Safety and efficacy of time restricted eating (TRE) in improving response to immunotherapy in patients with head and neck cancer (HNSCC).

Abstract

6051 Background: Advanced head and neck cancer continues to have very poor prognosis. Only a fraction of patients respond to immune checkpoint blocker therapy (ICB), with no effective treatment options for non-responders. Gut microbiome composition and diversity have been associated with response to ICB in many cancers including HNSCC. We, therefore hypothesized that modulation of gut microbiota may improve response to ICB. Time restricted eating represents an attractive strategy to favorably alter the microbiome and improve ICB responses. We conducted a phase I/II nonrandomized study to evaluate safety, feasibility and efficacy of TRE in improving response rates in treatment naïve metastatic HNSCC patients receiving single agent ICB. Methods: TRE is a form of circadian aligned nightly fast of 14 hours without reduction in calories. Patients were assigned to TRE vs control in 1:2 ratio, with first response evaluation at 12+/- 3 weeks as end point. Compliance to TRE was monitored via an app (“my circadian app”) and daily food logs. Blood and stool samples were collected at therapy initiation, at 1st response assessment and at 6 months. We evaluated stool metagenomics, untargeted stool/plasma metabolomics and serum cytokines as correlatives. We focused on microbial metabolites including indoles and short chain fatty acids (SCFA), which are potent aryl hydrocarbon receptor agonist and immunosuppressive. We also evaluated serum cytokines associated with insulin signaling pathways. Results: A total of 43 HNSCC patients receiving single agent pembrolizumab were accrued, 15 in the TRE arm and 28 in control arm. TRE was well tolerated without any study dropouts, weight loss or any adverse events attributable to the intervention. TRE was associated with significant improvement in disease control rate (DCR) at 3 months and at 6 months (3 month DCR 85% in TRE vs 50% in the control arm, p<0.0001, 6 month DCR 64% in TRE vs 39% in control arm, p=0.0004). Responders in both TRE and control arm had lower relative abundance of microbial immunosuppressive indole metabolites, butyric acid (NR vs R: 0.67 vs 0.58, p=0.05) and IGF1 (NR vs R: 0.42 vs 0.39, p=0.02). TRE led to a similar change in microbial metabolites including decrease in abundance of 5 Hydroxyindole acetic acid, Indole 3 acetic acid (IAA), Indole Carboxaldehyde, Indole acrylic acid, butyric acid (control vs TRE: 0.68 vs 0.58, p=0.03) and IGF1 (control vs TRE: 0.4 vs 0.3, p=0.03.) Conclusions: We report a novel phase I/II study of TRE in HNSCC patients and demonstrate excellent safety, feasibility and marked improvement in efficacy of ICB. The effect of TRE may be mediated via changes in microbial metabolome and insulin signaling pathways. TRE represents an attractive and safe strategy to improve treatment responses to ICB and needs to be explored in phase III expansion studies. Humaira Sarfraz, MD, and Shahla Bari, MD, both 1st authors/equal contributors. Clinical trial information: NCT05083416 .