Background: Immunotherapy is a developing paradigm in the treatment of head and neck squamous cell carcinoma (HNSCC). Most notably, blocking programmed cell death protein 1 (PD-1) has shown early promise as a single-agent in second-line, platinum resistant HNSCC (see for example: KEYNOTE-055). In effort to improve the outcome of PD-1 blockade, clinical trials are testing combinations of immunotherapy and standard of care drugs. Indeed, an unreported clinical trial (KEYNOTE-048) is one example combining a PD-1 checkpoint inhibitor (pembrolizumab) + platinum (Pt) + 5-fluorouricil (5-FU) versus cetuximab + Pt + 5-FU (EXTREME regimen) as a frontline treatment for patients with recurrent or metastatic HNSCC. The lack of effective pre-clinical models to profile immunotherapy has widened the search for effective combinations.

Methods: Here, with informed consent under IRB, fresh tumor biopsies were obtained from a randomly-selected cohort of HNSCC patients diagnosed with metastatic or locally advanced disease (N=39). We then employed CANscriptTM, a clinically-validated ex-vivo tumor platform that preserves the native, patient tumor and immune microenvironment. Using this platform we interrogated the following treatment schedules in every sample: (i) pembrolizumab as single agent, (ii) combination of pembro + Pt + 5-FU and (iii) EXTREME regimen. Indeed, these represent the same treatment arms from the yet unreported KEYNOTE-048. Next, using a strategy that integrates phenotypic reflex to drug, ex-vivo, with a proprietary algorithm-driven approach, we predicted clinical response to each treatment arm (M-Score). In addition to this, PDL-1 status was determined for every patient by immunohistochemistry, and RNA gene expression was performed to interrogate transcriptomic differences induced by each treatment.

Results: We predicted, based on M-Score, that single agent pembrolizumab results in an overall response rate of 12.8% (5/39), which closely matched previous single agent results from KEYNOTE-040 (14.6%). In contrast, we determined the EXTREME regimen resulted in a response rate of 30.7% (12/39), a finding consistent with reported clinical data (NCT00122460; ORR 36%). Interestingly, we predicted no benefit is achieved with the combination pembro/Pt/5-FU compared to EXTREME (30.7% vs. 30.7%). However, when samples were interpreted based on the type of platinum used (Carboplatin vs. Cisplatin) we determined that greater clinical response is achieved with the combination of Pembro/Carbo/5-FU (43.7% or 7/16) vs. Pembro/Cis/5-FU (21.7% or 5/23). In contrast, we predicted EXTREME regimen alone results in similar response rates regardless of the addition of Carbo or Cis (31.3% vs. 30.4%, respectively).

Conclusions: Here we prognosticate clinical trial results yet to be reported using an ex-vivo platform. Many of the retrospective clinical evidence closely matches our data. If prediction for response to the novel combination strategy in KEYNOTE-048 correlates, this would provide the first evidence that a human tumor model can be used to effectively profile immunotherapy response, while also providing translational insights into mechanisms of response and resistance.

Citation Format: Saravanan Thyiagarajan, Biswanath Majumder, Abhisekh Basu, Padhma Radhakrishnan, Manjusha Biswas, Munisha Smalley, Hans Gertje, Parker Cassidy, Mark Lawson, KS Sabitha, Govind Babu, Aaron J. Goldman. An ex-vivo tumor platform used to predict clinical trial efficacy: Pembrolizumab+Pt+5-FU vs. EXTREME in recurrent head and neck squamous cell carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr LB-286.