2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Researchers remain keenly interested in exploring PIK3CA as an anticancer target in breast cancer.
José Baselga, MD, PhD
Although PIK3CA is among the most commonly mutated genes in human breast cancer, developing drugs that effectively target this aberration has proved challenging. This year’s American Society of Clinical Oncology (ASCO 2018) Annual Meeting featured disappointing results from a phase III study into taselisib, one of the most advanced agents aimed at the mutation.
Nevertheless, researchers remain keenly interested in exploring PIK3CA as an anticancer target.PIK3CA is in the PI3K/AKT/mTOR network, which controls a variety of essential cellular processes such as proliferation, growth, metabolism, and motility.1 This pathway is activated via receptor tyrosine kinase or G protein—coupled receptors to undergo complex intracellular and extracellular interactions, involving other pathways to form a larger network that is not yet fully understood (Figure2).
The pathway is dysregulated in a variety of cancer types, where aberrations promote uncontrolled cell survival and propagation of abnormal cells and downregulate autophagy. The process implicates a number of oncogenes, including PTEN, AKT, MTOR, LKB1, TSC1, TSC2, PIK3R1, and PIK3CA.1 Of these, PIK3CA mutations in breast cancer are associated with both disease progression and endocrine therapy resistance.3
PIK3CA mutations have been observed across the spectrum of subtypes—notably, in hormone receptor—positive disease, which, in turn, makes up 75% of breast cancer cases.4 These mutations have been found in 30% to 40% of estrogen receptor (ER)—positive/HER2-negative, 20% to 25% of HER2-positive, and 10% to 15% of triple-negative tumors.5
Moreover, PIK3CA mutations appear to hold prognostic and predictive value in hormone receptor—positive, HER2-negative advanced or metastatic breast cancer. In an analysis of results from 12 studies, investigators found that patients who received non-PI3K inhibitor therapy had a lower median progression-free survival (PFS) if their tumors harbored PIK3CA mutations than those without the aberration. Among patients who received anti-PI3K therapy, those with PIK3CA mutations achieved a higher median PFS than those without the mutation.6
For each breast cancer subtype, PIK3CA alterations result in activation of PI3K-α, 1 of 4 PI3K isoforms. Of the isoforms, PI3K-α and PI3K-β are widely expressed, in contrast with PI3K-δ and PI3K-γ, which are mainly found in immune and hematopoietic cells.7 Although early trials suggested that an isoform-selective inhibitor of PI3K-α could be more effective and tolerable than a pan-PI3K inhibitor,7 the realities thus far have been disappointing. Still, existing research has at least shown proof of concept.
At ASCO 2018, Jose Baselga, MD, PhD, physician-in-chief at Memorial Sloan Kettering Cancer Center in New York, New York, presented results from the phase III SANDPIPER trial for taselisib. Baselga has pioneered the development of dual pathway inhibitors in breast cancer.
“Our findings are proof that targeting this pathway [PI3K] in breast cancer is effective,” Baselga said. “However, the benefit to patients was more modest than we had hoped for, and there is a risk of considerable side effects.”8
RTKs indicates receptor tyrosine kinases.
Zardavas Z, Phillips W, Loi S. PIK3CA mutations in breast cancer: reconciling ndings from preclinical and clinical data. Breast Cancer Res. 2014;16(1):201.
ASCO expert Harold J. Burstein, MD, PhD, of Dana-Farber Cancer Institute in Boston, Massachusetts, said the SANDPIPER trial is “a modest step forward but an important step forward, because it does suggest we can effectively target the pathway, and, hopefully, this is something we can build on.Difficulties in PIK3CA research can be divided into 2 broad categories. The first is tolerability: Inhibition of PI3K-α causes on-target hyperglycemia and other adverse events (AEs), particularly if selectivity is incomplete. The second is efficacy: PI3K inhibition may be insufcient to control PIK3CA-mutated breast cancer. This could be due to interactions downstream or with other mutated pathways. For example, PTEN mutations seem to nullify PI3K-α inhibitors, and mutated KRAS—of the KRAS/RAF/MEK/ERK pathway—appears to overcome a variety of PI3K/ AKT/mTOR inhibitors.9,10 Future research aims to investigate such interactions, possibly leading to new combination therapies or dual inhibitors.
More than 20 PI3K inhibitors have entered clinical trials, but many were abandoned before reaching phase III. So far, 2 have been approved: Copanlisib (Aliqopa), which is predominantly active against PI3K-α and PI3K-δ isoforms expressed in malignant B cells, is approved for patients with relapsed follicular lymphoma; idelalisib (Zydelig), a PI3K-δ inhibitor, is indicated in relapsed follicular lymphoma, chronic lymphocytic leukemia, and small lymphocytic lymphoma.Current breast cancer research concerning the PIK3CA mutation revolves around developing an effective PI3K-α inhibitor. The leading agent in this field has been taselisib. Unfortunately, recent results from the phase III SANDPIPER trial demonstrated limited benefit and poor tolerability, likely due to a lack of complete specificity for the alpha isoform.11
The double-blind, placebo-controlled SANDPIPER trial involved 516 postmenopausal women with ER-positive, HER2-negative locally advanced or metastatic breast cancer. In all, 340 patients had PIK3CA-mutant tumors. Each cohort received either taselisib 4 mg daily in combination with 500 mg of fulvestrant (Faslodex) or placebo plus fulvestrant.11
Patients receiving taselisib plus fulvestrant achieved a median PFS of 7.4 months (95% CI, 7.26-9.07), compared with 5.4 months (95% CI, 3.68- 7.29) for placebo plus fulvestrant (HR, 0.70; 95% CI, 0.56-0.89; P = .0037). Although this improvement was statistically significant, almost half (49.5%) of the participants receiving taselisib plus fulvestrant experienced grade 3 or greater AEs, compared with just 16.4% of patients receiving placebo plus fulvestrant. Diarrhea (60.1%) and hyperglycemia (40.4%) were most common for patients receiving taselisib plus fulvestrant.11
In response to these results, Baselga said, “Although [taselisib] is powerful against alpha [isoform], it also hits delta and gamma. The [adverse] effects we see that are limiting our patients staying on are mostly delta and gamma. I do think that in the case of breast cancer, what we need to do is to work on more specific alpha inhibitors that will be safer, because I think the negative factor here is actually the [adverse] effect profile that is based on the nonspecificity against the other isotopes.”
The results prompted Genentech to halt plans to pursue a breast cancer indication for taselisib. “The magnitude of benefit observed in SANDPIPER isn’t as strong as we had hoped for, and, given the challenging safety profile of this combination and the current clinical landscape, we will not be pursuing an FDA submission for taselisib based on the data presented at ASCO,” the company said in a statement.
This outcome is particularly frustrating because it mirrors that of buparlisib, a pan-PI3K inhibitor previously pursued by Baselga and colleagues.12 The phase III BELLE-2 trial compared fulvestrant, an ER agonist, with and without buparlisib in postmenopausal women with hormone receptor—positive, HER2-negative advanced breast cancer. The findings demonstrated that adding PI3K inhibition could extend PFS and was most effective for patients with PIK3CA mutations.
However, the buparlisib combination also came with more serious AEs than fulvestrant alone (23% vs 16%, respectively) and a higher incidence of all-grade AEs, particularly hyperglycemia, and elevated alanine aminotransferase and aspartate aminotransferase levels. This prompted a redirection of efforts from pan-PI3K inhibition to alpha isoform selectivity.Finding strategies to inhibit the PI3K pathway in Obreast cancer remains an active area of research. Several trials for buparlisib are continuing and other novel agents also are being studied (Table). Early findings include the following.
A selective PI3K-α inhibitor, alpelisib is under study in the phase III SOLAR-1 trial in combination with fulvestrant in patients with advanced breast cancer that has progressed during or after aromatase inhibitor therapy (NCT02437318). It also is being evaluated in phase II trials for a variety of solid tumor types, including breast cancer.
BC indicates breast cancer; ER, estrogen receptor; MBC, metastatic breast cancer; TNBC, triple-negative breast cancer.
aStudy is active but not recruiting participants.
bLSZ102 is a selective estrogen receptor degrader.
cDual inhibitor of PI3K and mTOR kinases.
In an ongoing phase II trial, 17 patients with advanced ER-positive, HER2-negative breast cancer who had received at least 1 previous systemic therapy in the advanced setting were given 350 mg of alpelisib daily in 28-day cycles. The objective response rate (ORR) was 41%, and 59% of patients achieved clinical benefit. Researchers also reported that coexisting MAP3K1/ESR1 mutations were associated with a more favorable outcome, whereas CCND1/FGFR1 mutations predicted unfavorable responses.Another phase I/II trial showed encouraging results when alpelisib was combined with nab-paclitaxel (Abraxane) for patients with advanced or metastatic HER2-negative breast cancer. Forty-three patients were given 250, 300, or 350 mg of alpelisib daily plus 100 mg/m2 nab-paclitaxel on days 1, 8, and 15 every 28 days. The ORR for this group was 60%, and the clinical benefit rate was 79%. Of patients with PI3K pathway—activated breast cancer, which consisted of those with either a PIK3CA or PTEN mutation, 100% (n = 19) showed clinical benefit. Additionally, the longest duration of response of 26 months was observed in a patient with PIK3CA mutation.14
Hyperglycemia was among the most frequently observed AEs in the study. Overall, 75% of patients experienced hyperglycemia of any grade, including 27% with grade 3 severity, and 32% required metformin to manage the AE. Other common all-grade AEs included diarrhea (88%), fatigue (68%), peripheral neuropathy (63%), and rash (63%).14
GDC-0077 has demonstrated greater than 300-fold selectivity for the alpha isoform over the beta, delta, and gamma isoforms in preclinical models. This effort may fulfill the call for more selective agents to improve tolerability. In PIK3CA-mutant breast cancer xenograft models, GDC-0077 induced tumor regressions as monotherapy and enhanced the combination of palbociclib (Ibrance) and fulvestrant.15 A phase I study under way involves patients with hormone receptor—positive, HER2-negative breast cancer (NCT03006172) as a single agent and in combination with hormone-targeting therapies.