Effect of home-based resistance training on chemotherapy relative dose intensity and tolerability in colon cancer. The FORCE Randomized Control Trial

Abstract

Background:

Many patients with colon cancer cannot fully adhere to postoperative chemotherapy due to dose-limiting toxicities, resulting in a lower relative dose intensity and potentially compromising overall survival. We examined whether home-based resistance training during adjuvant chemotherapy improves relative dose intensity and patient-reported toxicities versus usual care in colon cancer patients.

Methods:

Multi-center, randomized control trial (RCT) conducted at community and academic practices. Enrollment of patients receiving postoperative chemotherapy for colon cancer occurred between February 23, 2018, and September 29, 2021; final follow-up was March 21, 2022. Participants were randomized to resistance training(RT) (n=90) or usual care (n=91) for the duration of chemotherapy. Participants in the resistance training group engaged in twice weekly home-based progressive resistance training. At the end of the study, usual care(UC) was given an online exercise program.

Results:

Among 181 randomized patients (mean age, 55.2 [SD 12.8] years, 95 [52.5%] were men), there were no differences in the mean relative dose intensity (RDI) among those in RT (79% [SD 19%]) and those in UC(82% [SD 19%]); (mean difference −0.04 [95% CI, −0.09, 0.02]). Assignment to RT did not significantly reduce the number of moderate/severe symptoms per week across follow-up (relative rate: 0.94 [95% CI, 0.72, 1.22]). Additionally, time since randomization did not significantly modify the effect of RT on the overall number of symptoms (P=0.06).

Conclusions:

Among patients with colon cancer, these results do not support home-based resistance training as an adjunct to chemotherapy specifically to improve planned treatment intensity.

Précis:

This is the first large scale trial testing whether resistance training during chemotherapy for colon cancer patients can improve muscle mass, reduce toxicities and improve treatment completion. The results suggest that an at home-based resistance training with supportive coaching by an exercise trainer did not improve either relative dose intensity or overall patient reported symptoms, but may help decrease fatigue and nausea over time.

INTRODUCTION

Both The American Society of Clinical Oncology¹ and The American College of Sports Medicine² have released clinical guidelines that recommend patients undergoing chemotherapy do at least 2 sessions a week of resistance training (RT) to improve quality of life related symptoms. However, they note more research is needed on whether RT can reduce dose-limiting toxicities and improve Relative Dose Intensity(RDI), a global measure of chemotherapy completion^3–6 associated with survival outcomes.^{7, 8}

Each year, over 30,000 patients newly diagnosed with colon cancer who receive adjuvant therapy will experience dose-limiting toxicities, (most commonly, hematologic, gastrointestinal and neuropathic), that compromise optimal treatment outcomes.^{9, 10} One group at particularly high risk are those with sarcopenia (low muscle)or cancer cachexia ( a muscle and/ or fat wasting disease defined as > 5 % weight loss in the previous six months) . Not only is muscle wasting highly prevalent among newly diagnosed non-metastatic colon cancer patients (up to 40%), additional muscle losses (6.1% [95% CI, −8.4 to −3.8; P < .001])¹¹ have been reported during cancer treatment due to muscle atrophy, mitochondrial depletion,¹² and anabolic inhibition by cancer therapy.^{13, 14}

Low muscle at initiation of cancer treatment is associated with increased dose-limiting toxicites but whether increasing muscle mass through resistance training during chemotherapy treatment for colon cancer can reduce dose-limiting toxicities and increase relative dose intensity (RDI) has only been previously examined in two very small randomized controlled trials (RCT), both of which were null.^{15, 16} In the multi-center RCT Focus On Reducing Dose-limiting Toxicities in Colon Cancer With Resistance Exercise Study (FORCE) trial, we enrolled patients undergoing adjuvant chemotherapy for stage II and III colon cancer and randomized to a home-based resistance training program versus usual care (UC) to determine the effect of resistance training on RDI and patient-reported moderate/severe toxicities during treatment.

MATERIAL and METHODS

The institutional review board at all recruitment sites approved the study protocol. Participants provided written informed consent.

Patient Population

Recruitment details were previously described.¹⁷ Participants were enrolled from the Dana-Farber Cancer Institute (Boston, MA), multiple oncology centers housed within Kaiser Permanente Northern California, and Penn State Cancer Institute (Hershey, PA).

Eligibility for the trial was previously described.¹⁷ Briefly, eligible patients were diagnosed with stage II or III colon cancer, at least 18 years of age and had undergone curative-intent surgical resection at least 6 weeks prior to enrollment. Patients were to start one of four adjuvant chemotherapy regimens: FOLFOX (5-flurouracil [5-FU]/leucovorin [LV] and oxaliplatin), 5-FU/LV, CAPOX (capecitabine and oxaliplatin), or capecitabine alone, with plans of 3–6 months in duration. Patients needed to be English-speaking without medical or physical restrictions to be able to exercise.¹⁸ Patients were excluded if they had already received three or more chemotherapy cycles; had another concurrent, actively treated cancer (except non-melanoma skin cancer, in situ cervical cancer or localized prostate cancer with surveillance only); untreated hypertension; poor kidney function; metastatic disease; a current resistance training regimen of at least twice a week in the last 3 months; or if they were enrolled in other behavioral health (i.e., diet, physical activity, weight loss) clinical trials.

Study Design

FORCE is a two-arm randomized trial comparing study participants in a home-based resistance training intervention to a usual care group. Participants were randomized by the study manager and enrolled by the field interviewers to the resistance training, or usual care group in a 1 to 1 ratio using QMinim, a web based adaption of MinimPy,¹⁹ a covariate adaptive randomization procedure to allow balance on recruitment site, sex (male, female), cancer stage (II or III), chemotherapy regimen (FOLFOX, 5-FU/LV, CAPOX, capecitabine), and intended chemotherapy duration (3-months, 6-months). Investigators and interviewers were blinded to the outcome data until the end of the trial.

Intervention

Home-Based Resistance Training

The resistance training program was designed to promote muscle strength and hypertrophy using American College of Sports Medicine (ACSM) exercise prescription principles.^{2, 20, 21} Resistance training was prescribed twice per week, using a weight that ranged between 65–85% of the estimated one-repetition maximum. Patients were asked to do six to 10 repetitions for three to five sets, of each five large muscle multi-joint exercises. Details of the exercise prescription, delivery of the intervention and monitoring of intervention fidelity have been reported previously.¹⁷ Briefly, at the initial visit, a certified exercise professional who had national certification for exercise training, along with prior oncology expertise assessed each participants’ strength capacity with an estimated one-repetition maximum test to tailor the initial exercise prescription which was progressively increased in weight, sets, and reps over the course of chemotherapy. The intervention included 4–6 in-person visits at the beginning with an exercise professional (coinciding with their chemotherapy visits) and two days of progressive resistance training exercises at home each week for the duration of chemotherapy. There were five resistance training exercises included in the protocol (chest press, bent one arm row, squats, lunges, and deadlifts) using adjustable dumbbells (PowerBlock Inc., Burnsville, MN) that progressed over the course of the intervention.

Intervention participants were asked to consume two packets of supplemental protein powder (20 gms each, Agropur Inc., Eden Praire, MN) per day in an attempt to ensure that protein intake was at 1–1.5 g/kg/day, the recommended guideline for cancer survivors.²²

Participants were instructed to log the weight, number of repetitions, and number of sets they performed for each exercise performed during an exercise session, and the total amount of protein they consumed each day. Exercise professionals conducted semi-structured check-ins weekly to assess adherence from participant-kept logs and the participant’s ability to increase weights according to their exercise prescription as well as provide ongoing behavioral support.

Due to the COVID-19 pandemic, starting in March 2020, participants already on trial randomized to resistance training and newly randomized participants were switched to tele-coaching.

Usual Care

Participants randomized to usual care were instructed to maintain their usual activity. At the conclusion of participation in the study, usual care participants were offered a set of resistance bands; a 30-minute consult with a FORCE exercise trainer; an online exercise program, and samples of protein powder.

Study Outcomes and Assessments

Relative Dose Intensity (RDI) (Co-Primary)

RDI is a global measure of treatment completion that incorporates dose delays, dose reductions and early stoppage and is used to evaluate a patients overall received treatment compared to their prescribed or intended treatment. It was calculated as delivered dose intensity (DDI) / standard dose intensity (SDI) x 100% with the formula of Weycker²³ and as previously described.¹⁷ See supplementary methods for details on calculating RDI.

Patient Reported Adverse Events (Co-Primary)

To assess toxicities, nine symptoms specific to side effects of chemotherapy for colon cancer were measured using Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE).²⁴ Symptoms included nausea, vomiting, diarrhea, shortness of breath, hand-foot syndrome, numbness or tingling in hands or feet, pain, aching muscles, and fatigue. Participants were sent questionnaires before the beginning of every chemotherapy cycle and asked to report their symptoms over the past 7 days. Toxicities were grouped according to severity using the approach developed by Basch et, al.²⁴

Secondary Outcomes

Secondary outcomes include changes in body composition measured at baseline and follow-up by dual energy x-ray absorptiometry (DXA) scans.

Patient Characteristics

Race and ethnicity were collected as mandated by the National Institute of Health and based on self-report, per each institution’s standard for collection of these data. Height and weight were measured by a trained field interviewer or recorded from the electronic medical record, if unable to complete in person. Participants completed the short physical performance battery (SPPB) and grip strength to assess physical function.^{25, 26} Patients were considered frail if DXA appendicular lean mass was <7.25 Kg/m2 for men and <5.5 Kg/m2 for women, or baseline score on SPPB was <9.²⁷

Power Calculations

For the co-primary outcomes, we calculated that 90 patients randomized to each of the two study arms (N=180) would provide 80% power to detect a between group difference in mean percent RDI and mean number of grade 3 or 4 toxicities per week (at study end) of at least 0.42 standard deviation (SD) units, respectively (α=.05, two-sided t-test).²⁸ With an expected RDI SD of 17%, the minimum detectable difference in mean RDI was 7.3%.

Statistical Analysis

Multivariable linear regression was used to assess the difference between the resistance training and usual care groups in mean percent RDI and logistic regression was used to examine the planned RDI cut points of <85% and <70%^{7, 29, 30} in relation to study arm, with adjustment for all covariates used in randomization procedure. We conducted analyses stratified by a priori planned subgroups (treatment duration, sex, frailty status), and exploratory subgroups (age, randomization before or after chemotherapy start). Heterogeneity in intervention effect across levels of a stratification variable was assessed by inclusion of appropriate cross product terms in regression models.

The total number of self-reported moderate/severe symptoms per week and the relative rate of reporting a specific symptom each week in relation to study arm were examined using generalized linear models, with model parameter estimation via generalized estimating equations to accommodate the repeated measures of symptoms across multiple chemotherapy cycles (GEE; see supplementary methods for additional details). All models included an indicator for assigned resistance training, a function of time since randomization, and the covariates used in the randomization procedure. The main effect for assigned resistance training is interpreted as the average intervention effect across follow-up. Heterogeneity in the effect of intervention across follow-up was assessed by inclusion of appropriate cross-product terms in regression models. For the co-primary end points, statistical significance was assessed at the 5% level (2-sided).

In analyses of secondary endpoints, multivariable linear regression was used to assess the difference between the resistance training and usual care groups in mean BMI and body composition metrics at the end of the study intervention, adjusting for baseline value, age, sex, chemotherapy regimen, and clinical site.

Data analyses were conducted primarily in SAS version 9.4 and graphs were created in R version 4.1.3.

RESULTS

Participants were randomized between February 23, 2018, and September 29, 2021. The last follow-up visit was on March 21, 2022. Of 181 participants who were randomized, 90 were allocated to resistance training and 91 to usual care (Figure 1). The mean (SD) age was 55.2 (12.8) years, 95 (52.5%) were men. The usual care group had fewer patients receiving capecitabine only (4.4% vs 10.0%) and fewer patients who were sarcopenic (17.6 % vs 24.4%); other baseline characteristics were balanced (Table 1).

Figure 1. Consort diagram.

Ineligibility reasons (not mutually exclusive; do not add up to total above): Received treatment at non-study site location=121; Language barrier=100; Physical limitation=71; Other medical treatment or condition=57; Unable to recruit in protocol timeframe=44; Not new cancer, metastatic disease, or other cancer=42; Other reasons=42; Rectal cancer=37; No MD approval=35; Current resistance training=15; Unsafe to exercise=12; Not 6 mths FOLFOX (pre-amendment at DFCI)=11; Stopped chemotherapy=1; Enrolled in clinical trial=1; Cognitive impairment =1

Table 1.

Characteristics of trial participants

Characteristics	Overall	Usual Care	Resistance Training
N	(N=181)	(N=91)	(N=90)
Age
Mean (SD)	55.2 (12.8)	54.2 (12.6)	56.3 (12.9)
Gender
Men	95 (52.5)	49 (53.8)	46 (51.1)
Women	86 (47.5)	42 (46.2)	44 (48.9)
Race
White	127 (70.2)	69 (75.8)	58 (64.5)
African American	16 (8.8)	9 (9.9)	7 (7.8)
Asian	22 (12.1)	7 (7.7)	15 (16.7)
American Indian/Native Alaskan	2 (1.1)	1 (1.1)	1 (1.1)
Native Hawaiian or Pacific Islander	3 (1.7)	1 (1.1)	2 (2.2)
More than one	5 (2.8)	2 (2.2)	3 (3.3)
Other	6 (3.3)	2 (2.2)	4 (4.4)
Ethnicity
Not Hispanic or Latino	146 (80.7)	70 (76.9)	76 (84.4)
Hispanic or Latino	18 (9.9)	12 (13.2)	6 (6.7)
Unknown or Missing	17 (9.4)	9 (9.9)	8 (8.9)
Chemo regimen
FOLFOX 6m	94 (51.9)	51 (56.0)	43 (47.8)
FOLFOX 3m	7 (3.9)	2 (2.2)	5 (5.5)
CAPOX 3m	50 (27.6)	26 (28.6)	24 (26.7)
CAPOX 6m	14 (7.7)	6 (6.6)	8 (8.9)
CAP-only 6m	13 (7.2)	4 (4.4)	9 (10.0)
5FU-only 6m	3 (1.7)	2 (2.2)	1 (1.1)
AJCC stage
Stage 2	23 (12.7)	11 (12.1)	12 (13.3)
Stage 3	158 (87.3)	80 (87.9)	78 (86.7)
Randomized before initiation of chemotherapy
No	97 (53.6)	46 (50.5)	51 (56.7)
Yes	84 (46.4)	45 (49.5)	39 (43.3)
Sarcopenic*
No	125 (69.1)	65 (71.4)	60 (66.6)
Yes	38 (21.0)	16 (17.6)	22 (24.4)
Missing	18 (9.9)	10 (11.0)	8 (9.0)
Low physical function**
No	152 (84.0)	77 (84.6)	75 (83.3)
Yes	29 (16.0)	14 (15.4)	15 (16.7)

^*Sarcopenia = appendicular lean mass <7.25 for men or <5.5 for women on DXA scan

^**Baseline score <9 on the Short Physical Performance Battery

Intervention Adherence

Participants randomized to the resistance training group completed a median of 1.4 (0.6–1.7) sessions per week, using a weight that was 62% (53–70) of the estimated one-repetition maximum, and performed a median of 3 sets (2.6–3.6) with 7.5 repetitions (6.4–8.7). After adjusting for intended chemotherapy duration, the total weight lifted by participants in the resistance training group significantly increased by 0.7 kg per week (95% [CI: 0.4, 1.0]; P<0.0001), demonstrating a statistically significant increase in strength.

Co-Primary Outcomes (RDI and Patient Reported -Adverse Events )

Overall mean RDI was 80.6%. At the end of follow-up, (median 24.0 weeks [IQR 21.9–25.6] for 6-month adjuvant chemotherapy regimens; median 11.9 weeks [IQR 9.9–13.0] for 3-month regimens), there was no significant difference in RDI between those assigned resistance training vs. usual care (Figure 2) across all patients (estimated mean difference −0.04 [95% CI, −0.09, 0.02] P=0.19). In analyses stratified by chemotherapy regimen and individual drugs, we did not observe between-group differences in RDI.

Figure 2. RDI by intervention arm.

Models are adjusted for stage, sex, chemotherapy regimen and duration, and clinical site

There were no statistically significant interactions in the analyses of response to exercise by any subgroup (Figure 3). Resistance training had a higher prevalence of having an RDI <85% (estimated mean difference 21% [95% CI, 8%,35%]) (Table 2) compared to the usual care group. The prevalence of an RDI <85% among those assigned to resistance training relative to usual care did not significantly differ among analyses restricted to those receiving FOLFOX or CAPOX or individual drugs, nor were these analyses significant when the RDI cutoff was reduced to 70%.

Figure 3. RDI by intervention arm and a priori determined and exploratory subgroups.

Models are adjusted for stage, sex, chemotherapy regimen and duration, and clinical site

Table 2.

Observed prevalence of low RDI (<70%, <85%) by arm and estimated treatment difference.

	Usual Care		Resistance Training
	N	Observed %	N	Observed %	Estimated Treatment Difference (RD 95% CI)
Less than 70% of RDI
All patients	91	20%	89	24%	5% (−8%, 18%)
FOLFOX	53	25%	47	19%	−6% (−22%, 11%)
CAPOX	32	16%	32	22%	-- *
Oxaliplatin	85	28%	79	35%	11% (−4%, 25%)
5-FU	55	16%	48	13%	−3% (−19%, 14%)
Capecitabine	36	8%	41	24%	12% (−14%, 39%)
Less than 85% of RDI
All patients	91	40%	89	58%	21% (8%, 35%)
FOLFOX	53	49%	47	66%	17% (−3%, 36%)
CAPOX	32	25%	32	41%	19% (−9%, 47%)
Oxaliplatin	85	53%	79	59%	8% (−7%, 22%)
5-FU	55	35%	48	33%	0% (−19%, 18%)
Capecitabine	36	25%	41	41%	16% (−5%, 37%)

Estimated treatment differences are based on binomial regression with an identity link function, adjusting for stage, sex, chemotherapy regimen and duration, and clinical site.

^*Model didn’t converge

Seventy-seven percent of the expected number of PRO-CTCAE forms were returned and there were no differences in completion rate between resistance training (74%) and usual care (81%). Overall, resistance training did not reduce the average number of moderate/severe symptoms per week (Relative ratio [RR] 0.94 [95% CI, 0.72, 1.22]; P=0.62). Additionally, time since randomization did not significantly modify the effect of assigned resistance training on the average number per week of self-reported symptoms (P=0.6; Figure 3). Notably, only a small number of patients received an intended regimen of capecitabine or CAPOX >3 months (n=27), thus estimates for weeks 15 and 21 may be unstable (Figure 4).

Figure 4. The effect of resistance training (RT) vs usual care (UC) on the number of self-reported moderate/severe symptoms per week; a.) the average effect across time, and b.) each week since randomization.

a.) The number of moderate/severe symptoms was modeled as a quadratic function of time since randomization, assigned intervention group, and the characteristics used in the randomization procedure; b.) adds a statistical interaction between intervention group and time since randomization. The expected count (and 95% CI) for those assigned to UC (blue) and RT (red) was estimated at a.) week 12 and b.) each time point PRO-CTCAE questionnaires were distributed based on this model for our referent group (recruited from Kaiser, stage 3, female, on FOLFOX, and on a 6-month regimen). The number of symptoms per week in the RT group relative to the UC group (and 95% CI) was estimated in a.) our model without the interaction between time and RT, and b.) our model with this interaction, reporting the p-value for the statistical interaction between intervention group and our function of time since randomization. The bottom charts the number of participants contributing questionnaire data at a.) any time point, and b.) at each time point among the UC and RT groups.

Of the individual symptoms, two symptoms were reduced by resistance training over time; the effect of RT on fatigue, and nausea was significantly modified by time since randomization (P=0.03 and 0.05, respectively). Between 12 to 16 weeks post-randomization, the odds of reporting fatigue were approximately 50% lower among those assigned to RT relative to those assigned to UC, following which the effect attenuated towards the null (Supplemntal Figure 1).

Secondary Outcomes

There were no significant differences between resistance training vs. usual care in any change in body composition measure (Table 3). However, patients who were frail at baseline and in resistance training compared to usual care had considerable increases in their lean mass (1.01 kg [−0.2, 2.04], P=0.05; data not shown).

Table 3.

Changes in BMI and DXA outcomes in relation to study arm at the end of the study intervention

Endpoint	Randomized Group	N	Observed Baseline Mean (SD)	Estimated Treatment Difference (95% CI)	P
BMI	Usual Care	81	27.8 (6.0)	Reference
	Resistance Training	82	26.8 (4.8)	0.10 (−0.56, 0.75)	0.77
Total Lean Mass, kg	Usual Care	81	49.9 (11.2)	Reference
	Resistance Training	82	47.2 (10.9)	0.54 (−0.34, 1.41)	0.23
Total Fat Mass, kg	Usual Care	81	27.5 (10.5)	Reference
	Resistance Training	82	26.9 (9.5)	0.28 (−0.87, 1.42)	0.63
Appendicular Lean Mass, kg	Usual Care	81	21.5 (5.4)	Reference
	Resistance Training	82	20.2 (5.7)	0.01 (−0.44, 0.46)	0.96
Lean to Fat Ratio, %	Usual Care	81	2.0 (0.7)	Reference
	Resistance Training	82	1.9 (0.7)	0.06 (−0.04, 0.15)	0.24

Models are adjusted for baseline value of the dependent variable, age, sex, chemotherapy regimen, and clinical site

Exploratory Analyses

There was no difference in the effect of the intervention on RDI among those who gained >1kg of lean mass (estimated treatment difference −0.09 [−0.16, −0.02]) compared to those who maintained or lost lean mass (−0.07 [−0.18, 0.03], P interaction=0.83).

Twenty-six percent of patients had only tele-coaching with the trainer and no in-person instruction once the COVID-19 global pandemic limited in-person activities. There was no difference in RDI in those with tele-coaching only vs in-person visits (81%, mean SD [0.12] vs 79% [0.20], P=0.65).

No serious adverse events due to the intervention were reported.

DISCUSSION

Among patients with colon cancer treated with adjuvant chemotherapy, a home-based, progressive resistance training exercise intervention did not improve chemotherapy RDI. The number of self-reported toxicity-related symptoms from the PRO-CTCAE were similar between the two groups. In secondary analyses, resistance training reduced nausea and fatigue over time.

Only two previous small randomized controlled trials have examined effects of exercise on RDI in colon cancer. Our results are consistent with Van Vulpen et al. (n = 33) who reported no effect of supervised combined aerobic and resistance exercise on median RDI (82%) versus usual care (76%, p-value = 0.8) or the percentage of exercise participants achieving ≥85% RDI (35%) versus usual care (33%, p-value:  0.9).¹⁵ In the trial of by van Waart et al., (n = 23) the percentage of the planned chemotherapy dose received among exercisers (resistance and aerobic combined) was 87% and higher than usual care (78%). However, a statistical test for group comparisons was not performed due to the small sample size.¹⁶ Similarly, there have been only a few studies on the effect of resistance training during chemotherapy on dose-limiting toxicities, none of which were specifically in colon cancer. In one previous study in gastrointestinal cancer patients undergoing chemotherapy, resistance exercise training for 12 weeks significantly reduced lack of energy (P=0.011) and nausea (P=0.007), which our results confirm, as well as acid reflux (P=0.04), and back pain (P=0.001).³¹

There are several potential explanations for the lack of benefit from RT on RDI and on moderate/severe patient-reported toxicities. First, home-based resistance training with remote coaching may be less effective than supervised in-person resistance training.² However, a home-based exercise program was deemed most practical. Second, although adherence was comparable to other trials of exercise in cancer patients,² the frequency or duration of exercise, especially for those on 3-month regimens, may not have been adequate to promote a substantial change in RDI or symptoms. Third, the goal of the resistance training intervention was to maintain or build muscle mass. Our central hypothesis motivating this trial was that larger volumes of muscle mass for a given body surface area (BSA) may be beneficial to prevent dose-limiting toxicities by providing a larger volume of distribution for hydrophilic chemotherapy agents. While the resistance training group had a non-significant increase of 0.5 kg in total lean mass, an amount consistent with other home-based exercise intervention trials in cancer patients,³² this may not have been great enough to impact RDI. However, even among those who gained >1kg lean mass, there was no difference in effect of the intervention on RDI. Additionally, most patients in this trial may have already had sufficient muscle mass, thus masking any benefit: only 21% were sarcopenic based on commonly used DXA appendicular lean mass cut-offs.³³ Again of note, in secondary analyses, there was no difference in effect of the intervention on RDI in those who were frail at baseline (Figure 2). Lastly, and perhaps most importantly, our knowledge of how body composition impacts the pharmacokinetics of the combination regimens which most colon cancer patients receive is limited. Capecitabine is mainly lipophilic^34–36 and thus primarily distributes in the fat mass, while 5-FU is mainly hydrophilic²⁸ and primarily distributes in the lean mass. Oxaliplatin is often considered hydrophilic but may also have moderate lipophilicity.²⁹ Interestingly, for capatacibine, the usual care group had a non-significant higher RDI vs. the RT group. This could be explained by the higher gain in fat vs. lean mass seen in the usual care, compared to RT (Table 3). Further research is needed to examine the pharmacokinetic properties of these combination regimens in different body composition phenotypes before we fully understand how to optimize RDI through modulation of body size or composition.

Strengths

Strengths of this trial include a rigorous application of a standard exercise prescription which follows the principles of frequency, intensity, type, time, volume, progression (FITT-VP),³⁷ that was tailored specifically to each patient’s strength level and that was adapted weekly, as needed, by certified exercise trainers. RDI values observed in this trial are consistent with RDI values reported in the IDEA study,³⁸ the largest prospective pooling of 6 randomized trials of adjuvant therapy among patients with stage II and III colon cancer patients (nearly 13,000 patients), suggesting our results are generalizable to colon cancer patients undergoing chemotherapy.

Limitations

PRO-CTCAE forms were not completed by every patient at every chemotherapy cycle; however, there were no difference in percent missing by study arm. Furthermore, our modeling approach accommodated missing data, and our effect estimates assume data were missing at random.

Additionally, we tested only resistance training. The combination of aerobic and resistance training may be necessary to modulate body composition in order to impact RDI. Future research such as those trials which are part of the ENICTO (Exercise and Nutrition Interventions to Improve Cancer Outcomes) Consortium³⁹ will continue to test the impact of other and combined exercise modalities on RDI.

Conclusions

Among patients with colon cancer receiving chemotherapy, resistance training did not improve RDI or the overall number of patient-reported moderate/severe toxicities. Resistance training may provide numerous health benefits,^{1, 2} however, these results do not support home-based resistance training to specifically improve chemotherapy treatment intensity in colon cancer patients undergoing adjuvant chemotherapy.