Hideya Itagaki, Tomoyuki Endo
Gangrenous cholecystitis (GC) is a severe form of acute cholecystitis associated with perforation, sepsis, and increased mortality. Although computed tomography (CT) is useful for evaluating complicated cholecystitis, the relative diagnostic value of individual CT findings for identifying GC remains unclear. This review aimed to evaluate the association between individual CT findings and histopathologically confirmed GC and to summarize the diagnostic performance of these findings for preoperative assessment. We searched PubMed, Web of Science, Ichushi-Web, and the Cochrane Central Register of Controlled Trials from inception to March 1, 2026, and also screened Google Scholar for eligible studies. Eligible studies included those that enrolled patients with acute cholecystitis who underwent CT and had GC confirmed histopathologically. Two reviewers independently screened studies, extracted data, and assessed quality using the QUADAS-2 tool. Pooled odds ratios (ORs) were calculated with a DerSimonian-Laird random-effects model. When 2 × 2 data were available, we estimated pooled sensitivity and specificity using a bivariate random-effects model fitted by restricted maximum likelihood, from which we derived likelihood ratios and their confidence intervals; we examined model stability in sensitivity analyses. Ten retrospective studies involving 786 patients (310 GC and 476 non-GC) were included; eight studies (724 patients; 268 GC and 456 non-GC) contributed to the quantitative synthesis and two to the qualitative synthesis only. Absent gallbladder wall enhancement (OR 10.23, 95% CI 5.67-18.46) and wall irregularity (OR 10.26, 95% CI 5.80-18.17) were strongly associated with GC, as was pericholecystic stranding (OR 4.82, 95% CI 1.76-13.24). Absent wall enhancement (sensitivity 0.49, specificity 0.91, positive likelihood ratio 5.65) and wall irregularity (sensitivity 0.43, specificity 0.94, positive likelihood ratio 7.14) both combined high specificity with moderate sensitivity; their confidence intervals overlapped, and the findings were not compared directly. Pooled estimates were robust to the assumed between-study correlation and to omitting individual studies. Absent wall enhancement, wall irregularity, and pericholecystic inflammatory changes are useful CT findings for the preoperative assessment of suspected GC. Because the available evidence derives from surgically treated cohorts and the findings were not compared directly, these results should be regarded as hypothesis-generating and warrant prospective validation in unselected patients undergoing CT for suspected acute cholecystitis.