Stereotactic body radiotherapy (SBRT) for locally advanced pancreatic cancer : a systematic review and pooled analysis of 19 trials
A partir d'une revue systématique de la littérature (19 essais incluant au total 1 009 patients), cette méta-analyse évalue l'efficacité, du point de vue de la survie globale à 1 an, et la toxicité d'une radiothérapie corporelle stéréotaxique chez les patients atteints d'un cancer du pancréas non résécable ou à la limite de la résécabilité et de stade localement avancé
Introduction : While surgery is the standard of care for resectable pancreatic cancer (PC), standard dose chemoradiotherapy (CTRT) or CT alone are suitable for patients with unresectable disease. Stereotactic body radiotherapy (SBRT) is an alternative focused local therapy that delivers high RT doses in a few fractions to the cancer, sparing surrounding critical tissue. We have performed a systematic review and pooled analysis of published trials to evaluate the efficacy and safety of this emerging treatment modality. Materials and methods : We searched the Cochrane Central Register of Controlled Trials, Pubmed, EMBASE, SCOPUS, the Web of Science, and CINAHL for publications regarding SBRT for locally advanced PC. Overall survival (OS: 1-year OS %) was the primary endpoint and median OS, 2-year OS, 1-year loco-regional control (LRC) and G3-4 toxicities the secondary endpoints. A multivariate random-effects meta-analysis was performed to calculate the aggregated OS rates at 1 and 2 years as well as the 1-year LRC. Results : A total of 19 studies, encompassing 1009 patients, were included in the analysis. The pooled 1-year OS was 51.6% in n=13 trials with data available. Median OS ranged from 5.7 to 47 months (median 17 months). The LRC at 1-year was 72.3%. Overall, the occurrence of severe adverse events did not exceed 10%. LRC appeared to be correlated with the total SBRT dose and the number of fractions. Conclusion : The advantages of SBRT in terms of treatment time, satisfactory OS, and LRC means that it is an effective option for inoperable PC. However, a definitive validation of this treatment modality in large randomized studies is required, due to the non-randomized nature of the included studies and the limitations of small single center series that include mixed populations.