Meta-analysis of Exercise Training on Vascular Endothelial Function in Cancer Survivors.

Cancer and cardiovascular disease (CVD) are leading causes of morbidity and mortality in the United States. Vascular endothelial dysfunction, an important contributor in the development of CVD, improves with exercise training in patients with CVD. However, the role of regular exercise to improve vascular function in cancer survivors remains equivocal. We performed a meta-analysis to determine the effect of exercise training on vascular endothelial function in cancer survivors. We searched PubMed (1975 to 2016), EMBASE CINAHL (1937 to 2016), OVID MEDLINE (1948 to 2016), and Cochrane Central Registry of Controlled Trials (1991 to 2016) using search terms: vascular function, endothelial function, flow-mediated dilation [FMD], reactive hyperemia, exercise, and cancer. Studies selected were randomized controlled trials of exercise training on vascular endothelial function in cancer survivors. We calculated pooled effect sizes and performed a meta-analysis. We identified 4 randomized controlled trials (breast cancer, n=2; prostate cancer, n=2) measuring vascular endothelial function by FMD (n=3) or reactive hyperemia index (n=1), including 163 cancer survivors (exercise training, n=82; control, n=81). Aerobic exercise training improved vascular function (n=4 studies; standardized mean difference [95% CI]=0.65 [0.33, 0.96], I2=0%; FMD, weighted mean difference [WMD]=1.28 [0.22, 2.34], I2=23.2%) and peak exercise oxygen uptake (3 trials; WMD [95% CI]=2.22 [0.83, 3.61] mL/kg/min; I2=0%). Our findings indicate that exercise training improves vascular endothelial function and exercise capacity in breast and prostate cancer survivors.

Introduction

Cancer and cardiovascular disease (CVD) are the leading causes of mortality in the United States.[1] As a result of advances in cancer prevention, early detection, and novel treatment therapies, the cancer mortality rate has decreased by 23% during the past 3 decades.[1] Despite this benefit, cancer survivors are at increased risk for CVD-related mortality.[2,3] Indeed, Armenian et al.[3] recently reported that 8-year overall survival is significantly lower among cancer survivors who developed CVD than among cancer survivors without CVD.

In accordance with the “multiple-hit” hypothesis, the increased CVD mortality in cancer survivors is the result of underlying cardiovascular risk factors combined with direct effects of anticancer therapy (eg, cardiotoxicity) and direct effects of a sedentary lifestyle.[5] Cancer survivors’ cardiorespiratory fitness, measured objectively as peak aerobic capacity (peak VO2), is ~25% lower relative to healthy, age- and sex-matched individuals and is an independent predictor of mortality in advanced lung and breast cancer.[5-7] The reduced peak VO2 in breast cancer survivors is partly a result of impaired peripheral vascular function that results in reduced blood flow and oxygen delivery to the active muscles.[7-10] Studies in adult cancer survivors report better,[11-13] not different,[14-16] or impaired[8,10,17] vascular endothelial function relative to age-matched controls.

The endothelium plays a direct role in the balance of tissue oxygen supply and metabolic demand by regulating vessel tone at rest and during exercise.[18] Endothelial dysfunction impairs the endothelium’s ability to dilate in response to a variety of vasodilatory stimuli such as intra-arterial infusion of the endothelium-dependent vasodilator acetylcholine or shear stress. As such, endothelial dysfunction can limit blood flow and perfusion of skeletal muscle and organs. Furthermore, endothelial dysfunction has important clinical implications in that it represents an initial, reversible step in the development of atherosclerotic CVD.[19] Endothelial dysfunction leads to upregulation of adhesion molecules (increased leukocyte adherence), chemokine/cytokine secretion, increased vessel permeability, low-density lipoprotein oxidation, platelet activation, and vascular smooth muscle proliferation.[20-22] Flow-mediated dilation (FMD) is currently the standard for noninvasive assessment of large conduit artery endothelial function because of considerable clinical trial experience and validation and its association with cardiovascular events.[18,23,24] In this method, large conduit artery diameter (typically brachial) is measured by Doppler ultrasound before and after an increase in shear stress induced by reactive hyperemia, with vasodilation occurring through local endothelial release of nitric oxide (NO).[25]

In both healthy and diseased populations, exercise training is associated with favorable improvements in endothelial function[26,27] and peak VO2.[28-30] Moreover, studies in diseased populations found that an increase in peak VO2 correlated strongly with improvements in vascular endothelial function.[31,32] Functional vascular adaptations (improved ability to accommodate blood flow through vasodilation) elicited with exercise training appear to precede morphological changes (skeletal muscle hypertrophy, vascular and cardiac remodeling) and yield greater oxygen delivery to active muscles.[27,33]

Previous studies have demonstrated the role of exercise training for increasing peak VO2 in cancer survivors.[28] However, the role of regular exercise to improve vascular endothelial function in cancer survivors remains equivocal. Consequently, we performed a systematic review and meta-analysis to provide pertinent information on CVD risk modulation through exercise training on the primary end point, vascular endothelial function, and the secondary end point, peak VO2, in cancer survivors.

Methods

Data Sources

The authors (RIB and STB) searched PubMed (1975 to June 2016), EMBASE CINAHL (1937 to June 2016), OVID MEDLINE (1948 to June 2016), and Cochrane Central Registry of Controlled Trials (1991 to June 2016) using the following MESH terms and text words: vascular function, endothelial function, reactive hyperemia, flow-mediated dilation, cancer, and exercise. We also hand searched reference lists of all identified studies.

Study Selection

Two investigators (RIB and STB) independently reviewed the titles and abstracts of all citations to identify studies reporting the effects of exercise training on vascular endothelial function in cancer survivors, defined as any individual with a current or past diagnosis of cancer. Both investigators obtained the full text of potentially relevant articles and independently reviewed those using a pre-standardized data abstraction form and eligibility criteria defined a priori. We included randomized controlled trials in cancer survivors with an exercise training intervention (aerobic or resistance training, ≥3 weeks duration) and a measure of vascular endothelial function (FMD, reactive hyperemia).

Data Synthesis and Analysis

Data were analyzed using the change from the preperturbation baseline for both exercise and control groups. Standard deviation of the change was estimated assuming a correlation of 0.5 between baseline and end-of-study measures.[34] Study results were pooled using random effects models. Pooled statistics were calculated using weighted mean differences (WMDs) for outcomes (FMD and peak VO2) measured using the same methodology and standardized mean differences (SMDs) for vascular endothelial function measured using different scales across studies (ie, FMD and EndoPat). All results were calculated with 95% CIs. Statistical heterogeneity was assed using a χ2 test that considered a P value of less than .10 to indicate significant heterogeneity. I2 values were also calculated to quantify variability in effect size; values of 25%, 50%, and 75% described low, moderate, and high heterogeneity, respectively.[35,36]

Results

Study Selection, Evaluation, and Inclusion

We identified 310 citations from electronic databases and by manually scanning references of included studies. After removal of duplicate publications and initial screening for inclusionary criteria, 13 full articles were reviewed, of which 4 unique randomized controlled trials were identified (exercise training, n = 82; control, n = 81; Figure 1).[17,37-39] The trials included breast (n = 2) and prostate cancer survivors (n = 2) who were middle aged (mean age = 57 ± 7 years). Prostate cancer studies used cohorts of men on long-term androgen deprivation therapy[17] or following surgery (radical prostatectomy).[37] Breast cancer trials used cohorts of women undergoing neoadjuvant doxorubicin-cyclophosphamide therapy[39] or women with a history of invasive breast cancer within the previous 5 years.[38] All trials incorporated aerobic exercise training (n = 82) for 20 to 45 minutes per session at 55% to 75% peak VO2 (Table 1), 3 days a week, for 3 to 6 months, and one study included a resistance training component.[17] Control groups received usual care. In one study, intervention and control groups received the same dietary intervention,[38] and one study included a dietary education seminar biweekly for the exercise, but not the control group.[17] Risk-of-bias assessment is displayed in Figure 2.

Figure 1.

Flow of trials through the selection process.

Table 1.

Characteristics of Included Studies.

Study	Cancer Type	Sample Size (n)	Age (years)	BMI (kg/m2)	Treatment	Exercise Intervention	Vascular Measure
Giallauria et al,[38] 2016	Breast	EXT 25; UC 26	52; 54	27.3 28.2	<5 Years posttherapy	Cycle, 60%-70% VO2peak, 30 minutes, 3 d/wk for 3 months	EndoPAT
Gilbert et al,[17] 2016	Prostate	EXT 22; UC 20	70; 70	30.6 28.8	On ADT	Cycle, rowing, treadmill, 55%-75% HRmax, 30 minutes, 3 d/wk for 4 months, plus resistance and balance exercises	FMD
Jones et al,[39] 2013	Breast	EXT 10; UC 10	51; 46	28 27.3	On AC	Cycle, 55%-65% VO2peak, 20-45 minutes, 3 d/wk for 4 months	FMD
Jones et al,[37] 2014	Prostate	EXT 25; UC 25	58; 61	29 28	Postsurgery	Walking, 55%-75% VO2peak, 30-60 minutes, 5 d/wk for 6 months	FMD

Abbreviations: BMI, body mass index; EXT, exercise training; UC, usual care; VO2peak, peak oxygen uptake; ADT, androgen deprivation therapy; HRmax, maximal heart rate; FMD, flow-mediated dilation; AC, adriamycin-cyclophosphamide.

Figure 2.

Risk of bias assessment.a

a +, Low risk of bias; =, unclear risk of bias; -, high risk of bias.

Quantitative Data Synthesis: Exercise Training and Vascular Endothelial Function

Three studies quantified vascular endothelial function by FMD, and 1 study measured reactive hyperemia using the commercially available EndoPAT device to index blood flow response to tissue hypoxia. All studies were pooled, and exercise training was associated with a significant improvement in vascular function (SMD = 0.65 [0.33, 0.96], I2 = 0%; Figure 3). In the 3 studies measuring FMD, exercise training increased this outcome by approximately 1.3% (WMD = 1.28 [0.22, 2.34]; I2 = 23.2%) relative to the non–exercise training counterparts.

Figure 3.

Exercise training and vascular function.

Abbreviations: SMD, standardized mean difference; I-V, inverse variance method (fixed effects model); D+L, DerSimonian and Laird method (random effects model).

Additional End Points

Three studies examined the effect of exercise training on peak VO2 with 60 participants in the exercise groups and 61 participants in the control groups. Pooled data indicated that exercise training was associated with a statistically significant increase in peak VO2 (WMD = 2.22 mL/kg/min [0.83, 3.61], I2 = 0%; Figure 4).

Figure 4.

Exercise training and peak VO2.

Abbreviations: VO2peak, peak oxygen uptake; WMD, weighted mean difference; I-V, inverse variance method (fixed effects model); D+L, DerSimonian and Laird method (random effects model).

Discussion

The principal new finding of this meta-analysis is that exercise training is associated with a significant improvement in vascular endothelial function and peak VO2 in both breast and prostate cancer survivors. The absolute change in FMD that we found (1.3%) is similar in magnitude to that in other publications reporting the effect of exercise training on vascular function in healthy and diseased populations.[40-42] Moreover, a 1% increase in FMD is associated with an 8% to 13% reduction in future cardiovascular events[24]; improvements in FMD are indicative of improved CVD risk independent of changes in traditional risk factors such as BMI, blood pressure, or cholesterol.[27] As such, these improvements in vascular endothelial function and exercise capacity observed following exercise training may contribute to improved clinical outcomes in cancer patients.

There are several plausible mechanisms for this observed improvement in vascular endothelial function following exercise training in cancer patients. Prior studies demonstrate that exercise training has a direct effect on both large conduit and resistance artery endothelium-dependent vasodilatory function in humans.[27,33,43] Specifically, Hambrecht et al.[32] reported that 4 weeks of cycle training improved acetylcholine responses and adenosine-mediated blood flow in the left internal mammary artery of coronary artery disease patients.[44] Similarly, in populations without severe vascular disease, improvements in vascular function following exercise training are related to endothelial pathways, particularly NO, with little to no change in endothelial-independent smooth muscle function.[43] To date, only 1 study has assessed changes in endothelial-dependent and endothelial-independent vasodilation following exercise training in cancer patients.[17] Gilbert et al.[17] showed that 12 weeks of exercise training improved brachial artery FMD (endothelium-dependent vasodilation) by 2.2% in male prostate cancer patients on long-term androgen deprivation therapy, with no change in glyceryl trinitrate–mediated brachial artery vasodilation (endothelium-independent vasodilation). These findings suggest that improvements in peripheral vascular function may be endothelium mediated, with shear stress–mediated enhancements of NO bioactivity.[17]

NO release and magnitude of FMD appears to be dependent on the hyperemic blood flow response.[33,43] Exercise training increases the reactive hyperemia response to ischemia in both healthy and diseased populations.[33,45] In breast cancer patients, 3 months of cycle exercise training has been shown to improve reactive hyperemia index (RHI) by 19% (measured by peripheral artery tonometry).[38] Furthermore, these favorable improvements in RHI correlated significantly with improvements in VO2peak (ΔVO2peak vs ΔRHI: r = 0.47; P = .017).

Given that all prior exercise training studies have utilized either FMD or RHI as the measure of changes in vascular function, our mechanistic insights are currently limited to suggesting that exercise training may improve reactive hyperemic and endothelium-dependent pathways in cancer patients.

A consequence of the favorable change in peripheral vascular endothelial function is that it may be associated with increased oxygen delivery to the active muscles and concomitant improvement in peak VO2.[38,46,47] Indeed, in this meta-analysis, we found that peak VO2 was significantly higher (WMD = 2.2 mL/kg/min) after exercise training. Several prior meta-analyses found that aerobic training performed during or after adjuvant therapy significantly improved peak VO2 in cancer survivors, and the magnitude was similar to that in the present meta-analysis.[28,48,49] Importantly, the improvement in vascular endothelial function and peak VO2 with training may also have prognostic implications because a 3.5 mL/kg/min increase in peak VO2 is associated with a 12% to 17% decrease in mortality in men and women with and without CVD.[50,51] The mechanisms responsible for increased peak VO2 are not well known; however, they do not appear to be secondary to improved cardiac function.[9] Specifically, Haykowsky et al.[52] found that aerobic training performed during the first 4 months of adjuvant trastuzumab therapy did not improve left-ventricular ejection fraction in response to dobutamine stress in 17 women with HER2+ breast cancer. In a later study, Hornsby et al.[53] reported that 12 weeks of moderate- to high-intensity aerobic training did not change resting stroke volume, cardiac output, or left-ventricular ejection fraction in women with operable breast cancer during neoadjuvant chemotherapy. No studies have characterized changes in mitochondrial oxidative capacity or muscle fiber type in breast or prostate cancer, nor changes in autonomic function and blood flow distribution in response to exercise training. Decrements in autonomic function in breast cancer survivors and chemotherapy-treated survivors are reported[54,55]; it is unknown if exercise training can remediate these impairments. The present analysis supports a role of improved vascular endothelial function in improved muscle blood flow and peak VO2.

Limitations

As with most meta-analyses, our results are constrained by the select diversity of the trial participants, which was limited to breast and prostate cancer survivors. Moreover, our conclusions are drawn from 4 randomized controlled trials and are constrained by the quality of the trials reviewed. Because of the paucity of studies examining the effects of exercise training on vascular endothelial function in cancer survivors, we are unable to analyze modulating factors such as age, cancer type, stage, treatment, gender, time since diagnosis, mode, intensity, and duration of exercise training. Two studies included dietary interventions; it is well established that diet plays a role in vascular endothelial function, and thus, improvements cannot be attributed exclusively to exercise in these studies.[17,38,56,57] However, included studies encompassed diverse groups of survivors, including survivors on neoadjuvant chemotherapy, up to 5 years posttherapy, on hormone therapy, or postsurgery. Finally, the mechanisms responsible for improved vascular endothelial function were not examined. It is also currently unclear whether exercise training may improve other measures of vascular function (such as arterial stiffness) that increase oxygen delivery to the active muscles, potentially enhancing exercise tolerance.

Conclusion

Exercise training improves vascular endothelial function and peak VO2 in breast and prostate cancer survivors. Future studies are required to examine the biological mechanisms responsible for and the prognostic implications of the favorable improvement in vascular endothelial function in cancer survivors.