Purpose: Alpelisib, a selective oral inhibitor of the class I PI3K catalytic subunit p110α, has shown synergistic antitumor activity with endocrine therapy against ER+/PIK3CA-mutated breast cancer cells. This phase Ib study evaluated alpelisib plus letrozole's safety, tolerability, and preliminary activity in patients with metastatic ER+ breast cancer refractory to endocrine therapy.
Experimental Design: Twenty-six patients received letrozole and alpelisib daily. Outcomes were assessed by standard solid-tumor phase I methods. Tumor blocks were collected for DNA extraction and next-generation sequencing.
Results: Alpelisib's maximum-tolerated dose (MTD) in combination with letrozole was 300 mg/d. Common drug-related adverse events included hyperglycemia, nausea, fatigue, diarrhea, and rash with dose-limiting toxicity occurring at 350 mg/d of alpelisib. The clinical benefit rate (lack of progression ≥6 months) was 35% (44% in patients with PIK3CA-mutated and 20% in PIK3CA wild-type tumors; 95% CI, 17%–56%), including five objective responses. Of eight patients remaining on treatment ≥12 months, six had tumors with a PIK3CA mutation. Among evaluable tumors, those with FGFR1/2 amplification and KRAS and TP53 mutations did not derive clinical benefit. Overexpression of FGFR1 in ER+/PIK3CA mutant breast cancer cells attenuated the response to alpelisib in vitro.
Conclusions: The combination of letrozole and alpelisib was safe, with reversible toxicities. Clinical activity was observed independently of PIK3CA mutation status, although clinical benefit was seen in a higher proportion of patients with PIK3CA-mutated tumors. Phase II and III trials of alpelisib and endocrine therapy in patients with ER+ breast cancer are ongoing. Clin Cancer Res; 23(1); 26–34. ©2016 AACR.
This article is featured in Highlights of This Issue, p. 1
The combination of the PI3Kα-specific inhibitor alpelisib and the aromatase inhibitor letrozole is safe and clinically active in patients with ER+/PIK3CA-mutant breast cancer refractory to primary endocrine therapy. This result suggests a novel molecularly targeted strategy that may abrogate or delay the emergence of antiestrogen resistance in hormone-dependent breast cancer.
The PI3K pathway is the most frequently altered pathway in cancer, including mutation and/or amplification of the genes encoding the PI3K catalytic subunits p110α (PIK3CA) and p110β (PIK3CB), the PI3K regulatory subunit p85α (PIK3R1), AKT1-3, and the phosphatidylinositol-3,4,5 trisphosphate (PIP3) phosphatases PTEN and INPP4B, among others. PIK3CA mutations induce a transformed phenotype, including growth factor- and anchorage-independent growth, resistance to anoikis, and drug resistance (1–4). About 40% of ER+ breast cancers harbor PIK3CA mutations (5–7).
We and others have shown that aberrant activation of PI3K signaling is associated resistance to endocrine therapy (8). PI3K signaling has been shown to promote estrogen-independent growth of ER+ breast cancer cells (9, 10), and this growth is inhibited by the addition of PI3K inhibitors to antiestrogens (11). Additionally, inhibition of PI3K prevents the emergence of hormone-independent cells, which suggests that early intervention with antiestrogens and PI3K inhibitors could limit escape from endocrine therapy.
Drugs targeting multiple levels of the PI3K network have been developed (12, 13). Alpelisib (BYL719; Novartis Pharma AG) is an oral inhibitor that selectively targets p110α (14, 15). A phase I study of alpelisib, alone and in combination with fulvestrant, declared its maximum tolerated dose (MTD) as 400 mg/d (16).
The primary objective of this phase Ib trial (NCT01791478) was to determine the safety and tolerability of letrozole, an aromatase inhibitor, in combination with alpelisib in patients with ER+/HER2− metastatic breast cancer refractory to endocrine therapies. Secondary objectives included antitumor activity and correlation of clinical outcome with presence of PIK3CA mutations in tumor specimens. Preliminary data from other ongoing clinical trials of alpelisib with letrozole (NCT01872260) show no evidence of pharmacokinetic (PK) drug–drug interactions (17), therefore, no PK analysis was planned for this trial.
Patients and Methods
The patient population included postmenopausal patients with histologically confirmed ER+/HER2− metastatic breast cancer refractory to at least one line of endocrine therapy in the metastatic setting, or diagnosed with metastatic breast cancer during or within 1 year of adjuvant endocrine therapy; evaluable disease as defined by Response Evaluation Criteria In Solid Tumors (RECIST); age ≥18 years; life expectancy ≥6 months; ECOG performance status ≤1; adequate bone marrow, hepatic, and renal function; and fasting plasma glucose levels ≤140 mg/dL (7.8 mmol/L). A tumor specimen (primary or metastatic) from archival material or a fresh biopsy was required for study enrollment. Key exclusion criteria were CYP3A4 modifier drug treatment ≤2 weeks before starting alpelisib, clinically manifested diabetes mellitus, and prior treatment with PI3K inhibitors. Prior treatment with everolimus was allowed.
Approval was obtained from the ethics committees (IRB #101057, Vanderbilt University) at the participating institutions and regulatory authorities. All patients gave informed consent. The study followed the Declaration of Helsinki and Good Clinical Practice guidelines.
This phase Ib, multicenter, open-label study enrolled subjects using a standard 3+3 dose escalation design. All patients received letrozole 2.5 mg/d; alpelisib was initiated at 300 mg/d, 25% below the single-agent MTD. Both medications were administered daily on a 28-day cycle. In case of adverse events requiring dose adjustments, alpelisib was reduced to 250 mg/d. Intrapatient dose reductions were allowed after the initial 4 weeks of treatment. Patients were treated until disease progression, unacceptable toxicity, or withdrawal of consent.
Dose-limiting toxicities (DLTs) were defined as Common Terminology Criteria for Adverse Events (CTCAE) version 4.0 grade ≥3 toxicities. Exceptions were any grade ≥2 toxicity necessitating treatment interruption for more than 21 consecutive days, and non-CTCAE grade 2 hyperglycemia not resolved to grade 0 within 14 consecutive days of initiation of oral antidiabetic medications (metformin). Grade ≥3 anemia was not considered a DLT unless judged to be a hemolytic process secondary to study drug. Grade ≥3 lymphopenia was not considered a DLT unless clinically significant.
The MTD was defined as the highest dose of alpelisib in combination with letrozole not causing DLT in more than 33% of patients in the first treatment cycle. Twenty or more evaluable patients had to be treated before declaration of the MTD, with ≥6 evaluable patients treated at the MTD for one cycle. Criteria for evaluability were ≥21 days on alpelisib in cycle 1 or early discontinuation due to a DLT. The recommended phase II dose was defined as the highest dose at or below the MTD at which ≥75% of the patients could tolerate therapy for a minimum of 8 weeks without development of grade ≥2 hyperglycemia for more than 14 consecutive days despite initiation of oral antidiabetic medications; and grade ≥3 rash, grade ≥2 nausea, vomiting or diarrhea, and grade ≥2 rash, all for more than 14 consecutive days of optimal medical treatment.
Safety and radiographic assessments
Clinical and laboratory assessments were conducted at baseline and weekly during cycle 1; on days 1 and 15 of cycle 2; and on day 1 of subsequent cycles. Safety assessments included serial electrocardiograms and fasting plasma glucose. Radiographic responses were assessed every 2 months using RECIST version 1.1. Best response analysis was conducted in any patient that surpassed 8 weeks of treatment, therefore having at least one set of radiographic assessments. Clinical benefit rate was defined as any patient that did not have disease progression on their radiographic assessments for a set time frame (i.e., ≥ 6 or 12 months). Patients that discontinued study treatment due to toxicity prior to their first radiographic assessment were considered nonevaluable.
DNA was extracted using DNEasy or QiaAMP DNA tissue kits from formalin-fixed paraffin-embedded (FFPE) archival tumor sections or snap-frozen biopsies of metastases, respectively. Tumor cellularity was assessed by an expert breast pathologist (MES); specimens with ≥20% tumor nuclei were considered evaluable. In the few cases with paucicellular samples (<20% tumor cells), multiple sections were macro-dissected to achieve ≥20% tumor cellularity.
DNA was initially subjected to SNaPShot (18) analysis of 18 mutations in PIK3CA, PTEN, and AKT1, including the common hotspot mutations in exons 9 and 20 of PIK3CA, as previously described (19). SNaPShot is based on multiplex PCR, primer extension with fluorescently tagged dideoxy-nucleotides and capillary electrophoresis; it is a fast, high-throughput, multiplex profiling method based on the Applied Biosystems SNaPshot platform.
Subsequently, next-generation sequencing (NGS) was performed using the MSK-IMPACT (Memorial Sloan Kettering–Integrated Mutation Profiling of Actionable Cancer Targets; ref. 20), a hybridization capture-based assay for targeted deep sequencing of all exons and selected introns of 341 key cancer genes in DNA from FFPE tumor sections. Custom DNA probes targeting exons and selected introns of 341 genes were synthesized using the NimbleGen SeqCap EZ library custom oligo system and biotinylated to allow for sequence enrichment by capture using streptavidin-conjugated beads. Pooled libraries containing captured DNA fragments were subsequently sequenced on the Illumina HiSeq 2500 system (rapid-run mode) as 2 × 100-bp paired-end reads.
FGFR1 fluorescence in situ hybridization (FISH)
Three- to 4-μm tissue sections were mounted on sialinized slides and hybridized overnight with the ZytoLight SPEC FGFR1/CEN 8 Dual-Color Probe (ZytoVision). Briefly, deparaffinization, pretreatment, then the slides were denatured in the presence of 10-μL probe for 6 minutes at 73°C and hybridized at 37°C overnight in StatSpin (Thermobrite, Abbott Molecular, Inc.). Posthybridization saline sodium citrate washes were performed at 72°C, and the slides were next stained with DAPI before analysis. Slides were analyzed with Reflected light fluorescent microscope (Olympus BX60) at 100× for hotspots. Representative images of tumor cells were captured using Cytovision software. Thirty tumor cell nuclei from hotspots or random areas were individually evaluated with the 100× by counting green FGFR1 and orange centromer 8 (CEN8) signals. The average of FGFR1 copy number and the FGFR1/CEN8 was calculated. Cases were considered as FGFR1 positive (‘amplified’) under one of the following conditions:
The FGFR1/CEN8 ratio is ≥2.0.
The average number of FGFR1 signals per tumor cell nucleus is ≥6.
The percentage of tumor cells containing ≥15 FGFR1 signals or large clusters is ≥10%.
The percentage of tumor cells containing ≥5 FGFR1 signals is ≥50%, with (a–c) representing a high-level and (d) a low-level amplification (21).
MCF7 and CAMA-1 cells were obtained from ATCC and were cultured in DMEM supplemented with 10% FBS.
FGFR1- and GFP-expressing lentiviruses were generated by cotransfecting 4-μg proviral pLX302-FGFR1 or pLX302-GFP plasmids (Open BioSystems), 3 μg psPAX2 (plasmid encoding gag, pol, rev, and Tat genes) and 1 μg pMDG2 envelope plasmid (Sigma Aldrich) into 293FT cells using Lipofectamine 2000 (Thermo Fisher). 293FT cells were fed with growth media 24 hours after transfection; virus-containing supernatants were harvested 48 and 72 hours after transfection, diluted 1:4 and applied to target cells with 8 μg/mL polybrene (Sigma Aldrich). Target cells were selected with 1 μg/mL puromycin.
Cells were plated in phenol red-free IMEM + 10% dextran-charcoal–treated FBS and treated for 6 hours with drugs or vehicle. Stimulation with FGF2 (Sigma Aldrich) was performed 20 minutes before the lysis. Lysates were generated by removing media from cells, washing monolayers with cold PBS and lysis with RIPA: 50 mmol/L Tris, pH 7.4, 150 mmol/L NaCl, 1% Nonidet P-40, 0.5% Deoxycholate, 0.1% SDS, 1 mmol/L EDTA, 50 mmol/L NaF, 1 mmol/L NaVO4, and 1× protease inhibitor mixture (Roche). Lysates were clarified by centrifugation at 15,000 × g for 15 minutes. Protein concentration was determined by BCA assay (Thermo scientific). For immunoblot analysis, equal amounts of protein/lane were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and probed with the following antibodies: phospho-Akt (Ser473), total-AKT, phospho-p44/42 MAPK (Erk1/2), total p44/p42 MAPK (Erk1/2), and actin (Cell Signaling Technologies); FGFR1 antibody was obtained from Abcam. Immunoreactive proteins were visualized by enhanced chemiluminescence (Pierce).
For longer-term growth assays, CAMA-1 cells were seeded into 6-well plates in estrogen free-media (IMEM + 10% dextran-charcoal–treated FBS) plus 5 ng/mL FGF2 and treated with vehicle, alpelisib 2 μmol/L, lucitanib 2 μmol/L (provided by Clovis Oncology), and the combination. Media and inhibitors were replenished every 3 days, and cells were grown for 2 weeks until the untreated wells achieved 80% confluence. Cells were fixed and stained in 20% methanol with 0.5% crystal violet and washed with water. Dried plates were imaged on a flatbed scanner. Afterward, the crystal violet was solubilized with 20% acid acetic and quantitated by spectrophotometric detection at 490 nm using a plate reader (GloMax-Multi Detection System, Promega).
MCF-7 cells were plated in 96-well plates (3,000 cells per well) in a volume of 100 μL of phenol red-free IMEM + 10% dextran-charcoal–treated FBS. Twenty-four hours later, DMSO (vehicle) or alpelisib (0.05–4 μmol/L) ± lucitanib 1 μmol/L, all in the presence of 5 ng/mL FGF2 were added to the wells. After 72 hours, cell proliferation was measured using the CellTiter-Glo Luminescent Cell Viability Assay (Promega). Percentage inhibition was calculated relative to median signal from DMSO-treated wells.
Twenty-six patients were enrolled from April 2013 to February 2014 (Table 1). Overall, 85% of the patients had bone metastases and approximately 30% had visceral metastases. All patients not presenting with de novo metastatic disease had previously been exposed to an aromatase inhibitor in the adjuvant setting. In the metastatic setting, 85% had been exposed to at least one line of endocrine therapy, 69% to an aromatase inhibitor, and 31% to chemotherapy. Evaluable tumor samples were obtained in all 26 patients; 74% of these were FFPE blocks from the primary tumor.
Dose escalation and MTD
Alpelisib dose was started at 300 mg/d. DLT (grade 3 rash) occurred in two patients at the 350 mg/d dose. These rashes eventually resolved with medical intervention (high-dose antihistamines, topical, and/or oral corticosteroids) and drug interruption. One patient was able to restart at 300 mg/d. The alpelisib MTD was set at 300 mg/d; a total of 20 patients were treated at this dose (escalation and expansion phases). Less than 25% of patients required alpelisib interruption or dose reduction during the first 8 weeks of treatment.
Adverse events in this trial were similar to reports from the phase I single-agent alpelisib trial (16) and are summarized in Table 2. The most common side effects were gastrointestinal disorders (73%), hyperglycemia (62%), fatigue (54%), and rash (42%); all of these were dose dependent. Grade 3 adverse events associated with alpelisib were uncommon and generally cumulative (such as hyperglycemia and gastrointestinal disturbances), except for maculopapular rash, which tended to appear within the first 2 weeks of treatment. Of note, rash was less frequent in incidence and severity in patients that had been on antihistamines (i.e., due to seasonal allergies) prior to trial enrollment. Rashes (of any grade) responded well to high-dose antihistamines (twice-daily dosing), obviating the need for corticosteroids in the majority of patients developing this adverse event.
Twenty-three of 26 patients (88%) were evaluable for best response, three of 26 patients discontinued treatment due to toxicity prior to their first radiographic assessment (Table 3). Five patients (19%) achieved a partial response; four of these patients had ER+/PIK3CA-mutated breast cancer. Clinical benefit rate (lack of disease progression ≥6 months) was seen in nine patients (35%; 95% CI, 17%–56%), and eight patients (31%; 95% CI, 14%–52%) continued to have clinical benefit for more than 12 months. Of those eight patients, six had ER+/PIK3CA-mutant breast cancer, and all had previous exposure to an aromatase inhibitor in the metastatic setting (Fig. 1). Five patients are still on study for more than 12 months (data cutoff, June 2015).
Seventy-five percent (61%) of analyzed tumors were FFPE sections from the original breast cancer diagnosis (median time since original diagnosis 34 months; range, 0–99). We initially screened for PIK3CA mutations with SNaPShot (ref. 18; on all 26 tumors' DNA). We then followed with targeted capture NGS (23 tumors were evaluable for this analysis) in order to identify somatic alterations associated with clinical benefit or lack thereof. In all cases, findings by SNaPShot were confirmed by NGS. NGS analysis showed fewer genomic alterations in tumors from patients with clinical benefit compared with those without clinical benefit (41 vs. 79 genomic alterations; not shown). Most common alterations (>10%) were PI3KCA, TP53, GATA3, and BRCA1/2 mutations, MCL1, FGFR1, CCND1, FGF3/4/19, and CCND1 amplifications, and deletions/truncations in PTEN (Fig. 2A). The majority of PIK3CA mutations were in exons 9 and 20 (E545K, E542K, and Q546K, seven tumors; H1047R alone in five tumors). Interestingly, four of five (80%) patients with tumors with PIK3CAH1047R had durable clinical responses, whereas only two of seven (28%) tumors with mutations in exon 9 exhibited a clinical response in excess of 6 months. One of these tumors had a mutation in the Ras-binding domain (I273V) in addition to E545K. A tumor with a novel deletion in the C2 domain of PIK3CA (P447_L455del) exhibited a solid partial response in liver metastases (Fig. 1).
Seventeen of 23 patients (74%) with NGS data were evaluable for clinical benefit ≥6 months; three patients discontinued study treatment due to toxicity prior to their first radiographic assessment; three patients discontinued study treatment due to toxicity prior to their second radiographic assessment and were therefore considered nonevaluable for the NGS data/outcome analysis. Seven out of 17 patients (41%) had durable responses to the combination (Fig. 2A). Five of seven (71%) patients with clinical benefit had a PIK3CA-mutant tumor, in contrast to five of 10 (50%) patients without clinical benefit (Fig. 2B). Two patients with tumors harboring both PIK3CA and a known activating ESR1 mutation (Y537S and D538G, respectively) in their primary tumor exhibited stable disease or partial response. Interestingly, among patients evaluable for response, all those with cancers with FGFR1 and FGFR2 amplification, and KRAS and TP53 mutations progressed on treatment. FGFR1 amplification detected by NGS was confirmed by FISH analysis (Fig. 2C).
FGFR1 overexpression blocks the effect of alpelisib of ER+/PIK3CA mutant cells
To explore whether overexpression of FGFR1 is causal to resistance to alpelisib, we first examined the effect of alpelisib in the ER+ CAMA-1 breast cancer cells, which harbor FGFR1 amplification. FISH analysis of these cells showed a FGFR1/CEN8 ratio of 2.2, consistent with FGFR1 gene amplification (Fig. 3A). Growth of CAMA-1 cells in estrogen-free medium supplemented with FGF2 was completely insensitive to alpelisib, but inhibited by the FGFR1-3 tyrosine kinase inhibitor (TKI), lucitanib (Fig. 3B). Treatment with alpelisib inhibited p-AKT but not p-ERK, whereas lucitanib inhibited p-FRS2 and p-ERK but not p-AKT (Fig. 3C). We then stably transduced ER+/PIK3CA-mutant MCF7 cells with a FGFR1 expression vector. Compared with MCF7GFP control cells, FGFR1-overexpressing cells exhibited slightly higher levels of p-AKT and p-ERK (Fig. 3D). Similar to the results with CAMA-1 cells, treatment with alpelisib inhibited p-AKT but not p-ERK, whereas the addition of lucitanib mainly inhibited p-ERK but not p-AKT levels. Finally, MCF7GFP cells grown as monolayers had an IC50 to alpelisib <1 μmol/L, whereas this was at least 10-fold higher in MCF7FGFR1 cells. The addition of lucitanib resensitized MCF7FGFR1 cells to the PI3Kα inhibitor (Fig. 3E). Altogether, these data suggest a causal association between FGFR1 overexpression and activation of ERK with relative resistance to alpelisib.
The initial rationale for the development of isozyme-specific antagonists was to allow anti-p110α, anti-p110β, and anti-p110δ agents to be delivered at maximal target-inhibitory doses while potentially avoiding the side effects of pan-PI3K inhibitors. Specific inhibitors of p110α, such as alpelisib and MLN1117, and p110β-sparing inhibitors (e.g., taselisib, GDC-0032) are being developed with a focus on PIK3CA-mutant tumors. The results from this study provide evidence that alpelisib plus letrozole is safe, tolerable, and active in postmenopausal patients with ER+/HER2− metastatic breast cancer refractory to endocrine therapy. The MTD and recommended dose for phase II trials of alpelisib in combination with letrozole was defined as 300 mg/d. This finding was in contrast to the alpelisib single-agent MTD, which is 400 mg/d (16).
Most common side effects (hyperglycemia, rash, nausea, diarrhea, and fatigue) were of similar frequency to those seen in the phase I study of single-agent alpelisib (16). However, the rate of severe (grade 3) maculopapular rash was lower, most likely due to a reduced alpelisib accumulation with the 300 mg/d dose, and/or the unintended prophylactic use of antihistaminics (i.e., due to seasonal allergies) prior to trial enrollment. In view of the urticarial-like nature of the rash, we would endorse the prophylactic use of antihistaminics in order to minimize the incidence and severity of the alpelisib-associated rash. Consistent with the hyperglycemia seen with pan-PI3K inhibitors (22–24), more than 50% of patients developed mild to moderate hyperglycemia, an on-target toxicity common to all PI3K inhibitors, especially the ones with more sustained inhibition of p110α. Severe hyperglycemia was uncommon, with none of the patients requiring administration of insulin; all of them were successfully managed with oral hypoglycemic agents such as metformin. Of note, transaminitis and gastrointestinal side effects were less frequent and severe than those observed in trials with pan-PI3K inhibitors (22–24).
The combination of letrozole and alpelisib showed sustained clinical activity: 35% of patients remained on treatment ≥6 months and 31% remained on treatment ≥12 months. The clinical benefit rate ≥6 and ≥12 months seen in patients with ER+/PIK3CA-mutant metastatic breast cancer was numerically higher than in patients with wild-type PIK3CA tumors. Additionally, four of the five partial responses seen were in patients with a PIK3CA-mutant tumor. The significance of these findings is uncertain due to the small number of patients, but suggests that p110α-specific inhibitors may have greater activity against breast cancers of this genotype. The preliminary observation that the most frequent mutation in PIK3CA, H1047R, appears to be associated with higher clinical benefit from the p110α inhibitor compared with mutations in the helical domain further increases the enthusiasm for the development of PIK3CA mutant–specific inhibitors. Interestingly, a novel deletion in the C2 domain of PIK3CA (P447_L455del) was associated with a partial response. This domain is in contact with the iSH2 domain of p85, the regulatory subunit of PI3K (25). This deletion would be predicted to disrupt the interaction of p110α with p85, thus de-repressing p110α from p85-mediated inhibition. Confirmation of this hypothesis will require additional investigation beyond this report, but the partial response observed in this patient suggests that this mutation is oncogenic and, thus, may confer tumor dependence on PI3K function.
Several genomic alterations were detected in addition to PIK3CA mutations, and a higher proportion of those were seen in the group of patients without clinical benefit. FGFR1 and FGFR2 amplifications and mutations in TP53, BRCA1/2, and KRAS were found only in tumors from patients that progressed on therapy. Tumors with aberrant activation of the RAS/RAF/MEK/ERK pathway, such as KRAS-mutant cancers, do not respond to PI3K pathway inhibitors (26) and would not be expected to respond to alpelisib. Of note, the predominant pathway activated downstream of FGFRs has been shown to be ERK both in developmental models and in cancer cells (27, 28). Amplification of FGFR1 occurs in approximately 10% of breast cancers, predominantly in ER+ cancers where it has been associated with resistance to tamoxifen (29). Thus, to investigate whether an excess of FGFR1 is also associated with a lower response to PI3K inhibitors, we engineered ER+/PIK3CA-mutant cells with FGFR1 overexpression. In these cells, the IC50 to alpelisib was increased >10-fold compared with cells without FGFR1 amplification, suggesting a causal association between FGFR1 overexpression and drug resistance. Treatment with the FGFR TKI lucitanib completely restored the antitumor effect of alpelisib while inhibiting ERK but not AKT. Similar results were obtained with ER+/FGFR1-amplified CAMA-1 breast cancer cells, which were resistant to alpelisib but sensitive to lucitanib. These results are consistent with the lack of clinical benefit observed in the FGFR-amplified tumors in the trial herein and suggest a causal association between FGFR overexpression and both aberrant activation of ERK and resistance to PI3K inhibitors.
Two of the patients who did not have clinical benefit had amplification of ERBB2 (HER2) in a metastatic biopsy, consistent with the notion that this genetic alteration confers resistance to endocrine therapy (30, 31). Of note, the primary tumors in these two patients were not found to have HER2 amplification by FISH. Interestingly, two patients with clinical benefit were found to have ESR1 mutations in addition to PIK3CA mutations in metastatic biopsies. These ESR1 mutations (Y537S and D538G) are in the ligand-binding domain of ERα and have previously been shown to exhibit estrogen-independent transcriptional activity (32–34). They are rarely found in primary untreated tumors and are usually detected in metastases in patients who progress after prolonged endocrine therapy. The clinical benefit seen in these patients suggests the possibility of an interaction between mutant PIK3CA and ERα that requires further study.
The results of this study demonstrate that the combination of the p110α-specific inhibitor alpelisib with letrozole is safe, tolerable, and active in patients with endocrine therapy–resistant ER+ advanced breast cancer, particularly those that also harbor an activating mutation in PIK3CA. Several ongoing clinical trials are now exploring the use of alpelisib and endocrine therapy in patients with both ER+/PIK3CA-mutant or wild-type breast cancer in the neoadjuvant and metastatic settings.
Disclosure of Potential Conflicts of Interest
I.A. Mayer, D. Juric, and L.C. Cantley are consultant/advisory board members for Novartis. J.M. Balko is the co-inventor of a patent using MHC-II expression for prediction of response to anti-PD-L1 inhibitors, which is owned by Vanderbilt University. No potential conflicts of interest were disclosed by the other authors.
Conception and design: I.A. Mayer, D. Solit, Y. Li, L.C. Cantley, E. Winer, C.L. Arteaga
Development of methodology: I.A. Mayer, D. Solit.
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): I.A. Mayer, V.G. Abramson, L. Formisano, M.V. Estrada, M.E. Sanders, D. Juric, D. Solit, M.F. Berger, E. Winer, C.L. Arteaga
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): I.A. Mayer, L. Formisano, J.M. Balko, M.V. Estrada, D. Solit, M.F. Berger, H.H. Won, Y. Li, E. Winer, C.L. Arteaga
Writing, review, and/or revision of the manuscript: I.A. Mayer, V.G. Abramson, L. Formisano, J.M. Balko, M.V. Estrada, M.E. Sanders, D. Juric, D. Solit, Y. Li, L.C. Cantley, E. Winer, C.L. Arteaga
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.M. Balko, M.V. Estrada, M.E. Sanders, D. Solit, C.L. Arteaga
Study supervision: I.A. Mayer, D. Juric, L.C. Cantley, C.L. Arteaga
Other (funding of study): C.L. Arteaga
This work was funded by Stand Up to Cancer Dream Team Translational Research Grant (SU2CAACR-DT0209) – Stand Up To Cancer is a program of the Entertainment Industry Foundation administered by the American Association for Cancer Research, Breast SPORE P50 CA098131, VICC Support Grant P30 CA68485, a Breast Cancer Research Foundation grant (C.L. Arteaga), Susan G. Komen for the Cure Foundation SAC grant (SAC100013), K23 CA127469-01A2 (I.A. Mayer), and by R01 GM041890 (L.C. Cantley). This study was sponsored by Novartis Pharmaceuticals.
Violeta Sánchez was instrumental in the tissue handling and processing at the VICC. Rajmohan Murali and Nancy Bouvier were instrumental in the tissue handling, processing, and molecular analysis. The authors thank the patients for their participation, and their families for support throughout the study.
NOTE: Prior presentation: Presented in part at the American Association for Cancer Research (AACR) Annual Meeting, Philadelphia, PA, 2015.
- Received January 15, 2016.
- Revision received March 3, 2016.
- Accepted March 8, 2016.
- ©2016 American Association for Cancer Research.