Safety and Survival Rates Associated With Ablative Stereotactic Radiotherapy for Patients With Oligometastatic Cancer: A Systematic Review and Meta-analysis

Abstract

Importance: The oligometastatic paradigm postulates that patients with a limited number of metastases can be treated with ablative local therapy to each site of disease with curative intent. Stereotactic ablative radiotherapy (SABR) is a radiation technique that has become widely used in this setting. However, prospective data are limited and are mainly from single institutional studies.

Objective: To conduct a meta-analysis to characterize the safety and clinical benefit of SABR in oligometastatic cancer.

Data sources: A comprehensive search was conducted in PubMed/MEDLINE, Embase, Cochrane Database of Systematic Reviews, and Cumulative Index to Nursing and Allied Health Literature on December 23, 2019, that included prospective clinical trials and review articles that were published within the past 15 years.

Study selection: Inclusion criteria were single-arm or multiarm prospective trials including patients with oligometastatic cancer (ie, ≤5 sites of extracranial disease), and SABR was administered in less than or equal to 8 fractions with greater than or equal to 5 Gy/fraction.

Data extraction and synthesis: The Population, Intervention, Control, Outcomes and Study Design; Preferred Reporting Items for Systematic Reviews and Meta-analyses; and Meta-analysis of Observational Studies in Epidemiology methods were used to identify eligible studies. Study eligibility and data extraction were reviewed by 3 authors independently. Random-effects meta-analyses using the Knapp-Hartung correction, arcsine transformation, and restricted maximum likelihood method were conducted.

Main outcomes and measures: Safety (acute and late grade 3-5 toxic effects) and clinical benefit (1-year local control, 1-year overall survival, and 1-year progression-free survival).

Results: Twenty-one studies comprising 943 patients and 1290 oligometastases were included. Median age was 63.8 years (interquartile range, 59.6-66.1 years) and median follow-up was 16.9 months (interquartile range, 13.7-24.5 months). The most common primary sites were prostate (22.9%), colorectal (16.6%), breast (13.1%), and lung (12.8%). The estimate for acute grade 3 to 5 toxic effect rates under the random-effects models was 1.2% (95% CI, 0%-3.8%; I2 = 50%; 95% CI, 3%-74%; and τ = 0.20%; 95% CI, 0.00%-1.43%), and the estimate for late grade 3 to 5 toxic effects was 1.7% (95% CI, 0.2%-4.6%; I2 = 54%; 95% CI, 11%-76%; and τ = 0.25%; 0.01%-1.00%). The random-effects estimate for 1-year local control was 94.7% (95% CI, 88.6%-98.6%; I2 = 90%; 95% CI, 86%-94%; and τ = 0.81%; 95% CI, 0.36%-2.38%]). The estimate for 1-year overall survival was 85.4% (95% CI, 77.1%-92.0%; I2 = 82%; 95% CI, 71%-88%; and τ = 0.72%; 95% CI, 0.30%-2.09%) and 51.4% (95% CI, 42.7%-60.1%; I2 = 58%; 95% CI, 17%-78%; and τ = 0.20%; 95% CI, 0.02%-1.21%) for 1-year progression-free survival.

Conclusions and relevance: In this meta-analysis, SABR appears to be relatively safe in patients with oligometastatic cancer with clinically acceptable rates of acute and late grade 3 to 5 toxic effects less than 13% and with clinically acceptable rates of 1-year local control overall survival, and progression-free survival. These findings are hypothesis generating and require validation by ongoing and planned prospective clinical trials.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Trifiletti reported receiving support from Novocure for clinical trial research and personal fees from Springer Nature Inc. Dr Siva is supported by a National Health and Medical Research Council fellowship. Dr Zaorsky is supported by the National Institutes of Health LRP 1 L30 CA231572-01. Dr Zaorsky received personal fees from Springer Nature Inc and Weatherby Healthcare. No other disclosures were reported.

Figures

Figure 1.. Preferred Reporting Items for Systematic Reviews and Meta-analyses Flow Diagram

Search methods and screening process used to screen and select eligible articles. There were initially 442 articles screened; 21 articles were eligible for inclusion in the meta-analysis. LC indicates local control; OS, overall survival; PFS, progression-free survival; SABR indicates stereotactic ablative radiotherapy.

Figure 2.. Safety

A, Weighted random-effects model depicting acute grade 3 to 5 toxic effects for 12 studies. In the acute grade 3 to 5 toxic effects funnel plot, each blue circle represents 1 of the 12 included studies, with the study effect size proportions on the x-axis and corresponding SE on the y-axis. Publication bias was detected with P < .001 for the weighted linear regression test. B, Weighted random-effects model depicting late-grade 3 to 5 toxic effects for 12 studies. In the late grade 3 to 5 toxic effect funnel plot, each blue circle represents 1 of the 12 included studies, with the study effect size proportions on the x-axis and corresponding SE on the y-axis. Publication bias was not detected with P = .39 for the weighted linear regression test. The number of cases in each forest plot was rounded to the nearest whole number. fx indicates fraction; NR, not reported; NSCLC, non–small cell lung cancer; and Ph, phase.

Figure 3.. Clinical Benefit

A, Weighted random-effects model depicting 1-year local control (LC) for 13 studies. In the 1-year LC funnel plot, each blue circle represents 1 of the 13 included studies, with the study effect size proportions on the x-axis and corresponding SE on the y-axis. Publication bias was not detected with P = .45 for the weighted linear regression test. B, Weighted random-effects model depicting 1-year overall survival (OS) for 15 studies. In the 1-year OS funnel plot, each blue circle represents 1 of the 15 included studies, with the study effect size proportions on the x-axis and corresponding SE on the y-axis. Publication bias was not detected with P = .87 for the weighted linear regression test. The number of cases in each forest plot was rounded to the nearest whole number. fx indicates fraction; NR, not reported; NSCLC, non–small cell lung cancer; and Ph, phase.