In 2022, lung cancer had the highest incidence and mortality among all cancer types worldwide. Small cell lung cancer (SCLC) accounts for approximately 15% of all lung cancer cases globally [1]. As the most aggressive histological subtype, SCLC is characterized by poor survival [1]. Survival rates for this cancer type have improved in recent years, partly because of the introduction of immunotherapy [2–5]. Nevertheless, data from the US Surveillance, Epidemiology, and End Results (SEER) program indicate that the 5-year survival rate for SCLC was only 9% in 2017, compared with 28% for all histological lung cancer types combined [2]. Population-based studies conducted in Europe have also shown substantial improvement in SCLC survival over recent decades [6].

Small cell lung cancer is the histological subtype most strongly associated with tobacco use [7,8]. Consequently, most cases occur among current or former smokers, whereas only approximately 2% occur in individuals who have never smoked [9,10]. Trends in SCLC incidence therefore reflect changes in smoking prevalence. In the coming years, a reduction in new tobacco-attributable cases is expected because of declining smoking prevalence. Recent studies analyzing SCLC incidence in the United States from 2000 to 2019/2020 reported declining incidence rates according to sex and race since 2004 [11,12]. This trend highlights the need to explore additional factors that may influence SCLC occurrence and outcomes. Although previous studies have identified environmental exposures, such as air pollution and residential radon, as potential risk factors, current evidence remains limited [13]. Furthermore, the relatively low frequency of SCLC compared with other lung cancer subtypes means that studies specifically focused on SCLC remain scarce, resulting in incomplete epidemiological characterization.

Population-based studies from several countries have examined clinical characteristics and survival in SCLC, analyzing the impact of different risk factors on survival outcomes, including residential radon exposure and tobacco use [14]. For example, a study conducted in Spain assessing radon exposure did not identify significant differences in 1-year survival [15]. Regarding tobacco use, previous studies have described clinical and sociodemographic differences among patients with SCLC according to smoking status, suggesting that tobacco exposure may influence prognosis [16,17]. In Korea, nationwide registry analyses described prognostic differences according to smoking status and reported distinct characteristics among never smokers with SCLC [18].

Despite these findings, information regarding the epidemiological and clinical profile of SCLC remains limited, particularly in relation to smoking status. Therefore, the objective of this study was to describe SCLC cases diagnosed in Spain between 2016 and 2025, as recorded in the nationally representative Thoracic Tumor Registry, and to examine differences in sociodemographic, clinical, and survival characteristics according to tobacco use. A previous analysis of the Thoracic Tumor Registry of the Grupo Español de Cáncer de Pulmón, including SCLC cases diagnosed up to 2020, provided an initial description of clinical characteristics and outcomes [17]. The present study expands on that work by incorporating an additional 5 years of data (2020–2025), substantially increasing the sample size (from 956 to 4920 cases), and enabling the first estimates of long-term (5-year) survival in the Spanish population.

A total of 4920 SCLC cases diagnosed between 2016 and 2025 were included: 91 never smokers, 1872 former smokers, and 2957 current smokers at diagnosis. Overall, 72.9% of patients were men, and the median age was 66 years (IQR, 59–72 years). Among all patients, 68.2% had extensive-stage SCLC at diagnosis, and 17.4% received immunotherapy as first-line or subsequent-line treatment. Significant differences according to tobacco use were observed by sex, with a higher proportion of women among never smokers, and by age at diagnosis, as current smokers were younger at diagnosis. Comorbidities and symptoms at diagnosis also differed according to tobacco use: former smokers reported fewer comorbidities at diagnosis, whereas current smokers more frequently reported symptoms at diagnosis (Table 1). These patterns remained consistent after stratification according to disease stage at diagnosis (Supplementary Table 1).

Table 2 presents the distribution of symptoms at diagnosis. Overall, cough (36.7%) and dyspnea (36.0%) were the most frequently reported symptoms at diagnosis. Significant differences in symptoms according to tobacco use were observed. Weight loss was more common among current smokers, pain among never smokers, and dyspnea and hemoptysis among former smokers.

The overall 5-year survival rate was 9.4% (95%CI, 8.3–10.5). Fig. 1 shows survival curves according to tobacco use, including corresponding 95%CIs and risk tables. The 5-year survival rate was highest among never smokers at 15.4% (95%CI, 7.0–26.8%), followed by former smokers at 10.8% (95%CI, 8.9–12.9%) and current smokers at 8.2% (95%CI, 6.9–9.7%) (Table 3). Fig. 2 shows survival curves stratified according to stage at diagnosis, demonstrating significantly better survival among patients diagnosed with limited-stage SCLC compared with extensive-stage SCLC (Table 3).

Fig. 1.

Survival at 5 years after diagnosis according to tobacco use.

Fig. 2.

Survival at 5 years after diagnosis according to stage at diagnosis.

Analysis according to sex and tobacco use revealed that 5-year survival was higher in women than in men regardless of smoking status. Among current smokers, 5-year survival was 10.7% in women compared with 7.1% in men; among former smokers, it was 17.5% in women and 9.3% in men; and among never smokers, it was 20.7% in women and 0% in men (Supplementary Table 2). Survival according to sex and stage at diagnosis is shown in Supplementary Table 3. Overall, women with limited-stage SCLC showed better survival rates than men, whereas both sexes showed poorer survival in extensive-stage SCLC. In both stages, survival was higher among never smokers, particularly during the first years after diagnosis (Supplementary Table 4).

Cox regression analysis indicated that sex, tobacco use, age at diagnosis, and stage at diagnosis influenced mortality risk. Specifically, mortality risk was higher in men than in women (HR, 1.10; 95%CI, 1.01–1.20), in current smokers compared with never smokers (HR, 1.30; 95%CI, 1.01–1.68), and in patients diagnosed with extensive-stage SCLC compared with limited-stage SCLC (HR, 2.96; 95%CI, 2.72–3.22). Regarding age at diagnosis, the HR increased by 3% for each additional year of age (HR, 1.03; 95%CI, 1.02–1.03).

The findings of this study highlight several important issues related to the characteristics of SCLC at diagnosis and the impact of tobacco use on patient survival. First, in our cohort, never smokers accounted for 1.8% of all SCLC cases. This finding is consistent with previous studies [9,10,17], although recent clinical trials have reported a higher proportion of never smokers (7%–8%) [21,22]. Second, current smokers were diagnosed at a younger age than never smokers, with a notable difference of 8 years regardless of SCLC stage at diagnosis. Current and former smokers also had more symptoms and comorbidities at diagnosis than never smokers. Third, survival was significantly higher among never smokers, approximately double that observed among current smokers at 5 years after diagnosis. These differences persisted after stratification according to stage at diagnosis and sex, although the survival gap narrowed when comparing male and female current smokers. To our knowledge, this is the largest European study to date analyzing SCLC patient characteristics and survival.

The profiles of patients diagnosed with SCLC differed according to tobacco use. These findings are consistent with those reported by Uprety et al. [12]. Current smokers were diagnosed at a younger age than never smokers, suggesting a shorter induction period among smokers. This observation is consistent with several previous studies, including those by Tseng et al. [16] and Thomas et al. [23], who also observed that never smokers were older at diagnosis. Several factors may explain these findings. First, previous studies have shown that never smokers have a lower tumor mutational burden and fewer tobacco-specific mutations, because some mutations are directly associated with tobacco exposure [23]. Tobacco use may therefore induce earlier and more frequent mutations, leading to cancer development at younger ages. Second, diagnosis may occur earlier in current or former smokers because of the strong association between tobacco use and SCLC. When similar symptoms are present, clinicians may be more likely to suspect SCLC in smokers than in never smokers. A previous study using TTR data also reported longer diagnostic delays among never smokers than among smokers [24].

Another notable finding was that most patients were diagnosed with SCLC at an advanced stage, particularly among never smokers. According to SEER data, between 2013 and 2022, 70.9% of all SCLC cases were diagnosed at an advanced stage, substantially worsening prognosis [9]. Early diagnosis of SCLC remains challenging in the context of lung cancer screening. On the one hand, never smokers are generally not eligible for screening because they do not meet current criteria and are therefore not detectable through screening programs. In addition, screening sensitivity for early-stage SCLC is lower than that for other histological subtypes (8.8% for SCLC vs 56.6% for adenocarcinoma) [25]. This low sensitivity is attributable to the rapid progression and aggressiveness of SCLC. In the NLST study, 86% of screen-detected SCLC cases were diagnosed at stage III or IV disease [26], and screening provided no survival benefit compared with unscreened individuals [27]. Therefore, new strategies for early SCLC detection in both smokers and never smokers are urgently needed.

Current smokers with SCLC showed a higher prevalence of comorbidities and clinical symptoms at diagnosis compared with never smokers, regardless of disease stage. Current smokers most frequently presented with weight loss, dyspnea, and hemoptysis, whereas anorexia and pain were more common among never smokers. As mentioned above, smokers have a higher mutational burden, favoring accelerated tumor progression and more severe symptoms. Regarding comorbidities, a study conducted in the Netherlands among patients diagnosed with SCLC reported an increase in the prevalence of comorbidities between 1995 and 2012, particularly those associated with tobacco use, such as cardiovascular and pulmonary diseases [28].

Furthermore, our study observed differences in survival according to tobacco use and sex, particularly from the third year after diagnosis onward. Women consistently exhibited better prognosis than men regardless of smoking status, although these differences were attenuated among current smokers. Never smokers had a 5-year survival rate of 15.4%, whereas current smokers had a survival rate of 8.2%. Previous studies have reported conflicting findings. Tseng et al. [16] analyzed survival according to tobacco use using data from the Taiwan Cancer Registry and found that never smokers, particularly men, had lower survival rates. In contrast, studies by Sun et al. [29] in Korea and Liu et al. [30] in China observed, similarly to our findings, that never smokers had better survival and prognosis than smokers. Franco et al. [17], also using TTR data through 2020, reported a 2-year survival rate of 20.7% among never/former smokers and 11.5% among smokers. Another study conducted in the United States by Thomas et al. [26] did not identify survival differences according to tobacco use.

Differences among studies regarding survival outcomes may be related to stage at diagnosis. In our study, stage at diagnosis strongly influenced survival. At 5 years, survival for limited-stage SCLC was 23.3%, whereas survival for extensive-stage SCLC was only 2.6%. These results are consistent with previous studies [29]. Survival differences according to smoking status should be interpreted cautiously because smoking categories were not independent of sex distribution in our cohort, with a higher proportion of women among never smokers. Because women consistently showed better survival across smoking strata, residual confounding according to sex cannot be excluded. Overall, although survival remains poor in extensive-stage SCLC, outcomes were more favorable among women and never smokers.

This study has several strengths. First, the sample of patients with SCLC is representative of the Spanish population, as demonstrated in previous studies, and the sample size was large. In addition, the cohort included a substantial proportion of women (N=1333), which is notable because SCLC has traditionally been more strongly associated with men.

However, several limitations should be acknowledged. First, this study was primarily descriptive and was not designed to assess causal inference. The absence of detailed information regarding treatment modalities, molecular profiling, and temporal treatment patterns limited our ability to explore underlying mechanisms or evaluate survival changes over time associated with treatment. Treatments received were not specifically assessed, although there was no significant heterogeneity in therapeutic approaches, and treatment strategies were presumed to align with disease stage at diagnosis.

The introduction of immunotherapy into first-line therapy for SCLC has led to modest but statistically significant improvements in overall survival. For example, the IMpower133 trial demonstrated that the addition of atezolizumab to carboplatin-etoposide improved median overall survival compared with chemotherapy alone (12.3 vs 10.3 months; HR, 0.70; 95%CI, 0.54–0.91). Similarly, the CASPIAN trial demonstrated improved overall survival with durvalumab plus platinum-etoposide compared with platinum-etoposide alone (13.0 vs 10.3 months; HR, 0.73; 95%CI, 0.59–0.91). Although some patients in our study received immunotherapy, limited follow-up time after diagnosis precluded assessment of survival beyond 2 years. However, survival models were adjusted for variables strongly associated with treatment, such as disease stage at diagnosis.

Another limitation, related to the previous issue, is that 5-year survival could only be estimated for a subset of the cohort because follow-up data were available for only 180 patients due to loss to follow-up, death within 5 years after diagnosis, or recent diagnosis (after 2020). This limitation may reduce the precision of 5-year survival estimates, although annual survival after diagnosis was estimated. Additionally, molecular data were limited because only a small number of patients underwent genetic tumor analysis; only 1 patient had a TP53 mutation, and no RB1 mutations were recorded. These limitations should be considered when interpreting the findings and underscore the need for future studies incorporating detailed treatment, temporal, and molecular data to better understand survival determinants in SCLC.

In conclusions, this study demonstrates that tobacco use is strongly associated with SCLC. Current and former smokers are diagnosed with SCLC at younger ages and have survival rates approximately half those observed among never smokers. These findings have important public health implications because, although smoking prevalence has declined in some countries, it remains high in many regions worldwide, particularly in Europe, including Spain, and in countries such as China and India. Furthermore, differences in survival between men and women suggest the need to investigate potential molecular or biological mechanisms underlying these disparities, as well as possible differences in susceptibility to the harmful effects of tobacco exposure.

ARR: Conceptualization, methodology, supervision, writing – review and editing. MP: Conceptualization, methodology, supervision, writing – review and editing. CCP: Methodology, data curation, formal analysis, visualization, writing – original draft. The remaining authors: Investigation, resources, writing – review and editing.

The Thoracic Tumors Registry was registered in ClinicalTrials.gov (NCT02942458), and the study protocol was approved by the institutional committee of Hospital Universitario Puerta de Hierro-Majadahonda (Madrid, Spain) (No. PI 148/15).

No artificial intelligence tools were used in the conception, design, data collection, data analysis, data interpretation, or writing of the manuscript.

None declared.

Virginia Calvo de Juan has received consulting fees from Roche, AstraZeneca, MSD, BMS, Takeda, Regeneron, Amgen, GSK, Boehringer-Ingelheim, J&J, and BeiGene; speaker fees from Roche, AstraZeneca, MSD, BMS, Takeda, Regeneron, Amgen, Pfizer, J&J, and BeiGene; and support for meeting attendance and/or travel from AstraZeneca, Roche, MSD, Takeda, and Janssen. Manuel Cobo has served as a consultant or advisory board member for Novartis, AstraZeneca, Boehringer-Ingelheim, Roche, BMS, Lilly, MSD, Takeda, Pfizer, Kyowa, Sanofi, and Janssen; and has received fees from Novartis, AstraZeneca, Boehringer-Ingelheim, Roche, BMS, Lilly, MSD, Takeda, Kyowa, Pierre Fabre, Novocure, Sanofi, Janssen, and Daiichi. Ainhoa Hernández has received support for attending conferences from Roche, Sanofi, and MSD, as well as fees from Pierre Fabre and Roche. Rafael López Castro has received fees from AstraZeneca, Pierre Fabre, Roche, Pfizer, and Takeda, and support for attending meetings and/or travel from MSD, Takeda, Roche, and AstraZeneca. Rosario García Campelo has received consulting fees from Roche, Novartis, BMS, MSD, AstraZeneca, Takeda, Janssen, GSK, Amgen, Sanofi, Boehringer, Lilly, Johnson & Johnson, and Nuvalent; speaker fees from Roche, Novartis, BMS, MSD, AstraZeneca, Takeda, Janssen, GSK, Amgen, Sanofi, Boehringer, Lilly, and Johnson & Johnson; and support for meeting attendance and/or travel from Roche, MSD, and BMS. Joaquim Bosch-Barrera has received speaking fees from Regeneron, AstraZeneca, BMS, Takeda, Roche, Merck, and Pfizer; support for meeting attendance and/or travel from MSD, Roche, and AstraZeneca; and has participated in advisory boards for MSD and Roche. Manuel Dómine has received fees and consulting fees from AstraZeneca, BMS, MSD Oncology, Pfizer, Roche, and Takeda, as well as support for meeting attendance and travel from AstraZeneca, MSD Oncology, Pfizer, and Takeda. Mariano Provencio has received research funding from BMS, AstraZeneca, MSD, Roche, Takeda, Pfizer, Boehringer-Ingelheim, Amgen, Instituto de Salud Carlos III, Spanish Ministry of Science and Innovation, European Commission, Eli Lilly, F. Hoffmann-La Roche, Janssen, and Pierre Fabre. He has received consulting fees from BMS, AstraZeneca, MSD, Roche, Takeda, Eli Lilly, F. Hoffmann-La Roche, Janssen, Pfizer, and Amgen; speaker fees from BMS, AstraZeneca, MSD, Roche, Takeda, Eli Lilly, F. Hoffmann-La Roche, Janssen, Pfizer, and Amgen; support for meeting attendance and travel from BMS, AstraZeneca, MSD, Roche, Takeda, Eli Lilly, F. Hoffmann-La Roche, Janssen, Pfizer, Amgen, Boehringer-Ingelheim, and Pierre Fabre; and has participated in advisory boards for Amgen, Pfizer, Daiichi Sankyo, Johnson & Johnson, BMS, Takeda, AstraZeneca, Gilead, MSD, Guardant Health, Ipsen, Incyte Biosciences, Bayer, Pharmacosmos, Janssen, Astellas Pharma, and Aegean Pharmaceuticals. He is also a member of the Board of Directors of the Instituto de Investigación Sanitaria Puerta de Hierro-Segovia de Arana, the Grupo Español de Cáncer de Pulmón, and the Grupo Español de Linfomas. The remaining authors declared no conflicts of interest whatsoever.