Background: Anti-cancer therapies lead to chronic non-resolving inflammation and reduced immune function. One potential therapy is exercise training, but the effectiveness of these interventions to improve immune-related outcomes, the gaps in the literature, and recommendations to progress the field need to be determined.

Objectives: (1) to conduct separate meta-analyses in cancer survivors to determine the effects of exercise training on pro- and anti-inflammatory markers, and immune cell proportions and function; and (2) to perform subgroup analyses to determine whether exercise modality, cancer type, and specific markers help to explain heterogeneity in each meta-analysis.

Data sources: Electronic databases (PubMed/MEDLINE, EMBASE, CENTRAL, and CINAHL) from inception to March 2018. The reference lists of eligible articles and relevant reviews were also checked.

Study selection: Inclusion criteria were adult cancer survivors from randomized controlled trials performing structured exercise intervention (aerobic, resistance or combined training or Tai Chi/yoga) compared to usual care control group and included pro-inflammatory, anti-inflammatory, and/or immune cell outcomes.

Appraisal and synthesis methods: A total of 5349 potentially eligible articles were identified, of which 26 articles (27 trials) met the inclusion criteria. Effect sizes were calculated as standardized mean differences (SMD), where <0.2 was defined as trivial, 0.2-0.3 as small, 0.4-0.8 as moderate, and >0.8 as a large effect.

Results: Exercise training decreased pro-inflammatory markers (SMD: -0.2, 95% CI: -0.4, -0.1, p < 0.001). Sub-group analysis for the pro-inflammatory markers indicated that combined aerobic and resistance training had the greatest effect (SMD: -0.3, 95% CI: -0.5, -1.9, p < 0.001), that prostate (SMD: -0.5, 95% CI: -0.8, 0.1, p = 0.004) and breast cancer populations were most responsive (SMD: -0.2, 95% CI: -0.3, -0.1, p = 0.001), and that C-reactive protein (SMD: -0.5, 95% CI: -0.9, -0.06, p = 0.025) and tumor necrosis factor (SMD: -0.3, 95% CI: -0.5, -0.06, p = 0.004) were the most sensitive to change. Exercise training tended to decrease anti-inflammatory markers (p = 0.072) but had no effect on natural killer or natural killer T cell proportions or cytotoxic activity.

Conclusions: Exercise training reduces pro-inflammatory markers in cancer survivors, with the strongest evidence for combined training and for prostate and breast cancer survivors. Further research is warranted to determine if these changes are clinically relevant or are associated with improvements in symptoms. To strengthen future research, focusing on novel immune populations that include functional parameters and standardized reporting of key immune outcomes is recommended.