Effects and Mechanisms of Online Delivered Singing Training vs Usual Care in Patients With Persistent Symptoms 6-18 Months After Surgical Resection for Non-small Cell Lung Cancer (NSCLC) - a Multi-centre Cross-over RCT

Post-surgical rehabilitation should include combined aerobic training and strength training, alone or combined with respiratory muscle training. A Cochrane review (Cavalheri, et al., 2019) concluded that rehabilitation with exercise training improved physical capacity and physical fitness and suggested improved Health-Related QoL and dyspnoea. Additionally, a Core Outcomes Set (Edbrooke, et al., 2025) defined critical outcomes as: physical function, HRQoL, breathlessness, activities of daily living, emotional and mental well-being, pain.

In a recent Danish mapping study (Kaasgaard, et al., 2024), a large proportion experiences physiological-oriented symptoms: Fatigue (78%), hyper-vigilance (74%), dyspnoea (65%), and pain (47%); besides psychological-oriented symptoms: Worry about the future (40%), depression (35%), and anxiety (29%). In addition, 27%, experienced having vocal problems. This was confirmed in a subsequent Danish cross-sectional study, which specifically found impaired QoL (physical functioning) and symptoms (dyspnoea) 6-12 months after NSCLC surgery. This aligns with other recent Danish studies demonstrating persistent symptoms long-term after NSCLC surgery. In addition, a systematic review found poor mental health-related QoL, dyspnoea, and pain up to one year after surgery. Taken together, this implies signs of a neglected population with persistent symptoms, yet without a standard offer available to support them.

Singing has become increasingly popular in respiratory diseases, comprising elements of both physiological training and psychosocial aspects, therefore, likely to address the unmet needs in lung cancer. Previous studies have suggested physiological training effects and increased respiratory control, less dyspnoea, improved QoL, and reduced anxiety and depression. Moreover, singing is suggested to confer physiological training effects with work load equivalent to brisk walk pace. Furthermore, singing is perceived as a relevant and beneficial activity that builds joy, meaning, and social cohesion. However, studies have overall mostly been small and heterogenous, and with contradictory findings and a lack of knowledge about the underpinning mechanisms. Moreover, there are different approaches to singing, although a conventional choir approach is mainly adapted. The currently most well-documented, best-practice disease-specific approach is Singing for Lung Health (SLH), developed within a multidisciplinary healthcare setting in the UK since 2008. SLH aims to provide health-promoting benefits, uses exercises and songs as tools for purpose, and, further, includes movement and dancing while singing. In a recent multicentre-RCT in COPD - Sing-a-Lung 1.0 -, the study investigators demonstrated that PR with SLH improved functional capacity and QoL in a dose-response manner and was non-inferior to PR with usual care delivery of physical exercise. Furthermore, SLH was related to improved respiratory strength and dyspnoea control. This RCT was conducted within a real-life community-based setting, was the largest study so far, had sufficient power to detect change (n=270), and, thus, represents a proof-of-concept study.

Singing has not yet been investigated in lung cancer, but a qualitative study exploring singing for patients with breast or colorectal cancer suggests that singing may improve resilience, confidence, and mood, and builds social support. However, given that several epidemiological and common pathogenetic factors are shared between chronic obstructive pulmonary disease (COPD) and NSCLC patients, results from Sing-a-Lung 1.0 are likely to be transferrable specifically to people with lung cancer, and may address important needs of this overlooked group. However, the limited number of lung cancer patients in each municipality and the heterogeneity of current rehabilitation compromise the feasibility of an RCT equally to the one of Sing-a-Lung 1.0. Therefore, as it is expected that the patients will benefit from a singing intervention, singing will be compared with usual care using a cross-over RCT design. The singing intervention will be delivered online across all study sites, using video conferencing application, and delivered by three specialised singing teachers. Specifically, the study aims to investigate whether 10 weeks' online singing training (delivered as SLH) improves key physiological and psychosocial outcomes, and reduces symptom burden compared to usual care 6-18 months after curative intended NSCLC surgery, hypothesising that 1) SLH provides relevant and beneficial improvements in both physiological and psychological parameters (superior to usual care alone), and 2) SLH relieves symptom burden (superior to usual care alone). In an exploratory study of a sub-cohort, the study aims to investigate the underpinning mechanisms of singing compared to usual care, hypothesising that 3) SLH improves physical fitness, oxygenation, and inflammation. The overall project will run between Q3 2025 and Q4 2028. Intervention phase: Q1 2026 to Q3 2027. Recruitment and data collection at the following sites: Site 1) Dept. of Medicine, Zealand University Hospital, Roksilde and Næstved, Næstved Hospital, Ringstedgade 61, DK-4700 Naestved, Denmark; Site 2) Dept. of Respiratory Medicine, Bispebjerg Hospital, Bispebjerg Bakke 23, DK-2400, København NV, Denmark; Site 3) Dept. of Respiratory Medicine, Lillebaelt Hospital, Vejle, Kabbeltoft 25, DK-7100 Vejle, Denmark; Site 4) Dept. of Respiratory Medicine, Odense University Hospital, J. B. Winsløws Vej 4, DK-5000 Odense, Denmark. Specifically, recruitment at Site 2 (or from Site 1 if living close to Copenhagen): Data collection (both for the main study and exploratory sub-study) at site 5) August Krogh Sect. for Molecular and Human Physiology, Dept. of Nutrition, Exercise University of Copenhagen, and Sports, Faculty of Health and Medical Sciences, Universitetsparken 3, DK-2100 København Ø, Denmark. Participants will be recruited using the following methods: 1) Retrospectively: Participants from the recent cross-sectional study, conducted at Site 1 and 2 (Kaasgaard, et al., in peer-review) who indicated to be interested in being part of the study; 2) Retrospectively: Patient lists from the Danish Lung Cancer Registry (DLCR); 3) Prospectively: Ongoing identification in follow-up consultation. Included study participants will be block randomised to either: Group 1: 10 weeks' online SLH, or Group 2: 10 weeks' usual care followed by 10 weeks' online SLH. Study participants will be assessed at the following time points: Group 1: 1) T0: Baseline (at randomisation); 2) T1: Follow-up 11-14 weeks later). Group 2: 1) T0: Baseline (at randomisation); 2) T1: Follow-up 11-14 weeks later); 3) T2: Follow-up: Withing three weeks after the SLH intervention (appr. 11-14 weeks after T1). Primary study outcome: Change in physical capacity, assessed with the Six Minute Walking Test (6MWT). An a priori statistical power analysis calculation showed that the needed sample size is n=78 study participants and including an expected drop-out-rate of 20%, an overall total sample of n=100 study participants is needed. In the power analysis, the investigators considered data from previous studies on the primary study outcome, 6MWT distance, with a minimal important difference of ≥ 30 m, with the final calculation based on findings from our previous RCT on singing for COPD with a pre-post difference of mean 13.1±36.3 m after 10 weeks of singing in PR for COPD. Statistical power analysis calculations (primary study outcome: 6MWD): Estimated sample size for two-sample comparison of means: Test H0: n1 = n2, where n1 is the mean in population 1 and n2 is the mean in population 2. Assumptions: α (two-tailed) 0.05, β 0.2; Effect size Group 1: 5 m; Group 2: 15 m SD 15. Estimated required sample sizes (based on power: 80 %): n1 = 26; n2 = 52.