In addition to being associated with breast cancer, mutations of the BRCA1 and BRCA2 genes have also been implicated in cases of ovarian cancer. Mutations in these genes hinder the patient's ability to repair double-strand breaks in their DNA via the error-free homologous recombination (HR) pathway. Although this mutation renders the person having it more susceptible to developing cancer, if they do develop that disease, it is susceptible to therapies that cause double strand DNA damage (e.g., platinum-based therapies).
One of the main functions for the enzyme PARP is the repair of single-strand DNA breaks. If these breaks persist until DNA replication is ready to occur, then that process can result in double strand breaks. Inhibitors of this enzyme can prevent the repairs of these single-strand DNA breaks, ultimately causing cell death for those cancer cells lacking access to the HR DNA repair pathway. Additionally, some PARP inhibitors may exert another toxic effect upon cancer cells by preventing the dissociation of the PARP-DNA repair assembly. These PARP-DNA complexes are particularly toxic to cancer cells, as they prevent the process of DNA replication.
Rucaparib, the investigational treatment in this study, is a PARP inhibitor, and on Dec. 19, 2016, the FDA granted it expedited approval for the treatment of confirmed BRCA mutant (BRCAmut) ovarian cancer patients who had undergone at least two prior chemotherapy regimens. This approval was largely based upon the results obtained in the Study 10 (NCT01482715) and ARIEL2 (NCT01891344) clinical trials.
Data from the phase II ARIEL2 trial were presented at the Society of Gynecologic Oncology's (SGO) 2017 Annual Meeting on Women's Cancer. Gottfried Konecny, MD, of UCLA, presented data from the trial for BRCAmut patients (Abstract 1) while Elizabeth Swisher, MD, University of Washington, presented data for patients with BRCA-wild type ovarian cancer (BRCAwt) with various levels of loss of heterozygosity (LOH) (Abstract 7). This use of rucaparib for non-BRCAmut patients is considered off-label, as the current FDA approval is for usage in BRCAmut patients.
ARIEL2 Larger Description
Abstracts 1 & 7
The objective of the phase II ARIEL2 clinical study for BRCAmut (germline or somatic mutant) ovarian cancer patients (including those having primary peritoneal or fallopian tube cancer) was to assess the progression-free survival (PFS) and objective response rate (ORR). Additionally, the effects of platinum sensitivity status, as well as number of prior chemotherapy regimens upon ORR and PFS were also evaluated.
Key inclusion criteria for this study were an ovarian cancer diagnosis with RECIST-measurable disease and an ECOG score of 0-1. Patients were given 600 mg rucaparib orally bid. Treatment was administered until disease progression or discontinuation. Response status was investigator-assessed using RECIST version 1.1 criteria.
The larger ARIEL2 study, which included 493 ovarian cancer patients, was divided into two parts. Part 1 of the study included 206 ovarian cancer patients having BRCAmut (germline or somatic) or BRCAwt status. Key inclusion criteria for this part of the study was that the patient had one or more prior platinum-based lines of chemotherapy with platinum-based therapy as the last line of chemotherapy, as well as had high-grade ovarian cancer (HGOC) that was considered platinum-sensitive. Part 2 of this study included 287 ovarian cancer patients who had undergone 3-4 prior lines of chemotherapy. Additionally, patients included in this part of the study had platinum-sensitive, platinum-resistant, or platinum-refractory disease sensitivity status.
For the purposes of this study, platinum-sensitive HGOC was defined for those patients who had disease progression 6 months or more after their last platinum therapy. Patients who had disease progression less than 6 months after their last platinum therapy and a best response other than progressive disease were classified as having platinum-resistant HGOC. Patients were deemed to have platinum-refractory HGOC if progressive disease was their best response for their last platinum therapy which occurred during or up to 2 months after that treatment.
ARIEL2 Part1/2 for BRCAmut Patients
In the part of the study reported by Konecny, 134 high-grade HGOC patients were included and, of these, 41 had platinum-sensitive disease and were included in Part 1, while the 93 patients in part 2 had 3-4 prior lines of therapy and mixed platinum sensitivity status. For this group of 134 patients, the median age was 60, 64.2 percent had BRCA1 mutations, and 35.8 percent had BRCA2 mutations. Of these mutations, 58.2 percent were germline, 17.2 percent were somatic, and 24.6 percent were of unknown origin.
Among these patients, 76.1 percent had three or more prior lines of chemotherapy and 46.3 percent had three or more prior platinum-based regimens. In these BRCAmut patients, the largest number was deemed to have platinum-sensitive HGOC; 57 of these patients (42.5%) had no intervening therapies, while 14 did (10.4%). Of the remaining patients included, 49 (36.6%) had platinum-resistant HGOC and 14 (10.4%) had platinum-refractory disease.
BRCAmut Patient Results
The overall ORR for patients with platinum-sensitive disease whose last therapy was platinum-based was 70 percent. When this group is broken down by prior lines of therapy, those with only one prior line of therapy had an ORR of 83 percent, those having two prior lines of therapy had an ORR of 86 percent, while for those with three or more prior lines of therapy, that figure was 52 percent. Platinum-sensitive HGOC patients having three or more prior lines of therapy where the last therapy was non-platinum-based had an ORR of 43 percent. Platinum-resistant HGOC patients having three or more prior lines of therapy had an ORR of 25 percent, while those with platinum-refractory disease having three or more previous lines of chemotherapy had an ORR of 0 percent.
Platinum-sensitive HGOC patients whose last therapy was platinum-based had a mean PFS of 12.7 months, while those whose last therapy was non-platinum-based had a mean PFS of 7.4 months. Patients with platinum-resistant disease had a median PFS of 7.3 months, while those who had platinum-refractory HGOC had a mean PFS of 5.0 months.
When comparing the median PFS values for platinum-sensitive HGOC patients with differing types of BRCA mutational status (somatic, germline, or undetermined) whose last therapy was platinum-based, one immediately notes the median values for those having somatic mutations or germline mutations were almost identical (12.7 months vs. 12.8 months).
“In some instances, patients can develop a mutation in their BRCA genes, which restores the HR repair function, thus rendering the cancer more resistant to DNA-damaging agents such as platinum-based therapies,” Swisher explained. “In this study, of the 55 patients having platinum-resistant or platinum-refractory HGOC, eight were shown to have these reversion mutations via either next-generation sequencing of their biopsy or circulating tumor DNA analysis.” For the 47 patients in this category of platinum sensitivity without reversion mutations, the median PFS was 7.3 months, while the eight patients with these mutations had a median PFS of 1.7 months.
ARIEL2 BRCAmut Patients
BRCAmut platinum-sensitive HGOC patients included in this study had the highest ORRs among all patients tested, ranging from 52-86 percent, depending on the number of previous lines of chemotherapy. Those patients with platinum-resistant HGOC having three or more prior lines of therapy had an ORR of 25 percent, while patients with platinum-refractory HGOC had an ORR of 0 percent.
The median PFS for BRCAmut platinum-sensitive HGOC patients was 12.7 months, while those with platinum-resistant and platinum-refractory disease status had median PFS values of 7.3 months and 5.0 months, respectively. The most common treatment-related adverse events (AEs) encountered were nausea, fatigue, vomiting, and anemia. The most common grade 3 or 4 treatment-related AEs were anemia, increased AST/ALT levels, and fatigue.
When discussing these findings, Swisher noted, “Generally speaking, when we see increased resistance to platinum-based therapies, we also see increased resistance to PARP inhibitor treatment as well. The reversion mutations associated with increased platinum resistance are possibly the most common mechanism for PARP inhibitor resistance as well.”
Rucaparib monotherapy is currently being evaluated in a phase III clinical study of relapsed HGOC patients with somatic- or germline-BRCAmut status (ARIEL4; NCT02855944).
ARIEL2 Part1 for BRCAmut/wt Patients
The work presented by Swisher at the 2017 SGO conference included the results for 206 platinum-sensitive HGOC patients included in part 1 of the ARIEL2 study. Importantly, patients having BRCAwt status were included in this part of the study, which were not included in the FDA approval for treatment with rucaparib.
In this portion of the study, the responses of patients with BRCAwt HGOC were also evaluated. The BRCAwt patients were split into two groups for additional comparison. Those with 14 percent or more total genomic LOH were classified as BRCAwt-LOHhigh, while those who had less than 14 percent LOH were deemed BRCAwt-LOHlow. The patients having BRCAmut HGOC in this portion of the study had a median PFS of 12.8 months, while for the BRCAwt-LOHhigh and BRCAwt-LOHlow patients those values were 5.7 months and 5.2 months, respectively. Overall, response rate was 29 percent in BRCAwt-LOHhigh versus 10 percent in BRCAwt-LOHlow.
The methylation statuses of BRCA1 and RAD51C genes were determined using methylation-sensitive PCR. BRCA1-methylated tumors were observed in 12.7 percent of the patients, while RAD51C methylation was noted in 2.4 percent of the cancers. Of the patients having methylated BRCA1, 16 (80%) were considered LOHhigh while four were deemed LOHlow. Of the patients having unmethylated BRCA1, 58 were deemed LOHhigh while 59 were considered LOHlow. There were only four patients having methylated RAD51C HGOC, and they were all classified as LOHhigh. In the group having unmethylated RAD51C HGOC, 70 patients were classified as LOHhigh while 63 were considered LOHlow.
Investigator-assessed RECIST responses were confirmed in 52.4 percent of the patients with methylated BRCA1, 75 percent of those with methylated RAD51C HGOC and 29 percent who were classified as BRCAwt-LOHhigh. The median PFS for patients with methylated BRCA1 HGOC was 7.4 months and 9.5 months for those with methylated RAD51C cancer.
In this study, there were two patients with CDK12 mutants who had long, durable responses to their rucaparib therapy. One, having a somatic F89fs∗3 mutation, had a partial response with a PFS of 16.7 months, while another with a somatic homozygous deletion also had a partial response with a PFS of 29.3 months. Both of these patients were classified as BRCAwt-LOHhigh. “The responses that we saw with these patients and the long PFS intervals they had may indicate that some patients with CDK12 mutations may benefit from rucaparib therapy,” Swisher explained.
At the end of the study, the criteria used to classify patients as LOHhigh or LOHlow were reevaluated. “In our original work, we utilized a cutoff of 14 percent or higher as being LOHhigh; however, after some testing, we were able to get a better separation of the data using a value of 16 percent to define the LOHhigh group,” Swisher said. “When we use the 14 percent figure to define LOHhigh, we obtained a median PFS of 5.7 months for the LOHhigh versus 5.2 months for the LOHlow group; however, when the 16 percent figure was used, the median PFS for the LOHhigh group was 7.3 months while the value for the LOHlow group was 4.7 months.
“In this trial, both BRCA1- and RAD51C-methylated HGOC were associated with LOHhigh and sensitivity to rucaparib,” Swisher noted. “Even in patients with platinum-sensitive HGOC, the loss of BRCA1 methylation was common after platinum chemotherapy. Mutations in CDK12 and other genes relevant to HR may confer some sensitivity to PARP inhibitor therapy.
“Routine sequencing of HGOC biopsies could identify as many as 10-15 percent of cases with somatic mutations and 20 percent with germline mutations that would likely respond to treatment with a PARP inhibitor,” Swisher concluded. Going forward, the revised 16 percent cutoff figure for LOHhigh will be utilized in the phase III ARIEL3 trial (NCT: 01968213).
Richard Simoneaux is a contributing writer.