Clinical characteristics and survival analysis of non-smoking small cell lung cancer.

Introduction
Among malignant tumors, lung cancer is the leading cause of death (1) and a serious threat to global public health. Smoking is the most significant risk factor for lung cancer. Over the past few decades, the smoking rate has begun to decline. The proportion of lung cancer cases among non-smokers is gradually increasing. Overall, 15–20% of lung cancer cases are non-smokers (2,3). People now understand that lung cancer in smokers and non-smokers is different. It is recommended that non-smoking lung cancer should be investigated as a separate disease. Although the majority of lung cancers in non-smokers are adenocarcinomas, small cell lung cancer (SCLC) also accounts for approximately 8% of cases (4).
About 15% of all lung cancers are SCLC, a highly aggressive subtype with a dismal prognosis (5). It is generally believed that SCLC is closely related to the use of tobacco. However, non-smokers can also develop SCLC. The incidence of SCLC in non-smokers may be due to the use of second-hand tobacco, indoor radon, work-related fumes, and other factors (6-8). Patients with non-smoking SCLC were extremely rare in the past (9). However, recent global clinical studies on SCLC indicate that the proportion of non-smokers has increased (10-12). The proportion of non-smokers with SCLC varies among different regions and ethnic groups. Non-smoking SCLC in Caucasian, non-Asian populations is rare, accounting for only 1.8–3% (13,14), whereas non-smoking SCLC among Asians is higher, ranging from 13% to 26% (15-17). Therefore, research on non-smoking SCLC has become significantly important. Characteristics and prognosis of patients with non-smoking SCLC have been examined in previous studies. However, the results are inconsistent due to the small sample sizes (14,18). Moreover, the landscape for SCLC treatment is evolving with the continuous emergence of new drugs. However, the effect of these new treatment strategies on the survival of patients with non-smoking SCLC lacks relevant research. Therefore, we conducted a retrospective analysis of the clinical characteristics, prognosis, and the impact of treatment changes on survival of patients with non-smoking SCLC treated at our center over 12 years, from 2013 to 2025. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2594/rc).

Methods

Patient selection and data collection
We conducted a retrospective study of patients diagnosed with non-smoking SCLC at Jilin Cancer Hospital from January 1, 2013, to March 31, 2025. The main inclusion criteria were as follows: (I) SCLC must be diagnosed by histology or cytology; (II) non-smokers who have smoked fewer than 100 cigarettes in their lifetime. The primary exclusion criteria were as follows: (I) a history of other malignancies within the past five years; (II) transformed SCLC; (III) patients who have smoked in the past or are currently smoking; (IV) patients whose smoking history is unavailable. The electronic medical records provided us with the demographic characteristics of patients and the relevant clinical information. Additionally, survival data were gathered from eSuizhen (version 2.7.1) and the Smart Management Platform of Jilin Cancer Hospital. This study was approved by the Ethics Committee of Jilin Cancer Hospital (No. 202508-030-01). Informed consent was not required, as this study is retrospective. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Statistical analysis
Statistical Package for the Social Sciences software (version 26.0) was used to perform all statistical analyses. Descriptive statistics were used to summarize the baseline characteristics of the population of the study. The distribution of continuous variables was used to express them as median (range). Categorical variables are presented as numbers and percentages (n, %). Overall survival (OS), defined as the interval between the date of diagnosis and the date of death from any cause or the last follow-up, was used as the primary endpoint of this study. Progression-free survival (PFS) is the period from the initiation of first-line systemic treatment until the first documented disease progression (according to Response Evaluation Criteria in Solid Tumors [version 1.1; (RECIST 1.1)] or death from any cause. PFS was censored at the date of the last evaluable imaging follow-up for patients who were still alive and without progression at the data cutoff date. The Kaplan-Meier method was used for survival analysis, and survival curves were graphically represented for key patient subgroups. Differences in survival distributions between these groups were compared using the log-rank test. The Cox proportional hazards model was employed to assess the effects of various factors on both OS and PFS. Using the enter method, all selected variables were simultaneously added into the multivariable model without any automatic selection process. Hazard ratios (HRs) and their corresponding 95% confidence intervals (CIs) were calculated from this model. A two-sided P value of less than 0.05 was considered statistically significant.

Results

Patients selection and inclusion
We identified 7,651 patients with a histological or cytological diagnosis of SCLC after retrospectively reviewing the case system of Jilin Cancer Hospital. The following exclusion criteria were applied: (I) history of smoking (n=5,304), (II) concurrent other malignancies (n=89), (III) patients with SCLC transformed from other malignancies (n=43), and (IV) inability to obtain accurate smoking history (n=608). The patients were diagnosed between January 1, 2013, and March 31, 2025, with the follow-up cutoff date being September 30, 2025. Figure S1 provides a detailed flowchart of the patient selection process. For the pre-2019 cohort, the median follow-up period was 70.17 months (95% CI: 58.90–81.44 months); for the post-2019 cohort, it was 26.50 months (95% CI: 24.04–28.96 months); and for the entire population, it was 37.13 months (95% CI: 33.80–40.46 months).

Patients baseline characteristics
Overall, 1,607 (21%) non-smoking SCLC patients were included in this study. Of them, 980 patients (61.0%) had limited-stage SCLC (LS-SCLC), and 627 patients (39.0%) had extensive-stage SCLC (ES-SCLC). The median age was 59 years (range, 20–88 years), with a female predominance (58.6%). The majority of patients (81.1%) had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0–1. As for pathological subtype, pure SCLC constituted the majority (98.7%), while combined SCLC occurred in 21 cases (1.3%). Among patients with LS-SCLC, 65.5% received thoracic radiotherapy, 14.7% received prophylactic cranial irradiation (PCI), and 7.7% were treated with surgery; while in the ES-SCLC cohort, thoracic radiotherapy and PCI were performed in 22.5% and 3.3% of patients, respectively (Table 1). The median first-line PFS was 9.03 months (95% CI: 8.56–9.51 months) (Figure S2A), and the median OS was 25.47 months (95% CI: 23.99–26.94 months) (Figure S2B).

Clinical characteristics and prognosis of patients with LS-SCLC

Patients baseline characteristics and OS
Overall, 980 patients with LS-SCLC were included in this study. The median first-line PFS was 11.20 months (95% CI: 10.52–11.88 months) (Figure 1A), and the median OS was 34.27 months (95% CI: 30.17–38.37 months) (Figure 1B) at the time of the last follow-up.

Impact of local treatment modalities on prognosis
The intensity of local consolidative therapy has a significant association with survival outcomes. Treatment with thoracic radiotherapy was significantly linked to prolonged first-line PFS and OS (PFS: 11.7 versus 9.5 months, P=0.002; OS: 40.3 versus 23.1 months, P<0.001) (Figure 2A,2B). Additionally, PCI provided a significant survival benefit. Patients who received PCI had a median first-line PFS of 17.0 months and a median OS of 57.3 months, both of which were significantly improved compared to those who did not receive PCI (median PFS without PCI: 9.9 months; median OS without PCI: 28.3 months; all P<0.001) (Figure 2C,2D). Furthermore, 75 patients (7.7%) who underwent surgical resection demonstrated an excellent prognosis, with a median PFS of 20.0 months (95% CI: 14.40–26.9 months). At the time of this analysis, the median OS had not been reached (Figure S3A,S3B).
Early stage, thoracic radiotherapy, surgical intervention for SCLC, and PCI were associated with prolonged PFS according to a multivariate Cox regression analysis, which included age, gender, eastern cooperative oncology group performance status (ECOG PS) score, tumor, node, metastasis (TNM) stage, surgery for SCLC, time of diagnosis, and treatment regimen. Prolonged OS was linked to female gender, administration of thoracic radiotherapy, surgical intervention for SCLC, PCI, diagnosis after 2019, and treatment with immune checkpoint inhibitors combined with chemotherapy (Table S1).

Clinical characteristics and prognosis of patients with ES-SCLC

Patients baseline characteristics and OS
At the time of the last follow-up, the median first-line PFS was 6.67 months (95% CI: 6.29–7.04 months), and the median OS was 18.37 months (95% CI: 16.76–19.78 months) for the 627 patients with extensive-stage disease (Figure 3A,3B).

Efficacy comparison of first-line systemic treatment regimens
The median PFS for patients with ES-SCLC was 6.40 months (95% CI: 5.90–6.90 months), with chemotherapy alone, 7.00 months (95% CI: 6.24–7.76 months), with immunotherapy plus chemotherapy, 8.33 months (95% CI: 6.87–9.80 months) with antiangiogenic therapy plus chemotherapy, and 13.57 months (95% CI: 8.73–18.41 months) with immunotherapy plus antiangiogenic therapy and chemotherapy. Compared to chemotherapy alone, none of the combination regimens produced a statistically significant extension of PFS (P=0.11, 0.09, and 0.09, respectively).
In contrast, these first-line treatment groups displayed significant difference in OS. The median OS was 15.83 months (95% CI: 13.86–17.81 months) with chemotherapy alone, 23.93 months (95% CI: 16.32–31.55 months) with immunotherapy plus chemotherapy, 25.70 months (95% CI: 17.14–34.26 months) with antiangiogenic therapy plus chemotherapy, and 29.13 months (95% CI: 13.48–44.78 months) with immunotherapy plus antiangiogenic therapy and chemotherapy. Only immunotherapy plus chemotherapy significantly extended OS compared to chemotherapy alone (P<0.001). Statistically non-significant OS benefits were produced by the other combination regimens (P=0.07 and 0.11, respectively) (Figure 4A,4B).

Role of consolidative radiotherapy
Thoracic radiotherapy was associated with a significantly superior OS in ES-SCLC; the median OS in patients who received radiotherapy was 25.87 months, significantly better than the 16.80 months in those who did not (P=0.001) (Figure S4A). Additionally, PCI provided a significant survival benefit. Patients receiving PCI achieved a median OS of 31.47 months, which was significantly superior compared to that of patients who did not receive PCI (median OS without PCI: 17.53 months, P=0.02) (Figure S4B).
Thoracic radiotherapy, PCI, diagnosis after 2019, treatment with immunotherapy combined with chemotherapy, chemotherapy combined with antiangiogenic therapy, and chemotherapy combined with both immunomodulatory and antiangiogenic therapy were associated with prolonged PFS, as reveled by multivariate Cox regression analysis, that included age, gender, ECOG PS score, TNM stage, thoracic radiotherapy, surgery for SCLC, time of diagnosis, and treatment regimen. Additionally, prolonged OS was associated with age <65 years, ECOG PS score 0–1, thoracic radiotherapy, PCI, immunotherapy combined with chemotherapy (Table S2).

Temporal trend analysis
Significant difference in survival outcomes among patients with LS-SCLC were revealed by a temporal trend analysis between the two periods, stratified by the year 2019 (P<0.001). Due to limited follow-up time, the median OS for patients admitted after 2019 had not been reached; however, their survival curve demonstrated significant improvement compared to that of the pre-2019 cohort (median OS in the pre-2019 cohort: 29.30 months, 95% CI: 25.22–33.38 months) (Figure 5A).
With the median OS increasing from 15.97 months in the pre-2019 period to 19.17 months (P=0.004), temporal trend analysis indicated that the survival prognosis of patients with ES-SCLC significantly improved after 2019 (Figure 5B).

Molecular pathological characteristics
Eight patients with non-smoking SCLC underwent genetic testing. The results revealed that three cases with an adenocarcinoma component had EGFR exon 19 deletion mutations. Among the tested patients with pure SCLC, no clearly actionable targetable driver gene mutations were identified.

Discussion
To our knowledge, this study represents the most comprehensive examination of the clinical characteristics, treatment options, and prognosis of non-smoking patients with SCLC. We reviewed 7,651 cases of SCLC in our center; among them, 1,607 cases (21%) were non-smokers. The incidence rate of non-smoking SCLC in the current study was higher than that reported from non-Asian populations but was similar to that reported from Asian populations, including those from China. This suggests that non-smoking SCLC is not rare. It also suggests that it is crucial to examine the clinical pathological characteristics and prognosis of non-smoking SCLC as a separate subgroup.
Another feature of our study is that the patients included spanned a period of 12 years. More importantly, the patients covered both the periods before and after the transformation of standard treatment for SCLC. Atezolizumab combined with chemotherapy as the first-line treatment for ES-SCLC was approved by the Food and Drug Administration of the United States (FDA) in 2019 (19). Besides, in 2019, anlotinib was approved by the National Medical Products Administration (NMPA) for patients with SCLC who have received at least two previous treatment regimens (20). Subsequently, lurbinectedin was approved by the FDA (21,22) and the NMPA for the second-line treatment of SCLC in 2020 and 2024, respectively. Furthermore, for first-line treatment of ES-SCLC, the NMPA has approved multiple programmed death-1 or programmed death-ligand 1 inhibitors in combination with chemotherapy since 2023 (15,16,23,24). These approved drugs have transformed the treatment landscape for SCLC, improving the survival rates of patients with the disease. Consequently, we chose 2019 as the dividing point and analyzed OS. Although the median OS for LS-SCLC diagnosed after 2019 has not been reached, compared with SCLC diagnosed before 2019, the OS of patients diagnosed after 2019 has significantly increased, regardless of whether it is in the limited stage or the extensive stage. Our study revealed that the advancement of treatment provided a significant survival benefit for patients with non-smoking SCLC. Certainly, these comparisons are observational and are influenced by confounding factors such as ECOG PS, PCI, and chest radiotherapy.
According to our research, patients with non-smoking ES-SCLC are younger at diagnosis, and women are more common among them. Most of these patients have LS-SCLC. These findings align with previous studies in some aspects, while contradicting others. In our study, the median age of patients with non-smoking SCLC was 59 years, which was similar to studies from China (60 years) (18) and was younger than the studies in Korean (68–74 years) (25,26) or Caucasian populations (66–75 years) (8,27). The proportion of female patients is higher in non-smoking SCLC than in smoking SCLC, according to previous research. However, this proportion varies widely between studies, ranging from 31.9% to 87%. Overall, the proportion of females (31.9–80.2%) in studies from Asian populations was lower compared to that of Caucasians in non-smoking SCLC (70–87%) (8,18,25-28). In this study, 58.6% of patients with non-smoking SCLC were female, a figure that is similar to those reported from other retrospective studies from China (18) and other Asian populations. However, compared to studies conducted in South Korea (80.2%), the proportion of women in our study was significantly lower (26). This might be because both of these studies were retrospective and conducted in a single center, which could lead to selection bias. Additionally, it is impossible to rule out the regional factors. Therefore, further research is required to determine the proportion of female patients among non-smoking SCLC cases, including data from global and regional areas. The differences in the proportion of cases among women resulting in different treatment options remain unclear. Besides, the proportion of LS-SCLC in our study exceeded 60%. The proportion of LS-SCLC among non-smoking SCLC in studies from Korea and Spain was approximately 50% (8,26). ES-SCLC was higher in non-smoking SCLC in most studies (14,18,25,27,28). Particularly, from Taiwan, an analysis of 858 patients with non-smoking SCLC revealed that stage IV has been reached in 76.9% of patients (28). The differences in these research results suggest that the characteristics of non-smoking SCLC may vary among various races and regions apart from sample size. Moreover, environmental factors and work-related fumes in different regions might also contribute to the inconsistent results.
There is considerable concern over whether the prognosis for non-smoking SCLC differs from that of smoking SCLC. In an analysis of data from the Taiwan Cancer Registry, the median OS of patients with non-smoking SCLC and patients with smoking SCLC were 6.0 and 8.1 months, respectively. The two groups differed significantly (28). There was no difference in OS between non-smokers SCLC and smokers SCLC (9.0 versus 11.0 months, P=0.127) according to another analysis based on data from 280 cancer clinics representing 2.2 million American patients with cancer (14). Our study only assessed the survival outcomes of non-smoking SCLC. The prognostic differences between non-smoking SCLC and smoking SCLC are still unclear without concurrent data from smoking populations for comparison. The current results again suggest the need for more research on non-smoking SCLC, particularly in analyzing the differences between smoking and non-smoking SCLC at the levels of genome, transcriptome, and tumor microenvironment, as well as their impact on survival and implications for treatment choices.
In a retrospective study from Korea, objective response rate (ORR) of patients with non-smoking SCLC was higher than that of smoking patients with SCLC. However, OS did not differ between non-smokers and smokers (25). Frequently, ORR depicts no correlation with PFS and OS in SCLC. In the IMpower133 study, for instance, ORR of atezolizumab combination with chemotherapy was 60.2%, whereas ORR of the chemotherapy group was 64.4%. Compared to the chemotherapy group, PFS and OS of the atezolizumab combination with chemotherapy group were significantly longer (19). Similarly, ORR values of liposomal Irinotecan and topotecan were 44.1% and 21.6%, respectively, in a phase III study comparing their efficacy in recurrent SCLC. However, PFS and OS did not differ between the two groups (PFS: 4.0 versus 3.3 m, P=0.71, OS: 7.9 versus 8.3 m, P=0.31) (29). Given these results, we did not collect ORR data for patients with non-smoking SCLC, and therefore did not assess the correlation between ORR and OS in our study. We hypothesize that non-smoking and smoking SCLC patients have highly proliferative diseases that are responsive to initial treatment. However, recurrence and resistance are unavoidable, and the high ORR cannot result in OS benefits due to the lack of more effective treatment options after resistance. The key to resolving this issue is developing effective treatments targeting the resistance mechanism of SCLC. Additionally, larger sample size research is required to determine whether ORR of non-smoking SCLC is higher than that of smoking SCLC.
The treatment of non-smoking SCLC is an important clinical concern because it is a distinct subtype of lung cancer. Our research has revealed that local treatment and regimens containing immune checkpoint inhibitors help patients with non-smoking SCLC to survive. It should be noted that the combination of immunotherapy and chemotherapy as the standard treatment for ES-SCLC is also applicable to non-smoking patients. The advantages of local treatment suggest that multidisciplinary assistance will lead to better survival for patients with non-smoking SCLC. However, patients who received antibody drug conjugates, bispecific antibodies, or T-cell engagers were not included in our study. Thus, the efficacy of these new drugs in non-smoking SCLC patients remains unclear. The optimal treatment in patients with non-smoking SCLC can be determined only through molecular-level analysis.
Our research has several limitations. First, our research was a single-center retrospective study. Second, we are unable to gather precise information about exposure to second-hand smoke, radon gas in the home, and occupational factors that may be related to the occurrence of lung cancer, since we only extract data from electronic medical records. Particularly, SCLC exposed to second-hand smoke remains a smoking-related lung cancer. It may have a similar pathogenesis to smoking-induced SCLC and shares similar biological characteristics. Selection bias could result from the lack of this data. Third, we did not collect data for comparison from patients with smoking-related SCLC. Without a data comparison with the smoking population simultaneously, we cannot clearly determine the prognostic differences between non-smoking SCLC and smoking SCLC. The baseline characteristics and survival, as well as the genomic and transcriptomic profiles of smoking and non-smoking SCLC, will be analyzed in the future. Fourth, a molecular-level assessment of non-smoking SCLC was impossible to perform because only eight patients underwent genetic testing in our study, and the methods of genetic testing and genetic panels were different. Lastly, currently, there is no unified standard for defining smoking status. The definition by the Centers for Disease Control and Prevention and other research institutions (3,8,18) was used in our study, and it was not defined through biochemical indicators. This can also lead to selection bias.

Conclusions
Non-smoking SCLC is an important subgroup of SCLC with unique clinical and pathological characteristics. These characteristics include a younger age at diagnosis, a higher proportion of females, and a higher incidence of LS-SCLC. The current treatment approach is also applicable to non-smoking SCLC. Although the prognosis of patients with non-smoking SCLC has improved due to the advancements in treatment, a comprehensive molecular analysis of non-smoking SCLC in the future that reveals the mechanisms of its occurrence and development will enable the establishment of precise treatment strategies and significantly improve the OS of patients with non-smoking SCLC.

Supplementary
The article’s supplementary files as