A multi-omic analysis of high-grade serous ovarian cancer has uncovered molecular differences between clinical subgroups of the disease.

Ovarian cancer is common and aggressive, affecting about 22,500 people in the US each year, according to the National Cancer Institute. It also exhibits high levels of tumor heterogeneity and clinical outcomes.

Researchers led by the University of Texas MD Anderson Cancer Center analyzed tumor biopsies from 30 women with high-grade serous ovarian cancer (HGSC), some of whom had no visual evidence of disease after surgery, some of whom responded well to neoadjuvant chemotherapy, and some of whom responded poorly to neoadjuvant chemotherapy. This analysis indicated that tumors that underwent full resection differed molecularly from the other subgroups and that some of these differences between cancer subgroups could be applied to predict disease response.

"Our findings provide an understanding of the heterogeneity within HGSCs," MD Anderson's Anil Sood and his colleagues wrote in their paper, which appeared Tuesday in Cell Reports.

In the study, patients' tumors underwent laparoscopic assessment to determine whether complete gross resection (R0) was possible and to obtain tumor tissue. Primary R0, they noted, is associated with better clinical outcomes, though it has been unclear whether that is due to the surgery or the sort of tumor that lends itself to surgical resection.

They collected primary tumor tissue and tissue from metastatic sites, which they analyzed using whole-genome sequencing to an average somatic coverage of 118X and to germline coverage of 38X. They also conducted targeted deep-sequencing, RNA sequencing, and proteomic and immune profiling on these samples.

Across all both primary and metastatic samples, the most frequently mutated gene was TP53. However, the types of mutations affecting TP53 varied by tumor subgroup. For instance, the subgroups that underwent neoadjuvant chemotherapy (NACT) were more likely to have nonsense TP53 mutations, while the R0 group had missense mutations.

Other alterations varied by subgroup. About 54 percent of R0 group tumors had copy number loss of NF1, while the NACT groups had much lower levels of NF1 loss, about 18 percent. NF1 mRNA and protein levels were similarly decreased in the R0 group as compared to the NACT groups. NF1, the researchers noted, could be a marker to predict R0 and NACT response.

R0 subgroup tumors additionally exhibited a strong binding neoantigen signature, as compared to the NACT subgroups. This neoantigen signature, they noted, is associated with the increased infiltration of tumor immune cells. When they examined the immune profiles of the tumors, the researchers found the R0 group exhibited an increased number of infiltrated T cells, as compared to the NACT subgroups. The NACT subgroups, by contrast, had increased macrophages and B cells, as compared to the R0 group.

Meanwhile, R0 subgroup tumors also had lower levels of chromothripsis than the other subgroups. Further, when chromothripsis did occur in those tumors, it was more likely to affect chromosomes 6 and 19. Chromothripsis in NACT subgroup tumors was, by contrast, more likely to affect chromosomes 8 and 17.

These differences between tumor subgroups could potentially guide patient treatment in the future, the researchers said. "[I]f verified, our discoveries in this study — molecular, protein, and immune signatures between various HGSC subgroups — may enhance the prediction of R0 resection and have prognostic and therapeutic implications for patients with HGSC," they wrote in their paper.