CADTH Reimbursement Review

Selpercatinib (Retevmo)

Sponsor: Eli Lilly Canada

Therapeutic area: RET fusion–positive non–small cell lung cancer

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

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

Introduction

Lung cancer is the most frequently diagnosed cancer in Canada and the leading cause of cancer-related deaths in males and females,¹ with more than 29,600 new diagnoses (12.5% of new cases in males and 13.3% of new cases in females) and 21,000 disease-related deaths (24.2% of male cancer-related deaths and 25.8% of female cancer-related deaths) projected in 2021.¹ The adjusted 5-year net survival estimate in Canada for all forms of lung cancers is 22%¹ and the anticipated 5-year survival rate for patients with non–small cell lung cancer (NSCLC) is approximately 25% and 7% for patients with stage IV disease.² Smoking, which is an established risk factor for developing lung cancer, accounts for more than 72% of newly diagnosed cases in Canada.^1,3

Several treatments are available in practice for patients without prior testing for rearranged during transfection (RET) gene fusion. The drug plans and clinician input group highlighted the following treatment strategies: In patients who are treatment naive, first-line treatment combinations with platinum plus pemetrexed and pembrolizumab are preferred options for patients with programmed cell death ligand 1 (PD-L1) expression of less than 50%, and possibly in those with PD-L1 expression levels below 50% who are non-smokers, female, have increased disease, or symptom burdens. Pembrolizumab alone is preferred for those with PD-L1 expression below 50%. For patients who progressed on pembrolizumab as a first-line systemic therapy, treatment options involving platinum plus pemetrexed are preferred. Anti–PD-L1 therapies, including pembrolizumab, nivolumab, and atezolizumab, are available for patients who received platinum plus pemetrexed in the first-line setting, and docetaxel for those who progressed on platinum plus pemetrexed and pembrolizumab. These treatments were consistent with those highlighted by the clinical experts consulted by CADTH. The experts highlighted that the most preferred therapy used in the first-line setting for patients with RET fusion mutations across jurisdictions in Canada (except Prince Edward Island, where pembrolizumab is not funded) is the triple-therapy combination of platinum plus pemetrexed and pembrolizumab regardless of the PD-L1 tumour proportion score because of the limited activity reported in the literature for single-drug immunotherapy in the RET fusion population.

Selpercatinib (Retevmo) is a highly selective, adenosine triphosphate–competitive small-molecule inhibitor of the RET receptor tyrosine kinase, which is available in 2 oral formulations of 40 mg and 80 mg capsules. It received market authorization following the issuance of a Notice of Compliance with conditions (NOC/c) from Health Canada on June 16, 2021, for 3 indications: as a monotherapy in the treatment of metastatic RET fusion–positive NSCLC in adult patients, RET-mutant medullary thyroid cancer (MTC) in adult and pediatric patients 12 years of age and older with unresectable advanced or metastatic disease, and RET fusion–positive differentiated thyroid carcinoma in adult patients with advanced or metastatic disease (not amenable to surgery or radioactive iodine therapy) following prior treatment with sorafenib and/or lenvatinib. Initiation of treatment with selpercatinib is recommended only after testing for and confirmation of the RET gene mutation in patients. No past reviews have been submitted to CADTH for the RET fusion–positive NSCLC population.

The dosage recommended in the product monograph is 120 mg orally twice daily for patients who weigh less than 50 kg and 160 mg orally twice daily for patients who weigh 50 kg or more.

The objective of this CADTH review is to perform a systematic review of the efficacy and safety of selpercatinib as monotherapy for the first-line treatment of adult patients with metastatic RET fusion–positive NSCLC and as monotherapy for the treatment of adult patients with metastatic RET fusion–positive NSCLC who have received prior systemic therapy.

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

This section was prepared by CADTH based on the input provided by patient groups. The full patient group input is included in the Stakeholder Input section at the end of this report.

Three patient groups submitted input for the review. The Canadian Lung Cancer Advocacy Group Breathe Hope, the CanCertainty Coalition, and Lung Cancer Canada (LCC). A single respondent from the Canadian Lung Cancer Advocacy Group Breathe Hope. highlighted symptom burden management due to disease progression and treatment toxicity from chemotherapy as major drawbacks associated with the disease and available treatment options. Access to selpercatinib was considered valuable to the patient, who was willing to accept side effects from the treatment as a trade-off for reduced tumour growth.

Input provided by the CanCertainty Coalition highlighted potential limitations with access to treatments across jurisdictions in Canada, citing Ontario and the Atlantic provinces as jurisdictions where the level of access to oral cancer medications varies. The coalition recommended that CADTH examine equitable access to treatment across jurisdictions in Canada. They also cited potential issues associated with safety and the dispensing of take-home oral cancer treatments and recommended that these issues be considered during the review if the drug were to receive public funding.

In its input, LCC emphasized key concerns such as a lack of screening programs to detect disease in earlier stages. Lung Cancer Canada also noted the need for new treatments in the first and second line that improve patient-reported outcomes and overcome resistance to treatment, as well as the toxicity-related events, decreased functionality, and increased dependence of patients on caregivers associated with current chemotherapy and radiation therapy, which have provided less benefit to patients.

Clinician Input

Input From Clinical Experts Consulted by CADTH

Two clinical experts provided expert knowledge regarding treatment strategies in Canada. The clinical experts reported that no therapy is currently available for patients with RET fusion–positive NSCLC. Treatment goals identified by experts were similar to those highlighted by the clinician group. The clinical experts indicated that the most important goals are to achieve overall survival (OS), reduce symptom burden, delay disease progression, prolong life with improved quality of life, and decrease or eliminate hospital admissions and hospital stays. All these were considered valuable in this patient population.

The experts noted that if selpercatinib is approved for funding, it will likely be used as first-line therapy for patients with metastatic RET fusion–positive NSCLC. The experts described platinum plus pemetrexed and pembrolizumab (triplet therapy) as the preferred treatment option in the first line across jurisdictions (except the province of Prince Edward Island, where pembrolizumab is not funded) regardless of a patient’s PD-L1 tumour proportion score. Beyond the first line, the experts noted that docetaxel is funded and can be administered depending on whether the patient received triplet therapy as first-line treatment. Patients may also receive single-drug immunotherapies (pembrolizumab, nivolumab, or atezolizumab) in the second line if they had received platinum and pemetrexed in the first line; however, as noted in the clinical experts’ input, patients with RET fusion are known to respond poorly to immunotherapy, and docetaxel may be administered in place of an immunotherapy in next-line settings. The clinical experts recommended that treatment be made available to all RET fusion–positive patients with advanced or metastatic NSCLC.

The clinician experts indicated that response to treatment in practice is usually assessed using the same methods implemented in the LIBRETTO-001 trial. However, the frequency of assessments differs from those of the trial setting. As described by the experts, the standard will be to perform radiographic assessments every 8 to 12 weeks or sooner if the patient reports new symptoms or if physical findings indicate disease progression. The experts also noted that, in practice, symptom severity and adverse events (AEs) are generally reported every 3 to 4 weeks in patients receiving oral targeted therapies.

The experts noted that several molecular testing techniques are available to test RET fusion mutations across jurisdictions in Canada. Next-generation sequencing (NGS) was cited as the most commonly used technique, while NGS with RNA sequencing was considered the best test because of its sensitivity (100%) and specificity (99% using MSK IMPACT testing).

Clinician Group Input

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

The clinician group input was submitted by 2 groups: LCC and the Ontario Health (Cancer Care Ontario) Lung and Thoracic Cancers Drug Advisory Committee (DAC). Ontario Health drug advisory committees provide timely evidence-based clinical and health-system guidance on drug-related issues in support of Cancer Care Ontario’s mandate, the Provincial Drug Reimbursement Programs, and the Systemic Treatment Program. Twelve clinicians from LCC and 2 clinicians from the DAC provided input for this review. Both groups mentioned similar treatment goals for patients with advanced or metastatic NSCLC. Key goals noted by both clinician groups included: improvement in median OS in patients, rapid and prolonged improvement in symptoms, a median progression-free survival (PFS), and reduced toxicity-related AEs. Experts from LCC added that treatment goals include the prevention or treatment of brain metastases, reduction of resource utilization, and evaluation of the impact of COVID-19 on the safety of systemic therapy. Input from LCC emphasized that current treatments for patients with RET fusion–positive NSCLC have not improved OS in patients and are not associated with rapid and prolonged improvement in lung cancer symptoms as measured by median time to response, objective response rate (ORR), progressive disease rate, and median PFS.

Both groups mentioned that adding selpercatinib to the Canadian treatment paradigm will allow the drug to be administered as a first-line therapy in newly diagnosed patients with RET fusion–positive metastatic NSCLC. Clinicians in the LCC group noted that newly diagnosed patients with an Eastern Cooperative Oncology Group Performance Status (ECOG PS) of 0 to 3 could benefit from selpercatinib if it is approved for funding, although they expressed uncertainty about the best therapy for the second line and suggested that subsequent therapy could include docetaxel and anti–PD-L1 therapy for those who have not received such drugs in prior lines of therapy.

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. For the CADTH review of selpercatinib, the drug plans emphasized the different treatment strategies in place for patients with NSCLC and provided questions pertaining to the initiation of therapy, the prescribing of therapy, generalizability, funding algorithms, care provision, and system and economic issues. These questions were addressed by the clinician experts consulted for the CADTH review and their responses were based on the evidence presented by the sponsor in the LIBRETTO-001 trial. Clinician expert responses have been included in the Drug Program Input section (Table 4).

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

A single, ongoing, combined phase I and II, multi-centre, multi-cohort, open-label study met the criteria for the CADTH systematic review. The LIBRETTO-001 trial evaluated the safety and efficacy of selpercatinib in patients with advanced solid tumours, including RET fusion–positive solid tumours (e.g., NSCLC and thyroid, pancreatic, and colorectal cancers), RET-mutant MTC, and other tumours with RET activation (e.g., mutations in other tumour types or other evidence of RET activation). Patients recruited were 12 years or older (depending on the site and country).

The study was initiated in May of 2017 and has more than 84 participating centres across Australia, Canada, Denmark, France, Germany, Hong Kong, Israel, Italy, Japan, Singapore, Spain, South Korea, Switzerland, Taiwan, and the US. Three interim analyses were planned to support regulatory submissions in different jurisdictions. This review presents data obtained at the second and third interim data cut-offs of December 16, 2019, and March 30, 2020, respectively. Data obtained at interim 1 were updated at interim 2 and formed the basis of the Health Canada submission.

The study consisted of a dose-escalation phase (I) and a dose-expansion phase (II). The phase I portion was conducted initially as a 3 + 3 design (cohorts of 3 patients were assigned to increasing dose levels until 1 or more dose-limiting toxicities was observed) but was later updated to a Fibonacci dose-escalation design after the third escalation was implemented in patients in increments of approximately 67%, 50%, and 33%. The primary objective in phase I was to assess the maximum tolerable dose (MTD) and/or recommended phase II dose (RP2D) and any dose-limiting toxicities. The secondary objective at phase I was to evaluate the safety and tolerability of selpercatinib, characterize the pharmacokinetic (PK) properties, and assess the antitumour activity of selpercatinib.

The phase II portion, which is ongoing, has 5 cohorts, which include patients with a confirmed RET gene alteration in their tumours. Cohort 1 included patients with a RET fusion–positive solid tumour who progressed on or were intolerant to 1 or more prior standard first-line therapies. Cohort 2 was composed of patients with RET fusion–positive solid tumours without prior standard first-line therapy. The primary objective at phase II was to evaluate the antitumour activity of selpercatinib in patients recruited into the 5 cohorts. This was achieved by measuring the ORR using the Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST 1.1) or Response Assessment in Neuro-Oncology (RANO) guidelines, according to tumour type and as performed by an independent radiographic committee (IRC) and the study investigator. Other outcomes were assessed as secondary objectives in phase II. These included best change in tumour size from baseline, duration of response (DOR), central nervous system (CNS) ORR, CNS DOR, time to any and best response, clinical benefit rate, PFS, OS, safety and tolerability of selpercatinib, and the characterization of the PK properties. Health-related quality of life (HRQoL) was assessed as an exploratory outcome.

This CADTH review focuses on outcomes observed in patients with NSCLC with a confirmed RET fusion gene mutation enrolled in cohorts 1 and 2 at phase II of the LIBRETTO-001 trial. These patients were further subgrouped into 3 datasets based on clinically meaningful distinctions observed during the trial. These subgroups include the primary analysis set (PAS), which consists of the first consecutively enrolled patients previously treated with platinum-based chemotherapy; the integrated analysis set (IAS), which consists of patients treated with platinum-based chemotherapy; and the supplementary analysis sets (SASs), which consists of patient who are treatment naive (SAS1), patients treated with other systemic therapies that are not platinum-based (SAS2), and patients without measurable disease according to RECIST 1.1 (SAS3). These datasets supported the regulatory submission for marketing approval by Health Canada, the FDA, and the European Medicines Agency (EMA).

Efficacy Results

The key efficacy outcomes investigated in the LIBRETTO-001 trial are presented in Table 2. Data consist of findings obtained at interim analysis 2 (December 16, 2019) and interim analysis 3 (March 30, 2020).

Table 2: Summary of Key Results From the LIBRETTO-001 Trial (March 30, 2020, and December 16, 2019, Data Cut-Off Dates)

AE = adverse event; ALT = alanine transaminase; AST = aspartate transaminase; CI = confidence interval; IRC = independent radiographic committee; NE = not estimable; NR = not reported; NSCLC = non–small cell lung cancer; RET = rearranged during transfection; SAE = serious adverse event; SD = standard deviation.

Note: Data cut-off dates of December 16, 2019; March 30, 2020; and June 15, 2021.

Source: Minchom et al. (2021),⁷ Canadian product monograph, summary of product characteristics,⁴ Clinical Study Report,¹⁶ and sponsor’s submission package, Retevmo (selpercatinib).⁵

Overall Survival

Primary and integrated analysis sets (RET fusion–positive NSCLC with prior platinum chemotherapy): At the December 16, 2019, data cut-off, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

At the March 30, 2020, data cut-off, the median OS in the PAS was not estimable (NE) and the median follow-up was 19.94 months. In the IAS population, the median OS was NE, and the median follow-up was 14.26 months.

The sponsor conducted a follow-up analysis in the IAS population at a new cut-off date of June 15, 2021. The median OS was NE at this data cut-off. The median follow-up of survival was 26.4 months.

Treatment-naive RET fusion–positive NSCLC (SAS1): At the December 16, 2019, data cut-off, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

At the March 30, 2020, data cut-off, the median OS was NE and the median follow-up was 12.58 months.

The sponsor conducted a follow-up analysis in the treatment-naive population at a new cut-off date of June 15, 2021. The median OS was NE at this data cut-off. The median follow-up of survival was 21.9 months.

Prior other systemic therapy RET fusion–positive NSCLC (SAS2) and non-measurable disease RET fusion–positive NSCLC (SAS3): At the December 16, 2019, data cut-off, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

At the March 30, 2020, data cut-off, the median OS was 28.88 months (95% CI, 11.0 to NE) and the median follow-up was 17.05 months for prior other systemic therapy RET fusion–positive NSCLC (SAS2); the median OS was NE and the median follow-up was 10.48 months for non-measurable disease RET fusion–positive NSCLC.

Health-Related Quality of Life

The exploratory outcome of HRQoL was assessed in the entire NSCLC population (n = 253) using the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) disease-specific instrument. Scores were collected at baseline (cycle 1, day 1), every 8 weeks (until cycle 13), and every 12 weeks after cycle 13 (until end of treatment). A change of 10 or more points from baseline scores in the different domains was considered clinically meaningful.

No EORTC QLQ-C30 data were available at the March 30, 2020, data cut-off.

At the December 16, 2019, data cut-off, 3 subgroups were created, including the treatment-naive group (n = 39), the 1 prior line of therapy group (n = 64), and a group with 2 or more prior lines of therapy (n = 136). Lower scores obtained in functional subscales compared to the defined thresholds of 83 (physical function), 71 (emotional function), 58 (role function, social function), or 75 (cognitive function) were considered clinically meaningful problems for patients. Higher scores obtained in symptom subscales compared to the thresholds of 8 (nausea or vomiting), 39 (fatigue), 25 (pain), 17 (diarrhea, dyspnea, and financial difficulties), or higher than 50 (appetite loss, insomnia, and constipation) were considered clinically meaningful problems.

The mean score for global health status quality of life (GHS/QoL) at baseline in the overall population (all patients with NSCLC, n = 253) was 61.5. In the treatment-naive group, 1 line of prior therapy group, and 2 or more prior lines of therapy group, the obtained GHS/QoL mean scores were 60.2, 65.2, and 60.4, respectively.

The baseline scores for physical function in the overall population (all patients with NSCLC, n = 253) was 75.9. In the individual groups, the baseline scores for physical function were 72.6 in the treatment-naive group, 79.8 in the 1 prior line of therapy group, and 76.1 in the 2 or more prior lines of therapy group). As these were lower than the threshold score of 83, they were considered clinically important impairments at baseline. Improvements (≥ 10-point increase) in physical function were reported at cycle 3 in all 3 groups (naive = 43.5%; 1 prior line of therapy = 28.1%; and ≥ 2 prior lines of therapy = 29.8%).

Baseline scores for dyspnea exceeded the clinically meaningful threshold of 17 points in the overall population (all patients with NSCLC, n = 253) and in each subgroup (overall = 31.3; treatment-naive = 28.4; 1 prior line of therapy = 23.1; ≥ 2 prior lines = 37.7; standard deviation [SD] = 28.3) and were therefore considered clinically meaningful impairments. The proportion of patients who experienced a change in dyspnea from baseline by cycle of study treatment was higher in patients who reported improved symptoms compared to patients who reported worsened symptoms across cycles 3 to 13.

Because baseline scores for fatigue and insomnia did not meet a clinically meaningful threshold in the overall population (all patients with NSCLC, n = 253), they were not considered clinically meaningful impairments. The threshold was exceeded in the treatment-naive group (baseline mean = 41.6) and the group with 2 or more prior lines of therapy (baseline mean fatigue = 41.8), and these impairments were also considered clinically meaningful. The threshold was not met in the 1 prior line of therapy group, and was not considered a clinically meaningful impairment. More patients experienced improved outcomes in the change in baseline by cycle in insomnia scores compared to those who reported worsened outcomes across cycles 3 to 13. However, due to the decrease in the number of patients completing the questionnaires from baseline to cycle 13, these findings are uncertain. Data for the change from baseline by cycle in fatigue scores were not available.

Baseline scores for pain met a clinically meaningful threshold of 25 points in the overall population (mean = 29.4) and in all subgroups, and was therefore considered a clinically meaningful impairment. All line of therapy subgroups for nausea and vomiting exceeded the clinically meaningful threshold of 8 points. Data for the change from baseline by cycle in pain scores was not available.

Progression-Free Survival

Primary and integrated analysis sets (RET fusion–positive NSCLC with prior platinum chemotherapy): At the December 16, 2019, data cut-off, the median PFS in the PAS as assessed by the IRC was 16.53 months (95% confidence interval [CI],13.7 to NE). In the IAS population, the median PFS was 19.32 months (95% CI, 13.9 to NE).

At the March 30, 2020, data cut-off, the median PFS in the PAS as assessed by the IRC was 19.3 months (95% CI,13.9 to NE). In the IAS population, the median PFS was 19.3 months (95% CI, 16.5 to NE).

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The median PFS estimated in the IAS population was 24.94 months (95% CI, 19.3 to NE).

Treatment-naive RET fusion–positive: At the December 16, 2019, data cut-off, the median PFS by IRC assessment was NE and the median follow-up estimated was 9.17 months.

At the March 30, 2020, data cut-off, the median PFS by IRC assessment was NE and the median follow-up estimated was 10.84 months.

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The median PFS estimated in the treatment-naive population was 21.95 months (95% CI, 13.8 to NE).

Prior other systemic therapy RET fusion–positive NSCLC (SAS2) and non-measurable disease RET fusion–positive NSCLC (SAS3): At the December 16, 2019, data cut-off, the median PFS ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

At the March 30, 2020, data cut-off, the median PFS by IRC assessment was NE in both groups.

Objective Response Rate

Primary and integrated analysis sets (RET fusion–positive NSCLC with prior platinum chemotherapy): At the December 16, 2019, data cut-off, the ORR estimated by the IRC was 64% (95% CI, 53.9 to 73.0) in the PAS population, and in the IAS population, the ORR was 57% (104 of 184; 95% CI, 49.0 to 63.8).

At the March 30, 2020, data cut-off, the ORR by IRC was 63.8% (95% CI, 53.9 to 73.0) and 56.9% (95% CI, 50.0 to 63.6) in the PAS and IAS, respectively.

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The ORR estimated in the IAS population was 61.1% (95% CI, 54.7 to 67.2), which was consistent with previous analysis.

Treatment-naive RET fusion–positive NSCLC (SAS1): At the December 16, 2019, data cut-off, the ORR estimated by the IRC for the treatment-naive RET fusion–positive NSCLC population was 84.6% (95% CI, 69.5 to 94.1).

At the March 30, 2020, data cut-off, the ORR by IRC assessment was 85.4% (95% CI, 72.2 to 93.9).

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The ORR estimated in the treatment-naive population was 84.1% (95% CI, 73.3 to 91.8), which was consistent with previous analysis.

Prior other systemic therapy RET fusion–positive NSCLC (SAS2) and non-measurable disease RET fusion–positive NSCLC (SAS3): The ORR obtained by the IRC assessment in the prior other systemic therapy RET fusion–positive NSCLC (SAS2) was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| at the December 16, 2019, analysis date and 44.4% (8 of 18; 95% CI, 21.5 to 69.2) at the March 30, 2020, data cut-off.

The ORR obtained by the IRC assessment in the non-measurable disease RET fusion–positive NSCLC (SAS3) was |||||||||||||||||||||||||||||||||||||||| at the December 16, 2019, data cut-off and 33.3% (95% CI, 13.3 to 59.0) at the March 30, 2020, data cut-off.

Subgroup Analysis

Central Nervous System Metastasis at Baseline

At the December 16, 2019, data cut-off, the IRC identified 22 patients with measurable CNS disease out of the 80 patients with CNS metastasis at baseline. The CNS ORR in the 22 patients was 82% (18 of 22; 95% CI, 59.7 to 94.8). The CNS ORR in the 80 patients with measurable and non-measurable disease was 48% (38 of 80; 95% CI, 36.2 to 59.0).

At the March 30, 2020, data cut-off, 23 patients of 96 were assessed with measurable disease at baseline, and the CNS ORR was 87% (95% CI, 66.5 to 97.2). In the 96 patients with CNS disease at baseline (measurable and non-measurable disease), the CNS ORR was 46.9% (95% CI, 36.6 to 57.3).

Performance Status, Number of Prior Therapies, Prior Anti–Programmed Cell Death Protein 1 and Anti–PD-L1 Therapy, and Prior Multikinase Inhibitor Therapy

At the December 16, 2019, data cut-off, the ORRs in the following subgroups were:

• ECOG PS of 0: ORR 74.2% (95% CI, 55.4 to 88.1); ECOG PS of 1 or 2: ORR 59.5% (95% CI, 7.4 to 70.7)

• number of prior therapies: 1 to 2: ORR 58.7% (95% CI, 43.2 to 73.0); 3 or more: ORR 67.8% (95% CI, 54.4 to 79.4)

• prior anti–programmed cell death protein 1 (PD-1) or anti–PD-L1 therapy: yes: ORR 65.5% (95% CI, 51.9 to 77.5); no: ORR 61.7% (95% CI, 46.4 to 75.5)

• prior multikinase inhibitor therapy: yes: ORR 64.0% (95% CI, 49.2 to 77.1); no: ORR 63.6% (95% CI, 49.6 to 76.2).

No ORR subgroup data were available at the March 30, 2020, data cut-off.

Duration of Response

Primary and integrated analysis sets (RET fusion–positive NSCLC with prior platinum chemotherapy): At the December 16, 2019, data cut-off, a median DOR as assessed by the IRC of 17.5 months (95% CI, 12.0 to NE) was reported, with a median DOR follow-up of 12.1 months in the PAS population. In the IAS, a median DOR of 17.5 months (95% CI, 12.1 to NE) by IRC assessment, with an estimated median DOR follow-up of 9.2 months, was reported.

At the March 30, 2020, data cut-off, the median DOR as assessed by the IRC was 17.51 months (95% CI, 12.1 to NE) in the PAS and a median DOR by IRC of 17.51 months (95% CI, 12.1 to NE) in the IAS population.

The sponsor conducted a follow-up analysis at a new data cut-off date of June 15, 2021. The median DOR obtained in the IAS population was 28.6 months (95% CI, 20.4 to NE).

Treatment-naive RET fusion–positive NSCLC: At the December 16, 2019, data cut-off, the DOR was assessed by the IRC in 33 patients in the treatment-naive cohort. The median DOR was NE.

At the March 30, 2020, data cut-off, 65% of the responses were ongoing at 12 months by IRC assessment.

The sponsor conducted a follow-up analysis at a new data cut-off date of June 15, 2021. The median DOR estimated in the treatment-naive population was 20.2 months (95% CI, 13.0 to NE).

Prior other systemic therapy RET fusion–positive NSCLC (SAS2) and non-measurable disease RET fusion–positive NSCLC (SAS3): At the December 16, 2019, data cut-off, the median DOR was ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

At the March 30, 2020, data cut-off, the DOR was NE in both groups.

Harms Results

At the December 16, 2019 data cut-off, there were 329 patients in the NSCLC set and 702 patients in the overall safety set (OSAS). Overall, 98.8% of patients in the NSCLC set and 99.0% in the OSAS reported at least 1 adverse event (AE); 35.9% of patients in the NSCLC set and 33.3% in the OSAS reported at least 1 serious AE; and 6.4% patients in the NSCLC set versus 5.3% in the OSAS discontinued study due to an AE. At the March 30, 2020, cut off date, 746 patients were included in the OSAS, 99.2% reported at least 1 AE, 35.1% reported a serious AE, and 45 (6%), discontinued the study due to an AE.

At the December 16, 2019, data cut-off, the most common AEs reported in the patients with NSCLC population in the LIBRETTO-001 trial were dry mouth, diarrhea, hypertension, increased aspartate transaminase (AST), increased alanine transaminase (ALT), fatigue, constipation, peripheral edema, headache, and nausea. Serious events in the NSCLC population were commonly associated with pneumonia, increased ALT and AST, abdominal pain, pleural effusion, drug hypersensitivity, diarrhea, and acute kidney injury.

At the December 16, 2019, data cut-off, AEs leading to dose withdrawal, interruption, and dose reductions in the RET fusion–positive NSCLC population were consistent with those in the overall population. AEs commonly associated with treatment discontinuations in the NSCLC population were increased ALT, sepsis, increased AST, drug hypersensitivity, fatigue thrombocytopenia. Increases in ALT and AST were commonly associated with dose reductions and dose interruptions.

At the December 16, 2019, data cut-off, treatment was discontinued due to death in 6 patients, and 38 patients discontinued the study due to death. At the March 30, 2020, data cut-off, in the NSCLC population, treatment was discontinued due to death in 6 (1.7%) patients, and 55 (15.9%) patients discontinued the study due to death. In total, 36 patient deaths (10.4%) were attributed to disease progression, 13 deaths (3.9%) occurred due to AEs, and 6 deaths (1.7%) were attributed to other reasons. A single report of death (0.3%), which had occurred more than 28 days after the last selpercatinib dose, was identified in the NSCLC population. In the overall safety analysis set (OSAS) population, treatment was discontinued due to death in 11 patients (1.5%), and 103 patients (13.8%) discontinued the study due to death.

Notable harms reported in the NSCLC set were consistent with the OSAS at the March 30, 2020, and December 16, 2019, data cut-offs). At the December 2019 data cut-off, notable harms reported in the NSCLC set and the OSAS included electrocardiogram QT prolongation (16.7% versus 16.5%), increased AST (32.8% versus 29.9%), increased ALT (31.0% versus 28.6%), hypertension (31.9% versus 35.9%), and hypersensitivity (2.4% versus 4.3%), respectively. At the March 30, 2020, data cut-off, notable harms reported in the OSAS population included electrocardiogram QT prolongation (18.1%), increased AST (32.6%), increased ALT (32.6%), hypertension (37.4%), hypersensitivity (5.2%), and hemorrhage (2.4%).

Critical Appraisal

The open-label, non-comparative design of the LIBRETTO-001 trial, with no statistical testing, is the key limitation. In its statistical analysis plan, the sponsor did not provide hypothesis statements for statistical significance for the primary outcomes of the secondary and subgroup analyses.⁸ This design increases the risk of bias in estimating treatment effects because the potential for confounding effects related to variation in health status, and other unidentified prognostic factors could affect subjectively assessed outcomes (i.e., response, HRQoL, and AEs). The potential for bias was reduced by using an IRC assessment for key study outcomes such as ORR and DOR.

The lack of direct comparative data means there is uncertainty regarding the magnitude of effects obtained for the efficacy outcomes. Although the clinical experts consulted by CADTH emphasized that the safety profile of selpercatinib was favourable compared to the other therapies available as standard of care in Canada, in the absence of a comparative arm, the findings obtained from the safety analysis are uncertain, as the single-arm design does not allow for the differentiation of the symptoms of underlying NSCLC disease from treatment-related AEs.⁹ The sponsor agreed to provide results from the |||||||||||||||||||||||||||||||||||||||||||||| to confirm the clinical benefit of selpercatinib in patients with previously treated RET fusion–positive NSCLC according to the NOC/c issued by Health Canada. The sponsor had noted in its response that |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The sponsor submitted additional data with a new cut-off date of June 15, 2021, during the completion of this CADTH review. (Appendix 3 provides new data). Despite the results obtained from the updated (pre-specified or post hoc) analyses performed on June 15, 2021, whether the end points investigated are durable for the long-term in this patient population remains uncertain.

The primary objective investigated at the phase II portion of the LIBRETTO-001 study was the ORR as measured by RECIST 1.1. The FDA considers ORR alone a surrogate measurement when assessing treatment response in patients with advanced or metastatic NSCLC, and this may not correlate well with survival, unless the effect size of the ORR is large and the responses are durable. The sponsor hypothesized a true ORR of 50% or greater in the primary analysis of effectiveness, ruling out a lower limit of 30% for the ORR, which was considered clinically meaningful and consistent with the estimated response rates seen with approved targeted therapies in molecularly defined populations of patients who have failed prior therapies (e.g., osimertinib, crizotinib, alectinib). The ORR obtained by IRC in the PAS, IAS, and SAS1 sets were above the lower limit of 30% that the sponsor assumed in the sample size calculation for patients with RET fusion–positive NSCLC who progressed on or after receipt of platinum-based chemotherapy.⁵ Although the FDA review team noted that the magnitude of the ORR and DOR obtained in patients with RET fusion–positive NSCLC of the LIBRETTO-001 trial was large, and considered it sufficient to establish clinical benefit,¹⁰ |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.⁹ The sponsor |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||to provide results from |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| to confirm the clinical benefit of selpercatinib in patients with previously treated RET fusion–positive NSCLC according to the NOC/c issued by Health Canada. The sponsor also noted that ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.⁹ In the LIBRETTO-001 trial, radiographic scans were performed by an accredited laboratory and assessed by an independent radiology committee, which reduced bias. Results obtained from both IRC and investigator assessments did not differ greatly, which increases the validity of the ORR-related outcomes.

The time-to-events analyses, particularly the OS and PFS results, were considered exploratory by |||||||||||||||||, the FDA,^25,30 and CADTH due to the lack of a control arm. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||⁹ Due to immature data in the PAS, IAS, and SAS1 populations at the March 31, 2020, and December 16, 2019, cut-offs, the results were considered exploratory by CADTH. The sponsor submitted additional data to CADTH for a new cut-off date of June 15, 2021, later during the completion of the review. Although the sponsor reported a median PFS of 24.94 months (95% CI, 19.3 to NE) in the IAS and a median PFS of 21.95 months (95% CI, 13.8 to NE) in the SAS1 population, the information was considered insufficient to form concrete conclusions about PFS in this population because of the single-arm design of the LIBRETTO-001 trial and immature data. The median OS was NE in the IAS and SAS1 datasets at the June 15, 2021, cut-off, creating uncertainty about whether the observed magnitude of benefit related to tumour response with selpercatinib would be translated as OS in patients in the 2 groups.

The analysis sets — the PAS, IAS, and 3 SASs — were not predefined in the original statistical analysis plan; they were developed following consultation with the FDA and EMA. A key concern is that these were post hoc analyses and may have been susceptible to bias. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. CADTH reviewers agreed ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||, these were unlikely to introduce bias because the investigators remained blinded to results until after the revisions were made.⁹

The 3 planned interim analyses were pre-specified. However, the analyses were carried out using observed data only, with no formal hypothesis testing and only descriptive statistics provided, further complicating assessments of the magnitude of the effect observed in the different groups.

The exploratory outcome of HRQoL was evaluated. The validity of the findings is uncertain because the number of patients who completed questionnaires decreased from baseline through to cycles 13, resulting in considerable missing data at later time points. In the absence of a comparator arm and given an open-label design that introduced reporting bias, the impact of selpercatinib on patient-reported outcomes relative to other therapies is unknown. Although CADTH recognizes that the rarity of RET fusion–positive mutations in NSCLC may have contributed to the small sample size in the datasets at baseline, and may have influenced the number of patients available to complete the questionnaires at later stages of the trial, no strong, definitive conclusions can be made from the findings obtained for HRQoL in the different population sets of patients with NSCLC.

The clinical experts consulted during this CADTH review indicated that the findings obtained for the ORR, DOR, CNS ORR, and HRQoL outcomes investigated in the LIBRETTO-001 study are clinically meaningful for patients in practice. The LIBRETTO-001 trial recruited patients with an ECOG PS of 0, 1, or 2. The clinician experts consulted considered these findings generalizable to patients with an ECOG PS of 0 to 3 (but not for patients with an ECOG PS 4). As the experts also considered the baseline findings obtained in the trial to be similar to those observed in practice, the findings are generalizable to patients in Canada. The RET fusion mutation in patients was identified in the LIBRETTO-001 trial using polymerase chain reaction (PCR) testing and NGS. The clinical experts noted that NGS is available across several jurisdictions in Canada for testing oncogenic driver mutations at initial diagnosis.

Indirect Comparisons

Description of Studies

Two sponsor-submitted indirect treatment comparisons (ITCs) were summarized and critically appraised. Both aimed to evaluate the clinical efficacy of selpercatinib relative to other active treatments for RET fusion–positive NSCLC in patients with¹¹ or without prior systemic therapies.¹² All included studies enrolled patients with unknown RET fusion status, with the exception of LIBRETTO-001. The 3 outcomes that were analyzed were OS, PFS, and ORR.

Efficacy Results

The sponsor-submitted ITCs conducted a systematic review to identify relevant individual studies and used Bayesian network meta-analysis to evaluate the clinical efficacy of selpercatinib relative to other treatments for NSCLC. In both ITCs, a pseudo-control arm was needed due to the lack of a comparison arm in the study of selpercatinib. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

In the ITC of a treatment-naive population, selpercatinib was favoured over other treatments for OS (||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||), PFS (|||||||||||||||||||||||||||||||||||||||||||||||||||) and ORR (||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||). Selpercatinib was compared to monotherapy or a combination of platinum-based chemotherapy and immunotherapy drugs.

In the ITC of a population with treatment experience, results suggested that selpercatinib was favoured over other treatments for OS (|||||||||||||||||||||||||||||||||||||||||||||||||||), PFS (||||||||||||||||||||||||||||||||||) and ORR (|||||||||||||||||||||||||||||||||||||||||||||||||||). Selpercatinib was compared to monotherapy or combination of chemotherapy drugs and immunotherapy drugs, such as docetaxel, cabozantinib, atezolizumab, and nivolumab.

Harms Results

Harms were not assessed in the sponsor-submitted ITCs.

Critical Appraisal

These ITCs have a number of limitations that affect their internal and external validity, such as not being able to comprehensively assess the clinical heterogeneities across the included individual studies and their influence on the study results. Due to the lack of reporting certain patient characteristics, the treatment effect of selpercatinib, despite various adjustments, remains uncertain, and the generalizability of the study findings to patients with RET fusion–positive could be limited. In addition, other important outcomes, such as DOR, HRQoL and safety, were not assessed.

Other Relevant Evidence

Description of Studies

LIBRETTO-431: The CADTH review team identified an ongoing phase III, randomized, open-label study (LIBRETTO-431) comparing selpercatinib to platinum-based (carboplatin or cisplatin) and pemetrexed therapy with or without pembrolizumab in patients who are treatment naive and have locally advanced and/or metastatic RET fusion–positive non-squamous NSCLC. No results are currently available, as this trial is actively recruiting patients. The estimated primary completion date (at which the last participant in a clinical study will be examined or receive an intervention to collect final data for the primary outcome measure) and study completion date (when the last participant in a clinical study will be examined or receive an intervention or treatment to collect final data for the primary outcome measures, secondary outcome measures, and AEs) are January 15, 2023, and August 18, 2025, respectively.

SIREN: The CADTH review team identified another study analyzing the safety and efficacy of selpercatinib in a real-world setting (SIREN), in which the data were retrospectively collected and analyzed from patients with RET fusion–positive NSCLC participating in a selpercatinib access program.

The ORR, defined as a complete response (CR) or partial response (PR), was 68% (95% CI, 53 to 81), and the median PFS was 15.6 months (95% CI, 8.8 to 22.4) after a median follow-up of 9.4 months among all patients. In patients with untreated or previously progressed and measurable brain metastases (n = 8), the intracranial ORR reached 100%. In terms of AEs, 43 (88%) of 50 patients experienced treatment-related AEs of any grade, a large majority of them grade 1 or 2. The most frequent treatment-related AEs reported were fatigue or asthenia (40%), increased liver enzyme levels (34%), hypertension (26%), dry mouth (26%), and peripheral edema (20%). Treatment-related AEs of grade 3 or higher were reported in 12 patients (24%), with the most common being increased liver enzyme levels (10%), abdominal pain (4%), prolonged corrected QT time (4%), hypertension (4%), and fatigue or asthenia (4%).¹³

The following limitations were identified. First, the retrospective study design is prone to bias (e.g., reporting bias and nondifferential biases) and the patient population recruited may not be similar to that of the LIBRETTO-001 trial due to differences in the eligibility and exclusion criteria applied in the study (potential selection bias). The ORR, although measured using RECIST 1.1, was assessed by an unblinded review of practising physicians. There is also a potential measurement bias due to differences in the frequency and conduct of disease assessments in clinical practice versus the trial setting, the follow-up time frame in the study (which differs from that of the trial), and the therapies administered beyond disease progression. The small sample size of the study also limits the generalizability of the findings. Although the SIREN study provides additional data on the effectiveness and safety of selpercatinib in the real-world setting, the limitations identified introduces uncertainty.

Conclusions

The evidence supporting the funding request of selpercatinib was derived from an ongoing phase I and II, open-label, non-randomized, multi-cohort, single-arm study (LIBRETTO-001). The ORR observed in the LIBRETTO-001 trial suggested favourable tumour response in both the treatment-naive and treatment-experienced groups and was consistent with further follow-up analyses. The ORR and DOR, including the CNS ORR, obtained in both patient populations were considered clinically meaningful by the clinical experts consulted during the review. Time-to-event end points such as OS were NE at the March 30, 2020, and December 16, 2019, data cut-offs in the PAS, IAS, and SAS1 populations due to data immaturity. The median PFS was NE in the treatment-naive group at the March 30, 2020, and December 16, 2019, data cut-offs. Combined with the single-arm trial design, the evidence was considered insufficient to interpret OS and PFS findings. The sponsor provided additional data to CADTH for a new data cut-off of June 15, 2021. Although estimates obtained at the June 15, 2021, data cut-off suggested an improvement in median PFS in the IAS and SAS1 populations, the median OS was NE. CADTH considered these findings insufficient to provide concrete conclusions on the comparative treatment effect (PFS and OS) due to the single-arm trial design and immature data. It is therefore uncertain whether the observed magnitude of benefit related to tumour response with selpercatinib would be translated to OS in patients in the treatment-naive and treatment-experienced groups. As well, the limitations related to the single-arm, non-randomized design of the LIBRETTO-001 trial precluded drawing strong, definitive conclusions on the effects of selpercatinib on HRQoL, although HRQoL findings were described by the clinical experts consulted by CADTH as clinically meaningful. Safety information was reported for all patients who received a single dose of selpercatinib in the LIBRETTO-001 trial at both data cut-offs. Selpercatinib was associated with corrected QT prolongation, increased AST and ALT, hypertension, and drug hypersensitivity. These events have been labelled under the warnings and precautions section of the Canadian approved product monograph for selpercatinib. However, these notable harms were considered by the clinical experts as manageable and favourable compared to current standard of care treatment options. Although the SIREN study provided additional data on both the effectiveness and safety of selpercatinib in the RET fusion–positive NSCLC population, several limitations were identified with the study, and concrete conclusions could not be drawn to support the primary data obtained from the LIBRETTO-001 trial. The ITCs submitted to provide information on the comparative effects of selpercatinib were also associated with limitations that prevented drawing conclusions. The comparative effectiveness and safety of selpercatinib are therefore uncertain.

Introduction

Disease Background

Lung cancer is the most frequently diagnosed cancer in Canada and the leading cause of cancer-related deaths in males and females,¹ with more than 29,600 new diagnoses (12.5% of new cases in males and 13.3% of new cases in females) and 21,000 disease-related deaths (24.2% of male cases and 25.8% of female cases) projected in 2021.¹ The adjusted 5-year net survival estimate in Canada for all forms of lung cancers is 22%¹ and the anticipated 5-year survival for patients with NSCLC is approximately 25%, and 7% for patients showing stage IV disease.² Smoking is an established risk factor for developing lung cancer, accounting for more than 72% of newly diagnosed cases in Canada.^1,3

Lung cancers are classified into 2 types based on histology: small cell lung cancer and NSCLC. The latter are the most common, and are further categorized into 3 types based on cell types: adenocarcinomas, squamous cell carcinomas, and large-cell carcinomas. Adenocarcinomas are the most commonly diagnosed forms of NSCLC in Canada, accounting for 48% of new cases.¹

Early diagnosis improves prognosis and patient responsiveness to therapy. Diagnosis is based on histology and symptom presentation.^3,14 Patients may experience worsening coughs, chest pain, hemoptysis, malaise, weight loss, dyspnea, and hoarseness at clinical presentation or upon chest imaging.^1,3 In advanced or metastatic disease, patients experience additional symptom burdens such as trouble breathing, chronic cough and chest pain, pain in bones or the spine, yellowing of the skin or eyes, weakness or numbness of arms or legs, fatigue and unexplained weight loss depression, insomnia, and pain.^15,16 Staging at diagnosis is key in determining disease prognosis and facilitates treatment selection.^3,16 Late diagnosis is a significant contributing factor to early mortality and also challenging for disease management in real-world practice. Unfortunately, almost 50% of NSCLC diagnoses in Canada are made at stage IV, with only 23.1% of cases diagnosed at stage I.¹

The expression of oncogenic driver mutations on tumours plays a vital role in patient response to treatment.¹⁶ Several predictive drivers identified in recent years, including mutations of the EGFR (epidermal growth factor receptor), ROS1 (C-ROS oncogene 1), KRAS, and BRAF genes, ALK (anaplastic lymphoma kinase) fusions, and others have greatly influenced treatment strategies in practice, improved patient quality of life, and increased OS for patients.^5,16-18 The RET protein, a transmembrane tyrosine kinase receptor encoded by the RET gene, is known to play a substantial role in the development and maintenance of many systems (including the enteric nervous and genitourinary systems in neonates).¹⁹ Abnormal RET receptor activation by rearrangement or mutation was recognized as an oncogenic driver for many cancers, including NSCLC. These alterations were commonly associated with patients with adenocarcinoma histology, younger patients (usually ≤ 60 years), and those with a non-smoking or light smoking status.⁵ Prevalence estimates from studies show that only about 1% to 2% of NSCLC cases are RET fusion–positive.^5,20 Testing for driver mutations at initial diagnosis using molecular techniques such as NGS panel testing or PCR testing methods is available across jurisdictions in Canada.^16,17

Standards of Therapy

The clinician experts and clinician groups consulted during the review outlined similar treatment goals for patients with advanced or metastatic disease, which include improvement in median OS, rapid and prolonged improvement in cancer-related symptoms and improvement in quality of life (given that patients with advanced and metastatic disease experience greater symptom burden), reduced treatment-related toxicity and prevention, and treatment of brain metastasis.

Expert opinion from the clinician groups and drug plans consulted during the CADTH review highlighted treatment combinations funded in practice for patients without confirmed RET fusion. For the naive-treatment population, first-line treatment combinations with platinum plus pemetrexed and pembrolizumab were identified as the most preferred in patients with PD-L1 expression below 50% — and possibly in those with PD-L1 expression above 50% who are non-smokers, female, have increased disease, or symptom burdens — and pembrolizumab alone for those with PD-L1 expression above 50%. Among patients who progressed on prior systemic therapy, treatment options with platinum plus pemetrexed are preferred for those who had received pembrolizumab in the first-line therapy. Anti–PD-L1 therapy, including pembrolizumab, nivolumab, and atezolizumab, is recommended for those who had received platinum plus pemetrexed as first-line therapy (a small number of patients), and docetaxel is advised for those who have progressed on platinum plus pemetrexed and pembrolizumab.

The drug plans consulted by CADTH identified several treatment options with the potential for funding for first-line treatment, including pembrolizumab and atezolizumab (depending on the patient’s PD-L1 status). In the second-line setting, the drug plans noted that funded options may include immune checkpoint inhibitors, if no prior PD-1 inhibitor (pembrolizumab, nivolumab, or atezolizumab, depending on patient’s PD-L1 status) was administered to the patient, or chemotherapy following treatment with a PD-1 inhibitor (docetaxel or pemetrexed). The LCC clinician group pointed out that evidence from some cohort studies indicates that patients with RET fusion NSCLC are sensitive to pemetrexed, and that, in the absence of any randomized data, pemetrexed plus platinum will likely be the most efficacious therapy in patients with RET fusion NSCLC who had received only pembrolizumab as first-line therapy.

The clinician experts consulted identified treatments similar to those outlined by the clinician group and drug plans. The experts mentioned in their report that the most preferred therapy used in the first-line setting across jurisdictions in Canada (except Prince Edward Island, were pembrolizumab is not funded) is a triplet therapy combination of platinum plus pemetrexed and pembrolizumab, regardless of the PD-L1 tumour proportion score. This choice of therapy was based on available evidence that reported similar outcomes (PFS and OS). While they acknowledged that platinum doublets and single-drug immunotherapies are approved and available in practice, they noted that some studies have reported poor response rates to immunotherapy in the RET fusion population. The experts indicated in their input that patients with RET fusion (most likely with adenocarcinoma histology) have shown to respond to a combination of platinum and pemetrexed, but other platinum doublets are of limited significance.

The clinical experts also added that, beyond the first line (after administration of a triple therapy), single drugs such as docetaxel are the typical standard of care, although there is limited evidence for outcomes specific to patients with RET fusion using docetaxel. If a patient received pembrolizumab in the first line, the doublet combination of platinum and pemetrexed may be administered and if they received platinum plus pemetrexed in the first line, they may likely receive immunotherapy (e.g., pembrolizumab, nivolumab, or atezolizumab) in the second line. However, patients tested for RET fusion may be placed on docetaxel in the second line rather than immunotherapy (based on evidence that reports low response rates to immunotherapy in patients who are RET fusion positive). The experts also added gemcitabine and vinorelbine as available therapies in the second line and beyond. Other nonsystemic options outlined included radiation and surgical interventions employed as aggressive modalities in patients with oligometastatic disease or as palliative interventions with the goal of alleviating symptoms.

Drug

Selpercatinib is a highly selective, adenosine triphosphate–competitive small-molecule inhibitor of the RET receptor tyrosine kinases available in 2 oral formulations of 40 mg and 80 mg capsules. It received market authorization following the issuance of an NOC/c from Health Canada on June 16, 2021, for 3 indications: as a monotherapy in the treatment of metastatic RET fusion–positive NSCLC in adult patients, RET-mutant MTC in adult and pediatric patients 12 years of age and older with unresectable advanced or metastatic disease, and RET fusion–positive differentiated thyroid carcinoma in adult patients with advanced or metastatic disease (not amenable to surgery or radioactive iodine therapy) following prior treatment with sorafenib and/or lenvatinib.⁴ The sponsor is requesting funding for the following indications:

• as monotherapy for the first-line treatment of adult patients with metastatic RET fusion–positive NSCLC

• as monotherapy for the treatment of adult patients with metastatic RET fusion–positive NSCLC who have received prior systemic therapy.

Treatment initiation with selpercatinib is recommended only after an initial testing and confirmation of the RET gene mutation in patients. Selpercatinib has obtained regulatory approval and is marketed by other regulatory agencies, including the FDA (May 08, 2020) and the EMA (February 11, 2021) for indications similar to those outlined in the Canadian product monograph. Market approval in Canada was granted based on evidence generated from the LIBRETTO-001 trial, a phase I and II trial conducted in patients 18 years and older (in some countries in patients as young as 12) with advanced solid tumours, including RET fusion–positive solid tumours (e.g., NSCLC and thyroid, pancreatic, and colorectal cancer), RET-mutant MTC, and other tumours with RET activation (e.g., mutations in other tumour types or other evidence of RET activation).¹⁹

Dosing recommendations of selpercatinib are based on body weight. Patients weighing less than 50 kg are recommended to take 120 mg twice a day and patients with a body weight of 50 kg and above are recommended to take 160 mg of capsules twice a day every 12 hours. Dose adjustments are recommended for patients with severe hepatic impairment (a Child-Pugh score with a severity in the C range) irrespective of body weight.

Mechanism of Action

Selpercatinib demonstrated potent in vitro and in vivo activity as a selective inhibitor of both wild-type and oncogenically activated RET, including RET fusions, “founder” mutations, and anticipated acquired resistance mutations. The nonclinical development program also showed that selpercatinib treatment resulted in significant cytotoxicity only in human cancer cell lines that harbour endogenous RET gene alterations (e.g., fusions and mutations), with minimal cytotoxicity in human cancer cell lines without an endogenous RET gene alteration, as expected for a highly specific inhibitor of RET.⁵

Selpercatinib was found to be more than 250-fold more selective for RET than 98% of 329 non-RET kinases tested in a large in vitro screen. This high degree of selectivity was maintained against both kinase and non-kinase off-targets when validated in additional enzyme, cell-based, radio-ligand binding, and in vivo assays.⁵

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH based on input provided by patient groups. The full patient group input is included in the Stakeholder Input section at the end of this report.

Three patient groups submitted input for the review: the Canadian Lung Cancer Advocacy Group Breathe Hope, CanCertainty, and LCC.

The Canadian Lung Cancer Advocacy Group Breathe Hope is composed of Canadian lung cancer patients. The group is driven by the need to establish a Canadian lung cancer community of patients and caregivers; provide a supportive, confidential environment; raise awareness about lung cancer through personal contacts or events and connections with national, provincial, and local stakeholder individuals or organizations; share new, innovative lung cancer treatments and scientific research; and promote conference opportunities and web events. The input provided was based on a single patient’s experience with standard of care treatment options offered in Canada and selpercatinib.

The CanCertainty Coalition is a united voice of more than 30 Canadian patient groups, cancer health charities, and caregiver organizations from across the country working with oncologists and cancer care professionals to significantly improve the affordability and accessibility of cancer treatment.

Lung Cancer Canada is a registered national charitable organization that serves as Canada’s leading resource for lung cancer education, patient support, research, and advocacy. A member of the Global Lung Cancer Coalition, LCC is the only organization in Canada focused exclusively on lung cancer. Its input was based on 5 patient experiences. Three respondents lived in Canada and 2 patients lived in the US. All patients had completed RET fusion gene testing and had received selpercatinib treatment either in the first- or second-line setting. Patient information was collected through interviews (n = 4), questionnaires, and previous discussions with LCC (n = 1).

Most patient respondents from LCC (3 of 5) had been treated with chemotherapy, immunotherapy, or radiation before receiving selpercatinib. Two patients had received selpercatinib in the first-line setting. Patients with prior experience with other lines of therapy emphasized key undesirable effects associated with these therapies, which included treatment-associated toxicities, decreased functionality, and increased dependence on caregivers. The respondents identified several expectations for selpercatinib, which included improved management of their symptoms and improved quality of life (described by 1 patient as “allowing patients to have a full and worthwhile quality of life”). Most patients reported being hopeful the use of selpercatinib would result in gains in independence and functionality (which would reduce caregiver burden), more manageable treatment-related toxicities, improvements in OS, delayed disease progression, and long-term remission for improved survivorship.

Respondents reported positive experiences following the use of selpercatinib. Most respondents expressed noticeable improvements in symptoms and some improvement in quality of life after using selpercatinib. One patient described having experienced tumour shrinkage 2 months after commencing therapy and another patient mentioned discontinuing oxygen 2 weeks after treatment initiation and recovery of the ability to walk 1 month into treatment. Another respondent reported achieving long-term remission with 4 years of stable disease after using selpercatinib and another described experiencing noticeable improvement in brain metastasis. Treatment toxicity events following the use of selpercatinib were described by some respondents as minimal and manageable, while others stated that these events had less impact on their daily life compared to other treatment options (chemotherapy and radiation therapy). The most frequently reported AEs were fatigue, dry mouth, and edema. Other events such as constipation, bloating, and occasional abdominal pain were described as minor. No respondent reported experiencing a severe-grade treatment-related AE while on selpercatinib. Respondents indicated that they could return to functionality similar to states before disease diagnosis because they had experienced great improvement in their quality of life. Respondents reported revival of hope and dreams, which allowed them to set meaningful goals for the future.

CanCertainty’s input emphasized the need for equity of access to oral cancer treatments across jurisdictions in Canada. The group recommended addressing the variability in access to treatment in Ontario and the Atlantic provinces if selpercatinib was to receive public funding. One patient described a potential issue related to access to gene testing for RET fusion mutations at diagnosis, while another brought to light issues related to the absence of screening programs for early lung cancer detection across jurisdictions in Canada. Another patient emphasized the need for new targeted therapies that could improve patient-reported outcomes for patients expressing driver mutations such as RET fusion gene mutations.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). The following input was provided by 2 clinical specialists with expertise in the diagnosis and management of NSCLC.

Unmet Need

The clinical experts outlined similar treatment goals for patients with advanced or metastatic NSCLC disease. These included improved OS, reduced symptom burden, delayed disease progression, improved quality of life (ability to function independently and decreased caregiver burden), and a decrease in or elimination of hospital admissions and hospital lengths of stay, all of which would result in a reduction in health care utilization. As cited by the experts, patients with RET fusion tend to be younger than the average lung cancer patient, less likely to be retired, and more likely to have children as dependents. The experts also indicated a need for therapies that treat or prevent metastasis because of the high rate of metastasis in this population. The experts noted that approximately 40% of patients with advanced or metastatic disease will develop brain metastasis (the proportion may go as high as 60% in patients with the RET fusion mutation). They pointed out that brain metastasis has a devastating impact on patient survival, quality of life, and ability to function. The expert also emphasized the need for a treatment that will minimize risk of exposure to COVID-19 from in-person health care visits.

Place in Therapy

Both clinical experts indicated that selpercatinib is likely to be administered in the first-line setting and preferably as a single drug. The experts emphasized that platinum plus pemetrexed and pembrolizumab (triplet therapy) was the preferred treatment option in the first line across jurisdictions (except in the province of Prince Edward Island, where pembrolizumab is not funded). Beyond the first line, the experts noted that docetaxel is funded and can be administered depending on whether the patient received triplet therapy at the first line. Patients may also receive single-drug immunotherapies (e.g., pembrolizumab, nivolumab, or atezolizumab) in the second line if they had received platinum and pemetrexed in the first-line setting. However, as noted by the experts, patients with RET fusion are known to show low response rates to immunotherapy in practice, and docetaxel may be administered in place of an immunotherapy in next-line settings.

The clinical experts indicated that selpercatinib will address the underlying disease in patients because of its mechanism of action and its favourable safety profile. They added that selpercatinib should be prioritized in the first-line setting to ensure that all patients with the RET fusion mutation have access. The experts further explained that, in patients who had received prior therapy, selpercatinib may likely be used in the next line after progression, thereby shifting all other treatment options. The experts noted that it would not be appropriate to start patients with other therapies before considering selpercatinib if selpercatinib received public funding because other therapies are less effective, more toxic, and impose a larger burden on the health care system as a whole, and because IV therapies that have to be administered in a chemotherapy suite are more likely to require in-person (or in-hospital) supportive care for adverse effects.

Patient Population

Both experts consulted agreed that treatment should be made available to patients with metastatic RET fusion–positive mutations. The experts emphasized that patients with incurable disease due to RET fusion are likely to respond to selpercatinib. They added that, because patients are most likely to response well in first-line settings, therapy should be prioritized for patients in the first-line but also recommended to all patients as next-line options if they are currently on a different treatment.

However, the experts cautioned that patients with early-stage disease who are eligible for curative-intent therapy should not be offered selpercatinib. Patients are also ineligible to receive selpercatinib if they were previously treated with another selective RET inhibitor, have severely impaired performance status (e.g., an ECOG PS of 4), or have organ dysfunction that precludes safe administration of the drug; this included patients who are unable to take medications orally or absorb medications due to gastrointestinal tract complications.

The experts listed different molecular testing techniques that are available across jurisdictions in Canada to test RET fusion mutations in patients at initial diagnosis. The NGS technique was cited as the most commonly used technique, while NGS with RNA sequencing was considered the best test because of its 100% sensitivity and 99% specificity using the MSK IMPACT testing.

Assessing Response to Treatment

The clinical experts indicated that radiographic assessment methods used in practice (CT or MRI scans) are the same as those used in trials to assess response to treatment. However, the frequency of assessments differs in both settings (trial versus real-world). The standard frequency most likely to be implemented in practice for patients receiving selpercatinib involves conducting assessments every 8 to 12 weeks, or sooner if patients experience new symptoms or show physical findings that suggest progression. The experts added that the frequency of assessments for AEs, and the presence and severity of symptoms is every 3 to 4 weeks for patients receiving oral targeted therapy in clinical practice. Assessments may be carried out at shorter intervals at treatment initiation and subsequently spaced out after therapy has been established and patients are doing well in terms of tolerance and symptom control.

Discontinuing Treatment

Both clinical experts identified similar deciding factors for determining whether to discontinue treatment. These include disease progression (symptomatic disease progression — with the exception of oligoprogression amenable to a local intervention such as radiation or surgery to achieve disease control — or progression in the CNS that is only amenable to brain-targeted therapy such as radiation), the presence of unacceptable or unsafe adverse effects that cannot be managed using appropriate dose reductions and/or supportive care medications, and patient preference.

Prescribing Conditions

The clinical experts indicated that a medical oncologist would be best suited to oversee the administration of selpercatinib.

Clinician Group Input

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

Clinician input was submitted by 2 clinician groups: LCC and the Ontario Health (Cancer Care Ontario) DAC.

Ontario Health’s drug advisory committees provide timely evidence-based clinical and health-system guidance on drug-related issues in support of Cancer Care Ontario’s mandate, including the Provincial Drug Reimbursement Programs and the Systemic Treatment Program.

Twelve clinicians associated with LCC and 2 from the DAC provided input.

Unmet Needs

Both clinician groups identified similar treatment goals for patients with advanced or metastatic disease. Improvement in median OS, rapidity of and prolonged improvement in lung cancer–related symptoms measured by median time to response, ORR, progressive disease rate and median PFS, toxicity, prevention or treatment of brain metastases, resource utilization, and impact of COVID-19 on the safety of systemic therapy were emphasized by clinicians. The LCC clinician group input noted that current treatments for patients with RET fusion–positive NSCLC have not improved OS in patients and are not associated with rapid and prolonged improvement in lung cancer symptoms as measured by median time to response, ORR, or progressive disease rate and median PFS. Input from the DAC stated that current treatments fail for all patients within months to a few years.

Clinicians at LCC reported a need for a treatment that improves HRQoL in patients, given that patients with advanced or metastatic disease usually experience greater symptom burden, and that achievement of rapid and prolonged improvement in related symptoms (measured by median time to response, ORR, or progressive disease rate and median PFS) will translate to improved quality of life. The clinicians also pointed out that treatment-related toxicities associated with new treatments are important since they influence adherence to therapy and are directly related to patient quality of life. The prevention or treatment of brain metastasis was also relevant as the group presented evidence that demonstrates almost 40% of patients with advanced or metastatic disease present with brain metastasis during treatment. As noted, brain metastasis influences quality of life and carries a poor prognosis.

Assessing Response to Treatment

The LCC group described documentation of lung cancer–related symptom stabilization or improvement by frequency and severity with or without radiological evidence of tumour shrinkage or radiographic reduction of documented sites of known disease at baseline as clinically meaningful responses to treatment. The group noted that all available systemic therapy for metastatic NSCLC, including chemotherapy, anti–PD-L1 therapeutics, and their combinations have yet to demonstrate superiors outcomes and toxicity profiles in patients with RET fusion metastatic NSCLC.

Patient Population

The most appropriate candidates to be treated with selpercatinib among patients with metastatic NSCLC are those whose tumours or blood have documented RET fusion as validated molecular diagnostics, including but not limited to NGS or fluorescence in situ hybridization (FISH). Clinicians at LCC stated that eligible patients will be identified based on RET fusion testing of tumours or blood by NGS or FISH. The clinician group mentioned that the majority of the provinces already have testing in place or will be implementing NGS testing for fusion mutations at initial diagnosis. At the time their input was supplied, they had identified 7 jurisdictions in Canada that had NGS testing in place (New Brunswick, Ontario, Manitoba, Alberta and Northwest Territories, and British Columbia and Yukon).

Place in Therapy

Both groups noted that adding selpercatinib to the Canadian treatment paradigm will allow the drug to be administered as first-line therapy in newly diagnosed patients with RET fusion metastatic NSCLC. Clinicians in the LCC group mentioned that newly diagnosed patients with an ECOG PS of 0, 1, 2, or 3 could benefit from selpercatinib if the drug is approved for funding, although they expressed uncertainty about the best therapy for second-line treatment and suggested that decisions regarding subsequent therapy should consider docetaxel and anti–PD-L1 therapy for those who have not received such drugs in prior lines of therapy. They added that, in patients previously treated for RET fusion metastatic NSCLC, selpercatinib could be administered based on evidence drawn from data updates from the LIBRETTO-001 trial.

Discontinuing Treatment

Treatment may be discontinued in the following situations: multiple toxicities despite multiple dose reductions, patient preference, concurrent medical condition(s) that could jeopardize the safety of selpercatinib, and symptomatic disease progression (with the exceptions previously identified).

Drug Program Input

The drug plans 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.

The drug plans emphasized the different treatment strategies in place for patients with NSCLC. In patients with no test results for RET fusion–positive mutations, funded treatments may likely include pembrolizumab as a single drug (if PD-L1 levels are 50% or higher), platinum plus pembrolizumab and pemetrexed, or a platinum-based therapy based on a histology in patients who are treatment naive. In patients previously treated with systemic therapy, potential treatment options will include immune checkpoint inhibitors if patient with no prior experience with an PD-L1 inhibitor (pembrolizumab, nivolumab, or atezolizumab depending on PD-L1 status), or chemotherapy if the patient has experience with a PD-L1 inhibitor (docetaxel or pemetrexed). They also noted that patients being treated with selpercatinib may require different dosing schedules and dose intensity, as outlined in the Canadian product monograph.

According to the monograph, the recommended dose, to be administered approximately every 12 hours, is based on body weight. Patients weighing less than 50 kg should receive 120 mg twice a day and those weighing 50 kg or more should receive 160 mg twice a day. Patients with severe hepatic impairment (as indicated by a Child-Pugh score in the C range) should receive a reduced dose of 80 mg twice a day.

The drug plans also noted that if selpercatinib is approved for public funding, it may modify the treatment paradigm in practice and change treatments reimbursed for subsequent lines. The drug plans pointed out that access to companion testing is needed to identify eligible patients.

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

ECOG PS = Eastern Cooperative Oncology Group Performance Status; PD-1 = programmed cell death protein 1; PD-L1 = programmed cell death ligand 1; pERC = CADTH pan-Canadian Oncology Drug Review Expert Review Committee; RET = rearranged during transfection.

Clinical Evidence

The clinical evidence included in the review of selpercatinib 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 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 selpercatinib capsules, 40 mg and 80 mg, for oral administration:

• as monotherapy for the first-line treatment of adult patients with metastatic RET fusion–positive NSCLC

• as monotherapy for the treatment of adult patients with metastatic RET fusion–positive NSCLC who have received prior systemic therapy.

Methods

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

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.²¹

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 multi-file search. Duplicates were removed using Ovid deduplication for multi-file searches, followed by manual deduplication in EndNote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were selpercatinib (Retevmo/Retsevmo) and synonyms. Clinical trials registries searched included the US National Institutes of Health’s clinicaltrials.gov, WHO International Clinical Trials Registry Platform (ICTRP) search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register. No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. Appendix 1 provides detailed search strategies. The initial search was completed on November 4, 2021. Regular alerts updated the search until the meeting of the CADTH pan-Canadian Oncology Drug Review Expert Committee on March 9, 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 resource.²² Included in this search were the websites of regulatory agencies (FDA and EMA). 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 be included in the review, and differences were resolved through discussion.

A focused literature search for indirect comparisons dealing with selpercatinib for NSCLC was run in MEDLINE All (1946–) on November 3, 2021. No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results.

Findings From the Literature

One study was identified from the literature for inclusion in the systematic review (Figure 1). The study is summarized in Table 6.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Description of Studies

The LIBRETTO-001 trial is an ongoing phase I and II, multi-centre, multi-cohort, open-label study consisting of a dose-escalation phase and a dose-expansion phase.

The study was initiated in May of 2017 and has more than 84 participating centres, including centres in Canada and the US.

The median age of all patients enrolled in the LIBRETTO-001 trial as of December 16, 2019, who had received at least 1 dose of selpercatinib (safety analysis set, n = 702) was 59 years. More men (52.4%) than women (47.6%) participated in the study. Most patients recruited were White (69%), followed by patients of Asian descent (21.9%), patients who are Black (3.4%), and others (5.6%).¹⁹

Trial Design

Phase I: Dose Escalation and MTD Determination

The primary objective of phase I of the study was to use dose escalation to determine the MTD and RP2D of selpercatinib.

Secondary objectives for phase I included evaluation of the safety of selpercatinib, characterization of the pharmacokinetic (PK) properties, and assessment of the antitumour activity of selpercatinib (i.e., determining the ORR).

A “3 + 3” dose-escalation design with 3 or 6 patients enrolled in each dose cohort was used to assess the MTD and RP2D.

Phase II: Dose Expansion

The phase II portion of LIBRETTO-001 is ongoing and includes patients from phase I, enrolled into 1 of 5 pre-specified cohorts based on tumour type (Figure 2). Patients who were recruited into cohorts 1 through 4 required confirmed evidence of an RET gene alteration in a tumour before enrolment.

• Cohort 1: RET fusion–positive solid tumour progressed on or intolerant to 1 or more prior standard first-line therapy

• Cohort 2: RET fusion–positive solid tumour without prior standard first-line therapy

• Cohort 3: RET-mutant MTC progressed on or intolerant to 1 or more prior standard first-line cabozantinib and/or vandetanib

• Cohort 4: RET-mutant MTC without prior standard first-line cabozantinib or vandetanib or other kinase inhibitors(s) with anti-RET activity

A fifth cohort, cohort 5, included:

• cohorts 1 through 4 without measurable disease

• MTC not meeting the requirements for cohorts 3 or 4

• MTC syndrome spectrum cancers (e.g., MTC or pheochromocytoma), or poorly differentiated thyroid cancers with other RET alteration/activation could be allowed with

• prior sponsor approval

• cfDNA positive for a RET gene alteration not known to be present in a tumour sample.

The primary objective of the phase II study was to evaluate the antitumour activity of selpercatinib in patients recruited into each of the 5 cohorts. Secondary objectives included evaluating outcomes related to DOR, CNS ORR, CNS DOR, time to any and best response, clinical benefit rate, PFS, OS, safety, and PK properties.¹⁹

Figure 2: Study Design of the LIBRETTO-001 Trial

MTC = medullary thyroid cancer; MTD = maximum tolerable dose; RET = rearranged during transfection.

Source: Clinical Study Report (data cut-off of December 16, 2019).¹⁹

This CADTH review focuses on outcomes observed in patients with NSCLC and a confirmed RET fusion gene mutation enrolled into cohort 1 and 2 at phase II of the LIBRETTO-001 study. These patients were further subgrouped into 3 datasets to form the primary efficacy population for patients with NSCLC based on clinically meaningful distinctions observed during the trial. The primary efficacy population was initially negotiated by the sponsor and regulatory agencies (FDA and EMA) before the first data cut-off date of June 17, 2019,⁵ and supported the regulatory submission for new drugs at Health Canada.

Populations

Inclusion and Exclusion Criteria

Patients who were 18 years of age and older (12 years and older at some sites) with a locally advanced or metastatic solid tumour with progression or who were intolerant to standard therapy were included in phase I of the trial (Table 6). Patients were also included in cases where, in the investigator’s opinion, no standard therapy existed, the patients were not candidates for therapy or would be unlikely to tolerate or derive significant clinical benefit from standard therapy, or who declined standard therapy. The initial protocol for the phase I study did not focus on identifying specific genetic targets. However, patients with an RET alteration in a tumour (not just specific to detection in blood) confirmed by PCR or NGS testing were identified during phase I based on pre-clinical data of the inhibitory action of selpercatinib, and RET gene alteration was required for continued dose escalation once pre-specified PK criteria were met. Evidence of an RET gene alteration in a tumour was required for entry into phase II of the trial. Patients with an ECOG PS score of 0, 1, or 2 (age ≥ 16 years) or a Lansky Performance Status score of 40% or higher (age < 16 years) with no sudden deterioration 2 weeks before the first dose of study treatment were also enrolled.¹⁹

At the initial stages of the phase I study, patient enrolment did not require a confirmed diagnosis of an RET gene fusion or mutation in the solid tumour. New changes were made to the study protocol that incorporated the need for a confirmed RET fusion at screening for phase I and a confirmed RET fusion status before the patient was subsequently enrolled into phase II. These changes were proposed and implemented after the evaluation of clinical data for the first patients enrolled at phase I, in which selpercatinib exhibited favourable inhibitory properties in patients harbouring the RET mutation. Mandatory screening and confirmation of the RET mutation was implemented before enrolment in phase II.

Patients were excluded from phase II cohorts 1 to 4 if they had another validated oncogenic driver that could result in treatment resistance to selpercatinib. For both phases, those who had previously received RET inhibitor therapy, including investigational products, were ineligible. Patients were also ineligible if they had a clinically significant active cardiovascular disease or history of myocardial infarction within 6 months before the planned start of selpercatinib or prolongation of the QT interval corrected for heart rate.¹⁹

Baseline Characteristics in the RET Fusion–Positive NSCLC Population

At the data cut-off date of December 16, 2019, a total of 253 patients with NSCLC who met the eligibility criteria at phase I and II and had received at least 1 dose of selpercatinib at interim 1 (cut-off date of June 17, 2019), with at least 6 months of follow-up data, were included in the efficacy analysis. The NSCLC population consists of 3 subgroups (PAS, IAS, and SAS), which reflect the specified populations outlined in the reimbursement request. Table 7 presents the baseline characteristics observed in the NSCLC population at the December 16, 2019, data cut-off.⁵

The median age observed in the overall NSCLC population was 61 years (range = 23 to 83). More females were enrolled than males, and the majority of patients (51%) were White or of Asian descent (38%). The median body weight was 64.0 kg (range = 38.9 to 148). Most patients (61%) had a baseline ECOG PS score of 0 or 1, and 70% of patients had never smoked. Most patients had stage IV disease (67.2%) and 98% of patients had a history of metastasis. Overall, patients in the prior systemic therapy group had received platinum chemotherapy (100%) including other therapies (anti–PD-1 and anti–PD-L1 = 54.3%; multikinase therapy = 36.4%) in the first and second lines, and the median number of therapies observed was 3.⁵

Interventions

Treatment

Patients received selpercatinib as an oral capsule or suspension (introduced within study protocol version 5) once daily or twice daily, depending on the cohort assignment. The dosing regimen in the trial followed a fixed-milligram format and had a consistent timing in administration each day (approximately 12-hour intervals for the twice daily dosing schedule). A minimum period of 6 hours was required between administration of consecutive doses. Patients were required to skip doses that would have been administered more than 6 hours late and were expected to log them as missed in the patient diaries. Earlier versions of the study protocol (until version 7.0) discouraged patients from consuming any food at least 2 hours before and 1 hour after the administration of selpercatinib. However, the protocol was later amended to allow for the administration of selpercatinib with or without food based on new evidence from a food-effect study in healthy volunteers that showed minimal effects of food on the bioavailability of selpercatinib.¹⁹

Treatment Assignation

At phase I, patients received oral selpercatinib through a pre-specified dose-escalation plan, starting with 20 mg daily and progressing to 240 mg twice daily or until the safety review committee and sponsor determined that a suitable dose was achieved based on safety, PK exposure, and clinical activity. A dosage of 160 mg twice daily was selected at the end of phase I by the safety review committee as the RP2D.

All patients at phase II received selpercatinib 160 mg twice daily and continued treatment until disease progression, unacceptable toxicity, or other reasons for treatment discontinuation.¹⁹

Dose modifications, including reductions or schedule modifications, were approved at the investigator’s discretion for AEs in individual patients. Dose reductions were to be considered a dose-limiting toxicity if the dose was reduced during cycle 1 for a toxicity that could not be reasonably attributed to the patient’s underlying disease, other medical condition, or concomitant medications. Otherwise, no dose adjustments were to be made during the conduct of the trial. Re-escalation to a higher dose after a previous dose reduction was allowed; however, dose escalations above 160 mg twice daily were not authorized.¹⁹

Concomitant Medications

Concomitant medications were allowed only if patients had declared them at the screening phase and if they were considered acceptable before study enrolment. The most frequently reported therapeutic classes of concomitant medications were thyroid hormones (57.1%), natural opium alkaloids (37.3%), and anilides (35.2%). Standard supportive medications (e.g., hematopoietic growth factors [not for prophylaxis in cycle 1] and anti-emetics) were permitted.¹⁹

The sponsor also approved local treatment (palliative radiation therapy or surgery for bone metastases) while patients received selpercatinib but recommended holding selpercatinib for approximately 5 half-lives (approximately 2 to 3 days) before and after radiation therapy or surgery. Prohibited concomitant medications included systemic anticancer drugs, hematopoietic growth factors for prophylaxis in cycle 1, therapeutic monoclonal antibodies, drugs with immunosuppressant properties, or medications known to be strong inhibitors or inducers of CYP3A4 were prohibited as concomitant therapies. The use of proton pump inhibitors was also prohibited.¹⁹

Treatment Adherence

Treatment adherence was evaluated using patient diaries, visits to clinics, and medication counts from returned medication bottles. Administration of concomitant medications was also documented in diaries.¹⁹

Censoring

Patients could withdraw their consent to participate in the study whenever they regarded it best. Investigators received the sponsor’s approval to censor patients who showed disease progression from the RECIST 1.1 or RANO evaluations of tumours (except in cases of a demonstrated clinical benefit following drug use), demonstrated unaccepted toxicity, experienced intercurrent illness compromising their ability to fulfill protocol requirements, became pregnant, needed an alternative treatment (e.g., local radiation or surgery for disease that does not meet the definition of disease progression) unless treatment was temporary, showed significant protocol noncompliance, or withdrew consent, and in the event of patient loss to follow-up, death, study termination by the sponsor.¹⁹

Outcomes Investigated

A list of the efficacy end points assessed in the LIBRETTO-001 trial and included in this review is provided in Table 8.

Efficacy Measurement for Primary and Secondary Outcomes

Tumour Measurements and Assessments of Disease Response Using RECIST 1.1

Given the open-label and non-randomized nature of the LIBRETTO-001 trial, tumour evaluation for the primary end point using the RECIST 1.1 guidelines was planned as a third-party blinded central review to improve the internal validity of the study. However, at the interim analysis, tumour size was evaluated by both the investigator and an IRC.

Tumours were assessed at screening, within 28 days of cycle 1 at day 1, by CT or MRI of the neck, chest, or abdomen and pelvis, as indicated. Head and neck CT scans were not required for malignancies other than those involving the head and neck. Post-baseline assessments were conducted every 8 weeks (± 1 week) from cycle 3 through cycle 13, and 12 weeks onward until an identifiable disease progression, withdrawal of patient consent, initiation of a new cancer therapy, or if any discontinuation criteria were met. The protocol also allowed investigators to conduct an initial tumour evaluation 4 weeks after treatment initiation and a confirmatory evaluation 4 weeks after the tumour evaluation showed a CR or PR by RECIST 1.1 (or RANO) if consistent with the local regulatory requirements. Patients with an identified RET fusion–positive gene mutation in a tumour or a history of CNS metastases, or those with a clinical indication were required to undergo brain imaging at baseline at phase II. Post-baseline CNS imaging was also completed for patients presenting with baseline metastasis. Assessments were performed at the end-of-treatment visit if they had not been performed within the previous 2 cycles. Subsequent assessments followed the same radiographic methods as used during screening. During screening, tumour lesions were categorized as measurable versus non-measurable and target versus non-target.

Measurable Versus Non-Measurable

• Measurable: lesions that could accurately be measured in at least 1 dimension, the longest diameter in the plane of measurement to be recorded as:

  • tumour lesions: 10 mm or larger by CT scan

  • malignant lymph nodes: to be considered pathologically enlarged and measurable, the node must be 15 mm or greater on the short axis when assessed by CT scan. At baseline and in follow-up, only the short axis was measured and followed. Nodes that have a short axis shorter than 10 mm are considered nonpathological and should not be recorded or followed.

• Non-measurable: All other lesions, including small lesions (longest diameter < 10 mm or pathological lymph nodes with ≥ 10 to < 15 mm short axis), and truly non-measurable lesions.

Target Versus Non-Target

• Target: all measurable lesions up to a maximum of 2 lesions per organ and 5 lesions in total, representative of all involved organs, are to be identified as target lesions and measured and recorded at screening. Target lesions are to be selected based on their size (i.e., those with the longest diameter) and suitability for accurate repeated measurement. Lymph nodes may be selected as target lesions; they must be defined as measurable, and only the short axis of the node will contribute to the baseline sum. All other pathologic nodes with a short axis 10 mm or longer but shorter than 15 mm should be considered non-target lesions.

• Non-target: all other lesions not classified as target lesions (or sites of disease) are to be identified as non-target lesions and are to be recorded in the electronic case report form. Measurement of non-target lesions is not required.

The sum of the diameters (the longest for non-nodal lesions and the short axis for nodal lesions) for all target lesions were calculated. Disease response in target and non-target lesions were assessed by the investigator using RECIST 1.1, according to the categories and criteria.

Response Assessment in Neuro-Oncology Criteria for Primary CNS Malignancies

The RANO criteria was used to assess lower-grade primary CNS malignancies.

A measurable lesion was evaluated by contrast-enhancing MRI using the following definition:

• it has clearly defined margins

• it is visible on 2 or more axial slices, preferably less than 5 mm thick

• it is at least 10 mm in size if slice thickness is less than 5 mm (or twice the slice thickness if greater than 5 mm thick)

• it does not measure a cystic cavity

• non-measurable lesions are those that do not fit the criteria above, and specifically lesions that are cystic, necrotic, or include a surgical cavity should not be considered measurable

• measurements were calculated by summing the products of perpendicular diameters of all measurable enhancing lesions.

Follow-up

At phase II, a short-term follow-up assessment was designed to occur 4 weeks (at least 28 days [plus 7 days room allowed for the follow-up assessment]) after the last dose of the study drug. Patients were allowed to continue selpercatinib at the discretion of the investigator in cases of a documented progression of disease. Long-term follow-up assessments were scheduled every 3 months and could be conducted by phone. Long-term could include subsequent anticancer therapy and survival status. Assessments included periodic radiologic evaluation and ongoing safety assessments.¹⁹

Statistical Analysis

The primary efficacy dataset to support regulatory submission was initially negotiated by the sponsor and regulatory agencies (FDA and EMA) before the first data cut-off date of June 17, 2019. At the December 16, 2019, and the March 30, 2020, data cut-offs, the datasets supporting the evidence for the funding request differed from the population cohorts defined at phase I and II.

As a rationale for modifying the population sets, the sponsors noted that the analysis sets defined in the LIBRETTO-001 protocol and statistical analysis plan for the Clinical Study Report were designed for a study conducted in all tumour types. Subsequently, analysis sets were defined for each of the 3 predominant histologies of patient who had enrolled in the LIBRETTO-001 trial to understand efficacy by tumour type, rather than taking a tumour-agnostic approach, and to support the initial submissions to the FDA and EMA.

Various datasets were created to facilitate a regulatory review of the LIBRETTO-001 data in support of the proposed indications for marketing authorization. Information was maximized through consolidation of data from both the phase I and phase II parts of the LIBRETTO-001 trial, and groupings were based on clinically meaningful distinctions, resulting in similarity of patients within a group and facilitating interpretation of results.

The inclusion criteria and statistical considerations for the primary datasets supporting the proposed indications were agreed to by the EMA (July 09, 2019) and FDA (MTC: December 19, 2018; NSCLC: January 16, 2019).^5,9

Analysis Population

The second interim analysis on December 16, 2019, supported the FDA and Health Canada submissions. Patients included in this analysis had received at least 1 dose of selpercatinib at or before the first interim cut-off date of June 17, 2019. Data analyzed at a third interim data cut-off date of March 30, 2020, were also part of the submission request and this interim analysis was pre-planned to support the regulatory submission in Japan. Results obtained at the second and third cut-offs are presented in this report. The following datasets were defined for the efficacy and safety analysis of selpercatinib in the NSCLC population.^5,19

Efficacy Analysis Sets

At the December 16, 2019, and March 30, 2020, data cut-offs, the following subgroups were included in the analysis:

• The PAS (n = 105) is composed of the first consecutively enrolled patients previously treated with platinum-based chemotherapy (a subset of the IAS). This subset analysis includes the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–PD-1 and anti–PD-L1 antibody.

• The IAS (n = 184 and n = 218 for December 16, 2019, and March 30, 2020, data cut-off, respectively) is composed of all patients treated with platinum-based chemotherapy.

• The SASs include 3 subsets:

  • SAS1 (n = 39 and n = 48 for December 16, 2019, and March 30, 2020, data cut-off, respectively) is composed of patients who are treatment naive.

  • SAS2 (n = 16 and n = 18 for December 16, 2019, and March 30, 2020, data cut-off, respectively) is composed of patients treated with other systemic therapies that are not platinum-based.

  • SAS3 (n = 14 at the December 16, 2019, data cut-off, and n = 18 for at the March 30, 2020, data cut-off) consists of patients without measurable disease according to RECIST 1.1.

Central Nervous System Response Analysis Set

The CNS response analysis set included all treated RET fusion–positive NSCLC and other patients with a solid tumour who met the first criteria of PAS and had an investigator-assessed CNS metastases at baseline (reported as target or non-target lesion according to RECIST 1.1). The IRC reviewed all CNS scans and provided a CNS-only best overall response and DOR by RECIST 1.1 for each patient. The analysis was conducted by tumour type (NSCLC and other) as well as overall.

Health-Related Quality of Life Dataset

At the December 16, 2019, data cut-off, of the 253 patients enrolled in the NSCLC population, 239 were included in the HRQoL analysis. These patients were further subgrouped into 3 populations:⁷

• treatment-naive (n = 39)

• 1 prior line of therapy (n = 64)

• 2 or more prior lines of therapy (n = 136).

Overall Safety Analysis Set

The RET fusion–positive NSCLC safety analysis set includes all patients with documented RET fusion who were enrolled in the LIBRETTO-001 study and received 1 or more doses of selpercatinib as of the cut-off date. The safety analysis set at the December 16, 2019, data cut-off was composed of 702 patients treated with selpercatinib irrespective of the cohort assignation. At the March 30, 2020, data cut-off, 746 patients who had received at least 1 dose of selpercatinib were included in the OSAS, while at the December 16, 2019, data cut-off, 702 patients who had received at least 1 dose of selpercatinib were included in the OSAS.

Figure 3: Selpercatinib Enrolment and Analysis Populations

cabo = cabozantinib; CNS = central nervous system; IAS = integrated analysis set; MTC = medullary thyroid cancer; NSCLC = non–small cell lung cancer; PAS = primary analysis set; RET  = rearranged during transfection; SAS = supplemental analysis set; vande = vandetanib.

Note: Three analysis subsets are not shown: CNS efficacy in the NSCLC PAS, CNS efficacy in all patients with RET fusion–positive NSCLC, and a subset of PAS patients with prior programmed cell death protein 1 or programmed cell death ligand 1 therapy. Enrolment date: June 17, 2019.

¹ RET fusion–positive other tumours include pancreatic cancer, rectal neuroendocrine cancer, salivary gland cancer, carcinoid cancer, cancers of the colon and small intestine, and xanthogranuloma.

² Other solid tumours that do not fit the other disease cohorts.

³ Prior systemic therapy other than platinum-based chemotherapy.

⁴ Patients without measurable disease who were enrolled in phase I dose-expansion cohort 5 (per-protocol version 4.0 or earlier) or phase II cohort 5 (per-protocol version 5.0 and later).

⁵ Previously treated RET fusion–positive thyroid cancer.

Source: Sponsor’s submission package, Retevmo (selpercatinib),⁵ LIBRETTO-001.²⁶

Sample Size Definition

Primary Analysis Set

For the PAS, an ORR of 50% or more was hypothesized by the sponsor when selpercatinib is administered to patients with RET fusion–positive NSCLC who progressed on or after receipt of platinum-based chemotherapy. A sample size of 105 patients was estimated to provide more than 98% power to achieve a lower boundary of a 2-sided 95% exact binomial CI about an estimated ORR that exceeds 30%. Ruling out a lower limit of 30% for the ORR was considered clinically meaningful by the sponsor for the patient population. Under the primary analysis, the effectiveness of selpercatinib was established if the lower limit of the 95% CI will exceed 30% when the estimated ORR is 40% or greater (using the Clopper-Pearson method).⁵

The IAS included all patients with RET fusion–positive NSCLC enrolled in the LIBRETTO-001 study by the data cut-off date who met the PAS eligibility criteria.

Planned Analysis

There were 3 interim analyses planned to support regulatory submissions in different jurisdictions.

• The first interim analysis at the data cut-off of June 17, 2019, supported the original New Drug Application at the FDA but was later updated with results from the Day 60 Efficacy and Safety Update with a data cut-off of December 16, 2019.

• The second interim analysis at the data cut-off of December 16, 2019, provides an additional 6 months of follow-up information for patients who were enrolled and had received at least 1 dose of selpercatinib before or at the data cut-off date of June 17, 2019. This formed the basis of the Health Canada regulatory submission and FDA submission and therefore efficacy results related to the first analysis are not reported.

• The third interim analysis at the March 30, 2020, data cut-off date, which was planned to support the regulatory submission in Japan, provides additional information for relevant outcomes (ORR, DOR, OS, PFS, and safety) investigated in the LIBRETTO-001 trial. Efficacy results are presented for 2 data cut-off dates (March 30, 2020, and December 16, 2019) in this report for relevant outcomes in specific patient populations related to the reimbursement populations (i.e., patients with RET fusion–positive NSCLC).

Primary Outcome Analysis

The ORR was determined by the IRC and calculated based on the maximum likelihood estimator (i.e., crude proportion of patients with best objective response of CR or PR), and a 2-sided 95% exact binomial CI using the Clopper-Pearson method was provided. The ORR represents the proportion of patients with a best objective response of confirmed CR or confirmed PR based on RECIST 1.1.

The point estimate of the ORR was calculated based on the maximum likelihood estimator (i.e., crude proportion of patients in the PAS with best overall response of confirmed CR or confirmed PR). The point estimate was accompanied by a 2-sided 95% exact binomial CI using the Clopper-Pearson method. The effectiveness of selpercatinib was demonstrated if the lower limit of the 2-sided 95% CI exceeded 30%.⁵

Sensitivity Analyses

The IAS was used to perform sensitivity analysis on the data. No further details of the sensitivity analysis were provided.

Subgroup Analyses

Supportive analyses were performed on selected subgroups and special populations. These analyses were conducted on the PAS population and selected analysis sets as appropriate. The point estimates of the ORR (and 95% CI) and the median DOR (and range) based on IRC was calculated for the subgroups and special populations defined by the following:

• age at enrolment (< 65 years, ≥ 65 years)

• sex (male, female)

• race (White, Asian, other)

• ECOG PS at baseline (0, 1, or 2)

• smoking status (never smoked or smoker)

• type of molecular assay (NGS on tumour or PCR, NGS on plasma, FISH)

• RET fusion gene (KIF5B, non-KIF5B, unknown)

• history of metastatic disease (yes, no)

• CNS metastasis at baseline by investigator (yes, no)

• number of prior systemic therapies (0, 1 or 2, ≥ 3)

• Prior anti–PD-1 or anti–PD-L1 (yes, no)

• Prior multikinase inhibitor (yes, no).

The CADTH review protocol identified 4 subgroups of interest for which data will be presented. These include performance status at baseline, prior class or type of therapy, number of prior therapies, and CNS metastases at baseline.

Analysis of Secondary Outcomes

Duration of Response

The DOR was defined as the number of months from start date of CR or PR (whichever came first) and subsequently confirmed, to the first date that recurrent or progressive disease was objectively documented. If a patient died, irrespective of cause, without prior documentation of recurrent or progressive disease, then the date of death was used to denote the response end date. Calculations were performed only in patients who achieved a CR and PR. Kaplan–Meier methods were used to summarize the DOR, and median follow-up was estimated based on the Kaplan–Meier estimate of potential follow-up.⁵

The DOR was summarized descriptively using the Kaplan–Meier method. The Kaplan–Meier estimate with a 95% CI was calculated using the Brookmeyer and Crowley method to provide for the median. The event-free rate with a 95% CI was calculated using Greenwood’s formula for selected time points. The median follow-up for DOR was estimated according to the Kaplan–Meier estimate of potential follow-up.

• DOR (months) = (event or censoring date − response start date + 1)/30.4375

Progression-Free Survival

Progression-free survival was defined as the number of months that elapsed between the date of the first dose administration of selpercatinib and the earliest date of documented disease progression or death (whatever the cause).

Values for PFS were summarized descriptively using the Kaplan–Meier method and the median follow-up was estimated according to the Kaplan–Meier estimate of potential follow-up.

• PFS (months) = (event or censoring date − first dose date + 1)/30.4375

Overall Survival

Overall survival was defined as the number of months that elapsed between the date of the first dose administration of selpercatinib and the date of death (whatever the cause). Patients who were alive or lost to follow-up as of the data cut-off date were right-censored. The censoring date was determined from the date the patient was last known to be alive.

Values for OS were summarized descriptively using the Kaplan–Meier method and the median follow-up was estimated according to the Kaplan–Meier estimate of potential follow-up.⁵

• OS (months) = (death or censoring date − first dose date + 1)/30.4375

Central Nervous System Objective Response Rate

The CNS ORR was analyzed using the same method described for the analyses of ORR and DOR and summarized overall and by baseline disease measurability as assessed by IRC.⁵

Health-Related Quality of Life

Version 3 of the EORTC QLQ-C30 instrument was used to measure HRQoL in patients participating in the LIBRETTO-001 trial. This self-administered questionnaire is composed of 30 items, measuring patient health in 2 different domains, the functional subscales domain and the symptom subscale domain. It includes 9 items: nausea and vomiting (2 items), fatigue (3 items), pain (2 items), diarrhea (1 item), dyspnea (1 item), appetite loss (1 item), insomnia (1 item), constipation (1 item), and financial difficulties (1 item). The minimally important difference (MID) was defined as a “10-point difference from the baseline assessment value.”¹⁹ A definite change (improvement or worsening) was defined as an improvement or worsening, respectively, from baseline of 10 or more points without any further reduction or increase in score of 10 or more points, respectively.

Higher scores obtained in the EORTC QLQ-C30 signified better functioning. Scores were linearly transformed to a 0-to-100 scale. Missing values were handled based on the suggested techniques in the EORTC QLQ-C30 manual. Only raw scores with complete information on at least 50% of the items in the scale or subscale were computed. Scores values were not imputed. All subscales in the EORTC QLQ-C30 were scored from 0 to 100. Higher scores obtained in the quality of life and functional subscales represented improved functioning, while lower scores obtained in the symptom subscales represented fewer symptoms. Clinically important thresholds used in the assessment were obtained from published literature.⁷ Results were presented as descriptive statistics (median and quartile, mean and SD, and mean change and standard error from baseline) for each subscale at each study visit.

• Lower scores obtained in functional subscales were compared to the following defined thresholds: 83 (physical function), 71 (emotional function), 58 (role function, social function), or 75 (cognitive function) and were considered a clinically meaningful problem for patients.

• Higher scores obtained in symptom subscales were compared to the following thresholds: 8 (nausea or vomiting), 39 (fatigue), 25 (pain), 17 (diarrhea, dyspnea, financial difficulties), or higher than 50 (appetite loss, insomnia, constipation), and were considered clinically meaningful problems.

The MID considered was a change of at least 10 points from baseline. “Improvement” was defined as a change from baseline of 10 or more points without any further deterioration in score of 10 or more points. “Worsening” was defined as a change from baseline of at least 10 points without any further improvement in score of 10 points or more.

The Kaplan–Meier method was used to derive time to improvement and time to worsening. Patients with least 6 months of follow-up data and those who had experienced events before the 6-month time frame were also included in their analyses. Patients were censored at the last cycle for which they had completed the questionnaire and if their score did not vary by 10 points from baseline at any subsequent assessment.

Handling of Missing Data

No imputation was performed on missing data; all analyses were based on observed data only.

Protocol Amendments

Several updates to the protocol made during the study were identified and documented before data cut-off for the regulatory submission as presented in Table 9. Changes to the protocol did not affect the study’s efficacy and safety outcomes.¹⁹

Table 9: Protocol Versions and Dates

AE = adverse event; HRQoL = health-related quality of life; NA = not available; PK = pharmacokinetics; RET = rearranged during transfection; SAE = serious adverse event.

Source: Sponsor’s submission package, Retevmo (selpercatinib),⁵ LIBRETTO-001.¹⁹

Protocol Deviations

Important protocol deviations were reported in 40 patients (7.5%). The most frequently reported and important protocol deviations were those related to the investigational product in 17 patients, and inclusion criteria and serious adverse event (SAE) reporting, each in 8 patients.

Results

Patient disposition

Table 10 provides a summary of patient disposition in the NSCLC efficacy set as of December 16, 2019, data cut-off. A total of 253 patients with NSCLC were RET fusion–positive. Of the 105 patients included in the PAS, 40% of patients had discontinued treatment, with the most common reason for discontinuation being disease progression.⁵ Greater than 20% of patients in total and 31% of those in the PAS stayed on treatment post-progression at the investigator’s discretion.

Table 10: Patient Disposition (December 16, 2019, Data Cut-Off Date)

IAS = integrated analysis set; NSCLC = non–small cell lung cancer; PAS = primary analysis set; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: The PAS included a subset analysis of the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–programmed cell death protein 1 or anti–programmed cell death ligand 1 antibody. The IAS included all patients previously treated with platinum-based chemotherapy. SAS1 included patients who were treatment naive. SAS2 included patients previously treated with other systemic therapies that are not platinum-based. SAS3 consisted of patients without measurable disease according to the Response Evaluation Criteria in Solid Tumors Version 1.1. For patients with RET fusion–positive NSCLC, the PAS includes the first 105 patients of the IAS. The total column is the sum of the IAS, SAS1, SAS2, and SAS3 at the December 16, 2019, data cut-off.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Exposure to Study Treatments

At the December 16, 2019, data cut-off, most patients with RET fusion–positive NSCLC (93%) had received 160 mg of selpercatinib twice daily, either as the starting dose (82%), through protocol-allowed intra-patient dose escalation (10%), or through dose reductions from 240 mg twice daily (1%). Dosing was also consistent across the analysis sets.⁵ Table 11 presents information of starting doses, time on treatment, and time on study.

Table 11: Selpercatinib Starting Doses for Patients Who Received at Least 1 Dose of 160 mg of Selpercatinib Twice Daily, Time on Treatment, and Time on Study

b.i.d. = twice daily; IAS = integrated analysis set; NSCLC = non–small cell lung cancer; PAS = primary analysis set; q.d. = once daily; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: The PAS included a subset analysis of the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–programmed cell death protein 1 or anti–programmed cell death ligand 1 antibody. The IAS included all patients previously treated with platinum-based chemotherapy. SAS1 included patients who were treatment naive. SAS2 included patients previously treated with other systemic therapies that are not platinum-based. SAS3 consisted of patients without measurable disease according to Response Evaluation Criteria in Solid Tumors Version 1.1. For patients with RET fusion–positive NSCLC, the PAS includes the first 105 patients of the IAS. The total column is the sum of the IAS, SAS1, SAS2, and SAS3.

^aThe recommended phase II dose at the December 16, 2019, data cut-off was 160 mg b.i.d.

^bTime on study (months) = (study exit date − first dose date + 1)/30.4375 for patients who exited the study on or before the data cut-off date; time on study (months) = (data cut-off date − first dose date + 1)/30.4375 for patients who were still in the treatment phase as of the data cut-off date; time on study (months) = (last visit date − first dose date + 1)/30.4375 for patients who were in the long term follow-up as of the data cut-off date of December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Efficacy

Overall Survival

Primary and Integrated Analysis Sets (Patients with RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

Table 12 summarizes information on OS in the PAS and IAS populations for 2 cut-off dates. At the March 30, 2020, data cut-off, 73.3% of patients were still alive in the PAS population. The median OS was NE, and the median follow-up was estimated at 19.9 months. In the IAS population, 81.2% patients were still alive. The median OS was NE and the median follow-up obtained was 14.3 months.⁵ Figure 4 presents the Kaplan–Meier plot of OS in the PAS set (December 16, 2019, only).⁵

The sponsor conducted a follow-up analysis in the IAS population at a new cut-off date of June 15, 2021. The median OS was NE at this data cut-off. The median follow-up in months was 26.4.

Table 12: Overall Survival — PAS and IAS Sets (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

CI = confidence interval; IAS = integrated analysis set; NE = not estimable; NR = not reported; NSCLC = non–small cell lung cancer; OS = overall survival; PAS = primary analysis set; RET = rearranged during transfection.

Note: The PAS included a subset analysis of the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–programmed cell death protein 1 and anti–programmed cell death ligand 1 antibody. The IAS included all patients previously treated with platinum-based chemotherapy, at the December 16, 2019; March 30, 2020; and June 15, 2021, data cut-offs.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Figure 4: Kaplan–Meier Plot of Overall Survival — Primary Analysis Set (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

+ = censored; OS = overall survival; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: Data cut-off: December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Treatment-Naive RET Fusion–Positive NSCLC Cohort (SAS1)

Table 13 presents information on OS at both data cut-offs. Given the small number of events, the median OS was NE at either analysis time point.⁵

The sponsor conducted a follow-up analysis in the treatment-naive population at a new data cut-off date of June 15, 2021. The median OS was NE at this data cut-off. The median follow-up was 21.9 months.

Table 13: Overall Survival — Patients With Treatment-Naive RET Fusion–Positive NSCLC

CI = confidence interval; NE = not estimable; NR = not reported; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: Supplemental analysis set 1 included patients who were treatment naive with December 16, 2019; March 30, 2020; and June 15, 2021, data cut-offs.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Prior Other Systemic Therapy RET Fusion–Positive NSCLC Patients (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC Patients (Supplemental Analysis Set 3)

The OS results obtained in the SAS2 and SAS3 populations at both data cut-offs are presented in Table 14. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

Table 14: OS — Prior Other Systemic Therapy for Patients With RET Fusion–Positive NSCLC Patients (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC Patients (Supplemental Analysis Set 3)

NE = not estimable; NR = not reported; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: December 16, 2019, March 30, 2020, data cut-off.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Subgroup Analysis

Data on subgroup analysis were not reported.

Health-Related Quality of Life

Patients recruited in the LIBRETTO-001 trial were asked to complete the EORTC QLQ-C30 questionnaire at baseline (cycle 1, day 1), every 8 weeks (until cycle 13), and every 12 weeks after cycle 13 (until end of treatment). Patients were evaluated at odd-numbered cycles. A descriptive analysis of results was presented, and assessments of change were made from baseline. Lower scores obtained in functional subscales were compared to the following defined thresholds: 83 (physical function), 71 (emotional function), 58 (role function, social function), or 75 (cognitive function) and were considered a clinically meaningful problems for patients. Higher scores obtained in symptom subscales were compared to the following thresholds: 8 (nausea or vomiting), 39 (fatigue), 25 (pain), 17 (diarrhea, dyspnea, and financial difficulties), or higher than 50 (appetite loss, insomnia, and constipation) and were considered clinically meaningful problems. Table 15 provides a summary of the HRQoL findings.

Table 15: Summary of Findings for Health-Related Quality of Life (December 16, 2019, Data Cut-Off)

SD = standard deviation.

Source: Minchom et al. (2021).⁷

Global Health Status

The mean GHS/QoL in the overall population was 61.5 (SD = 23.6). The mean GHS at baseline was 60.2 (SD = 23.4) in the treatment-naive population and 65.2 (SD = 23.5) in the patient population with 1 line of prior therapy. The mean GHS at baseline for patients with 2 or more lines of therapy was 60.4 (SD = 22.6). The proportion of patients who showed improved GHS/QoL from baseline was high across cycles 3 to cycle 13 in all 3 subgroups. However, the sample size of patients who completed the questionnaires decreased at every cycle. Because fewer patients (n = 28) had completed the questionnaire at cycle 13 in the overall population compared to the other cycles, the results are uncertain and cannot be fully interpreted. Findings from the time-to-event analysis showed improvements in GHS/QoL in all 3 subgroups: 66.7% (18 of 27) in the treatment-naive population, 64.1% (25 of 39) in patients with 1 prior line of therapy, and 61.1% (33 of 54) in patients who had received 2 or more prior lines of therapy at study enrolment).⁷ Figure 5 presents the GHS/QoL data (from the EORTC QLQ-C30) at different treatment cycles.

Figure 5: Global Health Status/QoL (EORTC QLQ-C30) — Patients With RET Fusion–Positive NSCLC

EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; NSCLC = non–small cell lung cancer; QoL = quality of life; RET = rearranged during transfection.

Note: Data cut-off or December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Physical Function

The mean baseline obtained in the physical function scale for the overall population was 75.9 (SD = 22.2). The mean baseline score for physical function was 72.6 (SD = 24.1) in the treatment-naive subgroup, 79.8 (SD = 21.0) in patients with 1 prior line of therapy, and 76.1 (SD = 18.8) in the group with 2 or more lines of therapy. Baseline scores in the physical function scale in all 3 subgroups were lower than the threshold score of 83, and were considered clinically important impairments. The proportions of patients reported to have experienced improvement in physical function scores were higher across the subgroups and in the overall population compared to those who experienced worsened symptoms across cycle 3 to cycle 13. The majority of patients reported stable outcomes in physical function scores across cycles. However, due to the small sample size observed at later cycles, the interpretability of these findings is uncertain. An “improvement” (≥ 10-point increase) in physical function was observed at the start of cycle 3 (43.5%, 28.1% and 29.8% in the treatment-naive, 1 prior line, and 2 or more lines, respectively). The time-to-event analysis showed improvements in physical function in 55.6% (15 of 27), 41.0% (16 of 39) and 48.2% (26 of 54) of patients who were treatment-naive, had 1 prior line of therapy, or 2 or more prior lines of therapy, respectively.⁷ Figure 6 presents scores obtained in the physical function scale at different treatment cycles.

Figure 6: Physical Function (EORTC QLQ-C30) — Patients With RET Fusion–Positive

EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: Data cut-off: December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Symptom Subscales

Figure 7 presents the proportion of patients meeting a definite change in the RET fusion NSCLC population at the December 16, 2019, data cut-off.

Figure 7: Proportion of Patients Meeting “Definite” Change — RET Fusion–Positive NSCLC

NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: A definite change is defined as a ≥ 10 point change from their baseline score and no further change of ≥ 10 points in the score were observed at any subsequent assessment.

Data cut-off of December 16, 2019

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Dyspnea

The mean baseline scores obtained for dyspnea exceeded the clinically meaningful threshold of 17 points in the overall population and in each subgroup (overall = 31.3 [SD = 30.3]; treatment-naive = 28.4 [SD = 33.0]; 1 prior line of therapy = 23.1 [SD = 27.7]; ≥ 2 prior lines = 37.7 [SD = 28.3]) and were considered clinically meaningful impairments. The proportion of patients who experienced a change in dyspnea from baseline by cycle of study treatment was high in patients who reported improved symptoms compared to patients who reported worsened symptoms across cycles 3 to 13. The proportion of patients who reported stable outcomes was higher but constant across cycles assessed. Findings from the time-to-event analysis showed improvements in dyspnea (≥ 10-point decrease). In the treatment-naive group, 48.1% of patients (13 of 27) showed improvement, with 33.3% of patients in the 1 prior line of therapy group (13 of 39) and 61.1% in the 2 or more prior lines of therapy group (33 of 54) showing improvement.⁷

Insomnia

Baseline scores for insomnia did not meet a clinically meaningful threshold in any group and were not considered clinically meaningful impairments. The proportion of patients who reported stable symptoms in insomnia from baseline across cycles 3 to 13 was higher than that of patients who reported improved or worsened outcomes. More patients experienced improved outcomes compared to those who reported worsened outcomes across cycle 3 to 13. However, due to the decrease in number of patients completing the questionnaires from baseline to cycle 13, these findings are uncertain. The time-to-event analyses showed improvement (≥ 10-point decrease) in insomnia in 55.6% of the naive group (15 of 27), 38.5% of the 1 prior line of therapy group (15 of 39), and 50.0% in group 2 or more prior lines of therapy group (27 of 54). The median time to first improvement of insomnia was 4.1 months (95% CI, 1.9 to not reached) in the naive group, 13.8 months (95% CI, 3.7 to not reached) in the 1 prior line of therapy group, and 7.2 months (95% CI, 3.9 to not reached) in the 2 or more prior lines of therapy group.⁷

Fatigue

Baseline scores for fatigue did not meet a clinically meaningful threshold in the overall population (mean = 37.9; SD = 26.0) and so were not considered evidence of a clinically meaningful impairment. The 39-point threshold was exceeded in the naive group (baseline mean = 41.6; SD = 27.7) and the group with 2 or more prior lines of therapy (baseline mean fatigue = 41.8; SD = 26.3) and so was considered a clinically meaningful impairment. The threshold was not met in the 1 prior line of therapy group and so was not to be a clinically meaningful impairment. The time-to-event analyses showed improvements (≥ 10-point decrease) in fatigue in 77.8% of the naive group (21 of 27), 56.4% of the 1 prior line of therapy group (22 of 39), and 66.7% of the 2 or more prior lines of therapy group (36 of 54).

Pain

Baseline scores for pain met a clinically meaningful threshold of 25 points overall (mean = 29.4; SD = 28.5) and in all subgroups and was considered a meaningful impairment. Time-to-event analyses showed improvements (≥ 10-point decrease) in pain in 63.0% of the naive group (17 of 27), 46.2% of the 1 prior line of therapy group (18 of 39), and 50.0% of the 2 or more prior lines of therapy group (27 of 54). All line of therapy subgroups exceeded the clinically meaningful threshold of 8 points for nausea and vomiting; no other subscales met the threshold signifying a clinically meaningful problem.⁷

As fewer patients were available to complete questionnaires as cycles increased, the obtained findings should be interpreted with caution.

Progression-Free Survival

Primary and Integrated Analysis Sets (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

Table 16 summarizes information on PFS in the PAS and IAS populations. At the March 30, 2020, data cut-off, 52.4% of patients in the PAS population were alive and progression-free, with a median follow-up of 16.8 months by IRC. The median PFS in the PAS was estimated by the IRC to be 19.3 months (95% CI, 13.9 to NE). In the IAS population, 66.1% of patients were alive and progression-free, with a median follow-up of 13.6 months at the March 30, 2020, data cut-off. The median PFS was estimated as 19.3 months (95% CI, 16.5 to NE).

The sponsor conducted a follow-up analysis at a new data cut-off date of June 15, 2021. The median PFS estimated in the IAS population was 24.94 months (95% CI, 19.3 to NE).

Figure 8 and Figure 9 present the Kaplan–Meier plots of PFS based on IRC assessments of the PAS and IAS (December 16, 2019, data cut-off only).

Table 16: Progression-Free Survival — PAS and IAS (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

CI = confidence interval; IAS = integrated analysis set; IRC = independent radiographic committee; NE = not estimable; NR = not reached; PAS = primary analysis set.

Note: The PAS included a subset analysis of the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–programmed cell death protein 1 or anti–programmed cell death ligand 1 antibody. The IAS included all patients previously treated with platinum-based chemotherapy. Data cut-off dates were December 16, 2019; March 30, 2020; and June 15, 2021.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Figure 8: Kaplan–Meier Plot of Progression-Free Survival Based on IRC Assessment — Primary Analysis Set (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

+ = censored; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; PFS = progression-free survival; RET = rearranged during transfection.

Note:. December 16, 2019, data cut-off.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Figure 9: Kaplan–Meier Plot of Progression-Free Survival Based on IRC Assessment — Integrated Analysis Set (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

+ = censored; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; PFS = progression-free survival; RET = rearranged during transfection.

Note: December 16, 2019, data cut-off.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Treatment-Naive RET Fusion–Positive

Findings related to PFS in the treatment-naive population are presented in Table 17. At the March 30, 2020, data cut-off, 70.8% of patients were alive and progression-free at a median follow-up of 10.84 months. The median PFS by IRC assessment was NE.

At the December 16, 2019, data cut-off, 39 patients in the treatment-naive cohort had completed at least 6 months of follow-up from the first administered dose of selpercatinib. The median PFS was NE at the data cut-off date (95% CI, 13.8 to NE) according to the IRC assessment. Nine (23%) of 39 events had occurred, and a median follow-up of 9.2 months was reported. Figure 10 presents the Kaplan–Meier plot of PFS based on the IRC assessment.⁵

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The median PFS estimated in the treatment-naive population was 21.95 months (95% CI, 13.8 to NE).

Table 17: Progression-Free Survival — Treatment-Naive RET Fusion–Positive

IRC = independent radiographic committee; NE = not estimable; NR = not reported; NSCLC = non–small cell lung cancer; PFS = progression-free survival; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: SAS1 included patients who were treatment naive. Data cut-off dates of December 16, 2019, March 30, 2020, and June 15, 2021.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Figure 10: Kaplan–Meier Plot of Progression-Free Survival Based on IRC Assessment — Treatment-Naive RET Fusion–Positive NSCLC

+ = censored; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; PFS = progression-free survival; RET = rearranged during transfection.

Note: Data cut-off date of December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Prior Other Systemic Therapy RET Fusion–positive NSCLC (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC (Supplemental Analysis Set 3)

PFS findings obtained in the SAS2 and SAS3 populations are presented in Table 18. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

Table 18: PFS — Prior Other Systemic Therapy RET Fusion–Positive NSCLC (Supplemental Analysis Set 2) and Non-measurable Disease RET Fusion–Positive NSCLC (Supplemental Analysis Set 3)

NE = not estimable; NSCLC = non–small cell lung cancer; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: SAS2 included patients treated with other systemic therapies that are not platinum-based. SAS3 included patients with non-measurable RET fusion–positive NSCLC. Data cut-off dates of December 16, 2019, and March 30, 2020.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Subgroup Analysis

Data for the subgroup analyses were not provided.

Objective Response Rate

Primary and Integrated Analysis Sets (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

The ORR obtained at both cut-off dates (March 30, 2020, and December 16, 2019) in the PAS and IAS is summarized in Table 19.

At the December 16, 2019, data cut-off, an ORR of 64% (95% CI, 53.9 to 73.0) was estimated by the independent radiographic committee (IRC) in the PAS population, and in the IAS population, an ORR of 57% (104 of 184; 95% CI, 49.0 to 63.8) was estimated by the IRC.⁵

At the March 30, 2020, data cut-off, an ORR of 63.8% (53.9 to 73.0) and ORR of 56.9% (CI, 50.0 to 63.6) were reported by the IRC in the PAS and IAS, respectively. Of the patients with a confirmed best response of a CR or PR, 58.2% were ongoing in the PAS at a median follow-up of 15.67 months, and 69.4% of responses were ongoing in the IAS at a median follow-up of 11.99 months.

The sponsor conducted another analysis at a new data cut-off of June 15, 2021. The ORR estimated in the IAS population was 61.1% (95% CI, 54.7 to 67.2), which was consistent with previous analyses.

Table 19: Objective Response Rate in the PAS and IAS (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

CI = confidence interval; IAS = integrated analysis set; PAS = primary analysis set; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: The PAS includes a subset analysis of the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–programmed cell death protein 1 or anti–programmed cell death ligand 1 antibody. It also includes 5 patients with non-measurable disease according to the investigator, and 4 patients with non-measurable disease according to the IRC. The IAS includes all patients previously treated with platinum-based chemotherapy. Data cut-off dates of December 16, 2019; March 30, 2020; and June 15, 2021.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Treatment-Naive RET Fusion–Positive NSCLC (SAS1)

Table 20 summarizes the ORR obtained in the treatment-naive RET fusion–positive NSCLC population. At the December 16, 2019, data cut-off, the IRC-assessed ORR for the treatment-naive RET fusion–positive NSCLC population was 85% (33 of 39, 95% CI, 69.5 to 94.1).⁵ At the March 30, 2020, data cut-off, the ORR by IRC assessment was 85.4% for the treatment-naive set and consisted of 2.1% CRs and 83.3% PRs.

The sponsor conducted follow-up analysis with a new data cut-off of June 15, 2021. The ORR estimated in the treatment-naive population was 84.1% (95% CI, 73.3 to 91.8), which was consistent with previous analyses.

Table 20: Objective Response Rate — Treatment-Naive RET Fusion–Positive NSCLC

CI = confidence interval; NSCLC = non–small cell lung cancer; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: SAS1 included patients who were treatment naive.

^aTwo partial responses were unconfirmed.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Prior Other Systemic Therapy RET Fusion–Positive NSCLC (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC (Supplemental Analysis Set 3)

The ORR obtained by the IRC assessment in the prior other systemic therapy RET fusion–positive NSCLC (SAS2) was |||||||||||||||||||||||||||||||||| at the December 16, 2019, analysis and 44.4% (8 of 18; 95% CI, 21.5 to 69.2) at the March 30, 2020, analysis in patients with RET fusion–positive NSCLC who had received a previous systemic therapy other than a platinum-based chemotherapy. Table 21 summarizes ORR data obtained at both cut-offs in this subset.

The ORR obtained by the IRC assessment in the non-measurable disease RET fusion–positive NSCLC (SAS3) was |||||||||||||||||||||||||||||||||| at the December 16, 2019, data cut-off and 33.3% (95% CI, 13.3 to 59.0) at the March 30, 2020, data cut-off.

Table 21: ORR – Prior Other Systemic Therapy RET Fusion–Positive NSCLC (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC (Supplemental Analysis Set 3)

CI = confidence interval; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Subgroup Analysis

CNS Metastasis at Baseline (Overall NSCLC Population)

At the December 16, 2019, data cut-off, a total of 80 patients in the overall RET fusion–positive NSCLC population (independent of analysis set, n = 253) had undergone an investigator-assessed CNS metastasis at baseline and had accumulated at least 6 months of follow-up from the first administered selpercatinib dose. Table 22 presents an overview of the CNS ORR in the overall population of RET fusion–positive NSCLC with measurable and non-measurable CNS lesions at baseline.

A CNS measurable lesion (by RANO criteria) was defined in the LIBRETTO-001 trial as: having clearly defined margins; visible on 2 or more axial slices, preferably less than 5 mm thick; at least 10 mm in size if slice thickness is less than 5 mm (or twice the slice thickness if more than 5 mm thick); and not measuring a cystic cavity. Non-measurable CNS lesions were defined as those that did not fit the criteria for measurable lesion; more specifically, lesions that are cystic or necrotic or include a surgical cavity should not be considered measurable.

At the December 16, 2019, data cut-off, the IRC identified 22 patients with measurable CNS disease out of the 80 with CNS metastasis at baseline. Among the 22 patients, a CNS ORR of 82% (18 of 22; 95% CI, 59.7 to 94.8) was reported, with 5 patients (23%) reportedly exhibiting CRs and 13 patients (59%) exhibiting PRs. The median response was 9.4 months (95% CI, 7.4 to NE). Stable disease was reported in 18% of patients.^5,24 At the March 30, 2020, data cut-off, of the 23 patients assessed with measurable disease at baseline, the estimated CNS ORR was 87% (95% CI, 66.5 to 97.2) — 21.7% had a CR and 65.2% had a PR. The median CNS duration was 9.4 months (95% CI, 6.7 to 12.1).

Table 22: CNS ORR by IRC Assessment — CNS Analysis Set With a Measurable Lesion (RET Fusion–Positive NSCLC)

CI = confidence interval; CNS = central nervous system; CR = complete response; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; NR = not reported; ORR = objective response rate; PR = partial response; RET = rearranged during transfection.

Note: % is calculated based on the number of patients with CNS best response of CR or PR as denominator.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Table 23 presents an overview of the CNS ORR obtained in the overall RET fusion–positive NSCLC population with measurable and non-measurable CNS lesions at baseline at the 2 data cut-offs (March 30, 2020, and December 16, 2019). At the December 16, 2019, data cut-off, of 80 patients with CNS disease at baseline, a CNS ORR of 48% (38 of 80; 95% CI, 36.2 to 59.0) was reported by the IRC. At the March 30, 2020, data cut-off, of the 96 patients with CNS at baseline, the estimated CNS ORR was 46.9% (95% CI, 36.6 to 57.3).

The median CNS duration of response at the March 30, 2020, data cut-off was 9.36 months (95% CI, 6.7 to 12.1), with a follow-up duration of 12.98 months. At the December 16, 2019, data cut-off, the median DOR was NE (95% CI, 9.3 to NE) at a median duration of follow-up of 9.5 months. Twenty-five patients (31%) had a CNS DOR of at least 6 months, and 5 patients (6%) had a CNS DOR of at least 12 months. The assessment identified 25 patients (31%) with a CNS DOR of at least 6 months, and 5 patients (6%) with a CNS DOR of at least 12 months.⁵

Table 23: CNS ORR by IRC Assessment — Analysis Set with Measurable and Non-Measurable CNS Lesions

CI = confidence interval; CNS = central nervous system; CR = complete response; IRC = independent radiographic committee; NE = not estimable; NR = not reported; ORR = objective response rate; PR = partial response.

Note: Data cut-off dates of December 16, 2019, and March 30, 2020.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Performance Status, Number of Prior Therapies, Prior Anti–PD-1 and Anti–PD-L1 Therapy, and Prior Multikinase Inhibitor Therapy Subgroup Analysis

The ORR and DOR obtained in the subgroups (performance status, number of prior therapies, prior anti–PD-1 and anti–PD-L1 therapy, and prior multikinase inhibitor therapy) at the December 16, 2019, data cut-off are presented in Table 24.

Table 24: ORR and DOR — ECOG PS, Number and type of Prior Therapy Based on IRC Assessment

+= censored; CI = confidence interval; DOR = duration of response; ECOG PS = Eastern Cooperative Oncology Group Performance Status; IRC = independent radiographic committee; ORR = objective response rate; NR = not reached; PD-1 = programmed cell death protein 1; PD-L1 = programmed cell death ligand 1.

Note: Data cut-off date of December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Duration of Response

Primary and Integrated Analysis Set (RET Fusion–Positive NSCLC with Prior Platinum Chemotherapy)

The DORs obtained at both cut-off dates (March 30, 2020, and December 16, 2019) in the PAS and IAS are summarized in Table 25. At the December 16, 2019, data cut-off, a median DOR of 17.5 months (95% CI, 12.0 to NE) was reported by IRC, 24 (34%) of 67 events had occurred, and a median DOR follow-up of 12.1 months was estimated in the PAS population. In the IAS, a median DOR of 17.5 months (95% CI, 12.1 to NE), with 25 (24%) of 104 events, and a median DOR follow-up of 9.2 months was reported in the IRC assessment.

At the March 30, 2020, data cut-off, the median DOR reported by the IRC was 17.51 months (95% CI, 12.1 to NE) in the PAS and the median DOR was 17.51 months (95% CI, 12.1 to NE) in the IAS population.

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The median DOR obtained in the IAS population was 28.6 months (95% CI, 20.4 to NE).

The Kaplan–Meier estimates (December 16, 2019, cut-off) for the PAS population and the IAS population are presented in Figure 11 and Figure 12, respectively.⁵

Table 25: Duration of Response — PAS and IAS (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

CI = confidence interval; CR = complete response; IAS = integrated analysis set; IRC = independent radiographic committee; NE = not estimable, NR = not reported; NSCLC = non–small cell lung cancer; PAS = primary analysis set; PR = partial response; RET = rearranged during transfection.

^aIncludes censored patients whose disease has not yet progressed.

Note: The PAS includes the FDA breakthrough therapy designation population (n = 58) who received prior platinum-based chemotherapy and an anti–programmed cell death protein 1 or anti–programmed cell death ligand 1 antibody. The IAS included all patients previously treated with platinum-based chemotherapy. Data cut-off dates of December 16, 2019, March 30, 2020, and June 15, 2021.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Figure 11: Kaplan–Meier Plot of Duration of Response Based on IRC Assessment — Primary Analysis Set (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy

+ = censored; DOR = duration of response; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: Data cut-off date of December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Figure 12: Kaplan–Meier Plot of Duration of Response Based on IRC Assessment —Integrated Analysis Set (RET Fusion–Positive NSCLC With Prior Platinum Chemotherapy)

+ = censored; DOR = duration of response; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note: Data cut-off date of December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Treatment-Naive RET Fusion–Positive NSCLC Population

Table 26 summarizes information on the DOR obtained in the treatment-naive RET fusion–positive NSCLC population. At the December 16, 2019, data cut-off, the DOR in 33 patients in the treatment-naive cohort was assessed by the IRC. The median DOR was NE. At the data cut-off date of March 30, 2020, 65% of the responses were ongoing at 12 months, according to the IRC assessment. Figure 13 presents the Kaplan–Meier plot of the DOR based on the IRC assessment.⁵

The sponsor conducted a follow-up analysis at a new data cut-off of June 15, 2021. The median DOR estimated in the treatment-naive population was 20.2 months (95% CI, 13.0 to NE).

Table 26: Duration of Response — Treatment-Naive RET Fusion–Positive NSCLC Population

CI = confidence interval; CR = complete response; NE = not estimable; NR = not reached; NSCLC = non–small cell lung cancer; PR = partial response; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: SAS1 includes patients who were treatment naive. Data cut-off dates of December 16, 2019; March 30, 2020; and June 15, 2021.

^aIncludes censored patients whose disease has not yet progressed.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Prior Other Systemic Therapy RET Fusion–Positive NSCLC (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC (Supplemental Analysis Set 3)

The DOR obtained in the prior other systemic therapy RET fusion–positive NSCLC (SAS2) and non-measurable disease RET fusion–positive NSCLC (SAS 3) populations at both data cut-offs are presented in Table 27.

Figure 13: Kaplan–Meier Plot of Duration of Response Based on IRC Assessment — Treatment-Naive RET Fusion–Positive NSCLC Population

+ = censored; DOR = duration of response; IRC = independent radiographic committee; NSCLC = non–small cell lung cancer; RET = rearranged during transfection.

Note:.Data cut-off date of December 16, 2019.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Table 27: DOR — Prior Other Systemic Therapy RET Fusion–Positive NSCLC (Supplemental Analysis Set 2) and Non-Measurable Disease RET Fusion–Positive NSCLC (Supplemental Analysis Set 3)

CI = confidence interval; DOR = duration of response; NE = not estimable; NR = not reported; NSCLC = non–small cell lung cancer; RET = rearranged during transfection; SAS = supplemental analysis set.

Note: SAS2 included patients treated with other systemic therapies that are not platinum-based. SAS3 included patients with non-measurable disease RET fusion–positive NSCLC. Data cut-off dates of December 16, 2019, and March 30, 2020.

Source: Sponsor’s submission package, Retevmo (selpercatinib).⁵

Subgroup Analysis

Performance Status, Number of Prior Therapies, Prior Anti–PD-1 and Anti–PD-L1 Therapy, and Prior Multikinase Inhibitor Therapy Subgroup Analysis

The DORs obtained in the subgroups of interest identified in the CADTH protocol are presented in Table 24.

Harms

Only harms identified in the CADTH review protocol are reported here.⁵ AEs are presented for the OSAS (March 30, 2020, and December 16, 2019, data cut-offs) and for the NSCLC population (December 16, 2019, data cut-off) in the LIBRETTO-001 trial (Table 28). The focus for the review is the frequency of AEs on the NSCLC population.

Adverse Events

At the December 16, 2019, data cut-off, the most common AEs in the NSCLC population were dry mouth, diarrhea, hypertension, increased AST and ALT, fatigue, constipation, peripheral edema, headache, and nausea.⁵

At the March 30, 2020, data cut-off, AEs specific to the NSCLC population were not presented. In total, 99.7% of patients (n = 345) in the NSCLC population had at least 1 treatment-emergent AE and 61.2% had grade 3 or 4 events. The AEs reported in the NSCLC population were consistent with the overall safety set at both data cut-offs.

Serious Adverse Events

At the December 16, 2019, data cut-off, the most common SAEs in this population were pneumonia, increased ALT and AST, abdominal pain, pleural effusion, drug hypersensitivity, diarrhea, and acute kidney injury.⁵ At the March 30, 2020, data cut-off, SAEs specific to the NSCLC population were not reported.⁶ Table 28 presents SAEs reported in the NSCLC population at the December 16, 2019, data cut-off and in the overall safety population at the 2 data cut-offs (March 30, 2020, and December 16, 2019).

Withdrawals

At the December 16, 2019, data cut-off, AEs leading to dose withdrawal, interruption, and dose reductions in the RET fusion–positive NSCLC population were consistent with those observed in the overall safety set at both data cut-offs (March 30, 2020, and December 16, 2019). AEs commonly associated with treatment discontinuations in the NSCLC population were increased ALT, sepsis, increased AST, drug hypersensitivity, fatigue thrombocytopenia. Increased ALT and AST were commonly associated with dose reductions and dose interruptions. Table 28 presents data on the AEs leading to withdrawals, reductions, and interruptions in the OSAS (at the March 30, 2020, and December 16, 2019, cut-offs) and NSCLC populations.

Mortality

At the December 16, 2019 data cut-off, treatment was discontinued owing to death in 6 patients and 38 patients discontinued the study due to death. At the March 30, 2020 data cut-off, in the NSCLC population, treatment was discontinued owing to death in 6 (1.7%) patients and 55 (15.9%) patients discontinued the study due to death. In total, 36 patient deaths (10.4%) were attributed to disease progression, 13 of deaths (3.9%) occurred due to AEs, and 6 deaths (1.7%) were attributed to other reasons. One report of death (0.3%) identified in the NSCLC population had occurred more than 28 days after the last selpercatinib dose. In the OSAS population, treatment was discontinued due to death in 11 patients (1.5%) and 103 patients (13.8%) discontinued the study due to death.

Notable Harms

Notable harms reported in the NSCLC set were consistent with those in the OSAS (at the March 30, 2020, and December 16, 2019, data cut-offs). The most common AEs were increased ALT (32.8%), increased AST (31%), hypertension (31.9%), drug hypersensitivity (2.4%), and electrocardiogram QT prolongation (16.7%). Table 28 presents a summary of notable harms in the OSAS (at the 2 data cut-offs) and in the NSCLC population.

Table 28: Summary of Harms in the LIBRETTO-001 Trial for the Overall Safety Analysis Set (March 30, 2020, and December 16, 2019, Data Cut-Offs) and the NSCLC Population (December 16, 2019, Data Cut-Off)

AE = adverse event; ALT = alanine transaminase; AST = aspartate transaminase; NR = not reported; NSCLC = non– small cell lung cancer; OSAS = overall safety analysis set; NR = not reported.

Source: Sponsor’s submission package, Retevmo (selpercatinib)⁵ and European Medicines Agency report.⁶

Critical Appraisal

Internal Validity

The phase I and II design of the LIBRETTO-001 trial was used to determine the target population and the optimal drug dosage(s) for phase III.²⁷ The design was insufficient for evaluating long-term outcomes for new drugs because of potential biases (e.g., inflated type I error due to the small sample sizes and multiple comparison tests conducted).²⁸ The Notice of Compliance issued by Health Canada for the regulatory approval of selpercatinib outlined several conditions in accordance with the NOC/c, |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The sponsor has initiated a phase III trial (LIBRETTO-431) to assess the efficacy and safety of selpercatinib against active comparators currently used in clinical practice in the treatment-naive population (NCT04194944).²⁹ The clinical experts consulted by CADTH noted that there is limited evidence that patients with metastatic RET fusion–positive NSCLC benefit from currently available single-drug immunotherapy in practice, and not all patients respond to chemotherapy plus immunotherapy, indicating an unmet need for this patient population.

The open-label, non-comparative design of the LIBRETTO-001 trial, with no statistical testing, is the key limitation. The sponsor did not provide any hypothesis statements for statistical significance for the primary outcome, including secondary and subgroup analyses.⁸ The design increases the risk of bias in estimating treatment effects because the potential for confounding effects related to variations in health status and other unidentified prognostic factors that could affect subjectively assessed outcomes (i.e., response, HRQoL, and AEs). The potential for bias was reduced by using IRC assessments for key study outcomes, such as ORR and DOR. However, the lack of direct comparative data means the magnitude of effects obtained for the efficacy outcomes is uncertain. Although the clinical experts emphasized that the safety profile of selpercatinib was favourable compared to the other therapies available as standard of care in Canada, in the absence of a comparative arm, the findings obtained from the safety analysis are uncertain as the single-arm design does not allow for the differentiation of the symptoms of underlying NSCLC disease from treatment-related AEs.⁹ The sponsor agreed to provide results from ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| to confirm the clinical benefit of selpercatinib in patients with previously treated RET fusion–positive NSCLC according to the NOC/c issued by Health Canada. The sponsor noted in its response that ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The sponsor submitted additional data with a new cut-off date of June 15, 2021, during the completion of this CADTH review. Despite the results obtained from the updated (pre-specified or post hoc) analyses performed on June 15, 2021, whether the end points investigated are durable for long-term in this patient population remains uncertain.

The primary objective investigated in the phase II portion of the LIBRETTO-001 study was the ORR measured by RECIST 1.1. The FDA considers ORR alone to be a surrogate measurement when assessing treatment response in patients with advanced or metastatic NSCLC, and it may not correlate well with survival, unless the effect size of the ORR is large and the responses are durable. The sponsor hypothesized a true ORR of 50% or higher in the primary analysis of effectiveness and ruled out a lower limit of 30% for the ORR, which was considered clinically meaningful and consistent with the estimated response rates seen with approved targeted therapies in molecularly defined patient populations who have failed prior therapies (e.g., osimertinib, crizotinib, and alectinib). The ORRs obtained by the IRC for the PAS, IAS, and SAS1 were above the lower limit of 30% that the sponsor assumed in the sample size calculation for patients with RET fusion–positive NSCLC who progressed on or after receipt of platinum-based chemotherapy.⁵ Although the FDA review team noted that the magnitude of the ORR and DOR obtained in patients with RET fusion–positive NSCLC in the LIBRETTO-001 trial was large and considered it sufficient to establish clinical benefit,¹⁰ |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||⁹ The sponsor agreed to provide results from ||||||||||||||||||||||||||||||||||||||||||||||||||| to confirm the clinical benefit of selpercatinib in patients with previously treated RET fusion–positive NSCLC according to the NOC/c issued by Health Canada. The sponsor also noted that |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||⁹ In the LIBRETTO-001 trial, radiographic scans were performed by an accredited laboratory and assessed by an IRC, reducing bias. Results obtained from both the IRC and investigator assessments did not differ greatly, which increases the validity of the ORR-related outcomes.

The time-to-events analyses, particularly the OS and PFS results, were considered exploratory by |||||||||||||||||, the FDA,^10,30 and CADTH due to the lack of a control arm. Health Canada ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||⁹ Due to immature data in the PAS, IAS, and SAS1 populations at the March 31, 2020, and December 16, 2019, cut-offs, CADTH considered the results exploratory. The sponsor submitted additional data to CADTH for a new cut-off date of June 15, 2021, later during the completion of the review. Although the sponsor reported a median PFS of 24.94 months (95% CI, 19.3 to NE) in the IAS and 21.95 months (95% CI, 13.8 to NE) in the SAS1 population, the information was considered insufficient to form concrete conclusions on PFS in this population because of the single-arm design of the LIBRETTO-001 trial and immature data. The median OS was NE in the IAS and SAS1 datasets at the June 15, 2021, cut-off. Whether the observed magnitude of benefit related to tumour response with selpercatinib would be translated as OS in patients in the 2 groups is therefore uncertain.

The analysis sets — the PAS, IAS, and 3 SASs — were not predefined in the original statistical analysis plan; they were developed following consultation with the FDA and EMA. A key concern therefore is that these were post hoc analyses and may have been susceptible to bias. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| CADTH reviewers agreed |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||, these were unlikely to introduce bias because the investigators remained blinded to results until after the revisions were made.⁹

The 3 planned interim analyses were pre-specified. However, the analyses were performed on observed data only, with no formal hypothesis testing, and only descriptive statistics were provided, further complicating assessment of the magnitude of effects observed in the different groups investigated.

Many of the subgroup analyses that were identified as relevant for this review were pre-specified in the statistical analysis plan for LIBRETTO-001. However, the subgroups had limited sample sizes, resulting in imprecise results for many of the subgroups. There were no tests for statistical differences between subgroups.

HRQoL was evaluated using the EORTC QLQ-C30 questionnaire. This questionnaire has been validated in patients with NSCLC and is considered appropriate for advanced or metastatic forms of the disease. Although this was an exploratory outcome investigated in the LIBRETTO-001 trial, the clinical experts and clinician groups noted that findings for quality of life are clinically significant because patients with advanced disease usually experience greater symptom burdens due to disease progression. The experts consulted during the CADTH review noted that HRQoL findings were clinically meaningful to patients, particularly because available therapies have not been confirmed to improve HRQoL in patients. However, there is uncertainty regarding these findings because the number of patients who completed questionnaires decreased from baseline through to cycle 13, resulting in considerable missing data at later time points. The magnitude of effect may have been overestimated due to the small sample size in the overall population and in the subgroups investigated. In addition, whether these findings are generalizable to patients who underwent treatments beyond cycle 13 is uncertain. As indicated by the sponsor, fewer patients were available to complete questionnaires at later cycles. There is also a potential for selection bias over time given that long-term survivors in the trials tend to be healthier patients. In the absence of a comparator arm and given the possibility that the open-label design introduced reporting bias, the impact of selpercatinib on patient-reported outcomes in relation to other therapies is unknown. Although CADTH recognizes that the rarity of RET fusion–positive mutations in patients with NSCLC may have contributed to the small sample size in the datasets at baseline and may have influenced the number of patients available to complete the questionnaires at later stages of the trial, no strong, definitive conclusions can be made from the findings obtained for HRQoL in the different population sets of patients with NSCLC.

External Validity

The LIBRETTO-001 trial is a multi-centre study with more than 84 participating sites across different countries, including Canada and the US. The clinical experts consulted during the review acknowledged that the baseline characteristics observed in the trial were reflective of patients seen in Canadian practice settings, and the results are therefore generalizable to the Canadian population.

According to published literature, RET gene fusions are more common in patients with NSCLC and adenocarcinoma histology and are rare in patients with squamous cell carcinoma. The clinical experts noted that they would generally expect a patient who is RET fusion–positive to have adenocarcinoma, as driver mutations are rare in squamous cell carcinoma. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||^| The clinical experts indicated that, if a non-adenocarcinoma patient was to be found to harbour an RET fusion, they should still be offered selpercatinib, as the presence of the mutations far outweighs the histology in terms of biology and likely response to targeted therapy.

Most patients in the NSCLC population were between the ages of 45 and 74 years. The clinical experts consulted confirmed that the median age of patients in the NSCLC population of the LIBRETTO-001 trial is reflective of Canadian practice. The experts added that, among the patients with NSCLC seen in practice, those with RET fusion–positive status tended to be younger than average lung cancer patients. These patients are also likely to have had no beforebacco exposure. The population of the trial is therefore representative of the Canadian practice setting with respect to these characteristics.

The patients with RET fusion–positive NSCLC recruited in the trial had an ECOG PS of 0, 1, or 2; however, only approximately 2% of enrolled patients had an ECOG PS of 2. It is uncertain if the results are generalizable to patients with an ECOG PS of 2, 3 or 4 with worse performance status. The clinical experts consulted during the review stated that they would still use selpercatinib in patients with an ECOG PS of 2 or 3 because selpercatinib is well-tolerated and would improve symptom burden, and therefore may improve ECOG PS. The experts discouraged the use of selpercatinib in patients with an ECOG PS of 4 or organ dysfunction (e.g., liver failure), in patients lacking an RET fusion, or those who had previously received another selective RET inhibitor, including patients who are unable to take medications orally or absorb medications due to gastrointestinal complications.

Most patients had advanced or metastatic disease, most of which were stage IV. Few patents with stage I to III disease were enrolled in the group with prior systemic therapy. The clinical experts noted that the use of selpercatinib in the metastatic setting is not intended as a curative option for patients and advised against using selpercatinib in patients who are eligible to receive curative-intent therapies. The clinician experts emphasized that patients with early-stage disease may only be eligible to receive selpercatinib if they were confirmed as having incurable disease or were not eligible or refused potentially curative local therapies, such as surgery or radiation therapy.

The median number of therapies used in the prior systemic therapy group was 2 (half of the IAS population received at least 2 prior lines of therapy and a quarter had received more than 3 prior lines). Given that most evidence on the efficacy of selpercatinib was generated in the subset of patients who were RET fusion positive who had received prior systemic therapies (more patients in that subset compared to the treatment-naive subset), it was important to assess whether the number of prior lines of therapy administered to patients could influence patient response to selpercatinib. In addition, it is uncertain whether there will be differences in outcomes in the treatment-naive and the prior systemic therapy group in practice, given the study design and the small sample size of the group that was treatment naive. The clinical experts indicated that the number of lines of therapy a patient had received before receiving selpercatinib will not influence a patient’s response to selpercatinib based on the drug’s mechanism of action.

The clinical experts consulted during the CADTH review noted that most patients in the NSCLC population have comorbidities, some of which include cardiovascular disease, which is to be expected given that some patients are long-time smokers. The experts also emphasized that, in practice, patients with underlying cardiovascular disease will usually be referred to cardio-oncology subspecialists from cardiology who will treat the disease while the patient undergoes cancer treatments. As follow-up sessions will require electrocardiogram checks, the clinical experts concluded that that administering selpercatinib to these patients will not present an issue because they will be constantly followed throughout treatments.

Selpercatinib dosing and the drug administration interval in the trial align with Health Canada’s indication and are generalizable to the Canadian setting. Dose adjustments made in the trial for patients experiencing AEs and variation in body weight, including those using concomitant therapies, are explicitly described in the product monograph to ensure that patients receive adequate therapy throughout treatment.

In the trial, RET fusion mutations were identified by PCR or NGS techniques. The only approved molecular testing technique was FISH, which was considered an acceptable technique for detecting RET fusion tumours for the phase I dose-escalation portion of LIBRETTO-001 and for cohort 5, but not for cohorts 1 and 2 of phase II. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The sponsor recommended the NGS technique due to its accuracy in detecting oncogenic fusions and other alterations. The clinician group input and the clinical experts pointed out that NGS and PCR testing are available in Canadian practice, and testing for multiple oncogenic drivers is routinely conducted at initial diagnosis. According to the clinician experts consulted during the review, RET fusion will be included as part of the larger panel of targetable mutations for which testing is conducted upfront at diagnosis. As noted by the clinician, NGS is the most common molecular technique used for oncogenic testing among others available across jurisdictions in Canada (e.g., reverse-transcription PCR, FISH, immunohistochemistry, and NGS with DNA and RNA sequencing). Both reverse-transcription PCR and FISH were used in the trial.

Tumour assessments were conducted at screening within 28 days of cycle 1 at day 1 via CT or MRI. Post-baseline assessments were conducted every 8 weeks (± 1 week) from cycles 3 through 13. The clinician experts indicated that the frequency of assessments implemented in the trial was appropriate, but the frequency differed from real-world practice. The clinical experts consulted emphasized that it is standard practice to perform radiographic assessments every 8 to 12 weeks. However, this time frame may be accelerated if patients report new symptoms or their physical findings indicate progression. In addition, the clinical experts noted that, while patients receiving oral targeted therapies are assessed for the presence and severity of symptoms every 3 to 4 weeks, these timelines may be shortened at treatment initiation and spaced out after therapy has been established and patients are doing well in terms of tolerance and symptom control.

The majority of patients enrolled in the trial were evaluated for outcomes after completing 6 months of follow-up following the first administered dose of selpercatinib as of the first interim data cut-off date (June 17, 2019). |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| The clinical experts consulted by CADTH considered the duration of treatment implemented in the trial to be appropriate and the duration of follow-up sufficient to evaluate the tumour response in patients with RET fusion–positive NSCLC.

Indirect Evidence

Objectives and Methods for the Summary of Indirect Evidence

As there was no direct evidence comparing selpercatinib to other active therapies for the treatment of RET fusion–positive NSCLC in adult patients, a review of indirect evidence was undertaken. CADTH conducted a literature search to identify potentially relevant ITCs in patients with RET fusion–positive NSCLC. A focused literature search for ITCs dealing with selpercatinib or NSCLC was run in MEDLINE All (1946–) on November 3, 2021. No filters were applied to limit retrieval by study type. Retrieval was not limited by publication date or by language. Titles, abstracts, and full-text articles were screened for inclusion by 1 reviewer based on the population, intervention, comparator, and outcome criteria outlined in Table 28.

No potentially relevant ITCs were identified in the literature search.

Two sponsor-submitted ITCs were summarized and critically appraised, 1 for treatment-naive NSCLC¹¹ and the other for previously treated NSCLC.¹² These ITCs were used to inform the pharmacoeconomic models.

Description of Indirect Comparisons

Both sponsor-submitted ITCs included a systematic review of the literature (to identify trials investigating selpercatinib or comparator interventions in patients with RET fusion NSCLC or papillary thyroid cancer or RET-mutant MTC) and a network meta-analysis (NMA) that compared selpercatinib to other active treatments in patients with NSCLC who were treatment-naive or who had received previous treatments. In the sponsor-submitted ITCs, selpercatinib was compared to various chemotherapies or immunotherapies.

Methods of ITCs of Patients With NSCLC Who Were Treatment Naive and Treatment Experienced

Objectives

The objective of the sponsor-submitted report for patients with NSCLC who are treatment naive was to identify evidence from clinical trials and, if possible, an NMA, to evaluate the relative efficacy of selpercatinib versus other active treatments for this population.

The objective of the sponsor-submitted report for patients with NSCLC who are treatment experienced was to identify evidence from clinical trials and if possible, an NMA, to evaluate the relative efficacy of selpercatinib versus other active treatments for this population.

Study Selection Methods

Patients Who Were Treatment Naive

The randomized controlled trials (RCTs) that were used to inform the ITC were identified through a systematic literature search conducted by the ITC authors. Multiple databases were searched to identify clinical trials that evaluated the efficacy of drug therapies for RET fusion–positive NSCLC. It is unclear if the studies were selected independently by 2 reviewers. Data extraction was performed by 1 reviewer and verified by a second reviewer. It was unknown if the quality of the included studies was assessed.

Patients Who Were Treatment Experienced

The RCTs and single-arm trials that were used to inform the ITC were identified through a systematic literature search conducted by the ITC authors. Multiple databases were searched to identify clinical trials that evaluated the efficacy of drug therapies for RET fusion–positive NSCLC. It is unclear if the studies were selected independently by 2 reviewers. Data extraction was performed by 1 reviewer and verified by a second reviewer. Quality of the included RCTs was assessed using the standard National Institute for Health and Care Excellence checklist. Quality of the included single-arm trials was assessed using the Critical Appraisal Skills Programme cohort study checklist.

Inclusion and exclusion criteria for the clinical studies for each ITC are presented in Table 29.

Indirect Treatment Comparison Analysis Methods

Patients Who Were Treatment Naive

As LIBRETTO-001 was a single-arm study, a pseudo-control arm was needed to connect the first-line treatment arm in the LIBRETTO-001 study (selpercatinib) to the other comparators in the network. The ITC authors used individual patient-level data (IPD) available for the pemetrexed plus platinum chemotherapy arm from the KEYNOTE-189 study to simulate this pseudo-control arm. The KEYNOTE-189 study was an RCT comparing pembrolizumab combined with pemetrexed and platinum chemotherapy versus pemetrexed plus platinum chemotherapy alone in patients with advanced or metastatic NSCLC who had not previously received systemic therapy. Real-world data from the Flatiron database were used to adjust the pemetrexed plus platinum arm from the KEYNOTE-189 study for the prognostic impact of an RET fusion–positive status, and an adjustment factor was estimated for the prognostic effect of the RET fusion status. Subsequently, adjustment was done to match other prognostic factors of the RET-adjusted pemetrexed plus platinum arm from the KEYNOTE-189 study to the selpercatinib arm in the LIBRETTO-001 study. These prognostic factors included: |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Three statistical methods were used: propensity score matching (PSM) using logistic regression, propensity score weighting (PSW) using generalized boosted models, and targeted minimum loss-based estimation. For OS and PFS, Cox regression and nonparametric model were applied to the generated pseudo-control arm and selpercatinib to estimate the treatment effect of selpercatinib. The hazard ratios estimated for selpercatinib were included in the NMA to estimate the relative treatment effects of selpercatinib and comparators.

Bayesian NMAs were conducted. These analyses formed the base-case results and allowed for comparisons between treatments in the included trials. Frequentist NMAs were conducted as sensitivity analyses. Subgroup analyses based on the level of PD-L1 expression (≥ 50%) were conducted using a Bayesian approach. All the analyses were performed for each of the end points (OS, PFS, and overall response rate), using both fixed-effect and random-effect models.

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

When PSM by logistic regression was used, the results were more clinically plausible for pemetrexed plus platinum. This model was therefore chosen to match data from the pemetrexed plus platinum and selpercatinib arms.

Patients With Treatment Experience

For the patients who had received prior systemic therapies, a pseudo-control arm was also needed to connect the second- or later-line treatment cohort of the LIBRETTO-001 study to other active treatments. This pseudo-control arm was simulated for LIBRETTO-001 using IPD available for the docetaxel plus placebo arm from the REVEL study, which was an RCT assessing the effectiveness of docetaxel plus ramucirumab to docetaxel plus placebo in the treatment of patients with stage IV NSCLC in the second-line setting. To adjust the docetaxel plus placebo control arm for RET fusion–positive status, real-world data from the Flatiron database were used. Patients were selected from the Flatiron database if they had advanced or metastatic NSCLC, were RET fusion–positive, and had received prior systemic therapy. The size of the impact was estimated from Flatiron database information on both RET fusion–positive and –negative patients. A subsequent adjustment was done to match other prognostic factors of the RET-adjusted docetaxel control arm to the selpercatinib arm. These prognostic factors included ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Similar to the ITC of the treatment-naive population, 3 different methods were used: PSM with logistic regression, PSW with generalized boosted models, and targeted minimum loss-based estimation. For OS and PFS, Cox regression and nonparametric models were applied to the generated pseudo-control arm and selpercatinib arm to estimate the treatment effect of selpercatinib. The hazard ratios estimated for selpercatinib were included in the NMA to estimate the relative treatment effects of selpercatinib and relevant comparators. In the second-line setting, only survival data (OS or PFS) were adjusted for the REVEL study’s docetaxel plus placebo arm; ORR data were not adjusted for RET fusion–positive status and other prognostic factors because sufficient response data for the docetaxel plus placebo arm were not available in the Flatiron database.

Bayesian NMAs were conducted using both fixed-effect and random-effect models. A hierarchical exchangeable model was used to take into account PD-L1 expression as a class in the model, and subgroup analyses of patients by level of PD-L1 expression were therefore not performed. As the fixed-effect hierarchical exchangeable model can accommodate the outcomes by PD-L1 status — and based on model-fit statistics — it was considered the primary model. Sensitivity analyses were not performed in this study.

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

When PSM by logistic regression was used, the PFS results were more clinically plausible for docetaxel. This model was therefore chosen to match data from the docetaxel plus placebo and selpercatinib arms.

Heterogeneity across included trials was examined by using standard pairwise meta-analysis or closed loops. A meta-regression method was also adopted to explore heterogeneity, including the covariates of ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

In both ITCs for treatment-naive and pre-treated populations, efficacy outcomes were presented as hazard ratios with corresponding 95% credible intervals (CrI) for OS and PFS, and as odds ratios with corresponding 95% CrIs for ORR.

Table 30 below presents a summary of the methods used for the ITCs.

Table 30: Indirect Treatment Comparison Analysis Methods