Effects of Structured Exercise Programs on Self-Reported Health-Related Quality of Life in Patients with Advanced Non-Small Cell Lung Cancer: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Abstract

Current treatments for metastatic non-small cell lung cancer (NSCLC) have improved survival but remain non-curative, primarily aiming to control disease and to extend life. Structured exercise has demonstrated clinical and quality-of-life benefits in early-stage NSCLC. This systematic review assesses the impact of adjunctive exercise programs on health-related quality of life (HRQoL) in advanced NSCLC patients, with safety as a secondary outcome. Of 1168 studies screened, 13 met the inclusion criteria. All eligible studies were included in the systematic review, and three underwent meta-analysis. Using Synthesis Without Meta-analysis (SWiM), the findings were heterogeneous: four studies showed positive outcomes, two had mixed results, and seven showed negative outcomes. Meta-analysis of studies utilizing the EORTC-C30 tool demonstrated a positive mean difference of 1.57 (95% CI: 0 to 3.14), indicating a trend toward HRQoL improvement. Safety analyses largely revealed no major adverse events related to exercise interventions. Future studies must therefore be designed to account for confounders intrinsic to the underlying disease of study participants to better determine both the efficacy and the safety of structured, adjunctive exercise programs in this patient population.

1. Introduction

Lung cancer remains the leading cause of cancer-related mortality worldwide [1]. Patients diagnosed with this disease experience significant functional impairment from symptoms and deconditioning [2,3]. Related comorbidities, lifestyle choices such as smoking, and adverse effects of treatment compound this problem further [4]. Taken together, all these lead to poor quality of life (QoL) [3].

Supervised and structured exercise can reduce post-operative complications, shorten hospital stays, and improve quality of life in patients with resected lung cancer [4,5]. Current treatments for metastatic lung cancer, especially non-small cell lung cancer (NSCLC), have improved patient survival but are ultimately not curative and are mainly aimed at controlling disease and extending life [6,7].

Single-group studies on exercise training for patients with advanced lung cancer have shown improvements in exercise capacity, daily functionality, anxiety, and emotional well-being [8]. Exercise programs usually comprise moderate-intensity aerobic and resistance training. While physical training in patients with advanced lung cancer appears to be safe, some studies have reported a few minor musculoskeletal events, such as muscle pain, attributed to these interventions [8,9].

Cancer care needs to be holistic, and every opportunity to improve the disposition of patients’ needs to be taken advantage of to ensure the best possible quality of life for those undergoing active lung cancer treatment, regardless of whether they can hope for cure [3]. This includes employing adjunctive treatment approaches that can complement standard-of-care (SOC) medical interventions. Based on evidence in early-stage NSCLC and in small studies on patients with advanced NSCLC, exercise programs may be promising additions to the standard medical care of patients with metastatic NSCLC. A recent systematic review on the benefits of exercise in patients with advanced stage lung cancer showed very low to low certainty for the effectiveness of exercise training in this population. The review included only six published randomized controlled trials (RCTs) that had small sample sizes and an overall high risk of bias. Notably, several additional studies have since been published on this topic [8,9].

Wellness, which is central to the concept of quality of life, is multi-dimensional, and its definition cannot be distilled to the absence of illness alone [10]. Therefore, multidimensional metrics such as health-related quality of life (HRQoL) are best suited for assessing the overall well-being of patients. Health-related quality of life aims to measure the perceived physical and mental health of an individual or a group over time [10]. Several tools have been designed and validated to measure this, such as the CDC HRQOL–14 “Healthy Days Measure” and the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 [10,11].

The aim of this review is to revisit the value of exercise programs, specifically adjunctive exercise programs, in improving the HRQoL of patients with advanced/metastatic NSCLC.

2. Materials and Methods

2.1. Study Protocol and Registration

This systematic review protocol was prepared and conducted with guidance from the Cochrane Handbook for Systematic Reviews of Interventions ver. 6.3 and in accordance with Methodological Expectations of Cochrane Intervention Reviews (MECIR) standards [12,13]. In cases wherein quantitative meta-analysis could not be performed, Synthesis Without Meta-analysis (SWiM) guidelines were applied [14]. Reporting conformed with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. For the purposes of this protocol, the PRISMA 2020 checklist was used [15]. This protocol was registered with PROSPERO: CRD42022362150.

2.2. Eligibility Criteria

2.2.1. Study Characteristics

Studies that report HRQoL data from a validated measuring tool, such as the EORTC QLQ-C3012, Quality of Life Questionnaire [16], Functional Limitations Profile (FLP) [17], Short Form (SF-36) [18], EuroQol (EQ-5D) [19], FACT-G [20], FACT-L [21], and CDC HRQOL-14 [22], were included in this review. The included studies were all peer-reviewed. There were no restrictions on the period of publication. Only studies published in the English language were considered for inclusion. In the case of duplicated or shared data from the same population, such as in follow-up analyses from a common evolving dataset, the study with the largest sample size was included. Studies not meeting the inclusion criteria described above were excluded.

2.2.2. Population

The review included individuals diagnosed with advanced/metastatic (inoperable) NSCLC who were responding to ongoing cancer therapy. There were no restrictions for sex, age, or treatment type (systemic therapy or radiotherapy).

2.2.3. Intervention

Studies that described the effects of adjunctive (i.e., in addition to standard of care) exercise programs on the HRQoL of patients with advanced/metastatic NSCLC were included in this review. Examples of exercise programs were aerobics, calisthenics, stretching, strength training, yoga/Pilates, sports, or similar physical activities. Such programs needed to be supervised, structured, and time-bound prescriptions that were performed in or out of the hospital. Studies that simply “encouraged” patients to exercise instead of supervising their exercise were excluded. In addition, trials that described exercise programs being used exclusively as an alternative treatment for NSCLC or exercise programs in combination with alternative or non-SOC therapies were excluded from this review.

2.2.4. Comparator/Control

Standard of care in the context of lung cancer therapy referred to the administration of anti-cancer treatment (systemic therapy or radiation therapy), regularly scheduled clinic visits, and palliative and supportive care. Studies that provided SOC treatment alone for lung cancer for the treatment control group were included in this review. Moreover, if the studies involved physical activity/exercise encouragement without supervision as a component of the control arm, these studies were included. However, studies that only compared different exercise programs without an SOC arm were excluded. Studies that use alternative or non-SOC treatments as a control group were likewise excluded.

2.2.5. Outcome

The outcome measure of interest in this review was change in HRQoL indicators in patients with advanced/metastatic NSCLC undergoing adjunctive exercise programs. The effect measured was mean difference (MD) and its standard error. When possible, the standard mean difference (SMD) was determined, as this measure controlled for differences in the HRQoL instruments used across studies [23]. Wherever applicable, qualitative HRQoL effects of adjunctive exercise on NSCLC patients were determined for SWiM synthesis. For example, this approach was used to describe effects on domains of HRQoL that were not consistently reflected or measured across tools but were considered essential by the reviewers in assessing well-being in lung cancer patients. The secondary outcome of interest was the safety of supervised exercise programs in this population of NSCLC patients.

2.3. Study Design

The review focused on RCTs, with the intervention being a supervised and structured exercise program in addition to SOC and the control group being SOC, as defined previously. Due to difficulties associated with blinding studies on exercise interventions, open-label RCTs were considered for inclusion. Cluster-randomized trials and crossover trials were excluded in this review. Observational studies such as cohort, comparative cross-sectional, and case-control studies, as well as smaller studies such as case reports, monographs, technical reports, and editorials/commentaries, were excluded.

Search Strategy

Electronic Searches

The literature search included a broad range of terms and keywords related to non-small cell lung cancer, exercise, and RCTs. The search terms were restricted to PICO domains of population/participants, intervention, and study design. The following databases were searched for potential studies to be included in this review:

MEDLINE

Embase

Cochrane Library

PeDRO

For this review, a decision to exclude gray literature was made a priori. This is based on concerns around the potential for bias in these types of non-peer-reviewed works.

Manual Searches

Reference lists of included studies and reference lists of relevant reviews were manually searched for additional studies that may be included. In addition, the following high-impact clinical oncology journals were manually searched for relevant studies: Journal of Clinical Oncology, Annals of Oncology, JCO Global Oncology, and the Journal of Thoracic Oncology.

2.4. Data Collection and Analysis

2.4.1. Selection of Studies

The study was conducted in three stages. (1) Citations from the earlier search were managed and screened for duplicates. (2) Titles and abstracts were independently screened by two reviewers against the above-defined inclusion and exclusion criteria to identify candidate studies. Studies identified for possible inclusion based on title and abstract had their full-text version retrieved for further scrutiny. Works that did not meet the inclusion criteria were rejected at this stage, and the reason for rejection was recorded. (3) The full-text articles of all remaining studies were obtained and independently assessed for inclusion by two reviewers. Disagreements between the reviewers were resolved by discussion in consultation with a third reviewer. A Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram (Figure 1) was generated to summarize the characteristics of the included and excluded studies at each stage of the selection process described.

2.4.2. Data Extraction and Management

Data from the included studies were extracted independently by at least two reviewers using a standardized, Excel-based (Microsoft, USA) data extraction sheet that was piloted on a sample of five studies and then modified, if necessary, before full data extraction began. Any disagreements between the reviewers at any stage were directed to an arbitrator. For the primary outcome, the mean difference and attributed standard error between the intervention and control groups for global HRQoL and the respective scores for each domain measured were extracted for all included studies. The additional variables listed are clinically relevant data elements that relate to possible sources of heterogeneity. These will potentially guide subsequent sub-group analyses.

2.4.3. Quality and Risk of Bias Assessment

Two authors independently assessed the risk of bias in each included study, and discrepancies were discussed with a third author for consensus. Assessment employed version 2 of the Cochrane risk-of-bias tool for randomized trials (RoB 2) [12,24]. The following domains were assessed: risk of bias, allocation sequence generation, allocation concealment, blinding of outcome assessors, incomplete outcome data, selective outcome reporting, and other potential sources of bias. Included studies were not assessed for blinding of participants and study personnel, as the participatory nature of the intervention makes blinding impossible. If one or more of the seven bias components were classified as high-risk, the trial was graded as “high risk of bias” overall. Conversely, if the risk of bias of all seven terms was defined as “low risk of bias”, the trial was defined as “low risk of bias” as a whole; otherwise, the trial was graded as having “unclear risk”. Previous systematic reviews were assessed for risk of bias using the Assessing the Methodological Tool for Systematic Review-2 (AMSTAR 2) checklist [25].

2.4.4. Measurement of Treatment Effect

For continuous outcomes such as HRQoL scores, standardized mean differences (SMDs) with 95% confidence intervals (CI) were calculated to synthesize the effect of the intervention, assuming that these were normally distributed. SMD values of 0.2–0.5 were considered small effect sizes, values of 0.5–0.8 were considered medium effect sizes, and values >0.8 were considered large effect sizes [23]. If the units of the same outcome measure were consistent for different studies (i.e., the same HRQoL measurement tool is used for all), the mean differences with their respective 95% CIs were calculated to synthesize the effects [23]. If studies did not report standard difference (SDs), SDs were calculated from standard error (SE), including their respective 95% Cis, using methods proposed by the Cochrane Handbook [12,13]. An important consideration from the outset was correctly determining the directionality of the reported HRQoL scores. A positive change connoted an improvement in HRQoL following the intervention, while a negative change suggested a deterioration of the same following the intervention.

2.4.5. Data Synthesis

A meta-analysis was performed for at least two studies that investigated the same outcome. Exploration of the methodological and clinical heterogeneity of the included studies by comparing participants’ characteristics (biological and treatment-related), interventions (type of exercise program), and HRQoL tool used using information reported in the characteristics of included studies table was performed. Heterogeneity among studies was estimated using the I² statistic and Cochran’s Q test [13]. I² statistics of <25%, 25–50%, and >50% were considered indicative of low, moderate, and high heterogeneity, respectively [12,26]. Assessment of heterogeneity was performed using a Galbraith plot. In the event of heterogeneity, sub-group analysis using a leave-one-out approach was performed to control for the sources of these. A SWiM synthesis and GRADE analysis (for safety results) were conducted through qualitative analysis in case of a lack of data or substantial heterogeneity levels (i.e., >85%) [14,26,27,28]. Meta-analysis was performed using Stata 18 [29].

3. Results

Figure 1 provides a summary of the search results in accordance with the 2020 PRISMA guidelines.

A total of 1168 studies were retrieved based on the predetermined search terms. Removing duplicate records, 1161 studies were screened, with 47 of these studies assessed for eligibility. A total of 29 records were excluded for the following reasons: wrong outcomes (n = 4), wrong methods (n = 4), wrong population (n = 21), wrong intervention (n = 1), ongoing trial (n = 2), and conference abstract (n = 2). After screening, 13 records were found to be eligible for systematic review (n = 10) and meta-analysis (n = 3).

3.1. Risk of Bias

The risks of bias of the included studies are presented in Figure 2.

Thirteen original studies were included in the risk-of-bias analysis (Figure 2). As stated in the data collection and analysis portion, included studies were not assessed for blinding of participants and study personnel, as the participatory nature of the intervention makes blinding impossible. All RCTs were classified as low risk for selecting patients to their allocated intervention, due to the inherent characteristic of randomization. Most studies analyzed exhibit a moderate to high risk of bias, primarily attributable to the inherent challenges associated with blinding both participants and staff in the context of exercise interventions. The physical and observable nature of exercise made it inherently difficult to conceal allocation of the interventions, as well as to blind both personnel and participants effectively. The majority of the studies included made efforts to increase the objectivity of their studies by blinding the data assessors, homogenizing baseline characteristics, and accounting for patients dropped/lost to follow-up.

3.2. Synthesis Without Meta-Analysis (SWiM)

A myriad of different HRQoL tools have been used across different studies that have investigated the benefit of exercise programs for patients with advanced NSCLC. These include the European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC-C30) [30,32,35,39,40,41], the Functional Assessment of Cancer Therapy–Fatigue (FACT-F) [31,33,42], the Functional Assessment of Cancer Therapy–General (FACT-G) questionnaire [34], the Functional Assessment of Cancer Therapy–Lung (FACT-L) [36,37,38] MD Anderson Symptom Inventory–Lung Cancer [36] and the SF-36 Health Survey [38,42].

Analyzing the exercise regimens and patient characteristics across the included studies, several common eligibility criteria emerged. Most studies required good functional capacity, typically defined as an ECOG performance status of 0–1 [30,38,42] or 0–2 [33,34,35,36,37,41] or a Karnofsky Performance Score of ≥80% [40]. Additionally, many trials included patients with a life expectancy of at least 4 months [33,34,35,36] and some required the ability to tolerate moderate- to vigorous-intensity exercise [36,39]. Certain conditions also served as exclusion criteria. These included the presence of bone [34,35,36,38,41,42] or brain metastases [34,35,38,41,42], unstable comorbidities [30,32,34,37,38,40,41,42], and the need for enteral or parenteral feeding [35], reflecting concerns about safety and feasibility in patients with more advanced disease burden with comorbid illnesses. The types of exercise interventions employed were predominantly aerobic training [30,32,33,34,35,36,37,38,39,40,41,42] and strength or resistance exercises [31,32,34,35,36,37,38,40,41], often delivered in combination to maximize physical and functional outcomes in this vulnerable population (Supplementary Material I).

The results from the studies varied. Four studies reported improvements in the quality of life of NSCLC patients who participated in standardized exercise regimens compared to control groups, based on the EORTC, FACT-L, and SF-36 tools [32,36,37,41].

Meanwhile, two studies presented mixed results across different quality-of-life (QoL) measures. Cheville et al. [31] described improved fatigue in patients belonging to the intervention group using the FACT-F scale, but no significant improvement in general QoL based on the FACT-G scale. Similarly, the study by Mendizabal-Gallastegui et al. [42] reported improvement in QoL based on the EORTC QLQ-C30 and SF-36, suggesting meaningful QoL improvement as measured by this tool, but not when using the SF-36 and FACT-F scales. Notably, in two of these studies, improvements in self-reported QoL lasted over a follow-up period of 6–12 months [36,42]. The remaining studies included in the SWiM analysis do not show any statistically significant improvement in patient-reported health related quality of life [30,33,34,35,38,39,40].

3.3. GRADE Analysis of Safety Results

A total of eleven studies were included in the GRADE analysis of safety studies. Seven studies [30,35,36,37,38,40,41] reported no serious adverse events, while two studies [31,33] reported that there were neither any serious nor any minor adverse effects in their study results. In the work by Dhilon et al. [33], there were four minor adverse events in the exercise group, who had musculoskeletal soreness likely related to exercise, but the reported events resolved without any treatment intervention. The study by Cheung et al. [41] only reported one episode of limb numbness during exercise. In the study by Mendizabal-Gallastegui et al. [42], there were also reported adverse effects, aside from symptoms of fatigue, dizziness, vasovagal symptoms, and arrhythmia.

A comprehensive report on safety endpoints was reported by Mikkelsen et al. [41]. These authors reported three grade 3 events: one non-exercise-related distal radius fracture, one osteoporotic spinal compression, and one incidence of spinal stenosis. The rest of the reported events were mild, consisting of one knee swelling episode (grade 2) and five episodes of bruises and dizziness/feeling sick (grade 1). Another study [39] reported a detailed tally of adverse effects during their study period. Four severe adverse events were reported, comprising three hospitalizations (chronic obstructive pulmonary disease, pneumonia, and hyperthyroidism) and one emergency room consultation due to a fall. Two minor events were also reported (ankle pain and bronchitis), but these were deemed unrelated to the study intervention.

The authors of this systematic review recommend a low certainty of evidence regarding the aspect of safety for exercise interventions in advanced NSCLC. This is due to the heterogeneity of the study population in the included studies and the lack of a standardized methodology in reporting safety/adverse effects in the protocol. Caution should be taken in interpreting the safety results, since advanced NSCLC patients often possess poor baseline functional capacity, which could confound the reporting of results.

3.4. Meta-Analysis

Three studies [30,39,40] were included in the analysis (Figure 3). The point estimate of the overall positive mean difference of 1.57 (95% confidence interval of 0 to 3.14) suggests that adjunctive exercise programs are associated with a trend towards improvement of self-reported health-related quality of life (EORTC-C30) among patients with advanced/metastatic non-small cell lung cancer. Notably, although the point estimate is greater than zero, the 95% confidence interval of the interval estimate touches the line of no effect, which is likely the result of the small sample size of the pooled population. Moreover, the SMD across these studies is 0.81, indicating a large effect size of the intervention. The I² statistic of the forest plot is 2.75%. This shows low heterogeneity among the studies. This is further supported by the results of the test of θi = θj (p = 0.53), which fails to reject the H₀ of homogeneity across studies that employed the same EORTC tool.

3.5. Publication Bias

Figure 4 represents the risk of publication bias for the studies that employed the EORTC-C30 tool. None of the studies are situated outside of the pseudo 95% confidence interval, and all are distributed closer to the center, suggesting a reduced risk of publication bias.

3.6. Assessment of Heterogeneity

The Galbraith plot for the studies that employed the EORTC-C30 tool is shown in Appendix A. For this group of studies, an overall positive treatment effect is seen, as all studies are above the line of no effect. Moreover, all studies aligned along the regression line, indicating a homogenous trend across their respective results.

4. Discussion

The results of the systematic review show a benefit of the addition of structured exercise programs to the care of patients with advanced NSCLC, particularly those whose disease is responding to their systemic treatment.

Qualitative synthesis by SWiM demonstrates the benefit of multiple structured programs across different HRQoL tools. Notably, there is a variety of interventions and self-reporting tools used, with differences in results depending on the tool employed. Broadly, the qualitative benefit seen across the breadth of studies reviewed argues that structured programs—be they weight training, martial arts, etc.—can be beneficial for patients with advanced NSCLC. Our findings are consistent with the findings of Sweegers et al. [43], who reported that, for patients diagnosed with solid and hematologic cancers, including lung cancer, supervised exercise interventions lead to statistically significant improvements in quality of life compared to unsupervised exercise prescriptions. How much benefit can be derived from these programs, however, is a more challenging question to answer.

To this end, some clarity can be derived from our meta-analysis of studies that used the EORTC-C30 tool, which was performed to address this point. Across these studies, it is estimated that advanced NSCLC patients who undergo adjunctive, structured exercise programs experience an improvement in HRQoL score of 1.57 (95% CI: 0 to 3.14). Others have attempted to translate EORTC-C30 to clinical outcomes. For example, Osoba et al. [44] defined a change of 5–10 points as a small improvement within the EORTC-C30 framework. However, a comprehensive analysis by Cocks et al. [45] using expert opinions, systematic review, and meta-analysis to establish subscale-specific thresholds for defining clinically meaningful changes in EORTC QLQ-C30 scores found that standardized cut-off levels for clinically meaningful change may not be appropriate across all subscales and contexts, which emphasizes the need for a tailored interpretation depending on the domain and direction of change.

Nadler et al. [46] reported that patients with baseline QoL and functional scores significantly below the mean benefit the most from exercise interventions. Their findings suggest that these improvements, though often not statistically significant due to limited power, may still be clinically meaningful because they directly impact activities of daily living and independence.

Another important issue is the duration of intervention necessary to yield a benefit in NSCLC patients. Studies with follow-up under six months do not show benefit for exercise with respect to HRQoL [30,33,34,38]. An exception is the study by Mikkelsen et al. [41] that observed improvements at 12 weeks. Notably, this study included a broader cancer population beyond lung cancer. In contrast, studies with longer follow-ups show more promising results. For instance, Edbrooke et al. [36] reported a statistically significant improvement in HRQoL at six months. These findings suggest that the duration of the intervention impacts patients’ self-reported HRQoL.

A systematic review and meta-analysis by Peddle-McIntyre et al. [9] (Supplementary Material II) found that disease-specific quality of life was improved in patients who underwent exercise training when compared to SOC, with a standardized mean difference of 0.51 (95% CI 0.08–0.93). These authors recommend interpreting these results with caution due to significant heterogeneity, small sample sizes, and high risk of bias. The SMD for our meta-analysis of studies that employed the EORTC-C30 tool is computed to be 0.81, indicating a large effect size of the intervention. However, given that the population in our pooled analysis is small, the true magnitude of the effect may not be reflected.

Owing to the heterogeneity of the study population in the included studies and the lack of a standardized methodology in reporting safety/adverse effects, interpreting safety results needs to be performed cautiously. Moreover, since advanced NSCLC patients have poor baseline functional capacity, this may confound assessment of patient symptoms relating to intervention-associated safety signals. Additionally, cancer treatment–related adverse effects and any comorbid illness patients may have can certainly adversely impact exercise adherence and may in fact drive discontinuation of the same.

Despite the variability in methodology, tools for measurements, and results, the GRADE analysis concluded that exercise is still a safe intervention for advanced/metastatic NSCLC patients. Synthesizing all the information derived from the studies, the authors generally recommend the following:

• Prescribing aerobic and resistance exercises in advanced/metastatic NSCLC patients is generally safe for patients with good functional capacity (PS 0–2), controlled comorbidities, a life expectancy of at least 4 months, and an absence of bone metastases that could precipitate skeletal-related events.
• Any exercise regimen must be closely supervised by professionals, with regular follow-up periods.
• An individualized approach to the regimen, duration, intensity, and goals of exercise should be implemented based on the patient’s capacity and functional status.

Additional considerations in deciding whether to implement adjunctive structured and supervised exercise programs in a particular clinical scenario need to take into account social and health system-based considerations that may impact treatment as well (Figure 5). For example, patients with comorbid illness that impacts mobility may be initiated on a low-impact, home-based supervised regimen if healthcare providers able to conduct house visits are available in that patient’s jurisdiction.

This review has several important limitations. First, although the authors of this review decided a priori to exclude blinding in the assessment of study quality, it remains a potential source of bias. Second, because study participants have advanced disease, the risk for attrition is high. As HRQoL is a self-reported outcome, patients lost to follow-up will often need to be censored. This leads to the third limitation: small sample sizes across studies hamper the ability to detect smaller differences in HRQoL between study groups. In relation to the GRADE analysis for safety, standardization of reporting adverse effects is essential for uniformity of interpretation of results. The authors recommend established criteria for reporting adverse events, particularly the Common Terminology Criteria for Adverse Events (CTCAE) [47]. Lastly, there was a lack of uniformity in follow-ups. Included studies had follow-up periods that ranged from weeks to several months. As the duration of the intervention may influence its efficacy, shorter studies may be primed to yield negative results by virtue of their duration.

5. Conclusions

Adjunctive and structured exercise programs improve HRQoL in patients with advanced NSCLC; however, the magnitude of benefit may differ according to the measurement tool used, as well as the duration of the intervention. While these interventions are generally safe, future studies will need to be designed to account for confounders intrinsic to the underlying disease of the study participants.