Reimbursement Review

Mirvetuximab Soravtansine (Elahere)

Sponsor: AbbVie Corporation

Therapeutic area: Epithelial ovarian, fallopian tube, or primary peritoneal cancer

This multi-part report includes:

Clinical Review

TPA Review

Pharmacoeconomic Review

Clinical Review

Abbreviations

Executive Summary

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

Table 1: Background Information on the Application Submitted for Review

ABIW = adjusted ideal body weight; FR = folate receptor; NOC = Notice of Compliance; q.3.w. = once every 3 weeks.

Introduction

Epithelial ovarian cancer (EOC) is the most common type of ovarian cancer (OC) and the deadliest gynecologic malignancy. EOC can arise from the epithelium covering the fimbria of the fallopian tubes, the ovaries, or the peritoneal cavity.¹ Approximately 85% to 95% of OCs are epithelial in origin.^2-5 In most patients with advanced OC, the OC will initially respond to platinum-based chemotherapy; however, up to 70% experience disease recurrence.^6-11 OC that responds to platinum-based chemotherapy but subsequently relapses 6 months or more after completion of platinum-based therapy is considered “platinum sensitive” whereas OC that relapses within 6 months of completing platinum-based therapy is considered “platinum resistant.”¹² Up to 70% of patients with recurrent disease eventually develop platinum-resistant OC.^6-11 Compared to other late-stage cancers, platinum-resistant OC (PROC) is difficult to treat; patients have severe symptoms that impair their quality of life (QoL) and have a very poor prognosis with limited treatment options. PROC is characterized by low expected objective response rates (ORRs) (i.e., 10% to 15%),^12,13 short progression-free survival (PFS) (i.e., 3 to 6 months), and limited survival (i.e., median overall survival [OS]: 9 to 12 months).^12,14 In addition to poor survival, patients have severe impairments in health-related quality of life (HRQoL), including frequent bowel obstruction, abdominal pain, residual neuropathy from prior therapies, and anxiety and fear, all of which have direct, severe impacts on their daily lives.

There is no cure for PROC. Therefore, the conventional treatment goals are to maximize or maintain HRQoL while improving OS by controlling disease or delaying further progression.^15,16 Current therapies for patients with PROC consist primarily of non–platinum-based chemotherapy (e.g., weekly paclitaxel, pegylated liposomal doxorubicin [PLD], or topotecan), administered either as a single drug or in combination with bevacizumab.^17-19 Single-drug, non–platinum-based chemotherapy treatments are associated with many toxicities in the clinical context of residual effects of prior chemotherapy, including sensory and motor neurotoxicity.^20-25 Chemotherapy is also associated with numerous significant, severe adverse events (AEs);^26-28 these are often experienced without any improvement in disease-related symptoms. Non–platinum-based chemotherapy is associated with low ORRs of 4% to 13%, median PFS of 3 to 4 months, and median OS of approximately 9 months to 12 months, as well as significant adverse effects that negatively affect HRQoL.^17,26,27,29 Although bevacizumab in combination with chemotherapy shows modest improvements in PFS (median PFS = 6.7 months) and ORR (30.9%), it can be associated with higher rates of gastrointestinal (GI) perforation, hypertension, thrombotic complications, and proteinuria in patients with OC.^30-35 Additional challenges in the treatment of PROC include the lack of meaningful, predictive biomarkers, which impedes targeted treatment. There is a significant unmet need for a novel therapeutic option for PROC that takes advantage of molecular targets, such as FR alpha, which have yet to be leveraged.

Mirvetuximab soravtansine (MIRV) is a first-in-class antibody-drug conjugate (ADC) targeting FR alpha, a protein that is commonly overexpressed on ovarian carcinomas and minimally expressed on normal tissues. MIRV comprises an FR alpha–binding antibody, a cleavable linker, and the maytansinoid DM4, a potent tubulin-targeting drug. Single-drug MIRV has shown anticancer activity and a safety profile that primarily includes low-grade GI, neurosensory, rare interstitial lung disease (ILD), and reversible ocular AEs. MIRV is approved by Health Canada as monotherapy for the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens. The sponsor is requesting reimbursement as per the Health Canada indication.

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of MIRV (5 mg/mL, concentrate for solution for injection, IV infusion) in the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens.

Perspectives of Patient, Clinicians, and Drug Programs

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

Patient Input

Input for this review was submitted by 1 patient group, Ovarian Cancer Canada (OCC). Information for this input was gathered through 10 patient or caregiver interviews conducted between December 16, 2024, and February 28, 2025, and through an online survey distributed to patients living in Canada conducted from February 12, 2025, to March 3, 2025. The survey received responses from 41 participants, 34 of whom were patients living with PROC and 7 of whom were caregivers. Most survey respondents had EOC (59%), were diagnosed between 2022 and 2024 (52%) at stage III or IV (95%), and had experienced a recurrence (91%). A total of 3 respondents (1 patient and 2 caregivers) had experience with MIRV.

Patients highlighted that OC led to severe impacts on their self-esteem, ability to participate in physical activity, sleep, and ability to care for themselves and their families. Patients noted that their OC had resulted in an inability to plan their future, particularly due to the known poor prognosis with recurrence or platinum resistance. Patients also noted a profound fear of leaving their families and children. Regarding prior treatment, 24% strongly disagreed and 43% neither agreed nor disagreed that prior treatments were able to manage their cancer. Survey respondents ranked prolonged survival (70%), improved QoL (65%), and lengthening time to recurrence (70%) as the most important outcomes for a new treatment. Eighty percent of patients indicated that they had considered taking MIRV. According to the patients, they are willing to tolerate hair loss (100%), fatigue (95%), aching joints (79%), neuropathy (70%), and eye problems (55%) if MIRV prolonged their life and improved daily functioning. Most respondents (71%) felt that the benefits of MIRV would outweigh the risks.

OCC interviewed 1 patient (aged 69 years) who had been diagnosed with stage III OC in 2021 and was treated with MIRV in the US from 2023 to 2024 as well as a caregiver whose mother was diagnosed in 2022 at age 77 years with PROC who also received treatment in the US from June to August 2024. Both patients had paid out of pocket for treatment; 1 patient required a reverse mortgage on their home to pay the approximate $108,000 in treatment costs, excluding travel, to receive only 3 treatments.

A total of 2 patients and 2 caregivers had experience with FR alpha testing to confirm eligibility for treatment with MIRV. The patient group highlighted that to access FR alpha testing in the US from Canada, patients need the time, funds, ability to travel, and an understanding of how to access tests in the US, often having to obtain this information through self-advocacy and self-education. Patients reported the FR alpha testing process is costly, lengthy, confusing, and stressful. They also reported no side effects from testing. FR alpha testing in the US was an additional out-of-pocket cost (US$4,500). There were delays associated with receiving test results and starting treatment.

Clinician Input

Input From Clinical Experts Consulted for This Review

Two clinical experts with experience in the diagnosis and medical management of patients with platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer were consulted by Canada’s Drug Agency (CDA-AMC) for the review of MIRV.

The clinical experts consulted for the review noted that most patients become resistant to platinum-based therapy during their disease course and that PROC is characterized by high recurrence rates. The clinical experts noted that the treatment goals for PROC are to control disease, delay progression, manage symptoms, and maintain HRQoL. Current standard non–platinum-based chemotherapies, including paclitaxel, PLD, topotecan, and gemcitabine, with or without bevacizumab, offer limited clinical benefit, with low response rates (approximately 10% to 15%) and short expected durations of response (less than 6 months). As a result, new therapies that can meaningfully improve response rates, prolong survival, and address the underlying disease process while maintaining HRQoL are urgently needed.

The clinical experts noted that, compared to currently available therapies, which primarily offer symptom control, MIRV selectively targets tumour cells with high FR alpha expression and delivers a cytotoxic payload, offering a more targeted, disease-specific treatment. According to the clinical experts, MIRV should be considered as a first-line therapy upon confirmation of platinum resistance in eligible patients, given its response rate and survival benefit over existing available therapies. Both clinical experts noted that MIRV should be used as a single drug in patients with recurrent progressive PROC with high FR alpha expression, assessed using VENTANA FOLR1 (FOLR1-2.1) RxDx companion diagnostic assay (herein referred to as VENTANA FOLR1 Assay). Patients with low or absent FR alpha expression, or those unable to tolerate ADC-related toxicities, are less suitable candidates for treatment. According to the clinical experts, the diagnostic infrastructure for FR alpha expression is not yet standardized across Canada.

The clinical experts indicated that response to treatment should be assessed through a combination of symptom improvement, clinical and radiologic assessments, and standard outcome measures, including PFS, OS, and HRQoL. Treatment should be discontinued upon evidence of disease progression or the development of significant unresolved AEs, particularly ILD. The 2 clinical experts emphasized that MIRV should be administered in specialized oncology settings with multidisciplinary support, including optometry, ophthalmology, and pulmonology specialists, to monitor and manage drug-specific toxicities and ensure appropriate patient counselling and follow-up.

Clinician Group Input

Input for this review was submitted by 3 clinician groups: the BC Cancer Gynecologic Oncology Provincial Tumour Group, the Ontario Health (Cancer Care Ontario) Gynecologic Cancer Drug Advisory Committee (OH [CCO]), and the Society of Gynecologic Oncology of Canada (GOC). Information was gathered through published literature and the clinical experiences of physicians who treat patients with platinum-resistant cancer.

The groups noted that although initial OC treatment is typically effective with high response rates, most will subsequently relapse and develop recurrent disease, which is associated with poor prognosis and survival of 6 months to 12 months. The clinician groups highlighted that patients diagnosed with PROC represent a group with a high unmet need. According to clinical experts consulted for the review, treatment goals consist of improving PFS, delaying disease progression, and improving QoL by reducing symptom severity. The clinician groups noted the main treatment options are bevacizumab in combination with non–platinum-based chemotherapy, both of which have shown minimal impact on OS historically, have response rates less than 30%, and are associated with significant toxicity; in addition, OS rapidly becomes refractory to subsequent lines of treatment, which is in agreement with the input from the clinical experts consulted for the review.

The clinician group expressed differing opinions on the place in therapy for MIRV. OH (CCO) indicated it would provide an additional treatment option for patients for whom existing platinum-resistant options have failed; GOC indicated that MIRV would fit into the current treatment paradigm as an additional treatment option for platinum-resistant, recurrent, high-grade serous OC; and BC Cancer noted it would be used as the first treatment option in the setting of platinum resistance with high FR alpha expression, replacing non–platinum-based chemotherapy with or without bevacizumab. GOC noted that MIRV would be used as monotherapy and not combined with other treatments.

Clinician groups agreed that the patients best suited for MIRV are those with platinum-resistant, high-grade serous EOC with a high FR alpha expression (≥ 75% of cells with ≥ 2+ staining intensity). In clinical practice, the clinician groups noted that PFS and OS are the standard measures for assessing treatment response. The clinician groups agreed that a clinically meaningful treatment response would be defined as radiographic disease control (i.e., tumour response or stabilization as shown on CT or MRI) with improvement in cancer-related symptom burden and tolerable toxicity. Treatment discontinuation should be considered in the event of disease progression, intolerable AEs, or patient choice. Patients should receive treatment in an outpatient setting under the supervision (or guidance, in remote areas) of an oncologist. The clinician groups noted that access to optometry or ophthalmology is necessary to ensure ocular toxicities are appropriately assessed and managed.

Drug Program Input

Input was obtained from the drug programs that participate in the CDA-AMC reimbursement review process. Please refer to Table 4 for further information. The following were identified as key factors that could potentially affect the implementation of MIRV:

• relevant comparators

• considerations for initiation of therapy

• considerations for continuation or renewal of therapy

• considerations for prescribing of therapy

• generalizability

• funding algorithm (oncology only)

• care provision issues

• system and economic issues.

Clinical Evidence

Systematic Review

Description of Studies

One randomized controlled trial (RCT), the MIRASOL trial, was included in the sponsor’s systematic review. From February 2020 to July 2022, 453 patients (including patients from 8 Canadian sites) were enrolled in the MIRASOL trial. Patients were randomized at a ratio of 1 to 1 to receive single-drug MIRV (3-week regimen) (N = 227) or investigator’s choice (IC) chemotherapy (n = 226) (i.e., paclitaxel, PLD, or topotecan; 3-week or 4-week regimens). Randomization was stratified by the number of previous lines of therapy (1, 2, or 3) and the chemotherapy drug. Female patients aged 18 years of age or older with platinum-resistant disease and a confirmed diagnosis of high-grade serous EOC, primary peritoneal cancer, or fallopian tube cancer were included. Key inclusion criteria were radiographic progression on or after the most recent line of therapy; positive FR alpha expression, as assessed by the VENTANA FOLR1 Assay, prior treatment with at least 1 but no more than 3 prior systemic lines of anticancer therapy; and eligibility for single-drug therapy as the next appropriate line of treatment. All patients randomized to MIRV received MIRV at 6 mg/kg adjusted ideal body weight (AIBW) administered through IV every 3 weeks. For patients randomized to IC chemotherapy, the choice of chemotherapy (paclitaxel, PLD, or topotecan) was made before randomization; body weight from baseline was used to calculate body surface area to determine the required dose. No dose modifications were anticipated unless the patient’s body weight changed by plus or minus 10% from baseline. Patients continued to receive MIRV or the matching IC chemotherapy until disease progression, unacceptable toxicity, withdrawal of consent, or death, or until the sponsor terminated the study (whichever came first).

The primary objective of the MIRASOL trial was to compare the PFS of patients randomized to MIRV treatment versus IC chemotherapy. Key secondary objectives were to compare ORR, OS, and HRQoL assessed using the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Ovarian Cancer Module (EORTC QLQ-OV28) for patients randomized to MIRV versus patients randomized to IC chemotherapy. Safety outcomes were AEs, serious adverse events (SAEs), AEs of special interest, and deaths. Notable harms included peripheral neuropathy, fatigue, keratopathy, blurred vision, and pneumonitis.

Demographic and baseline characteristics were similar between the MIRV and IC chemotherapy groups. The median ages were 64.0 years in the MIRV group and 62.0 years in the IC group. In the MIRV group, 80% of patients had EOC, 12% had fallopian tube cancer, 7% had primary peritoneal cancer, and 13% were BRCA-positive (87% had negative or unknown BRCA status); in the IC group, these proportions were 81%, 10%, 9%, and 16% BRCA-positive, respectively. High-grade serous histology that was poorly differentiated (██%) or unknown (██% versus ██%) was common in both groups. The most frequent stages at diagnosis were stage IIIC (53%) and stage IV (29%), with a median time since diagnosis of █████ months.

The primary data cut-off (DCO) date was March 6, 2023, and the latest DCO date was September 26, 2024.

Efficacy Results

The efficacy outcomes in the MIRASOL trial were PFS, OS, ORR, duration of response (DOR), and EORTC QLQ-OV28. Results are reported at the latest DCO unless otherwise stated.

Progression-Free Survival

The primary end point was PFS per investigator assessment. PFS was assessed based on radiological imaging and determined using Response Evaluation Criteria in Solid Tumours Version 1.1 (RECIST 1.1). CT or MRI scans were collected and held for sensitivity analysis by the BICR and were tested using a stratified log-rank test. The PFS curve was estimated using the Kaplan-Meier estimator. The DCO date for the primary PFS analysis was March 6, 2023; by that time, 330 events had occurred. The median PFS was 5.62 months (95% CI, 4.34 months to 5.95 months) in the MIRV treatment group compared to 3.98 months (95% CI, 2.86 months to 4.47 months) in the IC chemotherapy group. The hazard ratio (HR) for PFS was 0.65 (95% CI, 0.521 to 0.808; P < 0.0001). The Kaplan-Meier estimated difference in PFS probabilities between treatment groups at 6 months was ████ (95% CI, ████ ██ ████) in favour of MIRV. At 12 months, the estimated difference in PFS probability was ████ (95% CI, ████ ██ ████) in favour of MIRV. The results of the subgroup analyses of PFS by blinded independent central review (BICR) and sensitivity analysis were consistent with those of the overall population analyses and the investigator assessment, respectively. The PFS results at the latest DCO date were consistent with those at the primary DCO date.

Overall Survival

OS was a key secondary outcome, and its curve was estimated using the Kaplan-Meier estimator. The comparison of OS between treatment groups was performed using Cox proportional hazard regression and log-rank test. The DCO date for OS analysis was September 26, 2024; by that time, 300 events had occurred. The median OS was 16.85 months (95% CI, 14.36 months to 19.78 months) in patients randomized to MIRV and 13.34 months (95% CI, 11.37 months to 15.15 months) in the IC chemotherapy group. The HR for OS was 0.68 (95% CI, 0.543 to 0.840; P = 0.0004). The Kaplan-Meier estimated difference in OS probabilities between groups at 6 months was ████ (95% CI, █████ ██ ████) in favour of MIRV. At 18 months, the estimated difference in OS probability was ████ (95% CI, ████ ██ ████) in favour of MIRV. The results of the subgroup analysis showed a trend for benefit of MIRV over IC chemotherapy in OS at both DCO dates and was consistent with the results of the primary analyses.

Objective Response Ratio

ORR refers to the proportion of patients who achieved a pre-established reduction in tumour volume as a complete response (CR) or partial response (PR). ORR was a key secondary outcome and was assessed using RECIST Version 1.1. At the latest DCO date, a higher percentage of patients treated with MIRV achieved an ORR compared to patients treated with IC chemotherapy (████% versus ████%), with a between-group difference of ████% (95% CI, ████ ██ ████) in favour of MIRV. The results of the subgroup analyses were consistent with those of the overall primary analysis at both DCO dates. ORR per BICR was evaluated as a sensitivity analysis, and the results were consistent with those of the investigator assessment.

Duration of Response

DOR was a secondary outcome. It was defined only for patients who had a confirmed best overall response (BOR) of CR or PR. The comparison of DOR between treatment groups was conducted using Cox proportional hazard regression and log-rank test. At the latest DCO date, among the ██ patients with events in the MIRV group, the median DOR was ██ months (95% CI, ████ ██ ████) compared to ███ months (95% CI, ████ ██ ████) in the ██ patients with events in the IC chemotherapy group, with a between-group difference of ███ months (95% CI, ████ ██ ████) in favour of MIRV. In a sensitivity analysis, the median DOR by BICR was ████ months (95% CI, ████ ██ ████ months) in the MIRV group compared to ████ (95% CI, ████ ██ █████) in the IC chemotherapy group. No subgroup analyses were performed.

Health-Related Quality of Life

HRQoL was assessed using the EORTC QLQ-OV28. The EORTC QLQ-OV28 Abdominal/GI Symptoms subscale at week 8 or 9 was a key secondary end point. The EORTC QLQ-OV28^36,37 consists of 28 items that assess QoL in patients with OC. Items are scored from 1 (not at all) to 4 (very much). Scores are derived by linearly transforming scale and/or standalone item raw scores to a 0-to-100 scale. Higher scores indicate worse symptoms. At the primary DCO date, 21% of patients in the MIRV group compared to 15% in the IC chemotherapy group achieved a 15-point symptom subscale decrease on the QLQ-OV28 Abdominal/GI Symptoms subscale at week 8 or 9, with a between-group difference of ████ (95% CI, █████ ██ █████) in favour of MIRV. In the change-from-baseline analysis, the difference in change from baseline in EORTC QLQ-OV28 Abdominal/GI Symptoms subscale at week 8 or 9 was –5.0 in favour of MIRV. These results were consistent with those at the primary DCO date.

Harms Results

A total of 210 patients (96%) treated with MIRV reported treatment-emergent adverse events (TEAEs) of any grade compared to 94% of patients in the IC chemotherapy group. Patients treated with MIRV reported fewer SAEs than those treated with IC chemotherapy (24% versus 33%). Compared to 16% of patients in the IC chemotherapy group, 9% of patients treated with MIRV reported TEAEs leading to discontinuation. Among patients treated with MIRV, ██ (████%) patients died compared to ███ (████%) patients in the IC chemotherapy. In both treatment groups, the primary cause of death was disease progression: ███% in the MIRV group versus ███% in the IC chemotherapy group.

Notable harms included peripheral neuropathy, fatigue, keratopathy, blurred vision, and pneumonitis. The proportions of patients with grade 3 peripheral neuropathy were similar in both treatment groups. Five patients in the MIRV group reported grade 3 fatigue compared to 11 patients in the IC chemotherapy group. Compared to IC chemotherapy, a greater proportion of patients treated with MIRV reported ocular TEAEs, including keratopathy (32% versus 0%) and blurred vision (41% versus 2%). The percentage of patients with pneumonitis was higher in the MIRV group than in the IC chemotherapy group (██% versus ██%).

Critical Appraisal

Internal Validity

The MIRASOL trial used stratified randomization of treatment based on the prognostic factors: number of previous lines of therapy (1, 2, or 3) and IC choice of chemotherapy (paclitaxel, PLD, or topotecan). However, details on allocation concealment were not provided by the sponsor. Baseline characteristics were balanced between treatment groups, and all analyses were conducted in the intention-to-treat (ITT) population. Missing data were low for efficacy outcomes, including PFS, OS, and ORR, while ██% (n = ██) and ██% (n = ██) of patients randomized to MIRV and IC chemotherapy had missing data on HRQoL assessed using the EORTC QLQ-OV28. The imputation methods for missing data were not fully described by the sponsor. The open-label design of the trial introduces the possibility of reporting, performance, and detection biases in the efficacy estimation of subjective harms and benefits. Among tumour efficacy outcomes (PFS, ORR), bias was minimized through the use of BICR, with high concordance between the investigator and BICR-assessed end points; OS is considered less susceptible to bias due to its objective nature. The trial reported that ██% of patients randomized to the MIRV group and ██% of patients randomized to IC chemotherapy received subsequent therapies. Notably, subsequent anticancer therapies were not formally balanced by protocol and were administered at investigator discretion. There was no formal sensitivity analysis adjusting OS estimates for postprogression therapies in the clinical study reports submitted by the sponsor. CDA-AMC notes that subsequent anticancer therapies could present a risk of bias for OS estimates if differential access or effectiveness of subsequent therapies existed between treatment groups. For the primary end point of PFS, the widths of the 95% confidence intervals (CIs) were not adjusted for multiplicity. For the key secondary end points of ORR, OS, and the EORTC QLQ-OV28 Abdominal/GI Symptoms scale, a hierarchical testing procedure was applied to control the family-wise type I error only if the null hypothesis for the primary end point was rejected at a 2-sided alpha level of 0.05.

External Validity

The population enrolled in the MIRASOL trial consisted of patients with platinum-resistant EOC who showed high FR alpha expression, defined as greater than or equal to 75% of tumour cells exhibiting greater than or equal to 2+ staining intensity, as assessed by the VENTANA FOLR1 Assay. According to the clinical experts consulted for the review, only 35% to 40% of patients with platinum-resistant EOC would meet this biomarker threshold, suggesting that MIRV is specific for this subgroup of patients with EOC. In addition, only patients with an Eastern Cooperative Oncology Group Performance Status (ECOG PS) of 0 or 1 were eligible, whereas in practice, a considerable proportion of patients with platinum-resistant disease have an ECOG PS of 2 or higher, particularly after multiple lines of therapy. The exclusion of patients with primary refractory disease, active ocular conditions, significant neuropathy, or serious comorbidities from the trial limits the generalizability of the efficacy results to patients with worse prognostic status who may still be considered eligible for treatment.

The clinical experts consulted for the review indicated that the comparator arm of IC chemotherapy, consisting of weekly paclitaxel, PLD, or topotecan, reflected the standard of care treatment options for platinum-resistant EOC in Canada and that the dosage schedules were consistent with Health Canada–approved labelling. Inputs from the clinical experts and patient and clinician groups suggested that the trial measured meaningful outcomes that are relevant to patients, including PFS, OS, and HRQoL as assessed through the EORTC QLQ-OV28. However, a key limitation was the incomplete measurement of HRQoL: the primary HRQoL analysis was assessed at week 8 or 9 only, with limited longer-term data reported. Given that HRQoL was identified as a key outcome by both patients and clinicians, the lack of long-term assessment of these data is a notable limitation.

GRADE Summary of Findings and Certainty of the Evidence

For the pivotal studies and RCTs identified in the sponsor’s systematic review, Grading of Recommendations, Assessment, Development and Evaluations (GRADE) was used to assess the certainty of the evidence for the outcomes considered most relevant to inform expert committee deliberations, and a final certainty rating was determined as outlined by the GRADE Working Group.^38,39

Following the GRADE approach, evidence from the RCTs started as high-certainty evidence and could be rated down for concerns related to study limitations (which refers to internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, or publication bias.

When possible, certainty was rated in the context of the presence of an important (nontrivial) treatment effect; if this was not possible, certainty was rated in the context of the presence of any treatment effect (i.e., the clinical importance is unclear). In all cases, the target of the certainty of evidence assessment was based on the point estimate and its location relative to the threshold for a clinically important effect (when a threshold was available) or to the null.

The reference points for the certainty of evidence assessment for PFS, OS, ORR, HRQoL assessed using EORTC QLQ-OV28, and notable harms (including peripheral neuropathy, fatigue, keratopathy, blurred vision, and pneumonitis) were set according to the presence or absence of an important effect based on thresholds informed by the clinical experts consulted for this review. The reference point for the certainty of the evidence assessment for the EORTC QLQ-OV28 Abdominal/GI Symptoms subscale was set according to the 15-point symptom subscale decrease threshold that was informed by the literature.

Results of GRADE Assessments

Table 2 presents the GRADE summary of findings for MIRV from the MIRASOL trial for the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens.

The selection of outcomes for the GRADE assessment was based on the sponsor’s Summary of Clinical Evidence, consultation with clinical experts, and input received from patient and clinician groups and public drug plans. The following list of outcomes was finalized in consultation with expert committee members:

• PFS

• OS

• ORR

• HRQoL measured using the EORTC QLQ-OV28

• notable harms (peripheral neuropathy, fatigue, keratopathy, blurred vision, and pneumonitis).

Table 2: Summary of Findings for MIRV vs. IC Chemotherapy for Patients With FR Alpha–Positive, Platinum-Resistant Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer (DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; DOR = duration of response; IC = investigator’s choice; EORTC QLQ-OV28 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Ovarian Cancer Module; FR = folate receptor; GI = gastrointestinal; LS = least squares; MID = minimal important difference; MIRV = mirvetuximab soravtansine; NR = not reported; OS = overall survival; ORR = overall response rate; PFS = progression-free survival; RCT = randomized controlled trial.

^aStudy limitation (which refers to internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, and publication bias were considered when assessing the certainty of the evidence. All serious concerns in these domains that led to the rating down of the level of certainty are documented in the table footnotes.

^bRated down 1 level for imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 5% (i.e., ≥ 50 fewer events per 1,000 patients) at 6 months and 12 months was considered clinically significant by the clinical experts. The 95% CI included both a meaningful benefit and the possibility of no effect.

^cIn the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 5% (i.e., ≥ 50 fewer events per 1,000 patients) at 6 months and 12 months was considered clinically significant by the clinical experts. The point estimate and entire CI exceeded the threshold.

^dRated down 2 levels for serious imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 5% (i.e., ≥ 50 fewer events per 1,000 patients) at 6 months was considered clinically significant by the clinical experts. The 95% CI included both a meaningful benefit and the possibility of no effect.

^eRated down 1 level for imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 5% (i.e., ≥ 50 fewer events per 1,000 patients) at 18 months was considered clinically significant by the clinical experts

^fIn the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 10% (i.e., ≥ 100 fewer events per 1,000 patients) was considered clinically significant by the clinical experts. The point estimate and entire CI exceeded the threshold.

^gRated down 2 levels for serious imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 10% (i.e., ≥ 100 fewer events per 1,000 patients) was considered clinically significant by the clinical experts. The 95% CI included both a meaningful benefit and the possibility of no effect.

^hRated down 2 levels for serious imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 10% (≥ 100 fewer events per 1,000 patients) was considered clinically significant by the clinical experts. The point estimate did not exceed this threshold, and the CI included the possibility of no effect, introducing uncertainty.

ⁱRated down 2 levels for serious imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 10% (i.e., ≥ 100 fewer events per 1,000 patients) was considered clinically significant by the clinical experts. The point estimate was lower than this threshold, and the CI included both a meaningful increase and no effect, introducing uncertainty.

^jIn the absence of a validated MID, the threshold was informed by the clinical experts consulted for this review. The point estimate and CI exceeded the 10% threshold (i.e., 100 per 1,000 patients) for clinical significance.

^kRated down 1 level for imprecision. In the absence of a validated MID, the threshold was informed by clinical experts consulted for this review. A between-group absolute risk difference of 10% (i.e., ≥ 100 fewer events per 1,000 patients) was considered clinically significant by the clinical experts. Although the point estimate and upper CI exceeded this threshold, the lower CI was slightly lower than this threshold, introducing uncertainty.

Source: MIRASOL trial Clinical Study Report.⁴⁰ Details included in the table are from the sponsor’s Summary of Clinical Evidence.

Long-Term Extension Studies

There were no relevant long-term extension studies submitted for this review.

Indirect Comparisons

In the absence of direct head-to-head trials evaluating the efficacy of MIRV compared to bevacizumab and chemotherapy in adult patients with PROC who have received 1 to 3 prior systemic treatment regimens, the sponsor conducted and submitted an indirect treatment comparison (ITC). The objective of this section is to summarize and critically appraise the sponsor-submitted ITC and to inform the pharmacoeconomic model.

Description of Studies

The sponsor conducted a matching-adjusted indirect comparison (MAIC) using individual patient data (IPD) from the MIRASOL study and published aggregate data from the AURELIA study (bevacizumab plus IC chemotherapy). The MAIC adjusted for the following baseline characteristics: IC chemotherapy, age, number of prior lines of therapy, platinum-free interval, ECOG PS, and cancer antigen 125 (CA-125) levels. It was not feasible to adjust additional baseline characteristics due to lack of testing or reporting in the AURELIA trial, more restrictive inclusion criteria in the MIRASOL trial, and impact on sample size.

Efficacy Results

The risk of progression or death in the MIRV arm was comparable to that in the bevacizumab and chemotherapy arm, both before matching (HR = ████ [95% CI, ████ ██████]) and after matching (HR = ████ [95% CI, ████ ██ ████]). Risk of death in the MIRV arm was lower than in the bevacizumab and chemotherapy arm, with an HR of ████ (95% CI, ████ ██ ████) before matching and an HR of ████ (95% CI, ████ ██ ████) after matching for MIRV versus bevacizumab and chemotherapy. For ORR, the OR of MIRV versus bevacizumab and chemotherapy was ████ (95% CI, ████ ██ ████) before matching; it increased to ████ (95% CI, ████ ██ ████) after matching, indicating numerically higher odds of overall response in the MIRV arm than in the bevacizumab and chemotherapy arm.

Harms Results

After matching, the odds ratio (OR) of grade 3 or higher TEAEs in the MIRV group versus the bevacizumab and chemotherapy group was ████ (95% CI, ████ ██ ████), indicating numerically lower odds of grade 3 or higher TEAEs in the MIRV arm than in the bevacizumab and chemotherapy arm. The ORs of discontinuation due to any TEAE in the MIRV arm versus the bevacizumab and chemotherapy arm were ████ (95% CI, ████ ██ ████) before matching and ████ (95% CI, ████ ██ ████) after matching, indicating lower odds of discontinuation due to any TEAE in the MIRV arm than in the bevacizumab and chemotherapy arm.

Critical Appraisal

The sponsor-submitted ITC was informed by a systematic literature review. The 2 included studies, the MIRASOL trial and the AURELIA trial, were phase III, global, multicentre, open-label RCTs among adult patients with epithelial ovarian, fallopian tube, or primary peritoneal cancer, with a focus on adult patients with platinum-resistant disease and similar median follow-up durations. The trials were conducted 10 years apart. The sponsor conducted a methodologically sound MAIC and reported on key aspects, including patient population matching, outcome definitions, statistical methods, and assessment of heterogeneity. Some key patient characteristics could feasibly be adjusted in the MAIC, including the number of prior lines of therapy, platinum-free intervals, choice of preassigned chemotherapy, and ECOG PS. Other key characteristics identified by the clinical experts consulted for this review — including tumour histology, FR alpha expression status, prior poly (ADP-ribose) polymerase (PARP) inhibitor use, and prior bevacizumab use — could not be adjusted due to lack of testing or reporting in the AURELIA trial, more restrictive inclusion criteria in the MIRASOL trial, or impact on sample size. The sponsor and the clinical experts have noted that the treatment landscape of OC evolved in the time between the trials; as a result, in the AURELIA trial, all patients had no prior exposure to PARP inhibitors and had prior antiangiogenic therapy at baseline. The clinical experts noted that the serous histology subtype, which was a required subtype for inclusion in the MIRASOL trial but not in the AURELIA trial, generally has a better prognosis than other subtypes. The sponsor and clinical expert noted that FR alpha expression has been shown to be an effect modifier for both MIRV and pooled chemotherapy; FR alpha was not reported in the AURELIA trial. As a result of this heterogeneity between patient populations in the trials, the clinical experts concluded that no meaningful cross-trial comparison could be made. In addition, due to poor overlap in patients’ baseline characteristics between the trials, the small effective sample size (ESS) after matching (██ for the MIRV arm and ██ for the chemotherapy arm in the MIRASOL trial) limited the power to detect statistically significant differences in treatment effects and constrained the ability to adjust for additional baseline characteristics.

ITC results were presented for OS, PFS, ORR, and harms outcomes; other outcomes of relevance to patients (e.g., HRQoL) were not reported.

Studies Addressing Gaps in the Evidence From the Systematic Review

No relevant studies addressing gaps in the evidence from the systematic review were submitted by the sponsor.

Conclusions

The MIRASOL trial demonstrated with moderate to high certainty that MIRV offers clinically meaningful improvement in PFS compared with IC chemotherapy in patients with platinum-resistant EOC whose tumours express high FR alpha, assessed according to the VENTANA FOLR1 Assay (i.e., ≥ 75% of viable tumour cells with ≥ 2+ staining intensity). MIRV likely led to an increase in OS compared to IC chemotherapy at 18 months. However, while MIRV demonstrated a higher ORR compared to IC chemotherapy, it resulted in little to no clinically important difference in the proportion of patients who reported a clinically meaningful improvement on the EORTC QLQ-OV28 Abdominal/GI Symptoms subscale compared to IC chemotherapy. MIRV was characterized by low-grade, reversible ocular, GI, and neurosensory AEs, including dry eye, abdominal pain, diarrhea, nausea, and constipation. We determined with low to moderate certainty that notable harms, including peripheral neuropathy, fatigue, keratopathy, and pneumonitis, showed little to no difference in patients treated with MIRV compared to IC chemotherapy. On the other hand, there was high-certainty evidence that treatment with MIRV resulted in blurred vision compared to IC chemotherapy.

The sponsor-submitted ITC provides mixed or comparable treatment effects of MIRV and of bevacizumab combined with chemotherapy on PFS and OS. Evidence from the ITC favours MIRV in the safety outcomes. As a result of the heterogeneity between patient populations in both trials and the evolution of the treatment landscape of OC in the time between the trials, the effects estimates could be biased and underestimate treatment differences.

Introduction

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of MIRV (5 mg/mL, concentrate for solution for injection, IV infusion) in the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens.

Disease Background

The contents within this section have been informed by materials submitted by the sponsor and by clinical expert input. The following has been summarized and validated by the review team.

OC is a highly heterogenous disease and the third most frequently diagnosed gynecological cancer globally.⁴¹ The age-standardized incidence rate of OC in Canada (excluding Québec) was 14.3 per 100,000 in 2019.⁴² In 2022, the 5-year prevalence of OC was estimated to be 54.6 per 100,000.⁴³ Estimates from 2024 suggests there were 3,000 new cases and 2,000 deaths that year due to OC, making it the leading cause of death from gynecological cancers.⁴⁴ EOC comprises the vast majority of all OCs and is the deadliest gynecologic malignant neoplasm. EOC arises from the epithelium covering the fimbria of the fallopian tubes, the ovaries, or the peritoneal cavity.¹ The main subtypes of epithelial ovarian tumours are high-grade serous, low-grade serous, endometrioid, clear cell, and mucinous tumours.^12,45,46 About 85% to 95% of OCs are epithelial in origin, and the majority (75% to 92%) are high-grade serous OCs.^2-5 The overall lifetime risk of developing EOC is 1.3%; this risk is elevated substantially (to 40% to 45%) for patients who carry a BRCA1 mutation. It is elevated to 15% to 20% for BRCA2 mutation carriers.^47,48

Most people with advanced OC will initially respond to platinum-based chemotherapy; however, up to 70% will experience disease recurrence, with subsequent treatment determined by the DOR following platinum therapy.^6-11 Although there is no universally accepted definition of PROC, OC that responds to platinum-based chemotherapy and subsequently relapses 6 months or more after completion is classified as platinum sensitive, while OC that relapses within 6 months is classified as platinum resistant.¹² Platinum-sensitive ovarian cancer (PSOC) will respond to further platinum-based chemotherapy, with response rates ranging from 30% to 90%.¹² However, up to 70% with recurrence will eventually develop platinum resistance.^6-11 Compared to other late-stage cancers, PROC is difficult to treat, and patients face a very poor prognosis with limited treatment options. PROC is associated with worse outcomes than PSOC, characterized by low expected ORRs (10% to 15%),^12,13 short PFS (3 to 6 months), and limited survival (median OS = 9 to 12 months for PROC versus 24 to 36 months for PSOC).^12,14 In addition to poor survival, patients have severe impairments in HRQoL, including frequent bowel obstruction, abdominal pain, residual neuropathy from prior therapies, and anxiety and fear, all of which have direct, severe impacts on their daily activities.

The diagnosis of OC uses a layered approach. Depending on the absence or presence of symptoms, staging is performed following clinical workup on the tumour and various assessments (e.g., ultrasound, CT, and/or MRI). Genetic testing for OC susceptibility genes is also recommended.⁴⁹ Genes that are tested include, but are not limited to, BRCA1 and BRCA2 and genes from the mismatch repair family. An additional challenge in the treatment of PROC is the lack of meaningful, predictive biomarkers, which can impede the continuation of treatment. FR alpha is a novel biomarker of therapeutic relevance. It is a 3,840 kDA glycosylphosphatidylinositol-anchored cell surface protein encoded by the FOLR1 gene.⁶ This protein is constitutively overexpressed in nearly 40% of OC solid tumours⁵⁰ and minimally expressed in nonmalignant tissues.^51-53 This allows targeted delivery of therapy to the malignant tissue while minimizing collateral toxic side effects.^53-57 Folate receptors have been shown to be expressed at elevated levels in more than 70% of primary and 80% of recurrent EOCs.^53,57 The VENTANA FOLR1 Assay is IHC assay used to identify patients with EOC who are eligible for FOLR1-targeted treatment. In EOC tissue, neoplastic cells labelled with the VENTANA FOLR1 Assay are evaluated for the percentage of tumour cell staining using immunohistochemistry.

Standards of Therapy

The contents within this section have been informed by materials submitted by the sponsor and by clinical expert input. The following has been summarized and validated by the review team.

The treatment goals for OC vary depending on the stage at which it is diagnosed and response to platinum-based chemotherapy. When diagnosed early (i.e., early-stage OC), the treatment goal is to eradicate disease through surgery and chemotherapy while adequately preserving HRQoL.^15,19 As previously noted, platinum resistance is a major factor contributing to mortality in patients, especially because PROC is more difficult to treat than other late-stage cancers. PROC is associated with worse outcomes. These are characterized by low ORRs (10% to 15%), short PFS (3 months to 6 months),^12,13 shorter survival compared with PSOC (median OS = 9 to 12 months for PROC versus 24 to 36 months for PSOC),^12,14 and significant impacts on patients’ QoL. There is no cure for PROC; therefore, the conventional treatment goals are to maximize or maintain HRQoL and to improve OS by controlling disease or delaying progression.^15,16

In the Canadian practice landscape, treatment options for PROC include single-drug, non–platinum-based chemotherapy (i.e., paclitaxel, PLD, topotecan, gemcitabine) with or without bevacizumab.^17-19 These treatments for PROC are associated with multiple limitations. The single-drug, non–platinum-based chemotherapy treatments are associated with many toxicities in the clinical context of residual effects of prior chemotherapy, including sensory and motor neurotoxicity following treatment.^20-25 Chemotherapy is also associated with many significant, severe AEs (grade ≥ 3 AEs are reported in 54% to 65.2% of patients treated with chemotherapy);^26-28 these are often experienced without any improvement in disease-related symptoms. Furthermore, non–platinum-based chemotherapy is associated with low ORRs (4% to 13%), median PFS (3 to 4 months), median OS of 9 months to 12 months, and significant adverse effects that negatively affect patients’ QoL.^17,26,27,29 The efficacy and safety of bevacizumab is limited to patients in the early-line treatment setting (i.e., 2 or fewer prior anticancer regimens); although bevacizumab in combination with chemotherapy shows modest improvements in PFS (median PFS = 6.7 months) and ORR (30.9%), it can be associated with higher rates of GI perforation, hypertension, thrombotic complications, and proteinuria in patients with OC.^30-34 There is a high unmet medical need in patients with no standard treatment option after their third-line treatment. The clinical experts consulted for this review indicated that the choice of non–platinum-based chemotherapy drug is based on various patient factors, preexisting residual chemotherapy side effects, and the anticipated toxicity profile for each chosen chemotherapeutic drug. The expected response rate in this clinical setting is 10% to 15%, with a DOR around 6 months; expected survival after diagnosis of platinum-resistant disease is 9 months to 12 months. Ongoing symptom management with disease progression is frequently needed. The limitations of current therapies suggest that there is a significant, urgent need for a novel therapeutic option to treat PROC that leverages biomarkers, such as FR alpha. Given the poor efficacy in terms of survival and response rates — as well as the notable impacts on QoL and the safety concerns associated with current treatments — effective and tolerable treatments that improve survival and QoL and reduce the symptom burden and toxicities associated with current treatments are needed.

Drug Under Review

The key characteristics of MIRV are summarized in Table 3 along with those of other treatments available for adult patients with platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer.

MIRV is indicated as monotherapy for the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens. MIRV is provided as concentrate for solution for IV infusion, 5 mg/mL per vial. The recommended dosage is 6 mg/kg AIBW administered once every 3 weeks (i.e., in 21-day cycles). MIRV is an ADC. The antibody is a chimeric immunoglobulin G1 directed against FR alpha. The small molecule, DM4, is a microtubule inhibitor attached to the antibody through a cleavable linker. Upon binding to FR alpha, MIRV is internalized; this is followed by the intracellular release of DM4 through proteolytic cleavage. DM4 disrupts the microtubule network within the cell, resulting in cell cycle arrest and apoptotic cell death. MIRV is approved by Health Canada as monotherapy for the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens. The sponsor is requesting reimbursement as per the Health Canada indication. The approved MIRV indication in the US aligns with the approved Health Canada indication.

Table 3: Key Characteristics of MIRV and Comparators

AIBW = adjusted ideal body weight; CHF = congestive heart failure; CV = cardiovascular; FR = folate receptor; GI = gastrointestinal; MIRV = mirvetuximab soravtansine; KDR = kill-to-death ratio ; PLD = pegylated liposomal doxorubicin; PRES = posterior reversible encephalopathy syndrome; q.2.w. = every 2 weeks; q.3.w. = every 3 weeks; TE = tracheoesophageal; VEGF = vascular endothelial growth factor.

Sources: MIRV Product Monograph;⁵⁹ Paclitaxel Drug Monograph;⁶⁰ PLD Drug Monograph;⁶¹ Topotecan Regimen Monograph;⁶² Bevacizumab Drug Monograph.⁶³

Perspectives of Patients, Clinicians, and Drug Programs

The full patient and clinician group submissions received are available in the consolidated patient and clinician group input document for this review on the project website.

Patient Group Input

This section was prepared by the review team based on the input provided by patient groups.

Input for this review was submitted by 1 patient group, OCC. OCC is a national charity that advocates for people affected by OC through research, education, and support. Information for this input was gathered through 10 patient or caregiver interviews from December 16, 2024, to February 28, 2025, and through an online Canadian survey distributed to patients in Canada available from February 12, 2025, to March 3, 2025. The survey received responses from 41 participants, 34 of whom were patients living with PROC and 7 of whom were caregivers. Survey and interview participants were from Alberta, British Columbia, Manitoba, Ontario, Québec, Saskatchewan, and California. Most had EOC (59%), had been diagnosed between 2022 and 2024 (52%) at stage III or IV (95%), and had experienced a recurrence (91%). A total of 3 interviewees (1 patient and 2 caregivers) had experience with MIRV.

Patients highlighted that OC had a severe impact on their self-esteem, sleep, ability to participate in physical activity, and ability to care for themselves and their families. Patients noted that their OC had resulted in an inability to plan for their future, particularly due to the known poor prognosis with recurrence or platinum resistance. Patients also noted a profound fear of leaving their families and children. Prior reported treatments included surgery (95%), chemotherapy (91%), PARP inhibitors (41%), bevacizumab (36%), and radiation (5%); however, 24% of those who responded strongly disagreed (and 43% neither agreed nor disagreed) that these treatments were able to manage their cancer. Fatigue, hair loss, bowel problems, brain fog, shortness of breath, neuropathy, and blood-related conditions (e.g., anemia) were among the reported side effects of treatment, with the greatest impact being poor QoL. Patients also reported significant barriers to accessing treatment, particularly due to travel burden and treatment cost. While most patients initially respond well to platinum-based chemotherapy, the patient group noted that many develop resistance and are left with no effective treatment options.

Survey respondents ranked prolonged survival (70%), improved QoL (65%), and lengthening time to recurrence (70%) as the most important outcomes for a new treatment. Most had considered MIRV (80%) and indicated that an associated moderate improvement (63%) in OC (as opposed to high improvement) would be sufficient for them to consider using it. According to patients, they would be willing to tolerate hair loss (100%), fatigue (95%), aching joints (79%), neuropathy (70%), and eye problems (55%) if MIRV prolonged their lives and improved their daily functioning. As a result, most respondents (71%) felt that the benefits of MIRV would outweigh the risks.

OCC interviewed 1 patient (aged 69 years) who had been diagnosed with stage III OC in 2021 and treated with MIRV in the US from 2023 to 2024. OCC also interviewed 1 caregiver whose mother was diagnosed in 2022 at age 77 years with PROC and received treatment in the US from June 2024 to August 2024. Both patients paid out of pocket for treatment, with 1 patient requiring a reverse mortgage on their home to pay the approximate $108,000 in treatment costs, excluding travel, to receive 3 only treatments. Both patients reported a decrease in CA-125 markers, with 1 patient experiencing a decrease to 1,800 from 3,700 after 1 treatment. One patient experienced additional improvements in energy and mobility, with minimal side effects except a slow onset of eye; this initially resolved with steroids, but progressed over time, and the patient will require cataract surgery in both eyes. The other patient reported a “somewhat negative” impact on QoL due to the stress of having to travel for treatment and secure financing (resulting in a 3- to 4-month treatment gap); this patient also reported fatigue and appetite issues. They were eventually required to stop treatment due to being too weak to travel to California.

A total of 2 patients and 2 caregivers had experience with FR alpha testing in the US. The patient group highlighted that to access FR alpha testing in the US from Canada, patients need time, funds, the ability to travel, and an understanding of how to access the tests; often, they must obtain this information through self-advocacy and self-education. Patients reported the FR alpha testing process was costly, lengthy, confusing, and stressful. However, they reported no side effects from testing. FR alpha testing in the US was an additional out-of-pocket cost (US$4,500), and there were delays associated with receiving results and starting treatment.

Clinician Input

Input From Clinical Experts Consulted for This Review

All CDA-AMC review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, 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 epithelial ovarian, fallopian tube, or primary peritoneal cancer.

Unmet Needs

The clinical experts consulted for this review highlighted that platinum-resistant epithelial ovarian, fallopian tube, and primary peritoneal cancers are characterized by high recurrence rates; most patients become resistant to platinum-based therapy during their disease trajectory. The clinical experts noted that the treatment goals for PROC are to control disease progression, manage symptoms, and maintain HRQoL. According to the clinical experts, current treatments, including non–platinum-based chemotherapies (such as paclitaxel, PLD, topotecan, and gemcitabine with or without bevacizumab) offer limited clinical benefit and duration. The response rates with these therapies are low (approximately 10% to 15%), and the DOR is about 6 months. The clinical experts emphasized that currently available therapies largely focus on symptom control and do not meaningfully alter the disease trajectory or significantly improve survival outcomes. Many patients become refractory to these options or experience cumulative additional toxicities that limit their ability to tolerate further treatment. As a result, new therapies that can meaningfully improve response rates, prolong survival, and address the underlying disease process while maintaining HRQoL are urgently needed.

Place in Therapy

The clinical experts noted that compared to current therapies, which primarily offer symptom control, MIRV targets tumour cells with FR alpha expression and delivers a cytotoxic payload, offering a more targeted, disease-modifying treatment. According to the clinical experts, MIRV should be considered as a first-line therapy upon confirmation of platinum resistance in eligible patients, given its response rate and survival benefit over existing therapies in PROC. The clinical experts noted that MIRV should be used as a single drug in the management of platinum-resistant disease. According to the clinical experts, MIRV will lead to a shift in the treatment paradigm for PROC, with an estimated 30% of patients potentially eligible based on high FR alpha expression. The clinical experts noted that MIRV should be prioritized before other nontargeted chemotherapy options to maximize clinical benefit and avoid cumulative treatment-limiting toxicities from other regimens, like weekly paclitaxel.

Patient Population

The clinical experts consulted for the review indicated that patients with recurrent, progressive PROC with high FR alpha expression are suitable for MIRV treatment. PROC can be identified through a combination of clinical and imaging assessments. FR alpha expression level can be assessed using the VENTANA FOLR1 Assay as a companion diagnostic test. The clinical experts indicated that MIRV provides significant response in patients with high FR alpha expression (PS2+ scoring method; i.e., ≥ 75% of viable tumour cells with moderate [2+] or strong [3+] staining intensity). Patients with low or absent FR alpha expression or those who cannot tolerate ADC-associated toxicities are not suitable for MIRV. According to the clinical experts, the diagnostic infrastructure for FR alpha expression is not yet standardized across Canada.

Assessing the Response Treatment

The clinical experts consulted for this review noted that in general, the outcomes used in clinical practice align with those observed in clinical trials of PROC treatments. Important outcomes for patients with PROC include PFS, OS, ORR, and HRQoL. Both clinical experts indicated that meaningful response to treatment should be assessed based on clinical improvement in patients' reported symptoms and on clinical and imaging assessments. This includes reduction in symptoms, radiologic evidence of tumour response, and clinical measures, including PFS and OS.

Discontinuing Treatment

According to the 2 clinical experts, disease progression or intolerable, severe toxicity should be an indication for discontinuation of treatment. The clinical experts also noted that the AEs that should lead to treatment discontinuation with MIRV are unresolved AEs, including rare ILD, blurred vision, and pneumonitis.

Prescribing Considerations

The clinical experts noted that MIRV should be prescribed and administered in specialized oncology settings by clinicians familiar with the unique safety profile of ADCs. Multidisciplinary team involvement, including ophthalmologists and/or optometrists and pulmonologists, is essential to monitor for toxicity and to provide patient counselling, given that these patients require ongoing frequent eye examinations and follow-ups during treatment to detect signs of keratopathy and manage pneumonitis.

Clinician Group Input

This section was prepared by the review team based on the input provided by clinician groups.

Input for this review was submitted by 3 clinician groups: the BC Cancer Gynecologic Oncology Provincial Tumour Group, the OH (CCO) Gynecologic Cancer Drug Advisory Committee, and GOC. The BC Cancer Provincial Gynecological Oncology Tumour Group comprises clinicians, pathologists, researchers, and allied staff involved in the care of patients with gynecological malignancies in the Province of British Columbia. OH (CCO)’s Drug Advisory Committees provide evidence-based clinical and health system guidance on drug-related issues. GOC is a national, nonprofit, multidisciplinary organization representing health care professionals involved in the treatment and prevention of gynecologic cancers, including physicians, nurses, pharmacists, and scientists. Information was gathered from published literature and the clinical experiences of physicians who treat patients with platinum-resistant cancer.

According to clinician groups, initial treatment for EOC is multimodal, typically involving a combination of surgical cytoreduction and systemic chemotherapy, most commonly carboplatin and paclitaxel. The groups noted that while initial OC treatment is typically effective, with high response rates, most patients will subsequently relapse and develop recurrent disease, which is associated with poor prognosis and survival of 6 months to 12 months. For patients with recurrence and PROC, treatment goals consist of improving PFS, delaying disease progression, and improving QoL by reducing symptom severity. In this setting, clinician groups noted that the main treatment options are bevacizumab and non–platinum-based chemotherapy, both of which have shown minimal impact on OS, have response rates of less than 30%, and are associated with significant toxicity; in addition, patients rapidly become refractory to individual lines of these treatments. Overall, the clinician groups highlighted that patients diagnosed with PROC represent a group of patients with a high unmet need for more efficacious treatment options.

The clinician group expressed differing opinions on the place in therapy for MIRV. OH (CCO) indicated that it would provide an additional treatment option for patients who have failed existing platinum-resistant options. GOC indicated that MIRV would fit into the current treatment paradigm as an additional treatment option for platinum-resistant, recurrent, high-grade serous OC. BC Cancer noted that it would be used as the first treatment option in the setting of platinum resistance with high FR alpha expression, replacing non–platinum-based chemotherapy with or without bevacizumab. GOC noted that MIRV would be used as monotherapy, not combined with other treatments. GOC anticipates that MIRV will cause a shift in the current treatment paradigm; the group’s position is that MIRV is currently the most effective treatment option available for this patient population. However, some OH (CCO) members felt that it would be difficult to justify its use as a first-line treatment in patients with platinum-resistant disease with no prior treatment with bevacizumab due to the high proportion of patients in the MIRASOL trial who had received prior bevacizumab and to the lack of head-to-head comparison with chemotherapy or bevacizumab. This organization also noted that evidence supporting the sequencing of bevacizumab followed by MIRV is limited, given that subgroup data suggest that MIRV may be less effective after bevacizumab. OH (CCO) suggested that if there are no access restrictions based on prior bevacizumab status, then real-world data should be reviewed to ensure effectiveness and improvements in OS in patients who have previously received bevacizumab.

Clinician groups agreed that the patients best suited for MIRV are those with platinum-resistant, high-grade serous EOC with high FR alpha expression (i.e., ≥ 75% of cells with ≥ 2+ staining intensity); OH (CCO) noted that patients with primary peritoneal or fallopian tube cancer would also be well suited to this treatment and suggested that eligibility should be restricted to this patient population. Those least suitable for treatment with MIRV are those with low FR alpha expression, poor functional status (e.g., ECOG 4), platinum-sensitive disease, or pre-existing conditions that may worsen the impact of AEs known to be associated with MIRV (e.g., severe liver or eye toxicities; peripheral neuropathy > grade 1; a chronic corneal disorder; history of cornea transplant; or an active ocular condition for which the patient is receiving ongoing treatment and monitoring).

In clinical practice, the clinician groups noted that PFS and OS are the standard measures for assessing treatment response. The clinician groups agreed that a clinically meaningful treatment response would be defined as radiographic disease control (tumour response or stabilization on CT or MRI) with improvement in cancer-related symptom burden and tolerable toxicity. For patients with a poor overall prognosis, GOC noted that improved PFS is clinically meaningful even in the absence of OS. GOC noted that safety and clinical assessment is performed every 3 weeks, before each treatment cycle, and radiologic assessment is performed every 9 weeks (3 cycles). Treatment discontinuation should be considered in the event of disease progression, intolerable AEs, or patient choice. Patients should receive treatment in an outpatient setting under the supervision (or guidance, in remote areas) of an oncologist. The clinician groups noted that access to optometry or ophthalmology is necessary to ensure ocular toxicities are appropriately assessed and managed.

Drug Program Input

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

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

CTCAE = Common Terminology Criteria for Adverse Events; ECOG = Eastern Cooperative Oncology Group; pERC = pan-Canadian Oncology Drug Review Expert Review Committee; PROC = platinum-resistant ovarian cancer.

Clinical Evidence

The objective of this Clinical Review Report is to review and critically appraise the clinical evidence submitted by the sponsor on the beneficial and harmful effects of MIRV (5 mg/mL, concentrate for solution for injection, IV infusion) in the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens. The focus is on comparing MIRV to relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of MIRV is presented in 4 sections, with our critical appraisal of the evidence included at the end of each section. The first section, the Systematic Review, includes pivotal studies and RCTs that were selected according to the sponsor’s systematic review protocol. Our assessment of the certainty of the evidence in this first section using the GRADE approach follows the critical appraisal of the evidence. The second section includes sponsor-submitted long-term extension studies; however, none were submitted by the sponsor. The third section includes indirect evidence from the sponsor. The fourth section includes additional studies that were considered by the sponsor to address important gaps in the systematic review evidence; however, no studies addressing gaps were submitted by the sponsor.

Included Studies

Clinical evidence from the following is included in the review and appraised in this document:

• one pivotal study (the MIRASOL trial) identified in the systematic review⁵⁰

• one ITC, including a MAIC.⁶⁴

Systematic Review

The contents within this section have been informed by materials submitted by the sponsor. The following has been summarized and validated by the review team.

Description of Studies

The MIRASOL trial was a randomized, open-label, phase III study designed to compare the efficacy and safety of MIRV to IC standard chemotherapy for patients with platinum-resistant, high-grade EOC, primary peritoneal cancer, or fallopian tube cancer whose tumours express a high level of FR alpha.

Between February 2020 and July 2022, a total of 453 patients (including patients from 8 Canadian sites) were enrolled in the study. The trial included a screening period (up to 28 days), a treatment period, an end-of-treatment visit, and a 30-day follow-up period (Figure 1). Patients were randomized at a ratio of 1 to 1 to receive single-drug MIRV (administered once every 3 weeks) or IC chemotherapy (paclitaxel [administered once weekly within a 4-week cycle], PLD [administered every 4 weeks], or topotecan [administered either on days 1, 8, and 15 every 4 weeks or for 5 consecutive days on days 1 to 5 once every 3 weeks]). Randomization was stratified by the number of previous lines of therapy (1, 2, or 3) and chemotherapy drug.

Treatment assignment was open-label due to distinct dosage schedules for each drug. However, radiographic tumour assessments were evaluated independently by a BICR that remained blinded to treatment allocation. All patients were deemed appropriate for single-drug therapy by the investigator.

FR alpha expression was assessed using to the VENTANA FOLR1 Assay (i.e., ≥ 75% of viable tumour cells with moderate [2+] and/or strong [3+] membrane staining). Disease progression was evaluated using RECIST 1.1. CT or MRI scans of tumour assessments were collected for sensitivity analysis by the BICR. Tumour assessments, including radiological assessments using CT or MRI scans, were performed at screening, every 6 weeks from cycle 1, day 1 (C1D1) for the first 36 days, and then every 12 weeks until disease progression, death, or initiation of anticancer therapy, whichever occurred first.

The primary objective of the MIRASOL trial was to compare the PFS of patients randomized to MIRV treatment versus IC chemotherapy. Key secondary objectives were to compare ORR, OS, and HRQoL assessed using the EORTC QLQ-OV28 Abdominal/GI Symptoms scale in patients randomized to MIRV treatment versus IC chemotherapy.

Characteristics of the included studies are summarized in Table 5.

Figure 1: Study Design of the MIRASOL Trial

EOT = end of treatment; Pac = paclitaxel; PLD = pegylated liposomal doxorubicin; Q3W = every 3 weeks; Topo = topotecan.

Source: MIRASOL Clinical Study Report.⁴⁰

Table 5: Details of Studies Included in the Systematic Review

AIBW = adjusted ideal body weight; CR = complete response; CTCAE = Common Terminology Criteria for Adverse Events; DOR = duration of response; ECHO = echocardiography; ECOG PS = Eastern Cooperative Oncology Group Performance Status; EOC = epithelial ovarian cancer; EORTC QLQ-OV28 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Ovarian Cancer Module; FR = folate receptor; IC = investigator’s choice; IHC = immunohistochemistry; ILD = interstitial lung disease; LVEF = left ventricular ejection fraction; MIRV = mirvetuximab soravtansine; MUGA = multigated acquisition; OS = overall survival; PARP = poly (ADP-ribose) polymerase; PD = progressive disease; PFS = progression-free survival; PLD = pegylated liposomal doxorubicin; PR = partial response; PRO = patient-reported outcome; RECIST 1.1 = Response Evaluation Criteria in Solid Tumours Version 1.1; TEAE = treatment-emergent adverse event; VENTANA FOLR1 Assay = VENTANA (FOLR1) (FOLR1-2.1) RxDx assay.

^aThis is a list of relevant end points included in this review. It is not an exhaustive list of end points included in the MIRASOL trial.

Source: MIRASOL trial Clinical Study Report.⁴⁰

Populations

Inclusion and Exclusion Criteria

The trial included female patients aged 18 years or older with platinum-resistant disease and a confirmed diagnosis of high-grade serous EOC, primary peritoneal cancer, or fallopian tube cancer. Key inclusion criteria were radiographic progression on or after the most recent line of therapy; positive FR alpha expression, as assessed by the VENTANA FOLR1 Assay; prior treatment with at least 1 but no more than 3 systemic lines of anticancer therapy; and eligibility for single-drug therapy as the next appropriate line of treatment. Patients were excluded if they had endometrioid, clear cell, mucinous, or sarcomatous histology, mixed tumours with the previously mentioned histologies, or low-grade or borderline ovarian tumour or if they had primary platinum-refractory disease or had progressed within 3 months of the last dose of first-line platinum-containing chemotherapy. Table 5 contains the detailed inclusion and exclusion criteria.

Interventions

All patients randomized to MIRV received single-drug MIRV at 6 mg/kg AIBW IV administered every 3 weeks. At C1D1, MIRV was administered at a rate of 1 mg per minute for 30 minutes; after 30 minutes, the rate was increased to 3 mg per minute if well tolerated. If well tolerated after 30 minutes at 3 mg per minute, the infusion rate was increased to 5 mg per minute. Subsequent infusions were delivered at the tolerated rate. The overall duration of infusion varied depending on dose and patient tolerance. After infusion, the IV line was flushed with 5% dextrose to ensure delivery of the full dose. Before the infusion of each MIRV dose, patients received 325 mg to 650 mg of acetaminophen (orally or through IV), 10 mg IV dexamethasone, and 25 mg to 50 mg diphenhydramine (IV or orally) (equivalent drugs of similar drug classes were also acceptable) for approximately 30 minutes. An antiemetic medication (e.g., 5-hydroxytryptamine type 3 serotonin receptor antagonists, such as palonosetron, granisetron, ondansetron, or an appropriate alternative) was recommended before each MIRV dose and used any time at the discretion of the treating physician. All patients received prophylactic corticosteroid eye drops (unless contraindicated by an ophthalmologist or physician) and daily lubricating artificial tears as directed.

For patients randomized to IC chemotherapy, the choice of chemotherapy (paclitaxel, PLD, or topotecan) was made before randomization. For all IC chemotherapy, body weight at C1D1 was used to calculate body surface area to determine the required dose; no dose modifications were anticipated unless the patient’s body weight changed by plus or minus 10% from baseline. Paclitaxel was administered at 80 mg/m² as a 1-hour IV infusion on days 1, 8, 15, and 22 of a 4-week cycle; PLD was administered at 40 mg/m² as a 1 mg per minute IV infusion on day 1 of a 4-week cycle. Topotecan was administered at 4 mg/m² as a 30-minute IV infusion on days 1, 8, and 15 of a 4-week cycle or alternatively at 1.25 mg/m² as a 30-minute IV infusion on days 1 to 5 of a 3-week cycle. Each IC chemotherapy drug was prepared as described in the prescribing information; no dose modifications were made unless the patient’s body weight changed by plus or minus 10% from baseline. Patients randomized to paclitaxel, PLD, or topotecan received premedication at the discretion of the investigator or according to institutional guidelines.

Patients continued treatment until disease progression, unacceptable toxicity, withdrawal of consent, death, or study termination by the sponsor (whichever came first). Patients who discontinued the study treatment for reasons other than progressive disease continued with tumour assessments until the start of a new anticancer therapy. All patients who discontinued the study drug were followed every 3 months (± 1 month) until death, loss to follow-up, withdrawal of consent for survival follow-up, or end of study.

Outcomes

A list of efficacy end points assessed in this Clinical Review Report is provided in Table 6, followed by descriptions of the outcome measures. Summarized end points are based on outcomes included in the sponsor’s Summary of Clinical Evidence as well as any outcomes identified as important to this review, according to the clinical experts consulted for this review and the input from patient and clinician groups and public drug plans. Using the same considerations, we selected the end points that were considered most relevant to inform expert committee deliberations and finalized this list of end points in consultation with members of the expert committee. All summarized efficacy end points were assessed using GRADE. Select notable harms outcomes considered important for informing expert committee deliberations were also assessed using GRADE.

Table 6: Outcomes Included in the Systematic Review

AE = adverse event; DOR = duration of response; EORTC QLQ-OV28 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Ovarian Cancer Module; ORR = objective response rate; OS = overall survival; PFS = progression-free survival; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

^aIf the primary end point of PFS by investigator assessment was met, then a hierarchical testing procedure was applied to key secondary end points (in the order listed previously) to control the study-wise type I error.

Source: Details included in the table are from the sponsor’s Summary of Clinical Evidence.⁶⁵

Progression-Free Survival

The primary end point of the MIRASOL trial was PFS per investigator assessment. PFS is the time from randomization until objective tumour progression or death, whichever occurs first. PFS was assessed based on radiological imaging and determined by the investigator (primary analyses) using RECIST 1.1. CT or MRI scans were collected and held for sensitivity analysis by the BICR. Clinical progression was not considered a progression end point in this trial. PFS was not considered a validated surrogate for OS because PFS has not demonstrated a benefit in OS among patients with platinum-sensitive or PROC.^66,67

Table 7: PFS Definitions in the MIRASOL Trial

PD = progressive disease.

Source: MIRASOL trial Clinical Study Report.⁴⁰

Objective Response Rate

ORR refers to the proportion of patients who achieved a pre-established reduction in tumour volume, indicating CR or PR. ORR was a key secondary outcome in the MIRASOL trial. ORR, assessed by RECIST 1.1, is an accepted end point for accelerated approval in OC.⁶⁸ Patients without postbaseline RECIST 1.1 assessment were as analyzed as those whose disease did not respond to treatment (i.e., these patients contributed to the denominator, but not the numerator).

Overall Survival

OS is the time from the date of randomization until the date of death from any cause. OS was a key secondary outcome of the trial. Patients who were alive or lost to follow-up at the analysis were censored at the last date they were known to be alive. When an analysis cut-off date was implemented, only deaths occurring on or before the cut-off date were counted as OS events. Patients without death events were censored on the latest date with a known alive status on or before the cut-off date. Patients whose date of death or last known date alive occurred after the DCO date were censored at the analysis cut-off date.

Duration of Response

DOR was defined as the time from the date of the first response (CR or PR) to the date of progression or death from any cause, whichever occurred first. DOR was a secondary outcome and was defined only for patients who had a confirmed BOR of CR or PR. Per the BOR definition, patients with an overall response of CR or PR must have had a repeat tumour assessment performed no fewer than 4 weeks after the criteria for response were first met. The first date at which a CR or PR response was noted (not the date of the confirmatory tumour assessment) was used to calculate DOR. DOR end dates and censoring rules were the same as those for PFS, as shown in Table 7.

Health-Related Quality of Life

The EORTC QLQ-OV28 was used to collect data on patients’ functioning, HRQoL, disease symptoms, and health status. The EORTC QLQ-OV28^36,37 is a disease-specific module of the Quality of Life Questionnaire Core 30 designed for patients with local or advanced OC who receive surgical treatment with or without chemotherapy. It consists of 28 items assessing abdominal and/or GI symptoms, peripheral neuropathy, other chemotherapy side effects, hormonal symptoms, body image issues, negative attitudes toward disease and/or treatment, and sexual dysfunction. Items are scored from 1 (not at all) to 4 (very much). The recall period for the first 24 questions was “during the past week”; for the last 4 questions (sexual dysfunction), it was “during the past 4 weeks.” Scores were derived by linearly transforming scale or standalone item raw scores to a 0-to-100 scale. Higher scores indicated worse symptoms or health states. The primary end point was the number of patients achieving at least a 15% (i.e., ≥ 15-point) absolute improvement on the QLQ-OV28 Abdominal/GI Symptoms subscale (items 31 to 36) at the week 8 or 9 assessment.

Harms

The numbers of patients experiencing AEs, SAEs, AEs of special interest, and mortality were assessed as secondary outcomes. An AE was defined as any noxious, pathological, or unintended change in anatomic, physiological, or metabolic function, as indicated by physical signs, symptoms, or laboratory abnormalities and occurring at any point during the clinical study. All AEs — including those attributed to study procedures — occurring from the time of informed consent until 30 days after the last dose of study treatment were reported, regardless of severity or relationship to the study drug. Investigators were responsible for managing AEs appropriately and monitoring patients until resolution. Any SAE, regardless of relationship to study medication, that occurred in a patient from the time of informed consent until 30 days after the last study treatment were recorded on the patient’s AE electronic case report form along with the investigator’s assessment of the relationship of the SAE to the study drug (i.e., MIRV, paclitaxel, PLD, or topotecan). TEAEs were defined as AEs with onset on or after the first dose of study drug and occurring within 30 days of the last dose or before the initiation of a new anticancer therapy, whichever occurred first.

Table 8: Summary of Outcome Measures and Associated Measurement Properties

EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EORTC QLQ-OV28 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Ovarian Cancer Module; GI = gastrointestinal; ICC = intraclass correlation coefficient; MID = minimal important difference.

Source: MIRASOL trial Clinical Study Report.⁴⁰

Statistical Analysis

Sample Size and Power Calculation

Sample size and power were determined based on the following assumptions: a median PFS for the IC chemotherapy group of 3.5 months, a median PFS for the MIRV group of 5.0 months, overall attrition of approximately 13% in both groups, uniform enrolment over an 18-month period, a 6-month follow-up duration, a median OS for the IC chemotherapy group of 12 months, and a median OS for the MIRV group of 17.5 months. Testing of group differences was conducted using the log-rank test.

The study was powered to test the null hypothesis of no difference in PFS between the MIRV arm and the IC chemotherapy arm against an alternative assumed HR of 0.7. To reject the null hypothesis with 90% power at a 2-sided type I error rate of 0.05, 430 patients (approximately 215 patients per treatment group) would be required. For the secondary end points of OS and ORR, it was estimated that, based on the chosen sample size of 430 patients (215 patients in each arm), an alternative OS HR of 0.6857 would have 90% power to reject the null hypothesis at a type I error rate of 0.05.

Statistical Testing

The primary outcome, investigator-assessed PFS, was estimated using the Kaplan-Meier method. Group differences in PFS were tested using the stratified log-rank test (stratified by the stratification factors used in randomization). The final analysis of PFS was assessed when at least 330 events had occurred at a DCO of March 6, 2023. If the primary end point of PFS was statistically significant at a 2-sided alpha level of 0.05, then the hierarchical testing procedure was used to test the key secondary end points in the following order at a 2-sided alpha level of 0.05 to control the study-wise type I error. The order was ORR, OS, and primary HRQoL outcome of the Abdominal/GI Symptoms scale of the EORTC QLQ-OV28. In the stratified analysis regarding the primary end point of PFS, the strata were IC chemotherapy (paclitaxel versus PLD versus topotecan) and number of prior lines of therapy (1 versus 2 versus 3). The restricted mean survival time (RMST) for investigator-assessed PFS at 12 months was compared between treatments.

For the secondary end point of ORR, differences between the MIRV treatment group and the IC chemotherapy were compared using the stratified Cochran-Mantel-Haenszel test for treatment comparison and the Clopper-Pearson method for 95% CI estimation. The secondary end points of OS and DOR were estimated using the Kaplan-Meier method. The comparison of OS and DOR between treatment groups was conducted using Cox regression and a log-rank test. The final analysis of OS was conducted when at least 300 events had occurred at the DCO of September 26, 2024. The stratified analyses were conducted according to randomization stratification factors, including the number of previous lines of therapy and IC chemotherapy. For the EORTC QLQ-OV28, the primary HRQoL outcome analysis was the number of patients achieving at least a 15% (i.e., ≥ 15-point) absolute improvement on the QLQ-OV28 Abdominal/GI Symptoms subscale at the week 8 or 9 assessment. Patients with missing week 8 or 9 questionnaires were excluded from the analysis unless they had progressive disease or death before the week 8 or 9 assessment, in which case they were included as unimproved. The secondary analysis was change from baseline in EORTC QLQ-OV28 scores, which was analyzed using a mixed model for repeated measures. The sponsor noted that interim analyses were conducted. The first interim analysis was futility only and was conducted when at least 110 PFS events had occurred. The study stopped for futility at interim analyses if the observed PFS HR (MIRV to IC chemotherapy) was greater than 1. This analysis was applied to PFS only. Another interim analysis was conducted for OS at the time of the final analysis of PFS, at which time approximately 180 deaths (60%) would have been observed. The final analysis of PFS occurred when at least 330 PFS events had occurred (DCO, March 6, 2023). A Lan-DeMets alpha-spending function using an O’Brien-Fleming stopping boundary was used to control overall type I error for OS at a 2-sided alpha level of 0.05.⁶⁹ The final analysis of OS was conducted when at least 300 deaths had occurred (DCO, September 26, 2024). It was projected that the final analysis of OS would be approximately 1 year after the final analysis of PFS. If the primary end point of PFS was statistically significant and the preliminary data suggested that 300 deaths would not occur within 1 year after the final analysis of PFS, then another interim analysis of OS would be conducted 1 year after the final analysis of PFS, and the stopping boundary would be adjusted.

Multiplicity Control

The PFS 95% CI reported was not adjusted for multiplicity; however, type I error was controlled for in the testing procedure. For the key secondary end points of ORR, OS, and the Abdominal/GI Symptoms scale of the EORTC QLQ-OV28, a hierarchical testing procedure was applied to control the family-wise type I error only if the null hypothesis for the primary end point was rejected at a 2-sided alpha level of 0.05.

Handling of Missing Data

Handling of missing data for the outcome measures is described in Table 9. For the primary and secondary end points, if the start date of subsequent anticancer therapy was missing, it was imputed using the earliest possible date that did not conflict with any other available data in the database. For records with a partial death date, the death date was imputed using the earliest possible date that did not conflict with any other available data in the database. For the HRQoL data, patients with missing data were excluded from the analysis unless otherwise described.

Sensitivity Analyses

A summary of the sensitivity analyses for select end points is provided in Table 9. If the primary end point of investigator-assessed PFS was statistically significant, PFS by BICR was analyzed using the same methods as in investigator-assessed PFS, and the concordance or discordance rate between the PFS outcomes for the BICR assessment and the investigator assessment were summarized. The RMST for PFS was estimated as a sensitivity analysis using an unstratified Kaplan-Meier estimator. Similarly, ORR and DOR as assessed by BICR were analyzed as sensitivity analyses using the same methods as in the investigator-assessed ORR and DOR, respectively. OS rates were summarized for the per-protocol population as sensitivity analyses. A series of sensitivity analyses was conducted on the EORTC QLQ-OV28 Abdominal/GI Symptoms subscale. First, analyses were repeated excluding patients who had died or progressed. Second, only patients who experienced progression before the week 8 or 9 assessment were classified as unimproved. Lastly, responder analyses were repeated using 2 alternative thresholds: 11.1 (representing the next smaller observable change score) and 22.2 (the next larger observable change score).

Subgroup Analyses

Investigator-assessed PFS, ORR, and OS were analyzed according to the following subgroups:

• BRCA status (positive versus negative or unknown)

• age (< 65 years versus ≥ 65 years)

• baseline ECOG PS (0 versus 1)

• prior exposure to bevacizumab

• number of prior lines of therapy

• primary platinum-free interval (≤ 6 months versus > 6 months)

• most recent platinum-free interval (≤ 3 months versus > 3 months)

• prior exposure to PARP inhibitor maintenance therapy (yes versus no versus uncertain)

• country (US versus rest of world)

• stage at diagnosis (I to III versus IV)

• weight at baseline (≤ 60 kg, 60 to 80 kg, or > 80 kg).

The summaries for time-to-event variables included the number and percentage of events, the median and its 95% CI, 2-sided P values, and the HR (MIRV to IC chemotherapy) and its 95% CI. All subgroup analyses were unstratified. No subgroup analyses were conducted for DOR.

Table 9: Statistical Analysis of Efficacy End Points in the MIRASOL Trial

BICR = blinded independent central review; BOR = best overall response; CI = confidence interval; CMH = Cochran-Mantel-Haenszel; CR = complete response; DOR = duration of response; eCRF = electronic case report form; EORTC QLQ-C30 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; EORTC QLQ-OV28 = European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Ovarian Cancer Module; HR = hazard ratio; IC = investigator’s choice; INV = investigator; ITT = intention to treat; LPP = longitudinal period population; MIRV = mirvetuximab soravtansine; NA = not applicable; OR = odds ratio; ORR = objective response rate; OS = overall survival; PFS = progression-free survival; PLD = pegylated liposomal doxorubicin; PR = partial response; QoL = quality of life; RECIST 1.1 = Response Evaluation Criteria in Solid Tumours Version 1.1; RMST = restricted mean survival time; vs. = versus.

Source: MIRASOL trial Clinical Study Report; MIRASOL trial Statistical Analysis Plan; MIRASOL trial patient-reported outcomes report.^65,70,71

Analysis Populations

The analysis populations in the MIRASOL trial are summarized in Table 10.

Table 10: Analysis Populations in the MIRASOL Trial

BICR = blinded independent central review; IC = investigator’s choice; ITT = intention to treat; LPP = longitudinal period population; MIRV = mirvetuximab soravtansine; ORR = objective response rate; QoL = quality of life.

Source: MIRASOL trial Clinical Study Report.⁶⁵

Results

Patient Disposition

Patient disposition in the MIRASOL trial is summarized in Table 11. Of the 227 patients randomized to MIRV treatment, 218 patients received treatment; 9 patients were not dosed due to various reasons, including patient exclusion at C1D1, pretreatment AEs, or withdrawal from treatment. Among those treated, 195 patients (86%) discontinued treatment, mainly due to disease progression (154 patients, 68%) or AEs (20 patients, 9%). The primary reason for permanent study discontinuation was death (██ patients, ██%). At the primary analysis, 32 patients (14%) remained on treatment. In the IC chemotherapy group, 207 of 226 randomized patients were treated, with 220 patients (97%) discontinuing treatment due to disease progression (145 patients, 64%) or AEs (28 patients, 12%). The IC chemotherapy group had ███ deaths (██%) as the main reason for study discontinuation. Only 6 patients (3%) remained on treatment at the primary analysis. Compared to IC chemotherapy, patients in the MIRV treatment group had lower rates of treatment discontinuation (86% versus 97%), permanent study withdrawal (██% versus ██%), and death (██% versus ██%), with more patients remaining on treatment at analysis (14% versus 3%).

Table 11: Summary of Patient Disposition in the MIRASOL Trial (DCO Date of March 6, 2023)

BICR = blinded independent central review; DCO = data cut-off; IC = investigator’s choice; ITT = intention to treat; MIRV = mirvetuximab soravtansine; NA = not applicable; PI = principal investigator.

^aPatients who withdrew consent to treatment continued to be followed.

Source: Table 10 and Table 11, MIRASOL trial Clinical Study Report.⁶⁵

Baseline Characteristics

A summary of baseline and disease characteristics for the ITT population in the MIRASOL trial are provided in Table 12. Demographic and baseline characteristics were similar between the MIRV and IC chemotherapy treatment groups. Among the 227 patients randomized to MIRV, the median age was 64.0 years, with most patients identifying as Asian (12%) or white (69%). In the IC chemotherapy group (n = 226), the median age was 62.0 years, with most patients identifying as Asian (11%) or white (64%). Among patients treated with MIRV, 80% had EOC, followed by fallopian tube cancer (12%) and primary peritoneal cancer (7%); 13% were positive for any BRCA mutations (BRCA status was negative or unknown for 87% of patients). In the IC chemotherapy group, 81% had EOC, 10% had fallopian tube cancer, and 9% had primary peritoneal cancer, with 16% BRCA-positive. All patients in the MIRV group had high-grade serous histology that was poorly differentiated (██%) or had an unknown histological grade (██%). Similarly, almost all patients (225 of 226) in the IC chemotherapy group had high-grade serous histology that was poorly differentiated (██%) or had an unknown histological grade (██%). Stage IIIC (53%) and stage IV (29%) were the most prominent cancer stages at initial diagnosis, with a median time from diagnosis to randomization of █████ months. In the MIRV treatment group, 48% of patients had received 3 prior lines of therapy, 55% had prior PARP inhibitor use, and 61% had received prior bevacizumab (compared to 46%, 56%, and 63% of patients, respectively, in the chemotherapy group). Overall, the number of patients who have had first and last lines of therapy were similar between the MIRV and chemotherapy treatment groups. The most recent platinum-free interval was greater than or equal to 3 to at least 6 months in 61% of patients in the MIRV group and in 55% of patients in the IC chemotherapy group. In both groups, most patients had an ECOG PS of 0 (57% versus 53%) or 1 (43% versus 45%).

Table 12: Summary of Baseline and Disease Characteristics of Patients (ITT Population; DCO Date of March 6, 2023)

AIBW = adjusted ideal body weight; APAC = Asia Pacific; C1D1 = cycle 1, day 1; DCO = data cut-off; ECOG = Eastern Cooperative Oncology Group; EOT = end of treatment; FR = folate receptor; IC = investigator’s choice; ITT = intention to treat; MIRV = mirvetuximab soravtansine; PARP = poly (ADP-ribose) polymerase; PLD = pegylated liposomal doxorubicin; PS = Performance Status; PS2 = moderate to strong staining intensity; SD = standard deviation.

^aPrimary diagnoses of “other” included tubo-ovarian primary (1 patient) and tubo-ovarian (1 patient).

^bPrimary diagnoses of “other” included cancers of tubo-ovarian origin (1 patient).

^cTime (months) was calculated as the number of days between the initial diagnosis date and the randomization date, divided by 30.4375.

^dFor patients who participated in double-blind trials evaluating PARP inhibitor vs. placebo for whom the actual treatment was not known.

^ePrimary platinum-free interval: 3 months to ≤ 12 months vs. > 12 months, defined as time from last dose of first-line platinum therapy to the date of disease progression and/or relapse following first-line therapy. Two patients were determined to be primary platinum-refractory (1 patient: primary platinum-free interval of 1.38 months; 1 patient: primary platinum-free interval of 0.59 months).

^fTime from last dose of latest line of platinum therapy to the date of disease progression and/or relapse following that line of therapy.

^gAll patients had an ECOG PS of 1 at screening; 3 patients shifted to an ECOG PS of 2 at C1D1.

^hThe 2 patients with missing ECOG PS were randomized but not dosed; therefore, their ECOG PSs were not recorded at screening or EOT.

Source: MIRASOL trial Clinical Study Report.⁶⁵

Exposure to Study Treatments

As of the DCO date of March 6, 2023, patients in the MIRV group had received a median of 7.0 cycles (1 to 39 cycles) of treatment. In the IC chemotherapy group, patients who received paclitaxel had a median of 4.0 cycles (1 to 9 cycles); those who received PLD had a median of 3.0 cycles (1 to 16 cycles); those who were treated with 1.25 mg/m² topotecan had a median of 6.0 cycles (1 to 19 cycles); and those who were treated with 4.0 mg/m² topotecan had a median of 2.0 cycles (1 to 12 cycles). The median duration of dosing was longer in the MIRV treatment group than in the IC chemotherapy group (4.98 months versus 2.96 months), and the number of cycles of treatment received was greater in patients who received MIRV (7.0 cycles versus 3.0 cycles). A larger proportion of patients treated with MIRV received a total of at least 2 cycles to at least 10 cycles of treatment compared with IC chemotherapy. If an incident of grade 1 medication error was reported on day 1, the patient continued MIRV treatment at the appropriate AIBW dosing without issue, and the dose was calculated using actual body weight instead of AIBW, resulting in a ██% greater dose at their first infusion.

Subsequent Anticancer Therapy

A total of ████% and ████% of patients treated with MIRV and IC chemotherapy received new anticancer therapies. In the MIRV treatment group, the most common new anticancer therapies received by patients were taxanes (████%), gemcitabine (████%), anthracyclines (████%), other chemotherapy (████%), bevacizumab (████%), and platinum compounds (████%). Of those treated with IC chemotherapy, the most common new anticancer therapies received were gemcitabine (████%), taxanes (████%), other chemotherapy (████%), platinum compounds (████%), and bevacizumab (████%).

Table 13: Summary of Patient Exposure in the MIRASOL Trial (ITT Population; DCO Date of March 6, 2023)

AIBW = adjusted ideal body weight; BSA = body surface area; DCO = data cut-off; IC = investigator’s choice; ITT = intention to treat; MIRV = mirvetuximab soravtansine; Pac = paclitaxel; PLD = pegylated liposomal doxorubicin; SD = standard deviation; Topo = topotecan.

^aFor MIRV: Dose (mg/kg) was calculated as the total cumulative dose (mg) divided by the number of valid drug administration records (performed or held infusions) and by AIBW (kg). For the remaining investigational products: Dose (mg/m²) was calculated as the total cumulative dose (mg) divided by the number of valid drug administration records (performed or held infusions) and by BSA (m²).

^bFor MIRV: calculated as (absolute dose intensity divided by 6) × 100. For Pac: calculated as (absolute dose intensity divided by 80) × 100. For PLD: calculated as (absolute dose intensity divided by 40) × 100. For Topo 1.25 mg/m² once daily for 5 days every 3 weeks.: calculated as (absolute dose intensity divided 1.25) × 100. For Topo 4 mg/m² on days 1, 8, and 15 every 4 weeks: calculated as (absolute dose intensity divided by 4) × 100.

Source: MIRASOL trial Clinical Study Report.⁶⁵

Concomitant Medications and Cointerventions

A summary of common (≥ 20%) concomitant medications is provided in Table 14. Overall, ███ (██%) and ███ (██%) of patients treated with MIRV and IC chemotherapy had at least 1 prior medication within 4 weeks of C1D1. ███% and ██% of patients treated with MIRV and IC chemotherapy received at least 1 concomitant medication, respectively. In the MIRV group, a total of ██ (██%) underwent prior and concomitant procedures compared to ██ patients (██%) in the IC chemotherapy group.

Efficacy

Progression-Free Survival

The DCO date for the primary PFS (investigator-assessed) analysis was March 6, 2023, by which time 330 events had occurred (Table 15). The median PFS was 5.62 months (95% CI, 4.34 to 5.95 months) in the MIRV treatment group compared to 3.98 months (95% CI, 2.86 to 4.47 months) in the IC chemotherapy group. The HR for PFS per investigator assessment was 0.65 (95% CI, 0.521 to 0.808; P < 0.0001). The Kaplan-Meier estimated differences in PFS probabilities between groups at 6 months was ████ (95% CI, ████ ██ ████) in favour of MIRV. At 12 months, the estimated difference in PFS probability was ████ (95% CI, ████ ██ ████) in favour of MIRV. The restricted mean PFS survival times at 12 months were ████ months (95% CI, ████ ██ ████) in the MIRV treatment group and ████ months (95% CI, ████ ██ ████) in the IC chemotherapy group. The Kaplan-Meier plot of PFS curves is shown in Figure 2.

Table 14: Common (≥ 20%) Concomitant Medications (ITT Population; DCO Date of March 6, 2023)

ATC = anatomical therapeutic chemical; DCO = data cut-off; IC = investigator’s choice; ITT = intention to treat; MIRV = mirvetuximab soravtansine.

Notes: Medications are coded using WHODrug Global-B3, March 2021 version.

Concomitant medications are defined as medications used during the course of study treatment, within 30 days of the last dose of the study drug, or before the start of a new anticancer treatment, whichever occurs first.

Source: MIRASOL trial Clinical Study Report (Table 14.1.5.2).⁶⁵

The PFS results at the latest DCO date of September 26, 2024, were consistent with those on March 6, 2023. The median PFS was 5.59 months (95% CI, 4.34 months to 5.88 months) in the MIRV group compared to 3.98 months (95% CI, 2.86 to 4.47 months) in the IC chemotherapy, with a between-group difference of ████ months (95% CI, ████ ██ ████) in favour of MIRV. The HR for PFS was 0.63 (95% CI, 0.51 to 0.78; P = 0.0001).

Results of the subgroup analyses of PFS by BICR at both DCOs were consistent with those of the overall population analyses with respect to prior number of lines of platinum and non–platinum-based lines of therapy before enrolment. PFS by BICR was evaluated as a sensitivity analysis at both DCOs; the results were generally consistent with those of the investigator assessment.

Figure 2: Kaplan-Meier Plot for PFS per Investigator (ITT Population; DCO of March 6, 2023)

DCO = data cut-off; IC = investigator’s choice; IMGN853 = mirvetuximab soravtansine; PD = progressive disease; PFS = progression-free survival.

Note: PFS is defined as the time from the date of the first dose of MIRV to the date of PD or death from any cause, whichever occurred first.

Source: MIRASOL trial Clinical Study Report (Figure 2).⁶⁵

Table 15: Summary of PFS Results From the MIRASOL Trial (ITT Population; DCO Date of September 26, 2024)

BICR = blinded independent central review; CI = confidence interval; DCO = data cut-off; HR = hazard ratio; IC = investigator’s choice; INV = investigator; ITT = intention to treat; IRT = interactive response technology; MIRV = mirvetuximab soravtansine; PFS = progression-free survival; RMST = restricted mean survival time.

Source: MIRASOL trial Clinical Study Report.⁶⁵

Overall Survival

The DCO date for the OS analysis was September 26, 2024, by which time 300 events had occurred. At a median follow-up period of 30.49 months, 300 patients had died, including ███ (████%) in the MIRV group and ███ (████%) in the IC chemotherapy group. The median OS was 16.85 months (95% CI, 14.36 months to 19.78 months) in patients randomized to MIRV compared with 13.34 months (95% CI, 11.37 months to 15.15 months) in the IC chemotherapy group. The HR for OS was 0.68 (95% CI, 0.543 to 0.840; P = 0.0004). The Kaplan-Meier estimated differences in OS probabilities between groups at 6 months was ████ (95% CI, █████ ██ ████) in favour of MIRV. At 18 months, the estimated difference in OS probability was ████ (95% CI, ████ ██ ████) in favour of MIRV. The Kaplan-Meier plot of OS curves at the DCO date of September 26, 2024, is shown in Figure 3. Additionally, the restricted mean OS (RMST) at 18 months was █████ months (95% CI, █████ ██ █████) in the MIRV group compared to █████ months (95% CI, █████ ██ █████) in the IC chemotherapy group.

The results of subgroup analyses of OS were consistent with those of the overall population analysis with respect to prior number of lines of platinum and non–platinum-based line of therapy before inclusion. A trend for benefit of MIRV in OS over IC chemotherapy, regardless of the subgroup including prior number of lines of therapy, was observed at both DCO dates. No sensitivity analyses were performed.

Figure 3: Kaplan-Meier Plot for OS (ITT Population; DCO Date of September 26, 2024) [Redacted]

Objective Response Rate

At the DCO date of September 26, 2024, █ (██%) patients randomized to MIRV and █ (██%) patients randomized to IC chemo achieved a BOR of CR, and ██ (██%) patients in the MIRV group and ██ (██%) patients in the IC chemotherapy group achieved a BOR of PR, respectively. A higher percentage of patients treated with MIRV achieved an ORR compared to patients who received IC chemotherapy (██% versus ██%), with a between-group difference of ██% (95% CI, ████ ██ ████) in favour of MIRV. The OR was ████ (95% CI, ████ ██ ███; ████████). Of the 227 patients randomized to MIRV who were included in the ITT population, ██ were nonevaluable for ORR, primarily due to not receiving the study drug; of the 226 patients in the IC chemotherapy group, ██ patients were nonevaluable for ORR, primarily due to not receiving the study drug. These ORR results were consistent with those at the DCO date of March 6, 2023.

Results of subgroup analyses were consistent with the overall analysis population for investigator-assessed ORR regardless of prior number of lines of platinum and non–platinum lines of therapy before enrolment at both DCOs. ORR per BICR was evaluated as a sensitivity analysis, and the results were consistent with those of the investigator’s assessment of ORR.

Duration of Response

At the latest DCO date of September 26, 2024, the percentages of patients who remained progression-free were similar between the MIRV and IC chemotherapy groups (██% versus ██%) Among the ██ patients with events in the MIRV group, the median DOR was ██ months (95% CI, ████ ██ ████ months) compared to ██ months (95% CI, ████ ██ ████) in the ██ patients with events in the IC chemotherapy group, with a between-group difference of ███ months (95% CI, ████ ██ ████) in favour of MIRV. These results were consistent with the results at the primary DCO.

DOR assessed by BICR was evaluated as a sensitivity analysis, and among patients randomized to MIRV, ██% had DOR events compared to ██% in the IC chemotherapy group. █████████████████████████████The median DOR by BICR was ██ months (95% CI, ████ ██ ████) in the MIRV group compared to ████ (95% CI, ████ ██ █████) in the IC chemotherapy group, with a between-group difference of ██ months (95% CI, ████ ██ ████) in favour of MIRV. No subgroup analyses were performed for DOR at either DCO date.

Health-Related Quality of Life (EORTC QLQ-OV28)

The proportions of patients who achieved a 15-point symptom subscale decrease (i.e., improvement) on the EORTC QLQ-OV28 Abdominal/GI Symptoms subscale at week 8 or 9 by March 6, 2023, were 21.0% in the MIRV group and 15.3% in the IC chemotherapy group. The between-group difference was ████ (95% CI, █████ ██ █████) in favour of MIRV.

In the analysis of change from baseline, the difference in change from baseline in EORTC QLQ-OV28 Abdominal/GI Symptoms subscale score at week 8 or 9 was –5.0 in favour of MIRV. In the subscale measuring attitude to disease, the difference in change from baseline at week 8 or 9 was 8.0 in favour of MIRV. No significant differences in change from baseline in peripheral neuropathy or sexual function at week 8 or 9 were observed between the MIRV and IC chemotherapy groups. In the attitude to disease and treatment subscale, the percentages of patients with improvement from baseline to week 8 or 9 were greater in the MIRV (████%) group than in the IC chemotherapy group (████%). The results at the DCO date of September 26, 2024, were consistent with those at the DCO date of March 6, 2023.

Table 16: Key Efficacy Results of the MIRASOL Trial (ITT Population; DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; IC = investigator’s choice; INV = investigator; IRT = interactive response technology; ITT = intention to treat MIRV = mirvetuximab soravtansine; ORR = objective response rate; OS = overall survival; Pac = paclitaxel; PH = proportional hazards; PLD = pegylated liposomal doxorubicin; RMST = restricted mean survival time; Topo = topotecan.

^aObjective response rate is defined as the proportion of patients with complete or partial response per investigator assessment.

^bStratification factors include the number of prior lines (1 versus 2 versus 3) of IC chemotherapy (Pac versus PLD versus Topo) chosen before randomization.

Source: Details included in the table are from the MIRASOL trial Clinical Study Report.⁶⁵

Table 17: Summary of HRQoL Results in the MIRASOL Trial (ITT Population; DCO Date of March 6, 2023)

CFB = change from baseline; CI = confidence interval; DCO = data cut-off; EORTC = European Organisation for Research and Treatment of Cancer; GI = gastrointestinal; HRQoL = health-related quality of life; IC = investigator’s choice; ITT = intention to treat; MIRV = mirvetuximab soravtansine; PN = peripheral neuropathy; SE = standard error; SF = sexual function.

^aEstimated using a mixed model for repeated measures.

Source: Sponsor-submitted MIRASOL trial patient-reported outcomes report. Details included in the table are from the MIRASOL trial Clinical Study Report.⁶⁵

Harms

Refer to Table 18 for harms data.

Adverse Events

A total of 210 patients (96%) treated with MIRV reported TEAEs of any grade, with 86% experiencing 1 TEAE of any grade compared with 94% of patients in the IC chemotherapy group. Notably, 42% of patients treated with MIRV reported TEAEs that were at least grade 3 in severity, and 24% reported at least 1 SAE related to MIRV. Compared to the IC chemotherapy group, a greater proportion of patients treated with MIRV reported ocular TEAEs, including dry eye (28% versus 2%), photophobia (18% versus < 1%), cataract (15% versus < 1%), and reduced visual acuity (12% versus 0%). Similarly, a higher proportion of patients treated with MIRV reported abdominal pain (30% versus 15%), arthralgia (██% versus ██%), and diarrhea (29% versus 17%). Additionally, fewer patients treated with MIRV reported neutropenia (11% versus 29%), anemia (10% versus 34%), and alopecia (1% versus 14%) compared with those who received IC chemotherapy.

Serious AEs

Patients treated with MIRV reported fewer SAEs compared to those treated with IC chemotherapy (24% versus 33%). In the MIRV treatment group, the most common SAEs of any grade were GI disorders (██%), pleural effusion (██%), and small intestinal obstruction (██%). Among patients treated with IC chemotherapy, the most common SAEs of any grade were GI disorders (██%), small intestinal obstruction (██%), and neutropenia (██%).

Withdrawals Due to TEAEs

Compared to 16% of patients in the IC chemotherapy group, 9% of patients treated with MIRV reported TEAEs leading to discontinuation. The most common events leading to discontinuation were respiratory, thoracic, and mediastinal disorders (██%), pneumonitis and blurred vision (████%). Among patients treated with IC chemotherapy, the most common AEs were GI disorders (██%), peripheral neuropathy (██%), thrombocytopenia, and fatigue (██%).

Mortality

The proportion of patients with TEAEs leading to death were similar between treatment groups (████████ ████). Among patients treated with MIRV, ██ (████%) patients died. ██ patients (███%) died while on the study drug or within 30 days of the last dose; the primary cause of death was disease progression (██ patients, ███%), followed by AEs (██ patients, ███%). ████████████ (████%) patients died 30 or more days after the last dose, with the most common causes being disease progression (██ patients, ████%), “unknown” cause (██ patients, ███%), and AEs and “other” causes (██████████ ████, ███%). Of the patients treated with IC chemotherapy, ███ (████%) patients died; ██ patients (███%) died while on the study drug or within 30 days of the last dose, with the primary causes of death being disease progression (██ patients, ███%) and AEs (██ patients, ███%). ███ ███████ (████%) patients died 30 or more days after the last dose; the most common cause of death was disease progression (██ patients, ████%).

Notable Harms

Notable harms included peripheral neuropathy, fatigue, keratopathy, blurred vision, and pneumonitis. The proportions of patients with grade 3 peripheral neuropathy were similar in both treatment groups (i.e., 3 patients in the MIRV group and 4 patients in the IC chemotherapy group). Five patients in the MIRV group reported grade 3 fatigue compared to 11 patients in the IC chemotherapy group. Compared to IC chemotherapy, a greater proportion of patients treated with MIRV reported ocular TEAEs, including keratopathy (32% versus 0%) and blurred vision (41% versus 2%). Pneumonitis was reported in ██ patients (██%) in the MIRV treatment group compared to ██ patients (██%) in the IC chemotherapy group.

Table 18: Summary of Harms Results in the MIRASOL Trial (Safety Population; DCO Date of March 6, 2023)

CTCAE = National Cancer Institute Common Terminology Criteria for Adverse Events; DCO = data cut-off; IC = investigator’s choice; MIRV = mirvetuximab soravtansine; Pac = paclitaxel; PLD = pegylated liposomal doxorubicin; TEAE = treatment-emergent adverse event; Topo = topotecan.

Source: Details included in the table are from the sponsor’s Summary of Clinical Evidence.

Critical Appraisal

Internal Validity

The MIRASOL trial used stratified randomization of treatments based on the prognostic factors: the number of prior lines of therapy (1, 2, or 3) and the IC choice of chemotherapy (paclitaxel, PLD, or topotecan). The method of allocation concealment was not explicitly described by the sponsor, but the use of centralized randomization suggests it was likely concealed. The use of an open-label design was justified because each treatment intervention has a unique dose, examination schedule, and safety profile. The baseline characteristics were generally similar between treatment groups. Although there were some differences in characteristics, such as histologic subtypes (EOC, fallopian tube cancer, and primary peritoneal cancer) or number of prior therapies, the clinical experts consulted by CDA-AMC did not expect these differences to influence the efficacy results significantly.

The open-label design of the trial introduces the possibility of reporting, performance, and detection biases in the efficacy estimation of subjective harms and benefits. Among tumour efficacy outcomes (PFS, ORR), bias was minimized through the use of BICR, with high concordance observed between the investigator- and BICR-assessed end points. OS is considered less susceptible to bias due to its objective nature. One rule in the PFS outcomes was to censor patients who received new anticancer therapy. CDA-AMC notes that this approach may inflate PFS estimates due to censoring. For example, patients who start new therapy often do so due to their prognosis or disease trajectory. The trial reported that ██% of patients randomized to MIRV and ██% in the IC chemotherapy group received subsequent therapies, taxanes, gemcitabine, or other chemotherapy. Notably, subsequent anticancer therapies were not formally controlled by the protocol; these were administered at the investigator’s discretion. We note that subsequent anticancer therapies could be a risk of bias for efficacy estimates of OS if differential access or effectiveness of subsequent therapies existed between treatment groups.

Missing data were minimal for the primary efficacy outcome of PFS and for the secondary outcomes of OS and ORR. All randomized patients were included in the ITT analysis set. The PFS and OS analysis had low attrition rates at both DCO dates. About ██% (n = ██) and ██% (n = ██) of patients randomized to MIRV and IC chemotherapy, respectively, had missing HRQoL data assessed using the EORTC QLQ-OV28. Imputation methods for missing data were not fully described. We note that this raises concerns about bias in the HRQoL results. Notably, the key secondary end point of the EORTC QLQ-OV28 did not meet the prespecified 15-point improvement in the Abdominal/GI Symptoms scale; however, a prespecified sensitivity analysis with a difference of 11.1 between groups favoured MIRV over IC chemotherapy.

The primary efficacy analyses for PFS and OS were based on stratified Cox proportional hazards models, which assume a constant HR between treatment groups over time. At the final analysis, the results demonstrated a statistically significant improvement in OS for patients treated with MIRV compared to those treated with IC chemotherapy. However, visual inspection of the Kaplan-Meier OS curves showed initial overlap between treatment arms, with notable separation beginning around 3 months and becoming more distinct by 6 months, suggesting potential nonproportional hazards. A similar pattern was observed in the PFS curves, which began to diverge around 2 to 3 months postrandomization. These patterns may indicate time-varying treatment effects, potentially violating the proportional hazards assumption. We note that neither the statistical analysis plan nor the Clinical Study Report specified whether formal tests or sensitivity analyses were conducted to assess this assumption. Although this introduces some uncertainty regarding the interpretation of HRs, the direction and magnitude of the treatment effects remained consistent across the prespecified subgroups and supported the OS benefit observed. The sponsor performed hierarchical testing beginning with PFS — then ORR, OS, and HRQoL — to control the family-wise type I error rate. Subgroup analyses were prespecified for prognostic variables, including number of prior lines of platinum- and non–platinum-based therapy. Overall, the benefits for PFS, OS, and ORR were consistent across subgroups. No interaction tests were performed.

External Validity

The population enrolled in the MIRASOL trial consisted of f patients with platinum-resistant EOC who showed high FR alpha expression (assessed using the VENTANA FOLR1 Assay as ≥ 75% of tumour cells exhibiting ≥ 2+ staining intensity). According to the clinical experts, only 35% to 40% of patients with platinum-resistant EOC would meet this biomarker threshold; this suggests that MIRV is specific to this subset of patients with EOC. In addition, only patients with an ECOG PS of 0 or 1 were eligible, whereas in practice, a considerable proportion of patients with platinum-resistant disease have an ECOG PS of 2 or higher, particularly after multiple lines of therapy. The exclusion of patients with primary refractory disease, active ocular conditions, significant neuropathy, or serious comorbidities limits the generalizability of the trial’s efficacy results to patients with worse prognostic statuses who may still be considered eligible for treatment.

The clinical experts noted that the administration of MIRV at 6 mg/kg AIBW every 3 weeks aligns with dosing in Canadian practice. However, MIRV is associated with specific ocular toxicities (e.g., blurred vision, keratopathy) that require specialized monitoring, including ophthalmologic assessments. Access to ophthalmology services is variable across Canada; wait times for specialist consultations could delay the detection and management of ocular AEs compared to the trial conditions. In the trial, systematic ophthalmologic evaluations were mandated and readily accessible.

The clinical experts consulted for this review indicated that the comparator arm of IC chemotherapy, consisting of weekly paclitaxel, PLD, or topotecan, reflected standard of care treatment options for PROC in Canada, and that the dosage schedules were consistent with Health Canada–approved labelling. The multinational nature of the trial, which involved sites in North America, Europe, Asia Pacific, and Canada, enhances the external validity, given that many participating centres were in countries with health care systems similar to Canada’s.

Input from the clinical experts and patient and clinician groups suggested that the trial measured meaningful outcomes relevant to patients, including PFS, OS, and HRQoL. However, a key limitation was the incomplete measurement of HRQoL: the primary HRQoL analysis was assessed at week 8 or 9 only, with limited longer-term data reported. Given that HRQoL was identified as a key outcome in patient and clinician input, the lack of long-term assessment of these data is a notable limitation.

GRADE Summary of Findings and Certainty of the Evidence

Methods for Assessing the Certainty of the Evidence

For the pivotal studies and RCTs identified in the sponsor’s systematic review, GRADE was used to assess the certainty of the evidence for the outcomes considered most relevant to inform expert committee deliberations, and a final certainty rating was determined as outlined by the GRADE Working Group:^38,39

• High certainty: We are very confident that the true effect lies close to that of the estimate of the effect.

• Moderate certainty: We are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. We use the word “likely” for evidence of moderate certainty (e.g., “X intervention likely results in Y outcome”).

• Low certainty: Our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. We use the word “may” for evidence of low certainty (e.g., “X intervention may result in Y outcome”).

• Very low certainty: We have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of the effect. We describe evidence of very low certainty as “very uncertain.”

Following the GRADE approach, evidence from the RCTs started as high-certainty evidence and could be rated down for concerns related to study limitations (which refer to internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, or publication bias.

When possible, certainty was rated in the context of the presence of an important (nontrivial) treatment effect; if this was not possible, certainty was rated in the context of the presence of any treatment effect (i.e., the clinical importance is unclear). In all cases, the target of the certainty of evidence assessment was based on the point estimate and its location relative to the threshold for a clinically important effect (when a threshold was available) or to the null.

The selection of outcomes for the GRADE assessment was based on the sponsor’s Summary of Clinical Evidence, consultation with clinical experts, and input received from patient and clinician groups and public drug plans. The following list of outcomes was finalized in consultation with expert committee members:

• PFS

• OS

• ORR

• HRQoL (measured using the EORTC QLQ-OV28)

• notable harms (peripheral neuropathy, fatigue, keratopathy, vision blurred, and pneumonitis).

The reference points for the certainty of evidence assessment for PFS, OS, ORR, HRQoL (assessed using EORTC QLQ-OV28), and notable harms (including peripheral neuropathy, fatigue, keratopathy, blurred vision, and pneumonitis) were set according to the presence or absence of an important effect based on thresholds informed by the clinical experts consulted for this review. The reference point for the certainty of the evidence assessment for the EORTC QLQ-OV28 Abdominal/GI Symptoms subscale was set according to the 15-point symptom subscale decrease threshold that was informed by the literature.

Results of GRADE Assessments

Table 2 presents the GRADE summary of findings for MIRV from the MIRASOL trial for the treatment of adult patients with FR alpha–positive, platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who have received 1 to 3 prior systemic treatment regimens.

Long-Term Extension Studies

No relevant long-term extension studies were submitted for this review.

Indirect Evidence

The contents within this section have been informed by materials submitted by the sponsor. The following has been summarized and validated by the review team.

Objectives for the Summary of Indirect Evidence

In the absence of direct head-to-head trials evaluating the efficacy of MIRV compared to bevacizumab and chemotherapy in adult patients with PROC who have received 1 to 3 prior systemic treatment regimens, the sponsor conducted and submitted a MAIC using IPD from the MIRASOL trial and published aggregate data from the AURLEIA trial (bevacizumab plus IC chemotherapy).⁶⁴ The objectives of this section are to summarize and critically appraise the sponsor-submitted ITC and to inform the pharmacoeconomic model.

Description of Indirect Comparison

Objective

The primary objective of this ITC was to compare the efficacy and safety of MIRV versus bevacizumab plus IC chemotherapy (AURELIA trial) in patients with previously treated, platinum-resistant epithelial ovarian, primary peritoneal, or fallopian tube cancers.

Study Selection Methods

The studies eligible for inclusion in the sponsor-submitted ITC were selected according to a sponsor-conducted literature review with a broader scope than the ITC objectives. The sponsor identified adult female patients with platinum-resistant or -refractory epithelial ovarian, primary peritoneal, or fallopian tube cancers who had received at least 1 prior systemic anticancer regimen as the broad population inclusion criteria and any pharmacological intervention as the broad intervention or comparator inclusion criteria. The sponsor’s systematic literature review was further refined by the population, intervention, comparators, outcomes, and study design (PICOS) described as relevant for the ITC in Table 19. The scope of the systematic literature review that is relevant for the ITC included trial evidence of adult patients with PROC who had received 1 to 3 prior systemic treatment regimens.

Table 19: Study Selection Criteria and Methods for the ITC Submitted by the Sponsor

AIBW = adjusted ideal body weight; HTA = health technology assessment; ITC = indirect treatment comparison; MIRV = mirvetuximab soravtansine; NMA = network meta-analysis; ORR = objective response rate; OS = overall survival; PFS = progression-free survival; PLD = pegylated liposomal doxorubicin; PROC = platinum-resistant ovarian cancer; RCT = randomized controlled trial; SLR = systematic literature review; TEAE = treatment-emergent adverse event.

Source: Details included in the table are from the sponsor’s Summary of Clinical Evidence.

ITC Analysis Methods

An indirect comparison of MIRV versus bevacizumab plus chemotherapy was conducted using a MAIC estimator. Before the analysis, a feasibility assessment was conducted to identify key areas of similarity and heterogeneity between the MIRASOL and AURELIA studies, including a comparison of inclusion and exclusion criteria, outcome definitions, and study designs. An anchored MAIC was conducted using IC chemotherapy as a common comparator in both studies, enabling a cross-trial comparison.

The MAIC was adjusted for the following baseline characteristics: IC chemotherapy, age, number of prior lines of therapy, platinum-free interval, ECOG PS, and CA-125 levels (descriptions provided in Table 20). These characteristics were identified by the sponsor as effect modifiers that were available in the published literature for the AURELIA study and in the MIRASOL trial’s IPD. Baseline characteristics were compared between trials using Wald tests and chi-square tests (or Fisher’s exact test for small frequencies) for continuous and categorical variables, respectively. Variables were matched by reweighting IPD patients in the MIRASOL trial. The sponsor identified key cross-trial differences that were not adjusted in the MAIC, including prior use of bevacizumab, positive FR alpha expression, histology type and grade, and prior use of a PARP inhibitor. The MIRASOL trial contains missing values for ECOG PS and CA-125 among all the matching characteristics. For ECOG PS, the AURELIA trial reported the number of missing values; thus, the available data in the MIRASOL trial was matched with the available data in the AURELIA trial by the specified category. However, the AURELIA trial did not report the number of missing values for CA-125; therefore, the CA-125 less than 100 U/mL category and missing values were combined into 1 category.

Individual patients in the MIRASOL study were assigned weights corresponding with the respective arms of the AURELIA study — specifically, the MIRV group was matched to the bevacizumab arm, and the chemotherapy arms were matched to each other. The weights represented the odds of being in the MIRASOL trial arm versus the corresponding AURELIA trial arm, calculated using a logistic regression model based on all matched baseline characteristics. The method of moments was used for parameter estimation in the logistic regression, and an ESS was calculated by dividing the squared sum of the weights by the sum of the squared weights. Patients in the MIRASOL trial with missing information on matching variables were removed from the analysis. The statistical analyses were conducted using R 4.2.2, with source codes adapted from the National Institute for Health and Care Excellence Decision Support Unit materials.⁷² The process is summarized in Supplementary Appendix 3.

The efficacy outcomes analyzed in the ITC of the MIRASOL and AURELIA trials included PFS per investigator, OS, and ORR per investigator, while the safety outcomes included the rates of grade 3 or higher TEAEs and treatment discontinuation due to TEAEs. These outcomes were selected based on availability and comparability across trials. The MAIC analysis methods are summarized in Table 21.

Table 20: Matching Variables in the MAIC and Definitions

CA-125 = cancer antigen 125; ECOG PS = Eastern Cooperative Oncology Group Performance Status; MAIC = matching-adjusted indirect comparison; PLD = pegylated liposomal doxorubicin; vs. = versus.

Source: Sponsor-submitted ITC report.⁶⁴

Table 21: MAIC Analysis Methods

DCO = data cut-off; ITC = indirect treatment comparison; MAIC = matching-adjusted indirect comparison; NR = not reported; ORR = objective response rate; OS = overall survival; PFS = progression-free survival; TEAE = treatment-emergent adverse event.

Source: Sponsor-submitted ITC report.⁶⁴ Details included in the table are from the sponsor’s Summary of Clinical Evidence.

Results of the ITC

Summary of Included Studies

The study designs of the MIRASOL and AURELIA trials were generally comparable. Some of the differences between trials were in the distributions of patient characteristics. Some of these characteristics were adjusted for in the matching procedure, including the number of prior lines of therapy, platinum-free interval, choice of preassigned chemotherapy, and ECOG PS. However, other characteristics were not feasible to include as adjustment variables due to lack of testing or reporting in the AURELIA trial, more restrictive inclusion criteria in the MIRASOL trial, or impact on sample size. FR alpha expression was considered a potential effect modifier for both MIRV and pooled chemotherapy;⁷³ however, FR alpha was not reported in the AURELIA trial. The 2 trials differed in the criteria used to measure response, progression, and AEs as well as in the censoring rules applied to PFS and OS. The assessment of homogeneity related to the MIRASOL and AURELIA trials is summarized in Table 22.

Table 22: Assessment of Homogeneity in the ITC

CA-125 = cancer antigen 125; CR = complete response; ECOG PS = Eastern Cooperative Oncology Group Performance Status; FR = folate receptor; GCIG = Gynecologic Cancer InterGroup; ITC = indirect treatment comparison; ITT = intention to treat; MAIC = matching-adjusted indirect comparison; NCI CTCAE = National Cancer Institute Common Terminology Criteria for Adverse Events; OC = ovarian cancer; ORR = objective response rate; OS = overall survival; PARP = poly (ADP-ribose) polymerase; PFS = progression-free survival; PLD = pegylated liposomal doxorubicin; PR = partial response; RECIST 1.0 = Response Evaluation Criteria in Solid Tumour Version 1.0; RECIST 1.1 = Response Evaluation Criteria in Solid Tumour Version 1.1; TEAE = treatment-emergent adverse event; vs. = versus.

Source: Details included in the table are from the sponsor’s Summary of Clinical Evidence.

Results

After excluding patients with missing characteristics, a total of ███ patients who received MIRV in the MIRASOL trial were matched with the ███ patients who received bevacizumab plus chemotherapy in the AURELIA trial, and ███ patients in the IC chemotherapy arm in the MIRASOL trial were matched with the ███ patients who received chemotherapy in the AURELIA trial. After matching on the baseline characteristics listed in Table 20, the ESSs were ██ for the MIRV arm and ██ for the chemotherapy arm. Weights ranged from ████ ██ ████ for the MIRV arm, with ██% of patients assigned to a weight less than or equal to 0.50 and ██% assigned to a weight greater than or equal to 1. In the chemotherapy arm, weights ranged from ████ ██ ████ with ██% of patients receiving a weight less than or equal to 0.50 and ██% receiving a weight greater than or equal to 1.

The baseline characteristics were compared between the MIRASOL and the AURELIA studies and are presented in Table 23 before and after matching. Before matching, preassigned chemotherapy, numbers of prior lines, and platinum-free interval durations were statistically significantly different between the studies. After matching, all baseline characteristics included in the adjustment were balanced between the studies.

Table 23: Baseline Characteristics Before and After Matching

BEV = bevacizumab; CA-125 = cancer antigen 125; ECOG PS = Eastern Cooperative Oncology Group Performance Status; ESS = effective sample size; MIRV = mirvetuximab soravtansine; PLD = pegylated liposomal doxorubicin.

^aIn the MIRASOL trial, patients with missing values in matched baseline characteristics were removed from both the before- and after-matching analyses.

^bIn the AURELIA trial, 3 patients had missing values for platinum-free interval (1 in the bevacizumab plus chemotherapy arm and 2 in the chemotherapy arm). In addition, 5 patients had missing values for ECOG PS (2 in the bevacizumab plus chemotherapy arm and 3 in the chemotherapy arm). Proportions were calculated among patients with nonmissing values.

^cFor CA-125, matching was based on CA-125 ≥ 100 U/mL vs. < 100 U/mL or missing. In the MIRASOL trial, there were 28 patients with missing baseline CA-125 values in the MAIC sample (11 in the MIRV arm and 17 in the chemotherapy arm of the ITT population). In the AURELIA trial, the number of patients with missing values was not reported.

Source: Sponsor-submitted ITC report.⁶⁴

Progression-Free Survival

The PFS curves after matching are shown in Figure 4, with median PFS results reported in Table 24. Before matching, the median PFS was ████ months (95% CI, ████ ██ ████) for the MIRV arm and ████ months (95% CI, ████ ██ ████) for the chemotherapy arm in the MIRASOL trial. After matching, the median PFS for the MIRV and chemotherapy arms in the MIRASOL study increased to ████ months (95% CI, ████ ██ █████) and ████ months (95% CI, ████ ██ ████), respectively. For the AURELIA trial, the median PFS was ████ months (95% CI, ████ ██ ████) for the bevacizumab and chemotherapy arm and ████ months (95% CI, ████ ██ ████) for the chemotherapy arm.

Figure 4: KM Curves of PFS After Matching (MIRASOL Trial, DCO Date of September 26, 2024) [Redacted]

Table 24: Median PFS Before and After Matching (MIRASOL Trial, DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; MIRV = mirvetuximab soravtansine; PFS = progression-free survival.

Source: Sponsor-submitted ITC report.⁶⁴

The HRs for PFS were estimated using a Cox proportional hazards model before and after matching. The risk of progression or death in the MIRV arm was comparable to that in the bevacizumab and chemotherapy arm, both before (HR = ████; 95% CI, ████ ██████) and after matching (HR = ████; 95% CI, ████ ██ ████). The proportional hazards assumption was not rejected based on the scaled Schoenfeld residuals when comparing PFS across the MIRV and chemotherapy arms in the MIRASOL trial (before matching: P value = ████; after matching: P value = ████). However, the proportional hazards assumption was rejected when comparing the bevacizumab plus chemotherapy and chemotherapy arms in the AURELIA trial (P value ████).

Table 25: PFS HR Before and After Matching (MIRASOL Trial, DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; HR = hazard ratio; ITC = indirect treatment comparison; MIRV = mirvetuximab soravtansine; PFS = progression-free survival; vs. = versus.

Source: Sponsor-submitted ITC report.⁶⁴

Overall Survival

The OS curves after matching are shown in Figure 5; median OS is shown in Table 26. The median OS periods were █████ months (95% CI, █████ ██ █████) for the MIRV arm and █████ months (95% CI, █████ ██ █████) for the chemotherapy arm in the MIRASOL trial. After matching, in the MIRASOL trial, the median OS for the MIRV arm increased to █████ months (95% CI, █████ ██ █████), while the median OS for the chemotherapy arm decreased to █████ months (95% CI, ████ ██ █████). For the AURELIA trial, the median OS periods were █████ months (95% CI, █████ ██ █████) for the bevacizumab plus chemotherapy arm and █████ months (95% CI, █████ ██ █████) for the chemotherapy arm.

Figure 5: KM Curves of OS After Matching (MIRASOL Trial, DCO Date of September 26, 2024) [Redacted]

Table 26: Median OS Before and After Matching (MIRASOL Trial, DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; ITC = indirect treatment comparison; MIRV = mirvetuximab soravtansine; OS = overall survival.

Source: Sponsor-submitted ITC report.⁶⁴

The HR estimates from the Cox proportional hazards model before and after matching are shown in Table 27. Risk of death in the MIRV arm was lower than in the bevacizumab and chemotherapy arm, with an HR of ████ (95% CI, ████ ██ ████) before matching and an HR of ████ (95% CI, ████ ██ ████) after matching for MIRV versus bevacizumab plus chemotherapy. The proportional hazards assumption based on the scaled Schoenfeld residuals was met both before and after matching for the MIRASOL trial (before matching: P value = ████; after matching: P value = ████) and the AURELIA trial (P value = ████).

Table 27: OS HR Before and After Matching (MIRASOL Trial, DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; HR = hazard ratio; ITC = indirect treatment comparison; MIRV = mirvetuximab soravtansine; OS = overall survival; vs. = versus.

Source: Sponsor-submitted ITC report.⁶⁴

ORR (MIRASOL Trial, DCO Date of September 26, 2024)

After matching, the estimated ORRs in the MIRASOL trial were ██% within the MIRV arm and ██% within the chemotherapy arm. The OR of ORR for MIRV versus bevacizumab and chemotherapy was ████ (95% CI, ████ ██ ████) before matching; it increased to ████ (95% CI, ████ ██ ████) after matching.

Table 28: ORR ORs Before and After Matching (MIRASOL Trial, DCO Date of September 26, 2024)

CI = confidence interval; DCO = data cut-off; ITC = indirect treatment comparison; MIRV = mirvetuximab soravtansine; OR = odds ratio; ORR = objective response rate; vs. = versus.

Source: Sponsor-submitted ITC report.⁶⁴

Harms

TEAEs of Grade 3 or Higher

In the AURELIA trial, the grade 3 or higher TEAE rates were ██% in the bevacizumab plus chemotherapy arm and ██% in the chemotherapy arm. After matching, the grade 3 or higher TEAE rates in the MIRASOL trial were ██% in the MIRV arm and ██% in the chemotherapy arm. The OR of grade 3 or higher TEAEs for MIRV versus bevacizumab and chemotherapy was ████ (95% CI, ████ ██ ████) before matching. After matching, the OR of MIRV versus bevacizumab and chemotherapy was ████ (95% CI, ████ ██ ████), indicating numerically lower odds of grade 3 or higher TEAEs in the MIRV arm than in the bevacizumab and chemotherapy arm, as shown in Table 29.

Table 29: Grade 3 or Higher TEAE Rate ORs Before and After Matching (MIRASOL Trial, DCO Date of September 26, 2024)