Opioid Dose and Survival of Patients with Incurable Nonsmall Cell Lung Cancer: A Prospective Cohort Study

Abstract

Background:

Preclinical studies show that opioids promote angiogenesis, tumor progression, and metastasis, resulting in shorter survival.

Objective:

To explore whether opioids are associated with the overall survival (OS) of patients with incurable nonsmall cell lung cancer (NSCLC).

Design:

Prospective cohort study of patients with NSCLC.

Setting:

We investigated patients newly diagnosed with advanced or post-operative recurrent NSCLC between April 2013 and December 2015 at a single institute.

Measurements:

We evaluated OS, opioid requirements, opioid doses, pain levels, and prognostic factors of advanced NSCLC. The effects of variables on survival were analyzed using univariable and multivariable models. Patients were stratified according to oral morphine equivalents (OMEs)/day (<60 or ≥60 mg) to assess the association between opioid dose and OS.

Results:

We analyzed 150 patients, including 64 who received opioid treatment during follow-up. The median OS was 242 days in the opioid group and 627 days in the no-opioid group (log-rank p < 0.001). Multivariable models revealed that the opioid requirement was an independent predictor of shorter OS, after adjustment for prognostic variables, including sex, histology, stage, history of systemic chemotherapy, and performance status (hazard ratio 1.73, 95% confidence interval 1.137–2.631). There was no significant difference in OS between patients who received ≥60 mg OME/day for 250 days versus <60 OME/day for 242 days.

Conclusions:

The opioid dose does not shorten the survival of patients with advanced NSCLC. The opioid requirement is associated with shorter survival when opioids are administered any time during the clinical course, independent of the influence of other key factors.

Introduction

Pain is a common symptom that adversely affects the quality of life of patients with cancer. There are many guidelines based on past studies for the use of opioids to alleviate the pain of cancer.^1,2 Evidence indicates that pain may induce cancer progression through several mechanisms,³ and opioids may be associated with longer survival by reducing pain.⁴ Opioids have beneficial as well as adverse effects. For example, patients suffer side effects such as constipation, nausea, and delirium. Furthermore, there are rare, but serious side effects such as opioid addiction, respiratory suppression, and prolongation of the cardiac QTc interval, which cause life-shortening torsades de pointes.^5,6 The chronic effects of opioids may include immunosuppression and tumor angiogenesis, leading to tumor progression and metastasis.^7–9 The evidence supporting these mechanisms was acquired from laboratory research and therefore may not apply to patients.

We lack sufficient evidence supporting the conclusion that opioid use independently influences the OS of patients with cancer. However, a previous study of patients with advanced nonsmall cell lung cancer (NSCLC) suggests that opioid treatment is associated with shorter survival.¹⁰ However, this study has certain methodological limitations such as its retrospective design, the lack of a control for a painfully progressive pattern of cancer, and an established prognostic factor such as a driver mutation.^10–12 We hypothesized that opioid treatment does not shorten the survival of patients with NSCLC. This study is the first to our knowledge to investigate the association between treatment with prescribed opiates and patient survival after diagnosis of incurable NSCLC.

Materials and Methods

Study design and patients

We prospectively studied patients with advanced or post-operative recurrence of NSCLC at the Department of Respiratory Medicine and Medical Oncology, Gifu Municipal Hospital. Eligibility criteria were as follows: diagnosis of stage IIIB not amenable to curative treatment, stage IV, or post-operative recurrent NSCLC; age ≥20 years; and no previous treatment for NSCLC. Exclusion criteria were as follows: diagnosis of other active and invasive malignancies and a patient's refusal to be included in the study. Patients with post-operative recurrence, who were treated with adjuvant therapy, were included if the interval between the end of intravenous adjuvant chemotherapy and registration exceeded six months. Consecutive patients were enrolled between April 2013 and December 2015 to determine whether an opioid prescription and the opioid dose were associated with survival. The data cutoff date was July 31, 2016. To chase missing data, the patients were followed using a survey sent to community medical institutions.

The Institutional Ethics Committee of Gifu Municipal Hospital approved the study protocol. Japanese law does not require investigators to obtain the informed consent of participants in a noninvasive observational trial such as this, which analyzes medical records. Therefore, we used an opt-out method rather than acquiring written or oral informed consent. This study is registered with the University hospital Medical Information Network (UMIN 000010363).

Opioid prescriptions

Data for opioid users included opioid dose, the cumulative prescribed opioid dose, and the mean daily opioid dose. The opioid requirement was defined as opioid use during the entire clinical course for cancer-related pain or dyspnea. The cumulative prescribed doses of different opioids administered to patients were converted to the oral morphine equivalent (OME) dose.^2,13,14 Pentazocine, buprenorphine, and codeine were not used for cancer-related pain management during the study. We calculated the mean daily opioid doses by dividing the cumulative prescribed opioid dose by the number of days any opioid was regularly administered. Opioids were administered by attending physicians who participated in the Palliative care Emphasis program on symptom management and Assessment for Continuous Medical Education.¹⁵ All physicians were pulmonologists associated with the Department of Respiratory Medicine and Medical Oncology, Gifu Municipal Hospital.

Pain levels

Cancer-related pain levels were obtained from medical records at the time of diagnosis of advanced NSCLC or recurrence. Pain severity was categorized according to the numerical rating scale (NRS) as follows: none (0), low (1–3), moderate (4–6), and severe (7–10). For patients who could not report their pain level using the NRS, ratings were obtained for the response options as follows: none, low, moderate, and severe. Regular pain assessment was performed by the attending physician.

Biomedical factors

Patients' characteristics, including age, sex, histology (including driver mutations such as EGFR mutations and ALK rearrangements), clinical stage, history of systemic chemotherapy, and Eastern Cooperative Oncology Group performance status (PS) were obtained from patients' records. These factors are associated with survival.^16–19 We established an age cutoff of 75 years according to clinical trials that included older patients with advanced lung cancer.²⁰ The clinical stage was determined according to the Union for International Cancer Control Tumor-Node-Metastasis classifications, 7th edition.²¹ Cancer treatment included systemic chemotherapy, molecularly targeted drugs, and immune-checkpoint inhibitors.

Statistical analysis

The primary endpoint of this study was overall survival (OS), which was assessed from the date of diagnosis of NSCLC or post-operative recurrence to death from any cause. Survival curves were calculated using the Kaplan–Meier method, and the data were compared using the log-rank test. Univariate and multivariate Cox regression analyses were used to assess the effects of clinical factors on OS. The analyses evaluated the prognostic importance of the variables as follows: age (<75 years vs. ≥75 years), sex, histology, including driver mutations (driver mutation-positive adenocarcinoma, and others), disease stage (IIIB or IV/post-operative recurrence), PS (0, 1/2/3, and 4), pain level (none, mild/moderate, and severe), and cancer treatment. Independent variables included in the multivariable Cox model were selected from these factors using a forward stepwise method, where the opioid requirement was forcibly administered. We used the log-rank test to assess the dose–response relationship between the average daily opioid dose and OS. We divided opioid users into groups that received a lower or higher dose of opioids (<60 mg OME/day or ≥60 mg OME/day, respectively).²²

The results of Cox regression analyses are presented as hazard ratios and 95% confidence intervals. All p-values are two sided, and p < 0.05 was considered statistically significant. All statistical analyses were performed using R version 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Patients' characteristics

We enrolled 151 consecutive patients, and excluded one because of another active and invasive malignancy. The patients' characteristics upon diagnosis and treatment are summarized in Table 1. The median opioid dose (interquartile range) was 35.8 mg OME/day (21.0–71.3 mg OME/day). The median follow-up (interquartile range) of censored patients was 603 days (281–819 days), and 61 and 41 patients who did or did not receive opioids, respectively, died during the study. Among patients who received opioids, 29, 23, and 10 had used nonsteroidal anti-inflammatory drugs, acetaminophen, and adjuvant drugs such as pregabalin or duloxetine when opioids were administered, respectively.

Table 1.

The Association of Patients' Characteristics with Opioid Treatment

	No opioid (n = 86) Number (%)	Opioid (n = 64) Number (%)	p
Age (years)
Median (range)	73 (47–95)	71 (43–91)	0.10
≥75	35 (41%)	21 (33%)	0.32
<75	51 (59%)	43 (67%)
Sex			0.15
Male	52 (60%)	46 (72%)
Female	34 (40%)	18 (28%)
Histology			0.001
Others	59 (69%)	58 (91%)
Driver mutation(+) Ad	27 (31%)	6 (9%)
Stage^a			0.66
IIIB+IV	76 (88%)	58 (91%)
Post-operative recurrence	10 (12%)	6 (9%)
ECOG PS			0.48
0–1	55 (64%)	36 (56%)
2	14 (16%)	10 (16%)
3–4	17 (20%)	18 (28%)
Pain level at diagnosis			0.25
None–mild	81 (94%)	57 (89%)
Moderate–severe	5 (6%)	7 (11%)
Anticancer treatment			0.57
Absent	31 (36%)	26 (41%)
Present	55 (64%)	38 (59%)

Based on the Union for International Cancer Control tumor-node-metastasis classifications, 7th edition.

Ad, adenocarcinoma; others, squamous cell carcinoma, not otherwise specified; large cell carcinoma and driver mutation-negative Ad; ECOG, Eastern Cooperative Oncology Group; PS, performance status.

Association with survival of prescribed opioids and other variables

Univariable analysis indicated that sex, histology, clinical stage, cancer treatment, PS, and opioid requirement were significantly associated with shorter OS, although there was no significant association with age and pain (Table 2). The median OS values were 242 and 627 days, respectively, for those administered opioids, or not (log-rank p < 0.001) (Fig. 1).

FIG. 1.

Kaplan–Meier plots of survival. (A) The median OS was 242 days in the opioid group and 627 days in the no-opioid group, log-rank p < 0.001. (B) Median OS was 250 days in the higher opioid dose group (≥60 mg OME/day) and 242 days in the lower dose group (<60 OME/day); log-rank p = 0.85. Opioid group (n = 64), no-opioid group (n = 86). OS, overall survival.

Table 2.

Univariable and Multivariable Analyses of Overall Survival

	Univariable			Multivariable
Variable	Hazard ratio	95% confidence interval	p	Hazard ratio	95% confidence interval	p
Opioid treatment
Absent
Present	2.81	1.86–4.24	<0.001	1.73	1.14–2.63	0.01
Age (years)
<75
≥75	1.48	0.99–2.21	0.055
Sex
Male
Female	0.45	0.29–0.71	<0.001	0.59	0.36–0.95	0.03
Histology
others
Driver mutation(+) Ad	0.12	0.06–0.26	<0.001	0.27	0.11–0.62	0.002
Stage
IIIB+IV
Post-operative recurrence	0.43	0.21–0.89	0.02	0.49	0.23–1.04	0.06
ECOG PS
0–1
2	4.84	2.80–8.38	<0.001	3.38	1.84–6.21	<0.001
3–4	6.32	3.88–10.30	<0.001	4.00	2.22–7.20	<0.001
Pain level at diagnosis
None–mild
Moderate–severe	0.99	0.46–2.14	0.98
Anticancer treatment
Absent
Present	0.20	0.13–0.31	<0.001	0.59	0.36–0.95	0.03

Multivariable models identified an opioid requirement as an independent significant predictor of shorter OS after adjustment for the clinical prognostic variables as follows: sex, histology, clinical stage, cancer treatment, and PS (hazard ratio 1.73 and 95% confidence interval, 1.14–2.63) (Table 2). However, there were no significant differences in survival among patients who received different doses of opioids (higher dose for 250 days vs. lower dose for 242 days, log-rank p = 0.85) (Fig. 1). There was no increased risk of mortality associated with the daily OME (hazard ratio 0.98 and 95% confidence interval, 0.82–1.17).

Discussion

This is the first prospective cohort study, to our knowledge, to investigate the association between treatment with prescribed opioids and survival after a diagnosis of incurable NSCLC. We found that the dose of prescribed opioids was not significantly associated with OS. Thus, the Kaplan–Meier curves were very close between the higher opioid dose group and the lower dose group, and an opioid dose–response relationship with OS was not observed. This result is consistent with previous studies of terminally ill cancer patients.^13,23 The chronic effects of opioids lead to immunosuppression,²⁴ and opioids may promote cancer progression and reduce survival. However, the use of opioids could improve pain control and may have contributed to the potential survival benefit.

The opioid requirement was independently associated with shorter OS among clinical prognostic factors. This finding may be explained by the onset of a painfully progressive pattern such as multiple bone metastasis and extensive pleural invasion during the entire clinical course.

Opioids may affect survival. The effects of opioids include respiratory depression, delirium, and addiction, leading to fatal adverse events. However, studies on the influence of opioids on the survival of end-of-life patients do not suggest a significant association of opioid use and survival.^13,23 Opioid addiction or a prolonged QTc interval that causes torsades de pointes was not observed in this study.

This study has several limitations. First, our study measured pain levels, PS, sex, stage, age, histology, and cancer therapy, but not other predictive factors such as body-weight loss, serum lactate dehydrogenase levels, comorbidities, and brain metastasis.^25–27 Second, this study was conducted at a single institution. The availability of opioids varies among institutions and settings, and our findings cannot be generalized to settings where opioids are rarely used and therefore may not be applicable to patients with other types and clinical stages of cancer. Third, no formal sample calculation was performed. With change in the standard therapy of advanced NSCLC, we had to terminate patient enrollment. The sample size of 150 patients with advanced NSCLC was adequate to explore the association between opioid requirement and survival. However, this study was not sufficiently powerful to determine whether high-dose opioids affected OS. Fourth, opioid treatment, particularly the initial dose, titration schedule, choice of opioid, and the reason for opioid use did not follow a standardized protocol. However, attending physicians were responsible for patient treatment, and we assumed therefore that opioid treatment was administrated according to each patient's condition according to the national clinical guideline.² The main reason for opioid use was pain management. We could not verify the long-term prognosis of patients who used opioids exclusively for management of dyspnea.

Furthermore, we were unable to conduct a placebo-controlled, randomized trial. Therefore, large data sets or a multicenter prospective design may provide the answer to the question of whether opioids influence survival. In this study, patients were not administered a neural blockade for cancer pain and did not receive comprehensive opioid drug management with the peripherally acting μ-opioid receptor antagonist (PAMORA).²⁸ Such methods may benefit cancer patients with a prognosis of longer survival.

Conclusion

Our data suggest that the opioid dose did not adversely influence the survival of patients with advanced NSCLC. The opioid requirement was associated with shorter survival when administered at any time during the clinical course, independent of the influence of other key factors. Moreover, prospective clinical trials are required, which are designed to determine whether PAMORA improves long-term survival of patients with advanced cancer. According to available data, oncologists should encourage patients to use opioids for the management of cancer-related symptoms.

Footnotes

Acknowledgments

The authors thank the study participants and their families for support; staff physicians Yohei Futamura, Akane Horiba, Takashi Ishiguro, and Tsutomu Yoshida; and research assistant Tamami Taniguchi for their contributions to the study. This study was supported by a grant-in-aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science, and Technology.

Author Disclosure Statement

No competing financial interests exist.

References

Caraceni

, Hanks

, Kaasa

, et al.: Use of opioid analgesics in the treatment of cancer pain: Evidence-based recommendations from the EAPC. Lancet Oncol, 2012; 13:e58–e68.

Yamaguchi

, Shima

, Morita

, et al.: Clinical guideline for pharmacological management of cancer pain: The Japanese Society of Palliative Medicine recommendations. Jpn J Clin Oncol, 2013; 43:896–909.

Herndon

2nd , Fleishman

, Kornblith

, et al.: Is quality of life predictive of the survival of patients with advanced nonsmall cell lung carcinoma?. Cancer, 1999; 85:333–340.

Halabi

, Vogelzang

, Kornblith

, et al.: Pain predicts overall survival in men with metastatic castration-refractory prostate cancer. J Clin Oncol, 2008; 26:2544–2549.

Farsi

, Mirafzal

, Hassanian-Moghaddam

, et al.: The correlation between prolonged corrected QT interval with the frequency of respiratory arrest, endotracheal intubation, and mortality in acute methadone overdose. Cardiovasc Toxicol, 2014; 14:358–367.

Pattinson

: Opioids and the control of respiration. Br J Anaesth, 2008; 100:747–758.

Lennon

, Mirzapoiazova

, Mambetsariev

, et al.: The Mu opioid receptor promotes opioid and growth factor-induced proliferation, migration and epithelial mesenchymal transition (EMT) in human lung cancer. PLoS One, 2014; 9:e91577.

Mathew

, Lennon

, Siegler

, et al.: The novel role of the mu opioid receptor in lung cancer progression: A laboratory investigation. Anesth Analg, 2011; 112:558–567.

Singleton

, Mirzapoiazova

, Hasina

, et al.: Increased mu-opioid receptor expression in metastatic lung cancer. Br J Anaesth, 2014; 113 Suppl 1:i103–i108.

10.

Zylla

, Kuskowski

, Gupta

et al.: Association of opioid requirement and cancer pain with survival in advanced non-small cell lung cancer. Br J Anaesth, 2014; 113 Suppl 1:i109–i116.

11.

Takano

, Fukui

, Ohe

, et al.: EGFR mutations predict survival benefit from gefitinib in patients with advanced lung adenocarcinoma: A historical comparison of patients treated before and after gefitinib approval in Japan. J Clin Oncol, 2008; 26:5589–5595.

12.

, Kuo

, Chang

, et al.: EML4-ALK translocation predicts better outcome in lung adenocarcinoma patients with wild-type EGFR. J Thorac Oncol, 2012; 7:98–104.

13.

Minami

, Fujimoto

, Ogata

, et al.: Opioids have no negative effect on the survival time of patients with advanced lung cancer in an acute care hospital. Support Care Cancer, 2015; 23:2245–2254.

14.

Saunders

, Messina

, Darwish

, et al.: Assessment of the relative potency of fentanyl buccal tablet to intravenous morphine in healthy volunteers using a thermally induced hyperalgesia pain model. J Clin Pharmacol, 2012; 52:870–879.

15.

Yamamoto

, Kizawa

, Nakazawa

, et al.: Outcome evaluation of the Palliative care Emphasis program on symptom management and Assessment for Continuous Medical Education: Nationwide physician education project for primary palliative care in Japan. J Palliat Med, 2015; 18:45–49.

16.

David

, Daly

, Li

, et al.: Increasing rates of no treatment in advanced-stage non-small cell lung cancer patients: A propensity-matched analysis. J Thorac Oncol, 2017; 12:437–445.

17.

Efficace

, Bottomley

, Smit

, et al.: Is a patient's self-reported health-related quality of life a prognostic factor for survival in non-small-cell lung cancer patients? A multivariate analysis of prognostic factors of EORTC study 08975. Ann Oncol, 2006; 17:1698–1704.

18.

Kogure

, Ando

, Saka

, et al.: Histology and smoking status predict survival of patients with advanced non-small-cell lung cancer. Results of West Japan Oncology Group (WJOG) Study 3906L. J Thorac Oncol, 2013; 8:753–758.

19.

Mitsudomi

, Morita

, Yatabe

, et al.: Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): An open label, randomised phase 3 trial. Lancet Oncol, 2010; 11:121–128.

20.

Maemondo

, Minegishi

, Inoue

, et al.: First-line gefitinib in patients aged 75 or older with advanced non-small cell lung cancer harboring epidermal growth factor receptor mutations: NEJ 003 study. J Thorac Oncol, 2012; 7:1417–1422.

21.

Goldstraw

, Crowley

, Chansky

, et al.: The IASLC Lung Cancer Staging Project: Proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM Classification of malignant tumours. J Thorac Oncol, 2007; 2:706–714.

22.

Bercovitch

, Adunsky

: Patterns of high-dose morphine use in a home-care hospice service: Should we be afraid of it?. Cancer, 2004; 101:1473–1477.

23.

Morita

, Tsunoda

, Inoue

, et al.: Effects of high dose opioids and sedatives on survival in terminally ill cancer patients. J Pain Symptom Manage, 2001; 21:282–289.

24.

Boland

, McWilliams

, Ahmedzai

, et al.: Effects of opioids on immunologic parameters that are relevant to anti-tumour immune potential in patients with cancer: A systematic literature review. Br J Cancer, 2014; 111:866–873.

25.

Akechi

, Okamura

, Okuyama

, et al.: Psychosocial factors and survival after diagnosis of inoperable non-small cell lung cancer. Psychooncology, 2009; 18:23–29.

26.

Denehy

, Hornsby

, Herndon

2nd , et al.: Prognostic validation of the body mass index, airflow obstruction, dyspnea, and exercise capacity (BODE) index in inoperable non-small-cell lung cancer. J Thorac Oncol, 2013; 8:1545–1550.

27.

Sorensen

, Hansen

, et al.: Brain metastases in adenocarcinoma of the lung: Frequency, risk groups, and prognosis. J Clin Oncol, 1988; 6:1474–1480.

28.

Janku

, Johnson

, Karp

, et al.: Treatment with methylnaltrexone is associated with increased survival in patients with advanced cancer. Ann Oncol, 2016; 27:2032–2038.