The association between changes in symptoms or quality of life and overall survival in outpatients with advanced cancer

Introduction

The ability of clinicians to predict survival time is vital for patients with advanced cancer and their families. Indeed, accurate survival prediction is important in clinical decision making, especially regarding the provision of palliative systemic therapies, other palliative interventions, timing of transitions to hospice care, and overall patient and family preparation for death (1). However, survival prediction is a very challenging task in advanced cancer scenarios (1). Several prognostic tools have been developed to help clinicians in the prediction of survival time, including: Glasgow prognostic score (GPS) (2) and modified Glasgow Prognostic Score (mGPS) (3). However, available prognostication schemas are of limited value in many common clinical scenarios.

Information about symptom burden and trajectories may provide additional data to inform prognostic estimates, but these data are not included in most prognostication schemas. Symptom assessment is an integral part of daily palliative care, and influences patients’ satisfaction. Generally, symptoms reflect disease progression in patients with advanced cancer (4). In addition, patient-reported symptoms and quality of life (QOL) accurately reflect survival-related patient functioning compared with traditional prognostic indicators (5). Symptoms such as dyspnea and anorexia are known to worsen as diseases progress (6,7) and cross-sectional symptom assessments are included in a few available prognostic tools (8,9). These symptoms are assessed in many QOL assessment tools as well (10). Current prognostic tools include usually symptoms, vital signs, laboratory data, and factors related to diseases. However, changes of symptoms or QOL are rarely included in established prognostic systems.

There is a paucity of studies examining the impact of symptom changes on survival prediction. While changes in symptoms can be approached with reliability using established outcome measures, there are only a few studies that embody them into prediction of survival time. Previous studies focused on patients admitted to palliative care units (11,12), and another study focused on patients with advanced cancer who have approximately 1-year median survival time (13). We hypothesized that longitudinal changes in symptoms and QOL could add useful information to prognostication schemas. Thus, the aim of this study was to explore the relationship of changes in symptoms and QOL on survival time in outpatients with advanced cancer. We present the following article in accordance with the STROBE reporting checklist (available at https://apm.amegroups.com/article/view/10.21037/apm-22-33/rc).

Methods

Participants

This is a preplanned sub-analysis of a prospective cohort study. The aim of the parent study was to develop a prognostic model to facilitate palliative care referral before the last 3 months of life in outpatients with advanced cancer at medical oncology clinics (14). Study subjects were derived from the subgroup of participants in a single-center parent study who completed longitudinal assessments of symptoms and QOL. Thus, participants for this study were selected conveniently. This study was conducted at a comprehensive cancer center, Seoul National University Bundang Hospital. Eligible patients with advanced cancer who were treated by the medical oncologists at the center were enrolled during the study period, between March 2016 and January 2019. The inclusion criteria were as follows: (I) a diagnosis of advanced cancer, (II) a clinician’s prediction of survival of 1 year or less, and (III) adult age (≥18 years). The definition of “advanced cancer” required recurrent or metastatic disease or progressive locally advanced disease without option for curative treatment. The exclusion criteria were as follows: (I) blood cancers, (II) clinicians’ prediction of survival less than 4 weeks, and (III) incompetency of the patient to communicate. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). We obtained written informed consent from each patient before enrollment. This study was approved by the Institutional Review Board (IRB) of Seoul National University Bundang Hospital (IRB number: B-1601/332-302).

Data collection

The instruments we evaluated were a part of the parent prospective cohort study’s protocol. We assessed patient performance status using the Eastern Cooperative Oncology Group (ECOG) performance status and Karnofsky performance status (KPS). Demographic data and clinical information were accessed from the electronic medical record, including age, sex, primary cancer site, and anticancer treatment. The patients’ symptoms were evaluated by the Korean version of the Edmonton Symptom Assessment System (K-ESAS) which has been validated in Korean patients with cancer (15). We assessed the patient’s QOL by using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) which has been validated in Korea (16). We assessed KPS, K-ESAS and EORTC QLQ-C30 at enrollment and follow-up at an approximately 3-month interval. The definition of survival time here was mortality either in or outside of hospital, calculated by subtracting the enrollment date from the death date.

We classified each patient into one of two groups, based on changes in KPS/ESAS/EORTC QLQ-C30 score from baseline (at enrollment). The KPS was rated as 11 levels that range from normal activity [100] to dead [0], by ten point increments. The ESAS ranged from 0 to 10, with a higher score indicating severe symptoms for each item, in one point increments. The EORTC QLQ-C30 scores were transformed into 0–100 points scale according to the scoring manual (17). Grouping was based on each change of three scales separately. According to previous studies (12,18), these groups were defined as follows. The “improved + stable” group included those who had at least a one level of improvement in the response scale or who did not show any change for KPS/ESAS/EORTC QLQ-C30 score at follow-up time. We defined the “Worsened” group as those who had at least a one level of worsening in these scales at follow up. Regarding EORTC QLQ-C30 score, we defined a change of 10 points as a one level change based on a previous study (19).

Statistical analysis

All analyses were performed using JMP version 16 for Windows (SAS, Cary, NC, USA). First, descriptive analyses were performed to show participants’ characteristics. Second, each patient was classified into the two pre-defined groups by changes in KPS/ESAS/EORTC QLQ-C30 scores from enrollment to follow-up time, categorized as: ‘improved + stable’ and ‘worsened’. We compared means of KPS/ESAS/EORTC QLQ-C30 score by paired t-tests from the enrollment to the follow-up time, within each group. Finally, the survival time of each group for each KPS/ESAS/EORTC QLQ-C30 score was assessed via the Kaplan-Meier method, and survival time was compared in the univariate analysis using a log-rank test. Statistical significance was set as a P value of <0.05.

Results

Study participants

Two hundred patients completed the baseline assessment in the parent study. Sixty patients (30% of the parent participants) completed follow-up assessment, due to functional decline and/or simple refusal. Therefore, a total of 60 patients were analyzed. The demographic information and baseline characteristics of patients are shown in Table 1. The mean patient age was 64.6 years [standard deviation (SD), 10.9], and 41 patients (68.3%) were male. The most common cancer sites were: lung (30.0%), colorectal/rectal (16.7%), stomach (11.7%) and breast (11.7%). Fifty patients (83.3%) were receiving palliative chemotherapy at enrollment and follow-up time. In total, 45 patients (75.0%) had an ECOG performance status of 0 or 1 at baseline. The median follow-up period after initial assessment was 92 days and the median overall survival time was 346 days (range, 118–1,103 days).

The changes in symptom scores

The mean KPS/ESAS/EORTC QLQ-C30 scores at enrollment and at follow-up are described in Table 2. The mean symptom scores at enrollment were similar to those at the follow-up assessment. Pain (ESAS: P=0.02), and the nausea/vomiting domain (EORTC QLQ-C30; P<0.01) were significant by paired t-tests.

Survival by changes in symptoms

The results of log-rank tests to compare survival times across the two groups appear in Table 3. Several ESAS items were associated with statistically significantly shorter overall survival in the “worsened” groups, including depression (P<0.01), and sleep disturbance (P<0.01). In addition, one QLQ-C30 item was significantly associated with reduced overall survival: dyspnea (P=0.03). The worsened groups showed poorer survival {median survival time 237 days [95% confidence interval (CI), 154–378 days] for depression in ESAS, 237 days (95% CI, 142–309 days) for sleep disturbance in ESAS, 252 days (95% CI, 176–350 days) for dyspnea in EORTC QLQ-C30} than that of improved + stable [median survival time 446 days (95% CI, 255–654 days) for depression in ESAS, 537 days (95% CI, 323–754 days) for sleep disturbance in ESAS, 432 days (95% CI, 284–654 days) for dyspnea in EORTC QLQ-C30].

Discussion

The aim of our study was to explore the influence of changes in symptoms and QOL on survival time in patients with advanced cancer. We found worsening of depression and sleep disturbance evaluated by ESAS, and dyspnea assessed by EORTC QLQ-C30, were statistically significantly associated with shorter survival time. Our findings are unique in terms of longitudinal assessments of symptoms for prognostic purposes, with most prognostic models either not incorporating patient-reported outcome measures at all, or only doing so in a cross-sectional manner. Our findings suggest that longitudinal data collection may add useful information to enhance the predictive accuracy of prognostic models, by allowing the incorporation of symptom and QOL changes thereto.

To contextualize our findings, one should recognize that previous meta-analyses showed that depression is related to mortality (20). In addition, several studies have found that supportive psychotherapeutic interventions, which aimed to decrease depression, have improved overall survival in patients with cancer (21,22). It has also been reported that early palliative care has resulted in longer survival time in patients with metastatic lung cancer, with improved QOL and depressive symptoms (23). Similar benefits have now been seen across multiple advanced solid tumor populations in various randomized clinical trials of integrated palliative care interventions (24). Our finding that worsening depression was related to shorter survival time further supports the use of integrated palliative care interventions to improve mood, as a way to improve outcomes among outpatients with advanced cancer, but also underscores the importance of more objectively assessing and longitudinally tracking depressive symptoms.

Similarly, sleep disturbance, which we found to be associated with overall survival in our cohort, can be an important indicator of depression in patients with serious illness. Insomnia in patients with cancer is known to be correlated with depressed mood (25). More than 90% of patients with depressed mood have been reported to experience abnormal sleep patterns (26). Negative thoughts such as uncertainty about treatment, fear of death, concerns about disease progression and diminished QOL can contribute to sleep disturbance as well (27). One might also expect increasing sleep disturbance to reflect significant symptom burden, such as from dyspnea and pain. Enrolled patients had median survival time of less than 1 year, therefore, they were more likely to have these negative cognitions and higher symptom burden. Additionally, sleep disturbance is common among individuals with anxiety disorders (28). Depression and anxiety are linked to maladaptive lifestyle habits, plus low compliance to medical treatments, which could lead to poor prognosis (29,30). Since 50% of our patients underwent chemotherapy, we assumed worsened sleep disturbance might be linked to difficulties adhering to chemotherapy.

We also found that change in dyspnea was significantly associated with survival in our population, and indeed many prognostic tools include dyspnea as a component of their schemas, but only in a cross-sectional manner (8,9). These tools are generally used for prediction of survival weeks before death, yet other data suggest that shortness of breath severity may increase in the last months before death (6). This lends support to our findings, which suggest that worsening of dyspnea farther upstream from death is still meaningfully linked with duration of survival. In general, symptoms are proportionate to tumor growth in patients with advanced cancer (4). Thus, one can expect that the worsening of symptoms would increase in the end-of-life and to otherwise be associated with shorter survival (31). Our findings point to a need to objectively measure and longitudinally track dyspnea as an important prognostically-relevant symptom, in addition to being one that can be improved with more targeted attention from palliative care specialists.

To the best of our knowledge, this is the first study to investigate the influence of changes in symptoms and QOL on survival time in ambulatory patients with advanced cancer who have more than three months of survival time. We thus propose a novel approach that symptom and QOL assessment be incorporated into prognostication schemas in advanced cancer, including longitudinal assessments. This approach has several favorable features, including that it is patient-reported and using validated scales, and should be easy to explain to patients and their families since symptoms are so often linked to the trajectory of diseases including responsiveness to treatment. Evidence from several patient-reported outcomes studies suggest that symptoms are generally under-recognized by oncologists, in both incidence and severity (32,33). Therefore, more routine collection of symptoms and QOL data could be beneficial to enhancing care and possibly even lead to prolonged survival if oncologists are able to pay more attention to worsening symptoms that are prognostically important, such as depression or sleep disturbance.

We recognize several limitations of this study. First, it was a single center study conducted at a comprehensive cancer center in a heterogeneous group of patients in South Korea, and all participating clinicians were medical oncologists trained in palliative care. In fact, majority of participants (30.0%) had been diagnosed with lung cancer and fifty patients (83.3%) were receiving chemotherapy at the time of enrollment and follow-up time. Patients with lung cancer may have shortened survival due to severe symptoms. Meanwhile chemotherapy could prolong survival time. We recognize this heterogeneity might influence our results. Therefore, additional studies are required in different settings worldwide. Second, the analysis cohort was limited to approximately one third of the larger study cohort, due to absent follow-up data for many participants. Our study subjects might have been feeling better overall and/or had better function compared to non-responders, to have replied to all questionnaires twice. Besides, this study was conducted in small sample. It is needed to investigate changes of symptoms and QOL for larger population in near future. Third, some important factors including laboratory data were excluded in the final analysis, and these may be meaningfully linked with survival time. Fourth, we regarded any changes of symptom or QOL scores similarly. However, a change in symptom or QOL score from 0 to 1 may be meaningfully different for a patient than a change from 7 to 8, and perhaps portend a different trajectory.

In conclusion, this study showed that worsening of depression, sleep disturbance and dyspnea predicted survival in our cohort of patients with advanced cancer. Thus, longitudinal assessment of symptoms appears to be useful in prognostication. This simple and objective approach can be helpful to differentiate patients with poorer prognosis (worsened symptoms and QOL) from those with better prognosis (improved or stable symptoms and QOL), using an approximately 3-month interval of assessment. Further studies are warranted to implement our findings to a variety of symptoms with more repeated assessments, and to other populations and settings.

Acknowledgments

We thank the participating patients and clinical research nurses, Jin Suk Kim and Esther Jeon, for supporting this study.

Funding: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science, Informatics, Communication and Technology; No. 2015R1C1A2A01053357).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://apm.amegroups.com/article/view/10.21037/apm-22-33/rc

Data Sharing Statement: Available at https://apm.amegroups.com/article/view/10.21037/apm-22-33/dss

Peer Review File: Available at https://apm.amegroups.com/article/view/10.21037/apm-22-33/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://apm.amegroups.com/article/view/10.21037/apm-22-33/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Written informed consent was obtained from each patient before enrollment. The protocol was approved by the Institutional Review Board (IRB) of Seoul National University Bundang Hospital (IRB number: B-1601/332-302).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Hui D, Paiva CE, Del Fabbro EG, et al. Prognostication in advanced cancer: update and directions for future research. Support Care Cancer 2019;27:1973-84. [Crossref] [PubMed]
2. Forrest LM, McMillan DC, McArdle CS, et al. Evaluation of cumulative prognostic scores based on the systemic inflammatory response in patients with inoperable non-small-cell lung cancer. Br J Cancer 2003;89:1028-30. [Crossref] [PubMed]
3. Proctor MJ, Morrison DS, Talwar D, et al. An inflammation-based prognostic score (mGPS) predicts cancer survival independent of tumour site: a Glasgow Inflammation Outcome Study. Br J Cancer 2011;104:726-34. [Crossref] [PubMed]
4. Echteld MA, Deliens L, Van Der Wal G, et al. Palliative care units in The Netherlands: changes in patients' functional status and symptoms. J Pain Symptom Manage 2004;28:233-43. [Crossref] [PubMed]
5. Quinten C, Coens C, Mauer M, et al. Baseline quality of life as a prognostic indicator of survival: a meta-analysis of individual patient data from EORTC clinical trials. Lancet Oncol 2009;10:865-71. [Crossref] [PubMed]
6. Seow H, Barbera L, Sutradhar R, et al. Trajectory of performance status and symptom scores for patients with cancer during the last six months of life. J Clin Oncol 2011;29:1151-8. [Crossref] [PubMed]
7. Hui D, dos Santos R, Chisholm GB, et al. Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to Acute Palliative Care Units. J Pain Symptom Manage 2015;50:488-94. [Crossref] [PubMed]
8. Maltoni M, Nanni O, Pirovano M, et al. Successful validation of the palliative prognostic score in terminally ill cancer patients. Italian Multicenter Study Group on Palliative Care. J Pain Symptom Manage 1999;17:240-7. [Crossref] [PubMed]
9. Morita T, Tsunoda J, Inoue S, et al. The Palliative Prognostic Index: a scoring system for survival prediction of terminally ill cancer patients. Support Care Cancer 1999;7:128-33. [Crossref] [PubMed]
10. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365-76. [Crossref] [PubMed]
11. Mori M, Parsons HA, De la Cruz M, et al. Changes in symptoms and inpatient mortality: a study in advanced cancer patients admitted to an acute palliative care unit in a comprehensive cancer center. J Palliat Med 2011;14:1034-41. [Crossref] [PubMed]
12. Suh SY, Won SH, Hiratsuka Y, et al. Assessment of Changes in Symptoms Is Feasible and Prognostic in the Last Weeks of Life: An International Multicenter Cohort Study. J Palliat Med 2022;25:388-95. [Crossref] [PubMed]
13. Suh SY, Leblanc TW, Shelby RA, et al. Longitudinal patient-reported performance status assessment in the cancer clinic is feasible and prognostic. J Oncol Pract 2011;7:374-81. [Crossref] [PubMed]
14. Kim YJ, Hiratsuka Y, Suh SY, et al. A Prognostic Model to Facilitate Palliative Care Referral in Oncology Outpatients. Cancer Res Treat 2021; [Epub ahead of print]. [PubMed]
15. Kwon JH, Nam SH, Koh S, et al. Validation of the Edmonton Symptom Assessment System in Korean patients with cancer. J Pain Symptom Manage 2013;46:947-56. [Crossref] [PubMed]
16. Yun YH, Park YS, Lee ES, et al. Validation of the Korean version of the EORTC QLQ-C30. Qual Life Res 2004;13:863-8. [Crossref] [PubMed]
17. Fayers PM, Aaronson NK, Bjordal K, et al. The EORTC QLQ-C30 Scoring Manual (3rd Edition). European Organisation for Research and Treatment of Cancer 2001.
18. Hui D, Shamieh O, Paiva CE, et al. Minimal clinically important differences in the Edmonton Symptom Assessment Scale in cancer patients: A prospective, multicenter study. Cancer 2015;121:3027-35. [Crossref] [PubMed]
19. Osoba D, Rodrigues G, Myles J, et al. Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol 1998;16:139-44. [Crossref] [PubMed]
20. Pinquart M, Duberstein PR. Depression and cancer mortality: a meta-analysis. Psychol Med 2010;40:1797-810. [Crossref] [PubMed]
21. Fu WW, Popovic M, Agarwal A, et al. The impact of psychosocial intervention on survival in cancer: a meta-analysis. Ann Palliat Med 2016;5:93-106. [Crossref] [PubMed]
22. Mirosevic S, Jo B, Kraemer HC, et al. "Not just another meta-analysis": Sources of heterogeneity in psychosocial treatment effect on cancer survival. Cancer Med 2019;8:363-73. [Crossref] [PubMed]
23. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med 2010;363:733-42. [Crossref] [PubMed]
24. Fulton JJ, LeBlanc TW, Cutson TM, et al. Integrated outpatient palliative care for patients with advanced cancer: A systematic review and meta-analysis. Palliat Med 2019;33:123-34. [Crossref] [PubMed]
25. Davidson JR, MacLean AW, Brundage MD, et al. Sleep disturbance in cancer patients. Soc Sci Med 2002;54:1309-21. [Crossref] [PubMed]
26. Geoffroy PA, Hoertel N, Etain B, et al. Insomnia and hypersomnia in major depressive episode: Prevalence, sociodemographic characteristics and psychiatric comorbidity in a population-based study. J Affect Disord 2018;226:132-41. [Crossref] [PubMed]
27. Bailey DE Jr, Wallace M, Mishel MH. Watching, waiting and uncertainty in prostate cancer. J Clin Nurs 2007;16:734-41. [Crossref] [PubMed]
28. Babson KA, Feldner MT. Temporal relations between sleep problems and both traumatic event exposure and PTSD: a critical review of the empirical literature. J Anxiety Disord 2010;24:1-15. [Crossref] [PubMed]
29. Strine TW, Mokdad AH, Dube SR, et al. The association of depression and anxiety with obesity and unhealthy behaviors among community-dwelling US adults. Gen Hosp Psychiatry 2008;30:127-37. [Crossref] [PubMed]
30. Kisely S, Crowe E, Lawrence D. Cancer-related mortality in people with mental illness. JAMA Psychiatry 2013;70:209-17. [Crossref] [PubMed]
31. Batra A, Yang L, Boyne DJ, et al. Associations between baseline symptom burden as assessed by patient-reported outcomes and overall survival of patients with metastatic cancer. Support Care Cancer 2021;29:1423-31. [Crossref] [PubMed]
32. Nekolaichuk CL, Bruera E, Spachynski K, et al. A comparison of patient and proxy symptom assessments in advanced cancer patients. Palliat Med 1999;13:311-23. [Crossref] [PubMed]
33. Laugsand EA, Sprangers MA, Bjordal K, et al. Health care providers underestimate symptom intensities of cancer patients: a multicenter European study. Health Qual Life Outcomes 2010;8:104. [Crossref] [PubMed]