Patient-reported outcomes after oesophagectomy in the multicentre LASER study

Abstract Background Data on the long-term symptom burden in patients surviving oesophageal cancer surgery are scarce. The aim of this study was to identify the most prevalent symptoms and their interactions with health-related quality of life. Methods This was a cross-sectional cohort study of patients who underwent oesophageal cancer surgery in 20 European centres between 2010 and 2016. Patients had to be disease-free for at least 1 year. They were asked to complete a 28-symptom questionnaire at a single time point, at least 1 year after surgery. Principal component analysis was used to assess for clustering and association of symptoms. Risk factors associated with the development of severe symptoms were identified by multivariable logistic regression models. Results Of 1081 invited patients, 876 (81.0 per cent) responded. Symptoms in the preceding 6 months associated with previous surgery were experienced by 586 patients (66.9 per cent). The most common severe symptoms included reduced energy or activity tolerance (30.7 per cent), feeling of early fullness after eating (30.0 per cent), tiredness (28.7 per cent), and heartburn/acid or bile regurgitation (19.6 per cent). Clustering analysis showed that symptoms clustered into six domains: lethargy, musculoskeletal pain, dumping, lower gastrointestinal symptoms, regurgitation/reflux, and swallowing/conduit problems; the latter two were the most closely associated. Surgical approach, neoadjuvant therapy, patient age, and sex were factors associated with severe symptoms. Conclusion A long-term symptom burden is common after oesophageal cancer surgery.


Introduction
Surgical resection is still the mainstay of curative treatment for oesophageal cancer. Centralization of oesophageal cancer surgery 1 , implementation of enhanced recovery protocols 2 , and adoption of minimally invasive techniques 3 have reduced morbidity and treatment-related mortality. Multimodal treatment has led to further improved survival rates 4,5 . Incremental improvements in long-term survival have highlighted the increasing importance of quantifying long-term health-related quality of life (HRQoL) of oesophageal cancer survivors 6 using validated disease-specific tools.
Validated questionnaires for the assessment of HRQoL during treatment of oesophageal cancer have been produced by the European Organization for Research and Treatment of Cancer (EORTC) 7,8 . These have been used to assess the impact of multimodal treatment 9 , and the effect of variation in surgical technique 10 and complication burden, on HRQoL. These questionnaires have, however, not been validated for assessment of HRQoL in the disease-free survivorship state. Longterm persistent adverse effects of oesophageal cancer surgery on disease-free HRQoL have been described 11 , but a robust assessment of HRQoL in the context of contemporary treatment strategies is lacking. A recent population-based cohort study 12 from England suggested that over 40 per cent of patients visit their primary-care physician after oesophagectomy with longterm symptoms as a result of the intervention. The presence of two or more symptoms was associated with an increased risk of new-onset depression or anxiety.
The multicentre LASER (LAsting Symptoms after Esophageal Resection) study was developed to evaluate the burden of longterm symptoms in survivors more than 1 year after initial treatment 13 . The aim was to identify the most prevalent symptoms affecting HRQoL, to establish whether symptoms cluster into domains, and to identify potential patient and treatment factors associated with development of long-term symptoms.

Methods
This was a multicentre study of consecutive patients who underwent oesophageal cancer surgery at 20 European centres between 1 January 2010 and 30 June 2016. Patients who had oesophagectomy with curative intent for either oesophageal or junctional (Siewert I and II) cancer and who were disease-free for at least 1 year after completion of initial treatment, either surgery or adjuvant therapy, were eligible. They were asked to complete the validated LASER questionnaire on a single occasion.
Patients who underwent salvage oesophagectomy after initial endoscopic or oncological treatment with persistent or recurrent oesophageal cancer were also included. Exclusion criteria were: endoscopic therapy or definitive chemoradiotherapy as sole therapy for oesophageal cancer, known cancer recurrence, ongoing surgical complications (such as tracheostomy), and non-oral nutrition (for example, jejunostomy catheter).
Local institutional review board and ethical approvals were obtained by each participating centre.

Intervention
The LASER questionnaire was used to evaluate the presence and severity of 28 symptoms in the 6 months preceding administration of the survey 13 . The development and collection of the LASER data set have been described previously 13 . In summary, the LASER questionnaire was developed through a Delphi consensus process of clinician experts in addition to input from international patient workshops. This questionnaire was used to collect data from all patients in the LASER study (Appendix S1) via either an electronic or paper format. These results were subsequently uploaded to a specifically created website (https:// www.laserstudy.org) for secure data storage and analysis.
Each symptom from the LASER questionnaire was graded according to its impact on quality of life, measured by means of validated EORTC QLQ-C30 6 and QLQ-OG25 7 questionnaires, and the frequency of the symptom was recorded, using a composite score from 0 to 5 (Table S1). A symptom grade of 3 or higher was indicative of either a frequent symptom or a symptom with a substantial impact on quality of life; these were therefore classified as severe symptoms.

Data collection
In addition to collection of data on quality of life, clinicians at participating centres were asked to provide additional clinical, pathological, and demographic metadata (full list provided in Appendix S2).

Statistical analysis
Basic frequency analysis was undertaken to identify the prevalence of severe symptoms. Principal component analysis was used to establish whether symptoms clustered into specific domains and whether these domains were associated. Multivariable logistic regression models were constructed to identify patient and clinical risk factors associated with the development of specific severe long-term symptoms (dependent variable). The independent variables included in these models were patient age (continuous), sex (female or male), time since surgery (continuous), neoadjuvant therapy (yes or no), surgical approach (open, hybrid minimally invasive oesophagectomy or totally minimally invasive oesophagectomy), access to the chest (transhiatal, Ivor Lewis, three-stage, or left thoracoabdominal), and postoperative 30-day complications (yes or no). Bonferroni correction was used to adjust for the effects of interaction between variables and multiple testing effects.
All statistical analyses were conducted by an experienced biostatistician using SAS V R version 9.4 (SAS institute, Cary, North Carolina, USA). P < 0.050 was considered statistically significant.

Results
A total of 876 of 1081 invited patients (81.0 per cent) responded to the questionnaire.
Of these, 586 (66.9 per cent) stated they had experienced symptoms in the past 6 months associated with the oesophagectomy. They filled out the questionnaire a median of 4.3 (i.q.r. 3.1-5.8) years after surgery. Patient and clinical characteristics are summarized in Table 1.

Clustering, frequency and severity of symptoms
The 28 symptoms from the LASER questionnaire appeared to cluster into six key domains: lethargy, musculoskeletal pain, dumping, lower gastrointestinal symptoms, regurgitation/reflux, and swallowing/conduit problems (Tables 2 and 3) (Fig. S1). Regurgitation/reflux symptoms and swallowing/conduit problems were closely associated, as were dumping and lower gastrointestinal symptoms (Fig. 1).
The symptoms most commonly classified as severe were reduced energy or activity tolerance (30.7 per cent), early satiety (30.0 per cent), tiredness (28.7 per cent), heartburn/acid or bile regurgitation (19.6 per cent), abdominal pain (14.0 per cent), bloating or cramping after eating (13.7 per cent), persistent cough (13.4 per cent), and difficulty getting food down (12.2 per cent) (Fig. 2). An early feeling of fullness after eating was noted to be the most common symptom in the first 2 years after surgery (34.4 per cent). Reduced energy or activity tolerance was the most common symptom 3-4 years after surgery (31.1 per cent) as well as beyond 5 years (32.9 per cent) (Appendix S3). Fig. 3 also highlights the dynamic nature of symptom progression over time, focusing on the five most prevalent symptoms.

Factors associated with symptoms
Results of multivariable analysis addressing the impact of demographics and clinicopathological characteristics on severe symptoms are available in Appendix S4. Results for the most prevalent four severe symptoms with a prevalence of least 15 per cent are shown in Table 4.
A totally minimally invasive approach to oesophagectomy was associated with a lower prevalence of reduced energy or activity tolerance, increased chest pain, loose bowel motions or diarrhoea after eating, dental problems, and reduced vomiting.
The presence of 30-day postoperative complications was associated with increased prevalence of severe chest pain.
Neoadjuvant therapy was associated with an increased prevalence of tiredness.
Increasing patient age was associated with increased early feeling of fullness after eating, tiredness, and heartburn/acid or bile regurgitation, increased prevalence of abdominal pain, pain from scars on the chest, difficulty in getting food down, regurgitation of food, nausea, vomiting, heart palpitations after eating, sweating after eating, bloating or cramping after eating, loose bowel motions or diarrhoea after eating, waking up during the night because of choking sensation, and diarrhoea unrelated to eating.
Female sex was associated with a reduced prevalence of heartburn/acid or bile regurgitation, reduced prevalence of chest pain, nausea, vomiting, heart palpitations after eating, sweating after eating, bloating or cramping after eating, loose bowel motions or diarrhoea after eating, waking up during the night because of choking sensation, and diarrhoea unrelated to eating.

Discussion
Two-thirds of patients experienced symptoms related to oesophagectomy more than 1 year after surgery. The most common severe symptoms were reduced energy or activity tolerance, early feeling of fullness after eating, tiredness, and heartburn/acid or bile regurgitation. This evidence about symptom frequency complements previous findings regarding symptoms that have the greatest impact on quality of life 13 . Symptoms tended to cluster into six domains. Choice of surgical approach, sex, and patient age were all factors in determining the symptom burden following treatment.
The extent of the hidden survivorship burden that patients in this cohort faced, even in their supposedly disease-free state, deserves comment. Reduced energy/activity tolerance and tiredness are well established associative factors in the development of major depressive episodes, which in turn have been linked with poor survival outcomes 14 . There is a growing body of survivorship-specific evidence among a range of cancers to suggest that chronic unattended symptoms have a sizeable impact on the HRQoL of cancer survivors 15 . This challenges healthcare professionals to address symptoms that they may perceive as benign during clinical consultations in order to prevent downstream effects that can manifest as poor patient outcomes.
This study may aid healthcare professionals in conceptualizing how these symptoms fit into symptom phenotypes. Knowledge of how symptoms cluster into domains, particularly when combined with personalized patient and clinical risk factor susceptibilities, will help to accurately stratify and predict survivorship care needs at an individual level 16 . Knowledge of modifiable (such as surgical approach) and non-modifiable (for   Many specialist centres currently lack dedicated follow-up survivorship programmes. As such, issues including health 17 and psychological well-being promotion, particularly in at-risk populations, are not managed systematically. Further effort should be made to counsel patients appropriately in the pretreatment phase. Future steps should include mechanistic investigation of symptom clusters in order to develop tailored interventions that may improve the long-term burden of survivorship. Specific analyses should include investigation of changes in gut hormone profiles 18  investigations alongside physical activity assessments will be invaluable in understanding how the effects of oesophagectomy may be linked to reduced energy symptom profiles. Strengths of the present study include the response rate, the multicentre international design, use of a refined and validated questionnaire, and a set-up allowing patients to complete the questionnaire outside the clinical setting.
This study had several limitations. The cross-sectional design, centred around a single episode of data collection after surgery, precluded comparison with preoperative baseline HRQoL and may have caused bias as patients did not fill out the form at a similar point in time after treatment. It is impossible to study longitudinal changes in symptoms adequately within the study design. Patients with recurrences were not included. BMI was recorded poorly in the data set. Input into the study from predominantly high-volume Western European centres may mean that the findings have limited generalizability to patients undergoing interventions in lowervolume centres or in other parts of the world.