Response to Immune Checkpoint Inhibitors in Recurrent/metastatic Esophageal Squamous Cell Carcinoma May Be Affected by Tumor Site and Lesion Size.

Jhe-Cyuan Guo,Chia-Chi Lin,Ta-Chen Huang,Min-Shu Hsieh,Li-Chun Lu,Chih-Hung Hsu
DOI: https://doi.org/10.1200/jco.2018.36.15_suppl.e16099
IF: 45.3
2018-01-01
Journal of Clinical Oncology
Abstract:e16099 Background: Heterogeneous tumor response has been reported in cancer patients receiving immune checkpoint inhibitors (ICIs). This study investigated whether or not tumor site and tumor size significantly affect the response to ICIs in patients with recurrent/metastatic esophageal squamous cell carcinoma (R/M ESCC). Methods: R/M ESCC patients, with measurable disease according to RECIST 1.1 and receiving anti-PD-1/PD-L1-based therapies at National Taiwan University Hospital, were enrolled. To evaluate the tumor response at different organ sites, we used an organ-specific response criteria that was modified from RECIST 1.1 by allowing a maximum of 5 measurable lesions per organ system. To evaluate the tumor response per individual tumor basis, we studied all measurable tumors and evaluate their individual tumor response following the definition of response of RECIST 1.1. The significance of tumor sites and target tumor sizes on the response were examined by logistic regression. Results: Thirty-eight R/M ESCC (M:F = 37: 1; median age of 58.0 years) were enrolled. The overall tumor response rate (RR) per RECIST 1.1 was 13.2%. Measurable tumors in lymph node (LN), lung, and liver were seen in 23, 16, and 10 patients, respectively. The organ-specific RRs were 30.4%, 31.3%, and 10.0% for LN, lung, and liver tumors, respectively (p= 0.040). Fifty-three, 34, and 32 measurable target lesions from LN, lung, and liver, respectively, were subject to individual tumor response evaluation. The mean sizes were 2.42, 1.79, and 4.22 cm for LN, lung, and liver target lesions, respectively (p< 0.001). In univariate logistic regression, tumor site (LN vs liver) was statistically significantly associated with response (p= 0.007) and lesion size (increased every 1 cm) trended to associate with response (p= 0.094). Conclusions: Our data suggest that response to ICIs is significantly different among R/M ESCC tumors at different organ sites. The underlying mechanism warrants further investigation. (Supported by the grants of MOST 105-2314-B-002 -186 -MY3, MOST 106-2314-B-002 -135 -, and 106-N3695)
What problem does this paper attempt to address?