Introduction
Regular lifetime exercise has been consistently associated with a reduced risk of developing postmenopausal breast cancer [
1]. Several mechanisms underlying the impact of exercise have been identified, including reductions in adiposity, sex hormone levels, insulin resistance and chronic inflammation [
1]. These pathways explain some portion of the association; however, additional mechanisms require identification and further investigation.
One additional mechanism of interest is the impact of exercise on growth factors known to drive cell proliferation and tissue growth [
2]. Vascular endothelial growth factor (VEGF) is a potent angiogenic cytokine in normal tissues and tumors. It has been shown to stimulate endothelial cell proliferation in vitro and induce angiogenesis in vivo [
3]. Fibroblast growth factors are a family of growth factors related to a broad range of physiologic functions including angiogenesis and epithelial proliferation [
4]. Fibroblast growth factor-2 (FGF2) is an interesting candidate in breast cancer pathology as a common single nucleotide polymorphism in the gene has been associated with breast cancer risk in large studies among Western populations [
5].
In a previous investigation from the Breast cancer and Exercise Trial in Alberta (BETA), we observed that higher levels of activity were associated with reduced body fat [
6]. Unhealthy fat pad expansion is characterized by inadequate vascularization and hypoxia – leading to increased fibrosis, inflammation and insulin resistance [
7]. We hypothesized that exercise-induced change in adipose tissue vascular remodeling may account for some of the observed exercise-related reduction in postmenopausal breast cancer risk. As part of ancillary investigations from BETA to evaluate novel mechanisms relating exercise to breast cancer risk, we investigated whether a higher versus a more moderate volume of exercise would lead to differential changes in VEGF and FGF2 levels.
Materials and methods
BETA study design and eligibility criteria have been described elsewhere [
8]. Briefly, we randomized 400 postmenopausal women, aged 50–74 years old, who were previously inactive but otherwise healthy to 150 (MODERATE) or 300 (HIGH) minutes/week of aerobic exercise at 65–75% of maximal heart rate reserve for 12 months. Participants completed three supervised sessions and two unsupervised sessions per week, with the HIGH volume exercise group completely exercise sessions that were twice as long as the MODERATE volume group. During exercise sessions, any aerobic activity that raised that participants’ heart rates to their target heart rate zones was permitted.
Fasting blood samples were drawn from study participants at baseline, 6 and 12 months, after a 24-h complete abstinence from exercise and alcohol intake and at least 10 h since the last food intake. FGF2 and VEGF levels in plasma were quantified using a custom-plex Human Circulating Cancer Biomarker Multiplex Assay. The multiplex assays were performed at Eve Technologies (Calgary, AB, Canada) using the Bio-Plex™ 200 system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The detection limits for FGF2 and VEGF in this assay were 3.6 and 6.4 pg/mL, respectively. The inter- and intra-assay coefficients of variation (CV) were 6.3 and 7.9% for FGF2, and 10.2 and 12% for VEGF, which are below or comparable to other CVs using the Bio-Plex platform [
9]. All lab personnel were blinded to the intervention assignment. Each participant’s samples at different time points were analyzed in the same batch and each batch included an equal number of MODERATE and HIGH blood samples.
Changes in concentrations of growth factors in the two intervention groups were compared in an intention-to-treat (ITT) analysis using linear mixed models that accounted for repeated measures of FGF2 and VEGF at 6 and 12 months, adjusting for baseline levels of the biomarker as a covariate. Treatment effect ratios (TERs) were calculated from these models and denoted the HIGH: MODERATE ratio of geometric mean biomarker levels over 12 months (reflecting changes from baseline to 6 and 12 months). The ITT analysis was repeated by stratifying on a priori potential effect modifiers: baseline body mass index (BMI), baseline fitness (estimated VO2max) and total percent body fat (full body dual X-ray absorptiometry scan).
A per-protocol analysis was undertaken for women who completed, on average, 90 to 100% of the exercise prescription for the MODERATE group or ≥ 90% of prescription for the HIGH group from weeks 13–52.
Results
A total of 386 women completed the trial with no statistically significant difference in dropout between arms. Baseline characteristics and exercise adherence of the BETA study participants have been previously described [
6]. We observed no statistically significant effects for increasing quartiles of exercise and change in circulating levels of VEGF or FGF2 with both groups combined (non-randomized, results not shown).
Results from linear mixed models suggested overall lower, but non-statistically significant group differences between baseline and 12 months for both FGF2 and VEGF for both arms of the trial. The HIGH group experienced slightly larger changes than the MODERATE group, but these changes were not clinically relevant (FGF2: TER 0.98, 95% CI 0.94–1.02; VEGF: TER 0.94, 95% CI 0.86–1.02; Table
1). Furthermore, the per-protocol analysis found essentially no differences between baseline and end of intervention effects with the TERs approximately equal to 1.0 for both biomarkers. Similarly, a per-protocol analysis stratified by exercise intensity, i.e., whether participants achieved 60% of the prescribed time in the target heart rate zone, showed no significant group differences in FGF2 and VEGF (Table
2).
Table 1
Intention-to-treat and per-protocol analysis of FGF2 and VEGF concentrations between high and moderate volume exercise groups in BETA (n = 386)
Intention-to-treat |
FGF2 |
High | 195 | 121.3 | 114.4–128.5 | 116.1 | 110.1–122.4 | 123.3 | 117.7–129.2 | 0.98 | 0.94–1.02 | 0.28 |
Moderate | 191 | 118.8 | 111.9–126.2 | 122.9 | 116.2–130.0 | 120.1 | 113.1–127.6 | | | |
VEGF |
High | 195 | 62.3 | 55.1–70.4 | 58.5 | 51.7–66.3 | 58.6 | 51.3–66.9 | 0.94 | 0.86–1.02 | 0.15 |
Moderate | 191 | 64.9 | 57.2–73.6 | 63.6 | 55.7–72.6 | 67.3 | 58.9–76.9 | | | |
Per-protocol d |
FGF2 | 0.64 |
High | 80 | 119.8 | 110.8–129.4 | 116.0 | 107.3–125.3 | 121.2 | 113.5–129.4 | 1.02 | 0.95–1.10 |
Moderate | 58 | 112.6 | 100.0–126.7 | 113.9 | 101.9–127.3 | 109.6 | 98.3–122.2 | | |
VEGF | 0.74 |
High | 80 | 60.6 | 50.8–72.3 | 60.6 | 50.0–73.4 | 58.0 | 47.6–70.7 | 0.98 | 0.84–1.13 |
Moderate | 58 | 70.3 | 54.2–91.0 | 64.9 | 49.9–84.6 | 74.0 | 56.9–96.3 | | |
Table 2
Stratified analyses of changes in FGF2 and VEGF between high and moderate volume exercise groups in BETA (n = 386)
FGF2 |
Time at high intensity c |
< 60% prescribed | MODERATE | 17 | 2.0 | 4.9 | 1.03 | 0.91–1.17 | 0.62 |
HIGH | 44 | 0.4 | 0.3 | | |
≥ 60% prescribed | MODERATE | 41 | 0.9 | −5.6 | 1.01 | 0.92–1.12 | 0.78 |
HIGH | 36 | −7.4 | 2.3 | | |
BMI d |
< 30 | MODERATE | 116 | 9.0 | 6.8 | 0.99 | 0.94, 1.05 | 0.83 |
HIGH | 121 | 4.0 | 5.2 | | |
≥ 30 | MODERATE | 75 | 5.1 | 2.2 | 0.95 | 0.88, 1.02 | 0.13 |
HIGH | 74 | −7.3 | 6.3 | | |
VO2max level e |
< 27.2 ml/kg min | MODERATE | 93 | 8.2 | 1.3 | 0.94 | 0.88, 1.00 | 0.05 |
HIGH | 98 | −2.2 | 8.2 | | |
≥ 27.2 ml/kg min | MODERATE | 98 | 6.7 | 8.5 | 1.02 | 0.96, 1.08 | 0.56 |
HIGH | 97 | 1.2 | 3.1 | | |
Total body fat f |
< 29.7 kg | MODERATE | 94 | 9.2 | 5.5 | 1.02 | 0.96, 1.08 | 0.53 |
HIGH | 99 | 1.2 | 3.0 | | |
≥ 29.7 kg | MODERATE | 97 | 5.8 | 4.5 | 0.94 | 0.88, 1.00 | 0.04 |
HIGH | 96 | −2.2 | 8.5 | | |
VEGF |
Time at high intensity |
< 60% prescribed | MODERATE | 17 | −6.5 | 11.0 | 1.04 | 0.83–1.31 | 0.72 |
HIGH | 44 | 5.0 | 7.2 | | |
≥ 60% prescribed | MODERATE | 41 | −8.0 | 3.2 | 0.90 | 0.73–1.10 | 0.28 |
HIGH | 36 | −6.0 | −16.8 | | |
BMI |
< 30 | MODERATE | 116 | −4.1 | 6.0 | 0.90 | 0.80, 1.01 | 0.06 |
HIGH | 121 | −10.4 | −10.6 | | |
≥ 30 | MODERATE | 75 | 1.5 | 0.3 | 1.01 | 0.88, 1.15 | 0.90 |
HIGH | 74 | 1.4 | 2.9 | | |
VO2max level |
< 27.2 ml/kg min | MODERATE | 93 | 4.0 | 1.9 | 0.93 | 0.82, 1.06 | 0.29 |
HIGH | 98 | −1.8 | −6.7 | | | |
≥ 27.2 ml/kg min | MODERATE | 98 | −7.2 | 5.7 | 0.94 | 0.83, 1.06 | 0.30 |
HIGH | 97 | −10.2 | −5.1 | | |
Total body fat |
< 29.7 kg | MODERATE | 94 | −7.7 | 4.5 | 0.92 | 0.80, 1.06 | 0.25 |
HIGH | 99 | −9.1 | −10.5 | | | |
≥ 29.7 kg | MODERATE | 97 | 4.0 | 3.1 | 0.95 | 0.85, 1.06 | 0.39 |
HIGH | 96 | −3.0 | −0.3 | | |
Stratifying results by baseline BMI, VO
2max, and total body fat did not reveal any sub-group effects for VEGF (Table
2). Conversely, the result for FGF2 varied by baseline VO
2max and total body fat. Specifically, women with lower baseline physical fitness had an overall lower concentration of FGF2 in the HIGH group (TER 0.94, 95% CI 0.88, 1.00). However, this result was largely driven by the opposite direction of change at six months (8.2% increase in MODERATE group vs. 2.2% decrease in HIGH group). At 12 months, the HIGH group had a greater increase in FGF2 from baseline (8.2% vs 1.3% in MODERATE group). Likewise, the statistically significant TER for FGF2 observed among women with greater total body fat at baseline (TER 0.94; 95% CI 0.88, 1.00) was driven by the opposite direction of change at 6 months (5.8% increase in MODERATE group vs 2.2% decrease in HIGH group), while FGF2 showed a greater increase in HIGH group (8.5% vs 4.5% in MODERATE group).
Discussion
Growth factors (i.e., FGF2 and VEGF) and their receptors (FGF2R and VEGFR), have been implicated in breast cancer tumorigenesis [
10,
11], progression [
4,
11], and metastasis [
4]. However, the expression of specific FGF2 and VEGF polymorphisms has also been shown to be protective against developing breast cancer [
12,
13]. In BETA, the greatest increases in FGF2 expression were observed with HIGH volumes of exercise in those with lower baseline physical fitness and higher total body fat whereas, VEGF expression was unaltered.
Our findings may be partially explained by the follow-up assessment timing and the somewhat differential nature of each growth factor’s pro-angio/arteriogenic action. In general, FGF2 and VEGF are important mediators of pro-angiogenic adaptations to repetitive exercise and act to increase tissue capillary density in response to hypoxic and inflammatory signaling [
14]. However, it is well known that VEGF signaling is more central to supporting angiogenesis and neovascularization whereas FGF2 signaling plays a larger role in arteriogenesis and the maintenance of mature vessels [
3,
15]. In general, chronically elevated growth factor levels may be considered pathologic – reflective of underlying tumor proliferative processes. However, in the context of exercise, transient growth factor increases can be viewed as normal physiologic hypoxia-driven adaptations – reflective of favorable vascular bed remodeling. While we would have expected to observe a transient increase in levels of these growth factors following the exercise intervention, in the context of the 12-month intervention within BETA, the six-month gap between follow-up assessments may have limited our ability to capture the transient spikes in the expression of these growth factors. Following an acute bout of exercise, the resulting increase in growth factor expression also appears to be transient – elevated after 7–10 days of training and returning to baseline after 28 days [
16,
17].
In conclusion, high doses of supervised exercise did not impact levels of VEGF or FGF2. Additional mechanisms related to decreased breast cancer risk through altered adipose tissue require investigation.
Acknowledgements
Calgary Study Coordinators were: Krista Carlson, Sana Fakih, Megan Farris, Quinn Harris, Sarah MacLaughlin, Erica Roberts, and Kristen Simone. Edmonton Study Coordinators were: Natalie Ilkiw, Ciara Kallal, and Dr. Amy Speed Andrews. Assistance with information sessions was provided in Calgary by Drs. Brigid Lynch and Fabiola Aparicio-Ting. Calgary Exercise Trainers were: Carrie Anderson, Alia Bharwani, Shannon Brown, Ashley Cuthbert, Sue Daniel, Julie Gowans, Margo Graham, Erin Korsbrek, Kathleen Kranenburg, Jessica Morrison, Jason Ng, Nicole Slot, Tania White, and Kaila Wright. Edmonton Exercise Trainers were: Arne Anderson, Lisa Belanger, Jennifer Crawford, Cindy Forbes, Alyssa Hindle, Corey Kuzik, Erin McGowan, Mary Norris, Janel Park, Julianne Symons, Linda Trinh, Stephanie Voaklander and Lynne Wong. Study recruiters were: Jennie Duke, Jasdeep Hayer, Trisha Kelly, Jasmine Lee, and Lilly Mah. Data entry was done by: Sinead Boyle, Barbara Mercer, Carla Quesnel and Trisha Kelly. Data management, including database creation, questionnaire design, data integrity and quality control, was done by: Dr. Steven Szarka, Farit Vakhetov, and Wendy Walroth. Qinggang Wang was responsible for the randomization procedures, sample size calculations and some data verification. The late Dr. Robert C. Millikan was a co-investigator on this trial and contributed to the study design and methods.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.