• aerobic exercise;
  • nonsmall cell lung cancer;
  • cardiopulmonary fitness;
  • quality of life


  1. Top of page
  2. Abstract


A feasibility study examining the effects of supervised aerobic exercise training on cardiopulmonary and quality of life (QOL) endpoints among postsurgical nonsmall cell lung cancer (NSCLC) patients was conducted.


Using a single-group design, 20 patients with stage I-IIIB NSCLC performed 3 aerobic cycle ergometry sessions per week at 60% to 100% of peak workload for 14 weeks. Peak oxygen consumption (VO2peak) was assessed using an incremental exercise test. QOL and fatigue were assessed using the Functional Assessment of Cancer Therapy–Lung (FACT-L) scale.


Nineteen patients completed the study. Intention-to-treat analysis indicated that VO2peak increased 1.1 mL/kg−1/min−1 (95% confidence interval [CI], −0.3-2.5; P = .109) and peak workload increased 9 W (95% CI, 3-14; P = .003), whereas FACT-L increased 10 points (95% CI, −1-22; P = .071) and fatigue decreased 7 points (95% CI; −1 to −17; P = .029) from baseline to postintervention. Per protocol analyses indicated greater improvements in cardiopulmonary and QOL endpoints among patients not receiving adjuvant chemotherapy.


This pilot study provided proof of principle that supervised aerobic training is safe and feasible for postsurgical NSCLC patients. Aerobic exercise training is also associated with significant improvements in QOL and select cardiopulmonary endpoints, particularly among patients not receiving chemotherapy. Larger randomized trials are warranted. Cancer 2008. © 2008 American Cancer Society.

Improvements in surgical techniques together with more effective chemotherapeutic regimens has led to significant survival gains for individuals diagnosed with localized (operable) nonsmall cell lung cancer (NSCLC).1 Given improving prognosis, acute and long-term adjuvant treatment sequelae are becoming recognized as important clinical endpoints in the multidisciplinary management of NSCLC.2

Surgery is the only curative-intent treatment for patients with localized NSCLC, but postoperative morbidity is considerable.3–5 Resection of the lung parenchyma reduces ventilatory capacity and reserve. Prospective studies have reported an average reduction in peak oxygen consumption (VO2peak) of 28% and 13% for pneumonectomy and lobectomy, respectively, up to 2 years after resection.3–5 In addition, NSCLC patients are typically older, are current or former smokers, are deconditioned, and commonly present with other concomitant cardiovascular diseases (eg, chronic obstructive pulmonary disease, ischemic heart disease, etc.). Also, up to 70% of lung cancer patients will receive either adjuvant locoregional and/or systemic therapy after resection. The sequential and often concurrent impact of these factors adversely affects the integrative ability of the heart, lungs, vasculature, and circulation to deliver oxygen to the metabolically active skeletal muscles for adenosine triphosphate synthesis to drive muscular contraction, which in turn reduces a patient's ability to tolerate exercise. Poor aerobic fitness may lead to increased susceptibility to other common age-related diseases, poor quality of life (QOL), and likely premature death.6, 7

Accordingly, we conducted a feasibility study examining the effects of supervised aerobic exercise training on aerobic fitness among NSCLC patients who had undergone surgical intervention. Secondary aims were to examine the effects of aerobic training on QOL and other cardiopulmonary endpoints. We hypothesized that supervised aerobic exercise training would be a feasible and safe intervention associated with beneficial effects on primary and secondary study endpoints.


  1. Top of page
  2. Abstract

Setting and Patients

Patients with histologically confirmed stage I-IIIB NSCLC being treated for curative or palliative intent at Duke University Medical Center (DUMC) were potentially eligible for this study. Other major eligibility criteria included 1) Karnofsky performance status ≥70%, 2) ≥30 days after surgical intervention, 3) absence of contraindications to adjuvant chemotherapy, 4) no contraindications to supervised aerobic exercise training based on cardiopulmonary exercise testing (CPET),8, 9 and 5) primary attending oncologist approval. The DUMC institutional review board approved the study, and written informed consent was obtained from all participants before initiation of any study procedures.


By using a prospective, single-group design, potential participants were identified and screened for eligibility via medical record review of patients scheduled for their new patient consultation at DUMC. Immediately after the consultation and oncologist approval, eligible patients were provided with a thorough review of the study by the study coordinator and asked if they were willing to participate. Interested participants completed a study questionnaire, pulmonary function test, and CPET. After the successful completion of the baseline assessments, all participants were scheduled for immediate supervised exercise training. After 14 weeks all baseline assessments were repeated except pulmonary function.

Exercise Training Intervention

The exercise training program was individually tailored to each patient and aimed specifically at increasing VO2peak. All exercise training sessions were supervised by American College of Sports Medicine (ACSM)-certified exercise specialists. Exercise training consisted of 3 aerobic cycle ergometry (Lifestyle Fitness 9500HR; Life Fitness, Franklin Park, Ill) sessions per week on nonconsecutive days for 14 weeks. In week 1, exercise intensity was initially set at 60% of baseline peak workload for a duration of 15 to 20 minutes. Duration and/or intensity were then subsequently increased throughout weeks 2 to 4 up to 30 minutes at 65% peak workload. In weeks 5 and 6, exercise intensity varied between 60% and 65% of peak workload for a duration of 30 to 45 minutes for 2 sessions; in the remaining session, patients cycled for 20 to 25 minutes at ventilatory threshold determined by a systematic increase in the pulmonary ventilation during exercise (VE)/VO2 ratio, whereas VE/VCO2 remained constant.10 From the 7th week onwards, patients performed 2 sessions at 60% to 70% peak workload with 1 threshold workout for 20 to 30 minutes. Finally, in weeks 10 to 14, patients performed 2 sessions at 60% to 70% peak workload with 1 interval session. Interval workouts consisted of 30 seconds at peak workload followed by 60 seconds of active recovery for 10 to 15 intervals.11 All exercise sessions included a 5-minute warm-up and 5-minute cool down. Exercise training intensity and safety were monitored continuously via heart rate, blood pressure, and arterial O2 saturation (SpO2).

Study Endpoints

The primary outcome was change in VO2peak (mL/kg−1/min−1) between baseline and postintervention (14 weeks). Secondary cardiopulmonary fitness endpoints were peak workload, ventilatory threshold, and O2 pulse. We also examined submaximal changes in select cardiopulmonary endpoints (ie, VO2, ventilatory parameters, and heart rate) at an isotime (ie, 75% of baseline peak workload) during the incremental exercise test. Secondary QOL endpoints were overall QOL, fatigue, and QOL subscales.

Study Endpoint Assessments

Incremental Cardiopulmonary Exercise Testing

To determine VO2peak, an incremental, physician-supervised CPET with 12-lead electrocardiogram (ECG) monitoring (Mac 5000, GE Healthcare) was performed at DUMC by ACSM–certified exercise specialists according to CPET guidelines for clinical8 and cancer populations.9 All tests were performed on an electronically braked cycle ergometer (Ergoline, Ergoselect 100, Bitz, Germany) with breath-by-breath died gas analysis (ParvoMedics TrueOne 2400, Sandy, Utah). Preceding exercise, 3 minutes of resting metabolic data were collected before participants began cycling at 20 W. Workloads were then increased 5 to 20 W/min until volitional exhaustion or until a symptom-limitation was achieved. Workload increments were determined by the medical history of the participant and metabolic responses to exercise during the first minute. During exercise SpO2 was monitored continuously using pulse oximetry (Hand-Held Pulse Oximeter, BCI, Waukesha, Wis), and blood pressure was measured noninvasively by manual auscultatory sphygmomanometry every 2 minutes.8 At the end of each workload, rating of perceived exertion was evaluated using the Borg Scale.12 Exercise was terminated if any of the following indications were observed: 1) chest pain, 2) ischemic ECG changes (S-T segment depression or elevation ≥0.1 mV), 3) abnormal blood pressure response (>250 mm Hg systolic; >120 mm Hg diastolic; drop in systolic pressure >20 mm Hg), 4) severe arterial oxygen desaturation (SpO2 ≤85%), and 5) dizziness and/or nausea. CPET procedures were standardized for all participants at baseline and postintervention; the metabolic measurement system was calibrated before and the calibration was checked after each test. All data were recorded as the highest 30-second value elicited during the CPET. Mean percentage of age- and sex-predicted peak heart rate and VO2peak was calculated from the equation provided by Jones et al13 and Fitzgerald et al14 (women) and Wilson and Tanaka15 (men), respectively.

Quality of Life

QOL was assessed using the Functional Assessment of Cancer Therapy–Lung (FACT-L) scale developed for the assessment of QOL in NSCLC patients.16 The FACT-L contains 4 general and 1 lung cancer symptom-specific subscales. General subscales include Physical Well-Being (PWB), Social/Family Well-Being, Emotional Well-Being, and Functional Well-Being (FWB). The 7-item Lung Cancer Subscale (LCS) assesses symptoms commonly reported by lung cancer patients (eg, shortness of breath, weight loss, tightness in chest). The trial outcome index was derived from adding scores on the PWB, FWB and LCS. Fatigue was assessed by the 13-item Fatigue Scale of the FACT measurement system developed specifically for the cancer population.17

Exercise Adherence

Exercise adherence was calculated as a percentage and is equal to the actual number of exercise sessions attended divided by the total number of sessions prescribed (ie, 42). Participants were not permitted to make-up exercise sessions after 14 weeks. Exercise volume was calculated as intensity (W) of each exercise session performed multiplied by the duration (minutes) for the total number of exercise sessions performed during the study.

Medical Characteristics

Medical and demographic data (ie, age, sex, weight, height, smoking history, tumor stage, tumor pathology, extent of resection, adjuvant therapy) were abstracted from medical records. Nonprotocol exercise was assessed by self-report.

Statistical Analysis

Under an intention-to-treat principle, analyses included all enrolled study participants regardless of adherence to the intervention. The dependent t test was used to test whether the mean change across time in the primary and secondary endpoints was significantly different from zero. The dependent t test was also used to test whether change in select primary and secondary endpoints was a function of adjuvant therapy (received chemotherapy vs no chemotherapy). A 2-sided alpha of .05 was used for all tests. Effects are summarized with means and standard deviations.


  1. Top of page
  2. Abstract

The study flow is presented in Figure 1. Participant recruitment took place between January 2006 and December 2007. In brief, 149 patients attended a new patient consultation at DUMC during the study period. Of these, 40 (40 of 149, 27%) met inclusion criteria and 20 (20 of 40, 50%) agreed to participate. Of these, 19 (19 of 20, 95%) completed all study procedures. The 1 patient lost-to-follow-up is excluded from all analyses.

thumbnail image

Figure 1. Study flow is shown.

Download figure to PowerPoint

Participant Characteristics

The baseline characteristics are shown in Table 1. Mean age was 62 ± 11 years, 53% were male, and mean body mass index was 26 ± 8 m/kg.2 Seventy-one percent underwent a lobectomy, 42% received adjuvant chemotherapy, and 80% presented with at least 1 concomitant comorbid disease (47% had hypertension, whereas 32% had type II diabetes mellitus). Mean forced expiratory volume in 1 second (FEV1), forced vital capacity, and diffusing capacity of the lung for carbon monoxide were equal to 71%, 89%, and 83% of predicted, respectively. Mean time from diagnosis was 30 ± 3 days. No adverse events were observed during the incremental CPET.

Table 1. Characteristics of the Participants (n=19)
VariableNo. (%)Mean±SD
  • *

    Numbers do not equal 100% because of overlap between categories.

  • BMI indicates body mass index; VATS, video-assisted thoracoscopic surgery; FEV1, forced expired volume; FVC, forced vital capacity; TLC, total lung capacity; RV, residual volume; DLCO, diffusion capacity of the lung for carbon dioxide.

Age, y 62±11
Men, %10 (53) 
Weight, kg 76±16
BMI, kg/m2 26±8
Smoking history  
 Current2 (11) 
 Former17 (89) 
Histologic features  
 Adenocarcinoma12 (63) 
 Squamous5 (26) 
 Undifferentiated2 (11) 
 IA8 (42) 
 IB3 (16) 
 IIA2 (11) 
 IIB2 (11) 
 IIIB4 (21) 
Extent of Resection*  
 Lobectomy12 (71) 
 Pneumonectomy1 (6) 
 Bilobectomy1 (6) 
 Wedge1 (6) 
 VATS1 (6) 
 Bronchoscopy3 (18) 
Adjuvant therapy  
 Received Chemotherapy8 (42) 
 Received Radiotherapy1 (5) 
Concomitant comorbidities*  
 Coronary artery disease3 (16) 
 Type II diabetes mellitus6 (32) 
 Hypertension9 (47) 
 Hyperlipidemia5 (26) 
 Asthma1 (5) 
 Atrial fibrillation2 (10) 
 Osteoarthritis2 (10) 
Pulmonary function data  
 Predicted FEV1, L (%) 2.2±0.5 (71)
 Predicted FVC, L (%) 3.7±0.9 (89)
 FEV1/FVC, % 62±11
 Predicted TLC, L (%) 6.3±1.6 (98)
 Predicted RV, L (%) 2.6±1.6 (115)
 Predicted DLCO, L (%) 19.2±5.8 (83)

Exercise Adherence

The overall adherence rate was 85% (range, 29%-100%), with patients completing a mean of 36 sessions from a total of 42 planned sessions. There was no change in nonprotocol exercise over the intervention period. No adverse events were observed during aerobic training sessions.

Intention-to-Treat Analyses

Changes in cardiopulmonary fitness endpoints are shown in Table 2. No significant changes in any cardiopulmonary endpoints at rest were observed from baseline to postintervention. Mean VO2peak increased 1.1 mL/kg−1/min−1 (95% confidence interval [CI], −0.3-2.5; P = .11), and peak workload increased 9 W (95% CI, 3-14; P = .003). For submaximal exercise isotime responses, VO2, ventilation, or heart rate were generally lower at the same relative workload (75% of baseline peak workload) in the postintervention incremental CPET; however, none of these changes reached statistical significance (P > .05).

Table 2. Mean Changes in Cardiopulmonary Endpoints (n=19)
VariableBaseline, Mean±SDPostintervention, Mean±SDMean Difference [95% CI]P
  1. Data are presented as mean ± standard deviation (SD).

  2. CI, confidence interval; VO2peak, peak oxygen consumption; METS, metabolic equivalent unit; RER, respiratory exchange ratio; VE, ventilation; RR, respiratory rate; SpO2, arterial oxygen saturation; RPE, regular pulse excitation; VT, ventilatory threshold.

 VO2peak, mL/kg−1/min−115.7±3.316.8±3.91.1 [−0.3 to 2.5].11
 VO2peak, L/min−11.16±0.31.26±0.30.10 [−0.01 to 0.19].10
 Predicted VO2peak, mL/kg−1/min−1, %62664 [−0.9 to 10].10
 Workload, W74±983±229 [3 to 14].003
 Heart rate, beats/min−1124±3.7130±196 [−0.4 to 13].06
 Predicted heart rate, beats/min−1, %76804 [−0.2 to 8].06
 Systolic blood pressure, mm Hg164±20162±20−2 [−14 to 10].74
 Diastolic blood pressure, mm Hg81±780±7−1 [−3 to 1].40
 O2 pulse, mLO2/beat13±213±20 [−0.6 to 1.2].52
 METS4.5±0.94.9±1.10.4 [−0.02 to 0.81].06
 RER1.06±0.041.08±0.090.02 [−0.02 to 0.06].28
 VE, L/min45±1049±124 [−0.1 to 7].06
 RR33±832±5−1 [−4 to 2].69
 Tidal volume, L1.4±0.41.6±0.40.2 [0.02 to 0.19].01
 SpO2, %95±495±30 [−1.4 to 1.1].80
 RPE16±217±21 [−0.5 to 2].23
 VO2, at VT, mL/kg−1/min−1, %71±970±10−1 [−6 to 3].58
 Workload at VT, W61±2564±183 [−6 to 12].52
 Reason for test termination, n (%)    
  Leg fatigue8 (42)9 (47)
  Dyspnea7 (37)6 (32)
  Both4 (21)4 (21)

Table 3 displays the changes in QOL endpoints. Mean FACT-L, FACT-General (FACT-G), and trial outcome index increased 10 points (95% CI, −1-22; P = .07), 8 points (95% CI, −2-19; P = .09), and 9 points (95% CI, 1-17; P = .03), respectively. Significant favorable changes were also observed for functional well-being (P = .007) and fatigue (P = .03), and the lung cancer subscale (P = .10) approached significance.

Table 3. Mean Changes in Quality of Life Endpoints (n=19)
VariableBaseline, Mean±SDPostintervention, Mean±SDMean Difference [95% CI]P
  1. Data are presented as means ± standard deviation (SD).

  2. CI, confidence interval; FACT, Functional Assessment of Cancer Therapy.

Global scores    
 FACT–Lung (0-136)98±18108±1410 [−1-22].07
 FACT–General (0-108)80±1688±138 [−2-19].09
 FACT trial outcome index (0-84)56±1264±109 [1-17].03
 Physical well-being (0-28)20±623±53 [−1-8].15
 Functional well-being (0-28)17±520±54 [1-6].007
 Social well-being (0-24)25±324±5−1 [−2-3].85
 Emotional well-being (0-28)18±520±32 [−1-5].27
 Fatigue (0-52)19±812±8−7 [−1 to −17].03
 Lung cancer subscale (0-28)19±421±22 [−1-5].11

Per Protocol Analysis

Changes in select primary and secondary study endpoints by chemotherapy (received chemotherapy vs no chemotherapy) are presented in Table 4. Exercise adherence was 93% and 72% for patients receiving and not receiving chemotherapy, respectively. For patients not receiving adjuvant chemotherapy (n = 11), VO2peak increased 1.7 mL/kg−1/min−1 (P = .008). Significant increases were also observed for peak heart rate (P = .05), peak workload (P < .001), and workload at ventilatory threshold (P = .05). Changes in submaximal exercise isotime responses were not statistically significant (P < .05). Concerning select QOL endpoints, for patients not receiving adjuvant chemotherapy, significant increases were observed for all QOL outcomes except the lung cancer subscale (P = .22). For patients receiving adjuvant chemotherapy (n = 8), there were no significant changes in any cardiopulmonary or QOL outcome (P > 0.05), except VO2peak at ventilatory threshold, which significantly decreased over the study period.

Table 4. Mean Changes in Cardiopulmonary and Quality of Life Endpoints by Adjuvant Therapy (Chemotherapy vs No Chemotherapy) (n=19)
VariableBaseline, Mean±SDPostintervention, Mean±SDMean Difference [95% CI]P
  1. Data are presented as mean ± standard deviation (SD). CI indicates confidence interval; VO2peak, peak oxygen consumption; VT, ventilatory threshold; FACT, Functional Assessment of Cancer Therapy.

Cardiopulmonary endpoints    
   Chemotherapy16.7±2.917.0±3.70.3 [−3 to 3].84
   No chemotherapy15.0±3.516.7±4.11.7 [0.6 to 3.0].008
  Heart rate, beats/min−1    
   Chemotherapy (n=8)135±20138±143 [−10 to 15].63
   No chemotherapy (n=11)116±18125±2011 [−1 to 19].05
  VO2peak, mL/min−1    
   Chemotherapy1.26±0.31.27±0.40.01 [−0.23 to 0.26].94
   No chemotherapy1.10±0.21.25±0.30.15 [0.05 to 0.24].007
  Workload, W    
   Chemotherapy83±2785±272 [−7 to 12].55
   No chemotherapy69±2382±2013 [8 to 19]<.001
  VO2 at VT, mL/kg−1/min−1, %    
   Chemotherapy69±863±7−6 [−11 to −1].03
   No chemotherapy72±974±112 [−0.7 to 4.5].48
  Workload at VT, W    
   Chemotherapy68±1658±20−10 [−21 to 3].12
   No chemotherapy55±1967±1912 [1 to 24].05
Quality of life endpoints    
 FACT–Lung (0-136)    
  Chemotherapy (n=8)99±799±110 [−15 to 14].97
  No chemotherapy (n=11)97±23114±1217 [−2 to 1].05
 FACT–General (0-108)    
  Chemotherapy80±579±10−1 [−14 to 11].82
  No chemotherapy78±2093±1015 [1 to 29].05
 FACT trial outcome index (0-84)    
  Chemotherapy56±559±103 [−9 to 15].52
  No chemotherapy55±1668±813 [1 to 25].04
 Fatigue (0-52)    
  Chemotherapy18±416±7−2 [−10 to 7].62
  No chemotherapy20±1010±7−10 [−18 to −2].03
 Lung cancer subscale (0-28)    
  Chemotherapy19±320±11 [−1 to 4].23
  No chemotherapy19±522±33 [−2 to 7].22


  1. Top of page
  2. Abstract

The principal finding of this pilot study was that a short-term, moderate to high-intensity supervised aerobic exercise training program was feasible, safe, and well tolerated among newly diagnosed NSCLC patients who had recently undergone surgical intervention. Moreover, analyses indicated significant improvements in QOL and select cardiopulmonary endpoints, particularly among patients not receiving chemotherapy. To our knowledge, this is the first study to examine the independent effects of aerobic training among this patient population in this setting.

Several recent randomized trials have examined the effects of exercise training as an adjunct supportive therapy in a broad range of cancer patients differing in terms of cancer diagnosis, disease stage, and treatment.18–21 Overall, the extant literature indicates that exercise training is safe and well tolerated among oncology patients. It is not clear, however, whether the low incidence of events reflects the true safety or less than optimal exercise test methodology and/or data reporting in clinical oncology research.9 Consistent with these findings, we observed no adverse events during the incremental CPET or supervised exercise training sessions. Moreover, adherence to exercise training was excellent (85% of planned sessions) and above conventionally accepted levels for exercise intervention trials in clinical populations.22 Patients in this study were older, had poor exercise tolerance, presented with a diverse range of concomitant comorbidities, had recently undergone surgical excision of lung tissue, and almost 1-third were receiving platinum-based chemotherapy. Thus, demonstration of the feasibility and safety of moderate- to high-intensity aerobic training in the present context is novel and important.

A second important finding of this study was the significant improvements in patient-reported outcomes (PROs). Results indicated that global QOL scores increased 8 to 10 points over the course of the intervention, with even stronger changes among patients not receiving chemotherapy. These findings may have important clinical significance. In a recent systematic review of 39 studies (12 were among lung cancer patients) involving 13,874 cancer patients, Gotay et al reported that PROs, especially QOL, provided prognostic information beyond conventional clinical assessments (eg, performance status, stage, etc.).23 Furthermore, a change of 4 points or more in the FACT-G is considered the minimal clinically important difference (CID).24 The CID for fatigue (ie, a 10-point change) was also achieved in the per-protocol analysis.24 The majority of, but not all, studies have also reported significant improvements in PROs, particularly global QOL and fatigue, after exercise training in the oncology setting.18–21 The mechanisms underlying the improvements in PROs with aerobic training are not known. Courneya et al reported that change in VO2peak was strongly correlated with change in QOL after aerobic training among breast cancer patients.25 In this study, changes in cardiopulmonary endpoints were not associated with changes in any PRO. Thus, other factors, including social interaction (between participants and study personnel), improvements in physical competence and self-confidence, positive feedback, coping with their cancer diagnosis and treatment, and distraction, may explain our results.

An intriguing finding of this study was that although intention-to-treat analyses indicated a nonsignificant improvement in VO2peak per-protocol analyses indicated that this improvement was largely restricted to those patients not undergoing chemotherapy. Improvements in VO2peak were 0.3 mL/kg−1/min−1 or 2% in patients receiving chemotherapy. Clearly, without a nonintervention control group, it is not known whether maintenance of VO2peak during chemotherapy is important. The direct effects of NSCLC chemotherapeutics on cardiopulmonary function are not fully known, although platinum-based regimens cause reductions in FEV1 and anemia,26, 27 which are expected to attenuate normal physiologic adaptations to aerobic exercise training. Accordingly, based on the current evidence, further study of aerobic training among patients undergoing adjuvant chemotherapy does not appear warranted at this time.

In contrast, the improvement in VO2peak was 1.7 mL/kg−1/min−1 or 11.3% among those not receiving chemotherapy. Following traditional aerobic exercise training guidelines (3-5 days/week at 50% to 75% of baseline VO2peak for 12-15 weeks), an ∼15% improvement is the generally accepted “clinically important” change in VO2peak in noncancer populations.28, 29 In noncancer populations, cardiopulmonary fitness has become established as a strong, independent predictor of mortality across a broad range of adult patients with chronic disease.6, 7, 30 No study to date has examined the prognostic value of cardiopulmonary fitness on survival in patients with lung cancer. Nevertheless, subjective measures of functional capacity (a surrogate of cardiopulmonary fitness) used in lung cancer management (ie, performance status scoring systems) are consistent predictors of mortality.31–34 Although the magnitude of VO2peak improvement was below the accepted “clinically important” change, given the well-established clinical value of VO2peak and Karnofsky performance status, we believe that these results warrant further investigation. Moreover, given that the prognostic value of VO2peak in lung cancer is not known, an improvement of ∼11% may in fact be meaningful, especially in the context of severe deconditioning and high postsurgical morbidity. Prospective, observational studies are required to fully address this question.

Of importance, although per-protocol analyses indicated significant improvements in VO2peak among patients not undergoing chemotherapy, the magnitude of benefit was modest (11%). Our group recently reported that presurgical aerobic exercise training (cycle ergometry, 5 days/week at 60%-100% VO2peak) was associated with a 15% to 22% improvement in VO2peak among 20 patients with suspected NSCLC.11 The contrasting findings may be explained by differences in exercise prescription and/or limitations to exercise between the 2 studies. In our presurgical study, aerobic training consisted of 5 cycle ergometry sessions/week over 4 to 6 weeks (∼30 exercise sessions) compared with 3 cycle ergometry sessions/week over 14 weeks (∼42 sessions) in this study. Although the total exercise volume was higher, the relative exercise “dose” was higher in the presurgical study, given the greater frequency of sessions/week over a shorter period. These results suggest that a high-exercise “dose” may be required to induce favorable adaptations in NSCLC patients who are severely deconditioned after surgical resection and related disease pathophysiology. The significant correlation between exercise volume and change in VO2peak (r = 0.54, P = .02) observed in this study support this notion (data not presented).

The contrasting findings may also be associated with greater exercise limitations among patients after extensive pulmonary resection. An obvious potential explanation is a ventilatory limitation or inadequate gas exchange after removal of a substantial portion of lung parenchyma. However, several elegant studies have demonstrated that VO2peak is not limited by ventilation or diffusion capacity.35–38 Indeed, our results corroborate these findings, as only 4 individuals demonstrated any evidence of ventilatory limitation,8 and only 1 had an oxyhemoglobin saturation at peak exercise <90%. Thus, improvements in other components involved in O2 transport39 must contribute to the modest improvements in VO2peak after aerobic training observed here.

Other potential limiting mechanisms include decreased cardiac output and/or peripheral muscle limitation.39 In this study, ∼65% of patients subjectively reported leg fatigue or a combination of leg fatigue and dyspnea as the major symptom(s) responsible for exercise termination. These findings indicate that O2 delivery and/or skeletal muscle limitation may contribute to the reduced VO2peak. Skeletal muscle limitation is well documented in chronic obstructive pulmonary disease (COPD) patients who exhibit similar disease etiology and symptoms as patients with NSCLC.40 However, it is currently not known whether skeletal muscle limitation is because of muscle dysfunction per se or muscle weakness because of deconditioning (disuse).41 Major contributors to skeletal muscle dysfunction in COPD include direct skeletal myopathy (from the use of oral corticosteroids), and high levels of systemic inflammation and oxidative stress (from underlying disease and therapy).41 Importantly, NSCLC patients are also deconditioned, receive corticosteroids, and have high levels of systemic inflammation.42 The contribution of these factors to exercise intolerance in NSCLC requires investigation.

Given all the above, exercise training programs that target both central and peripheral factors limiting exercise tolerance in NSCLC will be required to ensure optimal improvements in VO2peak. The combination of resistance and aerobic training may provide the optimal solution. In COPD, standard resistance training guidelines (ie, 3-5 times/week, 50%-75% of 1 repetition maximum for 12-24 weeks) have been demonstrated to improve skeletal muscle oxidative capacity, muscle endurance, and strength, as well as whole body exercise tolerance (VO2peak and 6-minute walk distance).40 It is postulated that the improvements in skeletal muscle function and strength from resistance training will not only have independent effects on aerobic fitness, but will allow patients to perform aerobic exercise training at higher intensities to elicit greater improvements in exercise tolerance and health-related QOL than either alone.

This study does have limitations, including the relatively small sample size, the nonrandomized design, and the exclusion of patients with contraindications to aerobic training and significant comorbid disease. Thus, significant patient selection bias likely exists because of the transparent nature of the study, with patients highly motivated to exercise and having better prognosis being more likely to participate. Indeed, only ∼13% of screened patients were actually recruited. However, although we likely recruited a highly motivated cohort of patients, only ∼10% of patients met national exercise guidelines (data not presented) and VO2peak was, on average, 38% below that for age- and sex-matched sedentary individuals14, 15 and comparable to that reported in large-cohort studies investigating the effects of surgical resection on VO2peak.4, 5 These findings suggest that our sample may be representative of a wider population of postoperative NSCLC patients. Finally, the main purpose of this study was to examine the feasibility and preliminary efficacy of aerobic training in this setting.

Another noteworthy limitation is that beneficial changes in VO2peak may be explained by the natural history of postoperative recovery. To minimize this issue, patients were recruited at least 3 weeks postresection (mean, 30 ± 3 days). Nevertheless, the time course of postsurgical ‘natural’ recovery in VO2peak has not been well characterized. Therefore, it is entirely possible that our observed findings may be partially explained by this phenomenon. Withstanding the significant selection bias and spontaneous postoperative recovery, we believe the present findings provide strong promising evidence for further investigation of exercise training in operable lung cancer.

In conclusion, this pilot study provided proof of principle that supervised aerobic training may be safe and feasible for select postsurgical NSCLC patients. Aerobic exercise training is also associated with significant improvements in QOL and select cardiopulmonary endpoints, particularly among patients not receiving systemic therapy. Larger randomized trials are now warranted.


  1. Top of page
  2. Abstract