Impact of Chronic Obstruction Pulmonary Disease with Acute Exacerbation on Survival in Current-Smoking Patients with Advanced stages Lung Squamous Cell Carcinoma undergoing Concurrent Chemoradiotherapy

BACKGROUND no data are available regarding the effect of chronic obstruction pulmonary disease (COPD) and COPD with acute exacerbation (COPDAE) on survival in patients with lung squamous cell carcinoma (SCC) receiving concurrent chemoradiotherapy (CCRT). METHODS


Introduction
Smoking is the leading cause of chronic obstructive pulmonary disease (COPD). [1] Smoking is also a trigger for COPD are-ups. Smoking damages the air sacs, airways, and the lining of lungs. [1,2] COPD and smoking are the independent risk factors for lung carcinoma, particularly squamous cell carcinoma (SCC). [3,4] COPD increases the risk of lung cancer by 6-13 times in the affected patients relative to individuals without COPD. [5,6] Lung cancer and COPD occur as comorbidities at a higher rate than would be expected if they were independently triggered. [6] Several mechanisms have been proposed to explain the association between COPD and lung cancer, including genetic risk factors, [7] common epigenetic processes, [8] and chronic local and systemic in ammatory processes. [9] In addition, lung cancer is the main cause of death among patients with COPD. [10] A detailed understanding of which clinical features of COPD increase the risk of lung cancer death is required.
[11] COPD is associated with chronic local and systemic in ammatory processes in lung tissue, [9] which might contribute to a low tolerance to curativeintent treatments, including radiotherapy (RT), lung lobectomy, or concurrent chemoradiotherapy (CCRT).
Nevertheless, no data are available regarding the survival effects of non-COPD, COPD, or COPD with acute exacerbation (COPDAE) in patients with lung cancer across curative-intent treatments, pathologic type of lung cancer, or lung cancer stage. COPDAE is contributes to long term decline in lung function of patients with moderate to severe COPD [12]. COPDAE might be a risk factor for the prognostic factor of survival in patients with lung cancer, although there has been no data to prove this.
For most patients with clinically evident mediastinal lymph node (N2) lung SCC (stage IIIA-IIIB according to American Joint Committee on Cancer [AJCC], Eighth Edition), the treatment approach is de nitive CCRT using platinum-based chemotherapy plus full-dose RT. [13][14][15] For N2 lung SCC, de nitive CCRT is the category 1 treatment approach as per the National Comprehensive Cancer Network (NCCN) guidelines [13][14][15]. Some studies have shown that COPD is associated with the risk of radiation-induced lung injury (RILI). [16,17] Among 80 patients with stage IIIA-IIIB non-small-cell lung cancer treated with cisplatinbased chemotherapy and RT, COPD was associated with an increased frequency of RILI including radiation pneumonitis (RP), an increased risk of severe pneumonitis, and late lung brosis. [18,19] COPD or severe COPD (COPDAE) associated with acute and late RILI might reduce survival in patients with lung SCC receiving de nitive CCRT.
Until now, no study has investigated the survival effects of non-COPD, COPD, and COPDAE in patients with lung SCC receiving de nitive CCRT. In the current study, we examined whether COPD or COPDAE is a risk factor for all-cause death, COPD death, and lung cancer death in patients with stage IIIA-IIIB lung SCC receiving standard CCRT.

Study population
In this cohort study, data were retrieved from the combination of two databases of National Health Insurance Research Database (NHIRD) and Taiwan Cancer Registry Database (TCRD). We enrolled current-smoking patients who had received a diagnosis of lung SCC with mediastinal lymph node positive (clinical N2 stage, AJCC stage IIIA-IIIB) between January 1, 2008, and December 31, 2017. The index date was the start date of de nitive CCRT, and the follow-up duration was from the index date to December 31, 2018. The TCRD of the Collaboration Center of Health Information Application contains detailed cancer-related information of patients, including the clinical stage, treatment modalities, chemotherapy regimens, dose of chemotherapy, pathology, radiation modalities and doses, and treatment protocols (CCRT or non-CCRT). [20][21][22][23][24][25] The study protocols were reviewed and approved by the Institutional Review Board of Tzu-Chi Medical Foundation (IRB109-015-B).

Inclusion and exclusion criteria
The diagnoses of the enrolled current-smoking patients were con rmed through a review of their pathological data, and patients with newly diagnosed lung SCC were con rmed to have no other cancers or distant metastases. Patients were included if they had received lung SCC diagnosis, were aged ≥ 20 years, and had AJCC clinical stages IIIA-IIIB without metastasis. Patients were excluded if they had a history of cancer before lung SCC diagnosis, had distant metastasis, had unknown pathologic type, had missing sex data, were aged < 20 years, had unclear staging, or exhibited non-SCC histology. In addition, we excluded patients with lung SCC if they did not receive any treatments within 3 months after diagnosis, received an insu cient CCRT dose (< 6000 cGy) after lung SCC diagnosis, received insu cient chemotherapy (concurrent chemotherapy with two agents containing platinum at least), or did not receive a platinum-based chemotherapy regimen. We also excluded those who received only sequential chemotherapy and RT, chemotherapy alone, or RT alone. Standard CCRT comprises concurrent chemotherapy with two agents containing platinum and thoracic RT with 6000 cGy in daily fractions. [26][27][28] Finally, we enrolled 3986 patients with AJCC stages IIIA-IIIB unresectable lung SCC who had received de nitive CCRT and assigned them into two groups based on their COPD status to compare overall survival outcomes: group 1 (patients with COPD before CCRT) and group 2 (patients without COPD before CCRT). We also investigated the effects of COPD severity (0, 1, or ≥ 2 hospitalizations for COPDAE within 1 year before the index date) on the survival outcomes in patients with stages IIIA-IIIB lung SCC receiving de nitive CCRT. In the current study, we retrieved data for all patients with diagnoses of COPD from Taiwan's NHIRD. The diagnoses of COPD were validated by the analysis of selected samples from the claims database of Taipei Veterans General Hospital (a 3035-bed tertiary referral hospital in Taiwan) using the same criteria in the previous study [29]. Almost all medical services provided by Taipei Veterans General Hospital are covered and recorded in the NHIRD [29]. The enrolled patients were all current smoking patients. However, the smoking intensity is not listed in Tables 1 and 2, because the strong collinearity of smoking intensity, COPDAE, histological degree of differentiation of the cancer, and smoking-associated complications. In statistics, multicollinearity (also collinearity) is a phenomenon in which one predictor variable in a multiple regression model can be linearly predicted from the others with a substantial degree of accuracy [30]. In this situation, the coe cient estimates of the multiple regression may change erratically in response to small changes in the model or the data [30].The incidence of comorbidities was scored using the Charlson comorbidity index (CCI). [31,32] Only the comorbidities observed in the 6 months before the index date were included; they were coded and classi ed according to the International Classi cation of Diseases, 10th Revision, Clinical Modi cation (ICD-10-CM) at the rst admission or after more than two repetitions of a code were issued at outpatient department visits. The supplemental Fig. 1 was the ow-chart of patient selection.

Study covariates and statistical analysis
Signi cant independent predictors, namely age, sex, cancer stages, histological degree of differentiation, diabetes, chronic bronchitis, asthma, emphysema, cardiovascular disease (CVD), acute myocardial infarction (AMI), stroke, CCI score, income level, and urbanization, were analyzed using the multivariate Cox proportional hazard model to determine hazard ratios (HRs). We applied inverse probability of treatment weighting (IPTW) to create a pseudo-study cohort; the weighted cohort avoids covariate bias and mimics randomized COPD or non-COPD assignment: IPTW for patients with COPD = 1/p(COPD); IPTW for patients without COPD = 1/(1 − p[COPD]). [33,34] The independent predictors were examined in univariate and multivariable analyses before and after IPTW adjustment. The independent predictors were controlled for and were strati ed in analysis. The primary endpoint was the mortality rate in the patients with COPD, with group 1 used as the control. According to the cause of death pro les in NHIRD, we also supplied the secondary endpoints of death of COPD, and death of cancer in our supplemental Tables 1 and 2.
The cumulative incidence of death was estimated using the Kaplan-Meier method, and differences in the frequency of hospitalization between patients with COPDAE, with COPD, and without COPD with lung SCC receiving de nitive CCRT were determined using the log-rank test. After adjustment for confounders, IPTW-adjusted models were used to determine the time from the index date to all-cause mortality in the patients with COPD, with COPDAE, and without COPD. Subsequently, in multivariate analysis, HRs were adjusted for age, sex, cancer stages, histological degree of differentiation, diabetes, chronic bronchitis, asthma, emphysema, CVD, AMI, stroke, CCI score, income level, and urbanization. All analyses were conducted using SAS (Version 9.3; SAS, Cary, NC, USA), and a two-tailed P value of < 0.05 was considered statistically signi cant.

Study cohort
We enrolled 3986 patients with stages IIIA-IIIB, unresectable lung SCC who had received standard de nitive CCRT and did not have distant metastases (Table 1). Among these patients, 1219 had COPD before de nitive CCRT (group 1), and 2767 did not have COPD before de nitive CCRT (group 2). The dosage distributions of radiation and chemotherapy between groups 1 and 2 was homogenous; the median irradiation dose in both groups was 6300 cGy. The median follow-up durations after the index date were 1.96 and 1.09 years for patients without and with COPD, respectively. The two groups differed signi cantly in age, follow-up duration, sex, CCI score, diabetes, asthma, emphysema, CVD, AMI, stroke, income level, and urbanization (Table 1). More patients in the COPD group were old and male; had diabetes, chronic bronchitis, asthma, emphysema, CVD, AMI, stroke, CCI ≥ 2, low income; and lived in rural regions than in the non-COPD group. However, the proportion of clinical stages IIIA and IIIB and histological degree of differentiation were balanced between group 1 and 2. The mortality was 68.12% and 76.95% in the non-COPD and COPD groups, respectively. The lung cancer death was 64.69% and 73.10% in the non-COPD and COPD groups, respectively. Survival curves of COPD and hospitalization for COPDAE in patients with lung SCC receiving de nitive CCRT Figure 1 presents Kaplan-Meier overall survival curves of propensity score weighted population of advanced stage lung SCC with or without COPD before CCRT. Speci cally, the 2-year overall survival rates across all clinical stages were 32.38% and 22.81% in the non-COPD and COPD groups, respectively ( Fig. 1); the overall survival rate was higher in the non-COPD group (log-rank test, P = 0.0472) than in the COPD group. Additionally, the 2-year overall survival rates in patients with lung SCC who were hospitalized for COPDAE 0, 1, and ≥ 2 times within 1 year before CCRT were 30.78%, 19.13%, and 17.93% respectively ( Fig. 2; log-rank test, P < 0.0001).

Discussion
COPD and its severity (the frequency of hospitalization for COPDAE) might play an important role in acute or late radiation-induced lung toxicity. [16][17][18][19] In lung cancer, the addition of chemotherapy to RT might be more toxic for irradiated normal lung tissue due to the risks of RILI. [35][36][37][38][39][40][41] Based on the NCCN guidelines, de nitive CCRT is considered category 1 curative-intent treatment for N2 nodal positive lung cancer.
[15] Owing to the different preferred regimens of squamous and nonsquamous cell carcinoma lung cancer based on the NCCN guidelines, [13][14][15] we only enrolled patients with stages IIIA-IIIB (clinical N2 positive) lung SCC to keep the dose and regimens of CCRT consistent. In our design of consistent clinic stages and pathologic type, the effect of COPD or COPDAE on survival outcomes was clari ed in the patients with stages IIIA-IIIB lung SCC receiving de nitive CCRT, with similar chemotherapy and irradiation dose and volume. This study is the rst to investigate the effects of COPD and COPDAE on the survival outcomes of patients with N2 stage lung SCC receiving de nitive CCRT. Our study showed COPDAE were not only the poor prognostic of all-cause death, but also the strongly poor prognostic factor of COPD death and lung cancer death for advance lung SCC receiving de nitive CCRT (Supplemental Tables 1 and 2). However, the preexisting COPD was the signi cant prognostic factor of all-cause death, instead if lung cancer death in the patients with advanced lung SCC receiving de nitive CCRT.
There were several small sample size, retrospective, inhomogeneous lung cancer types, different stages, and inconsistent treatments studies for the survival impact of COPD in treatments of lung cancers. [42][43][44][45] In the previous study, COPD is a risk factor for postoperative recurrence in non-small-cell lung cancer (NSCLC) patients, and moderate to severe COPD is an independent adverse prognostic factor for recurrence-free survival which is very different from our endpoint (overall survival). [ As shown in Table 1, more patients in the COPD group were male and old; had diabetes, chronic bronchitis, asthma, emphysema, CVD, AMI, stroke CCI ≥ 2, and low income; and resided in rural regions than in the non-COPD group. The distribution of patient characteristics was reasonable in this study and is compatible with that in previous studies, because patients with COPD have more comorbidities. [10,[46][47][48][49] Although matching creates a balanced dataset by making pairs between controls and treated patients on the basis of a similar propensity score, some patients may be excluded from the dataset, which is a major disadvantage. [50,51] In the current study, IPTW (Table 3) was conducted, which has advantages over matching of patients based on propensity scores when there are two groups to compare, when nding matches results in insu cient sample sizes, or when the data are censored. [50,51] Moreover, the data in Table 1 are real-world data; thus, we did not use propensity score matching to wash out too much sample of patients induced a deviation database. [50,51] To create a pseudo-study cohort, where the weighted version can avoid the covariate bias and mimic randomized COPD or non-COPD assignment, we used IPTW adjustment to identify the independent predictors of all-cause death in patients with lung SCC receiving CCRT. [33,34] In addition, we used the multivariate IPTW adjustment model to determine the HRs of all-cause death, COPD death, and lung cancer death in these patients for identifying the independent predictors of all-cause death.
In IPTW-adjusted models, the independent predictors of all-cause death or lung cancer death were similar in patients with lung SCC receiving CCRT (Table 2 and supplemental Table 2). In multivariate analysis through IPTW-adjusted the patients with COPD with stages IIIA-IIIB lung SCC receiving de nitive CCRT had worse survival compared with those without COPD receiving the same curative treatments (Tables 2). However, COPD is not a signi cant prognostic factor of lung cancer death in our study. This study is the rst to show that, among patients with lung SCC receiving de nitive CCRT, those with COPD had a high risk of all-cause death compared those without COPD, instead of a risk factor of lung cancer death.
Another contribution of our study is con rming that the severity of COPD is strongly associated with survival in the patients with lung SCC receiving CCRT. The frequency of hospitalization for COPDAE was signi cantly associated with worse survival in the patients with lung SCC receiving CCRT (Tables 2 and   Fig. 2). According to our literature review, this study is the rst to show that the frequency of hospitalization for COPDAE within 1 year before CCRT (similar to Global Initiative for Chronic Obstructive Lung Disease [GOLD] Classi cation 3-4)[10] is a signi cant risk factor for all-cause death, COPD death, and lung cancer death for the patients with lung SCC receiving de nitive CCRT. The potential mechanism might be severe COPD (GOLD 3-4) associated with severe lung in ammation before CCRT; [9] severe lung in ammation may be induced by COPD, resulting in worse survival after CCRT. According to a previous study, COPD in uences the risk of RP. [18] Therefore, COPD or COPDAE might be associated with worse survival in the patients with stages IIIA-IIIB lung SCC receiving de nitive CCRT. In addition, previous studies have demonstrated chemotherapy exposure, type of anticancer drug, and chemotherapy timing as risk factors for RP. [35][36][37][38][39][40][41] Animal studies and clinical reports have indicated the concomitant administration of RT and chemotherapy as risk factors for RP. [35][36][37][38][39][40][41]52] RILI might be associated with worse survival and poor pulmonary function in COPD and COPDAE patients, which might also contribute to poor survival in patients with lung SCC receiving de nitive CCRT. [10] Moreover, COPD or COPDAE related to poor pulmonary function and CCRT related to RILI might be additive or synergistic risk factors for all-cause death in patients with lung SCC receiving CCRT. Our ndings demonstrated that COPD and the frequency of hospitalization for COPDAE within 1 year before de nitive CCRT were signi cant risk factors for overall survival in the patients with stages IIIA-IIIB lung SCC. For lung cancer death, COPD with relatively better pulmonary function compared with COPDAE were contributed to better tolerance of RILI induced by de nitive CCRT. As a result, COPD is not the prognostic factor of lung cancer death in the patients with advance d lung SCC receiv-ing de nitive CCRT.
Very few studies have investigated the correlation between COPD, COPDAE, low income, and residence in rural regions and de nitive CCRT in patients with stages IIIA-IIIB lung SCC. However, some small, retrospective studies have shown similar outcomes compatible with ours that old age, male sex, moderate to high grade of differentiation, and high CCI scores are independent risk factors for all-cause death for lung cancer with nonhomogeneous stages, pathologic types, or various treatments. [53][54][55] Nevertheless, few studies have shown that COPD, COPDAE, income level, and urbanization are associated with the survival outcomes of patients with stages IIIA-IIIB lung SCC receiving CCRT. Our study revealed that COPDAE before CCRT are important risk factors for all-cause death and demonstrated the moderate to high grade of differentiation, old age, male sex, low income, and residence in rural regions are negative independent predictors of all-cause death and lung cancer death in the patients with lung SCC receiving CCRT (Table 2 and Supplemental Table 2).
Our results suggest that COPD and the frequency of hospitalization for COPDAE within 1 year before CCRT are associated with a high risk of all-cause death in patients with advanced stage (IIIA-IIIB) lung SCC receiving CCRT. The frequency of hospitalization for COPDAE within 1 year before CCRT is associated with a high risk of lung cancer death in patients with advanced stage (IIIA-IIIB) lung SCC receiving CCRT. According to our literature review, no peer-reviewed and large cohort study has investigated patients with COPD, COPDAE, and IIIA-IIIB lung SCC receiving standard CCRT regimens. This is the rst study to examine the effects of COPD and COPDAE on the survival outcomes of patients with stage IIIA-IIIB lung SCC receiving de nitive CCRT. In the future, COPD or hospitalization for COPDAE before de nitive CCRT should be considered as valuable predictors in patients with stage IIIA-IIIB lung SCC receiving de nitive CCRT. COPD and COPDAE are simple, easily measurable, and valuable predictors of all-cause death in patients with stage IIIA-IIIB lung SCC receiving CCRT; they would enable shared decision-making between physicians and patients. In addition, although patients with COPD or COPDAE with stages IIIA-IIIB lung SCC receive de nitive CCRT, radiation oncologists might consider more precision RT techniques, including image-guided radiation therapy, respirator gating, and other respiratory control techniques, for reducing RILI and ensuring low irradiation to normal lung, heart, and esophageal tissues to prevent acute or late radiation-induced pulmonary, cardiac, and esophageal toxicities in these patients. [56,57] This study has some limitations. First, toxicity induced by different treatments could not be determined; therefore, treatment-related mortality estimates may have been biased. Second, because all the patients with lung SCC were enrolled from an Asian population, the corresponding ethnic susceptibility remains unclear; hence, our results should be extrapolated to non-Asian populations with caution. Third, diagnoses of all comorbidities were completely dependent on ICD-10-CM codes. Nevertheless, the Taiwan Cancer Registry Administration randomly reviews charts and interviews patients to verify the accuracy of the diagnoses, and hospitals with outlier chargers or practices may undergo an audit and subsequently receive heavy penalties if malpractice or discrepancies are identi ed. Therefore, for obtaining crucial information concerning population speci city and disease occurrence, a large-scale randomized trial comparing carefully selected patients undergoing suitable treatments is essential. Finally, TCRD does not contain information regarding dietary habits or body mass index, and these factors may be risk factors for mortality. However, considering the magnitude and statistical signi cance of the observed effects in this study, these limitations are unlikely to affect the conclusions. Consent for publication: Not applicable. Patient consent was waived because data les are de-identi ed by scrambling the identi cation codes of both patients and medical facilities and sent to the National Health Research Institutes to form the original les of NHIRD. We used data from the National Health Insurance Research Database (NHIRD). The authors con rm that, for approved reasons, some access restrictions apply to the data underlying the ndings. The data utilized in this study cannot be made available in the manuscript, the supplementary les, or in a public repository due to the "Personal Information Protection Act" executed by Taiwan's government, starting from 2012. Requests for data can be sent as a formal proposal to obtain approval from the ethics review committee of the appropriate governmental department in Taiwan. Speci cally, links regarding contact info for which data requests may be sent to are as follows: http://nhird.nhri.org.tw/en/Data_Subsets.html#S3 and http://nhis.nhri.org.tw/point.html.
Availability of data and material: We used data from the National Health Insurance Research Database (NHIRD). The authors con rm that, for approved reasons, some access restrictions apply to the data underlying the ndings. The data utilized in this study cannot be made available in the manuscript, the supplementary les, or in a public repository due to the "Personal Information Protection Act" executed by Taiwan's government, starting from 2012. Requests for data can be sent as a formal proposal to obtain approval from the ethics review committee of the appropriate governmental department in Taiwan. Speci cally, links regarding contact info for which data requests may be sent to are as follows: http://nhird.nhri.org.tw/en/Data_Subsets.html#S3 and http://nhis.nhri.org.tw/point.html.  Kaplan-Meier overall survival curves of propensity score weighted population of advanced stage lung squamous cell carcinoma with or without chronic obstruction pulmonary disease before concurrent chemoradiotherapy.

Figure 2
Kaplan-Meier overall survival curves of propensity score weighted population of advanced stage lung squa-mous cell carcinoma according to the frequency of hospitalization for chronic obstruction pulmonary disease with acute exacerbation within 1 year before concurrent chemoradiotherapy