High-intensity interval training and thromboembolic events during chemotherapy for testicular cancer: a retrospective analysis from the Body & Cancer cohort

Abstract Background: Men with testicular cancer receiving platinum-based chemotherapy have an increased risk of thromboembolic events, with incidence rates between 8–24%. A recent trial evaluating the effect of high-intensity interval training (HIIT) prematurely closed as three out of nine participants (33%) in the intervention group developed a thromboembolic event. The purpose of this retrospective cohort study was: 1) (primary) to evaluate the incidence of thromboembolic events in men receiving chemotherapy for testicular cancer who had participated in HIIT during a 6-week exercise program (Body & Cancer) 2) to describe the feasibility of this program. Material and methods: Forty men who had participated in at least one HIIT session from February 2007 to February 2020 were included. Electronic medical records were searched for incident thromboembolic events (arterial and venous) during Body & Cancer and up to one-year post-chemotherapy. Attendance, cardiorespiratory fitness (VO2-peak), and upper and lower extremity muscular strength (1 repetition maximum (RM)) were obtained from the Body & Cancer database. Results: One participant developed a thromboembolic event during Body & Cancer. No participants developed a thromboembolic event in the follow-up period. In all, data represent 160 HIIT sessions with a median attendance of eight sessions [range 1–19]. Statistically significant increases in upper and lower extremity strength were observed (8.6 (4.2 to 13.0) and 26.0 (14.9 to 37.0) kg, respectively). No significant increase in cardiorespiratory fitness was found (0.14 (−0.03 to 0.31) l/min). Conclusion: While conclusions on the safety of HIIT cannot be drawn, data from the present study do not support previous findings cautioning avoidance of HIIT due to a possible added risk of thromboembolic events in men receiving platinum-based chemotherapy for testicular cancer. Considering the potential for positive effects on cardiovascular outcomes associated with HIIT, future studies with robust design should be performed in this population to confirm these observations.


Background
Testicular germ cell cancer (TGCC) is the most common type of malignancy in young adult males of European descent. It is one of the most curable malignancies with a 5-year survival rate of approximately 95%. This is in large part due to the implementation of platinum-based chemotherapy in patients with metastatic disease [1,2].
Standard chemotherapy for metastatic TGCC consists of a combination cisplatin (20 mg/m 2 ) and etoposide (100 mg/m 2 ) daily for 5 days, and bleomycin (15,000 IE/m 2 ) weekly, administered in a 3-week schedule (BEP), alongside supportive treatment. However, BEP is associated with toxicities resulting in the adverse side-and late effects including fatigue [3], reduced well-being [4,5] and physical capacity [6], as well as an overall increased risk of cardiovascular disease [7,8].
In an effort to prevent acute and long-term adverse events such as cardiovascular disease, exercise has been proposed as an intervention [6,9]. Recommendations for cancer survivors include avoidance of inactivity and promotion of physical activity during and after treatment [10]. However, while specific exercise prescriptions have been proposed for individual outcomes (e.g., fatigue) [10] generalizability across cancer populations is limited as the predominance of studies has been performed in breast and prostate cancer survivors.
To the authors' knowledge, four exercise studies have exclusively included TGCC survivors [11][12][13][14][15]. Of these, three studies were performed during chemotherapy. One examined the effects of resistance exercise [11], and one the feasibility of an individual, low-threshold physical activity intervention [12]. Both studies found the interventions to be feasible and safe. The third study, performed by Thorsen et al. [15], aimed to evaluate the effect of 2 x weekly, highintensity interval training (HIIT) on cardiorespiratory fitness ('Testicular cancer and aerobic and strength training trial' (TAST-trial)). However, an unexpectedly high number of thromboembolic events (TE) among the patients randomized to the intervention arm was observed (3/9, 33%) vs. 0/10 in the control arm). While the risk of TE during BEP is a known adverse event, with observed incidence rates ranging between 8%-24% [15], the study was prematurely closed due to a perceived increased risk of TE. As such, findings from this study raises uncertainty as to the safety of HIIT during chemotherapy for TGCC [9].
Against this backdrop, a multimodal exercise program, including HIIT (Body & Cancer) [16] has been offered to cancer patients receiving chemotherapy in the Copenhagen area since 2007. The primary purpose of this study is therefore to evaluate the incidence of TE in the cohort of patients receiving chemotherapy for TGCC who have participated in HIIT during Body & Cancer. Further, we aim to describe the feasibility of this program, specifically related to attendance as well as explorative analyses of the effect on cardiorespiratory fitness and muscular strength.

Material and methods
This retrospective study is reported in line with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [17]. Data was primarily retrieved by the first author and supplemented by the last author.

Setting
The study took place at the Center for Health Research (UCSF), Copenhagen University Hospital, Rigshospitalet Participants Forty patients receiving chemotherapy for metastatic TGCC, who had participated in at least one high-intensity exercise session of the Body & Cancer program between February 2007 through to February 2020, were identified in the Body & Cancer database, and included in the study. In this interim no major changes in the treatment of TGCC took place. All participants were treated at the Department of Oncology, Rigshospitalet and were eligible for Body & Cancer if they had initiated their first cycle of chemotherapy and had a WHO performance status of 0 or 1. Patients with symptomatic congestive heart failure and/or myocardial infarction within the previous three months were excluded. Primary thromboprophylaxis adjunct to chemotherapy was not administered as standard treatment.

The Body & Cancer program
Details of the program have been described previously [16]. In brief, Body & Cancer is an ongoing six-week, group based (10-15 participants), multimodal exercise program comprised of both low-and high-intensity components (Table 1).
Of particular interest for the present study are the three high-intensity days. Here participants engage in an aerobic warm-up, followed by resistance exercise (6 machines targeting large muscle groups, loads up to 5-8 repetition maximum (RM)), and HIIT predominantly using stationary bicycles, with target peak intensities corresponding to 85 -95% of each participant's maximum heart rate. In general, HIIT sessions increase in length over the six-week period, (from approximately 15 to 30 min), and peak intervals progressively increase in both frequency and length (from 30 s and up to two minutes). Exercise sessions are supervised by a cancer nurse specialist and a physical therapist. Prior to each high-intensity exercise session, participants are screened (e.g., musculoskeletal issues, temperature, blood pressure, and recent blood samples) to ensure the safety of participation [16].
Cardiorespiratory fitness and muscular strength testing are assessed at baseline (Body & Cancer commencement) and after six weeks. Muscular strength is ascertained using the 1 repetition maximum test (1RM) [18] in all six resistance exercise machines (Technogym (Gambettola, Italy)). Cardiorespiratory fitness (VO 2 -peak) is estimated by the use of a Watt-max test (stepwise work capacity) [19,20] on a stationary exercise cycle (Monark Ergomedic 839E, Sweden). Prior to testing, participants work at a steady state level for approximately 5 min. Hereafter, the Watt-max test commences at 47-87 Watt, increasing by 10-20 Watt with each consecutive minute until volitional fatigue. VO 2 -peak is then estimated using the formula VO 2 -peak ¼ 0.16þ (0.0117 Â maximal power output), where maximal power out is measured in watts. Test protocols used at baseline (e.g., 67 W, increasing 15 W/min) are used again at six weeks for comparability. In addition, patient-reported outcomes, including self-reported physical activity levels using a modified version of the Saltin-Grimby Physical Activity Level Scale [21] are obtained. Demographic, patient-reported outcomes, and objective physiological outcomes, as well as attendance, are prospectively recorded in a Body & Cancer database.

Outcomes and data sources
Electronic medical records Incident TE's during the six-week Body & Cancer program was the primary outcome. Electronic medical records were individually searched for TE diagnoses as well as ordination Further, data regarding cancer and treatment characteristics were obtained as well as other variables related to thromboembolic risk. Specifically this included: Cancer characteristics (histology, metastasis and location, long axis diameter of largest retroperitoneal lymph node, International Germ Cell Cancer Collaborative Group (IGCCCG) prognosis group [22], tumor markers prior to first chemotherapy, hematology within one week of Body & Cancer start (hemoglobin, thrombocytes, leukocytes)); Treatment characteristics (date of orchiectomy and chemotherapy commencement, completed chemotherapy regimen, use of vascular access device); Comorbidities.

Body & Cancer database
Age, body mass index (BMI), and smoking status at Body & Cancer initiation were retrieved. If these were not accessible in the database, they were obtained from the medical records. Also, self-reported leisure time physical activity [21] at baseline was recorded. Finally, attendance as well as cardiorespiratory fitness (VO 2 -peak) and upper and lower extremity muscular strength (1 RM chest press and leg press, respectively) at baseline and after 6 weeks, were obtained.
For comparability to the TAST trial [15], the cancer stage was determined using the Royal Marsden Hospital staging system [23]. Khorana scores were calculated using available blood tests and BMI data [24].

Statistical methods
Descriptive statistics were performed in MicrosoftV R Excel (2016). Continuous variables are presented as mean values ±standard deviations (SD) or median [range]. Categorical variables are presented as frequency (percentages). To explore mean changes in cardiorespiratory fitness and muscular strength, paired t-tests (two-tailed and with a significance level set at 0.05) were conducted in participants with both baseline and six-week data. These analyses were conducted using Statistical Analysis Software (SAS) version 9.4.

Results
In all, data represents participation in 160 high-intensity exercise sessions. Of the forty participants included in this study, one developed a TE during Body & Cancer. Two participants were receiving anticoagulants (Heparin and Innohep) prior to participation in Body & Cancer due to venous thromboses and indications of pulmonary emboli. In addition, one participant was receiving life-long anticoagulant treatment (Xarelto) due to a previous ischemic stroke. No participants developed

Physiological outcomes
Statistically significant increases in upper and lower extremity muscular strength were seen after participation in Body & Cancer (Table 4). While no significant increase in cardiorespiratory fitness was observed, the results indicate that mean VO 2 -peak levels were maintained. Mean attendance rates to the high-intensity sessions for participants providing full data for upper and lower extremity strength and VO 2peak measures were 60.5% ± 19.0, 61.1% ± 19.9, 59.3% ± 22.2, respectively.

Comparability between Body & Cancer and TAST-trial populations
These results stand in contrast to the findings of Thorsen et al. who reported a TE incidence of 33% (n ¼ 3/9) in the intervention group that participated in the TAST trial during chemotherapy for TGCC [15]. Participants in these two cohorts were similar related to TE risk factors such as age, BMI and smoking status. Further, treatments received are comparable as standard chemotherapy regimens for TGCC are similar in Norway and Denmark. However, the Body & Cancer cohort had a higher burden of disease. For example, all participants in the TAST trial had stage ll TGCC and were IGCCCG good prognosis group. In comparison, 25% (n ¼ 10) of Body & Cancer participants had stage III or IV TGCC, and 20% were classified as either intermediate (n ¼ 7) or poor (n ¼ 1) prognosis group. This is of relevance as an increased incidence of TE has been observed with higher-stage TGCC and prognosis group status [25,26]. Further, Khorana scores [24] (used to predict future risk of venous TE) were 0-2 for all TAST-trial participants, while approximately 35% (n ¼ 12) in the Body & Cancer cohort had scores !3 just prior to initiating the program, indicating an increased risk for venous TE's in these individuals. Additionally, available data indicate that the Body & Cancer cohort was less physically active with mean cardiorespiratory fitness levels at baseline considered low (38 ml/kg/min, n ¼ 28) vs. 43 ml/kg/min in the TAST trial, considered average for Scandinavian men between the age of 30-39 years [27].
Considering that sedentary or low levels of physical activity are associated with an increased risk of cardiovascular morbidity [28][29][30], disparities in physical activity levels between the two study populations should also be taken into account. Taken together, it seems reasonable to conclude that the Body & Cancer population had an a priori elevated risk for developing TE's compared to the TAST-trial intervention group, yet the incidence of TE in the Body & Cancer cohort was considerably lower. Indeed, observed incidence rates of TE in the Body & Cancer population were in the lower range when considering that observed incidence of TE's lay in the interim between 8-24% in TGCC populations receiving chemotherapy (without exercise) [15]. Arguably, this adds credibility to the results indicating that TGCC patients receiving platinum-based chemotherapy can participate in HIIT without additional risk of developing TE's.

TE risk
TGCC patients are at an increased risk of experiencing TE's [31][32][33]. This is in part due to the prothrombotic state induced by cancer itself. [25,31,32] Additionally, though not fully understood, cisplatin-based chemotherapy can cause intravasal damage, potentially triggering intravasal coagulation affecting both arterial and venous systems, further increasing risk of TE's [25,31,32,34,35]. In addition, it has been hypothesized that HIIT can amplify chemotherapy-induced hypercoagulability [15]. This is supported by evidence from studies in non-cancer populations finding a greater susceptibility to arterial thrombus formation and acute coronary events at high compared to light and moderate intensity exercise [30,36,37]. This is particularly seen in sedentary individuals with cardiovascular disease performing unaccustomed strenuous exercise. At the same time, compelling evidence indicates that regular participation in HIIT induces superior cardiovascular adaptations, including cardiorespiratory fitness, and metabolic function [30,36]. In cancer populations, three systematic reviews and meta-analyses have recently found that HIIT is associated with favorable effects including cardiorespiratory fitness and cancer-specific health-related outcomes such as fatigue, both during and after cancer treatment [38][39][40]. In support of our findings, no incidence of TE's was reported in the three studies evaluating HIIT in participants receiving platinum-based chemotherapy [16,41,42]. Two of these studies [41,42] were performed in patients receiving treatment for lung cancer (n ¼ 17), and while not directly comparable (e.g., different cancer, the dosage of platinum-based chemotherapy), these observations do lend credibility to the results of the present study. The third study was a randomized controlled trial evaluating Body & Cancer. Here none of the 135 participants randomized to the intervention group experienced TE's, including seventeen participants receiving platinum-based chemotherapy for ovarian cancer (n ¼ 10) and TGCC (n ¼ 7) [16].

Strengths and limitations
Our study has several limitations that should be taken into consideration. First, both the TAST trial and Body & Cancer utilized HIIT with target peak intensities corresponding to 85-95% of maximal heart rate. However, though heart rate monitors are worn in Body & Cancer, it is primarily as a precautionary measure and not to monitor compliance to the individual HIIT session. Thus, we cannot be certain that HIIT target intervals were met. Despite this uncertainty, our data represent participation in 160 high-intensity sessions, of which Watt-max tests were performed in 47 of these sessions. Therefore, at a minimum, the results are based on the test sessions where the intensity was monitored and registered. Relatedly, median attendance was eight sessions (range [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19]. While higher attendance would increase the credibility of our findings, it should be noted that no apparent trend in HIIT participation was observed in the TAST-trial with TE cases observed after 4, 9, and 12 sessions. Second, the external validity of this study should also be taken into consideration as the Body & Cancer cohort, similar to participants in exercise trials (such as the TAST trial), likely represents a select group of men receiving treatment for TGCC. As such, both cohorts were likely more active and motivated for exercise compared to the background population. Nonetheless, the Body & Cancer cohort presented with low mean cardiorespiratory fitness levels at baseline and, comparative to the TAST-trial, a higher risk for TE's. As such, though the results of this study are likely generalizable to the select group of men receiving chemotherapy for TGCC that are motivated for participating in HIIT during treatment, caution should be taken when extending these findings to the background population.
Finally, this study was a retrospective analysis of a pragmatic exercise program in a relatively small cohort. Accordingly, it does not offer the possibility of drawing conclusions regarding causality. Thus, though results do not indicate an association between HIIT and the development of TE's, conclusions regarding the safety of HIIT cannot be drawn.

Feasibility and future directions
The average attendance rate to Body & Cancer was just under 50%. Additionally, approximately 50% of the cohort did not provide 6-week follow-up data, indicating that this six-week exercise program is not ideal for this population. This is not surprising considering the significant barriers (e.g., five days of infusion every three weeks, considerable side effects, additional blood tests, and scans) to participation in a six-week, time-intensive program. Nonetheless, despite small numbers and high drop-out, statistically significant increases in upper and lower extremity muscular strength were observed. Further, while no significant increase in cardiorespiratory fitness was found, the results indicate that mean VO 2 -peak levels were maintained for participants providing full data. Notably, as the delivery of daily exercise sessions and collection of data were performed by the same staff, the potential for bias (favoring positive effect) exists. Accordingly, studies with robust methodologies (e.g., randomized design, blinded assessment) should be performed to confirm these observations.
Within this context, a recent randomized controlled trial explored the effect of a 12-week HIIT intervention in participants who had been treated for TGCC (mean time since diagnosis of 8 years). Among positive findings, the study found significant positive effects on cardiorespiratory fitness, fatigue, and cardiovascular disease risk in the intervention group vs. the usual care group, with no reported adverse events [13,14]. While these results are encouraging, there was a considerable variation in the time since diagnosis (range 1 to 20 years), and only a little over a third (n ¼ 23/63) of the study population had received chemotherapy. As such, considering the observed results and the positive effects of exercise (and specifically high-intensity exercise) during chemotherapy found in other cancer populations [10,[38][39][40]43], a considerable rationale exists to ascertain the safety and effect of HIIT specifically tailored to TGCC patients during treatment with BEP.

Conclusion
Results from the present study do not indicate an association between HIIT participation and the development of TE's in patients receiving platinum-based chemotherapy for TGCC. However, future studies with robust designs should be performed to confirm this observation. In the interim, until this knowledge is available, exercise should be encouraged during chemotherapy. In regard to HIIT, based on the available data, neither promotion of-nor blanket statements warning against, HIIT are appropriate in this population. Instead, a more pragmatic approach whereby patients, who are motivated for or find meaning in HIIT, are supported in making informed decisions to do so. Importantly, considering the inherent increased risk of TE's associated with cancer, all patients should be educated to be aware of signs indicating a TE.