Abemaciclib in recurrent/metastatic head and neck squamous cell carcinoma (RM-HNSCC) harboring CDKN2A loss, and/or CCND1 and/or CDK6 amplification: A phase II multicenter trial.

6044 Background: More than half of HPV negative HSNCC harbor genomic aberrations (i.e. CDKN2A loss, cyclin D amplification (ampl.)) that activate the cyclin-dependent kinase 4 and 6 (CDK4/6) – retinoblastoma protein (Rb) signalling pathway. Hyperactivated CDK4/6-Rb pathway leads to acceleration of G1/S transition of cell cycle and ultimately to uncontrolled cellular proliferation. We hypothesized that abemaciclib, a CDK4/6 inhibitor, could be a potent targeted strategy by inhibiting this commonly dysregulated pathway in HNSCC. Methods: This single arm, multicentre Phase II evaluated the efficacy of abemaciclib (200mg/d orally BID) in molecularly-selected, HPV negative (HPV - ), RM-HNSCC patients (pts) progressing after at least 1 prior line of platinum and cetuximab. A molecular screening step through centralised CGH-assay was required before treatment start: only pt with tumor harboring CDKN2A loss and/or CCND1 and/or CDK6 ampl. without homozygous deletion of RB1 were eligible to the therapeutic step. The primary endpoint was 8-week non-progression rate (PFR 8W ) as per central imaging review according to RECIST V.1.1. Secondary endpoints included best overall response rate (BORR), progression free survival (PFS), overall survival (OS) and adverse event (AE) according to NCI-CTCAE V5.0. The study used a Fleming A’Hern design with an inefficacy bound of 15% and a target PFR 8W of 40% (α = 5% one sided, 1-β = 90%). Results: Twenty-six HPV - RM-HNSCC pts (M: 23, F: 3, median age: 59 [range, 42-78], heavily pre-treated [84.6% ≥ 2 prior lines]) received at least one dose of abemaciclib. According to central CGH-assay, molecular alterations were CDKN2A loss + CCND1 ampl. + CDK6 ampl. (n = 1), CDKN2A loss + CCND1 ampl. (n = 10), CDKN2A loss (n = 6) or CCND1 ampl. (n = 9), all with intact Rb. Among the 24 evaluable pts, 7 pts were progression-free at 8 weeks (PFR 8w :29.2% (95% CI 14.6-)). BORR according to central imaging review was stable disease for 7 pts (31.8%) and progressive disease for 11 pts (50%); no objective response was observed. Median PFS and OS were 7.1 weeks (95% CI: 6.9-10.4) and 4.8 months (95% CI: 2.4-7.3), respectively. Most common related AEs (≥20% of pts, all grade) were diarrhea/nausea/vomiting, fatigue and hematological toxicity (anemia, lymphocyte and neutrophil count decreased). Two fatal serious AE potentially related to abemaciclib according to both investigator and sponsor were reported: a case of pulmonary hemorrhage also possibly related to abemaciclib-induced tumor necrosis and a case of myocardial infarction. Conclusions: Abemaciclib had limited antitumor activity in RM-HNSCC harboring molecular alteration in CDK4/6 pathway. Clinical trial information: NCT03356223 .