Background: In 11.2018, the PIK3CA-inhibitor alpelisib was made available in France through an early access program (EAP), in combination with fulvestrant in pre-treated PIK3CA-mutant, HR-positive, HER2-negative advanced breast cancer (ABC) patients. Patients had to received two or more prior systemic treatments for ABC, including an aromatase inhibitor and a CDK4/6 inhibitor in the absence of contraindications. This retrospective real-life, EAP-based study aimed to assess the efficacy and safety of alpelisib/fulvestrant combination in the post CDK4/6 inhibitor setting. Methods: The IRB-approved protocol and call for data were sent on 10.2020 to the cancer centers which participated the most in the EAP prospective registry. Eligible patients were women who started alpelisib/fulvestrant between 11. 2018 and 10.2020 as part of the EAP (which excluded patients with visceral crisis or inflammatory BC). Alpelisib and fulvestrant were used at standard doses. Primary endpoint was PFS by local investigators using RECIST1.1. Secondary endpoints included objective response rate and safety (NCI CTCAE v5.0). Results: 10 centers provided individual data regarding 209 consecutive patients. Patients had received a median number of 4 (1-14) previous systemic treatments for ABC, including CDK4/6 inhibitors, chemotherapy, fulvestrant (alone or in combination) and everolimus for 206 (98.8%), 159 (76.1%), 163 (78%) and 123 (58.8%) patients, respectively. With a median FU of 7.0 months, median PFS was 4.0 months (95%CI [3.5;5.0]) and 35.4% of 164 evaluable patients had an objective response. After stratification on the number of prior lines of treatment, prior exposure to everolimus had no impact on PFS (mPFS in the 123 patients pretreated with everolimus: 4.0m, 95%CI [3.5-5.5]). Of note, this population was enriched in patients who had a long disease control by everolimus (median time spent on everolimus: 7.0m, range (6.5-9.0)). In multivariable analysis, characteristics significantly associated with longer PFS were PS < 3 (HR = 0.03, 95%CI [0.02-0.29]) and prior treatment with fulvestrant (HR = 0.53, 95%CI [0.32-0.89]). N = 81(38.8%) patients discontinued alpelisib due to adverse events (AEs). Most frequent grade 3/4 AEs were hyperglycemia, skin rash, diarrhea and fatigue occurring in 13.4, 8.1, 4.8 and 1.9 % of patients, respectively. Conclusions: Despite heavy pre-treatments, alpelisib +fulvestrant had a clinically relevant efficacy in the French EAP population. Interestingly, prior treatment with either everolimus or fulvestrant did not overtly impair alpelisib-fulvestrant efficacy. The best treatment sequence for PI3KCA/mTOR inhibitors could be examined in future trials in PIK3CA-mutant ER+/HER2- ABC patients.