Rationale for Requested Change: The clinical validity of FoundationOne Liquid CDx assay as a companion diagnostic device for identifying PIK3CA mutations in locally advanced or metastatic breast cancer patients who may benefit from treatment with inavolisib in combination with palbociclib and fulvestrant was established using clinical data from the INAVO120 clinical trial. [1-3]
INAVO120 (WO41554/NCT04191499) is a randomized, double-blind, placebo-controlled trial evaluating the efficacy of the triplet combination of inavolisib plus palbociclib and fulvestrant versus placebo plus palbociclib and fulvestrant in patients with PIK3CA-mutated, Hormone Receptor-Positive (HR+), HER2-negative (HER2-) locally advanced or metastatic breast cancer, who have not received prior systemic therapy for locally advanced or metastatic disease and whose disease progressed during or within 12 months of completing adjuvant endocrine therapy. PIK3CA mutation (PIK3CAm) status was prospectively determined in a central laboratory using the FoundationOne Liquid CDx assay, a ctDNA NGS-based comprehensive genomic profiling (CGP) assay, on plasma-derived ctDNA or in local laboratories using various validated PCR or NGS assays on tumor tissue or plasma. All patients were required to provide both a freshly collected pre-treatment blood sample and a tumor tissue sample for central evaluation (FoundationOne Liquid CDx assay) and determination of PIK3CAm status.
The major efficacy outcome measure was investigator (INV)-assessed progression-free survival (PFS) per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Additional efficacy outcome measures included overall survival (OS), INV-assessed objective response rate (ORR), and INV-assessed duration of response (DOR).
A total of 325 patients were included in the New Drug Application (NDA) efficacy population for inavolisib or placebo plus palbociclib and fulvestrant. Of the 325 patients from the INAVO120 NDA population (after excluding 43 patients from China who were not available for F1LCDx testing), 1.4% (4/282) of patients had unevaluable F1LCDx test results and 98.6% (278/282) of patients had evaluable F1LCDx test results, including 95.4% (269/282) that were positive for one or more study-eligible PIK3CA mutation(s) by F1LCDx (i.e., F1LCDx-positive) and 3.2% (9/282) with no PIK3CA mutation detected (i.e., F1LCDx-negative). One additional F1LCDx-positive patient was excluded from the efficacy analyses due to ultimately not receiving their planned treatment with inavolisib in combination with palbociclib and fulvestrant.
In the efficacy analysis for the NDA population, the median PFS (months) with 95% 2-sided CI for the inavolisib cohort was 15.0 [11.3, 20.5] months, and the median PFS for the placebo cohort was 7.3 [5.6, 9.3] months. The PFS hazard ratio (HR) for inavolisib in combination with palbociclib and fulvestrant vs. placebo in combination with palbociclib and fulvestrant and the associated 2-sided 95% CI was estimated as 0.43 [0.32, 0.59].
The clinical validity of F1LCDx was demonstrated by assessing clinical efficacy in the F1LCDx PIK3CA mutation-positive population based on progression free survival (PFS) as the primary efficacy endpoint. The efficacy analysis dataset consists of PIK3CA mutation(s) positive, HR+, HER2- locally advanced or metastatic breast cancer patients from the NDA efficacy population, enrolled by the central F1LCDx assay or a local assay (and confirmed PIK3CA-positive by F1LCDx).
For each treatment arm, Kaplan-Meier curves were used to estimate the time-to-event distributions. The 50th percentile of Kaplan-Meier estimates were used to estimate the median duration of PFS and are reported with a 2-sided 95% CI in Table 81 (section 10.13 in F1LCDx Technical Information).[3] The HR was calculated using a Cox proportional hazards model.
Table 81. Summary of PFS for Each Treatment Group for F1LCDx-positive Population*
*Treatment groups were defined by actual received treatment indicator (TRT01A) for the clinical validation study.
Total PIK3CAm population (n=325)
F1LCDx+ PIK3CAm population (n=268)
Efficacy Endpoint
ITOVEBITM + Palbociclib + Fulvestrant
N=161
Placebo + Palbociclib + Fulvestrant
N=164
ITOVEBITM + Palbociclib + Fulvestrant
N=137
Placebo + Palbociclib + Fulvestrant
N=131
Progression-Free Survivala,b
Patients with event, n (%)
82 (51)
113 (69)
68
92
Median, months (95% CI)
15.0 (11.3, 20.5)
7.3 (5.6, 9.3)
16.6 (13.4, 24.2)
7.3 (5.8, 10.1)
Hazard ratio (95% CI)
0.43 (0.32, 0.59)
0.42 (0.30, 0.59)
CI = confidence interval; CR = complete response; PR = partial response
a Per RECIST version 1.1.
b Based on investigator assessment.
Compared to the placebo cohort, in which the median PFS was 7.3 [5.8, 10.1] months, the median PFS for the inavolisib treatment cohort was 16.6 [13.4, 24.2] months. The PFS hazard ratio (HR) for inavolisib in combination with palbociclib and fulvestrant vs. placebo in combination with palbociclib and fulvestrant and the associated 2-sided 95% CI was estimated as 0.42 [0.30, 0.59], which were comparable to that in the total PIK3CA positive population.
A sensitivity analysis was performed to assess the robustness of the clinical efficacy estimated after accounting for the F1LCDx-unevaluable (failed FoundationOne Liquid CDx testing or not tested by FoundationOne Liquid CDx) results. Following imputation, the estimated HR for PFS based on the Cox proportional hazards model using the complete dataset was 0.44 (95%CI: [0.43, 0.44]) in the PIK3CA+ patients as determined by FoundationOne Liquid CDx testing. Comparable results were seen with those from the observed dataset, which demonstrated the efficacy results were robust after accounting for the missing FoundationOne Liquid CDx results.
