NEW YORK – Data from early research suggests that oncologists should be on the lookout for metastatic prostate cancer patients who have both BRCA1/2 mutations and high microsatellite instability, because they are likely to fare well on a checkpoint inhibitor and not respond to a PARP inhibitor.
An analysis of about 1,000 samples from prostate cancer patients, published earlier this month in JCO Precision Oncology, suggested that about 16 percent of patients have BRCA1/2 mutations and are MSI-high. Initial research on these samples suggests that the BRCA1/2 alterations they harbor may just be bystander mutations, rather than driver mutations, and therefore, doctors should use the co-occurring MSI-high status to direct therapy.
After Emmanuel Antonarakis, senior author of the paper and a genitourinary oncologist, treated two metastatic prostate cancer patients at the University of Minnesota Masonic Cancer Center who were both positive for BRCA1/2 mutations and MSI-high, he decided to research the frequency of this co-occurrence in other prostate cancer patients. There was a clinical urgency to this research, since these two patients did not respond to treatment with AstraZeneca's PARP inhibitor Lynparza (olaparib), which is approved in the US for metastatic castration-resistant prostate cancer patients who have progressed on anti-androgen drugs and have mutations in genes that cause homologous recombination repair deficiency (HRD), such as BRCA1/2 mutations.
On the other hand, patients with advanced or refractory solid tumors that are MSI-high can also receive checkpoint inhibitors, such as Merck's Keytruda (pembrolizumab).
"When I was confronted with a clinical decision for these patients, I was not sure which part to believe more, the BRCA1/2 inactivation or the microsatellite instability, and I didn't know which one was driving the patient's cancer," said Antonarakis, a professor of medicine at the University of Minnesota Medical School.
Given his two choices, a PARP inhibitor or a checkpoint inhibitor, Antonarakis went with the former. "In both cases, the response was not favorable," he recalled.
But when he gave these patients a PD-1 inhibitor after they failed on the PARP inhibitor, both responded. "That led me to the question: How often does this occur? And does this occur in BRCA-associated cancers other than prostate cancer?" Antonarakis said.
The researchers analyzed comprehensive genomic profiling data from more than 213,000 tumor samples. They homed in on the tumor types that have approved PARP inhibitor indications, including prostate, breast, ovarian, and pancreatic cancers harboring a BRCA1/2 mutation or other genetic mutations associated with HRD.
They found little overlap of BRCA1/2 mutations and MSI-high status among breast and ovarian cancers, with less than 2 percent of tumors with co-occurence. In pancreatic cancer, the overlap was slightly higher, with 1.4 percent of BRCA1-mutant and 3.8 percent BRCA2-mutant tumors also having MSI-high status.
However, in prostate cancer these biomarkers appeared to co-occur much more frequently, with 12.8 percent of BRCA1 and 3.4 percent of BRCA2 alterations showing up in tumors that are also MSI-high. In a further look at 14 genes related to HRD, including PALB2, ATM, CDK12, RAD51, and others, 46.3 percent of MSI-high prostate cancer samples also harbored at least one HRD gene mutation.
The researchers also examined the association of MSI-high status and tumor mutation burden (TMB). They found that microsatellite instability "leads to the accumulation of numerous genome-wide mutations, particularly insertion/deletion events, and to a high TMB," they wrote, concluding that the prevalence of BRCA1/2 mutations in MSI-high tumors were bystander mutations, a product of microsatellite instability and high TMB, rather than driver mutations.
"When a cancer has a deficiency in mismatch repair or microsatellite instability, the consequence of that is that those cancers develop hundreds or even thousands of mutations throughout the genome," Antonarakis said. "These mutations occur randomly, so one of the downstream genes that could be mutated as a consequence of microsatellite instability might happen to be BRCA1 or BRCA2."
That means that MSI-high tumors can also have mutated BRCA1/2 genes, but Antonarakis noted that "just defining a mutation as pathogenic doesn't tell you if it's a driver mutation or a passenger or bystander mutation."
Antonarakis and colleagues also found the BRCA1/2 mutations co-occurring with MSI-high and high TMB tumors were monoallelic, or expressed on only one of the gene copies, versus biallelic, or expressed in both copies of the gene. These monoallelic mutations were also associated with low genome-wide loss-of-heterozygosity scores compared to biallelic BRCA1/2 mutations, suggesting that they do not result in HRD, which is important for PARP inhibitor response.
There are two PARP inhibitors, Lynparza and Clovis Oncology's Rubraca (rucaparib), approved in the US for prostate cancer patients with mutations associated with HRD.
In a statement, an AstraZeneca spokesperson said that Lynparza's FDA approval in prostate cancer was based on results from the Phase III PROfound trial, "which did not specifically evaluate patients with co-occurring MSI-high and BRCA-mutated metastatic castration-resistant prostate cancer." The spokesperson did not comment further on the potential implications of the latest study on Lynparza's prostate cancer indication. Clovis did not respond to a request for comment about the implications of the study on Rubraca's use before press time.
Meanwhile, Neeraj Agarwal, director of the genitourinary oncology program at the University of Utah's Huntsman Cancer Institute, noted that these study findings could be "practice influencing."
"The article suggests we should be using immune checkpoint inhibitors in these patients as the preferred line of therapy," said Agarwal, who was not involved with the study. "However, the lack of sensitivity or lack of efficacy in these patients with PARP inhibitors would ideally need to be validated prospectively."
Agarwal noted that previous research has shown that about half of metastatic prostate cancer patients do not receive a subsequent line of therapy. Therefore, choosing the most appropriate initial treatment for these advanced prostate cancer patients is critical to their overall survival.
"These are elderly patients, and once they progress on one treatment, there's approximately a 50 percent probability that they will not see subsequent lines of therapy," Agarwal said. "In a way, by choosing a wrong therapy upfront, we're potentially depriving them of a therapy that could have been more effective for them."
Although few metastatic prostate cancers tend to be MSI-high, occurring in only about 3 percent of patients, according to Agarwal, these patients often respond well to checkpoint inhibitors such as Keytruda, just like the two MSI-high prostate cancer patients Antonarakis treated with immunotherapy after they failed to respond to a PARP inhibitor. Agarwal and colleagues tested this hypothesis out in a separate small study published in 2020, and more than half of MSI-high metastatic prostate cancer patients prescribed Keytruda responded.
Antonarakis noted that in light of these findings, oncologists should look carefully at their prostate cancer patients' genomic sequencing results to see whether patients have both a BRCA1/2 mutation co-occurring with MSI-high status. Commonly used next-generation sequencing panels will report out both patients' MSI and BRCA mutation status for prostate cancer patients. Ideally, it would be helpful, he suggested, if genomic profiling reports have a note to make oncologists aware that these alterations can co-occur.
Based on his group's initial findings, Antonarakis is now putting together a consortium of cancer institutions to study the efficacy of PARP inhibitors in MSI-high and BRCA-mutated prostate cancer patients at a larger scale.
"It's still in preliminary stages, but I'm hoping to get 20 to 50 institutions to pull together their data from these patients and then to try to report clinical outcomes," he said. "We encouraged the scientific community in our paper to try to develop a national or international consortium to capture patients that have both, and then to try to report their clinical outcomes. Because the prevalence of the combination is small, I don't think this could be done from one cancer center or one company."