NEW YORK – Data from two randomized trials have demonstrated the frontline benefit of PARP inhibitors combined with steroids and hormone therapy for metastatic castration-resistant prostate cancer, but questions remain as to the degree of benefit in biomarker-defined patient subgroups.
During the American Society of Clinical Oncology's Genitourinary Cancers Symposium on Thursday, back-to-back presentations focused on two Phase III trials of competing PARP inhibitors: PROpel involving AstraZeneca/Merck's Lynparza (olaparib) and MAGNITUDE involving Zejula (niraparib). (Of note, Janssen bought the rights to develop niraparib for prostate cancer in 2016, and prefers to use its scientific name for the investigational prostate cancer indication to differentiate it from GlaxoSmithKline's approved ovarian cancer indications).
Both PROpel and MAGNITUDE compared the activity of their PARP inhibitors, the hormone therapy abiraterone acetate, and the steroid prednisone against just hormone therapy and prednisone in first-line mCRPC patients. Both studies set primary endpoints of radiographic progression-free survival, and in both studies, patients benefited from the addition of the PARP inhibitor.
There was, however, a key difference in the findings when it came down to biomarker stratification. In Merck and AstraZeneca's study, the addition of Lynparza to the treatment combination benefited patients regardless of their homologous recombination repair, or HRR, gene mutation status, whereas in Janssen's study, the addition of niraparib only benefited patients with these mutations.
In PROpel, 399 patients with and without HRR gene mutations treated with the Lynparza-containing combination lived for a median of 24.8 months without their cancer progressing versus a median of 16.6 months among 397 patients treated with abiraterone and prednisone alone. The addition of Lynparza translated to a 34 percent reduction in the risk of disease progression or death.
When Fred Saad, director of urologic oncology at the University of Montreal Hospital Center, broke down PROpel's findings by HRR gene mutation status, they found a 24 percent reduction in the risk of disease progression or death with the Lynparza combo specifically for the HRR non-mutated subgroup and a 46 percent reduction in the risk of disease progression or death with the Lynparza regimen for those who did indeed have an HRR gene mutation.
Meanwhile, Janssen's MAGNITUDE study was designed a bit differently. Kim Chi, the associate director of clinical research at the University of British Columbia's Vancouver Prostate Center, reported that in a preplanned futility analysis of 233 patients without HRR gene mutations, there was no progression-free survival benefit at all of added niraparib. Meanwhile, in a subgroup of patients harboring HRR gene mutations — which included mutations in ATM, BRCA1, BRCA2, BRIP1, CDK12, CHEK2, FANCA, HDAC2, or PALB2 — 212 patients treated with the niraparib combination lived a median of 16.5 months without their disease progressing, versus 13.7 months among those treated with just the hormone therapy and steroid. This translated to a 27 percent reduction in the risk of disease progression or death in patients who received the niraparib combo.
In another, narrower subgroup analysis, MAGNITUDE study investigators found that patients harboring BRCA1 or BRCA2 mutations benefited to a greater extent than the full group of patients with HRR gene mutations. Among patients harboring BRCA1/2 mutations, the niraparib-containing regimen reduced patients' risk of disease progression or death by 47 percent. In this subgroup, patients lived a median of 16.6 months with the niraparib combination versus 10.9 months on just abiraterone and prednisone.
Both studies have yet to report overall survival data but are trending toward showing added benefit with the PARP inhibitor combinations. In PROpel this benefit was seen in patients irrespective of HRR biomarkers, and in MAGNITUDE the benefit was in the biomarker-defined patient populations.
Biomarker controversies
In prostate cancer, as with other PARP inhibitor-treated cancers, the efficacy of drugs like Lynparza and niraparib has been historically intertwined with patients' HRR biomarker status. PARP inhibitors work by preventing one of the paths that cancer cells can take to repair DNA breaks, and these drugs work particularly well against cancer cells harboring mutations that hobble their ability to repair damaged DNA. These biomarkers remain a key consideration for oncologists prescribing these therapies even as drugmakers have increasingly pursued regulatory approvals for broader indications that don't require biomarker testing.
Several PARP inhibitors are already commercially available for mCRPC in the later-line treatment setting, with varying biomarker eligibility requirements. Lynparza, for instance, is US Food and Drug Administration-approved for patients with germline or somatic HRR gene-mutated mCRPC who have progressed after prior hormone therapy. With this indication, of note, there were lingering questions as to whether the biomarker-defined indication was too broad, and several regulators abroad, including in Europe and China, subsequently approved the agent specifically for BRCA1/2 mutated mCRPC in the treatment refractory setting.
Clovis Oncology's Rubraca (rucaparib), meanwhile, is FDA-approved as a third-line treatment for mCRPC patients with germline or somatic BRCA1/2 mutations, and Clovis is currently evaluating the drug in the second-line treatment setting in patients with either BRCA1/2 or ATM mutations.
Janssen's niraparib does not have an FDA-approved prostate cancer indication at this time, despite having received breakthrough therapy designation for the treatment in a later-line setting based on the ongoing single-arm GALAHAD trial. Even though that non-randomized Phase II study showed benefit, Mark Wildgust, Janssen's VP of global oncology medical affairs, said that the firm has shifted away from the already crowded later-line mCRPC space, where regulators would likely require a head-to-head comparison of niraparib against currently available options. Instead, the company is focusing on the first-line treatment setting now.
Lynparza, meanwhile, already has a leg up on niraparib in the mCRPC space and stands to clinch an even greater lead should AstraZeneca and Merck pursue a biomarker-agnostic indication in the first-line setting — a tantalizing prospect for the drug's sponsors. In a statement, Roy Baynes, Merk's senior VP and head of global clinical development, said the firm "looks forward to discussing these important results with global health authorities as quickly as possible."
Scot Ebbinghaus, Merck's VP of oncology clinical research, said that the PROpel data do support a regulatory filing in an all-comer, first-line mCRPC population. If AstraZeneca and Merck do pursue this broader indication, however, they will still have to contend with uncertainty in the minds of oncologists and regulators about Lynparza's benefit in the non-HRR gene-mutated patient population.
Unpacking discrepancies
Given that Lynparza and niraparib are both PARP inhibitors and have the same fundamental mechanism of action, it is worth asking why one drug would benefit patients regardless of their biomarker status while the other treatment's efficacy would be limited to the biomarker subgroups. Possible explanations for the conflicting results in the HRR mutation-negative group have to do with diverging biomarker testing approaches.
In the MAGNITUDE study, Janssen investigators prospectively tested patients for HRR gene mutations using tumor tissue, blood, and saliva.
"We really wanted to make sure that patients were indeed HRR-positive or HRR-negative," Janssen's Wildgust said, explaining that patients who were HRR gene mutation-negative according to circulating tumor DNA in plasma samples were confirmed using tissue sequencing. "We know that … almost a third of patients who are tested with ctDNA and are HRR mutation-negative are actually positive for DNA repair gene alterations."
As such, testing for these gene mutations using tissue, saliva, plasma, and whole blood helped rule out missed HRR gene alterations in the MAGNITUDE study.
Moreover, PARP inhibitors have been shown to benefit only biomarker-defined patients in the later-line mCRPC setting when used as a single agent. According to Wildgust, Janssen built this extra biomarker confirmation into its study wanting to home in on the key question of whether PARP inhibitors have synthetic lethality when combined with abiraterone, making the combination beneficial regardless of biomarker status.
"To really answer that question and test that hypothesis, we needed to be very clear that those patients who were negative were indeed negative and those patients who were positive were indeed positive," he said. "And the answer is that … in those patients who we confirmed are HRR mutation-negative, there is no clinical efficacy benefit in using niraparib with [abiraterone]."
Because adding the PARP inhibitor brings additional toxicity, both physical and financial, Wildgust said ruling out this benefit in the biomarker-negative subgroup was valuable from a risk-benefit perspective. He questioned how HRR status was confirmed in other studies, such as PROpel. "And that's a question that those other companies and the physicians involved in those studies should answer," Wildgust said.
Merck assessed HRR mutation status in PROpel via either ctDNA or tissue testing, and felt that this had important real-world implications, since not all prostate cancer patients are candidates for tissue biopsy.
"We know from prior experience with the [later-line] PROfound trial that there's a sizable minority of patients that aren't going to have tumor tissue available for testing," Ebbinghaus said. "Being able to test for HRR mutations in blood is something that we've been keenly interested in."
Additionally, Ebbinghaus said he is not all too concerned about the potential for missed HRR gene mutations in ctDNA-tested patients. "Obviously, no test is ever perfect, but it seems to me implausible that missed gene mutations are going to swing the outcome," he said, arguing that the data do point to a synthetic lethality effect of combining Lynparza with abiraterone. "There's basic science behind combining these two."
In a discussion of the data presented Thursday, Saad also seemed eager to brush off the missed-mutation theory as an explanation for why the Lynparza regimen benefited the HRR mutation-negative subgroup. He emphasized that the percentage of HRR gene mutation-positive patients in PROpel regardless of testing method — 27.8 percent in the Lynparza arm and 29 percent in the comparator arm — aligned closely with what other studies have shown, making it unlikely that there was any significant sway from missed mutations.
"Even if we used only tissue [to screen for HRR mutation status], that's what we would have expected," he said. "Every single analysis that has been done so far has always turned out to be in the range of 25 to 30 percent." He acknowledged there is the potential to miss some mutation-positive patients with ctDNA analysis but maintained that it wouldn't make much difference in the end result.
In a discussion following the PROpel and MAGNITUDE presentations, Celestia Higano, an oncologist in the department of urologic sciences at the University of British Columbia, adamantly cautioned against trying to compare Lynparza and niraparib head to head because the study designs were so different.
"From a bird's-eye view, these trials appear to be quite similar," she said. "They're both Phase III, they're both studying first-line mCRPC, they're looking at a PARP inhibitor with abiraterone versus placebo and abiraterone, and the primary endpoint in both was radiographic progression-free survival, and overall survival was a secondary endpoint. [But] these apparently similar trials are actually very different."
The main thing setting these trials apart, she argued, is that MAGNITUDE set HRR gene mutation status as part of the eligibility criteria and confirmed the lack of benefit in HRR mutation-negative patients in a preplanned futility analysis, whereas PROpel enrolled all-comers and then looked at HRR gene mutation status retrospectively.
Instead of trying to parse the biomarker differences across two differently designed trials, Higano suggested that oncologists focus on the studies' immediate real-world implications. In light of PROpel, she said it would likely be best to wait for the overall survival data to read out in the full, biomarker-agnostic patient population and not risk the drug and financial toxicities of the combination in a front-line mCRPC population.
Indeed, hematologic toxicities with the Lynparza regimen were more than double that of the abiraterone and prednisone-alone regimen in PROpel, and analyses in other cancer indications have estimated the drug's monthly cost to be well over $7,000.
Meanwhile, in light of the benefit seen in the BRCA1/2-mutated subgroup in the MAGNITUDE study, Higano was willing to "hedge a little bit." Since mCRPC patients with BRCA1/2 mutations are known to have worse prognoses with abiraterone treatment alone, she said, "maybe it would be reasonable to treat patients [with the niraparib combination] even before overall survival [data] is available." In the full HRR gene-mutated population, however, she would still await overall survival data before making a change in clinical practice.