Abstract
Background/Aim: The aim of this study was to investigate the impact of drug-induced interstitial lung disease (DILD) on the mortality of patients with lung cancer. Patients and Methods: Japanese patients with lung cancer who had received chemotherapy in Fujita Health University Hospital from January 2017 to December 2018 were enrolled in this study. The primary outcome was to identify independent factors associated with patient mortality. The secondary outcome was to identify the risk factor of DILD. Results: Four hundred and fifty-seven patients were assigned to the current study. The multivariable analysis revealed that being aged 75 years or older, small cell lung carcinoma, cancer stage IV, and DILD event were risk factors of mortality. Male sex was identified as a risk factor of DILD. Conclusion: DILD event has the same degree of risk for mortality as age 75 years or older in lung cancer patients.
In the past 20 years, the mortality of lung cancer has decreased with the development of kinase inhibitors; however, the risk factors of cancer death have been largely unchanged. Aging, history of smoking, male sex (1-4), small cell lung cancer (SCLC) (5), and cancer stage Ⅳ (6) have been reported as risk factors of mortality. Although acquisition of drug resistance in cancer cells is the cause of poor prognosis, severe side effects should not be ignored. Drug-induced interstitial lung disease (DILD) has been recognized as a severe side effect of chemotherapy (7, 8), and the mortality of DILD is over 20% (9-11). DILD is commonly recognized as a risk factor of mortality in patients with lung cancer, however there are no reports available comparing DILD with the abovementioned “well-known” risk factors.
To prevent the development of DILD, the discontinuation of chemotherapy (12) and initiation of corticosteroid therapy should be considered (13). Since these therapeutic interventions increase the cost of treatment and might decrease the efficacy of chemotherapy in patients with lung cancer, it is important to predict DILD. Male sex and a history of smoking have been reported as risk factors for DILD induced by gefitinib (14). In addition, poor performance status (PS) (10, 15), old age, history of smoking (15), previous or concomitant interstitial lung disease (ILD) (15-17) are risk factors of DILD induced by anticancer chemotherapy. However, these previous studies did not distinguish between de novo DILD and exacerbation of pre-existing ILD. In addition, cancer stage Ⅳ, as opposed to less severe cancer stages, might be a confounding factor of poor PS. To identify the risk factors of DILD, studies considering these factors are needed. Thus, patients with a history of ILD should be excluded in future studies, and cancer stage Ⅳ should be evaluated as a factor of DILD.
Here, to evaluate the impact of DILD on the therapeutic outcome of patients with lung cancer, we investigated the risk factors of DILD and compared the relative contributions of various risk factors to patient mortality.
Patients and Methods
Data source and study design. Japanese patients with lung cancer who had received chemotherapy in Fujita Health University Hospital from January 2017 to December 2018 were enrolled in this retrospective cohort study. The follow-up period was until September 2019. All data were collected from the medical records of the Fujita Health University Hospital. The exclusion criteria were patients a) aged <20 years, b) received chemotherapy for cancer other than lung cancer, c) undergoing renal replacement therapy, and d) with a history of ILD. DILD induced by anticancer chemotherapy was diagnosed using the following criteria (18): a) interstitial pulmonary infiltrates identified on computed tomography (CT) by at least two physicians, b) elimination of other disease diagnoses (e.g., tumor progression, pulmonary infection, and cardiovascular disease), and c) no exposure to other medicines or materials that might cause DILD. Patients with lung cancer were divided into non-DILD or DILD groups (Figure 1).
Outcome measures. We measured overall survival time from the initiation of chemotherapy to the end of the follow-up period as a primary outcome to evaluate the prognosis of patients with DILD. Cox proportional hazards model was used to identify independent factors associated with patient mortality as a primary outcome. This model applied for the following covariates: age 75 years or older, male sex, history of smoking, diagnosis of SCLC, cancer stage Ⅳ, and DILD event. To identify the risk factors of DILD as a secondary outcome, we conducted multivariable analysis applied for the following covariates: male sex and history of radiation.
Statistical analyses. All data were analyzed using their mean value and range. Student’s t-test and chi-square test as univariate analyses were used to analyze continuous and nominal data, respectively. Logistic regression analysis using a multivariable model was used to identify the risk factors of DILD. Cox proportional hazards models were used to identify the risk of mortality with hazard ratios (HRs) and 95% confidence intervals (CIs). The fitness of the logistic regression model was evaluated by Hosmer-Lemeshow test. The prediction ability was evaluated by the area under the curve of the receiver operating characteristic (AUROC). A two-sided p-value of <0.05 was considered significant in all statistical analyses. SPSS version 22.0 (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses.
Ethics approval. This study was approved by the ethics board of Fujita Health University Hospital (ethics committee approval number: HM20-364) and conducted in accordance with the appropriate guidelines. Because it was a retrospective cohort study, an opt-out approach of informed consent was used as approved by the ethics board.
Results
Patient characteristics. The total number of patients enrolled in this study was 519. Thirty-four patients who received chemotherapy for cancer other than lung cancer and 14 patients with unclear histories of chemotherapy were excluded. To distinguish DILD and pre-existing ILD, 14 patients with histories of ILD were excluded from the current study. There were no patients <20 years of age or undergoing renal replacement therapy. Four hundred and fifty-seven patients were included in the current study (Figure 1). Baseline characteristics are shown in Table I. Patients with or without DILD event were divided into non-DILD (n=410) and DILD (n=47) groups, respectively. The number of male patients was significantly higher in the DILD group (p=0.021). The number of patients with a history of radiation showed an increasing trend in the DILD group compared to that in the non-DILD group (p=0.071). There were no significant differences in other characteristics between two groups (Table II). Drugs suspected to cause DILD were osimertinib (n=5), gefitinib (n=2), afatinib (n=2), alectinib (n=1), erlotinib (n=1), nivolumab (n=9), pembrolizumab (n=7), docetaxel (n=4), pemetrexed (n=3), docetaxel+ ramucirumab (n=3), amrubicin (n=2), irinotecan (n=1), tegafur/gimeracil/oteracil (TS-1) (n=2), abraxane (n=1), cisplatin+etoposide (n=1), carboplatin+gemcitabine (n=1), docetaxel+nedaplatin (n=1), and carboplatin+abraxane (n=1).
Risk factors of mortality in patients with lung cancer. The multivariable analysis revealed that age 75 years or older (HR=1.667, 95%CI=1.174-2.368, p=0.004), small cell lung carcinoma (HR=2.528, 95%CI=1.688-3.788, p<0.001), cancer stage IV (HR=2.302, 95%CI=1.653-3.205, p<0.001), and DILD event (HR=1.877, 95%CI=1.239-2.846, p= 0.003) were risk factors of mortality (Table III). DILD event was a higher risk factor of mortality than male sex and history of smoking.
Risk factors of DILD in patients with lung cancer. Since the number of male patients and patients with histories of radiation in the DILD group were higher than those of the non-DILD group, we conducted multivariable analysis by using these characteristics. Male sex was identified as a risk factor of a DILD event (OR=2.297, 95%CI=1.037-5.092, p=0.041) (Table IV), and this multivariable model was fitted in a logistic analysis model (Hosmer-Lemeshow test, p=0.628). However, no significant difference was found in the ROC analysis (sensitivity=83.0%, specificity=33.7%, AUCROC=0.583, p=0.062).
Discussion
Our results suggest that a DILD event is a more severe risk factor of mortality in patients with lung cancer than history of smoking and male sex. In addition, DILD has the same degree of risk for mortality as old age. Interstitial lung abnormalities were reported as a high-risk factor of mortality in four separate research cohorts: the Framingham Heart Study, the Age Gene/Environment Susceptibility-Reykjavik Study, the chronic obstructive pulmonary disease (COPD) Gene Study, and the Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points study (19). Because the associations between interstitial lung abnormalities and mortality were not attenuated after adjustment for smoking or cancer in this previous research, interstitial lung abnormalities had a more severe impact on mortality than other characteristics. These cohorts investigated the therapeutic outcome of patients with cardiovascular disease, COPD, and did not focus on DILD induced by anticancer chemotherapy. Although these results could not be directly compared with the results of previous studies, our study supported the results of previous research cohorts.
In the present study, univariate and multivariate analyses suggested that male sex was a predictor of DILD event. However, old age, history of smoking, history of radiation, and cancer stage IV were not identified as risk factors of DILD in univariate analyses. Thus, our results differed from the previous reports that included patients with pre-existing ILD as subjects. Since we did not evaluate histopathological findings and severity of DILD, it was unclear whether pre-existing ILD was a confounding factor in the previous studies. However, male sex was identified as a risk factor of DILD in the present study; hence our results partially supported a previous report (14). Although history of smoking might associate with a DILD event in male patients, to elucidate this mechanism, further studies are needed. Our findings suggest that a preventive approach of DILD is important for decreasing the mortality of patients with lung cancer and continued monitoring of male patients undergoing chemotherapy is needed.
A limitation of the present study was that the pathological characteristics of DILD were not evaluated. Difficulty in diagnosing ILD has been reported as a limitation in previous studies (20, 21). As our study was a retrospective cohort study, a prospective study should be conducted to overcome this limitation. In conclusion, a DILD event has the same risk for mortality as an age of 75 years or older in lung cancer patients.
Acknowledgements
The Authors would like to thank the patients and their families for their contribution to this study.
Footnotes
↵* These Authors contributed equally to this work.
Authors’ Contributions
Kohei Iwashita, Tomohiro Mizuno, and Satomi Kumazawa designed this study. Kohei Iwashita and Satomi Kumazawa carried out the survey of the electronic records. Kohei Iwashita and Tomohiro Mizuno performed the statistical analyses. Kohei Iwashita, Tomohiro Mizuno, Kazuyoshi Imaizumi, and Shigeki Yamada drafted the manuscript. All Authors approved the final manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
- Received February 6, 2021.
- Revision received February 14, 2021.
- Accepted February 15, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.