Tumor Gene Methylation May Predict Ovarian Cancer Response to PARP Inhibitor


Hypermethylation of 2 wild-type tumor-associated genes increased ovarian cancer responsiveness to the PARP inhibitor rucaparib, according to a subgroup analysis of the randomized ARIEL2 trial presented at the 2017 Society of Gynecologic Oncology Annual Meeting on Women’s Cancer.

Elizabeth Swisher, MD

Hypermethylation of 2 wild-type tumor-associated genes increased ovarian cancer responsiveness to the PARP inhibitor rucaparib (Rubraca), according to a subgroup analysis of the randomized ARIEL2 trial presented at the 2017 Society of Gynecologic Oncology Annual Meeting on Women’s Cancer.

Objective responses with rucaparib occurred in more than half of patients whose tumors exhibitedBRCA1promoter methylation and in 3 of 4 patients withRAD51Cpromoter methylation. Methylation of either gene was mutually exclusive ofBRCAmutations or other homologous recombination genes.BRCA1andRAD51Cpromoter methylation correlated strongly with loss of heterozygosity (LOH), which is associated with prolonged progression-free survival (PFS).

“BRCA1andRAD51Cmethylation in ovarian carcinomas are associated with high loss of heterozygosity and sensitivity to rucaparib,” said senior study author Elizabeth Swisher, MD, a medical oncologist at the University of Washington. “Loss ofBRCA1methylation is common after exposure to platinum chemotherapy, even in platinum-sensitive patients. If methylation were to be used as a predictor of PARP inhibitor sensitivity, it would need to be assessed in a pretreatment, rather than archived specimen.”

“Routine sequencing of high-grade ovarian cancer would identify at least 10% to 15% of cases with somatic mutations and 20% with germline mutations that are likely to respond to PARP inhibition,” she added.

These findings stemmed from the subgroup analysis of the phase II ARIEL2 trial, which evaluated rucaparib in patients with relapsed, platinum-sensitive high-grade ovarian cancer. Swisher and colleagues sought to determine whetherBRCA1andRAD51Cpromoter hypermethylation might potentially correlate with markers of homologous recombination deficiency and response to rucaparib.

The analysis had its genesis in several lines of evidence. Promoter-region hypermethylation is associated with gene downregulation, andBRCA1andRAD51Cpromoter hypermethylation has been associated with decreased gene expression in ovarian cancer. However, attenuated gene expression in ovarian cancer was not associated with improved survival, according to data from The Cancer Genome Atlas (TCGA).

Results of the ARIEL2 trial showed improved progression-free survival (PFS) inBRCA-mutant andBRCAwild-type/high LOH disease as compared withBRCAwild-type/low LOH. Additionally, patients with ovarian cancer with damaging mutations in other homologous recombination genes responded to rucaparib. However, some patients with mutated homologous recombination genes did not respond to the PARP inhibitor.

In the ARIEL2 analysis Swisher reported at the SGO meeting, there were 165 participants who had evaluable tumor samples. The results showed that 12.7% of tumors hadBRCA1promoter methylation and 2.4% hadRAD51Cpromoter methylation, and the promoter methylation was mutually exclusive ofBRCA1mutation and other homologous recombination genes (P= .015). All 4 tumors with methylatedRAD51Cpromoter and 16 of 20 evaluable specimens with methylatedBRCA1promoter had high LOH (80% association overall).

BRCA1andRAD51Cmethylation was assessed in 90 and 99 paired archival and pretreatment biopsy specimens, respectively. Of 77 cases withoutBRCA1methylation in archival tissue, one exhibited methylation in a pretreatment biopsy specimen. Of 13 cases withBRCA1methylation in archival specimens, 4 were unmethylated in pretreatment biopsy specimens.RAD51Cmethylation was concordant between archival and pretreatment biopsy, but the sample included only 2RAD51Cmethylated cancers.

Response data showed investigator-assessed responses to rucaparib in 11 of 21 (52.4%) of theBRCA-methylated cases, 3 of 4 (75%)RAD51C-methylated cases, and 29% ofBRCAwild type/LOH-high cases. The median duration of response was 6.1 months forBRCA1methylated cases and 9.5 months forRAD51Cmethylated cases.

Additionally, 2 patients withCDK12-mutated tumors had prolonged responses to rucaparib.CDK12is involved in regulation of RNA splicing, and its loss leads to downregulation of many DNA repair genes and possibly homologous repair deficiency, said Swisher.

CDK12mutations occur in approximately 3% of cases of high-grade serous ovarian cancers, according to data from TCGA. Approximately 56% of high-grade serous ovarian cancers are not associated with homologous repair deficiency, suggesting that a number of other genetic/molecular alterations could result in homologous repair deficiency in the disease, said Swisher.

The ARIEL2 trial had a cutoff for high LOH of 14%, which was associated with a 42% reduction in the hazard for progression or death among patients withBRCAwild-type tumors treated with rucaparib (P= .0027). An analysis of PFS curves suggested an optimal cutoff of £6%, which will be evaluated in the phase III trial.

In December 2016, the FDA granted an accelerated approval to rucaparib as a treatment for patients withBRCA-positive advanced ovarian cancer who have received at least 2 prior lines of chemotherapy.


Swisher EM, Harrell MI, Lin K, et al. BRCA1 and RAD51C promoter hypermethylation confer sensitivity to the PARP inhibitor rucaparib in patients with relapsed, platinum-sensitive ovarian carcinoma in ARIEL2 Part 1. In: Proceedings from the Society for Gynecologic Oncology Annual Meeting on Women’s Cancer; March 12-14, 2017; National Harbor, MD.

Related Videos
Related Content