SAN ANTONIO (GenomeWeb) – A study involving Novartis' investigational PI3K inhibitor alpelisib in combination with chemotherapy has shown promising activity in breast cancer patients with PIK3CA mutations detected by tissue-based and circulating tumor DNA tests.
The SOLAR-1 study involved 572 men and women with hormone receptor-positive, HER2-negative advanced breast cancer characterized by PIK3CA mutations. Data presented at the San Antonio Breast Cancer Symposium yesterday showed that breast cancer patients with these mutations in tumor tissue lived nearly twice as long without their disease progressing when treated with the combination of alpelisib plus fulvestrant versus just fulvestrant (11 months versus 5.7 months). Patients without such mutations in tumor tissue had median progression-free survival of 7.4 months on alpelisib and fulvestrant compared to 5.6 months on fulvestrant.
More than 76 percent of patients on the alpelisib arm saw their tumors shrink, compared to 44 percent in the fulvestrant-only arm. At the time of data analysis, the overall survival data were immature. "However, we're seeing an overall survival trend," said Dejan Juric from the Massachusetts General Hospital, presenting data from SOLAR-1 at the meeting.
PIK3CA mutations occur in approximately 40 percent of hormone receptor-positive, advanced breast cancer patients, overactivating the PI3K pathway and driving tumor growth. Patients with these mutations tend to be resistant to endocrine treatment and have poor outcomes, but there are currently no approved treatments in this setting.
"The determination of PIK3CA [mutation] status in the tissue was integral to the conduct of this trial," said Juric.
In addition to tissue-based testing, patients in SOLAR-1 also had their plasma samples collected at baseline for ctDNA analysis by PCR to retrospectively assess PFS according to PIK3CA mutation status. This analysis showed that median PFS was 10.9 months in mutation-positive patients on alpelisib plus fulvestrant — nearly identical to when mutation status was determined is tissue samples — versus 3.7 months on fulvestrant. Meanwhile, in mutation-negative patients by ctDNA analysis, median PFS was 8.8 months versus 7.3 months on alpelisib/fulvestrant and fulvestrant, respectively.
"This analysis clearly illustrates the potential clinical value of this easily accessible biospecimen, allowing for a rapid determination of PIK3CA mutation status right at the time of initiation of therapy, as opposed to relying on distant archival tumor material," Juric said.
There is data suggesting that PIK3CA mutations occuring in different regions of the gene may not impact the activation of PI3K pathway the same way, which may in turn impact patients’ response to the inhibitor. Researchers tested patients for 11 PIK3CA mutations in exons 7, 9, and 20, but there were no mutation-specific differences in terms of response to the alpelisib/fulvestrant combination in this study.
Qiagen developed the tissue and plasma clinical trial assays researchers used to determine PIK3CA mutation status in SOLAR-1. A spokesperson for the company said that both tests operate on a single platform, the RT-PCR-based Rotor-Gene Q MDx, which the US Food and Drug Administration has previously approved to run the Therascreen KRAS, EGFR and JAK2 assays.
Alpelisib is in late-stage development and Novartis has begun discussions with regulators about the drug. In addition to working with Qiagen, Novartis said it will also work with Foundation Medicine to develop a companion test to identify patients who are likely to benefit from the alpelisib and fulvestrant combination.
Qiagen yesterday announced its companion diagnostic partnership with Novartis for alpelisib. Foundation also recently announced it would collaborate with Novartis to develop companion tests on its next-generation sequencing FoundationOne CDx for targeted and immuno-oncology drugs.
"The companion diagnostics being planned to support alpelisib provide multiple options for testing for PIK3CA gene mutations," a Novaris spokesperson said. "We are developing both PCR- and NGS-based CDx, which can either work from tissue or plasma."
PI3K has been a difficult pathway to tackle and even when drugs hit the pathway, they tend to induce a challenging toxicity profile. Earlier this year, for example, Roche decided not to pursue development of its PI3K inhibitor taselisib in metastatic breast cancer after studies showed lackluster efficacy and safety issues. Another Novartis PI3K inhibitor, buparlisib, also ran into toxicity issues and as a result the drugmaker sold the drug off to a Chinese biopharma. While presenting the data on SOLAR-1, Juric noted that alpelisib hits only alpha-isoform of PI3K, and is different from these earlier agents that target multiple isoforms.
The most common adverse event seen in the trial is hyperglycemia, which "is inextricably linked" with on-target activity of alpelisib, Juric said, adding that insulin resistance is also a hallmark of PI3K inhibition. Although hyperglycemia resulted in alpelisib dose adjustments and treatment discontinuations, this can be identified early and managed with glucose monitoring and use of anti-diabetic medicine, such as metformin, he said.
Among the main takeaways from this study, according to Juric, is the consistency of the benefit seen in PIK3CA mutated patients via plasma and tissue testing, which highlights the importance of this biomarker, and "potentially bringing the precision oncology paradigm to breast cancer."
However, there is more work to be done in terms of pinning down the capabilities of the tissue and plasma testing, since there was some discordance between which patients the clinical trial assays deemed PIK3CA mutation positive and negative. Preliminary analysis showed that when patients are negative in tumor tissue for PIK3CA mutations, the blood analysis is also negative in 97 percent of cases, while approximately 60 percent of tissue-positive patients are also mutation-positive in blood.
Achieving concordance with the two assays when determining mutation-positive status is more complicated, Juric explained, since there are more tumor cells in tissue, and blood analysis is less sensitive due to the presence of fewer circulating tumor cells. "You also have the biological component," he said. "Perhaps what you're actually capturing is the proliferation rate and the shedding of disease. We're trying to precisely define these numbers."