INTRODUCTION

In 2020 approximately 1.8 million new cases of cancer were diagnosed, appending the 16.9 million cancer survivors (CS) currently living in the United States (1). Cancer survivors endure cardiovascular, metabolic, and/or musculoskeletal toxicities culminating in common comorbidities such as hypertension, hyperlipidemia, osteoarthritis, diabetes mellitus, and coronary artery disease (2). Side effects of cancer and its concurrent treatments can linger months to years after treatment is concluded (3). In addition, negative lifestyle factors such as inactivity, poor nutrition, and weight gain are common following cancer diagnosis (4). Collectively, treatment-related toxicities, comorbidities, and poor lifestyle contribute to declinations in cardiorespiratory function (CRF). Cardiorespiratory function, as measured by maximal volume of oxygen consumption (Vo₂max), is an independent predictor of all-cause mortality and is recognized as a vital sign which should b routinely assessed in clinical practice (2, 4–11). To combat low CRF, exercise can be prescribed to improve physiological function and reduce treatment effects; and in turn improve mortality rates, prognosis, recurrence, and incidence of certain cancers (10,12,13).

The design of tailored exercise prescriptions requires assessment of CRF yielding accurate Vo₂max values (4,9,12,14). A graded exercise test (GXT) performed to volitional fatigue with concurrent gas analysis via a metabolic cart is the gold standard method of measurement of Vo₂max (15). When conducting GXTs, treadmill protocols are the preferred mode as they demonstrate consistently higher and more accurate Vo₂max values compared to cycle ergometry (16–20). The most common treadmill protocol in North America is the Bruce (19,21,22). However, the Bruce's large and abrupt increases in speed and grade can be challenging for participants such as CS who likely present with neuropathy, cachexia, lymphedema, pain, and/or fatigue (1,6–9,23). It has been reported that Vo₂max measured by GXT may underestimate CRF in CS most likely because of physiological limitations and reduced confidence to perform maximal effort (14,17). This protocol may specifically contribute to these risks of Vo₂max underestimation in CS.

Administration of Vo₂max tests are costly, time-consuming, and requires expensive equipment and qualified technicians to perform. To save time and money, peak volume of oxygen consumption (Vo₂peak) tests are used in place relying on validated predictive equations to estimate the metabolic cost of exercise dependent on termination time without the use of a metabolic cart (24). Predictive equations are accepted as accurate estimations of Vo₂max during steady-state exercise when Vo₂max criteria such as Vo₂ plateau, lactate, or maximal heart rate (HR) are not measured or cannot be met (24). The American College of Sports Medicine (ACSM) Foster equation for sedentary men, the Pollock equation for active and sedentary women, and the McConnell & Clark equation for cardiac patients and elderly persons are correlated with Vo₂max and used to estimate CRF using the Bruce treadmill protocol (24). However, these prediction equations have been reported to underestimate Vo₂peak, especially in those with low fitness (19). The standard error of these estimations is ±1 metabolic equivalent of task as reported by Heyward and Gibson or 4 mL·kg⁻¹·min⁻¹ as reported by Pollock and colleagues, which is significant in populations with reduced exercise capacity such as CS (19,24).

Currently, there is only 1 cancer-specific (CANCER) treadmill protocol validated for CS (23). To account for cancer-specific toxicities, this treadmill protocol increases in speed and grade gradually with shorter stages, allowing CS suffering from significant and debilitating side effects to progress further in the test, allowing for a greater and more accurate measurement of Vo₂peak (20). The ACSM metabolic equations are validated predictive equations for the CANCER treadmill protocol as they yield equivalently accurate Vo₂peak values as Vo₂max when measured via gas analysis (23).

In summary, a valid, cost effective, and timely method to estimate CRF is needed for clinical and rehabilitative purposes in CS. Although the CANCER treadmill protocol has been established, the Bruce treadmill protocol is commonly used despite lacking validation of its predictive equations in CS. Additionally, the Bruce protocol's inherent difficulty and frequent inaccuracies when estimating Vo₂peak in nonathletic populations suggests it may be unsuitable to measure CRF for the cancer population (20). Therefore, the purpose of this study was to compare the Bruce protocol and standard prediction equations to the CANCER protocol using both gas analysis and the validated prediction equations in a group of CS. A secondary purpose investigated systolic cardiovascular work, e.g., peak values of HR, blood pressure (BP), and rate of pressure product between the Bruce and CANCER protocols.

METHODS

CANCER Protocol

This protocol consisted of 21, 1-minute stages. Speed and/or grade were increased at the completion of each stage. Details of this protocol are presented in Table 1.

TABLE 1. CANCER protocol.

View Fullscreen

Measured Vo₂peak during the CANCER protocol (CANCERmet) was obtained using a Cosmed research grade metabolic cart (Cosmed, Chicago, Illinois) Expired gases were continuously collected where Vo₂ and carbon dioxide output were recorded once every 3 seconds. Calibration of the metabolic cart was performed before each test with a 3L syringe and precision gas mixtures. Before each test each subject received an explanation as to how the test was conducted and why the metabolic cart was being used. A respiration mask was attached to the subject's face with tubes connecting the mask and metabolic cart.

Estimated Vo₂peak of the CANCER protocol (CANCERest) was calculated using ACSM walking and running equations on the last completed stage of the CANCERmet protocol. This methodology has been previously validated as yielding accurate Vo₂peak values (23). The participant was either walking or running by the last completed stage, which determined the specific equation used. If the subject was walking when the test was terminated, the following equation was used: Vo₂peak = (0.1 × S) + (1.8 × S × G) + 3.5; where S = speed and G = grade (15). If the subject was holding onto the handrails and walking at the termination of the test, the following correction equation was used: Vo₂peak = 0.694 ([0.1 × S] + [1.8 × S × G] + 3.5) + 3.33 (15,25). If the subject was running when the test was terminated, the following equation was used: Vo₂peak = (0.2 × S) + (0.9 × S × G) + 3.5 (15). If the subject was running at the end of the test and holding on to the handrails the following correction equation was used: Vo₂peak = 0.694 ([0.2 × S] + [0.9 × S × G] + 3.5) + 3.33 (15,25).

RESULTS

Subject demographics are summarized in Table 2. Participants included 38 females and 9 males; all common cancer types were represented in our sample. There were no significant differences between any resting or descriptive characteristics. All participants were able to achieve Vo₂peak criteria, and no adverse effects were observed during or after any of the tests.

TABLE 2. Patient demographics and treatment characteristics.

View Fullscreen

All mean peak exercise values are presented in Table 3. Vo₂peak (mL·kg⁻¹·min⁻¹) was significantly lower in CANCERmet compared to the Bruce (Figure 1), significantly lower in CANCERest compared to the Bruce, and there was no significant difference between CANCERmet and CANCERest).

View Figure

• Download as Powerpoint
• Download Figure

TABLE 3. Mean peak exercise values for cancer survivors. ^a

View Fullscreen

Mean peak HR and SBP were significantly lower in the Bruce compared to CANCERmet, respectively (P = 0.01). Rate-pressure product, which is the product of both HR and SBP and a reflection of systolic myocardial workload and oxygen consumption, was significantly lower in the Bruce compared to the CANCER protocol (Figure 2). Average total treadmill time was significantly lower on the Bruce compared to the CANCER (Figure 3).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

The purpose of this study was to determine whether the Bruce protocol accurately estimates exercise capacity in CS, when compared against the CANCER protocol. Both protocols can obtain Vo₂peak using a metabolic cart or estimate Vo₂peak through standard prediction equations. The Bruce protocol has several validated prediction equations (22,27,28,30) including the standardized set recommended by ACSM (15). Likewise, the CANCER protocol has also been validated as producing accurate Vo₂peak values when using prediction equations in CS, which is confirmed in this work (23). Our principal finding indicated that Vo₂peak when estimated using the Bruce is significantly higher than both the actual and estimated values yielded by CANCERmet and CANCERest, respectively. Concerning, this larger estimation of aerobic function did not correspond to peak physiological responses representative of systolic myocardial work. This suggests that subjects exercised at a higher capacity during the CANCER protocol yet yielded a Vo₂peak value significantly lower than those measured by the Bruce. Thus, the Bruce protocol overestimated Vo₂peak in this sample of CS.

GXTs require linearly increasing work rates in order to elicit the highest physiological responses to exercise (e.g., HR, BP, rating of perceived exertion, and lactate threshold) and thus corresponds with maximal Vo₂ (30,31), Specifically, GXTs rely on the Fick Principle and the relationship between Vo₂, cardiac output, and oxygenated arterial-venous difference (17). Gas analysis itself allows measurement of Vo₂, while vitals such as HR and BP allow clinicians insight into the linear relationship of cardiac output, the cardiovascular component of the Fick equation. Our results demonstrate that significantly higher HR, SBP, and rate-pressure product values were achieved in the CANCER protocol, reflective of true maximal cardiac output. Of note, 47% of CS exceeded age-predicted HR maximum estimates during the CANCER protocol, in comparison to only 23% during the Bruce. Likewise, treadmill time was greater when subjects completed the CANCER. These outcomes are logical as the CANCER protocol was specifically designed to accommodate treatment-related aftereffects with the design of shorter and less intense stages, allowing the CS to continue further into the protocol and elicit higher physical responses.

Although Bruce Vo₂peak values were significantly higher, the protocol yielded significantly lower peak physiological responses compared to the CANCER protocol. These observations suggest that most participants did not reach maximal effort during the Bruce, which may be attributed to the protocol design itself. The Bruce protocol's workload was designed for highly functioning individuals, which has been reported to be too intense for even the average individual to perform (24). Sharp inclines between stages could lead to a greater reliance on anaerobic metabolism, resulting in subjects fatiguing before reaching true maximal volition (32–35). In addition, CS may have cancer cachexia and/or other treatment-related muscular toxicities which may adversely affect systemic mitochondrial function and subsequent adenosine triphosphate generating capacity, which can cause severe fatigue with or without physical exertion (36,37). The higher intensities and more dramatic workload changes of the Bruce protocol may have exacerbated these decrements, resulting in lower physiological responses and early test termination due to muscular fatigue, not cardiovascular fatigue, which is the assumption and goal of a GXT. This response to the Bruce is not exclusive to CS. Others have reported difficulties achieving target HR in a general patient population because of their physical inability to keep up with the large incremental changes in workload (20,38,39). Similarly, Pollock and colleagues demonstrated small but significant differences in Vo₂max, HR, and SBP between the Balke-Ware and Bruce treadmill protocols in healthy women (19). Myers and colleagues also reported reduced Vo₂peak values as measured by an individualized ramp test, a protocol with gradual work increments, compared to tests using standard increments like the Bruce protocol in patients with reduced oxygen kinetics (20). Likewise, Pollock and colleagues found significant differences between Vo₂ as measured by the Balke-Ware, Bruce, Ellestad, and a continuous multistage running protocol in a population of older adults. Despite a lower Vo₂ value, the Balke-Ware, the protocol with the most gradual rate of progression in metabolic equivalent of task cost, elicited the maximum physiological responses to exercise as compared to the other 3 protocols (30).

In this study, the Bruce yielded lower peak physiological responses despite estimating higher values of aerobic function, i.e., Vo₂peak values. These results agree with other studies which demonstrated that Bruce Vo₂peak equations inaccurately estimate CRF in active individuals (17) as well as in sedentary or chronically diseased individuals (14,20,30,35,40). In addition to the design of the Bruce, Aguiar and colleagues reported that the ACSM metabolic equations inaccurately predicted Vo₂peak using the Bruce because of the steep increases in workload irrespective of steady-state oxygen consumption necessary for estimation purposes (41). Myers and colleagues similarly demonstrated that the accuracy of estimation of Vo₂ from the Bruce is poor in those with heart disease or reduced oxygen kinetics (20). They concluded that protocols with large and unequal increments between stages result in overestimation of Vo₂ and greater variability due to a nonlinear relation between oxygen uptake and work rate (20). Our results mirror findings in other chronic diseased populations who performed treadmill protocols with smaller work rate increments and yielded reduced Vo₂peak as compared to the Bruce (20). Thus, the CANCER may have yielded a lower, but more accurate Vo₂peak compared to the Bruce because of the decreased intensity and shorter stages.

CLINICAL IMPLICATIONS

Cardiorespiratory fitness is inversely correlated with all-cause mortality and cancer recurrence. A decline in CRF and Vo₂peak is commonly observed after cancer treatment because of treatment-related side effects and toxicities on physiological systems as well as inactivity (8,42,43). Exercise-based rehabilitation programs are becoming more widely used in comprehensive cancer care with the mission to improve physiological functioning, specifically CRF. For this reason, it is imperative that Vo₂peak be measured with accuracy so clinicians have confidence in prescribing exercise from these results. GXTs are the most common method of assessing aerobic function, and of these, estimation of Vo₂peak from prediction equations is the most economic and feasible way to quantify the result (38). When conducting GXTs, the Bruce protocol and its corresponding prediction equations are used most often, but our data suggests that not only is the Bruce protocol too intense for use in CS but resulted in overestimated and inaccurate Vo₂peak values. Any miscalculation of Vo₂peak is deleterious when prescribing exercise in this population. Underpredicting Vo₂peak could potentially prevent the CS from exercising at the minimum threshold required for improvements in function (14). In contrast, and more concerning, overpredicting Vo₂peak could lead to overtraining at too high of an intensity, which could compromise the patient's health and the efficacy of the rehabilitation program (44). Data presented in this study demonstrates that unlike the Bruce, the CANCER protocol provides accurate Vo₂peak values, higher physiological responses to exercise, and allows proper CRF assessment in patients who have limited functional capacity. We suggest that clinicians discontinue the use of the Bruce treadmill protocol to measure CRF in CS, as it overestimates Vo₂peak and yields lower physical exertion. Instead, we propose that the CANCER protocol be used as the standard method to measure aerobic capacity in the design, implementation, and surveillance of exercise-based rehabilitation programs for CS.