Diagnosis of invasion depth in resectable advanced gastric cancer for neoadjuvant chemotherapy: An exploratory analysis of Japan clinical oncology group study: JCOG1302A.

Background: Neoadjuvant chemotherapy (NAC) is increasingly used for resectable locally advanced gastric cancer (LAGC). JCOG1302A investigated the diagnostic criteria of LAGC patients with cT3-4/N1-3 to minimize contamination of pathological stage I as a candidate for NAC. In JCOG1302A, 77.2% of cT3-4 tumors diagnosed via a combination of endoscopy and computed tomography (CT) were pT3-4. However, the role of endoscopic ultrasonography (EUS) and additional diagnostic procedures/modalities remains unclear. Here, we investigated whether EUS, thin-slice CT, and foaming agent (FA) in CT contribute to accurate diagnosis of AGC invasion depth.

Methods: Using JCOG1302A study data, we compared positive predictive value (PPV), negative predictive value (NPV), and kappa index (KI) between conventional and additional diagnostic procedures to identify pT3-4: conventional endoscopy (CE) with versus without EUS, 1-mm versus 5-mm CT slice, and CT with versus without FA.

Results: We analyzed 1232 patients' data. PPV, NPV, and KI were 79.2%/73.7%, 59.2%/58.8%, and 0.38/0.39 (CE alone/CE with EUS), 77.8%/75.5%, 62.9%/71.2%, and 0.38/0.39 (5-mm CT/1-mm CT), and 78.6%/75.1%, 60.9%/69.7%, and 0.38/0.40 (CT without FA/CT with FA), respectively. Overall, there were no remarkable differences in any comparison. More specifically, PPV and KI were slightly higher with CE alone rather than CE with EUS. Although NPV was higher for 1-mm CT and CT with FA, PPV was rather higher for 5-mm CT and CT without FA.

Conclusion: Additional diagnostic procedures/modalities, like EUS, 1-mm slice CT, or FA in CT may not improve the diagnostic accuracy of invasion depth in resectable LAGC.