Alpelisib for PIK3CA-Mutated, Hormone Receptor–Positive Advanced Breast Cancer
F. André, E. Ciruelos, G. Rubovszky, M. Campone, S. Loibl, H.S. Rugo,
H. Iwata, P. Conte, I.A. Mayer, B. Kaufman, T. Yamashita, Y.-S. Lu, K. Inoue,
M. Takahashi, Z. Pápai, A.-S. Longin, D. Mills, C. Wilke, S. Hirawat, and D. Juric, for the SOLAR-1 Study Group*

BACKGROUND
PIK3CA mutations occur in approximately 40% of patients with hormone receptor (HR)– positive, human epidermal growth factor receptor 2 (HER2)–negative breast cancer. The PI3Kα-specific inhibitor alpelisib has shown antitumor activity in early studies.
METHODS
In a randomized, phase 3 trial, we compared alpelisib (at a dose of 300 mg per day) plus fulvestrant (at a dose of 500 mg every 28 days and once on day 15) with placebo plus fulvestrant in patients with HR-positive, HER2-negative advanced breast cancer who had received endocrine therapy previously. Patients were enrolled into two cohorts on the basis of tumor-tissue PIK3CA mutation status. The primary end point was pro- gression-free survival, as assessed by the investigator, in the cohort with PIK3CA-mu- tated cancer; progression-free survival was also analyzed in the cohort without PIK3CA- mutated cancer. Secondary end points included overall response and safety.
RESULTS
A total of 572 patients underwent randomization, including 341 patients with con- firmed tumor-tissue PIK3CA mutations. In the cohort of patients with PIK3CA-mutated cancer, progression-free survival at a median follow-up of 20 months was 11.0 months (95% confidence interval [CI], 7.5 to 14.5) in the alpelisib–fulvestrant group, as com- pared with 5.7 months (95% CI, 3.7 to 7.4) in the placebo–fulvestrant group (hazard ratio for progression or death, 0.65; 95% CI, 0.50 to 0.85; P<0.001); in the cohort with- out PIK3CA-mutated cancer, the hazard ratio was 0.85 (95% CI, 0.58 to 1.25; posterior probability of hazard ratio <1.00, 79.4%). Overall response among all the patients in the cohort without PIK3CA-mutated cancer was greater with alpelisib–fulvestrant than with placebo–fulvestrant (26.6% vs. 12.8%); among patients with measurable disease in this cohort, the percentages were 35.7% and 16.2%, respectively. In the overall population, the most frequent adverse events of grade 3 or 4 were hyperglycemia (36.6% in the alpelisib–fulvestrant group vs. 0.7% in the placebo–fulvestrant group) and rash (9.9% vs. 0.3%). Diarrhea of grade 3 occurred in 6.7% of patients in the alpelisib– fulvestrant group, as compared with 0.3% of those in the placebo–fulvestrant group; no diarrhea of grade 4 was reported. The percentages of patients who discontinued alpelisib and placebo owing to adverse events were 25.0% and 4.2%, respectively.
CONCLUSIONS
Treatment with alpelisib–fulvestrant prolonged progression-free survival among patients with PIK3CA-mutated, HR-positive, HER2-negative advanced breast cancer who had received endocrine therapy previously. (Funded by Novartis Pharmaceuticals; SOLAR-1 ClinicalTrials.gov number, NCT02437318.)

The authors’ full names, academic de- grees, and affiliations are listed in the Appendix. Address reprint requests to Dr. André at Institut Gustave Roussy, 114 Rue Edouard Vaillant, Villejuif, 94805, France, or at [email protected]; or to Dr. Juric at Massachusetts General Hospi- tal Cancer Center, 55 Fruit St., Boston, MA 02114, or at [email protected].
*A complete list of the investigators in the SOLAR-1 Study Group is provided in the Supplementary Appendix, available at NEJM.org.
N Engl J Med 2019;380:1929-40. DOI: 10.1056/NEJMoa1813904
Copyright © 2019 Massachusetts Medical Society.

T he new engl a nd jour na l of medicine
M

ore than 70% of breast cancers are hormone receptor (HR)–positive and human epidermal growth factor
receptor 2 (HER2)–negative.1,2 Approximately 40% of patients with HR-positive, HER2-negative breast cancer have activating mutations in the gene PIK3CA, inducing hyperactivation of the alpha isoform (p110α) of phosphatidylinositol 3-kinase (PI3K).3-5 Endocrine therapy, with or without the use of a cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitor, is the standard treat- ment for patients with HR-positive, HER2-nega- tive advanced breast cancer.6-8 However, acquired resistance to endocrine-based therapy remains a challenge.9,10
Alpelisib (BYL719) is an orally bioavailable,
small-molecule, α-specific PI3K inhibitor that selectively inhibits p110α approximately 50 times as strongly as other isoforms.11 PIK3CA-mutated cancers have been shown to be sensitive to alpelisib in preclinical tumor models11 and in a phase 1 trial of alpelisib in patients with ad- vanced solid tumors.12 The combination of al- pelisib with fulvestrant had synergistic antitumor activity as compared with either agent alone in PIK3CA-mutated, estrogen-receptor–positive xeno- graft models.13,14 In a phase 1b trial, alpelisib plus fulvestrant led to a complete or partial re- sponse in 29% of patients with heavily pretreat- ed PIK3CA-altered, HR-positive advanced breast cancer, as compared with no complete or partial response in patients without PIK3CA-mutated tumors.15 The most frequent adverse events of grade 3 or 4 that were reported with alpelisib were hyperglycemia and maculopapular rash.15 Here, we present the results of the primary analysis of SOLAR-1 (Clinical Studies of Alpelis- ib in Breast Cancer 1), a phase 3 trial to evaluate the efficacy and safety of an α-specific PI3K in- hibitor plus fulvestrant in patients with PIK3CA- mutated, HR-positive, HER2-negative advanced breast cancer who had received endocrine ther- apy previously.

Trial Design
We conducted this randomized, double-blind, placebo-controlled, phase 3 trial in 34 countries and enrolled patients at 198 trial centers into two cohorts on the basis of tumor-mutation
status (PIK3CA-mutated vs. not PIK3CA-mutated). Within each cohort, patients were randomly as- signed in a 1:1 ratio to receive oral alpelisib (at a dose of 300 mg to be taken with food [one 200-mg tablet and two 50-mg tablets], regard- less of body weight, with continuous daily dos- ing) plus fulvestrant (administered as a 500-mg intramuscular injection on days 1 and 15 of cy- cle 1 and on day 1 of subsequent 28-day cycles) or placebo plus fulvestrant. Within each cohort, randomization was stratified according to the presence or absence of lung or liver metastases and according to previous receipt of CDK4/6 in- hibitor treatment. Patients received treatment until disease progression, an unacceptable level of toxic effects, withdrawal of consent, loss to follow-up, or death. Dose reductions of alpelisib (or matching placebo) were permitted (according to the schedule of changes in the daily dose from 300 mg to 250 mg to 200 mg) to help man- age adverse events (see the Supplementary Ap- pendix, available with the full text of this article at NEJM.org). No dose reductions of fulvestrant were allowed. Patients who discontinued alpelisib or placebo could continue receiving fulvestrant.
patients
Enrollment was open to men and postmeno- pausal women who had locally confirmed HR- positive, HER2-negative advanced breast cancer, were eligible to receive further endocrine therapy after relapse or progression, and were receiving or had received aromatase inhibitor treatment in the context of neoadjuvant or adjuvant therapy or for advanced disease. Patients had to have ade- quate tumor tissue for central analysis of PIK3CA mutational status. For postmenopausal women, previous radiation therapy to the ovaries or pre- vious treatment with a luteinizing hormone– releasing hormone agonist for induction of ovar- ian suppression was prohibited. Patients were excluded if they had received chemotherapy previ- ously for advanced disease, had received fulves- trant therapy previously, or had received any PI3K, AKT, or mTOR (mechanistic target of ra- pamycin) inhibitor.
Patients had either measurable disease (at
least one measurable lesion according to the Response Evaluation Criteria in Solid Tumors [RECIST], version 1.1) or one or more predomi- nantly lytic bone lesions, an Eastern Cooperative

1930

n engl j med 380;20

nejm.org

may 16, 2019

alpelisib for advanced Breast Cancer

Oncology Group performance-status score of 0 or 1 (on a scale from 0 to 5, with higher num- bers indicating greater disability), and adequate organ and bone marrow function. Patients were excluded if they had inflammatory breast cancer, uncontrolled central nervous system metastases, concurrent cancer or cancer within 3 years be- fore randomization (except for adequately treated basal-cell or squamous-cell carcinoma, nonmela- nomatous skin cancer, or curatively resected cervical cancer), type 1 diabetes or uncontrolled type 2 diabetes (fasting plasma glucose level,
>140 mg per deciliter [7.7 mmol per liter], or a
glycosylated hemoglobin level of >6.4%), or cur- rently documented pneumonitis.
Primary resistance was defined as relapse within 24 months while the patient was receiv- ing adjuvant endocrine therapy or progression within 6 months while the patient was receiving endocrine therapy for advanced disease. Second- ary resistance was defined as relapse after at least 24 months of adjuvant endocrine therapy, relapse within 12 months after ending adjuvant endocrine therapy, or progression after at least 6 months of endocrine therapy for advanced dis- ease. Patients whose disease relapsed at least 12 months after the completion of adjuvant endo- crine therapy and who were not treated for ad- vanced disease were considered to have endo- crine-sensitive disease. A subsequent protocol amendment (on August 30, 2016) specified that these patients were ineligible for enrollment, in order to focus the trial on the endocrine-resis- tant population.
End points
The primary end point was progression-free sur- vival, as assessed by the investigator, according to RECIST, version 1.1, in the cohort of patients with PIK3CA-mutated cancer. The key secondary end point was overall survival in the cohort with PIK3CA-mutated cancer. Additional secondary end points included progression-free survival and overall survival in the cohort without PIK3CA- mutated cancer, progression-free survival accord- ing to the level of circulating tumor DNA (ctDNA), overall response, clinical benefit (defined as a complete or partial response or as stable disease for >6 months), and safety. (The ctDNA-related and overall survival analyses are not report- ed here.)
Assessments
Before enrollment, cohort status was centrally determined according to the presence or absence of any PIK3CA mutation by means of polymerase- chain-reaction analysis of mutation hot spots in the C2, helical, and kinase domains of PI3K (cor- responding to exons 7, 9, and 20, respectively) with the use of a tumor-tissue sample, preferably a sample obtained during the most recent pro- gression. Imaging (computed tomography, mag- netic resonance imaging, or both) was performed at screening within 4 weeks before randomiza- tion, every 8 weeks for the first 18 months, and then every 12 weeks until disease progression or withdrawal for any other reason. Vital signs and hematologic and biochemical laboratory tests were performed at screening, every 2 weeks for the first 8 weeks, and then every 4 weeks. The fasting glucose level was also assessed on day 8. Adverse events (assessed according to the Na- tional Cancer Institute Common Terminology Criteria, version 4.03) were recorded continu- ously until 30 days after the last dose of trial treatment.
Trial Oversight
The trial protocol, which includes the statistical analysis plan, is available at NEJM.org. The origi- nal trial protocol and subsequent amendments were approved by an independent ethics com- mittee and institutional review board at each site. The SOLAR-1 trial was designed and over- seen by a steering group of medical oncology experts, including representatives from the trial sponsor (Novartis). The protocol was designed by the steering committee, which included the primary investigator. Written informed consent for trial participation and biomarker-sample col- lection was obtained from all the participants. An independent data and safety monitoring committee reviewed unblinded efficacy and safe- ty data. The trial was conducted in accordance with Good Clinical Practice guidelines and the principles of the Declaration of Helsinki. All the authors had access to all trial data and were in- volved in the development of the manuscript and approved its submission for publication. The authors confirm that the trial conformed to the protocol and statistical analysis plan, and they vouch for the accuracy and completeness of the reported data, which were analyzed by a statisti-

n engl j med 380;20

nejm.org

may 16, 2019
1931

T he new engl a nd jour na l of medicine

cian employed by Novartis. A professional medi- cal writer, funded by the sponsor, assisted with the development of the manuscript.

Statistical Analysis
The primary end point (progression-free survival in the cohort with PIK3CA-mutated cancer) was compared between groups by means of a strati- fied log-rank test at a one-sided 2.0% signifi- cance level. We calculated that 243 events of disease progression or death would be required for the trial to detect a hazard ratio of 0.6 with 83.8% power. Two interim analyses were con- ducted after 42% (in a futility analysis) and 78% (in an efficacy analysis) of the expected numbers of events of progression or death were document- ed. The overall type I error rate for this group- sequential design was controlled with the use of a Haybittle–Peto boundary. Of the overall alpha level of 0.02 for the PIK3CA-mutation cohort, 0.0001 was spent at the interim efficacy analysis; this left an alpha level of 0.0199 remaining to de- clare statistical significance at the final analysis. Proof-of-concept criteria, designed to assess whether a treatment benefit was obtained in the biomarker-negative control cohort, required a hazard ratio of 0.60 or less and a posterior prob- ability of at least 90% that the true hazard ratio was less than 1.00; data from the cohort without PIK3CA mutations were analyzed with the use of a one-sided 0.5% significance level. A separate O’Brien–Fleming alpha-spending function, which was independent of the Haybittle–Peto boundary that was used for the primary efficacy analysis, guaranteed protection of the overall type I error (at an alpha level of 2.5%, equivalent to a two- sided level of 5%, on the basis of a Bonferroni adjustment) across all hypotheses and repeated testing of the overall survival hypotheses at the
interim and the final analyses.
A stratified Cox regression model was used to estimate the hazard ratio and 95% confidence interval in the analysis of progression-free sur- vival. To provide supportive evidence regarding the primary end point, progression-free survival was also assessed in an audit-based random subgroup of 50% of the cohort of patients with PIK3CA-mutated cancer by an independent review committee whose members were unaware of the trial-group assignments. Within each cohort, overall survival was tested only if there was a significant difference between the trial groups with regard to progression-free survival. Efficacy
analyses were performed with the use of data from all the patients in the two cohorts (with and without PIK3CA-mutated cancer) who under- went randomization, and safety analyses included all the patients who received at least one dose of any trial agent. Additional details regarding the trial design are included in the Supplementary Appendix.

Characteristics of the patients
Between July 26, 2015, and July 21, 2017, a total of 572 patients underwent randomization. A total of 1244 patients were tested for PIK3CA mutation status, and interpretable results were available for 1173 (94.3%). A total of 341 patients had PIK3CA-mutated disease, including 169 who were assigned to receive alpelisib plus fulvestrant and 172 who were assigned to receive placebo plus fulvestrant (Fig. S1 in the Supplementary Appen- dix). The characteristics of the patients in the co- hort with PIK3CA-mutated cancer were balanced between the two trial groups at baseline (Table 1). The median age of these patients was 63 years. Lung or liver metastases were present in 170 patients (49.9%), and 77 (22.6%) had bone-only disease. A total of 20 patients (5.9%) had re- ceived CDK4/6 inhibitor therapy previously. At randomization, 292 patients (85.6%) had endo- crine-resistant disease. An additional cohort of
231 patients without PIK3CA-mutated cancer
underwent randomization for the proof-of-con- cept analysis (Table 1).

Treatment in the Cohort with PiK3CA-Mutated Cancer
At the data cutoff (June 12, 2018), in the cohort with PIK3CA-mutated cancer, the trial intervention was ongoing in 42 patients (24.9%) receiving alpelisib–fulvestrant and in 32 (18.6%) receiving placebo–fulvestrant. The median duration of ex- posure to alpelisib was 5.5 months (interquartile range, 1.6 to 13.0), and the median duration of exposure to placebo was 4.6 months (interquar- tile range, 1.9 to 13.1). The most common rea- sons for discontinuation of a trial agent were progressive disease (in 93 patients [55.0%] in the alpelisib–fulvestrant group and 117 [68.0%] in the placebo–fulvestrant group) and decision by the patient or the patient’s guardian (in 16 pa- tients [9.5%] and 6 patients [3.5%], respectively). The median relative dose intensity was 82.7% for

1932

n engl j med 380;20

nejm.org

may 16, 2019

alpelisib for advanced Breast Cancer

Table 1. Characteristics of the Patients at Baseline.*
Characteristic Cohort with PIK3CA-Mutated Cancer Cohort without PIK3CA-Mutated Cancer
Alpelisib– Placebo– Alpelisib– Placebo– Fulvestrant Group Fulvestrant Group Fulvestrant Group Fulvestrant Group (N = 169) (N = 172) (N = 115) (N = 116)
Age — yr
Median 63 64 62 63
Range 25–87 38–92 39–82 32–88
Female sex — no. (%) 168 (99.4) 172 (100) 115 (100) 116 (100)
ECOG performance-status score — no. (%)†
0 112 (66.3) 113 (65.7) 84 (73.0) 79 (68.1)
1 56 (33.1) 58 (33.7) 30 (26.1) 37 (31.9)
Missing data 1 (0.6) 1 (0.6) 1 (0.9) 0
Sites of metastases — no. (%)‡
Breast 1 (0.6) 3 (1.7) 5 (4.3) 4 (3.4)
Bone only 42 (24.9) 35 (20.3) 26 (22.6) 23 (19.8)
Visceral site
Any 93 (55.0) 100 (58.1) 66 (57.4) 74 (63.8)
Liver 49 (29.0) 54 (31.4) 41 (35.7) 36 (31.0)
Lung 57 (33.7) 68 (39.5) 37 (32.2) 55 (47.4)
Lung or liver 84 (49.7) 86 (50.0) 56 (48.7) 56 (48.3)
No. of metastatic sites — no. (%)
0 0 1 (0.6) 0 0
1 63 (37.3) 52 (30.2) 44 (38.3) 33 (28.4)
2 58 (34.3) 60 (34.9) 35 (30.4) 38 (32.8)
≥3 48 (28.4) 59 (34.3) 36 (31.3) 45 (38.8)
Previous treatment — no. (%)§
Any CDK4/6 inhibitor 9 (5.3) 11 (6.4) 7 (6.1) 8 (6.9)
Chemotherapy¶ 101 (59.8) 107 (62.2) 78 (67.8) 72 (62.1)
Line of treatment in advanced disease — no. (%)‖
First line 88 (52.1) 89 (51.7) 71 (61.7) 62 (53.4)
Second line 79 (46.7) 82 (47.7) 42 (36.5) 53 (45.7)
Endocrine status — no. (%)**
Primary resistance 23 (13.6) 22 (12.8) 31 (27.0) 26 (22.4)
Secondary resistance 120 (71.0) 127 (73.8) 66 (57.4) 65 (56.0)
Sensitivity 20 (11.8) 19 (11.0) 16 (13.9) 20 (17.2)
* Any differences between the two trial groups were less than 10% in the cohort of patients with PIK3CA-mutated cancer. The gene PIK3CA encodes for the alpha isoform of phosphatidylinositol 3-kinase (PI3Kα). Percentages may not total 100 because of rounding. Further data regarding the baseline characteristics of the patients are provided in Table S10 in the Supplementary Appendix. CDK denotes cyclin-dependent kinase.
† Eastern Cooperative Oncology Group (ECOG) performance-status scores are assessed on a scale from 0 to 5, with higher numbers indicating greater disability.
‡ One patient in the placebo group in each cohort had locally advanced disease with no metastases.
§ All patients had previously received treatment with an aromatase inhibitor.
¶ Chemotherapy was for patients receiving neoadjuvant or adjuvant therapy only. One patient in the placebo group of the cohort with PIK3CA- mutated cancer received chemotherapy for advanced disease (which was a protocol deviation).
‖ Three patients in each trial cohort (two patients in the alpelisib–fulvestrant group and one in the placebo–fulvestrant group in each cohort) were excluded because of protocol deviations.
** Primary endocrine resistance was defined as relapse within 24 months while the patient was receiving adjuvant endocrine therapy or pro- gression within 6 months while the patient was receiving endocrine therapy in the context of metastatic disease. Secondary endocrine re- sistance was defined as relapse that occurred after at least 24 months while the patient was receiving adjuvant endocrine therapy, relapse that occurred within 12 months after the end of adjuvant endocrine therapy, or progression that occurred after at least 6 months while the patient was receiving endocrine therapy in the context of metastatic disease. After enrollment began, the trial protocol was updated to ex- clude patients who had a relapse at least 12 months after the completion of neoadjuvant or adjuvant endocrine therapy and had not been treated for metastatic disease (endocrine sensitive).

n engl j med 380;20

nejm.org

may 16, 2019
1933

T he new engl a nd jour na l of medicine

alpelisib and 100% for placebo. Dose interrup- tions for alpelisib or matching placebo occurred in 125 patients (74.0%) receiving alpelisib–fulves- trant and in 55 (32.2%) receiving placebo–ful- vestrant, and dose reductions occurred in 108 (63.9%) and 15 (8.8%), respectively.

Efficacy of Alpelisib–fulvestrant
in the Cohort with PiK3CA-Mutated Cancer
In the cohort with PIK3CA-mutated cancer, the median duration of follow-up from randomiza- tion to data cutoff was 20.0 months (range, 10.7 to 33.3). The median progression-free survival was 11.0 months (95% confidence interval [CI],
7.5 to 14.5) in the alpelisib–fulvestrant group, as compared with 5.7 months (95% CI, 3.7 to 7.4) in the placebo–fulvestrant group (hazard ratio for progression or death, 0.65; 95% CI, 0.50 to 0.85; P<0.001) (Fig. 1A). At 12 months, the per- centage of patients with progression-free survival was 46.3% in the alpelisib–fulvestrant group and 32.9% in the placebo–fulvestrant group. These results were supported by the blinded independent review, which showed a median progression-free survival of 11.1 months (95% CI, 7.3 to 16.8) among 85 patients who had been assigned to receive alpelisib–fulvestrant, as compared with 3.7 months (95% CI, 2.1 to 5.6) among 88 patients assigned to receive pla- cebo–fulvestrant (hazard ratio, 0.48; 95% CI,
0.32 to 0.71). Analyses of progression-free sur-
vival according to stratification criteria and im- portant demographic and prognostic factors showed consistent benefit of treatment with alpelisib–fulvestrant across prespecified subgroups (Fig. 2).
Overall response among all the patients in this cohort was greater with alpelisib–fulvestrant than with placebo–fulvestrant (26.6% vs. 12.8%), and clinical benefit was also greater with alpelisib– fulvestrant (61.5% vs. 45.3%) (Table 2). Among patients with measurable disease, overall re- sponse was 35.7% in the alpelisib–fulvestrant group and 16.2% in the placebo–fulvestrant group; the percentages of patients with clinical benefit were 57.1% and 44.1%, respectively (Ta- ble 2). The trial-group assignments remained concealed from the investigators and patients during follow-up for the assessment of overall survival at the time of the primary end-point analysis.
Efficacy of Alpelisib–fulvestrant in the Cohort without PiK3CA-Mutated Cancer
Proof-of-concept criteria were not met in the co- hort of patients without PIK3CA-mutated cancer at the final efficacy analysis. The median pro- gression-free survival was 7.4 months (95% CI,
5.4 to 9.3) in the alpelisib–fulvestrant group and
5.6 months (95% CI, 3.9 to 9.1) in the placebo– fulvestrant group (hazard ratio for progression or death, 0.85; 95% CI, 0.58 to 1.25; posterior probability of true hazard ratio <1.00, 79.4%) (Fig. 1B). At 12 months, the percentage of pa- tients with progression-free survival was 28.4% in the alpelisib–fulvestrant group and 22.2% in the placebo–fulvestrant group. The median dura- tion of follow-up was 7.4 months (range, 0.1 to 16.4) at the time of data cutoff (December 23, 2016). A majority of patients in this cohort went on to receive either chemotherapy or hormonal therapy plus a targeted therapy as their next treatment after progression.

Safety
The total safety population included 284 patients who received alpelisib–fulvestrant and 287 who received placebo–fulvestrant. The adverse events of any grade that occurred in at least 35% of the patients in either group were hyperglycemia (in 63.7% of the patients who received alpelisib–ful- vestrant and 9.8% of those who received placebo– fulvestrant), diarrhea (in 57.7% and 15.7%, respec- tively), nausea (in 44.7% and 22.3%), decreased appetite (in 35.6% and 10.5%), and rash (in 35.6% and 5.9%) or maculopapular rash (in 14.1% and 1.7%) (Table 3). The most common adverse events of grade 3 or 4, occurring in at least 5% of patients in either group, were hyperglycemia (in 36.6% of the patients who received alpelisib– fulvestrant and 0.7% of those who received pla- cebo–fulvestrant), rash (in 9.9% and 0.3%, re- spectively), maculopapular rash (in 8.8% and 0.3%), and diarrhea (in 6.7% and 0.3%).
Permanent discontinuation of alpelisib or pla-
cebo due to adverse events occurred in 71 pa- tients (25.0%) receiving alpelisib–fulvestrant and in 12 (4.2%) receiving placebo–fulvestrant. The most frequent adverse events leading to the dis- continuation of alpelisib were hyperglycemia (in 18 patients [6.3%]) and rash (in 9 [3.2%]); no patients discontinued placebo owing to hyper- glycemia or rash.

1934

n engl j med 380;20

nejm.org

may 16, 2019

Probability of Progression-free Survival
alpelisib for advanced Breast Cancer

8 10 12 14 16 18 20 22 24 26 28 30 31
Month
85 75 62 50 39 30 17 14 5 3 1 1 0
67 58 48 37 29 20 14 9 3 2 0 0 0

No. at Risk
Alpelisib+fulvestrant 115 110 86 76 48 48 31 29 14 12 7 5 3 0
Placebo+fulvestrant 116 110 79 72 43 42 31 30 20 20 8 5 1 0

Serious adverse events occurred in 99 patients (34.9%) receiving alpelisib–fulvestrant and 48 (16.7%) receiving placebo–fulvestrant (Table S3 in the Supplementary Appendix). There were 19 deaths during the trial (including during the
day postintervention safety period): 7 deaths (2.5%) in patients receiving alpelisib–fulvestrant and 12 (4.2%) in those receiving placebo–fulves- trant. A total of 5 patients receiving alpelisib– fulvestrant and 8 receiving placebo–fulvestrant

Probability of Progression-free Survival

n engl j med 380;20

nejm.org

may 16, 2019
1935

T he new engl a nd jour na l of medicine

Subgroup No. of Patients Hazard Ratio for Progression or Death (95% CI)
All patients 341 0.65 (0.50–0.85)
Lung or liver metastases
Yes 170 0.62 (0.44–0.89)
No 171 0.69 (0.47–1.01)
Bone-only disease
Yes 77 0.62 (0.33–1.18)
No 264 0.66 (0.49–0.88)
Previous CDK4/6 inhibitor treatment
Yes 20 0.48 (0.17–1.36)
No 321 0.67 (0.51–0.87)
Previous chemotherapy
Neoadjuvant 46 0.37 (0.17–0.80)
Adjuvant 161 0.63 (0.42–0.95)
None 133 0.87 (0.58–1.29)
Endocrine status
Primary resistance 45 0.64 (0.31–1.32)
Secondary resistance 247 0.66 (0.49–0.90)
Sensitivity 39 0.87 (0.35–2.17)
Line of treatment in advanced disease
First line 177 0.71 (0.49–1.03)
Second line 161 0.61 (0.42–0.89)
No. of metastatic sites
<3 234 0.59 (0.43–0.83)
≥3 107 0.77 (0.50–1.20)
PIK3CA mutation subtype
E542K 60 0.60 (0.29–1.23)
E545X 105 0.61 (0.37–1.00)
H1047X 193 0.68 (0.48–0.95)
Geographic region
Europe 173 0.56 (0.39–0.81)
North America 43 0.41 (0.19–0.91)
Asia 70 0.76 (0.42–1.37)
Latin America 31 1.43 (0.54–3.79)
Other 24 0.93 (0.25–3.45)
0.1 1.0 10.0
Alpelisib+Fulvestrant Better Placebo+Fulvestrant Better
Figure 2. Subgroup Analysis of Progression-free Survival in the Cohort with PIK3CA-Mutated Cancer.
Confidence intervals have not been adjusted for multiplicity. Inferences drawn from the confidence intervals may not be reproducible. The previous chemotherapy subgroup was based on the last line of chemotherapy received. Patients may have received chemotherapy in the context of both neoadjuvant and adjuvant therapy. Patients may have had more than one PIK3CA mutation. There were multiple subtypes of E545 and H1047 mutations. CDK de- notes cyclin-dependent kinase.

died from underlying breast cancer. Other deaths in the alpelisib group were due to cardiorespira- tory arrest and a second primary cancer. Safety profiles in the two trial groups were similar in the two cohorts (Tables S4 through S6 in the Supplementary Appendix).

These results show improvements in patients’ outcomes with the addition of an α-specific PI3K
inhibitor to standard treatment for PIK3CA- mutated, HR-positive, HER2-negative advanced breast cancer, findings that validate PIK3CA as an important treatment target in this population. Patients with PIK3CA-mutated, HR-positive, HER2- negative advanced breast cancer that had pro- gressed during or after the receipt of endocrine therapy had significantly longer progression-free survival when they received alpelisib–fulvestrant than when they received placebo–fulvestrant, with an estimated 35% lower risk of progression

1936

n engl j med 380;20

nejm.org

may 16, 2019

alpelisib for advanced Breast Cancer

Table 2. Best Overall Response, According to Local Assessment, in the Cohort with PIK3CA-Mutated Cancer.
Response Alpelisib–Fulvestrant Group Placebo–Fulvestrant Group
All patients
No. of patients 169 172
Confirmed best overall response — no. (%)
Complete response 1 (0.6) 2 (1.2)
Partial response 44 (26.0) 20 (11.6)
Stable disease 58 (34.3) 63 (36.6)
Neither complete response nor progressive disease* 38 (22.5) 25 (14.5)
Progressive disease 16 (9.5) 53 (30.8)
Unknown status 12 (7.1) 9 (5.2)
Overall response†
No. of patients 45 22
Percentage of patients (95% CI) 26.6 (20.1–34.0) 12.8 (8.2–18.7)
Clinical benefit‡
No. of patients 104 78
Percentage of patients (95% CI) 61.5 (53.8–68.9) 45.3 (37.8–53.1)
Patients with measurable disease at baseline
No. of patients 126 136
Confirmed best overall response — no. (%)
Complete response 1 (0.8) 2 (1.5)
Partial response 44 (34.9) 20 (14.7)
Stable disease 58 (46.0) 63 (46.3)
Progressive disease 13 (10.3) 45 (33.1)
Unknown status 10 (7.9) 6 (4.4)
Overall response†
No. of patients 45 22
Percentage of patients (95% CI) 35.7 (27.4–44.7) 16.2 (10.4–23.5)
Clinical benefit§
No. of patients 72 60
Percentage of patients (95% CI) 57.1 (48.0–65.9) 44.1 (35.6–52.9)
* In this category, the best overall response was evaluated only in patients who had no measurable disease at baseline according to the Response Evaluation Criteria in Solid Tumors, version 1.1.
† Overall response was defined as a complete or partial response.
‡ Clinical benefit in the overall population was defined as a complete or partial response, stable disease lasting at least 24 weeks, or the status of having neither a complete response nor progressive disease for at least 24 weeks.
§ Clinical benefit in patients with measurable disease at baseline was defined as a complete or partial response or as stable disease lasting at least 24 weeks.

or death. A clinically relevant treatment benefit was not observed for alpelisib–fulvestrant in the cohort without PIK3CA-mutated cancer. In the cohort with PIK3CA-mutated cancer, alpelisib– fulvestrant was also associated with significantly higher percentages of patients with tumor re- sponse than was placebo–fulvestrant, a finding
that is consistent with observations from the phase 1b study.15 Progression-free survival was similar in the placebo groups in the two cohorts defined according to PIK3CA mutation status.
In previous studies of PI3K inhibitors, patients with PIK3CA-mutated breast cancer had prolon- gation of progression-free survival that was sig-

n engl j med 380;20

nejm.org

may 16, 2019
1937

T he new engl a nd jour na l of medicine

Table 3. Most Frequent Adverse Events, According to Single Preferred Term and Regardless of Relationship to Intervention, in the Overall Patient Population.*
Adverse Event Alpelisib–Fulvestrant Group (N = 284) Placebo–Fulvestrant Group (N = 287)

Any Grade Grade 3 Grade 4 Any Grade Grade 3 Grade 4

number of patients (percent)
Any adverse event 282 (99.3) 183 (64.4) 33 (11.6) 264 (92.0) 87 (30.3) 15 (5.2)
Hyperglycemia† 181 (63.7) 93 (32.7) 11 (3.9) 28 (9.8) 1 (0.3) 1 (0.3)
Diarrhea‡ 164 (57.7) 19 (6.7) 0 45 (15.7) 1 (0.3) 0
Nausea‡ 127 (44.7) 7 (2.5) 0 64 (22.3) 1 (0.3) 0
Decreased appetite 101 (35.6) 2 (0.7) 0 30 (10.5) 1 (0.3) 0
Rash§ 101 (35.6) 28 (9.9) 0 17 (5.9) 1 (0.3) 0
Vomiting‡ 77 (27.1) 2 (0.7) 0 28 (9.8) 1 (0.3) 0
Weight loss 76 (26.8) 11 (3.9) 0 6 (2.1) 0 0
Stomatitis 70 (24.6) 7 (2.5) 0 18 (6.3) 0 0
Fatigue 69 (24.3) 10 (3.5) 0 49 (17.1) 3 (1.0) 0
Asthenia 58 (20.4) 5 (1.8) 0 37 (12.9) 0 0
Alopecia 56 (19.7) 0 0 7 (2.4) 0 0
Mucosal inflammation 52 (18.3) 6 (2.1) 0 3 (1.0) 0 0
Pruritus 51 (18.0) 2 (0.7) 0 16 (5.6) 0 0
Headache 50 (17.6) 2 (0.7) 0 38 (13.2) 0 0
Dysgeusia 47 (16.5) 0 0 10 (3.5) 0 0
Arthralgia 32 (11.3) 1 (0.4) 0 47 (16.4) 3 (1.0) 0
* Safety analyses included all the patients who received at least one dose of any trial agent; one patient who was randomly assigned to the placebo–fulvestrant group did not receive either placebo or fulvestrant. The events that are listed were reported as a single term in at least 15% of the patients for any grade in either group. Three adverse events of special interest (pancreatitis, severe cutaneous reactions, and pneumonitis) fell below the reporting threshold listed here. Hypersensitivity, which occurred in 16.5% of the patients in the alpelisib–fulves- trant group (grade ≥3 in 1.8%) and in 4.2% of those in the placebo–fulvestrant group (grade ≥3 in none), was not reported as any single preferred term that reached the reporting threshold listed here.
† Adverse events of any grade related to hyperglycemia (including diabetes mellitus, hyperglycemia, insulin resistance, and metabolic syndrome [preferred terms] and others [see the Methods section in the Supplementary Appendix for a complete list]) were reported in 65.8% of the patients in the alpelisib–fulvestrant group (grade ≥3 in 38.0%) and in 10.5% of those in the placebo–fulvestrant group (grade ≥3 in 0.7%).
‡ Gastrointestinal toxic effects of any grade (including nausea, vomiting, and diarrhea [preferred terms] and others [see the Methods section in the Supplementary Appendix for a complete list]) were reported in 75.4% of the patients in the alpelisib–fulvestrant group (grade ≥3 in 8.8%) and in 34.8% of those in the placebo–fulvestrant group (grade ≥3 in 1.0%). Diarrhea was assessed at a maximum grade 2 severity in 18.3% of the patients.
§ Adverse events of any grade related to rash (including rash, rash follicular, rash generalized, and rash maculopapular [preferred terms] and others [see the Methods section in the Supplementary Appendix for a complete list]) were reported in 53.9% of the patients in the alpelisib– fulvestrant group (grade ≥3 in 20.1%) and in 8.4% of those in the placebo–fulvestrant group (grade ≥3 in 0.3%).

nificant but not clinically meaningful. These include the Buparlisib Breast Cancer Clinical Evaluation (BELLE) 2 and 3 clinical trials of pan-PI3K inhibition with buparlisib16,17 and the SANDPIPER clinical trial of the β-sparing PI3K inhibitor taselisib.18 However, further develop- ment of pan-PI3K and β-sparing PI3K inhibitors has been limited by their narrow therapeutic index, which results in frequent treatment dis- continuation and low on-target bioactivity. Spe-
cific inhibition of PI3Kα may represent im- proved biologic targeting, a finding supported by the observed incidence of hyperglycemia of grade 3 or 4 (10.8% with taselisib vs. 36.6% with alpelisib).18
The safety profile in the SOLAR-1 trial was similar to that in previous trials of alpelisib plus fulvestrant.15 The most frequent adverse events were hyperglycemia, gastrointestinal toxic effects, and rash. Adverse events were generally revers-

1938

n engl j med 380;20

nejm.org

may 16, 2019

alpelisib for advanced Breast Cancer

ible and, with the exclusion of hyperglycemia, mostly of low grade. Hyperglycemia, an on-target effect of alpelisib, led to the permanent discon- tinuation of alpelisib in 6.3% of the patients. Because this toxic effect may be inextricably linked with α-specific PI3K inhibition,5 rigorous safety monitoring was performed during the trial to minimize permanent treatment discon- tinuations and optimize potential benefit. Ad- verse events were managed by means of dose modifications and early concomitant medical intervention as indicated (Tables S7 through S9 in the Supplementary Appendix).
Alpelisib has activity in patients with PIK3CA-
mutated, HR-positive, HER2-negative advanced breast cancer that has progressed during or after treatment with an aromatase inhibitor. There- fore, the integration of genomic testing for PIK3CA mutation into routine clinical practice may be useful in the selection of therapy; validated diag- nostic testing procedures are not yet available. This trial shows that treatment with alpelisib– fulvestrant can provide an extension of progres- sion-free survival among patients with PIK3CA- mutated disease. This effect was observed across various subgroups. Preliminary analysis of pro- gression-free survival on the basis of ctDNA re- sults shows a similar effect.19 With the availabil- ity of ribociclib, palbociclib, and abemaciclib for the treatment of HR-positive, HER2-negative advanced breast cancer, there is potential for an increasing number of patients to receive CDK4/6 inhibitors with endocrine therapy in the context of first-line and second-line treatment of ad- vanced breast cancer. However, PI3K-driven treat- ment resistance remains a problem.20,21 The BYLieve (Alpelisib [BYL719] in Patients with PIK3CA-Mutant, HR+, HER2− Advanced Breast Can-
cer) trial (ClinicalTrials.gov number, NCT03056755) is recruiting patients who have had disease pro- gression during or after treatment with a CDK4/6 inhibitor in order to further assess the efficacy of alpelisib in this context. In the SOLAR-1 trial, there appeared to be a strong treatment benefit in patients receiving second-line therapy and in patients who had received neoadjuvant or adju- vant chemotherapy previously, which supports the use of alpelisib–fulvestrant for the population of previously treated patients.
Preclinical studies have shown that some tu- mors with reduced sensitivity to alpelisib have sustained or increased levels of retinoblastoma protein and that the combination of PI3Kα and CDK4/6 inhibitors overcomes intrinsic and adap- tive resistance in PIK3CA-mutated xenografts.22 In addition, loss of phosphatase and tensin homo- logue protein (PTEN) has been shown to confer clinical resistance to alpelisib, which is reverted by PI3Kβ blockade in PTEN-null xenografts and cell lines.23
In conclusion, this phase 3 trial showed a sig- nificant prolongation of progression-free survival and greater overall response with alpelisib–fulves- trant than with placebo–fulvestrant among pa- tients with PIK3CA-mutated, HR-positive, HER2- negative advanced breast cancer who had disease that had relapsed or progressed during or after the receipt of previous endocrine therapy. There was a higher incidence of hyperglycemia, rash, and diarrhea with alpelisib than with placebo.
A data sharing statement provided by the authors is available with the full text of this article at NEJM.org.
Supported by Novartis Pharmaceuticals.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
We thank the trial participants, their families, and the staff at each trial site; and John Munro, of Articulate Science, for edito- rial assistance with an earlier version of the manuscript.

Appendix
The authors’ full names and academic degrees are as follows: Fabrice André, M.D., Eva Ciruelos, M.D., Gabor Rubovszky, M.D., Mario Campone, M.D., Sibylle Loibl, M.D., Hope S. Rugo, M.D., Hiroji Iwata, M.D., Pierfranco Conte, M.D., Ingrid A. Mayer, M.D., Bella Kaufman, M.D., Toshinari Yamashita, M.D., Yen-Shen Lu, M.D., Kenichi Inoue, M.D., Masato Takahashi, M.D., Zsuzsanna Pápai, M.D., Anne-Sophie Longin, M.Sc., David Mills, M.Sc., Celine Wilke, M.D., Samit Hirawat, M.D., and Dejan Juric, M.D.
The authors’ affiliations are as follows: Institut Gustave Roussy, INSERM Unité 981, Université Paris-Sud, Villejuif (F.A.), Institut de Cancérologie de l’Ouest, St. Herblain (M.C.), and Novartis Pharma, Paris (A.-S.L.) — all in France; Hospital Universitario 12 de Octubre, Madrid (E.C.); National Institute of Oncology (G.R.) and Duna Medical Center (Z.P.), Budapest, Hungary; German Breast Group, Neu- Isenburg, and Center for Hematology and Oncology Bethanien, Frankfurt — both in Germany (S.L.); UCSF Helen Diller Family Com- prehensive Cancer Center, San Francisco (H.S.R.); Aichi Cancer Center, Nagoya (H.I.), Kanagawa Cancer Center, Yokohama (T.Y.), Saitama Cancer Center, Saitama (K.I.), and National Hospital Organization Hokkaido Cancer Center, Sapporo (M.T.) — all in Japan; Istituto Oncologico Veneto and the Departments of Surgery, Oncology, and Gastroenterology, University of Padua, Padua, Italy (P.C.); Vanderbilt University, Nashville (I.A.M.); Chaim Sheba Medical Center, Tel Hashomer, Israel (B.K.); National Taiwan University Hospi- tal, Taipei (Y.-S.L.); Novartis Pharma, Basel, Switzerland (D.M., C.W.); Novartis Pharmaceuticals, East Hanover, NJ (S.H.); and Massa- chusetts General Hospital Cancer Center, Boston (D.J.).

n engl j med 380;20

nejm.org

may 16, 2019
1939

alpelisib for advanced Breast Cancer

References
Setiawan VW, Monroe KR, Wilkens LR, Kolonel LN, Pike MC, Henderson BE. Breast cancer risk factors defined by estro- gen and progesterone receptor status: the Multiethnic Cohort Study. Am J Epidemiol 2009;169:1251-9.
Howlader N, Altekruse SF, Li CI, et al. US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status. J Natl Cancer Inst 2014; 106(5):dju055.
Cancer Genome Atlas Network. Com- prehensive molecular portraits of human breast tumours. Nature 2012;490:61-70.
Mollon L, Aguilar A, Anderson E, et al. A systematic literature review of the prev- alence of PIK3CA mutations and mutation hotspots in HR+/HER2-metastatic breast cancer. Cancer Res 2018;78:Suppl 13:1207. abstract.
Goncalves MD, Hopkins BD, Cantley LC. Phosphatidylinositol 3-kinase, growth disorders, and cancer. N Engl J Med 2018; 379:2052-62.
National Comprehensive Cancer Net- work. NCCN clinical practice guidelines in oncology: breast cancer, version 2. 2018 (https://www.nccn.org/professionals/ physician_gls/pdf/breast.pdf).
Rugo HS, Rumble RB, Macrae E, et al. Endocrine therapy for hormone receptor- positive metastatic breast cancer: Ameri- can Society of Clinical Oncology Guide- line. J Clin Oncol 2016;34:3069-103.
Cardoso F, Senkus E, Costa A, et al. 4th ESO-ESMO international consensus guidelines for advanced breast cancer (ABC 4). Ann Oncol 2018;29:1634-57.
Shah PD, Dickler MN. Endocrine ther- apy for advanced breast cancer. Clin Adv Hematol Oncol 2014;12:214-23.

Liu CY, Wu CY, Petrossian K, Huang TT, Tseng LM, Chen S. Treatment for the endocrine resistant breast cancer: current options and future perspectives. J Steroid Biochem Mol Biol 2017;172:166-75.
Fritsch C, Huang A, Chatenay-Rivau- day C, et al. Characterization of the novel and specific PI3Kα inhibitor NVP-BYL719 and development of the patient stratifica- tion strategy for clinical trials. Mol Can- cer Ther 2014;13:1117-29.
Juric D, Rodon J, Tabernero J, et al. Phosphatidylinositol 3-kinase α-selective inhibition with alpelisib (BYL719) in PIK3CA-altered solid tumors: results from the first-in-human study. J Clin Oncol 2018;36:1291-9.
Miller TW, Hennessy BT, González- Angulo AM, et al. Hyperactivation of phos- phatidylinositol-3 kinase promotes escape from hormone dependence in estrogen receptor-positive human breast cancer. J Clin Invest 2010;120:2406-13.
Bosch A, Li Z, Bergamaschi A, et al. PI3K inhibition results in enhanced estro- gen receptor function and dependence in hormone receptor-positive breast cancer. Sci Transl Med 2015;7:283ra51.
Juric D, Janku F, Rodón J, et al. Alpelisib plus fulvestrant in PIK3CA- altered and PIK3CA-wild-type estrogen receptor-positive advanced breast can- cer: a phase 1b clinical trial. JAMA On- col 2018 December 13 (Epub ahead of print).
Di Leo A, Johnston S, Lee KS, et al. Buparlisib plus fulvestrant in postmeno- pausal women with hormone-receptor- positive, HER2-negative, advanced breast cancer progressing on or after mTOR in- hibition (BELLE-3): a randomised, double-

blind, placebo-controlled, phase 3 trial. Lancet Oncol 2018;19:87-100.
Baselga J, Im SA, Iwata H, et al. Bupar- lisib plus fulvestrant versus placebo plus fulvestrant in postmenopausal, hormone receptor-positive, HER2-negative, advanced breast cancer (BELLE-2): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2017;18:904-16.
Baselga J, Dent S, Cortés J, et al. Phase III study of taselisib (GDC-0032) + fulves- trant (FULV) in patients (pts) with estro- gen receptor (ER)-positive, PIK3CA-mutant (MUT), locally advanced or metastatic breast cancer (MBC): primary analysis from SANDPIPER. J Clin Oncol 2018;36: Suppl:LBA 1006. abstract.
Juric D, Ciruelos E, Rubovszky G, et al. Alpelisib + fulvestrant for advanced breast cancer: subgroup analyses from the phase III SOLAR-1 trial. Presented at the San An- tonio Breast Cancer Symposium, San An- tonio, TX, December 4–8, 2018. abstract.
Turner NC, Neven P, Loibl S, André F. Advances in the treatment of advanced oestrogen-receptor-positive breast cancer. Lancet 2017;389:2403-14.
O’Leary B, Cutts RJ, Liu Y, et al. The genetic landscape and clonal evolution of breast cancer resistance to palbociclib plus fulvestrant in the PALOMA-3 trial. Cancer Discov 2018;8:1390-403.
Vora SR, Juric D, Kim N, et al. CDK 4/6 inhibitors sensitize PIK3CA mutant breast cancer to PI3K inhibitors. Cancer Cell 2014;26:136-49.
Juric D, Castel P, Griffith M, et al. Convergent loss of PTEN leads to clinical resistance to a PI(3)Kα inhibitor. Nature 2015;518:240-4.
Copyright © 2019 Massachusetts Medical Society.

SPECIALTIES AND TOPICS AT NEJM.ORG
Specialty pages at the Journal’s website (NEJM.org) feature articles in cardiology, endocrinology, genetics, infectious disease, nephrology, pediatrics, and many other medical specialties.

1940

n engl j med 380;20

nejm.org

may 16, 2019