Angeles Alvarez Secord

Purpose: Evaluate efficacy and safety of carboplatin plus mirvetuximab soravtansine (MIRV) in folate receptor alpha (FRa)-expressing recurrent PSOC. Patients and Methods: This phase 2, single arm trial enrolled patients with recurrent FRa-expressing (≥25% of tumor cells with ≥2+ membrane intensity) PSOC and 1 prior platinum-based therapy. Patients received up to 6-8 cycles of carboplatin (AUC 5) plus MIRV (6mg/kg AIBW) Q3W, followed by single-agent MIRV. Primary and key secondary endpoints were investigator-assessed objective response rate (ORR) per RECIST v1.1 in the FRa ≥50% group and FRa ≥25% group (ie, overall population), respectively. Results: Among 125 enrolled patients (102 FRα ≥50%; 61 prior PARPi), ORR by end of 6 combination cycles was 62.7% (95% CI, 52.6–72.1) in the FRα ≥50% group and 62.4% (95% CI, 53.3–70.9) in the overall population. Most patients (n=101; 81%) transitioned to single-agent MIRV continuation with durable responses. Patients with FRa ≥25% had median duration of response of 11.2 months (95% CI, 8.5–13.6) and median progression-free survival of 11.0 months (95% CI, 9.8–13.1). Most common adverse events (AEs) were blurred vision (81%) and corneal events (74%); 96.7% of corneal cases resolved or had final grade 1. Treatment-emergent AEs led to MIRV and carboplatin discontinuation in 15% and 9%, respectively. Adjudicated treatment-related pneumonitis occurred in 19 (15.2%) patients: 14 grade 1–2; 3 grade 3–4, and 2 fatal cases. Conclusions: Carboplatin plus MIRV was manageable with dose modifications for AEs and showed meaningful activity, including in PARPi-exposed patients, supporting further evaluation in PSOC.

PERK and other potential biomarkers of therapy response in LGSOC. A, pERK IHC examples demonstrating negative (0), weak (1+), moderate (2+), and intense (3+) positivity for histoscore calculation. Scale bars, 50 µm. B, Response to trametinib (top) and physician’s choice SOC (bottom) according to KRAS/BRAF/NRAS mutation status (left), pERK status (middle), and chr1 abnormalities (right). C, pERK histoscore according to KRAS/BRAF/NRAS mutation type. Labels specify the numbers within each group that were evaluable for pERK expression levels. D, pERK histoscore by chr1p-loss status. E, Frequency of pERK status between chr1p-loss and -intact groups. Chr1pq-aberrant, chr1p-loss with concurrent chr1q-gain.

Study case flow diagram for WES of translational specimens from patients enrolled in GOG281. One hundred thirty-four samples underwent successful WES, of which 112 were also evaluable for pERK via IHC. Additionally, 36 samples were evaluable for pERK IHC that did not have matching WES.

Molecular landscape of tumor samples from patients enrolled in GOG281/LOGS. MAPK-associated, MAPK-associated genes as defined in the Gene Ontology Term GO:0000165; Chr1pq-aberrant, chr1p-loss with concurrent chr1q-gain.

Purpose:Low-grade serous ovarian carcinoma (LGSOC) is a distinct form of ovarian cancer characterized by younger patient age and relative chemoresistance. The GOG281/LOGS trial (NCT02101788) investigated the efficacy of the MEK inhibitor trametinib compared with physician’s choice standard-of-care (SOC) in patients with LGSOC with persistent/recurrent disease. The study demonstrated significantly improved progression-free survival (PFS) in the trametinib-treated arm.Experimental Design:Two hundred and sixty patients with recurrent/persistent LGSOC were enrolled and randomly assigned in GOG281. We performed molecular analysis of 170 patients with available tumor specimens, comprising whole-exome sequencing and phospho-ERK (pERK) IHC, to identify biomarkers of clinical benefit from trametinib. The demographics of the translational cohort (n = 170) were comparable with those of the total trial cohort.Results:High tumor pERK expression (greater than the median histoscore of 140) was associated with significantly prolonged PFS with trametinib treatment versus SOC (median 20.1 vs. 5.6 months, log-rank P < 0.0001; test for interaction P = 0.023). Tumors harboring canonical RAS–RAF–MAPK mutations (KRAS/BRAF/NRAS: 44/134, 32.8% of cases) had a higher response rate to trametinib (50.0% vs. 8.3%; Barnard’s P = 0.0004; test for interaction P = 0.054), but KRAS/BRAF/NRAS status was not predictive of prolonged PFS (test for interaction P = 0.719). KRAS amplification (n = 5 without KRAS/NRAS/BRAF mutation) and mutation of MAPK-associated genes (n = 25 without KRAS/NRAS/BRAF mutation or KRAS copy number gain) expanded the number of cases with identifiable MAPK defects to 55.2%, but consideration of these events did not improve the discrimination of trametinib responders. Chr1p loss (49% of cases) was associated with lower pERK expression (P = 0.021).Conclusions:This exploratory analysis suggests that pERK expression and mutation of KRAS/BRAF/NRAS are candidate biomarkers of improved PFS and response to trametinib, respectively.

Demographics of translational research population.

Impact of pERK status and KRAS/BRAF/NRAS mutation status on GOG281 patient outcome. A, PFS of GOG281/LOGS patients with low-pERK tumors (pERK ≤ 140) in trametinib versus SOC arms. B, PFS of GOG281/LOGS patients with high-pERK tumors (pERK > 140) in trametinib versus SOC arms. C, PFS of GOG281/LOGS patients with KRAS/BRAF/NRAS WT tumors in trametinib versus SOC arms. D, PFS of GOG281/LOGS patients with KRAS/BRAF/NRAS-mutant tumors in trametinib versus SOC arms. Labeled HR refer to comparison of trametinib versus SOC arm.