Background: Evidence on the efficacy of alpelisib in combination with fulvestrant after progression on CKD4/6 inhibitors (CDK4/6i) is derived from a single non-comparative prospective study. Conversely, the effectiveness of everolimus plus exemestane on PIK3CA-mutant metastatic breast cancer (BC) after CDK4/6i failure has never been investigated in a prospective study. In this retrospective study, we compared alpelisib plus endocrine therapy (ET) with everolimus plus exemestane in patients with PIK3CA-mutant metastatic BC post-CDK4/6i progression. Methods: We tracked 40 patients treated with alpelisib plus ET and 22 treated with everolimus plus exemestane. We further identified 42 patients who did not harbor PIK3CA mutations (PIK3CA-wild-type group) and received everolimus as a subsequent treatment after progression on CDK4/6i. The timeframe spanned from 1st March 2020 to 30th November 2024. Results: The median progression-free survival (PFS) for the alpelisib group was 4.9 months compared to 4.5 months for the everolimus group [Hazard ratio (HR), 1.22; 95% CI, 0.65–2.28; p-value = 0.53]. The median overall survival (OS) was 9.6 months and 18.3 months for alpelisib and everolimus, respectively (HR, 0.67; 95% CI, 0.25–1.76; p-value = 0.47). Median PFS in the PIK3CA-mutant everolimus plus ET group was 4.5 months (95% CI, 2.8–6.7) compared to 5 months (95% CI, 3.5–6.9) for the PIK3CA-wild-type everolimus plus ET group (HR, 0.77; 95% CI, 0.46–1.29; p-value = 0.32). The most common side effects in the alpelisib group were hyperglycemia (57.5%), rash (27.5%), and anorexia (22.5%), while the most common side effects in the everolimus group were fatigue (40.9%) and stomatitis (27.3%). Conclusions: Our results regarding the efficacy of alpelisib plus ET were inferior to those reported in the current literature. Conversely, outcomes of everolimus plus exemestane were consistent with the current literature, denoting that the combination is an acceptable treatment option for patients with PIK3CA-mutant metastatic BC.

With the advent of modern statistical software, complex experimental designs are now routinely employed in many areas of research. Failing to correctly identify the structure of the experimental design can lead to incorrect model selection and misleading inferences. This paper describes the hassediagrams package in R that determines the structure of the design, summarised by the layout structure, and generates a Hasse diagram of the layout structure. By considering the randomisation performed, in conjunction with the layout structure, a set of randomisation objects can be defined that form the restricted layout structure. This structure can also be visualised using a generalisation of the Hasse diagram. Objects in the restricted layout structure can be used to identify the terms to include in the statistical model. The use of the procedure thus ensures consistency of model selection due to the systematic approach taken to generate the model.