NEW YORK — Researchers have identified three different states of an aggressive type of ovarian cancer based on differences in clonal heterogeneity and active pathways.
These findings could contribute to personalizing treatments for ovarian cancer patients and also help elucidate why most patients develop resistance to chemotherapy and relapse, said Jaana Oikkonen, a postdoctoral researcher in molecular oncology at the University of Helsinki and co-corresponding author on the new study.
High-grade serous carcinoma (HGSC) is the most common epithelial ovarian cancer subtype and is typically diagnosed at a metastasized state.
While 80 percent of the patients with HGSC respond well to the current standard of care, almost all patients experience relapse within a few years after diagnosis, leading to a five-year survival rate of less than 40 percent, the authors noted.
For years, researchers have suspected that relapse rates are high among patients with HGSC possibly because the tumors contain genetically heterogeneous clones.
"However, so far it had not been studied whether this tumor heterogeneity is present in all patients and what level of heterogeneity the tumors have," Oikkonen said.
For this prospective and longitudinal study, published in Cancer Cell on Thursday, Oikkonen and colleagues used whole-genome sequencing data from tumors of 510 treatment-naive patients and post-treatment samples from 148 patients with HGSC. They also analyzed mutation and transcriptomic data among the samples.
Based on the genomic, pathway, and phenotypic differences, the researchers were able to categorize the pre-treatment samples into three different classes: evolving, maintaining, and adaptive states. These three evolutionary states were characterized by "distinct temporal order and enriched aberrations in the oncogenic signaling cascades," the authors noted.
The evolving state was estimated to be the youngest state in evolutionary time and showed aberrations in MAPK and ERBB2 signaling, the researchers noted. In this state, the cancers exhibited only a few subclones and showed the best response to therapies, as was noted in post-treatment samples.
In the adaptive state, which was estimated to be the most evolved among the three evolutionary states, various signaling cascades, including driving NOTCH and WNT pathways, were affected.
However, the maintaining state tumors were the most insidious. Tumors in this state had aberrations in the PI3K/AKT pathway and had multiple subclones, whose frequencies remained similar between different metastatic sites. In addition, patients with tumors in this state were associated with the shortest therapy response, the authors noted.
Coauthor Alexandra Lahtinen, also a postdoctoral researcher at University of Helsinki, said that the findings suggest that when the tumor metastasizes, it might be important to collect tumor samples from multiple sites in order to decide which targeted treatments may work best for patients.
Furthermore, the researchers developed and performed a so-called nested pathway analysis, which tests mutational relationships at the pathway level between the evolutionary states to estimate relatedness and identify changed pathways. This analysis showed that tumors could transition from evolving state to adaptive either directly or through the maintaining state.
"Patients whose tumors evolve directly to the adaptive state from the evolving state have better survival than patients whose tumors evolve to adaptive state via the maintaining state," the authors wrote. The latter was characterized by PI3K/AKT pathway alterations, they added.
In subsequent experiments, the researchers used PI3K inhibitors on five cancer organoids, and highlighted that the inhibitor alpelisib — currently being evaluated in the EPIK-O Phase III clinical trial in combination with olaparib in patients with platinum-resistant HGSC — may give good results.
Lahtinen said that her team would like to include more post-treatment samples to understand how chemotherapy affects tumors in each of the three states identified in this study. "However previous studies have hinted that there are no huge differences before and after chemotherapy," she added.