RSS-Feed abonnieren
DOI: 10.1055/s-0040-1704644
A SYSTEMATIC REVIEW AND META-ANALYSIS OF ENDOSCOPIC AND SURGICAL RESECTION FOR AMPULLARY LESIONS
Publikationsverlauf
Publikationsdatum:
23. April 2020 (online)
Aims Ampullary lesions (ALs) can be treated by endoscopic-ampullectomy (EA), surgical-ampullectomy (SA) or pancreaticoduodenectomy (PD). EA reveals high risk of incomplete resection and surgical interventions lead to substantial morbidity and mortality. To date, there are no prospective comparative trials. We performed a systematic review and meta-analysis to analyze complete resection (R0), adverse events (AEs) and recurrence between EA, SA and PD.
Methods Electronic databases (Medline, EMBASE, SCOPUS) were searched for publications analyzing ALs from 1990 to 2018. Studies that provided data for R0, AEs and/or recurrence were eligible. PRISMA guidelines were followed and papers were evaluated by Newcastle-Ottawa-Scale. R0, AEs and recurrence were pooled by means of a fixed and random-effects model. Proportions were transformed by Freeman-Tukey-Double-Arcsine-Proportion-model and compared by students’ t-test.
Results We identified 59 independent studies. The pooled R0 was 76.6% (CI:71.8%-81.4%, I2=91.38%, p< 0.001) for EA, 96.4% (CI:93.6%-99.2%, I2=37.8%, p=0.107) for SA and 98.9% (CI:98.0%-99.7%, I2=0%, p=0.531) for PD. AEs were 24.7% (CI:19.8%-29.6%, I2=86.4%, p< 0.001) for EA, 28.3% (CI:19.0%-37.7%, I2=76.8%, p< 0.001) for SA and 44.7% (CI:37.9%-51.4%, I2=0%, p=0.653) for PD. Recurrences were registered in EA in 13.0% (CI:10.2%-15.6%, I2=91.3%, p< 0.001), in SA in 9.4% (CI:4.8%-14%, I2=57.3%, p=0.007) and in PD in 14.2% (CI:9.5%-18.9%, I2=0%, p=0.330). Differences between proportions were significant in R0 for EA compared to SA (p=0.007) and PD (p=0.022), for complications between EA and PD (P=0.049) and not for recurrence.
Conclusions Our data indicate an enhanced rate of complete resection in surgical interventions but accompanied with clearly higher risk of complications. Nevertheless, studies showed various sources of bias, limited quality and a significant homogeneity, particularly in EA studies. High quality studies are necessary to determine the standard in therapy for AL.
#