The diagnostic performance of probe-based confocal laser endomicroscopy in the detection of gastric cancer: a systematic review and meta-analysis
Annals of gastroenterology
Probe-based confocal laser endomicroscopy; gastric adenocarcinoma; gastric cancer; screening
Background: Gastric cancer (GC) represents a significant global health burden with high morbidity and mortality, especially when diagnosed at advanced stages. Therefore, early detection of GC is critical. Probe-based confocal laser endomicroscopy (pCLE) is a new evolving technology that uses real-time, high-resolution imaging to inspect the mucosa at the cellular and microvascular level, using a confocal probe. Widespread studies using pCLE are limited at the current time. We aimed to investigate the diagnostic efficacy of this modality for the detection of GC. Methods: Multiple databases were searched from inception until November 2021. The diagnostic performance of pCLE was assessed by calculating its sensitivity, specificity and accuracy for the detection of GC, using pooled proportions and 95% confidence intervals (CI) with a random-effects model. Heterogeneity was assessed using I. Results: Seven studies were included, with a total of 567 patients (mean age 61.7 years, 364 males). Pooled performance metrics of pCLE included a sensitivity of 87.9% (95%CI 81.4-92.4; P<0.001; I=0%), specificity 96.5% (95%CI 91.5-98.6; P<0.001; I=51.84%), and an accuracy of 94.7% (95%CI 89.5-97.4; P<0.001; I=65.44%). Conclusions: pCLE is a highly effective diagnostic modality for detecting GC. Larger, randomized controlled studies are needed to determine its role in daily practice compared to conventional endoscopic practices.
Canakis, Andrew; Deliwala, Smit S.; Kadiyala, Jagannath; Bomman, Shivanand; Canakis, Justin; and Bilal, Mohammad, "The diagnostic performance of probe-based confocal laser endomicroscopy in the detection of gastric cancer: a systematic review and meta-analysis" (2022). GW Authored Works. Paper 1682.
School of Medicine and Health Sciences Resident Works