Meta-analysis of dose-fractionation radiotherapy trials for the palliation of painful bone metastases.

PURPOSE: To compare pain relief among various dose-fractionation schedules of localized radiotherapy (RT) in the treatment of painful bone metastases.

METHODS AND MATERIALS: A systematic search for randomized trials of localized RT on bone metastases using different dose fractionations was performed using Medline (1966 to February 2001) and other sources. The primary outcomes of interest were complete and overall pain relief. The studies were divided into three groups: comparisons of doses given as a single fraction, single vs. multiple fractions, and comparisons of doses given as multiple fractions. The complete and overall pain responses for studies comparing single vs. multiple fractions were pooled. Exploratory analyses of the dose-response relationship, using the biologic effective dose (alpha/beta = 10), were performed using results from all three groups of trials.

RESULTS: Two trials comparing single vs. single, eight trials comparing single vs. multiple, and six trials comparing multiple vs. multiple fractions were included. The complete and overall response rates from studies comparing single-fraction RT (median 8 Gy, range 8-10 Gy) against multifraction RT (median 20 Gy in 5 fractions, range 20 Gy in 5 fractions to 30 Gy in 10 fractions) were homogeneous and allowed pooling of data. Of 3260 randomized patients in seven studies, 539 (33.4%) of 1613 and 523 (32.3%) of 1618 patients achieved a complete response after single and multifraction RT, respectively, giving a risk ratio of 1.03 (95% confidence interval 0.94-1.14; p = 0.5). The overall response rate was in favor of single-fraction RT (1011 [62.1%] of 1629) compared with multifraction (958 [58.7%] of 1631; risk ratio 1.05, 95% confidence interval 1.00-1.11, p = 0.04), reaching statistical significance. However, when the analysis was restricted to evaluated patients alone, the overall response rates were similar for single fraction and multifraction RT, at 1011 (72.7%) of 1391 and 958 (72.5%) of 1321, respectively (risk ratio 1.00; p = 0.9). Exploratory analyses by biologic effective dose did not reveal any dose-response relationship among the fractionation schedules used (single 8 Gy to 40 Gy in 15 fractions). Of the other results and observations reported in the trials, only the re-irradiation rates were consistently different between the treatment arms (more frequent re-irradiation in lower dose arms among trials reporting re-irradiation rates).

CONCLUSION: Meta-analysis of reported randomized trials shows no significant difference in complete and overall pain relief between single and multifraction palliative RT for bone metastases. No dose-response relationship could be detected by including data from the multifraction vs. multifraction trials. Additional data are needed to evaluate the role of re-irradiation and the impact of RT on other treatment end points such as quality of life.