CADTH Reimbursement Review

Selumetinib (Koselugo)

Sponsor: Alexion Pharma GmbH

Therapeutic area: Neurofibromatosis type 1

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

Neurofibromatosis type 1 (NF1) is an autosomal dominant genetic disorder associated with progressive cutaneous, neurologic, skeletal, and neoplastic manifestations.^1-3 Approximately half of all NF1 cases are familial, while half arise from spontaneous mutations in the NF1 gene.^3,4 Currently, the incidence of NF1 in Canada is unknown, although it is estimated to occur in 1 in 2,500 to 3,000 births.^2,4-6 Patient group input received by CADTH for this review noted that there are currently more than 12,000 cases of NF1 in Canada.

The most common manifestations of NF1 include abnormally coloured patches of skin (café-au-lait macules [CALMs]), freckling under the arms and in the inguinal region, and benign tumours predominantly in the skin and nerves known as neurofibromas. Other manifestations may include bone dysplasia, scoliosis, ocular problems, and neurologic complications with impacts such as cognitive impairments and learning disabilities. Neurofibromas are histologically benign nerve sheath tumours that typically originate in the terminal nerve branches of the skin. Plexiform neurofibromas (PNs) are the most common type of tumour in patients with NF1, occurring in up to 50% of patients.^7-10 One or multiple PNs may grow along large nerves and plexuses anywhere in the body, with varying manifestations that continue to develop to early adulthood, and multiple PNs may be both symptomatic and asymptomatic in the same individual.^11-13 Additionally, PNs have a complex shape and can reach large sizes, resulting in clinical symptoms such as disfigurement, motor dysfunction (weakness and restricted range of motion [ROM]), pain, and neurologic dysfunction. The severity of symptoms from PNs may range from mild to severe; however, the presence of symptoms may depend on their location and impact on surrounding structures. PNs grow most rapidly during early childhood, although growth rates vary among patients.^14-16

Treatment and clinical management options for NF1-associated PNs are extremely limited and depend on symptomatology. For symptomatic patients, treatments aim to relieve symptoms caused by the individual PNs. Currently, the only available options to treat and manage NF1-associated PNs are pain management and surgical excision to remove as much of the tumours as possible. However, surgery is not a viable option for many patients as most PN are not amenable to complete resection due to encasement of, or proximity to, vital structures.^7,17,18

Selumetinib (Koselugo) is an orally available, selective inhibitor of mitogen-activated protein kinases (MEK) 1 and 2. Selumetinib blocks MEK activity and inhibits growth of RAF-MEK-ERK pathway–activated cell lines, thereby leading to an inhibition of cellular proliferation and PN growth. Selumetinib is available as 10 mg or 25 mg oral capsules. The recommended dosage of selumetinib is 25 mg/m² twice daily based on body surface area (BSA). The Health Canada indication for selumetinib is for the treatment of pediatric patients with NF1 aged 2 years and older who have symptomatic, inoperable PNs. The notice of compliance was granted on August 31, 2022. Selumetinib has not been previously reviewed by CADTH.

The objective of the current report is to review the beneficial and harmful effects of selumetinib 10 mg and 25 mg twice daily for the treatment of pediatric patients with NF1 aged 2 years and older who have symptomatic, inoperable PNs.

Stakeholder Perspectives

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

Patient Input

CADTH received input from 2 patient groups: the Tumour Foundation of BC (TFBC) and the Canadian Organization for Rare Disorders (CORD). The TFBC provides essential information and support services to patients with neurofibromatosis and their families. CORD works with governments, researchers, clinicians, and industry to promote research, diagnosis, treatment, and services for all rare disorders in Canada.

Both patient groups conducted online surveys in November 2022, recruiting patients with NF1 and their caregivers. Additionally, the TFBC conducted a Zoom videoconference focus group. The TFBC group recruited 25 patients and caregivers, and CORD recruited 8 caregivers. Key themes identified by patients and caregivers included limitations on daily living, functional, and social activities; moderate to severe chronic pain; dependency on caregivers into adulthood; financial stress because of the diagnosis; and the lack of treatment options, which negatively affects the emotional well-being of patients and families.

Respondents from both groups described difficulties obtaining a diagnosis of NF1, as well as significant impacts on both affected children and their families in terms of managing physical and mental disability. Additionally, substantial negative mental health impacts were reported, with most patients living with anxiety and fear over their diagnosis, and some patients experiencing suicidal feelings or actions. Respondents to surveys and interviews were surprised and disappointed with the lack of available treatment options and support, with 46% not having been offered any kind of treatment, and only 17% of patients who were offered treatment experiencing minimal improvement in symptoms.

No patients in the TFBC survey had experience with selumetinib. The half (n = 4) of the CORD respondents who had experience with selumetinib through clinical trials described it as “miracle drug” that was “life-changing” due to substantial improvements in pain level, functional abilities including speaking clearly and chewing food, and softening and shrinking tumours that were previously disabling and/or disfiguring.

Numerous outcomes were identified as important to patients, reflecting the heterogenous nature of the disease, but common themes included an overall improved quality of life (QoL), a desire for reduction in pain, reduction or prevention of tumour size or growth, improved function and emotional well-being, greater independence from caregivers, and fewer health care visits.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The information in this section is based on input received from a panel of 6 clinical specialists consulted by CADTH for the purpose of this review.

The clinical experts indicated that the main limitations and unmet needs of pediatric patients with NF1 with symptomatic PNs is the lack of access to disease-modifying medical interventions that can reduce the burden of disease or stabilize symptomatic PNs. The clinical experts noted that there are currently no established practice guidelines for this heterogenous disease. For patients with symptomatic PNs, the only available treatment option, surgery, is either aimed at excising tumours if possible, or debulking if complete excision is not achievable. The experts noted that surgery is not curative for most large or extensive PNs and is associated with significant risks of secondary injuries depending on the number, location, size, and vascularity of tumours, particularly for PNs involving large arteries or nerves. The experts added that multiple invasive surgeries may be required, as tumours may regrow or increase in size, further increasing the risks to patients. Aside from surgery, current treatment strategies consist of “watch and wait” for patients with PNs that are not currently symptomatic. Otherwise, treatment for patients with symptomatic PNs focuses on relieving pain, reducing functional impairment, and improving overall QoL. The panel emphasized the availability of MEK inhibitors through managed access programs, noting that selumetinib is the first and only Health Canada–authorized MEK inhibitor available for the treatment of PNs outside of clinical trials. The panel concluded that selumetinib is expected to cause a shift in the current treatment paradigm given the absence of other medications available for this population. The experts stated that, should selumetinib be recommended for reimbursement, it would likely be the initial therapy of choice.

The clinical experts noted that only a minority of NF1 patients have symptomatic PNs. Patients with NF1 are diagnosed based on standard, well-established, and recently updated clinical diagnostic criteria, including clinical characteristics such as CALMs, and the presence of neurofibromas. Although recently updated diagnostic criteria include genetic testing, the experts noted that genetic testing is not required for diagnosis of NF1, and that the results of genetic testing do not affect treatment decisions once a clinical diagnosis of NF1 has been established. The clinical diagnosis of NF1 is relatively straightforward in older children, adolescents, and adults, but can be challenging in younger children due to the absence of clinical characteristics such as CALMs or PNs. However, the updated diagnostic criteria, which include genetic testing in patients without a family history, have improved confirmatory diagnosis in young children before they manifest other clinical features of NF1. Diagnosis of large, extensive, or rapidly growing PNs generally requires more clinical expertise, and possibly more complex tumour characterization, including MRI, and sometimes a biopsy if there is concern about malignant transformation.

The experts also highlighted that, in terms of natural history, there is a trend for tumours to appear and grow rapidly in early childhood (before the age of 6 to 8 years), and then slow down or remain static in adulthood. Rapid growth of a PN and transformation to a malignant peripheral nerve sheath tumour (MPNST) is a concern. The experts also discussed the uncertainty regarding treatment decisions for asymptomatic patients, as none have been established. In addition, no evidence is available regarding whether treatment with MEK inhibitors such as selumetinib can prevent growth of new PNs.

The experts emphasized the heterogeneity of the disease in NF1 patients, with cutaneous neurofibromas and PNs often occurring throughout the body and ranging in severity from asymptomatic to severely debilitating due to pain, functional impairment, or disfigurement. One clinical expert highlighted that disfigurement due to large, visible PNs is a source of anxiety and concern due to public fear and social stigmatization. The panellist also highlighted the potential for ongoing problems to persist into adulthood due to large PNs that may result in severe disfigurement and displacement of joints and bones; however, there is no clear evidence that treating asymptomatic PNs with selumetinib in children will prevent the development of symptoms in adults. Other concerns for the NF1 population raised by the experts include deficient social skills, frequent learning disabilities, autism, and attention-deficit/hyperactivity disorder, further highlighting the vulnerability and marginalization of these patients.

Treatment with selumetinib is the only available medical treatment for NF1 patients whose extensive inoperable PNs are causing significant pain, functional impairment, and/or disfigurement. Although it is difficult to determine which patients are most likely to respond to treatment, 1 clinical expert currently treating pediatric patients via compassionate access to selumetinib stated that about 80% of patients will respond to treatment. The experts noted that most NF1 patients with PNs are asymptomatic, and the benefit of treatment for these patients has not yet been established. Clinical trials of selumetinib are currently being conducted in the adult population, and these should provide insight into similarities or differences in effectiveness by age. The experts agreed that the lack of knowledge about both the potential benefits and harms associated with long-term selumetinib treatment is a concern, given that NF1 is a life-long disease. The experts also noted that the life expectancy of NF1 patients has been reported to be reduced by 10 to 15 years, although estimates of life expectancy with currently available medical management are unknown.

The clinical experts noted that current clinical trials aim to address important outcomes; however, given the heterogeneity of the disease, standardizing subjective measures (such as pain perception) across this population is problematic, and interpreting the results relies heavily on clinical judgment. The clinical experts agreed that the most important outcomes in the management of pediatric patients with NF1 and symptomatic, inoperable PNs is the reduction or improvement of symptoms (i.e., reduced pain and improved function), as well as overall improvements in QoL and disease stabilization. The experts noted that volumetric MRI, although used in the clinical trial to define disease progression, is only used by the National Institutes of Health (NIH) for research purposes and is not available in Canadian clinical practice. The experts considered a change in planar tumour size of 20% to 25% to be indicative of response to treatment. One expert discussed the potential for symptomatic disease progression despite no evidence of progression on imaging studies and for improvement in symptoms without reduction of tumour size on imaging studies. In addition, the panellists emphasized that it is not always clear which tumours are the cause of symptoms when patients have large numbers of PNs, making it difficult to know when the disease is progressing. The experts also noted that tumours are frequently irregular in shape, making measurements of changes in tumour size difficult. As a result, the panel agreed that response to treatment is multidimensional and must consider reductions in tumour sizes, changes in symptoms, and improvements in function and disfigurement.

The experts stated that young children with NF1 and symptomatic PNs may initially be followed with an MRI every 3 months, in addition to annual follow-ups with NF1 specialists to assess other features of the disease. According to the panel of experts, upon initiating treatment, patients would be seen weekly for a month, then monthly, and if treatment is well tolerated or disease stabilizes, follow-up intervals would be extended to 6 months. The experts also noted that imaging in young children often requires a general anesthetic. While no firm treatment duration for selumetinib has been determined, the experts suggested that, similar to the SPRINT phase II trial, in clinical practice patients would continue treatment until disease progression or toxicity. The experts expected that the initial treatment authorization period for selumetinib should be 18 months. The clinical experts agreed that selumetinib would be discontinued in patients who are not responding (i.e., tumour growth, lack of stabilization, or improvement of symptoms), or in patients with severe adverse events (AEs) that are cannot be managed. The experts also noted that the need for surgery to further debulk tumours could indicate that treatment is not working and should therefore be discontinued. One clinical expert suggested that selumetinib may be used in conjunction with debulking surgery, although there is currently no evidence for this approach.

The experts indicated that expertise in the use of selumetinib in Canada is limited to pediatric oncologists and neurooncologists in tertiary care hospitals. Currently, only pediatric oncologists are prescribing treatment with selumetinib, as they have the experience and know-how to manage these patients. However, the experts highlighted that, with further insight and growing experience, NF1 experts who are pediatricians could manage this oral treatment. Given the heterogeneity in the disease and the individualized approach to treatment, decisions often involve a multidisciplinary team of pediatricians, NF1 experts, neurooncologists, and nurse practitioners. The experts also emphasized the importance of consulting other specialists, including surgeons, cardiologists, ophthalmologists, dermatologists, and pharmacists, on the management of selumetinib, adverse effects, and drug interactions. The expert panel also anticipated access to specialty clinics may be a limiting factor for patients in remote areas, as patients would be required to attend in-person appointments for treatment initiation and imaging follow-up as well as to assess safety.

Clinician Group Input

Input for this review was received through shared clinical experiences from 1 clinician group, the Canadian Pediatric Brain Tumour Consortium (CPBTC), which included 27 pediatric neurooncologists across Canada.

Overall, the clinician group input was aligned with that given by the clinical expert panel convened by CADTH, highlighting that no systemic therapies exist for treating NF1-associated PNs, which represents the major unmet need in this patient population, and that surgical resection, if feasible, is the only option currently available for patients. The clinician group emphasized that selumetinib has clearly shifted the current treatment paradigm and emerged as the standard-of-care, first-line therapy for patients with inoperable, symptomatic PNs. They described the patients most in need of intervention as those in whom PNs are invading critical structures, causing a deformity, or causing functional impairment in activities of daily living such as walking, swallowing, or eating. The clinician group also noted that, in Canada, treatment initiation with selumetinib is currently limited to pediatric oncologists, neurooncologists, or pediatric neurologists with an expertise in neurooncology. The CPBTC suggested that treatment with selumetinib in Canada could be initiated by oncologists and followed up remotely in conjunction with local clinicians.

Finally, the CPBTC group noted that many parents of children with NF1 also have NF1 themselves and are likely to have lower socioeconomic standing in part because of the disease. It was therefore the CPBTC’s opinion that many patients and parents of patients are more likely to lack private insurance that covers selumetinib, which could result in unequitable access in some parts of the country. The CPBTC group emphasized that children without private insurance who are also not eligible for the provincial public drug plans will need special consideration and that the drug in question urgently requires reimbursement and equitable access as a standard-of-care treatment for patients with NF1 and symptomatic PNs.

Drug Program Input

The drug programs identified the following jurisdictional implementation issues: relevant comparators, considerations for initiation of therapy, considerations for continuation or renewal of therapy, considerations for discontinuation of therapy, considerations for prescribing of therapy, generalizability, care provision issues, and system and economic issues. Table 4 provides more details.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

SPRINT phase II is a phase II, open-label, single-arm, multicentre study that aimed to evaluate the efficacy of 25 mg/m² selumetinib twice daily in 50 pediatric patients with NF1 and inoperable PN. The primary outcome of the SPRINT phase II study was the objective response rate (ORR) determined by change in PN volumes through volumetric MRI. Secondary outcomes included patient-reported outcomes (PROs) and functional evaluations to determine the effect of selumetinib on pain, motor function, and health-related quality of life (HRQoL). Two data cut-offs (DCOs) were submitted for the SPRINT phase II trial.^19,20 The primary DCO occurred on June 29, 2018, and an updated DCO occurred on March 31, 2021, providing a maximum follow-up of 5.6 years.¹⁹

At baseline, patients included in the SPRINT phase II trial were mostly white (42 [84.0%]) and male (30 [60.0%]), with a mean age of 10.3 years (standard deviation [SD] = 3.92 years). The median number of target PNs causing morbidity was 3 (range = 1 to 4), and the mean target PN volume was 837.11 mL (SD = 925.011), ranging from 5.6 mL to 3,820.0 mL. Pain was present in the target PNs in 26 patients (52.0%). The most common locations of target PNs were the neck and trunk, and the trunk and extremity (12 [24.0%], each), and || ||||||| patients had at least 1 prior PN- or NF1-related surgical procedure.^19,20

Efficacy Results

Key results of the efficacy analyses of the SPRINT phase II trial are summarized in Table 2.

Pain

Evaluation of pain was a secondary end point of the SPRINT phase II trial. Pain intensity was measured by the Numeric Rating Scale-11 (NRS-11); a self-evaluation of pain in patients aged 8 years and older consisting of 4 questions scored on a scale of 0 (no pain) to 10 (worst pain imaginable). A threshold of 2 points was suggestive of clinically meaningful change according to the literature. The interference of pain on daily functioning was measured by the Pain Interference Index (PII), a 6-item scale that assesses the extent to which pain has interfered with daily activities in the past 7 days (0 = not at all to 6 = completely). Higher scores for both scales indicate greater impact of pain on patients.

At the June 29, 2018, DCO, the mean adjusted change from baseline score for target tumour pain intensity measured by the NRS-11 was reduced at precycle 13 by −2.07 points (95% confidence interval [CI], −2.84 to −1.31).¹⁹ At the March 31, 2021, DCO, representing a longer follow-up period, the NRS-11 target tumour pain was reduced at precycle 13 with an adjusted mean change from baseline of ||||| points (95% CI, ||||| || |||||).²⁰

For the PII, the self-reported adjusted mean change from baseline score at precycle 13 was reduced by −0.65 points (95% CI, −0.89 to −0.42), and the adjusted mean change from baseline in parent-reported PII scores at precycle 13 was reduced by −0.82 points (95% CI, −1.17 to −0.47) at the June 29, 2018, DCO.¹⁹ At the March 31, 2021, DCO, results were consistent with the primary analysis, with a reduction in the adjusted mean change from baseline at precycle 13 of ||||| (95% CI, ||||| || |||||) for the self-reported total score, and ||||| (95% CI, ||||| || |||||) for the parent-reported score.²⁰

Motor Function

Motor function was evaluated in patients with motor morbidity using the strength of muscle groups and ROM tests, as well as the Patient-Reported Outcomes Measurement Information System (PROMIS) mobility and upper extremity domains. The PROMIS was completed by both the patient and the parent. Higher scores indicate better physical functioning.

The baseline score for the self- and parent-reported assessments in the mobility domains of PROMIS were 46.57 (SD = |||||) and 37.43 (SD = |||||), while the baseline scores for the self- and parent-reported assessments in the upper extremity domain were 45.95 (SD = ||||||) and 38.15 (SD = ||||||), with higher scores indicating better physical functioning. At the March 31, 2021, DCO, self-reported mobility and self-reported upper extremity improved, with adjusted mean changes from baseline at precycle 13 of |||| points (95% CI, ||||| || ||||), and |||| points (95% CI, ||||| || ||||), respectively. In the parent-reported assessments, the adjusted mean change from baseline at precycle 13 improved in the mobility and upper extremity domains by |||| points (95% CI, |||| || ||||), and |||| points (95% CI, ||||| || ||||), respectively.²⁰

Strength using the manual muscle test (a Medical Research Council 5-point Likert scale) was assessed in the 33 patients who had motor morbidity in any body quadrant at enrolment. At the March 31, 2021, DCO, || patients had evaluable strength assessments at baseline and precycle 13, with a mean strength score of |||| (SD = |||||) at baseline, and an adjusted mean change from baseline of |||| points (95% CI, |||| || ||||). For ROM, the mean ROM sum of all joints was |||||| degrees (SD = |||||||), and the adjusted mean change from baseline at precycle 13 was an increase of ||||| degrees (95% CI, ||||| || ||||||).²⁰

Health-Related Quality of Life

HRQoL, a secondary end point of the SPRINT phase II study, was measured using the Pediatric Quality of Life Inventory (PedsQL) tool, which assesses function in 4 domains: physical (8 items), emotional (5 items), social (5 items), and school (5 items) on a 5-point Likert scale (0 = never a problem; 4 = almost always a problem), with scores reverse-transformed to a 0-to-100 scale, with higher scores indicating better HRQoL. No minimally important difference threshold was identified in the literature. The observed mean score at baseline was 73.91 (SD = ||||||) in the self-reported version, and 60.79 (SD = ||||||) in the parent-reported version. At the March 31, 2021, DCO, the adjusted mean changes from baseline in the self-reported version of the PedsQL were |||| points (95% CI, |||| || |||||) and ||||| points (95% CI, |||| || |||||) in the parent-reported version, suggesting improvements in HRQoL.²⁰

Objective Response Rate

The ORR was the primary end point of the SPRINT phase II study. At the June 29, 2018, DCO, 33 patients (66.0% [95% CI, 51.2 to 78.8]) achieved an ORR, according to the Response Evaluation in Neurofibromatosis and Schwannomatosis (REiNS) criteria. The ORR achieved in the sensitivity analysis based on an independent central review (ICR) was |||||. Differences in the ORR between the primary central analysis and the ICR analysis were primarily due to differences in categorization of confirmed partial response (PR) versus stable disease (based on the chosen threshold of 20% shrinkage to determine response), where |||||| patients were considered to have a confirmed PR despite reductions in tumour size being slightly below the threshold of 20%.¹⁹ At the later, March 31, 2021, DCO, the ORR was 68.0% (95% CI, |||| || ||||).²⁰ At both DCOs, the ORR was based on confirmed PRs.

An exploratory ICR analysis using modified Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST 1.1), which used a 30% volume reduction for PR as opposed to 20% with REiNS, was also conducted at the June 29, 2018, DCO. Based on the RECIST 1.1 assessment, the ORR was only ||||, with | |||||| patients having an unconfirmed PR, and || ||||||| patients having stable disease.¹⁹

Change in target PN volume was also assessed as part of the volumetric MRI and application of the REiNS criteria. At the March 31, 2021, DCO, the mean percent change from baseline in target PN volume at precycle 13 was ||||||| (SD = ||||||), corresponding to a mean absolute change of ||||||| mL (|||||||). The proportion of patients with a maximum reduction from baseline of 20% or greater was identical to the June 29, 2018, DCO, at 77.1%, and | ||||||| with a maximum reduction from baseline of 40% or greater.²⁰

Harms Results

Nearly all patients (49 [98.0%]) in the SPRINT phase II trial experienced a treatment-emergent adverse event (TEAE). The most frequent TEAEs reported at the March 31, 2021, DCO were vomiting (|| |||||||), increased blood creatine phosphokinase (CPK) (|| |||||||), diarrhea (|| |||||||), nausea (|| |||||||), and dry skin (|| |||||||). Grade 3 or higher TEAEs were reported in || ||||||| patients, with the most frequent being diarrhea || ||||||||| hypoxia || |||||||| and pyrexia || |||||||. Overall, || ||||||| patients had at least 1 TEAE leading to a dose interruption.²⁰

At the March 31, 2021, DCO, || ||||||| patients experienced serious adverse events (SAEs), the most frequent being infections and infestations || ||||||||| and gastrointestinal disorders || ||||||| constipation, abdominal pain, and diarrhea).²⁰

A total of | ||||||| patients discontinued selumetinib due to AEs; |||||| of which were grade 3 (acute kidney injury, diarrhea| |||||| paronychia, and increased weight) and |||||| were grade 4 |||||| |||||||||| ||||||||| and skin ulcer), with ||| patient experiencing a grade 3 and grade 4 AE leading to withdrawal.²⁰

No AEs with a fatal outcome were reported in the SPRINT phase II study; however, after the March 31, 2021, DCO, |||||| patients died due to progressive neurofibrosarcoma after selumetinib treatment was terminated. These deaths were not attributed to treatment with selumetinib.²⁰

The most frequent notable harm associated with selumetinib was paronychia, occurring in || ||||||| patients. The majority of cases were grade 1, |||||| were grade 2, and |||||| was grade 3. One patient discontinued treatment due to grade 3 paronychia at the earlier (June 2018) DCO. Other notable harms included ||||||| |||||| ||| |||||||| ||| |||||||||||||| |||||| ||| |||||||||²⁰

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

CFB = change from baseline; CI = confidence interval; DCO = data cut-off; NRS-11 = Numeric Rating Scale-11; SD = standard deviation.

^aThe model included terms for precycle, baseline score, age, the number of morbidities at baseline and baseline × precycle interaction.

^bThe model included terms for precycle, baseline score, age, the number of clinical complications at baseline and baseline × precycle interaction.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Critical Appraisal

SPRINT phase II was a phase II, open-label, single-arm, multicentre study. The choice to conduct a single-arm study has implications for the overall strength and interpretability of the results. As a single-arm study, there is an increased risk of bias in the estimation of treatment effects due to the potential for confounding related to natural history, and other unidentified prognostic factors that could affect all study outcomes. The noncomparative design of the SPRINT phase II trial precludes an assessment of the therapeutic benefit or safety of selumetinib. In a single-arm trial, because all patients received the same treatment, treatment effect on time-to-event end points are uninterpretable and were only considered exploratory and supportive. Awareness of treatment assignment by both patients and parents or caregivers increases the risk of detection bias and performance bias and may lead to systematic overestimates or underestimates of the overall treatment effect. As such, the open-label trial design limits interpretability of the clinical outcome assessments such as the PRO and functional end points, as well as AEs. The already small sample size (N = 50) was further restricted for secondary end points, including PROs and functional evaluations, as these were based on patients with target PNs in specific locations or limited to patients of a certain age. The outcome of the SPRINT phase II study was the ORR and was considered appropriate by the clinical experts consulted by CADTH and the CADTH review team as an objective measure to assess the activity of selumetinib. Secondary clinical outcome assessments (PROs and functional evaluations) were considered appropriate to evaluate the wide range of PN-related morbidities; however, based on the design of the SPRINT phase II study, and the lack of statistical tests or imputation of missing data, the results should only be viewed as supportive of the overall effect of selumetinib.

There is a lack of standardized end points for trials in NF1. As previously noted, multiple outcomes were included in the SPRINT phase II trial, including response and time-to-event outcomes based on volumetric MRI using the REiNS imaging criteria. The clinical experts consulted by CADTH noted that volumetric MRI is not used in routine clinical practice as it is not standard of care in Canada, and that evidence of disease progression is multifactorial, based on standard imaging techniques, although they emphasized the importance of clinical symptomatology and physical assessment in determining progression and response. As such, patients in Canadian clinical practice would be evaluated for progression slightly differently than in the SPRINT phase II trial, potentially affecting the generalizability of the results. Patient-reported outcomes (NRS-11, PII, PROMIS, and PedsQL) and functional outcomes (strength and ROM) were also evaluated in the SPRINT phase II trial. The clinical experts consulted by CADTH noted that the outcome scales reported in the trial were not used in routine clinical practice and may not be generalizable to the typical patient in Canada. The clinical experts noted that a gestalt-type approach is considered in clinical practice for overall improvement or deterioration in symptomatology overall, as opposed to specific changes in certain domains (e.g., grooved pegboard test and key pinch grip), although variation and heterogeneity by patients and caregivers is significant in this population.

Indirect Comparisons

No appropriate comparators are available to conduct a standard indirect treatment comparison (ITC), and a placebo-controlled trial design was considered unethical by National Cancer Institute (NCI) Pediatric Oncology Branch (POB) investigators due to significant PN-related morbidity and promising results shown in the phase I trial.²¹ Indirect comparisons were therefore necessary to estimate the relative benefit of selumetinib. The NCI POB conducted 2 additional studies, a Natural History (NH) study to better understand and quantify NF1 manifestations and to allow more sensitive end points to be developed for clinical studies, and Study 01-C-0222, a randomized, crossover, double-blinded, placebo-controlled phase II study of tipifarnib in children and young adults with NF1 and progressive PN. Given the lack of direct comparative evidence for selumetinib, the sponsor conducted naive qualitative comparisons of the results from the SPRINT phase II trial, with the NH study and the placebo arm of Study 01-C-0222 serving as external control arms. The sponsor also conducted a propensity score modelling analysis of progression-free survival (PFS) compared to the NH study.

Description of Studies

The sponsor conducted a naive, side-by-side comparison of results from stratum 1 of the SPRINT phase II trial versus patients with PNs from the NH study using the outcomes of tumour growth (absolute and annual rates) based on the full NH cohort as well as an age-matched NH cohort. The “age-matched” NH cohort included patients who were aged 3 to 18 years and had at least 1 volumetric MRI within this age range and at least 1 subsequent volumetric MRI. A naive, side-by-side qualitative comparison was also conducted for the outcome of PFS between stratum 1 of the SRINT phase II trial and the placebo arm of Study 01-C-0222.

In the propensity scoring analysis, PFS from stratum 1 of the SRINT phase II trial was compared to the age-matched cohort of the NH study. Prognostic factors were identified based on data from the NH study. The univariate and multivariate Cox models (covariates: study, sex, race, target PN location, PN status, age, weight, height, and target PN volume) were fitted to estimate an unadjusted and adjusted hazard ratio (HR), respectively. Age, weight, height, and target PN volume were kept as continuous variables in the model. Three different matching algorithms were explored (matching 1:1 without replacement, inverse probability of treatment weighting [IPTW], and matching 1:2 with replacement).

Efficacy Results

Plexiform Neurofibromas Growth Rate, Naive Comparison: SPRINT Phase II Stratum 1 Versus Natural History Study

Data on the natural history of NF1-related PNs, based on the patients from the selected external controls, demonstrated that the majority of PNs grow continuously over time or, at best, remain stable in size (i.e., < 20% increase in volume from baseline). In contrast to the median annual volume changes of |||||| and ||||| seen in SPRINT (2018 and 2021 DCOs, respectively), the median annual volume changes in the NH study (an age-matched cohort with maximum follow-up aligned to each DCO of SPRINT) were |||||| and ||||||| respectively.

Over the full duration of the studies, the mean percentage change from baseline in SPRINT was |||||| compared to ||||||| in the NH study. The follow-up duration and included patients differed notably in these populations.

Patients who enrolled in the NH study and later went on to participate in stratum 1 of the SRINT phase II trial (n = ||||||) experienced PN growth before selumetinib (median = ||| per year; maximum = ||| per year), and a median volume reduction of |||||| per year after selumetinib treatment (median follow-up = ||| years; range = ||| || |||). Of these patients, |||||| had a reduction of at least 20% in their target PN and the response was sustained for |||||| patients at the latest DCO.

Progression-Free Survival, Naive Comparison: SPRINT Phase II Stratum 1 Versus NH Study

At the time of the March 32, 2021, DCO, disease progression was experienced by ||||| of patients in the NH study compared to ||||| of patients in stratum 1 of the SRINT phase II trial over a 5.6-year period. Median PFS in the NH study’s age-matched cohort was ||| years (95% CI, 1.1 to 1.6) and was ||| ||||||| in stratum 1 of the SRINT phase II trial. The probability of remaining without progression in stratum 1 of the SRINT phase II trial and the NH study was ||||| (95% CI, ||||| || |||||) and ||||| (95% CI, ||||| || |||||), respectively.

Progression-Free Survival, Naive Comparison: SPRINT Phase II Stratum 1 Versus Study 01-C-0222

Because Study 01-C-0222 required progressive disease for enrolment, a subgroup analysis was conducted for the earlier DCO (i.e., 2018) of stratum 1 of the SRINT phase II trial, including only those with progressive PNs at enrolment. In this subgroup, the probability of remaining without progression at 2 years was 94.7% (95% CI, 80.6% to 98.7%), compared to 20.6% (95% CI, 7.7% to 37.8%) in the placebo arm of Study 01-C-0222. The sponsor did not update this comparison for the 2021 DCO.

Progression-Free Survival, Propensity Scoring Analysis: SPRINT Phase II Stratum 1 Versus NH Study

The univariate Cox analysis identified age, weight, height, and PN status at baseline (i.e., progressive, nonprogressive, or unknown) as associated with PFS; younger patients with progressive PN at baseline had a higher risk of progression. The multivariate analysis identified only PN status as correlated with PFS.

After matching, the sample sizes were small, and some standardized differences remained unbalanced (> 0.1 to > 0.2) in the 1:1 and 1:2 matching analyses. In the IPTW analysis, no baseline characteristics differed by a standardized difference of more than 0.1. However, the effective sample size after IPTW was not reported. Across all 3 methods of propensity scoring analysis, the HR for PFS ranged from |||| || |||| in favour of selumetinib, with P values of less than 0.001.

Harms Results

Safety outcomes were not assessed in the ITCs.

Critical Appraisal

Because the NCI POB investigators deemed it unethical to conduct placebo-controlled trials in this population, only unanchored ITCs were possible. Unanchored naive comparisons are subject to substantial inherent limitations as there is no method of controlling for inherent differences in the study design and patient populations, and differences seen in clinical outcomes may be confounded by underlying differences in the compared trials.

In the naive comparison, results were only reported for mean annual change in target PN volume, absolute and percent change in target PN volume from baseline, and PFS. In the propensity scoring analysis, only PFS was assessed. Patient and clinician input suggests that tumour volume or change in volume does not always correlate directly with symptomatology, in part because it is highly dependent on the location of the PNs with respect to important structures. Outcomes related to symptoms, morbidity, disability, HRQoL, and disfigurement were not assessed. No safety outcomes were evaluated.

Notable differences were evident between the patient populations of the 2 external controls in comparison to the SPRINT trial with regards to baseline age, race, target PN location, PN status (i.e., progressive, nonprogressive, or unknown), target PN volume, and treatment history. The study designs also differed with respect to follow-up and frequency of imaging. The risk of bias and imprecision is inherently high due to small study sizes, observed clinical heterogeneity, and the unanchored and naive approach to the comparison, but the direction of potential bias as a result of these differences is unknown. The sample size of the before-and-after analysis (n = |) of patients participating in both the NH study and SPRINT was particularly small, limiting the interpretation of results.

Propensity scoring analysis was conducted using 3 standard methods. Although this was an appropriate approach to mitigate the impact of between-trial differences in baseline patient characteristics, it is unknown whether all key treatment effect modifiers and prognostic factors were accounted for. The methodology for selecting baseline characteristics was not explained or justified. Of the 3 methods of propensity score analysis, only IPTW demonstrated balance in every baseline characteristic examined, while in the 1:1 and 1:2 matching analyses some standardized differences were still greater than 0.1 or greater than 0.2 in important characteristics. The sample size of all analyses were small as a result of the studies informing the comparisons, but the 1:1 and 1:2 matching analyses also resulted in further drops in sample size. The effective sample size of IPTW was not reported, limiting interpretation.

Overall, interpretation of the ITCs is substantially compromised by important limitations. From the naive comparisons and the propensity scoring analyses, the results suggest selumetinib confers a benefit in terms of reduction in the rate of tumour growth and improvement in PFS. However, the magnitude of the benefit is uncertain, and no conclusions can be drawn from the ITCs regarding other clinically important outcomes or harms.

Other Relevant Evidence

No long-term extension studies or other relevant studies were included in the sponsor’s submission to CADTH.

Conclusions

There is an unmet need for disease-modifying treatment options for the rare population of patients with NF1-associated, symptomatic, inoperable PNs. Patients and clinicians highlighted the need for treatments that reduce pain and disfigurement and improve function, while also preventing the growth of new PNs and shrinking existing PNs. One ongoing, phase II, open-label, single-arm, multicentre study (SPRINT phase II) was included in this review. Notable concerns were associated with the internal and external validity of the SPRINT phase II study, driven primarily by the single-arm, open-label design, which precludes the ability to attribute the study results to treatment with selumetinib as opposed to disease natural history or concomitant interventions, and introduces significant bias to all subjective clinical outcome assessments evaluated. These are considered of critical importance given that measurement of disease progression and treatment response in clinical practice relies on available imaging techniques coupled with clinical symptomatology, which may vary from the methods and outcomes used in the SPRINT phase II trial.

The data submitted to CADTH were clinically relevant in this setting, given the variability of location and extent of PNs between patients. Clinical outcome assessments, including PROs and functional evaluations, were supportive overall of the primary imaging findings of the SPRINT phase II trial, reducing PN-associated morbidity and improving HRQoL. However, given their evaluation as secondary outcomes, small sample sizes, lack of statistical testing, and heterogeneity in the location and size of target PNs, results for PROs and functional evaluations can only be interpreted as supportive of the overall effect of selumetinib. For the primary end point in the SPRINT phase II trial, the clinical experts consulted by CADTH agreed that the ORR of 68.0% was clinically meaningful, although the clinical experts contended the observed responses were underestimated, based on their experience and the definitions used for response and progression. While selumetinib treatment also resulted in reductions in PN volume, the correlation between PN volume changes and improvements in symptoms or function remains uncertain, and the experts noted that tumour size may not always reflect morbidity. While the time-to-event end points, duration of response (DOR) and PFS, appeared to be supportive of the observed ORR, the nonrandomized design of the SPRINT phase II trial made attributing these events to selumetinib challenging.

ITCs included a naive side-by-side comparison and propensity scoring analysis against external controls as representations of natural history. The results suggest selumetinib confers a benefit in terms of reduction in the rate of tumour growth and improvement in PFS. However, due to important between-trial differences in study design and populations, and major uncertainties inherent in the methodologies applied, the magnitude of the benefit is uncertain. The relative efficacy of selumetinib was not assessed with regard to any other important clinical outcomes, such as HRQoL, morbidity, and disfigurement, which may not be directly correlated with changes in tumour volume. No safety outcomes were assessed in the indirect comparisons.

Aside from the AEs known to be associated with MEK inhibitors, selumetinib was generally well tolerated in the SPRINT phase II trial, with limited grade 3 or serious AEs, and an overall toxicity profile that can generally be managed with supportive care or dose interruptions. Although the results of the SPRINT phase II trial were generally positive, it is difficult to draw firm conclusions about the magnitude and the generalizability of the clinical benefit and safety of selumetinib given the identified limitations in the available evidence, which is inherent in the complexity of the disease and trial conduct.

Introduction

Disease Background

Of the 3 distinct forms of neurofibromatosis, the most common is NF1. An autosomal dominant genetic disorder associated with progressive cutaneous, neurologic, skeletal, and neoplastic manifestations,^1,2 NF1 affects members of each sex and each ethnic group equally.³ The incidence of NF1 in Canada is unknown, although it is estimated to occur in 1 in 2,500 to 3,000 births.^2,4-6 The patient group input received by CADTH for this review estimated that there are currently over 12,000 cases of NF1 in Canada.

The disease is caused by germline mutations in the NF1 tumour suppressor gene (17q11.2), which encodes the tumour suppressor protein neurofibromin 1. An activating protein that promotes the conversion of active RAS GTP proteins to inactive RAS guanosine 5′-diphosphate, neurofibromin 1 is a negative regulator of the RAS proto-oncogene, a key signalling molecule controlling cell growth, resulting in overactivation of the RAS/RAF/MEK/ERK mitogen-activated protein kinase cascade pathway, leading to abnormal cell growth.^22-24 Approximately half of all NF1 cases are familial, while half arise from spontaneous mutations in the NF1 gene.^3,4

Neurofibromatosis type 1 is a highly heterogenous disease with hallmarks and clinical features that may be evident from birth and can affect a wide range of organ systems. The most common manifestations of NF1 include abnormally coloured patches of skin (CALMs), freckling under the arms and in the inguinal region, and benign tumours predominantly in the skin and nerves, known as neurofibromas. Other possible manifestations include bone dysplasia, scoliosis, ocular problems, and neurologic complications with impacts such as cognitive impairments and learning disabilities. Neurofibromas are histologically benign nerve sheath tumours, typically originating in the terminal nerve branches of the skin. PNs are the most common type of tumour in patients with NF1, occurring in up to 50% of patients.^7-10 One or multiple PNs may grow along large nerves and plexuses anywhere in the body, with varying manifestations continuing to develop to early adulthood, and multiple PNs may be both symptomatic and asymptomatic in the same individual.^11-13 PNs have a complex shape and can reach large sizes, resulting in clinical symptoms such as disfigurement, motor dysfunction (weakness and restricted ROM), pain, and neurologic dysfunction. The severity of symptoms from PNs range from mild to severe; however, the presence of symptoms may depend on their location and impact on surrounding structures. PNs grow most rapidly during early childhood, although growth rate is highly variable between patients.^14-16 In children, rapid PN growth cannot be attributed to the anticipated growth rate of a child, as the PN growth rate does not correlate with increases in body weight or body mass index.^14,16 Spontaneous shrinkage of PNs over time has been reported, but mainly in adults.^13,15,25

PNs have the potential for malignant transformation.^23,26 An MPNST is a type of cancer that forms in the cells of the sheath that covers and protects the peripheral nerves. The risk of developing MPNSTs is greater in patients with NF1, with 1 study citing an incidence of MPNSTs in patients with NF1 of 4.6% compared to 0.001% in the general population,²⁷ and the lifetime risk estimated to be between 8% and 15.8%.^28-30 Other tumours associated with NF1 include low-grade gliomas, with optic pathway gliomas occurring in approximately 15% of NF1 patients.²³

Diagnostic criteria for NF1 were established in 1987 and updated most recently in 2021.³¹ Diagnostic criteria are based predominantly on clinical manifestations, although the revised criteria incorporate additional clinical features and genetic testing. A diagnosis of NF1 in patients who do not have a parent diagnosed with NF1 requires the presence of 2 or more of the following:

• 6 or more CALMs larger than 5 mm (by greatest diameter) in prepubertal individuals and 15 mm in postpubertal individuals

• freckling in the axillary or inguinal region

• 2 or more neurofibromas of any type or 1 PN

• optic pathway glioma

• 2 or more iris Lisch nodules identified by slit lamp examination or 2 or more choroidal abnormalities, defined as bright, patchy nodules imaged by optical coherence tomography/near-infrared reflectance imaging

• a distinctive osseous lesion such as sphenoid dysplasia, anterolateral bowing of the tibia, or pseudarthrosis of a long bone

• a heterozygous pathogenic NF1 variant with a variant allele fraction of 50% in apparently normal tissue such as white blood cells.³¹

A child of a parent who meets the diagnostic criteria merits a diagnosis of NF1 if 1 or more of these criteria are present.

Standards of Therapy

Treatment and clinical management options for NF1-associated PNs are extremely limited and depend on symptomatology. Treatment strategies consist of “watch and wait” with frequent monitoring for patients with asymptomatic PN, which generally does not require treatment. For symptomatic patients, treatments aim to relieve symptoms caused by the individual PNs. Currently, the only available options to treat and manage NF1 include pain management and surgical excision to remove as much of the tumours as possible. However, for many patients, surgery is not a viable option as most PNs are not amenable to complete resection due to encasement of, or proximity to, vital structures.^7,17,18 Several phase I and II clinical studies for progressive PN in children and young adults with NF1 have failed to demonstrate consistent or durable decreases in PN volume.^13,32

Drug

Selumetinib (Koselugo) is an orally available, selective inhibitor of MEK 1 and 2. Selumetinib blocks MEK activity and inhibits growth of RAF-MEK-ERK pathway–activated cell lines, leading to inhibition of cellular proliferation and PN growth.³³

Selumetinib is available as 10 mg or 25 mg oral capsules. The recommended dosage of selumetinib is 25 mg/m² twice daily. Recommended dosages are summarized in Table 3. Dosing is individualized based on BSA (mg/m²) and rounded to the nearest achievable 5 mg or 10 mg dose (up to a maximum single dose of 50 mg). Selumetinib capsules of different strengths may be combined to attain the desired dose.³³

Table 3: Recommended Dosage of Selumetinib Based on Body Surface Area

Source: Selumetinib product monograph.

Treatment with selumetinib should continue as long as a clinical benefit is observed, or until PN progression or unacceptable toxicity.³³

The reimbursement request for selumetinib is in line with the approved Health Canada indication: for the treatment of pediatric patients aged 2 years and older with NF1 who have symptomatic, inoperable PNs. Health Canada granted a Notice of Compliance on August 31, 2022.³⁴ Selumetinib has not previously been reviewed by CADTH. Selumetinib has also been approved by other major regulatory bodies including the US FDA (2020), European Medicines Agency (2021), and Australia’s Therapeutic Goods Administration (2021).

To support patients requiring treatment with selumetinib, the sponsor indicated that the Alexion OneSource Patient Support Program provides necessary resources and support by phone and/or email across Canada to patients and caregivers on follow-up touch points for disease education, appointment reminders, and/or adherence support, as well as reimbursement navigation.³⁴

Stakeholder Perspectives

Patient Group Input

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

CADTH received input from 2 patient groups: the TFBC and CORD. The TFBC provides essential information and support services to patients with neurofibromatosis and their families. CORD works with governments, researchers, clinicians, and industry to promote research, diagnosis, treatment, and services for all rare disorders in Canada.

Both patient groups conducted online surveys in November 2022, recruiting patients with NF1 and their caregivers. Additionally, the TFBC conducted a Zoom focus group. All patients had a diagnosis of NF1. The TFBC group recruited 25 patients and caregivers, and CORD recruited 8 caregivers. A total of 8 patients (32%) included in the TFBC survey were adults, while all patients represented in the CORD survey were younger than 18. Key themes identified by patients and caregivers with NF1 included limitations on daily living, functional, and social activities; moderate to severe chronic pain; dependency on caregivers into adulthood; financial stress because of the diagnosis; and lack of treatment options, which negatively affects the emotional well-being of patients and families.

Respondents from both patient and caregiver groups described difficulties obtaining a diagnosis of NF1, as well as significant impacts on both affected children and their families in terms of managing physical and mental disability, with 96% of TFBC survey respondents indicating that they live with chronic pain rated at 5 or greater on a 0-to-10 pain scale. Additionally, substantial negative mental health impacts were reported, with most patients living with anxiety and fear over their diagnosis, and some patients experiencing suicidal feelings or actions. Respondents also cited the financial burden of out-of-pocket expenses and time lost from work and school, the negative toll of multiple surgeries, and hospitalizations with limited benefits. Respondents who had previous treatment experience described out-of-pocket expenses for psychological and physical supports (e.g., scoliosis braces) and treatment or diagnosis options accessed in the US. Some respondents had experience with repeated surgeries that provided minimal or temporary improvement in key outcomes with substantial recovery time. A total of 92% of TFBC respondents reported expenses related to the care of their neurofibromatosis (such as prescription or nonprescription drugs, medical equipment, physiotherapy, counselling, or travel for medical care), and 40% indicated that they fund their own medical expenses without any public or private benefits. Respondents to both surveys and interviews indicated they were surprised and disappointed by the lack of treatment options or support available, with 46% not having been offered any kind of treatment, and only 17% of patients who were offered treatment experiencing minimal improvement in symptoms.

Numerous outcomes were identified as important to patients, reflecting the heterogenous nature of the disease, but common themes included an overall improved QoL, a desire for reduction in pain and reduction or prevention in tumour size or growth, improved function and emotional well-being, greater independence from caregivers, and a reduced number of health care visits.

No patients in the TFBC survey had experience with selumetinib, however, all of them indicated that they would consider taking selumetinib if given the opportunity to access it. The half (n = 4) of the CORD respondents who had experience with selumetinib through clinical trials described it as “miracle drug” that was “life-changing” due to substantial improvements in pain level, functional abilities including speaking clearly and chewing food, and softening and shrinking tumours that were previously disabling and/or disfiguring. The respondents described either no side effects or mild side effects, such as sensitive skin around the toenails that can be easily infected. Although interviews conducted by CORD pointed out that the long-term benefits of selumetinib have yet to be demonstrated, the patient organizations felt that selumetinib should be available to all appropriate patients.

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 selumetinib review, a panel of 6 clinical experts from across Canada was convened to characterize unmet therapeutic needs, help identify and communicate situations in which gaps in the evidence could be addressed through the collection of additional data, promote the early identification of potential implementation challenges, acquire further insights 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

The clinical experts indicated that NF1 is a highly heterogeneous, multisystem, genetic condition manifesting from early childhood with numerous needs that are currently not being met. In patients with NF1 with symptomatic, inoperable (i.e., that cannot be completely removed surgically) PNs, tumours are often located throughout the body, may cause pain or discomfort, and result in impairment in mobility, vision, breathing, or other functions. Moreover, the experts also highlighted that, in terms of natural history, there is a trend for tumours to appear and grow rapidly in early childhood, particularly before ages of 6 to 8 years, and then for growth to become much slower or stop in adulthood.

The experts agreed that the main limitations and unmet needs of pediatric patients with NF1 with and symptomatic PNs that cannot be completely removed surgically is the lack of access to disease-modifying medical interventions that can reduce the burden of disease or stabilize symptomatic PNs. Given the multisystemic, heterogenous nature of the disease, the expert panel emphasized the need for systemic treatment of multiple PNs that cause issues in this population.

Place in Therapy

The clinical experts noted that there are currently no established practice guidelines for this heterogenous disease. Recently, a panel of multidisciplinary NF1 experts published consensus recommendations for the management of NF1-associated PNs (Fisher et al. [2022]). For patients with symptomatic PNs, surgery is the only available treatment option, which is either aimed at excising tumours if possible, or debulking if complete excision is not achievable. The experts noted that surgery is not curative for most large or extensive PNs and is associated with significant risks of secondary injuries based on many factors, including the number, location, size, and vascularity of tumours, particularly for PNs involving large arteries or nerves. Additionally, the experts pointed out that multiple invasive surgeries may be required, as tumours can regrow or increase in size, adding to the risks to patients.

Aside from surgery, current treatment strategies consist of “watch and wait” for patients with PNs who are not currently symptomatic. Otherwise, treatment for patients with symptomatic PNs focuses on relieving pain, reducing functional impairment, and improving overall QoL. The panel emphasized the availability of MEK inhibitors through managed access programs, noting that selumetinib is the first, and only Health Canada–authorized MEK inhibitor available outside of clinical trials for the treatment of PNs. The panel concluded that selumetinib is expected to cause a shift in the current treatment paradigm, given the absence of other medications available for this population. The experts stated that, should selumetinib be recommended for reimbursement in this population, it would likely be the initial therapy of choice, and there is no evidence supporting the use of other treatments before initiation of selumetinib for PNs that cannot be completely excised surgically in children with NF1.

Patient Population

The clinical experts noted that only a small minority of NF1 patients have symptomatic PNs. Patients with NF1 are diagnosed based on standard, well-established, and recently updated clinical diagnostic criteria, including clinical characteristics such as CALMs, and the presence of neurofibromas. Although the recently updated diagnostic criteria include genetic testing, the experts noted that genetic testing is generally not required for a diagnosis of NF1. The availability of genetic testing varies by province, and it may not be funded through provincial health plans for many patients. The experts also noted that the results of genetic tests do not affect treatment decisions once a clinical diagnosis of NF1 has been established.

The diagnosis of NF1 is relatively straightforward in older children, adolescents, and adults, but can be challenging in younger children. However, the updated diagnostic criteria, which include genetic testing in patients without a family history, have improved confirmatory diagnosis in young children before they manifest other clinical features of NF1. The diagnosis of large, extensive, or rapidly growing PNs generally requires more clinical expertise, and may require more complex tumour characterization, including MRI, and sometimes a biopsy if there is concern for malignant transformation. The experts added that, in terms of natural history, there is a trend for tumours to appear and grow rapidly in early childhood (before the age of 6 to 8 years), and then slow down or remain static in adulthood. Rapid growth of a PN is a concern for transformation to a MPNST, which can be metastatic and often fatal despite treatment. The experts also discussed the uncertainty regarding treatment decisions for asymptomatic patients, which have not been established. In addition, there is no evidence available regarding whether treatment with MEK inhibitors such as selumetinib can prevent growth of new PNs.

The experts emphasized the heterogeneity of the disease in NF1 patients, with cutaneous neurofibromas and PNs often occurring throughout the body and ranging in severity from asymptomatic to severely debilitating due to pain, functional impairment, or disfigurement. One expert noted that disfigurement due to large, visible PNs can be a source of anxiety and concern due to public fear and social stigmatization. The panellist also highlighted the potential for continuous problems into adulthood due to large PNs, which may result in severe disfigurement and displacement of joints and bones, but added that there is no clear evidence that treating asymptomatic PNs in children will prevent the development of symptoms in adults. Other concerns raised by the experts for the NF1 population include deficient social skills, frequent learning disabilities, autism, and attention-deficit/hyperactivity disorder, further highlighting the vulnerability and marginalization of these patients.

According to the experts consulted for the review, the patients who are most likely to benefit from treatment with selumetinib are those whose extensive inoperable PNs are causing significant pain, functional impairment, and/or disfigurement. Although it is difficult to determine which patients are most likely to respond to treatment, 1 expert currently treating pediatric patients via compassionate access to selumetinib stated that about 80% of patients will respond to treatment.

Regarding patients least suitable for treatment, the experts noted that most NF1 patients with PNs are asymptomatic, and the benefit of treatment for these patients has yet to be established. Almost all PNs persist throughout childhood and into adulthood, and the panellists noted that other MEK inhibitors may turn out to be just as effective as selumetinib in both children and adults. Clinical trials in the adult population currently being conducted for selumetinib should provide insight into similarities or differences in effectiveness by age. The experts agreed that the lack of knowledge about both the potential benefits and harms associated with long-term selumetinib treatment is a concern, given that NF1 is a life-long disease. The experts also noted that the life expectancy of NF1 patients has been reported to be reduced by 10 to 15 years, although estimates of life expectancy with currently available medical management are unknown.

Assessing Response to Treatment

The clinical experts noted current clinical trials aim to address important outcomes that are used in clinical practice. However, given the heterogeneity of the disease, 1 expert highlighted that standardizing subjective measures (such as pain perception) across this population is an issue; thus, interpreting the results relies heavily on clinical judgment.

The clinical experts agreed that the most important outcomes in the management of pediatric patients with NF1 and symptomatic, inoperable PNs is the reduction or improvement in symptoms (i.e., reduced pain, improved function), as well as overall improvements in QoL and disease stabilization due to the potential for rapid, progressive growth of the tumours. One expert added that, due to disfigurement, cosmetic improvements are also likely important; however, the panel noted that current surgical management for cosmetic removal of most neurofibromas is not funded by provinces, except in special cases.

The clinical experts noted that volumetric MRI measurements are used in clinical trials to define disease progression; however, volumetric MRI is only used by the NIH for research studies, and is not available in Canadian clinical practice. The experts considered a change in planar tumour size of 20% to 25% to be indicative of response to treatment. One expert discussed the potential for symptomatic disease progression despite the lack of imaging evidence of progression and improvement in symptoms without reduction of tumour size. In addition, the experts emphasized that it is not always clear which tumours are the cause of symptoms when patients have large numbers of PNs, making it difficult to know when the disease is progressing. The experts also highlighted that tumours are frequently irregular in shape, making measurements about changes in tumour size difficult. As a result, the panel noted that response to treatment is multidimensional and must consider reductions in tumour sizes, changes in symptoms, and improvements in function and disfigurement. One clinical expert also emphasized the additional challenges of conducting MRI on young children. This often requires an anesthetic and can pose difficulties when determining clinically important growth based on child size compared to adolescent or adult size.

The experts stated that young children with NF1 and symptomatic PNs may initially be followed with MRI imaging every 3 months, in addition to annual follow ups with NF1 specialists to assess other features of the disease. Upon initiating treatment, the experts stated that patients would usually be seen weekly for a month, then every month, and if treatment is well tolerated, or disease stabilizes, then follow ups would be prolonged to every 6 months. The clinical experts also noted that imaging in young children often requires a general anesthetic.

Discontinuing Treatment

When deciding to discontinue treatment, the clinical experts agreed that treatment would be discontinued in patients who are not responding (as indicated by tumour growth, lack of stabilization, or lack of improvement of symptoms), or in patients with severe AEs that cannot be managed. The experts identified the risk of cardiotoxicity and retinal issues associated with the use of selumetinib as examples of significant AEs that would result in discontinuation. The experts also noted that the need for surgery to further debulk tumours could indicate that the treatment is not working and should therefore be discontinued. However, 1 clinical expert pointed out that selumetinib may be used in conjunction with debulking surgery, although the available evidence does not support this approach.

The experts noted that no firm treatment duration or date of discontinuation for selumetinib has been determined, but suggested that, similar to the SPRINT phase II trial, in clinical practice, patients would continue treatment until disease progression or toxicity. The experts agreed that the initial treatment authorization period for selumetinib should be 18 months.

Given the approved Health Canada indication for use in patients aged 2 to 18 years, the clinical experts discussed the use of selumetinib in patients aged 18 or older. Currently, 1 clinical expert has been providing selumetinib and other MEK inhibitor treatments to adults with PNs on an off-label basis through a philanthropic clinic and compassionate access to treatment. As a result, the clinical experts felt that patients may continue to receive benefits beyond the age of 18 years, although this requires further study.

Prescribing Conditions

The experts indicated that expertise in Canada in using selumetinib is sparse and limited to pediatric oncologists and neurooncologists in tertiary care hospitals, although the clinical setting is still evolving. The experts reported that selumetinib should be restricted to use in patients who are being followed in specialized centres, and that the usual approach to disease management begins with discussions with families about natural history, and shared decision-making in creating individualized treatment approaches.

According to the clinical experts, in Canada, only pediatric oncologists are prescribing selumetinib, as they have the experience and know-how to manage these patients. However, the experts noted that, with further insight and growing experience, NF1 experts who are pediatricians could likely continue to manage this oral treatment. Given the heterogeneity in the disease and the individualized approach to treatment, decisions often involve a multidisciplinary team of pediatricians, NF1 experts, neurooncologists, and nurse practitioners. The experts emphasized the importance of consulting other specialists, including surgeons, cardiologists, ophthalmologists, dermatologists, and pharmacists, to manage adverse effects and drug interactions.

The expert panel added that access to specialty clinics may be a limiting factor for patients in remote areas, emphasizing that patients would be required to attend in-person appointments for treatment initiation and imaging follow-up, as well as to assess safety. However, the experts considered the potential for remote monitoring, local bloodwork, or eye exams to be acceptable.

Clinician Group Input

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

Input for this review was received input from 1 clinician group, the CPBTC, which included 27 pediatric neurooncologists across Canada representing 16 children’s hospital who provide oncological care for children with neoplasms of the central nervous system and peripheral nerves. Information was gathered through shared clinical experiences of the included clinicians.

Overall, the clinician group input aligned with that given by the clinical expert panel convened by CADTH. It pointed out that no systemic therapies exist for treating NF1-associated PNs, which represent the major unmet need in this patient population, with surgical resection, if feasible, the only option currently available for patients. The clinician group emphasized that selumetinib has clearly shifted the current treatment paradigm and emerged as the standard of care as first-line therapy for patients with inoperable, symptomatic PNs. In terms of patients most in need of intervention, the clinician group agreed with the clinical expert panel that these patients include those in whom PNs are invading critical structures, causing a deformity, or causing functional impairment in activities of daily living such as walking, swallowing, or eating, although the clinician group noted it is impossible to determine which patients will and will not respond.

The clinician group also noted that, in Canada, treatment initiation with selumetinib is currently limited to pediatric oncologists, neurooncologists, or pediatric neurologists with an expertise in neurooncology. However, the CPBTC noted that general neurologists and pediatricians in jurisdictions outside of Canada have been able to prescribe selumetinib, and suggested that this practice would improve access for patients who live in remote areas. The CPBTC suggested that treatment with selumetinib in Canada could be initiated by oncologists and followed remotely in conjunction with local clinicians.

Finally, the CPBTC noted that many parents of children with NF1 also have NF1 themselves and are likely to have lower socioeconomic standing in part because of the disease. It was therefore the CPBTC’s opinion that many patients and parents of patients are more likely to lack private insurance to cover selumetinib, which could result in unequitable access in some parts of the country. The CPBTC group emphasized that children without private insurance who are also not eligible for the provincial public drug plans will need special consideration and that the drug in question urgently requires reimbursement and equitable access as a standard-of-care treatment for patients with NF1 and symptomatic PNs.

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 4.

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

AE = adverse event; b.i.d. = twice daily; CALM = café-au-lait macule; CDEC = CADTH Canadian Drug Expert Committee; MEK = mitogen-activated protein kinase; MPNST = malignant peripheral nerve sheath tumour; NF1 = neurofibromatosis type 1; NF2 = neurofibromatosis type 2; PN = plexiform neurofibroma; TPGS = alpha-tocopherol polyethylene glycol succinate.

Clinical Evidence

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

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of selumetinib 10 mg and 25 mg twice daily for the treatment of pediatric patients aged 2 years and above with NF1 who have symptomatic, inoperable PNs.

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 listed in Table 5. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Table 5: Inclusion Criteria for the Systematic Review

NF1 = neurofibromatosis type 1; PN = plexiform neurofibroma; vs. = versus.

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.³⁵

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) via Ovid and Embase (1974–) via Ovid. All Ovid searches were run simultaneously as a multifile search. Duplicates were removed using Ovid deduplication for multifile searches, followed by manual deduplication in EndNote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. Three component searches were conducted. The first used the search concepts Koselugo and NF1. The second search used the concepts of Koselugo and pediatrics. A third search was conducted on Koselugo and CADTH-developed search filters were applied to limit retrieval to randomized controlled trials or controlled clinical trials. Clinical trials registries searched included 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.

Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. Appendix 1 provides 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 March 22, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.³⁶ Included in this search were the websites of regulatory agencies (US FDA and European Medicines Agency). Google was used to search for additional internet-based materials. Appendix 1 provides more information on the grey literature search strategy.

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 include in the review, and differences were resolved through discussion.

Findings From the Literature

One study from the literature was selected for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 6. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 6: Details of Included Studies

CPAP = continuous positive airway pressure; CR = complete response; CTCAE v4 = Common Terminology Criteria for Adverse Events Version 4; DCO = data cut-off; IOP = intraocular pressure; MEK = mitogen-activated protein kinase; MPNST = malignant peripheral nerve sheath tumour; NA = not applicable; NCI = National Cancer Institute; NF1 = neurofibromatosis type 1; NR = not reported; NYHA = New York Heart Association; PN = plexiform neurofibroma; POB = Pediatric Oncology Branch; PR = partial response; PROMIS = Patient-Reported Outcomes Measurement Information System; QTc = corrected QT interval; QTcF = QT interval corrected using Fridericia’s correction formula; ULN = upper limit of normal.

Source: SPRINT Phase II Stratum I Clinical Study Report (June 29, 2018, DCO).¹⁹

Description of Studies

The SPRINT phase II study is a phase II, open-label, single-arm, multicentre study that evaluated the efficacy of selumetinib in pediatric patients with NF1 and inoperable PN. The primary outcome was the ORR (confirmed PR and complete response [CR]) determined by change in PN volumes through volumetric MRI.¹⁹

For the SPRINT phase II trial, patients were enrolled in 1 of 2 strata based on their PN-related morbidity as determined by the clinical team at the time of study entry:¹⁹

• stratum 1: PN-related morbidity present at enrolment

• stratum 2: no significant PN-related morbidity present at enrolment, but potential for development of PN morbidity.

Based on the reimbursement request and approved Health Canada indication, only results for stratum 1 are summarized in this report, which includes only patients with symptomatic PNs. A total of 50 patients were enrolled into stratum 1 and received 25 mg/m² selumetinib twice daily on a continuous 28-day cycle.¹⁹

Two DCOs were submitted for the SPRINT phase II trial.^19,20 The primary DCO occurred on June 29, 2018, and an updated DCO occurred on March 31, 2021, providing a further 2 years and 9 months of follow-up, 4 years and 7 months after the last patient enrolled in stratum 1 of the SRINT phase II trial, for a maximum follow-up of 5.6 years.

The SPRINT phase II study was sponsored by the Cancer Therapy Evaluation Program, conducted by the NCI POB at 4 study centres in the US, and supported by AstraZeneca through a cooperative research and development agreement.¹⁹ No Canadian investigative sites were included,¹⁹ and it was unclear if any patients from Canada were enrolled.

Populations

Inclusion and Exclusion Criteria

The inclusion and exclusion criteria for stratum 1 of the SPRINT phase II trial are summarized in Table 6. Briefly, eligible patients were those between the ages of 2 and 18 years with a diagnosis of NF1 and inoperable PNs, defined as those that could not be completely surgically excised without risk substantial morbidity due to proximity to vital structures, invasiveness, or high vascularity. Patients in stratum 1 were those with at least 1 clinically relevant PN-related morbidity as determined by the clinical team at the time of study entry, which was defined as (but not limited to) head and neck PN that could have compromised the airway or great vessels, paraspinal PN that could have caused myelopathy, brachial or lumbar plexus PN that could have caused nerve compression and loss of function, PN that could have resulted in major deformity (e.g., orbital PN) or were significantly disfiguring, PN of the extremity that could have caused limb hypertrophy or loss of function, and painful PNs. Volumetric MRI was used to classify patients’ PNs as typical (nodular component of PN was < 30%), nodular (nodular component of PN was ≥ 30%), or solitary nodular (target PN was a single nodular PN). Patients with volumetric MRI scans within 3 years before enrolment were to have these submitted to the NCI POB to estimate the target PN rate of growth status at enrolment (progressive, nonprogressive, or unknown). Patients were also required to be able to swallow whole capsules.¹⁹

Baseline Characteristics

Baseline characteristics for patients enrolled in stratum 1 of the SPRINT phase II are summarized in Table 7. The mean age of patients was 10.3 years (SD = 3.92), ranging from 3.5 to 17.4 years of age. Most patients were white (42 [84.0%]) and male (30 [60.0%]). The mean time between diagnosis of NF1 to start of selumetinib treatment was 9.03 years (range = 2.0 to 16.5 years), and the mean time between diagnosis of PN to start of selumetinib treatment was 7.55 years (range = 0.7 to 16.5 years). The target PN was chosen at study entry by the investigator as the most clinically relevant PN causing morbidity at baseline. Most target PNs were classified as typical (45 [90.0%]), and the mean target PN volume was 837.11 mL (SD = 925.011), with values ranging from 5.6 to 3,820.0 mL. The median number of morbidities arising from each target PN was 3 (range = 1 to 4), with || patients having | morbidities. Disfigurement as a result of PNs was observed in 44 patients (88%), with 33 patients (66.0%) also having motor dysfunction. The most common location of target PNs was the neck and trunk, and the trunk and extremity (12 [24.0%], each). At baseline, 21 patients (42.0%) and 15 patients (30.0%) were classified as having progressive and nonprogressive PN status, respectively, while 14 patients (28.0%) had unknown PN status.^19,20

Forty-seven patients (94.0%) had at least 1 disease-related symptom at baseline. The most commonly reported symptoms at baseline were muscular weakness (34.0%), limb asymmetry (32.0%), tumour pain (32.0%), joint ROM decreased (26.0%) and scoliosis (26.0%). In total, || ||||||| of patients had previous disease-related treatment modalities, primarily interferons and imatinib, and || ||||||| patients had at least 1 prior PN- or NF1-related surgical procedure.^19,20

Table 7: Summary of Baseline Characteristics (Full Analysis Set)

DCO = data cut-off; NF1 = neurofibromatosis type 1; PN = plexiform neurofibroma; SD = standard deviation.

^aLansky performance status was assessed in patients who were aged 16 years or younger and Karnofsky performance status was assessed in patients who were older than 16. A total of 47 patients were less than 16 years, while only 3 patients were 16 or older.

^bClassification based on imaging.

^cThe “other dysfunction” category included patients with PN pain, swallowing, disfigurement and sensory neuropathy. Patients with PN pain were captured under the pain morbidity category (i.e., pain present = yes) and sometimes also captured in the “other dysfunction” field.

^dDoes not include symptoms recorded as “other.” Pain was included as a PN-related morbidity.

^eMedical therapy directed at other NF1 tumours may also be included.

^fSurgery included 4 patients who had 1 or 2 biopsies only and no other prior PN-related surgical procedures. Other NF1-related surgeries may also be included.

Source: SPRINT Phase II Stratum I Clinical Study Report (June 29, 2018, DCO).¹⁹

Interventions

The recommended phase II dose for selumetinib was determined during the SPRINT phase I trial. In the SPRINT phase II trial, selumetinib was administered orally at a dosage of 25 mg/m² twice daily (approximately every 12 hours) based on BSA, continuously for 28-day cycles with no rest periods between cycles. Selumetinib was supplied as 10 mg (white) and 25 mg (blue) capsules. Patients were instructed to take the dose of selumetinib on an empty stomach (no food or drink other than water for 2 hours before and 1 hour after dosing) with water only. The capsules were not to be crushed or broken and were to be swallowed whole. Doses were rounded to the nearest 5 to 10 mg using a dosing nomogram. Selumetinib dosing was capped at 50 mg when BSA was 1.9 m² or greater. Selumetinib dosing was adjusted for changes in BSA according to the dosing nomogram.¹⁹

The duration of treatment with selumetinib was determined to ensure that patients had the opportunity to derive a benefit but was limited based on the following:¹⁹

• For patients with documented disease progression within approximately 1.5 years before study entry (defined as ≥ 20% increase in the size of PN or ≥ 13% increase in the product of the longest 2 perpendicular diameters, or ≥ 6% increase in the longest diameter), there was no limit to the duration of treatment if the patient met the requirements for further treatment.

• For patients with no previous documented history of disease progression within the 1.5 years before study entry, the duration of the study was to be limited to 2 years if no imaging response (i.e., volume decreased by ≥ 20%) was observed.

  • Patients who were removed from treatment after 2 years for reasons other than toxicity or progression and who had stable disease were continued to be monitored with a volumetric MRI analysis every 4 to 6 months. If the PN demonstrated some growth (volume increase ≥ 15%) within approximately 2 years of stopping selumetinib, treatment with selumetinib could be restarted with the goal of stopping further PN growth. Patients were re-consented on the study and had to meet all eligibility criteria with the exception of prior treatment with selumetinib or another specific MEK 1 or 2 inhibitor before restarting therapy. In these patients, treatment could continue as long as the PN remained stable or responsive (< 20% increase in the PN volume).

  • For patients who showed an imaging response, the treatment duration was not to be limited unless the patient experienced subsequent disease progression or met other off-treatment criteria. In all cases, treatment could be discontinued earlier at the discretion of the institutional principle investigator if this was considered to be in the best interest of the patient.

In all cases, treatment could be discontinued earlier at the discretion of the institutional principle investigator if this was considered to be in the best interest of the patient. Patients were to be followed up for 5 years after completing selumetinib, or for a total duration of 7 years, whichever was longer.¹⁹

Dose Modifications, Reductions, or Interruptions

Selumetinib was to be withheld in patients with toxicities requiring dose modification. Selumetinib doses held while recovering from toxicity were not made up and the cycle remained 28 days or until recovery from toxicity. If the toxicity resolved to meet study parameters or grade 1 or higher on the Common Terminology Criteria for Adverse Events (CTCAE) Version 4 within 21 days of drug interruption, selumetinib was resumed at a dose reduced by 25% to 33%. Doses reduced for toxicity were not re-escalated, with the exception that the dose could be increased to account for an increase in BSA at the time of on-study evaluations (dosing was rounded to the nearest 5 to 10 mg). If toxicity did not resolve to meet study parameters within 21 days, the patient was removed from selumetinib treatment, except in the event that the patient was receiving clear clinical benefit and recovery occurred within 3 months of discontinuation.¹⁹

If target PN volume measurements subsequent to a dose reduction demonstrated an increase of 10% or greater from best response in target PN volume, but had not met criteria for progressive disease, and selumetinib was well tolerated at this reduced dose, dosing could be resumed at the dose level before the dose reduction. In this instance, selumetinib was administered twice daily every 5 days, followed by 2 days of rest.¹⁹

If toxicity recurred in any patient who had resumed treatment at the reduced dose, the dose could be reduced a second time using the same criteria, with the exception of cardiotoxicity, for which only 1 dose reduction was allowed. If toxicity recurred after 2 dose reductions, the patient was removed permanently from selumetinib. Patients removed from selumetinib for toxicity were followed until resolution of toxicity and off-study criteria were met.¹⁹

Treatment Discontinuation and Withdrawal From Study

Patients in the SPRINT phase II trial could discontinue treatment and assessments at any time, or at the discretion of the investigator. Patients could be discontinued from selumetinib in the following situations:¹⁹

• nonmedical or administrative reasons:

  • patient refusal

  • investigator decision in the interest of the patient

  • serious protocol violation as determined by the investigator

  • noncompliance with study requirements

• toxicity requiring dose modification, which did not resolve per study parameters within 21 days of selumetinib interruption, except when the patient had shown clear clinical benefit

• patients with clinical or imaging evidence of progressive disease on treatment following any treatment cycle

• patients who developed a concurrent serious medical condition that could have precluded or contraindicated further administration of selumetinib

• pregnancy was to result in immediate removal from selumetinib

• patients who underwent complete surgical resection of their PN (thus rendering them with no evidence of disease)

• other reasons:

  • death

  • lost to follow-up

  • withdrawal of consent for any further data submission

  • completion of long-term safety evaluations

  • screen failure.

Concomitant Medications

Restricted, prohibited, and permitted concomitant medications in the SPRINT phase II study included:

• Corticosteroids were allowed for control of symptoms related to the underlying NF1 or for other reasons.

• Throughout the study, patients were instructed to avoid changes to, or the addition of concomitant medications. Unless considered clinically indicated, patients were to avoid taking other additional nonstudy medications that could interfere with selumetinib. Patients were to avoid medications that are known to either induce or inhibit the activity of hepatic microsomal isoenzymes CYP1A2, CYP2C19, and CYP3A4, as this may interfere with the metabolism of selumetinib.

• Appropriate antibiotics, blood product support, anti-emetics and general supportive care were to be used as indicated.

• Selumetinib was to be administered with caution in patients who were also receiving concomitant coumarin anticoagulant medications (e.g., warfarin). These patients were to have their international normalized ratio monitored or anticoagulant assessments conducted more frequently, and the dose of the anticoagulant adjusted accordingly.

• No other cancer chemotherapy, radiation, immunotherapy, biologic therapy, hematopoietic growth factors, or investigational agents was permitted while receiving selumetinib.

• Use of supplemental vitamin E, including any multivitamin containing vitamin E, was also restricted.¹⁹

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 and summarized in the following section. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 4.

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

6MWT = 6-minute walk test; DVQ = Dysfunctional Voiding Questionnaire; PFT = pulmonary function test.

Primary Efficacy End Point

The primary efficacy end point of the SPRINT phase II study was the ORR, defined as the percentage of patients with a CR or confirmed PR (defined as target PN decrease ≥ 20% compared with baseline; confirmed when observed again within 3 to 6 months). The ORR was based on the NCI POB central analysis (i.e., the primary analysis) of volumetric MRI of the target PN. Response was assessed before cycles 5, 9, 13, 17, 21, 25 and then after every 6 cycles (precycles 31, 37, and 43) until discontinuation of selumetinib.¹⁹ Response was measured using the REiNS criteria. Definitions of response per REiNS criteria were:³⁷

• CR: disappearance of the target PN

• PR: decrease in the volume of the target PN by 20% or more compared with baseline; the PR was considered unconfirmed at the first detection and confirmed when observed again within 3 to 6 months

• Stable disease: insufficient volume change to qualify for either PR or progressive disease

• Progressive disease: increase in the volume of the target PN by 20% or more compared with baseline or the time of best response after documenting a PR; the appearance of new PN (with the exception of new discrete subcutaneous neurofibromas) or unequivocal progression of existing clinically relevant nontarget PN was also considered progressive disease; the clinical appearance of new discrete subcutaneous neurofibromas did not qualify for disease progression.

Additional assessments of ORR included best objective response (BOR), defined as the best response recorded from the start of treatment until progression or the last evaluable volumetric MRI assessment in the absence of progression, which was summarized by category (confirmed PR, unconfirmed PR, stable disease, progressive disease or not evaluable [NE]), and ORR by PN status at enrolment (progressive, nonprogressive, or unknown).¹⁹

Secondary Efficacy End Points

Secondary efficacy outcomes of the SPRINT phase II trial included DOR, PFS, and time to response (TTR), and were based on the NCI POB central analyses of volumetric MRIs of target PN.¹⁹

Other secondary end points included clinical outcome assessments or PROs to evaluate pain, motor function, bowel and bladder function, HRQoL, and functional measures for pain, respiratory function, visual function, physical function, and changes in pain intensity or other morbidities.¹⁹ A summary of PROs, and respondents for specific measures is provided in Table 9, with a detailed discussion of these outcomes supplied in Appendix 4.

Duration of Response

Only patients who had a CR or confirmed PR were included in the analysis of DOR. The DOR was derived from precycle volumetric MRI assessments and was defined as the first documented response (subsequently confirmed) until documented progression on treatment or death in the absence of disease progression (i.e., precycle volumetric MRI assessment of PFS event or censoring — precycle volumetric MRI assessment of first response, where each cycle was 28 days). If a patient did not progress following a response, then their DOR used the PFS-censoring MRI assessment.¹⁹

Progression-Free Survival

PFS was a secondary outcome of the SPRINT phase II trial and was defined as the time from initiation of treatment until the precycle volumetric MRI assessment of objective disease progression on treatment or death. Patients who did progress or died by the time of analysis were censored at their last evaluable precycle MRI assessment. However, if the patient progressed or died after 2 or more missed precycle MRI assessments, the patient was censored at the time of the latest evaluable precycle MRI assessment.¹⁹

Time to Response

TTR was a secondary outcome of the SPRINT phase II trial and was defined as the time from study treatment initiation until the precycle volumetric MRI assessment of the first documentation of CR or a subsequently confirmed PR. The precycle volumetric MRI assessment of the first documented response was to coincide with that used for the DOR end point. Only patients who had achieved a CR or a confirmed PR were evaluated for TTR. Patients who had not progressed at the time of analysis were censored at their last evaluable precycle MRI assessment.¹⁹

Clinical Outcome Assessments

Table 9: Summary of PROs and Functional Evaluations by Respondent and Age

DCO = data cut-off; PRO = patient-reported outcome; PROMIS = Patient-Reported Outcomes Measurement Information System; Y = yes.

^aSeparate child: 8 to 12 and adolescent: 13 to 18 forms.

^bSeparate toddler: 2 to 4; young child: 5 to 7; child: 8 to 12; and adolescent: 13 to 18 forms.

Source: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO).¹⁹

Pain: The effect of selumetinib on pain was assessed through self-evaluation of pain intensity on the tumour selected by the physician, on the tumour selected by the patient, overall tumour pain, and other pain, as assessed by the NRS-11, and the interference of pain with daily functioning as rated by the patient and by the parent, using the PII.¹⁹

Numerical Rating Scale-11: The NRS-11 is a self-evaluation of pain intensity and consists of 4 questions scored on a scale of 0 (no pain) to 10 (worst pain imaginable). The NRS-11 was only used in patients aged 8 years and older. The primary outcome for the self-report NRS-11 was the rating of pain in 1 specific PN (e.g., the “target tumour”). Patients who had their baseline evaluation using the earlier version of the NRS-11 (Version 1), which did not specifically indicate the physician-selected target PN, were included in the primary outcome analysis if the self-selected PN and physician-selected PN matched. The revised NRS-11 (Version 2) was used for all patients enrolled from November 2015. Pain was rated across all study precycles for the same PN each time. Pain palliation was also defined for the primary outcome and was based on reduction in pain intensity, and stability or reduction in analgesic use.¹⁹

Clinically meaningful thresholds (CMTs) of change for the NRS-11 were based on literature for other populations and were defined as a decrease of 1 or 2 points.^34,38-41 No published data focusing on the reliability, validity, responsiveness to change, or minimal important difference (MID) for the NRS-11 was identified for pediatric patients with NF1. Additional information on the measurement properties in NRS-11 is provided in Table 39 of Appendix 4.

Pain Interference Index: The PII is a 6-item scale that assesses the extent to which pain has interfered with an individual’s daily activities in the past 7 days. Items are rated on a 7-point Likert scale (0 = not at all to 6 = completely), and the total score is the mean of the completed items. The total score was computed if at least 50% of the items were answered. Higher scores indicated more interference with daily activities.¹⁹

Reliability and validity of the PII were assessed in patients with NF1 or cancer, as well as in caregivers.⁴² Cronbach alphas were 0.84 for the patient PII and 0.94 to 0.96 for the parent PII. Additionally, patient and parent scores on the PII were correlated (r = 0.62). The PII scale demonstrated good validity based on correlation with other pain scales. In the subgroup of NF1 patients (N = 31), the correlation between PII and NF1 Disease Severity Scale ratings was assessed. Patients with pain rated as moderate to severe scored significantly higher on PII than did those with pain rated as mild. Responsiveness to change was measured in a small study of adolescents and young adults with NF1 (N = 12), although patient-reported pain did not significantly decrease.⁴³ No published data focusing on the MID was identified. Additional information on the measurement properties in PII is provided in Table 39 of Appendix 4.

Motor function: Analysis of motor function included only patients with motor morbidity at enrolment, with the exception of the leg length discrepancy analysis and the grooved pegboard analysis. These analyses included only patients with lumbosacral plexus or lower limb PN and patients aged 5 years and older at enrolment with cervical, upper thoracic, or upper limb PN. Motor function in all patients with motor morbidity was assessed using the PROMIS Pediatric Short Form Version 1.0 – Mobility 8a and PROMIS Pediatric Short Form Version 1.0 – Upper Extremity 8a, and included a recall period of 7 days.¹⁹

Patient-Reported Outcome Measurement Information System: The PROMIS was completed by both the patient and the parent. Physical functioning scales assessed the domains of mobility and upper extremity function and included mobility items such as “I could walk upstairs without holding on to anything” and upper extremity items such as “I could button my shirt or pants.” The short forms consisted of 8 items using a 5-point Likert scale (0 = unable to do, 4 = can do without any difficulty), with higher scores indicating better physical functioning. Raw scores were converted to t scores, which were based on reference data from the US general population (mean = 50 and SD = 10).¹⁹ CMTs used for the analysis of PROMIS were:

• mobility self-report: 2.26 (raw score), 3.27 (transformed)

• mobility parent-report: 3.36 (raw score), 3.66 (transformed)

• upper extremity self-report: 3.73 (raw score), 6.45 (transformed)

• upper extremity parent-report: 4.64 (raw score), 6.18 (transformed).

No published data focusing on the reliability, validity, responsiveness to change, or MID for the PROMIS were identified for pediatric patients with NF1. Additional information on the measurement properties of the PROMIS are described in Table 39 of Appendix 4.

Functional Evaluations

The primary outcomes for assessing motor function were strength for each muscle group in all patients with PN causing motor dysfunction, weakness, or cord compression were strength evaluations (Medical Research Council 5-point Likert scale) for all muscle groups, and ROM (measured in degrees) of all joints at baseline. Each joint (ROM) and muscle group (strength) was allocated to a location quadrant based on the anatomic location of the PN (upper or lower; right, left, or bilateral). A strength score was calculated as the average strength of all muscles in the same body quadrant as the target PN. Similarly, the ROM score was calculated as the sum of all the degrees of movement for each of the joints in the same quadrant as the target PN; higher ROM scores indicate more degrees of movement.¹⁹

Airway function: Analysis of airway function included only patients with airway morbidity at enrolment. All patients with airway PN (upper airway or extrathoracic, and lower airway or intrathoracic) underwent functional evaluations, including sleep studies, evaluation of endurance using the 6-minute walk test (6MWT) and pulmonary function tests (PFTs). The primary outcome for this analysis was the Apnea-Hypopnea Index (AHI) as measured in events per hour, forced expiratory volume in the first second (FEV₁) measured in litres, and respiratory resistance at 20 Hz (R20). For pre-school children, forced expiratory volume in the first 0.75 seconds (FEV_0.75) was used in place of FEV₁.¹⁹

A change of 12% or more in FEV₁ or FEV_0.75 was classified as an improvement, as recommended by the REiNS functional group.⁴⁴ Functional improvement was defined as a decrease of 20% or more in R20. (The REiNS functional group recommended the threshold of 20% or more for R10, but R20 was used in the SPRINT phase II study.¹⁹)

Bowel and bladder function: Analysis of bowel and bladder function included only patients with bowel and/or bladder morbidity at enrolment. Bowel and bladder functionality was measured with the Dysfunctional Voiding Questionnaire (DVQ) and was only completed by patients aged 8 years or older or by the parent or guardian of a patient with bowel and/or bladder morbidity.⁴⁵ The DVQ consists of 14 items measuring bowel and/or bladder dysfunction on a 5-point Likert scale. The last question requested feedback on the ease of completing the questionnaire and was not included in the total score. Scores ≥ 11 (out of 52) were the threshold for bowel and bladder dysfunction. The absolute change in DVQ score from baseline was calculated as each post-baseline value minus the baseline value.^19,45

Vision function: Analysis of vision function included only patients with vision morbidity at enrolment and included measurements of visual acuity and the extent of exophthalmos. Visual acuity was measured using the HOTV chart, or Teller acuity cards if the patient was too young to reliably perform HOTV testing. The HOTV results were reported as the logarithm of the minimum angle of resolution (logMAR), and Teller acuity was recorded in cycles per centimetre, which was converted to a logMAR by the study team.¹⁹

Exophthalmos was measured using exophthalmometry (in millimetres). Change from baseline in exophthalmometry and using the logMAR were classified as improvement, no change, or deterioration according to the score at each precycle visit.

A decrease in the logMAR of more than 0.2 and a decrease in exophthalmos of more than 2 mm were considered clinically meaningful improvements, as recommended by the REiNS functional group.¹⁹

Global Impression of Change

Global impression of change (GIC) in tumour pain, overall pain, and tumour-related morbidities compared to baseline were measured by the GIC scale, consisting of 3 questions scored on a 7-point scale (1 = very much improved to 7 = very much worse), and was performed by patients and parents. The patient self-report and parent-report ratings for each of the 3 items were analyzed and reported separately.¹⁹

Pediatric Quality of Life Inventory

General HRQoL was measured using the PedsQL Version 4.0. The PedsQL assesses function in 4 domains:¹⁹

• physical functioning (8 items)

• emotional functioning (5 items)

• social functioning (5 items)

• school functioning (5 items).

Each item is scored on a 5-point Likert scale (0 = never a problem; 4 = almost always a problem). For patient-reported and parent-reported measures, items were reverse-scored and linearly transformed to a 0-to-100 scale, with higher scores indicating better HRQoL. Scale scores were computed as the sum of the items divided by the number of items answered. If more than 50% of the items in the scale were missing, the scale score was not computed.⁴⁶ A total scale score was also derived from the sum of all the items divided by the number of items answered on all the scales. The primary outcomes for HRQoL were the total scale score of the patient-reported PedsQL for children older than 8 years and the total scale score of the parent-reported PedsQL administered to parents of children aged 2 years and older. Secondary outcomes for HRQoL were the mean scores of the 4 domains (physical, emotional, social, and school) of the patient-reported scale completed by children older than 8 years and the 4 domain mean scores from the parent-reported PedsQL administered to parents of children aged 2 years and older.¹⁹

Patients were classified with impaired global HRQoL (yes/no) at each precycle visit, using the transformed scores if their total or domain scores fell 1 SD below the population sample mean as reported by Varni et al. (2003).^19,46 No published data focusing on the reliability, validity, responsiveness to change, or MID for the PedsQL were identified for pediatric patients with NF1. Additional information on the measurement properties of the PedsQL are described in Table 39 of Appendix 4.

Quality of Life Background Form

A QoL Background Form was completed by the parent of all patients aged 2 to 18 years to document information about demographics, use of pain medication, and NF1 disease severity. Descriptive statistics (counts and percentages) were provided for each question asked.¹⁹

6-Minute Walk Test

All patients aged 5 years or older at enrolment, with a lower extremity PN, cord compression, or airway PN (including patients with tracheostomy, providing they can walk independently) underwent endurance testing using the 6MWT. The absolute change from baseline was calculated as each post-baseline value minus the baseline value. CMT was defined as a change of 30 m (improvement or deterioration) or no change (otherwise).¹⁹

General PN Symptoms

General PN symptoms were measured by the PN symptoms checklist, which consisted of 36 symptoms scored on a 5-point Likert scale (0 = not at all; 1 = a little; 2 = some; 3 = pretty much; 4 = a lot) for each patient at each assessment.¹⁹

Harms Outcomes

Safety and tolerability were assessed in terms of AEs, SAEs, deaths, laboratory data, vital signs, electrocardiogram, left ventricular ejection fraction, physical examination, and performance status (Lansky and Karnofsky), ophthalmology, and bone density examination. Adverse events and SAEs were summarized by system organ class and preferred term. Severity of AEs was defined using the Medical Dictionary for Regulatory Activities (Version 21.0) and CTCAE Version 4.0.¹⁹

Statistical Analysis

Sample Size and Power Calculation

The sample size for the primary objective of ORR in patients with PN-related morbidity was based on a target response rate to exclude 15% with a lower 2-sided 95% CI. With 50 total evaluable symptomatic patients, an exact binomial test with a nominal 1-sided 2.5% significance level had 90% power to detect the difference between a null hypothesis response rate of 15% and an alternative hypothesis response rate of 36%. With 14 or more responses out of the 50 patients, the lower limit of the exact 2-sided 95% CI for the response rate would be 16.2% or greater. Thus, 14 or more responses among 50 evaluable patients was consistent with results that would statistically significantly exceed a 15% true response rate based on a 2-sided CI.¹⁹

Statistical and Analytical Plans

At baseline, the most clinically relevant tumour was selected by the treating physician as the target lesion and was used to determine response to treatment. All volumetric MRIs were sent to the NCI POB expert reader for central review, which was the primary analytical method. The NCI POB central analysis was carried out on all volumetric MRIs by a single, expert reader. No site-investigator review was performed.

Disease progression was evaluated according to REiNS criteria, which involved the analysis of 1 target PN and up to 2 nontarget PNs, for which disease progression could be based solely on progression of a clinically relevant nontarget PN. However, because no clinically relevant nontarget PN were reported for stratum 1 of the SRINT phase II trial, the use of target PNs only for assessment of disease progression was referred to as “modified REiNS criteria.”¹⁹

Following a regulatory agency request, a retrospective ICR of the SPRINT phase II stratum 1 volumetric MRI scans was conducted, according to modified REiNS (sensitivity analysis) and modified RECIST 1.1 criteria (exploratory analysis).¹⁹ The ICR analysis was conducted using the same modified REiNS criteria used by the NCI POB central review. The modification in standard REiNS and RECIST 1.1 criteria was based on use of only the target PN for derivation of response. The ICR analysis was performed with volumetric MRIs assessments performed by either an independent whole-body radiologist or a neuro-radiologist (2 radiologists and 1 neuro-radiologist were available). The ICR was used as a sensitivity analysis for the primary objective of ORR and secondary objective of DOR.¹⁹ Analyses based on the ICR of the volumetric MRI images were not performed in the March 31, 2021, DCO. Interreader variability analyses were conducted to determine the margin of error for target PN measurements.

A summary of outcomes and analytical methods for the SPRINT phase II study is outlined in Table 10.

Table 10: Summary of Statistical Analysis of Efficacy End Points in SPRINT Phase II

6MWT = 6-minute walk test; CMT = clinically meaningful thresholds; DCO = data cut-off; DVQ = Dysfunctional Voiding Questionnaire; ICR = independent central review; MMRM = mixed model for repeated measures; NA = not applicable; NRS-11 = Numeric Rating Scale-11; PedsQL = Pediatric Quality of Life Inventory; PFS = progression-free survival; PFT = pulmonary function test; PII = Pain Interference Index; PROMIS = Patient-Reported Outcomes Measurement Information System; RECIST = Response Evaluation Criteria in Solid Tumors; REiNS = Response Evaluation in Neurofibromatosis and Schwannomatosis.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Efficacy Analyses

All efficacy analyses were performed on the full analysis set (FAS).

Objective response rate: For the primary end point of ORR assessed by NCI POB volumetric MRI, results were presented with corresponding 2-sided exact 95% CIs based on the Clopper-Pearson method. The overall response (at each DCO) was also summarized by category (unconfirmed PR, confirmed PR, CR, stable disease, progressive disease, or NE).³⁴

The BOR, defined as the best response recorded from the start of the treatment until progression or the last evaluable MRI assessment in the absence of progression, was also be summarized by category (confirmed PR, CR, stable disease, progressive disease, or NE).³⁴ Patients in the FAS with no imaging assessments were counted as nonresponders.¹⁹

• Subgroup analysis: The ORR was also summarized by PN status at enrolment (progressive, nonprogressive, or unknown). Progressive disease at trial entry was defined as an increase of 20% or greater in neurofibroma volume no more than 15 months before enrolment.

• Sensitivity analyses: A sensitivity analysis based on the ICR of volumetric MRI data was also performed for the primary end point of ORR using the REiNS criteria to assess the robustness of the central single reviewer on analysis of ORR. Cross-tabulation summaries of PN response by the NCI POB central analysis versus the ICR analysis were produced to assess the concordance between the NCI POB central analysis target PN responses and the ICR analysis responses. Finally, the ICR review, based on REiNS, was used to evaluate the interreader agreement in the volume measurements at baseline and in changes in target PN volume over time.

  The modified RECIST 1.1 assessments conducted by the ICR were used as an exploratory analysis of ORR.

• Plexiform neurofibroma volume: Changes in PN growth were evaluated descriptively by summaries of percentage and absolute change in PN volume from baseline to scheduled precycle assessments. The average tumour volume and 95% CI were displayed graphically across time (precycle assessments). The best percentage change from baseline was summarized descriptively and presented graphically using waterfall plots. For patients who had PN progression at study entry, the type of PN (“typical,” “nodular,” or “solitary”) was marked on the waterfall plot.

Duration of response: DOR was summarized by the Kaplan-Meier method. Plots of DOR were presented, along with median DOR and 95% CI, to summarize the number and percentage of patients remaining in response for at least 4, 8, 12, 16, 20, and 24 cycles. The Kaplan-Meier method was also used to calculate the median (and 95% CI) time to onset of response from initiation of study treatment and summarize the number and percentage of responding patients with an onset by or within at least 4, 8, 12, 16, 20 and 24 cycles. Swimmer plots showing the profile of each patient, classified by PN status at enrolment (progressive, nonprogressive, or unknown), were produced.

• Sensitivity analysis: Sensitivity analyses for DOR included assessment based on the ICR according to REiNS criteria (as for the outcome of ORR) using the same statistical method used for the NCI POB central analysis. A second sensitivity analysis using the actual dates of the NCI POB–based volumetric MRI assessments or death was conducted to assess the impact of dose interruptions on the DOR.

  The modified RECIST 1.1 assessments conducted by the ICR were used as an exploratory analysis of DOR.

Other secondary end points (PFS and TTR): PFS and TTR were summarized using the Kaplan-Meier method. The median PFS and TTR (and 95% CIs) were calculated from the Kaplan-Meier plot. The percentage of patients who are progression-free at cycles 5, 9, 13,17, and 25, and the percentage of patients not in response at cycles 5, 9, 13, 17, and 25 were summarized.

Interreader variability: For outcomes based on volumetric MRI (ORR, DOR, PFS, and TTR), interreader variability was assessed by 2 independent radiologists to determine the amount of agreement between the target PN volume measurements at baseline and changes in volume over time using the intraclass correlation coefficient.

Clinical Outcome Assessments

For analysis of the PROs that include patient-reported and parent-reported questionnaires (PII, PROMIS, DVQ, GIC, general PN symptom checklist, and PedsQL), results were analyzed and presented separately. Summary measures of disposition and compliance at each scheduled assessment were derived for all PROs. The primary analysis of the PROs and functional outcomes were based on the descriptive statistics and mixed model for repeated measures (MMRM) to summarize changes over time. Change from baseline in the PRO and functional evaluation scores at precycle 13 was the primary analysis.¹⁹ Evaluation at precycle 13 was chosen based on the results of SPRINT phase I, which suggested that the majority of patients achieve their first response within the first 12 months of treatment.^19,21 Changes from baseline were analyzed using a restricted maximum likelihood–based MMRM analysis, including terms for precycle visit, baseline score, age, the number of morbidities at baseline and baseline visit × precycle interaction. The model presented least squares mean estimates, standard errors, 95% CIs and P values for mean changes from baseline to each precycle visit. At each post-baseline assessment, the absolute change in scores from baseline was calculated as the post-baseline value minus the baseline value for each item, domain, and primary PRO outcome as applicable. The change from baseline values were classified as worsening (≥ 3 points, 2, or 1 points compared to baseline), stable, or improved (1, 2 or ≥ 3 points compared to baseline):¹⁹

In addition, supportive analyses using CMTs were conducted to help with interpretation of clinical benefit using published literature, and both distribution- and anchor-based approaches. For the anchor-based approach, the GIC was used as an anchor as it asks patients (or parents) to assess changes in pain and other morbidities that can be linked to relevant questions or outcomes from the PROs or functional evaluations. The correlation between the anchor and the PRO or functional evaluation was reported. An anchor was considered adequate if the correlation coefficient was greater than 0.30.⁵⁰ For the distribution-based approach, the half value of the SD of the baseline scores was used.¹⁹

Correlation Analysis (Exploratory)

Correlation and association analyses between changes in PROs and functional outcomes and percent target PN volume change from baseline to precycle 13 were explored to characterize the effect of selumetinib on percent target PN volume change and key PRO and functional outcomes. Correlations between clinical outcome assessments and PN volume at baseline were explored using scatterplots and Spearman rank correlation coefficients (r), 95% CIs, and P values. The Spearman rank correlation coefficients were assessed according to Cohen (1988):⁵¹

• r = 0 to 0.3 equates to a weak correlation

• r = 0.3 to 0.5 equates to a moderate correlation

• r = 0.5 to 1.0 equates to a strong correlation.

A positive correlation was defined as both an improvement in the clinical outcome assessment and a reduction in target PN volume. In addition, the association between PN response at precycle 13 and clinically meaningful change in clinical outcome assessments (yes or no) at precycle 13 were evaluated by a Fisher exact test.¹⁹

Missing Data

The PRO and functional end points were analyzed using the MMRM under a missing-at-random assumption. In addition, the rates of missing data by precycle assessment for individual items, domains, and end points were summarized. If, after inspection of the patterns and reasons for these missing data, the data were believed to be not missing at random, the impact of missingness on the analyses was explored by making assumptions about the missing items, domains, or days.^19,34

Safety Analyses

Safety data were not formally analyzed but summarized descriptively by count and percentage.

Analysis Populations

The following analysis populations of interest were defined in the SPRINT phase II trial:

• FAS: all patients who received at least 1 dose of selumetinib. The FAS was used for all efficacy analyses, unless otherwise reported

• safety analysis set: the same as the FAS in this study, consisting of all patients who received at least 1 dose of selumetinib

• Pharmacokinetic analysis set: patients in the FAS with at least 1 postdose sample taken for pharmacokinetic analysis.

Protocol Amendments, Deviations, and Changes to Planned Analyses

Protocol Amendments

The original clinical study protocol was finalized on April 6, 2011. A total of 18 protocol amendments were reported. The first 8 amendments (A to H) were for the phase I portion of the study (SPRINT I). Amendment I was implemented to expand the SPRINT study to include a phase II evaluation of selumetinib in the same population. Nine additional protocol amendments were recorded for the SPRINT phase II trial, though Amendments O to R occurred following the DCO for the SPRINT phase II trial. A list of protocol amendments and key changes for the SPRINT phase II study are summarized in Table 11.

Table 11: Protocol Amendments and Key Changes to SPRINT Phase II

AE = adverse event; BiPAP = bilevel positive airway pressure; CAEPR = Comprehensive Adverse Event and Potential Risks; CPAP = continuous positive airway pressure; CSP = clinical study protocol; CTEP = Cancer Therapy Evaluation Program; DCO = data cut-off; ERK = extracellular signal related kinase; IOP = intraocular pressure; PN = plexiform neurofibroma; QTc = corrected QT interval; QTcF = QT interval corrected using Fridericia’s correction formula; SPEER = Specific Protocol Exceptions to Expedited Reporting.

Source: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO).¹⁹

Protocol Deviations

The number of patients with important protocol deviations in each treatment group are summarized in Table 12. Prior to database lock for the June 29, 2018, and March 31, 2021, DCOs, at least 1 important protocol deviation occurred in 2 patients (4.0%), and 4 patients (8.0%) in the SPRINT phase II trial, respectively.

Table 12: Important Protocol Deviations (Full Analysis Set)

DCO = data cut-off.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Changes to Planned Analyses

Changes from the planned analysis in the protocol before the final outcome results included:

• The statistical analysis plan (version 4.0) describes the analysis of phase I and stratrum 1 of the phase II SPRINT trial.

• Analysis of the SPRINT phase II secondary objective “To determine the effect of selumetinib on the PN growth rate based on volumetric analysis of MRI studies obtained prior to enrolment (if available and amenable to volumetric analysis)” was not performed.

• Analysis of the SPRINT phase II secondary objective “In patients who enroll on this study with presence of an optic pathway tumour or other glioma not requiring treatment with chemotherapy or radiation therapy, to evaluate the effect of selumetinib on changes in the size of the optic pathway tumour or other glioma” was not performed.

Changes from the planned analysis in the protocol after the availability of the final, validated outcome results included:

• The threshold for clinically significant improvement for exophthalmometry was corrected from greater than −2 mm to less than −2 mm.

• The threshold for clinically significant improvement for HOTV or Teller acuity cards was amended from greater than or equal to 0.2 logMAR to less than −0.2 logMAR.

• The threshold for clinically significant deterioration for HOTV or Teller acuity cards was amended from equal to or greater than −0.2 logMAR to greater than 0.2 logMAR.

• The margin of error and intraclass correlation coefficient were assessed both for the PN volume at baseline, and for the percentage change in PN volume postbaseline.

• The association between post-baseline longitudinal changes in PRO or functional outcomes and changes in tumour volume was assessed explaining the change in clinical outcome by change in tumour volume rather than the opposite.

• Growth rates associated 95% CI were not reproduced based on the bootstrap percentile interval technique as individual patient data were available, allowing for a precise estimate of the standard error and 95% CI.

• Use of local laboratory ranges instead of project ranges as the local reference ranges were markedly different from the project reference ranges for certain parameters (increased lipase, increased CPK, and conversion to the latter) would have resulted in erroneous grade shifts.

  Dose reductions are not presented according to the statistical analysis plan as there was no specific case report form field in the clinical database to explicitly record them.

Results

Patient Disposition

The disposition of patients enrolled in stratum 1 of the SPRINT phase II study is summarized in Table 13. A total of 50 patients were enrolled and received at least 1 dose of selumetinib. At the June 29, 2018, DCO, 16 patients (32%) discontinued selumetinib, primarily due to AEs (6 [12%]), disease progression (3 [6.0%]), and investigator discretion (3 [6.0%]). Of the 3 patients with disease progression, treatment was discontinued at |||| years, |||| years, and |||| years. Four patients who discontinued treatment also terminated the study, primarily due to voluntary discontinuation by patient (2 [4.0%]), loss to follow-up (1 [2.0%]) and “other” unspecified reasons (1 [2.0%]). Three of the patients who discontinued treatment due to AEs also discontinued from the study.¹⁹

At the March 31, 2021, DCO, |||||| additional patients discontinued treatment (for a total of || |||||||), with no new additional discontinuations due to AEs, and |||||| additional discontinuations due to disease progression (total of | |||||||) and investigator discretion (a total of 6 [12.0%]). ||||| additional patients who discontinued treatment also terminated the study (total of || |||||||), primarily due to “other” reasons (total of | |||||||) which were not specified, voluntary discontinuation by the patient (total of | ||||||, switch to alternative treatment (total of | ||||||), disease progression (total of | |||||]), and loss to follow-up (total of | ||||||).²⁰

Table 13: Patient Disposition

DCO = data cut-off.

^aIn the June 29, 2018, DCO, 2 patients discontinued selumetinib treatment due to a completed treatment period and disease progression on study. Since the June 29, 2018, DCO, both patients have been re-treated with selumetinib and were therefore no longer included in these categories.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Exposure to Study Treatments

Exposure in the safety population of SPRINT phase II is summarized in Table 14. At the June 29, 2018, DCO, the total treatment duration of selumetinib was ||||| days (range = || || |||| days; approximately |||| months or ||| years). The actual treatment duration (sum of days study dose was administrated) of selumetinib was |||||| days (range = || || |||| days).¹⁹

At the March 31, 2021, DCO, the total treatment duration of selumetinib was |||||| days (range = || || |||| days; approximately |||| months or ||| years). The median treatment duration was |||||| days (range = |||||| to |||| days; approximately |||| months or ||| years).²⁰

Dose Interruptions

At the June 29, 2018, DCO, all patients had at least 1 dose interruption, primarily due to patient compliance ||| ||||||||| “other” ||| ||||||||| AEs ||| ||||||||| or medication error || |||||||. Single missed doses were counted as interruptions. One patient had ||| interruptions recorded; however, interruptions did not exceed |||||| days per interruption. For patients with interruption due to AEs, | |||||| required more than 1 dose reduction. “Other” reasons for interruption included logistical issues such as surgery, travel issues, and participation in a sleep study.¹⁹

Table 14: Duration of Exposure Including Both Initial Treatment and Re-Treatment (Safety Analysis Set)

DCO = data cut-off; SD = standard deviation.

^aTotal treatment duration = (last dose date – first dose date + 1).

^bFor patients without re-treatment, total treatment duration = (last dose date of initial treatment – first dose date of initial treatment + 1). For patients with re-treatment, total treatment duration = (last dose date of initial treatment – first dose date of initial treatment + 1) + (last dose date of re-treatment – first dose date of re-treatment + 1).

^cActual treatment duration = sum of days of study dose administered.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

At the March 31, 2021, DCO, the numbers of interruptions due to “other” reasons, patient compliance, AEs, and medical error were || |||||||, || |||||||, || |||||||, and | |||||||, respectively. Between the June 29, 2018, and the March 31, 2021, DCOs, |||||| additional patients had dose interruptions due to missed doses, which were recorded as medication errors, and |||||| additional patients had dose interruptions due to AEs. One patient who had a medication error at the time of the June 29, 2018, DCO experienced an additional medication error as of the June 29, 2018, DCO, which resulted in study treatment overdose.²⁰

Concomitant Therapy

As of the June 29, 2018, DCO, all patients received at least |||||| concomitant medication for the treatment of medical conditions or AEs. The most common concomitant medications ||||| || ||||||||| |||| ||||||||||| |||||||| ||||||||||| |||||||| ||||||||| |||||||| ||||||||||| |||||||| ||||||||| ||||||| ||| ||||||||| |||||||. In total, || ||||||| patients received || ||||| ||||||||||| |||| ||| ||| ||||||| ||| ||| ||||| ||||||||| ||| |||| |||||| ||||| |||||||||||| ||||||||¹⁹

At the time of the March 31, 2021, DCO, the most common concomitant medications were ||||||||||| |||||||| ||||||||||| |||||||| ||||||||| |||||||| ||||||||||| |||||||| ||||||||| |||||||| ||||||||| |||||||| |||||||||| |||||||| ||| ||||||||||||| |||||||.²⁰

Efficacy

Only efficacy outcomes and analyses of subgroups identified in the review protocol are reported here. Appendix 3 provides detailed efficacy data.

Symptom Improvement

Pain

Evaluation of pain intensity measured by the NRS-11 and interference of pain on daily functioning as measured by the PII were secondary end points of the SPRINT phase II trial and are summarized in Table 15 (June 29, 2018, DCO) and Table 16 (March 31, 2021, DCO).

NRS-11: At the June 29, 2018, DCO, 34 patients were eligible to complete the NRS-11 based on age (aged ≥ 8 years); however, only 26 had baseline data on the physician-selected target tumour. Eight patients did not have baseline data as they either received the first version that did not contain this item or did not complete the physician-selected target tumour question in the second version, or there was a difference between the self-selected and physician-selected tumours. Overall, only 20 patients (58.8%) completed all questions of the NRS-11 at both baseline and precycle 13, while 33 patients (97.1%) and 29 patients (85.3%) completed at least 1 question of the NRS-11 at baseline and precycle 13, respectively. In total, 24 patients had sufficient data for NRS-11 for physician-selected target tumour pain, at baseline and cycle 13. The mean adjusted change from baseline score for target tumour pain intensity at precycle 13 was −2.07 (95% CI, −2.84 to −1.31). Of the 24 patients with physician-selected target tumour scores who completed both baseline and precycle 13 assessments, 12 (50.0%) reported an improvement of 2 or more points on the NRS-11. No patients reported deterioration at precycle 13. A total of 14 patients with a baseline NRS-11 score of 2 or more had a reduction in pain intensity by at least 2 points without increased analgesic use.¹⁹

At the March 31, 2021, DCO, the adjusted mean change from baseline at precycle 13 was ||||| (95% CI, ||||| || |||||), which was maintained throughout the analysis period.²⁰

• Sensitivity Analysis: Using the predefined sensitivity analysis threshold of 1 point for the NRS-11, the adjusted mean change from baseline was deemed to be clinically meaningful from precycle 3 through precycle 25 at the June 29, 2018, DCO.¹⁹

  The anchor-based approach to derive a CMT resulted in a threshold of |||| points. However, the sponsor noted that the results of the NRS-11 were poorly correlated with the GIC and were therefore not presented. The distribution-based approach resulted in a threshold of |||| points, which was not included in the results as it lies between the primary CMT of 2 points, and the original sensitivity analysis of 1 point.¹⁹

  Sensitivity analyses were not conducted at the March 31, 2021, DCO.

Pain Interference Index: At the June 29, 2018, DCO, 34 patients were eligible to complete the PII self-report based on age (≥ 8 years); however, because 1 patient had severe cognitive impairment, the self-report was not administered. In total, 32 patients (94.1%) and 28 patients (82.4%) completed all questions of the self-reported PII at baseline and precycle 13, respectively, while 33 (97.1%) and 29 (85.3%) completed at least 50% of the questions. The adjusted mean change from baseline in self-reported PII score at precycle 13 was −0.65 (95% CI, −0.89 to −0.42). Of the 29 patients with precycle 13 data, || ||||||| reported an improvement greater than the CMT of 0.75 points or greater,|| ||||||| reported no change, and | |||||| patient-reported deterioration.¹⁹

For the parent-reported PII, 48 parents were eligible to complete the parent-reported based on their child’s age (≥ 5 years), although there was no baseline value for 1 patient. In the parent-report PII, 46 parents (95.8%), and 42 parents (87.5%) completed all PII questions and precycle 13 questions, respectively, and 47 (97.9%) and 43 (89.6%) completed at least 50% of the questions. The adjusted mean change from baseline in parent-reported PII score at precycle 13 was −0.82 (95% CI, −1.17 to −0.47). Of the 42 parents who completed both baseline and precycle 13 assessments, || ||||||| patients reported an improvement greater than the CMT of 1.78 points or greater, || ||||||| reported no change, and | |||||| reported deterioration.¹⁹

At the March 31, 2021, DCO, results were consistent with the primary analysis, with an adjusted mean change from baseline at precycle 13 of ||||| (95% CI, ||||| || |||||) for the self-report total score, and ||||| (95% CI, ||||| || |||||) for the parent-report total score. Improvements from baseline in self-reported and parent-reported PII scores were observed as early as precycle 3 and were maintained through precycle 49.²⁰

Table 15: Change From Baseline for NRS-11 Pain Intensity and PII Primary Outcomes, MMRM (Full Analysis Set; June 29, 2018, DCO)

CFB = change from baseline; CI = confidence interval; DCO = data cut-off; MMRM = mixed model for repeated measures; NR = not reported; NRS-11 = Numeric Rating Scale-11; PII = Pain Interference Index; SD = standard deviation; SE = standard error.

^aObserved values have not been adjusted.

^bThe model included terms for precycle, baseline score, age, the number of morbidities at baseline and baseline × precycle interaction.

Source: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO).¹⁹

Table 16: Change From Baseline for NRS-11 Pain Intensity and PII Primary Outcomes, MMRM (Full Analysis Set; March 31, 2021, DCO)

CFB = change from baseline; CI = confidence interval; DCO = data cut-off; MMRM = mixed model for repeated measures; NRS-11 = Numeric Rating Scale-11; PII = Pain Interference Index; SD = standard deviation; SE = standard error.

^aObserved values have not been adjusted.

^bThe model included terms for precycle, baseline score, age, the number of morbidities at baseline and baseline × precycle interaction.

Source: SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Motor Function

Motor function was evaluated using the strength of muscle groups and ROM in patients with motor morbidity.¹⁹ Additional assessments of motor function included the PROMIS physical functioning scales, grooved pegboard, grip and key pinch test and leg length evaluations. Results for PROMIS and overall strength and ROM are summarized in the following section, with results for grooved pegboard, grip and key pinch test, and leg length evaluations included in Appendix 3.

Patient-Reported Outcomes Measurement Information System (PROMIS): Results for the PROMIS mobility and upper extremity domains as assessed by the patient and the parent at the June 29, 2018, and March 31, 2021, DCOs are summarized in Table 17. In total, 33 patients had a motor PN-related morbidity and were aged 5 to 18 years. A total of 16 patients (66.7%) completed all questions and at least 50% of questions at baseline, and 20 patients (83.3%) completed all questions and at least 50% of questions at precycle 13 of the PROMIS mobility self-report. For the PROMIS upper extremity self-report, 19 patients (79.2%) completed all questions at both baseline and precycle 13 assessments, while 22 (91.7%) and 20 (83.3%) completed 50% of questions at baseline and precycle 13, respectively. For parent-report mobility, 25 (75.8%) and 29 (87.9%) completed all questions and at least 50% of questions at baseline and precycle 13, respectively. For parent-reported upper extremity, 27 (81.8%) and 29 (87.9%) completed all questions at baseline and precycle 13, while 31 (91.2%) and 29 (87.9%) completed at least 50% of questions at baseline and precycle 13.^19,20

At both DCOs, the baseline scores for the self- and parent-report assessments in the mobility domain were 46.57 (SD = |||||) and 37.43 (SD = |||||), while the baseline scores for the self- and parent-report assessments in the upper extremity domain were 45.95 (SD = ||||||) and 38.15 (SD = ||||||). The adjusted mean change from baseline at precycle 13 in the self-reported mobility domain was 1.75 points (95% CI, −0.70 to 4.19) at the June 29, 2018, DCO, and |||| points (95% CI, ||||| || ||||) at the March 31, 2021, DCO, neither of which met the CMT of |||| points. In the parent-reported assessment, the adjusted mean change from baseline at precycle 13 in the mobility domain was |||| points (95% CI, |||| || ||||) at the June 29, 2018, DCO, and |||| points (95% CI, |||| || ||||) at the March 31, 2021, DCO, which also did not meet CMT of |||| points.^19,20

The adjusted mean change from baseline at precycle 13 in the self-reported upper extremity domain was |||| points (95% CI, −|||| || ||||) at the June 29, 2018, DCO, and |||| points (95% CI, ||||| || ||||) at the March 31, 2021, DCO, while the adjusted mean change from baseline at precycle 13 in the upper extremity domain was |||| points (95% CI, ||||| || ||||) at the June 29, 2018, DCO, and |||| points (95% CI, ||||| || ||||) at the March 31, 2021, DCO, neither of which met the CMTs of |||| points and |||| points.^19,20

Table 17: Change From Baseline for PROMIS, MMRM (FAS With Motor PN-Related Morbidity)

CFB = change from baseline; CI = confidence interval; DCO = data cut-off; FAS = full analysis set; MMRM = mixed model for repeated measures; NR = not reported; PN = plexiform neurofibroma; PROMIS = Patient-Reported Outcomes Measurement Information System; SD = standard deviation; SE = standard error.

^aPatients aged 8 to 18 years at enrolment were expected to complete self-report measures of the PROMIS (N = 24).

^bObserved values are for transformed scores and have not been otherwise adjusted.

^cThe MMRM included terms for precycle, baseline score, age, the number of morbidities at baseline and baseline × precycle interaction.

^dParents or legal guardians of patients aged 5 to 18 years at enrolment expected to complete the parent proxy PROMIS (N = 33).

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Strength and Range of Motion: A total of 33 patients had motor morbidity in any body quadrant at enrolment; however, only 31 patients had baseline values. Results for the MMRM analysis including patients with motor morbidity at enrolment, regardless of the location of their target PN at the June 29, 2018, and March 31, 2021, DCOs were nearly identical; thus, only the results for the most recent DCO (March 31, 2021) are summarized in Table 18. At the March 31, 2021, DCO, 27 patients had evaluable strength assessments at baseline and precycle 13, with an adjusted mean change from baseline of |||| (95% CI, |||| || ||||), which was the same as the June 29, 2018, DCO.^19,20

Table 18: Change From Baseline for Strength MMT, MMRM (Full Analysis Set With Motor PN-related Morbidity; March 31, 2021, DCO)

CFB = change from baseline; CI = confidence interval; DCO = data cut-off; MMRM = mixed model for repeated measures; MMT = manual muscle test; NA = not applicable; NR = not reported; PN = plexiform neurofibroma; SD = standard deviation; SE = standard error.

^aAverage strength score from all muscles in the same body quadrant (right upper, right lower, left upper, left lower, upper bilateral, lower bilateral) as the target PN.

^bThe model included terms for precycle, baseline score, age, the number of morbidities at baseline and baseline × precycle interaction.

Source: SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Results of the MMRM analysis for change from baseline in ROM in patients with target PN in any body quadrant are summarized in Table 19. At the June 29, 2018, DCO, at baseline (N = 33), the mean ROM sum of all joints was |||||| degrees (SD = |||||||). There was an increase in ROM from baseline at precycle 13 of ||||| degrees (95% CI, ||||| || ||||||).¹⁹ At the March 31, 2021, DCO, the adjusted mean change from baseline at precycle 13 was ||||| degrees (95% CI, ||||| || ||||||).²⁰

Table 19: Change From Baseline for ROM MMT, MMRM (Full Analysis Set With Motor PN-Related Morbidity)

CFB = change from baseline CI = confidence interval; DCO = data cut-off; MMRM = mixed model for repeated measures; NA = not applicable; NR = not reported; PN = plexiform neurofibroma; ROM = range of motion; SD = standard deviation; SE = standard error.

^aSum of all the degrees of movement for each of the joints in the same body quadrant (right upper, right lower, left upper, left lower, upper bilateral, lower bilateral) as the target PN.

^bThe model includes terms for precycle, baseline score, age, the number of morbidities at baseline and baseline × precycle interaction.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Grooved Pegboard, Key Pinch Grip, and Leg Length Evaluations: Results for grooved pegboard, key pinch grip, and leg length evaluations were only available at the June 29, 2018, DCO and are summarized in Figure 6, Figure 7, Figure 8, and Figure 9 of Appendix 3. Overall, results for these measures were consistent with the PROMIS and overall strength findings.^19,20

Airway Function

In stratum 1 of the SRINT phase II trial, 16 patients had airway dysfunction at baseline. Results of tests of airway function, including FEV₁ or FEV_0.75 (in pre-school children), R20, and AHI, for the June 29, 2018, and March 31, 2021, DCOs are summarized in Table 20.

The mean change from baseline at precycle 13 in FEV₁ or FEV_0.75 was ||||| L (SD = ||||||), representing a mean percent change from baseline of |||||| (SD = |||||||| || ||) at both DCOs, where | ||||||| patients showed improvement, | ||||||| patients showed no change and no patients showed deterioration in FEV₁ or FEV_0.75.^19,20

At precycle 13 for both DCOs, the mean change from baseline in R20 was |||||| (SD = ||||||), representing a ||||| (SD = ||||||) change from baseline.^19,20

At baseline, no patients in SPRINT phase II had an AHI of more than 5 events per hour. The mean baseline AHI was ||||| (SD = ||||||). At both DCOs, the mean change from baseline in AHIs at precycle 13 was ||||| (SD = ||||||), with a mean number of AHI at precycle 13 of ||||| (SD = ||||||).^19,20

Table 20: Airway Function Test Scores Over Time and Change From Baseline Over Time (Full Analysis Set With Airway PN–Related Morbidity)

CFB = change from baseline; DCO = data cut-off; FEV₁ = forced expiratory volume in the first second; FEV_0.75 = forced expiratory volume in the first 0.75 seconds; PN = plexiform neurofibroma; R20 = respiratory resistance at 20 Hz; SD = standard deviation.

Sources: SPRINT Phase II Stratum 1 Clinical Study Report (June 29, 2018, DCO)¹⁹ and SPRINT Phase II Stratum 1 Clinical Study Report (March 31, 2021, DCO).²⁰

Bowel and Bladder Function

Results for bowel and bladder function as assessed by the parent-reported DVQ are summarized in Table 21. A total of 10 patients had bowel and bladder PN–related morbidity at baseline, of whom, only 2 had baseline data for self-reported scores. The mean change from baseline at precycle 13 was therefore not calculable at both DCOs.

In the parent-reported assessment of bowel and bladder function, 6 ||||| and 7 ||||| parents completed all questions at baseline and precycle 13, respectively. Insufficient data were available to determine the adjusted mean change from baseline at both DCOs for the parent-reported assessment. The observed mean change from baseline at precycle 13 was |||| (SD = ||||) at both DCOs.^19,20

Table 21: Bowel and Bladder Function Parent-Report Scores Over Time and Change From Baseline Over Time (Full Analysis Set With a Bowel and/or Bladder PN–Related Morbidity)