A Literature Review of Difficult Gallstone Treatment: Results, Critical Analysis, and Limits of Subgroup Analysis

OBJECTIVES: Peroral cholangioscopy enables the use of laser (LL) or electrohydraulic (EHL) lithotripsy to treat difficult gallstones. With the future objective of incorporating cholangiocopy-guided LL/EHL into the Brazilian health system, we conducted a literature review comparing effectiveness and safety of the latter vs. conventional treatments (ERCP and EPLBD) to treat difficult gallstones.

METHODS: Research until April 2024 was carried out at online databases, obtaining 1,233 titles (duplicates excluded), of which 40 were read in full. Risk of bias and quality of evidence by outcome was assessed through ROBINS-1 (RCTs), AMSTAR-2 (SLR with meta-analysis) and GRADE accordingly. Results were critically analyzed.

RESULTS: 4 RCTs and 2 SRL with meta-analysis were selected. According to the RCTs, successful clearance of difficult gallstones varied from 77.1% to 100% for cholangioscopy, and from 67% to 72% for conventional treatment (high quality evidence). Franzinni, 2019 (RCT) also showed a 23.4% increase in success rate for post-failed EPBL cholangioscopy. Regarding RSL with meta-analysis, Facciorusso 2023 showed that cholangioscopy with direct visualization resulted in statistically higher stone clearance rates compared to conventional ERCP (p = 0.006). In contrast, Galetti 2020 stated there is no superior treatment. The least study initially included 3 RCTs but drew its conclusion from a 2-RCTs subgroup analysis after removing Angsuwatcharakon 2019. Controversially, the excluded study was rated at the highest risk of bias among the 3 RCTs, and the authors themselves recognize key points that makes it more in line with current times than the others.

CONCLUSIONS: The results obtained in the subgroup analysis were erroneously used as a conclusion by Galetti 2020, making it biased. Subgroup analysis should be used to understand heterogeneity and not to make emphatic statements. This situation reinforces the need to integrate statistics with critical analysis to better understand and represent a clinical context.