An immune molecular subgroup of patients with high-grade serous ovarian cancer have significantly improved progression-free and overall survival compared with patients classified as having “proangiogenic” subtypes of high-grade serous ovarian cancer, according to the results of a study (Abstract 5502) presented at the Gynecologic Cancers Oral Abstract Session Saturday.
However, this same immune molecular subgroup had their OS halved if they are treated with first-line bevacizumab using the ICON7 scheduling. These patient subgroups were identified using a 63-gene signature developed by Charlie Gourley, PhD, chair of medical oncology and honorary consultant in medical oncology at the Edinburgh Cancer Research Centre, and colleagues.
“Bevacizumab is widely prescribed for the treatment of [high-grade serous ovarian cancer] yet clinical trials have failed to show a benefit in terms of improved overall survival,” Dr. Gourley said in an interview with ASCO Daily News. “Based on this, bevacizumab has not been approved for use in ovarian cancer by the FDA, yet is widely prescribed in the United States, off label.”
Dr. Gourley and colleagues conducted this analysis in search of an appropriate biomarker that might help to stratify patients who will derive the greatest benefit from antiangiogenic drugs like bevacizumab.
In this study, they extracted mRNA from 265 formalin-fixed paraffin-embedded samples taken from Scottish patients with high-grade serous ovarian cancer (treated with primary debulking then platinum-based chemotherapy) and performed transcriptional analyses.
Using unsupervised hierarchical clustering, the researchers identified three major subgroups of patients: two with angiogenic gene upregulation (the proangiogenic groups) and one with angiogenic gene repression and immune gene upregulation (the immune molecular subgroup). A survival analysis of this group of patients indicated that those patients in the immune subgroup had a significantly improved overall survival (OS) compared with the combined proangiogenic groups (p = 0.001 for both).
Based on these results, a 63-gene signature was developed to distinguish patients in the immune subgroup from those in the two proangiogenic subgroups. Using this gene signature, Dr. Gourley and colleagues applied their findings to the Tothill dataset in silico. This analysis confirmed that patients in the immune subgroup had significantly improved progression-free survival (PFS; hazard ratio [HR] 0.661, 95 % CI 0.439, 0.996; p = 0.048) and OS (HR 0.357, 95% CI 0.219, 0.582; p < 0.001) compared with the proangiogenic groups.
The results were further validated in 284 samples taken from the ICON7 study, which assigned patients to paclitaxel/carboplatin with or without concomitant and maintenance bevacizumab.
Looking at just those patients in the control arm of ICON7, patients in the immune subgroup again had significantly improved PFS (HR 0.47, 95% CI 0.32, 0.71; p < 0.001) and OS (HR 0.45, 95% CI 0.26, 0.79; p = 0.005) compared with the proangiogenic groups.
Dr. Gourley suggested two possible explanations for these improved outcomes.
“Firstly, as this subgroup has repression of angiogenic processes the survival may be better because the blood supply to the tumor is less well developed,” Dr. Gourley said. “Secondly, it is possible that in these patients the active engagement of the tumor by the host immune system results in more indolent disease.”
However, in the group of patients treated with bevacizumab, those patients in the immune subgroup had significantly worse PFS (p = 0.015), with a median PFS of 18.5 months when treated with bevacizumab compared with 35.8 months on carboplatin/paclitaxel alone. In contrast, the proangiogenic group had a trend toward a better PFS when treated with bevacizumab, with a median PFS of 12.3 months on carboplatin/paclitaxel alone compared with 17.4 months with the addition of bevacizumab.
In his discussion of the abstract, Jonathan A. Ledermann, MD, FRCP, of University College London Cancer Institute, United Kingdom, said that very little is currently understood about the mechanism of action by which these results might occur, but that it is very important moving forward to study the interaction further.