Search

There is limited knowledge on the benefit of the α-subunit–specific PI3K inhibitor alpelisib in later lines of therapy for advanced estrogen receptor–positive (ER+) HER2− and triple-negative breast cancer (TNBC). We conducted a phase II multicohort study of alpelisib monotherapy in patients with advanced PI3K pathway mutant ER+HER2− and TNBC. In the intention-to-treat ER+ cohort, the overall response rate was 30% and the clinical benefit rate was 36%. A decline in PI3K pathway mutant circulating tumor DNA (ctDNA) levels from baseline to week 8 while on therapy was significantly associated with a partial response, clinical benefit, and improved progression-free-survival [HR 0.24; 95% confidence interval (CI), 0.083–0.67, P = 0.0065]. Detection of ESR1 mutations at baseline in plasma was also associated with clinical benefit and improved progression-free survival (HR 0.22; 95% CI, 0.078–0.60, P = 0.003). Significance: Alpelisib monotherapy displayed efficacy in heavily pretreated ER+ breast cancer with PIK3CA mutations. PIK3CA mutation dynamics in plasma during treatment and ESR1 mutations detected in plasma at baseline were candidate biomarkers predictive of benefit from alpelisib, highlighting the utility of ctDNA assays in this setting. This article is highlighted in the In This Issue feature, p. 2007

Supplementary Data from Alpelisib Monotherapy for PI3K-Altered, Pretreated Advanced Breast Cancer: A Phase II Study

Supplementary Figure from Alpelisib Monotherapy for PI3K-Altered, Pretreated Advanced Breast Cancer: A Phase II Study

Supplementary Methods

Supplementary Tables S1-5

Supplementary Table S8

Supplementary Figures

Background: The PI3Kα subunit specific inhibitor alpelisib in combination with fulvestrant is approved in advanced breast cancer harbouring PIK3CA mutations, but predictive biomarkers are lacking. We performed a phase II single arm study of alpelisib monotherapy in previously treated advanced hormone receptor-positive HER2-negative (HR+HER2-) and triple negative breast cancer (TNBC) cohorts. Methods: Eligible patients with genomic aberrations in the PI3K pathway determined via tumour sequencing or ctDNA were given alpelisib 350mg oral daily as monotherapy. The primary endpoint was RECIST objective response rate (ORR) by central review. Secondary endpoints were clinical benefit rate (CBR), defined as complete or partial response or stable disease for ≥24 weeks, and progression free survival (PFS). Positron emission tomography (FDG-PET) was performed at baseline and 8 weeks. Efficacy was analysed according to baseline tumour alterations and serial ctDNA assays. Analyses were conducted separately for the HR+HER2- and TNBC cohorts. Stated p values are nominal without multiplicity adjustment. Results: 43 patients were recruited in total (33 in ER+HER2-, 10 in TNBC). Median follow-up was 44.6 months. In the TNBC cohort ORR and CBR were 0%, hence the cohort was stopped early. Data will be presented here for the HR+HER2- cohort. Of 33 ER+HER2- patients recruited, 28 (85%) were evaluable and 26 (79%) were evaluable for ORR after central review. PI3K pathway aberrations at study entry were mutations in PIK3CA (26/28; 93%) and PTEN (2/28; 7%). The median age was 60 (41 - 77yrs) and 27/28 of patients were female. 85.7% of cases had visceral disease, and 57% had >3 metastatic sites. 100% of patients had received prior endocrine therapy and 78.6% prior chemotherapy, with a median of 3 prior lines (range 1 - 9). All patients were endocrine refractory. The ORR was 38% and CBR 43%. Median PFS was 4.6 months (95% CI, 3.7 - 7.3) and median OS 16.0 months (95% CI, 8.5 - 34.4). In patients with PET scans at baseline and 8 weeks, 67% (17/25) achieved a partial metabolic response but this was not associated with clinical benefit. In patients with partial response by RECIST, serial ctDNA levels at week 8 declined from baseline (p = 0.008) compared to patients with stable or progressive disease (p = 0.17), with similar results seen for patients with clinical benefit. Decline in ctDNA at week 8 was also associated with improved PFS (HR 0.24 [95% CI 0.08-0.67], p = 0.0065; 5.6 v 3.6 months). Multiple PIK3CA mutations were detected in plasma in 31% (8/26) of patients but had no impact on ORR/CBR. 39% (11/28) of patients had mutations in ESR1 detected in plasma at baseline, and this was associated with clinical benefit (p = 0.019) and improved PFS (HR=0.22 [95% CI 0.078-0.60], P=0.0032, 8.21 v 4.53 months). ctDNA assays detected 11 cases with ESR1 mutations compared to tumour sequencing which detected 5 cases only. No other baseline alterations were significantly associated with outcomes, although no patients (0/4) with PTEN alterations achieved clinical benefit. End of treatment ctDNA analysis identified multiple new alterations including MAP3K1 in 39% (11/28), TP53 in 32% (9/28) and PTEN in 25% (7/28). Adverse events included 20.9% grade ≥3 hyperglycaemia and 25.6% ≥grade 3 rash consistent with reported data. Permanent discontinuation of alpelisib for toxicity occurred in 15.2% (5/33) of patients. Conclusions: Alpelisib monotherapy in pre-treated ER+HER2- disease showed evidence of clinical efficacy. Decline in the levels of PIK3CA mutations detected with ctDNA on therapy are significantly associated with clinical benefit, as are the presence of ESR1 mutations at baseline. ctDNA improves the yield of ESR1 mutation detection and warrants further study as a biomarker of alpelisib monotherapy benefit in endocrine-refractory populations. Citation Format: Peter Savas, Louisa Lisa Lo, Stephen James Luen, Elizabeth Fay Blackley, Courtney Templeton van Geelen, Yi-An Ko, Kate Moodie, Jason William Callahan, Chen-Fang Weng, Andjelija Zivanovic Bujak, Miriam Manning Yeung, Sarah Ftouni, Prudence Anne Francis, Sarah-Jane Dawson, Sherene Loi. Alpelisib monotherapy for PI3K-altered, pre-treated advanced breast cancer: A phase II study [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-18-14.

Venetoclax, a potent and selective BCL2 inhibitor, synergizes with endocrine therapy in preclinical models of ER-positive breast cancer. Using a phase Ib 3 + 3 dose-escalation and expansion study design, 33 patients with ER and BCL2-positive metastatic disease (mean prior regimens, 2; range, 0–8) were treated with daily tamoxifen (20 mg) and venetoclax (200–800 mg). Apart from uncomplicated “on-target” lymphopenia, no dose-limiting toxicities or high-grade adverse events were observed in the escalation phase (15 patients), and 800 mg was selected as the recommended phase II dose (RP2D). In the expansion phase (18 patients), few high-grade treatment-related adverse events were observed. For 24 patients treated at the RP2D, the confirmed radiologic response rate was 54% and the clinical benefit rate was 75%. Treatment responses were preempted by metabolic responses (FDG-PET) at 4 weeks and correlated with serial changes in circulating tumor DNA. Radiologic responses (40%) and clinical benefit (70%) were observed in 10 patients with plasma-detected ESR1 mutations. Significance: In the first clinical study to evaluate venetoclax in a solid tumor, we demonstrate that combining venetoclax with endocrine therapy has a tolerable safety profile and elicits notable activity in ER and BCL2-positive metastatic breast cancer. These findings support further investigation of combination therapy for patients with BCL2-positive tumors. See related commentary by Drago et al., p. 323. This article is highlighted in the In This Issue feature, p. 305

BackgroundMetastatic breast cancer (mBC) is a heterogenous disease with increasing availability of targeted therapies as well as emerging genomic markers of therapeutic resistance, necessitating timely and accurate molecular characterization of disease. As a minimally invasive test, analysis of circulating tumour DNA (ctDNA) is well positioned for real-time genomic profiling to guide treatment decisions. Here, we report the results of a prospective testing program established to assess the feasibility of ctDNA analysis to guide clinical management of mBC patients.Methods and findingsTwo hundred thirty-four mBC patients (median age 54 years) were enrolled between June 2015 and October 2018 at the Peter MacCallum Cancer Centre, Melbourne, Australia. Median follow-up was 15 months (range 1-46). All patient samples at the time of enrolment were analysed in real time for the presence of somatic mutations. Longitudinal plasma testing during the course of patient management was also undertaken in a subset of patients (n = 67, 28.6%), according to clinician preference, for repeated molecular profiling or disease monitoring. Detection of somatic mutations from patient plasma was performed using a multiplexed droplet digital PCR (ddPCR) approach to identify hotspot mutations in PIK3CA, ESR1, ERBB2, and AKT1. In parallel, subsets of samples were also analysed via next-generation sequencing (targeted panel sequencing and low-coverage whole-genome sequencing [LC-WGS]). The sensitivity of ddPCR and targeted panel sequencing to identify actionable mutations was compared. Results were discussed at a multidisciplinary breast cancer meeting prior to treatment decisions. ddPCR and targeted panel sequencing identified at least 1 actionable mutation at baseline in 80/234 (34.2%) and 62/159 (39.0%) of patients tested, respectively. Combined, both methods detected an actionable alteration in 104/234 patients (44.4%) through baseline or serial ctDNA testing. LC-WGS was performed on 27 patients from the cohort, uncovering several recurrently amplified regions including 11q13.3 encompassing CCND1. Increasing ctDNA levels were associated with inferior overall survival, whether assessed by ddPCR, targeted sequencing, or LC-WGS. Overall, the ctDNA results changed clinical management in 40 patients including the direct recruitment of 20 patients to clinical trials. Limitations of the study were that it was conducted at a single site and that 31.3% of participants were lost to follow-up.ConclusionIn this study, we found prospective ctDNA testing to be a practical and feasible approach that can guide clinical trial enrolment and patient management in mBC.

Background The increasing availability of targeted agents for treatment of metastatic breast cancer (mBC) necessitates accurate and timely molecular characterisation of disease. As a minimally invasive test, circulating tumour DNA (ctDNA) is well positioned to overcome many of the limitations associated with traditional tumor biopsies. Here, we established a program to assess the feasibility of routine prospective ctDNA testing for the clinical management of mBC patients. Methods Detection of somatic mutations from patient plasma was performed using a multiplexed droplet digital PCR (ddPCR) approach to identify hotspot mutations in PIK3CA, ESR1, ERBB2 and AKT1. In parallel, a subset of samples were also analysed via next generation sequencing (targeted amplicon (TA) sequencing and low-coverage whole-genome sequencing). Results were discussed at a multidisciplinary breast cancer meeting prior to therapy selection. Results 234 mBC patients were enrolled on this study, with a median age at diagnosis of 54 years (28-80) and a median of 2 lines of prior therapy. The average turnaround time for ctDNA testing using ddPCR was 9 days (1-49). Using ddPCR, 80/234 (34.2%) patients had ≥1 mutation identified, with 52/234 (22.2%) patients having an alteration in PIK3CA, 35/234 (15.0%) in ESR1, 9/234 (3.8%) in AKT1 and 2/234 (0.9%) in ERBB2. TA sequencing performed in the first 159 patients, identified actionable mutations (classified using the OncoKB database) in 63 patients (39.6%) and showed that a mean variant allele fraction of > 5% was significantly associated with inferior overall survival (Hazard ratio: 1.8; 95% Confidence interval: 1.1-3.1; p Conclusions Prospective ctDNA testing of mBC patients is a practical and feasible approach to guide clinical trial enrolment and patient management. Legal entity responsible for the study The authors. Funding National Health and Medical Research Council Australia. Disclosure S.Q. Wong: Travel / Accommodation / Expenses: Bio-Rad Laboratories. S. Dawson: Research grant / Funding (self): Genentech. All other authors have declared no conflicts of interest.

1055Background: Mutations of genes involved in the PI3K signalling pathway occur frequently in BC. We performed a phase II study of BYL719, a selective PI3Kα inhibitor, in BC to evaluate its effica...