Dosimetric comparison of photon and particle radiotherapy devices for prostate cancer: a rapid systematic review with frequentist and Bayesian network meta-analysis
li y, xie r, lei l, li y, liu q, zhang h (+1 more)
Abstract
Abstract Background Multiple radiotherapy modalities are applied for prostate cancer treatment, yet their dosimetric performance of photon and particle beam techniques lacks systematic network meta-analysis (NMA) comparison. Methods This rapid systematic review performed frequentist and Bayesian NMA to rank six radiotherapy technologies: IMRT, VMAT/RapidArc, CyberKnife, Tomotherapy, IMPT and carbon ion radiotherapy (CIRT) based on dosimetric plan quality. We searched PubMed (2000–2026, English), enrolling studies comparing ≥ 2 modalities with extractable dosimetric data. Frequentist random-effects NMA was the primary analysis, and Bayesian NMA acted as sensitivity analysis; subgroup analysis was stratified by fractionation regimen. Results Nine studies containing 186 dosimetric data points were included, and 13 indicator-fractionation combinations within the SBRT subgroup supported network modeling. CIRT obtained top rankings for PTV conformity, homogeneity, V95, rectal and bladder Dmean but the worst rectal Dmax. CyberKnife performed best in PTV D98 and rectal Dmax; IMRT, VMAT and Tomotherapy had respective optimal dosimetric indicators. Frequentist and Bayesian results reached 97.9% directional agreement, with substantial heterogeneity (I²>80%). Conclusions CIRT provides superior target metrics and minimal mean OAR doses, but higher rectal Dmax requires caution. Findings should be interpreted considering limited particle therapy literature and high heterogeneity.