Rucaparib in ovarian cancer: an update on safety, efficacy and place in therapy

Treatment

Study 10 was a phase I–II trial that evaluated rucaparib in patients with germline BRCA1/2-mutated ovarian carcinoma or other solid tumors. The phase II part enrolled 42 patients with platinum-sensitive, high-grade, predominately serous ovarian carcinoma associated with a germline BRCA1/2 mutation who received two to four prior regimens and had a progression-free interval of 6 months or more following their most recent platinum therapy. The larger proportion of patients had a BRCA1 mutation (71.4%), and BRCA2 mutation was seen in 28.6% of patients. The investigator-assessed objective response rate (ORR) by Response Evaluation Criteria in Solid Tumors (RECIST) was 59.5% and the median duration of response was 7.8 months [95% confidence interval (CI), 5.6–10.5].²⁵

ARIEL 2, a two-part phase II trial²⁶ was conducted to assess the safety and efficacy of rucaparib in patients with platinum-sensitive, high-grade ovarian cancer patients with one or more chemotherapy regimen (part 1) or three or four prior chemotherapy regimens (part 2; ClinicalTrials.gov identifier: NCT01891344).³⁰

In part 1, a total of 204 patients were enrolled. The primary endpoint was PFS and secondary endpoints were ORR, duration of response, safety and pharmacokinetics. Rucaparib was given orally at 600 mg twice per day for continuous 28-day cycles until disease progression or any other reason for discontinuation. Tumor samples were analyzed to identify HRD. The biomarker chosen for HRD was the genomic loss of heterozygosity (LOH), and the prespecified (prospectively defined) cutoff to define LOH as high was ⩾14%. Based on HRD, patients were classified in three subgroups: ‘BRCA-mutant’ [deleterious germline (capped at N = 15 patients) or somatic], ‘BRCA-wildtype and LOH high’ (LOH high group), or ‘BRCA wildtype and LOH low’ (LOH low group).

Of 204 patients, 192 were classified into three groups based on HRD status: BRCA-mutant (n = 40/20.8%), LOH high group (n = 82/42.8%) and LOH low group (n = 70/36.4%). The median PFS (months; 95% CI) was: BRCA-mutant (12.8; 9.0–14.7), LOH high group (5.7; 5.3–7.6) and LOH low group (5.2; 3.6–5.5). PFS was significantly longer in the BRCA-mutant subgroup (hazard ratio 0.27, 95% CI 0.16–0.44, p < 0.0001) and LOH high subgroup (hazard ratio 0.62, 0.42–0.90, p = 0.011) than in the LOH low subgroup. The ORR by RECIST were: BRCA-mutant (80%), LOH high (29%) and LOH low (10%). The proportion of patients who achieved a response was similar irrespective of whether the BRCA mutation was germline or somatic or whether a patient had a BRCA1 or BRCA2 mutation. The median duration of response (months; 95% CI) was longer in the BRCA-mutant (9.2; 6.4–12.9) and LOH high (10.8; 5.7–NR) groups compared with the LOH low group (5.6; 4.6–8.5).

The ARIEL2 part 2 included high grade serous ovarian cancer (HGSOC) platinum-sensitive, platinum-resistant and platinum-refractory patients. A secondary analysis of the population of ARIEL 2 part 2 with germline or somatic BRCA mutation was performed to assess ORR, disease control rate and PFS and to determine the effect of platinum sensitivity status and prior lines of chemotherapy on these endpoints. A total of 134 patients were stratified in four groups: platinum-sensitive with platinum as immediate prior treatment (n = 57/42.5%), platinum-sensitive with nonplatinum therapy as an immediate prior treatment (n = 14/10.4%), platinum-resistant (n = 49/36.6%) and platinum-refractory (n = 14/10.4%). There was no difference in PFS between germline and somatic BRCA mutations in the platinum-sensitive with platinum therapy as an immediate therapy: 12.8 and 12.7 months, respectively. The results are summarized in Table 1.³¹ The genomic molecular signature established in part 1 will be prospectively applied to part 2. Completed part 2 data are not yet available.

In a review of the safety and efficacy of rucaparib, the US Food and Drug Administration (FDA) approved rucaparib to treat women with advanced ovarian cancer who have already been treated with at least two chemotherapies and have a BRCA1 or BRCA2 mutation identified by an approved companion diagnostic test. The agency also approved the FoundationFocus CDxBRCA (Cambridge, MA, USA) test to detect BRCA alterations on 19 December 2016 associated with the use of rucaparib.³²

Maintenance

The findings of ARIEL 2 have been further expanded in the phase III ARIEL 3 trial. This was a randomized, double-blind, placebo-controlled trial that included patients with platinum-sensitive high-grade serous or endometrioid ovarian, primary peritoneal, or fallopian tube carcinoma. All patients received at least two previous platinum-based chemotherapy regimens and achieved complete or partial response to their last platinum-based regimen. This trial included 564 patients, who were randomized 2:1 to receive oral rucaparib 600 mg twice daily or placebo in 28-day cycles. Patients were stratified by LOH status, progression-free interval after the penultimate platinum-based regimen, and best response to the most recent platinum-based regimen. In this study, LOH ⩾16% was the discriminant for ‘LOH high’. The statistical analytical plan evaluated three sequential cohorts of patients: BRCA-mutated (germline or somatic), BRCA-wild type with HRD (with LOH high) and intention to treat (ITT) population (BRCA-wild type and low LOH or BRCA-wild type and indeterminate LOH). The step-down procedure evaluated rucaparib (versus placebo) on PFS in the BRCA-mutated group first; if this analysis was statistically significant, the BRCA-wild type LOH high cohort was added and analyzed; if this analysis was significant, the remaining patient cohort was added and analyzed. This latter analysis, if reached, would assess the entire ITT population. A similar procedure was enacted for the patient-reported outcomes (PROs) as assessed by the FOSI-18.²⁷

Of the 564 patients enrolled, 196 patients were BRCA-mutated (130 with germline mutation and 56 with somatic), 158 patients were BRCA-wild type with LOH high and 210 patients were included in the ITT population (161 patients were BRCA-wild type and low LOH and 49 with indeterminate LOH). The median PFS (months; 95% CI) in the BRCA-mutant group was 16.6 (13.4–22.9), compared with placebo with 5.4 months [3.4–6.7, (hazard ratio 0.23, 95% CI 0.16–0.34, p < 0.0001)]. The benefit of rucaparib in the HRD cohort was similar to the BRCA-mutant group, with a PFS of 13.6 (10.9–16.2) versus 5.4 months in the placebo [5.1–5.6; 0.32 (0.24–0.42); p < 0.0001]. In the ITT (BRCA-wild type and low LOH or BRCA-wild type and indeterminate LOH), the PFS was 10.8 (8.3–11.4) versus 5.4 in the placebo group [5.3–5.5; 0.36 (0.30–0.45); p < 0.0001]. All subgroups had a PFS benefit for rucaparib versus placebo, irrespective of volume of disease, response to chemotherapy, status of LOH or BRCA mutation. The overall survival data are not mature at this point. Most patients in ARIEL 3 (n = 374/66%) had achieved an objective response by RECIST prior to enrolment, however 207 (37%) had measurable disease at the time of randomization and were assessable for objective response in a prespecified exploratory analysis. The ORR in the BRCA-mutant group with rucaparib was 38% (95% CI 23–54) versus 9% in the placebo group (95% CI 1–28). The ORR in the HRD cohort was also increased with rucaparib (27%; 95% CI 23–54) compared with the placebo group (7%, 95% CI 1–20). Even in the ITT population, there was a benefit in ORR with rucaparib (18%, 95% CI 12–16) versus placebo (8%, 95% CI 2–17); however, the majority of responses were accounted for in the previous two subsets. No differences in PROs were discovered between the treatment cohorts in the BRCA-mutant group so no further analysis was undertaken.

ARIEL 3 trial validated the next generation sequence HRD assay used in the previous trial. An additional exploratory endpoint was done to assess outcomes in women whose tumors had mutations in non-BRCA HRR genes. Overall, deleterious mutations were detected in 43/564 (7.6%) patients. In the rucaparib group with a non-BRCA HRD gene mutation (n = 28), the most common gene mutations were RAD51C (n = 6), RAD51D (n = 4), and RAD54L (n = 3). Patients with RAD51C/D-mutant were associated with rucaparib sensitivity, with only 2 of 10 patients having disease progression and 7 having a PFS duration of ⩾1 year (median PFS, 16.4 months; range, 5.4–30.4 months).³³

Use of a HRD score to predict efficacy of PARP inhibitors

Clinical data with PARP inhibitors indicate that there is an ovarian cancer patient population beyond those with germline BRCA mutations that may benefit from treatment with a PARP inhibitor.³⁴ Deficiency in HR DNA repair leads to loss or duplication of chromosomal regions. Next generation sequencing assays can quantify this genomic instability by measuring the percentage of genome-wide allelic imbalance or LOH, as a surrogate marker for HRD.^35,36 In the phase II and III trials of rucaparib in epithelial ovarian cancer, the HRD signature was chosen based on an association between the extent of LOH in tumor’s samples and the clinical benefit from rucaparib treatment.^26,27 Tumor was evaluated for HRR mutation and LOH using the Foundation Medicine T5 next generation sequence assay (Cambridge). One of the main advantages of detecting tumor genomic LOH is that it can identify HRD-positive tumors regardless of the underlying mechanisms, which include both known (i.e. BRCA mutations) and unknown genetic and other mechanisms.²³ In the ARIEL2 part 1 trial [ClinicalTrials.gov identifier: NCT01891344], the prespecified cutoff to define LOH high was of 14% or higher.²⁶ Review of these data enabled optimization of the degree of LOH associated with clinical efficacy and were prospectively defined in ARIEL 3 [ClinicalTrials.gov identifier: NCT01968213] as ⩾16%.²⁷

However, even in the ITT population of ARIEL 3, a benefit with rucaparib in PFS was seen. The limitation of using a genomic scars it that past signatures may be present and confound the interpretation of clinical results if the tumor became resistant.³⁷

Other measures of HRD have been evaluated in prospective trials, including that assessed in the phase III switch maintenance trial of niraparib in women with platinum-sensitive recurrent ovarian cancer response to induction platinum combination therapy (NOVA trial, ClinicalTrials.gov identifier: NCT01847274).³⁸ In that study, HRD was defined using a composite of factors associated with genomic instability, including LOH, telomeric allelic imbalance and large-scale state transitions (Myriad MyChoice, Salt Lake City, UT, USA).³⁸ Of note, in the NOVA trial, patients with nongermline BRCA mutation were randomized in a cohort that included HRD (defined as somatic BRCA mutation and nongermline BRCA mutation-associated HRD) and non-HRD genomic assessment. In this study niraparib demonstrated significant improvement over placebo among patients with HRD as assessed by this test. However, a lesser but statistically significant benefit was seen in the HRD-negative cohort suggesting that these biomarkers may not be sufficiently precise to predict absence of benefit on an individual basis.³⁸

The optimal method for the identification of which BRCA wildtype cancers are most likely to respond to a PARP inhibitor is still unknown. However, efforts to validate an HRD tool as a predictive biomarker with high sensitivity and specificity is a highly desired goal for future studies. Given the dynamic alterations that occur in the tumor microenvironment in response to intrinsic and extrinsic stress,³⁹ real-time surveillance tools are likely to be necessary to help clinicians in directing treatment decisions.

Safety and tolerability of rucaparib

Several different PARP inhibitors have been approved in the last few years and although long-term follow up is required for a precise assessment of safety, rucaparib appears to have a manageable toxicity profile. In the phase II part of Study 10, the most common treatment-emergent adverse events (AEs) of grades 3 or 4 were anemia (38.1%), asthenia/fatigue (26.2%) and alanine aminotransferase or aspartate transferase elevations (14.3%). The grade 3 or 4 AEs were managed with treatment modification or supportive care. In the phase 2 ARIEL2 trial, rucaparib displayed a satisfactory safety profile and toxicity similar to Study 10.²⁶ All patients had at least one treatment-emergent AE and the most frequent grade 3 or greater AEs were anemia or decreased hemoglobin in 22% of patients and elevations in alanine aminotransferase or aspartate aminotransferase in 12% of patients. An integrated analysis of the antitumor activity and safety of Study 10 and ARIEL2 trials revealed that rucaparib has a manageable safety profile.⁴⁰

No new safety signals appeared in the much larger ARIEL 3 trial.²⁷ Aligned with previous experience and noted as probable class effects, the most common AEs were: nausea (75%), asthenia/fatigue (69%), constipation (37%) and vomiting (37%). Grade 3 or 4 AEs were seen in 56% of patients in the rucaparib group. AEs leading to dose reduction were observed in 55% and 4% of patients in the rucaparib and placebo group, respectively. Treatment interruption due to a treatment-emergent AE was common and occurred in 64% of patients in the rucaparib group and 10% in the placebo group. However, treatment discontinuation due to a treatment-emergent AE was far lower (13%) and compared with 2% in the placebo group. The most common serious AEs were anemia (4% of patients), pyrexia (2%), vomiting (2%) and small intestinal obstruction (1%).

Considering laboratory abnormalities, decreased hemoglobin concentration (anemia) was the most common AE, occurring in any grade in 37% of patients and in 19% grade 3 or greater. A decline in hemoglobin concentration occurred generally in the first few cycles. Compared with other PARP inhibitors, anemia is a common side effect for this class of drug and it was noticed as grade 3 or greater in 19% of patient using olaparib⁴¹ and in 25.3% of patients using niraparib.³⁸ Another AE noticed in ARIEL 3 that was seen less frequently with other PARP inhibitors was an increase in alanine aminotransferase or aspartate aminotransferase concentration (grade 3–4: 34% of rucaparib-treated patients versus 10% with placebo). These transaminase alterations were transient, self-limiting, and not associated with any parameters of hepatic dysfunction. Grade 1 or 2 creatinine increases were observed within the first few weeks of rucaparib treatment and then stabilized with continued rucaparib treatment.

Rucaparib and others PARP inhibitors

Others PARP inhibitors have also been approved for the treatment of ovarian cancer. Olaparib was the first PARP inhibitor approved by the US FDA. In the US, olaparib was first approved as monotherapy with the capsule formulation for patients with high-grade ovarian carcinoma and germline BRCA1 or 2 mutation who have received three or more prior chemotherapies on 19 December 2014.¹⁶ The SOLO-2 trial was a phase III study that evaluated olaparib as a maintenance treatment of patients with high-grade serous epithelial ovarian, fallopian tube, or primary peritoneal cancer with platinum-sensitive, relapsed, BRCA-mutated who had received at least two lines of previous chemotherapy and had an objective response to platinum-based chemotherapy. In this study, the investigator-assessed median PFS was significantly longer in the olaparib arm [19.1 months (95% CI 16.3–25.7)] than with placebo [5.5 months (5.2–5.8); hazard ratio 0.30 (95% CI 0.22–0.41), p < 0.0001].⁴¹ The results allowed the approval of olaparib as maintenance therapy on 17 August 2017 in the US and on 16 December 2014 in the European Union. The first agent to receive US FDA approval as a maintenance therapy in platinum-sensitive recurrent ovarian cancer patients was niraparib (27 March 2017) based on results from the NOVA trial. This phase III, double-blinded, placebo-controlled randomized trial of niraparib monotherapy maintenance showed an increase in PFS in the germline BRCA mutation group of 21 months with niraparib versus 5.5 months in the placebo group (hazard ratio 0.27, 95% CI 0.17–0.41). The non-gBRCA cohort with HRD had also a benefit in PFS of 12.9 months versus 3.8 (hazard ratio 0.38, 95% CI 0.24–0.59) as the overall non-gBRCA cohort of 9.3 months versus 3.9 months (hazard ratio 0.45, 95% CI, 0.34–0.61).³⁸

Veliparib is a PARP inhibitor that is not approved but has been evaluated in a phase II trial for the treatment of persistent or recurrent epithelial ovarian, fallopian tube, or primary peritoneal cancer in patients who carry a germline BRCA1 or BRCA2 mutation. The ORR in platinum-resistant and platinum-sensitive patients was 20% and 35%, respectively.⁴² An ongoing placebo-controlled phase III trial in newly diagnosed ovarian cancer patients, which is combining veliparib/placebo and chemotherapy, followed by veliparib/placebo maintenance has recently completed enrolment [ClinicalTrials.gov identifier: NCT 02470585].

Currently, there are no head-to-head trials of PARP inhibitors in the setting of high-grade recurrent platinum-sensitive ovarian cancer. However, it is important to recognize that despite the fairly homogeneous outcomes across these three phase III trials, several design, population, and assessment differences exist. These are summarized in Table 2.

There are still many unanswered questions regarding the use of PARP inhibitors in ovarian cancer, such as the optimal timing and duration of administration of these drugs, the long-term effects of PARP inhibition, how PARP inhibitor resistance develops, whether PARP inhibitors can be used again if previously administered, whether there are combinations that can overcome HR resistance (innate or induced), whether PARP inhibitor response can be augmented in cohorts where they are already known to be clinically active, and how somatic events can be evaluated and addressed in ‘real-time’. The answers to these questions will have a profound impact on the broader clinical utility of this class of agent in patients with solid tumors. A summary of the results of single-agent trials of PARP inhibitors in ovarian cancer is presented in Table 3.