were documented with a pathological growth velocity rate of less than 2 cm per year for knee height (61.3% versus 52.9%) and a pathological growth velocity rate of less than 1.5 cm per year for tibial length (60.7% versus 50.0%).

Among patients aged younger than 18 years, the proportions with any epiphyseal growth plate abnormalities documented at month 12 were similar, at 45.8% in both the MOVE trial and the NHS (Table 28).

Table 26: Summary of Harms in the MOVE Trial (Principal Safety Population)

AE = adverse event; NHS = Natural History Study; TEAE = treatment-emergent adverse event.

Notes: Patients may have experienced more than 1 event or type of event. The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

^aGreater than or equal to 5% of patients in any group.

^bGreater than 1% of patients in any group.

^cIncludes rash, rash (generalized), and rash (maculopapular).

Source: Clinical Study Report for the MOVE study.¹⁹

Table 27: Summary of Harms Related to Growth Measurements for Patients Aged Younger Than 18 Years in the MOVE Study and the NHS (Principal Safety Population)

NHS = Natural History Study; SD = standard deviation.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

^aThe z scores were calculated with reference to growth charts from the US Centers for Disease Control and Prevention.

^bClinically meaningful postbaseline leg length discrepancies were defined as a femur plus tibia length difference of less than –1.5 cm or more than 1.5 cm.

Table 28: Summary of Growth Plate Abnormalities for Patients Aged Younger Than 18 Years (Principal and Supplementary Safety Population)

EGP = epiphyseal growth plate; NHS = Natural History Study.

Note: The data cut-off dates were December 4, 2019, for patients aged younger than 14 years and January 24, 2020, for patients aged 14 years or older.

Source: Clinical Study Report for the MOVE study.¹⁹

Critical Appraisal

Internal Validity

The open-label, nonrandomized study design of the MOVE trial makes it challenging to interpret the efficacy and safety of palovarotene. In randomized study designs, where treatment allocation is by chance, systematic baseline differences in known and unknown prognostic factors between groups may be ruled out. However, in nonrandomized study designs, decisions to provide treatment are made by researchers or by physicians and patients together; therefore, between-group imbalances in known and unknown prognostic factors are likely, and these can increase the uncertainty of any observed treatment effects. One known difference between the MOVE study and the NHS was prior exposure to palovarotene. Some patients entered the MOVE trial treatment-naive, while others entered having previous exposure to palovarotene from ongoing phase II studies. To address this consideration, an amendment to the study design restricted the analysis of all efficacy end points to only those patients who were treatment-naive. Imbalances in baseline characteristics between the MOVE trial and the NHS were observed for hearing loss; symptoms of pain, lethargy, and change in mood and behaviour at last flare-up; unknown or reported cause of last flare-up; method for reporting general medical history; and prior use of and new onset of medication. Although baseline total HO volume divided by age at baseline was adjusted for in the wLME model comparative analysis for the outcome of new annualized HO volume, none of the mentioned characteristics was adjusted in either the prespecified or post hoc analyses. Although the clinical experts consulted by CADTH for the purpose of this review noted that the previously mentioned characteristics were not considered to be important prognostic factors, the presence of these imbalances suggests the groups may be imbalanced on unknown or unmeasurable prognostic factors. As a result, confounding cannot be ruled out. Sensitivity analyses — including a matched pair analysis, a propensity score weighting analysis, and a tipping point analysis for missing data conducted on the overall trial population — were found to be generally consistent with the main analyses.

Of note, age at baseline differed between patients in the MOVE trial and the NHS, with a greater proportion of patients in the NHS aged 18 years or older compared to those in the MOVE trial (51.5% versus 30.4%). Similarly, compared to patients in the MOVE trial, patients in the NHS were older at the time of FOP diagnosis (7.5 years versus 6.6 years) and had had more time elapse between the time of FOP diagnosis and study enrolment (13.4 years versus 11.4 years). Based on input from the clinical experts, older age may be a marker of more severe disease (given that HO tends to worsen with age). However, the clinical experts added that is it difficult to ascertain the prognostic implication of age at diagnosis or the effect that age may have on treatment. The MOVE trial and the NHS were also imbalanced in terms of the proportion of patients who discontinued from the studies. Overall, a greater proportion of patients discontinued the NHS than the MOVE study (71.1% versus 17.8%). Of those patients who discontinued from the NHS, 45.6% did so to enrol in intervention studies, including the MOVE trial. It is uncertain whether or how those patients who discontinued from the NHS to enrol in other intervention studies differed from those who remained. Accordingly, attrition bias cannot be ruled out.

In open-label studies, knowledge of treatment or group assignments by patients, clinicians, and assessors may influence reporting or measurement of outcome and bias the treatment effect,⁴⁹ especially for subjective outcomes.^50-52 The open-label nature of the MOVE trial increases the risk of bias in the measurement of both the subjective outcomes, such as HRQoL, and the subjective harms. The primary outcome of annualized change in HO volume was considered by the clinical experts to be an objective outcome, which would mitigate the bias associated with open-label studies. Radiologic scans were independently assessed at an independent, central imaging laboratory. Interpretation of the scans followed a predefined procedure, and steps were taken to standardize the demarcation of anatomic regions to guide WBCT scans to minimize measurement variability. All reviewers were blinded to whether scans originated from the MOVE trial or the NHS. Although the visit schedules differed between the MOVE trial and the NHS, for the purpose of comparative analysis, measurements were aligned at month 12. However, the quantification of HO only occurred after new HO was detected qualitatively by the reader, which introduces a subjective component to the process. Evidence for validity of this measure is limited, and no evidence was found for its reliability and responsiveness.

Of concern, the sponsor was unblinded to the data at the time of the second interim analysis. Failure to maintain the blind may have introduced observer bias for outcomes related to HRQoL or based on clinical judgment. The direction and magnitude of bias related to the unblinding of the sponsor is uncertain.

Regarding the statistical analysis of the treatment effect, the initial analysis of efficacy results followed a defined statistical analysis plan. However, unanticipated events led to deviation from the statistical analysis plan. First, dosing of patients aged younger than 14 years was paused due to increased rates of premature epiphyseal closure. Second, at the time of the second interim analysis, statistical futility was determined based on the prespecified analysis using a Bayesian model that included all enrolled patients. On inspection of the data related to the overall study population, it was noted that 47.4% of patients in the MOVE trial and 23.8% of patients in the NHS had at least 1 region with negative new HO (i.e., new HO volume less than 0 mm³). At the time futility was declared, dosing for patients in the FOP palovarotene program was paused. In response, the Data Monitoring Committee and sponsor were unblinded to all study data. After unblinding and review of the post hoc efficacy analysis, the Data Monitoring Committee recommended that “palovarotene be continued in skeletally mature children ≥ 14 years.” Subsequently, a non-prespecified subgroup analysis was conducted in female patients aged 8 years and older and male patients aged 10 years and older that formed the basis of evidence for the approved indication. In addition to analysis that was consistent with the prespecified primary analysis for the original Bayesian approach, the assessment of the target population included additional data analyses to accommodate these negative HO values. These additional analyses included a wLME model that adjusted for baseline HO volume divided by age and an unadjusted Wilcoxon rank sum test alongside the Bayesian analysis specified as the primary analysis in the original study. The sponsor noted that the wLME model approach was the original primary efficacy analysis for HO volume before amendment 1 of the study protocol. The sponsor also noted that because the Wilcoxon rank sum test depends only on the numeric rank order of the observed volume of new HO rather than on the magnitude of HO, it is less influenced by extreme values compared to the wLME model; therefore, it is less influenced by the unanticipated negative HO volumes previously observed.

However, the following limitations for each of these analytical approaches should be considered. Regarding the Bayesian model, the priors used on parameters of this model appear to be informative priors, which likely influenced the estimates from the posterior distribution. Informative priors were likely necessary due to the complexity of the model and the relatively small sample size of the study. However, no clear justification was provided for the priors used; nor were sensitivity analyses performed using alternative priors to evaluate the influence of these priors on the final result. The assumed priors most likely biased results conservatively for comparing patients from the MOVE trial to those in the NHS. In addition, several details for estimating the results from the Bayesian model were not included in study reports (e.g., methods for evaluating convergence, the amount of burn-in and iterations performed for estimation, and whether multiple chains were used); these exclusions limit the ability to verify the validity of the reported estimates from the Bayesian model. Missing data also pose a concern for the Bayesian model because the random-effects model imposes a MAR assumption in the model, which assumes that missing observations for a given patient can be explained by observed outcomes from the same patient. However, given that the rate of missing data was higher at later time points (34% at 18 months for the MOVE trial and 38% at 24 months for the NHS, based on the original trial population), much of the missing data are likely related to patient discontinuation. As noted previously, the reason for discontinuation for a large portion of patients in the NHS was enrolment in an intervention study, which is likely informative for the missing HO volume data. If we assume that patients who enrol in intervention studies are more likely to have severe disease (and as a result, higher HO volume), then ignoring this missing data mechanism would tend to bias results conservatively versus patients from the MOVE trial. In general, the major limitations of the Bayesian modelling approach seem to suggest that results from this analysis are biased conservatively.

For the MOVE trial itself, due to the interruption in the administration of the study drug, interim analysis 3 served as the final analysis; the primary data analysis included assessments collected on or before December 4, 2019, for patients aged younger than 14 years, and on or before January 24, 2020, for patients aged 14 years or older. Questions remain about the impact of the clinical hold on the study data — specifically, whether the data collected after a prolonged period of interruption are interpretable. This interruption suggests that a large portion of patients with missing data from the MOVE study were aged younger than 14 years because follow-up was ended for these patients. The likely impact of this mechanism of missing data is unknown. However, the results included in the present report are restricted to those from interim analysis 3, which served as the final analysis.

Regarding the wLME approach, which relied on a single observation (the last observed follow-up), these results are limited by the loss of information, which likely limited power for this analysis. This analysis is also likely biased due to missing data, as previously described. Although the use of a single observation likely avoided patients not being included in the analysis due to missing data (though this was not reported on), as noted, for patients who discontinued from either study, the last observed data value is not likely representative of the final value that might have resulted if these patients had been followed. As in the Bayesian analysis, the overall impact of this limitation likely biases the results conservatively. Another limitation arises from the adjustment for baseline HO volume divided by age: it has been shown that adjusting for the ratio of 2 variables in a regression analysis can result in a spurious correlation between an exposure and an outcome that undermines the validity for inference and estimation under this approach.⁵³ Due to this limitation, the overall reliability of the wLME model results for drawing conclusions is uncertain.

Regarding the Wilcoxon rank sum test, the major limitation of this approach is that it is an unadjusted comparison. Covariate adjustment was limited for all analyses presented, but the lack of adjustment for age and baseline HO volume in the Wilcoxon test is particularly concerning because these elements are likely to have strong influences on the change in HO volume over time. Age is of particular concern because there were notable imbalances in age between the MOVE study and the NHS, with patients in NHS generally being older. When measuring physical attributes, such as HO volume, a patient’s age is often a key consideration, particularly for the pediatric population.

Finally, conclusions from each of the analyses described are limited by the post hoc nature of the analyses based on a target population that was not prespecified in the original trial. Drawing conclusions based on reported P values from these analyses is not recommended because these are not suitable for inference in this context. Instead, evaluation of the evidence should focus on the reported effect estimates and measures of uncertainty around those estimates along with the aforementioned key limitations of these estimates. Although all of the analyses employed in the MOVE trial are limited in this fashion, the Wilcoxon test, in particular, does not provide an interpretable effect estimate that can be used to inform decision-makers.

External Validity

Population

The clinical experts consulted by CADTH for the purpose of this review noted that FOP is variable and heterogenous in its presentation and progression. However, the clinical experts felt that the study populations in both the MOVE trial and the NHS were reflective of what is observed in the clinical setting. The clinical experts added that in the clinical setting, a clinical diagnosis of FOP is sufficient to initiate treatment with palovarotene, and that treatment should not be predicated on specific FOP variants.

Comparator

Currently, there are no other treatments for this patient population. Based on input from the clinical experts, the NHS was an appropriate comparator to the MOVE trial. The clinical experts noted that during the NHS period (December 2014 to April 2020), there were no changes in treatment practices that may have influenced patient outcomes. The clinical experts also added that because there are no approved treatments for FOP — and no prior therapies have shown to affect HO volume — it is unlikely that any observed differences in standard of care or background therapy could bias the assessment of change in HO volume in either the MOVE trial or the NHS.

Outcome

The clinical experts agreed that while HO volume is a good outcome for proof of concept, clinically, overall HO volume alone is not relevant. The clinical experts noted that relevant clinical outcomes for this patient population are related to HRQoL, physical function, ROM, and frequency of flare-ups.

Other Relevant Evidence

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

In addition to the pivotal MOVE trial, the following studies were considered for inclusion in this section of the report: Study 201 and Study 202. Study 201 was a phase II, randomized, double-blind, placebo-controlled trial; Study 202 was an open-label extension of Study 201. Given that the Health Canada–recommended dosage for palovarotene was used in Study 202 and not Study 201, only the relevant results from Study 202 are summarized here as supportive evidence. In addition, the study populations were not aligned with the indicated population, given that patients 6 years or older with FOP were included. The phase II study results informed the selection of the dosing regimen studied in the MOVE trial in terms of extending the duration of flare-up treatment and increasing the daily doses.¹⁹

Study 201

Study 201 was a phase II, randomized, double-blind, placebo-controlled trial evaluating the ability of different doses of palovarotene to prevent HO at the flare-up site in patients with FOP compared to placebo. For inclusion in the study, patients had to be 6 years of age or older with clinically diagnosed classic FOP and had to have a symptomatic onset of a distinct flare-up within 7 days of day 1 of the study. A total of 40 patients with FOP were randomized to receive treatment according to the 10 mg plus 5 mg palovarotene regimen (i.e., 10 mg for 14 days followed by 5 mg for 28 days), the 5 mg plus 2.5 mg palovarotene regimen (i.e., 5 mg for 14 days followed by 2.5 mg for 28 days), or placebo for 6 weeks. The 6-week treatment period was followed by a 6-week follow-up period.

The primary end point in Study 201 was the percentage of patients with response (defined by no or minimal new HO at the flare-up site) at week 6 compared with baseline. The key secondary end point was change from baseline in HO volume at week 6 and week 12 as assessed by low-dose CT scan. Detailed study results were not reported for Study 201 because the study population and intervention were not aligned with the population and intervention of interest identified by the CADTH review protocol.

Study 202

Description of Studies

Detailed information for Study 202 is summarized in Table 29. Study 202 was designed as a phase II, multicentre, open-label study. Its main objective was to evaluate the safety and efficacy of different palovarotene dosing regimens in patients with FOP. Efficacy was assessed based on the ability of palovarotene to prevent the formation of new HO as assessed by low-dose WBCT scan (excluding head). The study was conducted in 3 parts: part A, part B, and part C. In part A, all pediatric and adult patients who had successfully completed Study 201 were enrolled and followed for up to 36 months. A total of 40 patients were enrolled, including 20 patients with 28 treated flare-ups and 20 untreated patients. Patients who had an eligible flare-up received 10 mg palovarotene daily for 14 days, followed by 5 mg palovarotene daily for 28 days (or weight-based equivalent). The primary end points and key secondary end points for part A were the amount of reduction in each of the following at week 12: HO volume, frequency of flare-ups, and long-term postexposure persistence of effect. The Health Canada–recommended dosage for palovarotene was used in parts B and C of Study 202, but not in part A; therefore, part A is not relevant and is not described further in this report. In part B, patients who successfully completed Study 201 (including any patient who participated in part A of Study 202) as well as up to 20 new adult patients were followed for up to 24 months. In total, part B enrolled 54 pediatric and adult patients who had a total of 79 treated flare-ups. The adult cohort included all patients with at least 90% skeletal maturity, regardless of age. The pediatric cohort included all patients with less than 90% skeletal maturity. Any patient in the pediatric cohort who achieved greater than or equal to 90% skeletal maturity during part B was considered for enrolment into the adult cohort at the investigator’s discretion. Part B added a 5 mg palovarotene daily chronic treatment regimen administered between flare-ups for participants in the adult cohort for up to 24 months. Part B also increased the flare-up dosing to 20 mg palovarotene daily for 28 days, followed by 10 mg palovarotene daily for 56 days (or weight-adjusted equivalents in the pediatric cohort). Treatment could be extended if the flare-up was ongoing. The primary end point for part B was the proportion of flare-ups with no new HO. In part C, 48 patients from part B were followed for up to an additional 48 months. There have been no new participants in part C. All eligible patients, including skeletally immature participants, received a chronic treatment regimen of 5 mg palovarotene daily (with weight-adjusted doses for skeletally immature participants). The primary end point for part C was the annualized change in new HO volume.

Table 29: Details of Study 202

AE = adverse event; CAJIS = Cumulative Analogue Joint Involvement Scale; FOP = fibrodysplasia ossificans progressiva; FOP-PFQ = fibrodysplasia ossificans progressiva Physical Function Questionnaire; HO = heterotopic ossification; ROM = range of motion; PROMIS = Patient-Reported Outcomes Measurement Information System; SAE = serious adverse event; WBCT = whole-body CT.

^aOnly outcomes identified as relevant efficacy outcomes in the CADTH systematic review protocol and reported in the MOVE trial are listed.

Source: Clinical Study Report for Study 202.^5.4

Baseline Characteristics

The majority of patients in part B were white (75.9%), and most were female (57.4%). The median age of patients was 19.0 years (range, 7 years to 54 years), with a median weight of 52.3 kg (range, 21 kg to 108 kg) and a median height of 165.3 cm (range, 118 cm to 189 cm).

The majority of patients treated with palovarotene in part C were white (76.1%), and half were female (50.0%). The median age of treated patients was 20.0 years (range, 9 years to 48 years), with a median height of 165.2 cm (range, 124 cm to 190 cm) and a median weight of 60.1 kg (range, 22 kg to 100 kg). There were 2 patients who did not receive treatment in part C.

Results

This section focuses on the relevant outcomes for part B and part C that are identified in the CADTH systematic review protocol and reported in the MOVE trial.

Efficacy

Detailed information for the primary and key secondary end points of part B and part C in Study 202 is summarized in Table 30 and Table 31.

The primary end point in part B was the proportion of flare-ups with no new HO at week 12. At week 12, 72.5% patients had flare-ups with no new HO. The primary end point in part C was annualized change in new HO volume. The changes from baseline in mean volume of new HO assessed by WBCT scan (excluding head) were 9,332 mm³, 62,836 mm³, and 89,487 mm³ for months 12, 24, and 36, respectively. Change in ROM (assessed using the CAJIS) and change in physical function (assessed using the FOP-PFQ and PROMIS Global Physical Health scale) were the key secondary end points for parts B and C. The changes in CAJIS score from baseline were 1.1 at month 12 and 15.0 at month 24 in part B. In part C, the changes were 0.5, 2.7, 3.9, and 2.0 for months 12, 24, 36, and 60, respectively. The changes in FOP-PFQ total score from baseline were 3.34 at month 12 and 0.89 at month 24 in part B, and in part C, the changes were 2.79, 8.68, 13.75, and 10.24 for months 12, 24, 36, and 60, respectively. The changes in adult PROMIS Global Physical Health T-scores from baseline were –0.3 at month 12 and 2.5 at month 24 in part B, and for part C the changes were 0.33, –1.78, –2.90, and –3.18 for months 12, 24, 36, and 60, respectively.

Table 30: Summary of Primary Outcomes in Study 202, Part B and Part C

HO = heterotopic ossification; NR = not reported; SD = standard deviation.

Note: Patients could contribute more than 1 flare-up. Any flare-up treated with palovarotene 20 mg and 10 mg was included, regardless of whether the patient received chronic treatment.

^aPart B flare-up population set.

^bThe WBCT population set (patients who received chronic palovarotene 5 mg once daily) was used for part B, and the enrolled population set was used for part C (i.e., all patients enrolled into part C).

Source: Study 202 Clinical Study Report.^5.4

Table 31: Summary of Key Secondary Outcomes in Study 202, Part B and Part C