<h3>Background</h3> In the multicohort phase 2 KEYVIBE-005 study (NCT05007106), the antitumor activity of vibostolimab (anti-T-cell immunoreceptor with Ig and ITIM domains [TIGIT]) coformulated with pembrolizumab (anti–PD-1; vibostolimab/pembrolizumab) was demonstrated in patients with previously treated advanced mismatch repair–deficient (dMMR) endometrial cancer (cohort B1; n = 40; ORR, 64%) and in patients with previously untreated advanced esophageal cancer treated with vibostolimab/pembrolizumab plus 5-fluorouracil and cisplatin (cohort E; n = 40; ORR, 53%). In this exploratory analysis, we evaluated the association between biomarkers and response to study treatment. <h3>Methods</h3> Evaluable pretreatment tumor samples from patients enrolled in cohorts B1 and E were analyzed. Immunohistochemistry was used to measure TIGIT (clone SP410, formulation locked assay) and PD-L1 (per combined positive score using PD-L1 IHC 22C3 pharmDx) expression. RNA NanoString was used to assess T-cell–inflamed gene expression profile (Tcell_infGEP),<i> TIGIT,</i> and poliovirus receptor (<i>PVR</i>) expression. A Spearman correlation was used to assess the relationship between the proportion of total immune cells (ICs) and TIGIT-positive ICs, and association was evaluated using a box plot of each biomarker by responder status. Area under the receiver operating characteristics curve (AUROC) was also estimated when sample size was sufficient. Clinical data cutoff was May 8, 2023. <h3>Results</h3> In cohort B1, TIGIT IHC was evaluable in 35 patients (88%), PD-L1 in 39 patients (98%), Tcell_infGEP in 20 patients (50%), and <i>TIGIT</i> RNA and <i>PVR</i> in 19 patients (48%). There was a correlation between total ICs and TIGIT-positive ICs (ρ = 0.62). The AUROC for discriminating TIGIT IHC, PD-L1, and Tcell_infGEP as predictors of response were 0.51 (95% CI, 0.27–0.76), 0.65 (0.44–0.85), and 0.38 (0.08–0.69), respectively. No clear relationships between <i>TIGIT</i> RNA and <i>PVR</i> and response were observed. In cohort E, TIGIT IHC was evaluable in 34 patients (85%), PD-L1 in 35 patients (88%), and Tcell_infGEP, <i>TIGIT</i> RNA, and <i>PVR</i> in 11 patients (28%). A correlation was observed between total ICs and TIGIT-positive ICs (ρ = 0.64). The AUROCs for discriminating TIGIT IHC and PD-L1 as predictors of response were 0.49 (95% CI, 0.27–0.70) and 0.62 (0.41–0.82), respectively. No relationships between Tcell_infGEP,<i> TIGIT</i> RNA, and <i>PVR</i> and response were observed. <h3>Conclusions</h3> We observed a trend between PD-L1 expression and response in patients with advanced dMMR endometrial cancer treated with vibostolimab/pembrolizumab as well as patients with advanced esophageal cancer treated with vibostolimab/pembrolizumab plus 5-fluorouracil and cisplatin. TIGIT IHC, Tcell_infGEP, <i>TIGIT</i> RNA<i>,</i> and <i>PVR</i> were not associated with response. <h3>Acknowledgements</h3> Medical writing and/or editorial assistance was provided by Mehak Aggarwal, PharmD, and Holly C. Cappelli, PhD, CMPP, of ApotheCom (Yardley, PA, USA). This assistance was funded by Merck Sharp &amp; Dohme LLC, a subsidiary of Merck &amp; Co., Inc., Rahway, NJ, USA. <h3>Trial Registration</h3> clinicaltrials.gov, NCT05007106. <h3>Ethics Approval</h3> The study protocol and all amendments were approved by the institutional review board or ethics committee at each institution. <h3>Consent</h3> All patients provided written informed consent.