ImmunoGen presents full data and exploratory analyses from phase III FORWARD I study of mirvetuximab soravtansine in ovarian cancer at ESMO.

ImmunoGen, Inc. announced full data and additional exploratory analyses from the Phase III FORWARD I study evaluating mirvetuximab soravtansine compared to chemotherapy in women with folate receptor alpha (FRalpha)-positive, platinum-resistant ovarian cancer during an oral presentation at the European Society for Medical Oncology (ESMO) 2019 Congress in Barcelona, Spain. �While it is disappointing that FORWARD I did not meet the primary endpoint of progression-free survival, mirvetuximab demonstrated consistent and meaningful efficacy signals in patients with high levels of FRalpha expression and was well tolerated with a differentiated safety profile in both the ITT and FRalpha high populations,� said Dr. Kathleen Moore, Associate Director of Clinical Research at the Stephenson Cancer Center at the University of Oklahoma. �Despite recently reported advances in frontline treatment with the addition of PARPi maintenance therapy, the majority of patients will unfortunately develop platinum-resistant disease with limited therapeutic options characterized by low response rates, short progression-free survival, and significant toxicities. The encouraging data from FORWARD I suggest the potential for a significant improvement over single-agent chemotherapy in the FRalpha high population and I look forward to the continued development of mirvetuximab for these patients in the upcoming Phase III study.�

The FORWARD I Phase III trial randomized 366 patients 2:1 to receive either mirvetuximab or the physician's choice of single-agent chemotherapy (pegylated liposomal doxorubicin, topotecan, or weekly paclitaxel). Eligibility criteria included patients with platinum-resistant ovarian cancer that expressed medium or high levels of FRalpha, who had been treated with up to three prior regimens. The primary endpoint of this study was progression-free survival (PFS), which was assessed using the Hochberg procedure in the entire study population and in the subset of patients with high FRalpha expression. The Hochberg procedure enables the simultaneous testing of two overlapping populations. Under this statistical analysis plan, if the p-value of the primary endpoint in either population is greater than 0.05, the p-value in the other population needs to be less than or equal to 0.025 to achieve statistical significance.

Key Findings from the Phase III FORWARD I Study : In the entire study population, the confirmed overall response rate (ORR) was higher for mirvetuximab than for chemotherapy (22% vs 12%, p-value 0.015), without a significant difference in the primary endpoint of PFS (HR 0.981, p-value 0.897) or overall survival (OS) (HR 0.815, p-value 0.248). In the pre-specified FRalpha high subgroup (218/366, 60%): Median PFS (mPFS) was longer in patients who received mirvetuximab compared with chemotherapy (4.8 months vs 3.3 months, HR 0.693, p-value 0.049). Given that the p-value in the entire study population exceeded 0.05, the statistical analysis plan for the study required the p-value in the high subset to be less than or equal to 0.025 to achieve statistical significance.Confirmed ORR was higher for mirvetuximab than for chemotherapy (24% vs 10%, p-value 0.014). OS was longer in patients who received mirvetuximab compared with chemotherapy (HR 0.618, p-value 0.033). The trend in improved OS in patients who received mirvetuximab compared with chemotherapy persisted with an additional 6 months of follow-up (updated through August 2019: HR 0.678, with median OS [mOS] 16.4 months vs 12.0 months, p-value 0.048).

Mirvetuximab was well-tolerated, with fewer patients experiencing grade 3 or greater treatment emergent adverse events (TEAEs) (46% vs 61%), fewer dose reductions (20% vs 31%), and fewer discontinuations due to drug-related TEAEs (5% vs 8%) compared with chemotherapy. The safety profile of mirvetuximab was confirmed, with the most common drug-related adverse events including nausea (46% all grades; 1% grade 3 or greater), blurred vision (42% all grades; 2% grade 3 or greater), and keratopathy (33% all grades; 1% grade 3 or greater). Over twice the percentage of patients who received mirvetuximab compared with chemotherapy reported improved quality of life, as measured by at least a 15-point improvement in the abdominal/GI symptom subscale of the EORTC-QLQ OV28 (32% vs 14%)

Exploratory Analyses : �While FORWARD I generated promising outcomes in the FRalpha high subgroup, the anti-tumor activity did not reach the levels we have observed in our previous studies with mirvetuximab. Accordingly, we have undertaken a comprehensive assessment of the factors that may have contributed to the outcomes in FORWARD I. These exploratory analyses demonstrate that the use of a simplified scoring method to assess tumor samples for FRalpha expression inadvertently introduced a population of patients into FORWARD I with lower levels of FRalpha than intended,� said Anna Berkenblit, M.D., Senior Vice President and Chief Medical Officer of ImmunoGen. �When we reassessed the FORWARD I tumor samples using the scoring method from our previous studies, we determined that a significant percentage of patients included in FORWARD I had low levels of FRalpha expression that should have precluded enrollment. For those patients with medium or high levels of FRalpha expression upon rescoring, we observed efficacy outcomes for mirvetuximab much more in line with our previous experience, with improved activity correlating with FRalpha expression and the strongest treatment effect for all efficacy endpoints in the intended FRalpha high patient population. These findings have informed the design of our planned Phase III registration trial in FRalpha high patients.�

Previous studies with mirvetuximab have used a PS2+ scoring method to assess tumor samples for FRalpha expression to determine eligibility. The PS2+ scoring method assesses both intensity of staining (0, 1+, 2+, or 3+) and percentage of tumor cells staining at each intensity, with at least 50% of cells with at least 2+ staining considered FRalpha medium and at least 75% of cells with at least 2+ staining considered FRpalpha high. In preparation for launch of a companion diagnostic for commercial use, a simplified scoring method to assess FRalpha expression, known as 10X, was implemented prior to the start of FORWARD I. Eligibility was determined by scoring the percentage of tumor cells with positive membrane staining by less than 10X magnification without the need to separately assess level of intensity. A bridging study indicated that the 10X scoring method was sufficient for patient selection: staining visible at less than 10X magnification correlated with higher intensity staining (2+ and 3+), with lower intensity staining visible only at higher magnification. Comparison to the much larger dataset from patients enrolled in FORWARD I, however, suggested a significant population shift towards increased prevalence of FRalpha expression under the 10X scoring method as compared to the PS2+ scoring method. Rescoring of the FORWARD I tumor samples by an independent pathologist, blinded to treatment assignment, using the PS2+ method demonstrated that 34% of patients enrolled in FORWARD I had FRalpha expression below the intended level. In addition, the FRalpha high subset enrolled in the study also contained a mixture of FRalpha expression when scored using the PS2+ method.

Key Findings from Exploratory PS2+ Scoring for FRalpha Determination in Phase III FORWARD Study :Exploratory efficacy analyses of the FORWARD I patients scored using the PS2+ method demonstrate improved outcomes correlated with FRalpha expression, with the strongest treatment effects for all efficacy endpoints in the PS2+ FRalpha high population (n=116). Compared with chemotherapy, mirvetuximab was associated with: Longer PFS (mPFS 5.6 months vs 3.2 months, HR 0.549 [95% CI 0.336, 0.897]); Higher confirmed ORR (29% vs 6%); and Longer OS (updated through August 2019: mOS 16.4 months vs 11.4 months, HR 0.678 [95% CI 0.410, 1.119]). �With the results of these exploratory analyses, we have developed a clear view of which patients benefit most from mirvetuximab and how to best identify those patients,� said Mark Enyedy, ImmunoGen�s President and Chief Executive Officer. �We are working closely with FDA to finalize the design of a Phase III registration trial for mirvetuximab, which we call MIRASOL, and believe that the robust data generated from the FORWARD I analyses increase the likelihood of a positive outcome with this next study. We anticipate enrolling the first patient by the end of the year with topline readout in the first half of 2022.�