Development and validation of a nomogram for predicting overall survival in small cell lung cancer: a multicenter analysis integrating multimodal therapy and molecular pathology

Introduction

Small cell lung cancer (SCLC) is a highly aggressive neuroendocrine subtype of lung cancer, accounting for approximately 15–20% of all lung malignancies. It is characterized by rapid proliferation, early dissemination, and high metastatic potential, resulting in dismal survival outcomes (1). At diagnosis, more than 70% of patients present with extensive-stage disease accompanied by distant metastases (2). The current standard treatment consists of etoposide combined with platinum-based chemotherapy (CT), often in conjunction with radiotherapy. Although this regimen achieves high initial response rates, relapse and chemoresistance typically develop within one year, severely limiting long-term survival (3).

In recent years, the emergence of immune checkpoint inhibitors has provided new therapeutic opportunities for SCLC. Agents targeting programmed cell death protein 1 (PD-1) or programmed death-ligand 1 (PD-L1), such as atezolizumab and durvalumab, have demonstrated modest but clinically meaningful improvements in overall survival (OS) among patients with extensive-stage SCLC (ES-SCLC) in pivotal trials (4). Recent evidence suggests that neoadjuvant immunochemotherapy may be a feasible treatment strategy for selected patients with potentially resectable SCLC, highlighting the growing interest in multimodal and individualized management in this disease (5). Nevertheless, despite these advances, challenges persist in optimizing immunotherapy for SCLC—particularly in identifying predictive biomarkers, refining combination strategies, and tailoring treatment to the disease’s unique molecular profile.

Over the past decade, several prognostic models and nomograms for SCLC have been proposed. These models have been derived from single-center cohorts, population-based databases, and, more recently, inflammatory or hematologic marker-based analyses (6-8). While some of these models have shown better predictive performance than conventional staging alone, important limitations remain. Many were developed in the pre-immunotherapy era, some were restricted to selected patient subsets, and many relied predominantly on anatomical stage or systemic inflammatory indices without accounting for differences in treatment exposure or tumor pathological heterogeneity. As a result, their applicability to contemporary real-world SCLC populations may be limited.

Traditional staging systems in SCLC are limited by their stage-centered and anatomy-based nature, which may be insufficient for individualized prognostic assessment. tumor, node, metastasis (TNM) and Veterans Administration Lung Study Group (VALG) staging are essential for disease classification and treatment planning, but they mainly describe tumor burden and disease spread rather than differences in treatment modality, biological phenotype, or molecular-pathological characteristics. In the immunotherapy era, patients with similar anatomical stages may still have substantially different outcomes under different multimodal treatment strategies. Therefore, prognostic assessment in SCLC should extend beyond stage alone toward a more comprehensive framework.

Advances in molecular pathology have introduced biomarkers such as Ki-67, chromogranin A (CgA), synaptophysin II (Syn2) (the term used in the original pathology reports, corresponding to synaptophysin staining), and cluster of differentiation 56 (CD56), which reflect neuroendocrine differentiation and proliferative activity in SCLC (9). However, the reported prognostic value of these markers remains inconsistent across studies. For example, Pei et al. found that Ki-67 expression was not significantly associated with survival in ES-SCLC (10), whereas Zhang et al. reported poorer prognosis in patients with high Ki-67 expression (11). Similarly, Gkika et al. observed that synaptophysin expression was associated with better OS (12), whereas He et al. reported that positive Syn expression was associated with worse progression-free survival (PFS) and OS and that CD56 positivity correlated with shorter OS in pulmonary high-grade neuroendocrine carcinomas (13). These findings suggest that individual pathological markers may capture only one dimension of tumor biology and may show unstable prognostic associations across cohorts. Therefore, aggregating several routinely assessed pathological markers into a composite score may provide a more robust summary of neuroendocrine differentiation and proliferative phenotype than any single marker alone.

In this study, we conducted a multicenter real-world study to develop a nomogram-based prognostic model for patients with SCLC by comprehensively incorporating clinical features, treatment modalities—including CT, radiotherapy, and immunotherapy—and pathological biomarkers. We aimed to establish a practical model for individualized prediction of OS and to provide a tool that may support risk stratification and treatment planning in contemporary clinical practice. We present this article in accordance with the TRIPOD reporting checklist (14,15) (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1487/rc).

Methods

Study design and patient selection

This multicenter retrospective study enrolled patients diagnosed with SCLC between April 2020 and April 2024 from three tertiary hospitals: Union Hospital, Tongji Medical College, Huazhong University of Science and Technology; Xiangyang No. 1 People’s Hospital; and 900th Hospital of Joint Logistics Support Force. Eligible patients met the following inclusion criteria: (I) pathologically confirmed SCLC; (II) receipt of at least one cycle of anticancer treatment, including systemic therapy and/or radiotherapy; and (III) availability of complete clinicopathological and follow-up data. Exclusion criteria included (I) a history of other malignancies and (II) receipt of fewer than one cycle of planned anticancer treatment. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective cohort study was approved by the institutional review boards and local ethics committees of the following hospitals: Union Hospital, Tongji Medical College, Huazhong University of Science and Technology (IRB No. 0464), Xiangyang No. 1 People’s Hospital (IRB No. XYYYE20240076), and 900th Hospital of Joint Logistics Support Force, Fujian Medical University (IRB No. 2023-097). Written informed consent was obtained from all patients for publication of this article.

Treatment modalities and molecular pathology

Treatment information was retrieved from electronic medical records and verified by attending oncologists. CT regimens primarily included etoposide, irinotecan, or paclitaxel in combination with platinum-based agents. Immunotherapy comprised PD-L1 inhibitors (e.g., atezolizumab, durvalumab) and PD-1 inhibitors (e.g., adebrelimab, pembrolizumab). Targeted therapy, when applied, followed standard institutional protocols. Molecular pathology was evaluated using four immunohistochemical (IHC) biomarkers—Ki-67, CgA, Syn2, and CD56. Each marker was semi-quantitatively scored as negative [0], weakly positive [1], or strongly positive [2]. A composite pathological score was calculated as the sum of these individual marker scores, as previously described (16).

Study endpoints and definitions

The primary endpoint was OS, while PFS and short-term treatment efficacy served as the secondary endpoints. OS was defined as the time from the start of initial therapy to death from any cause. PFS was defined as the interval from the initiation of first-line therapy to documented disease progression or death from any cause. Outcome assessment was performed independently and blinded to predictor variables. OS and PFS were determined by two investigators who were not involved in predictor collection. Survival status and dates of death were obtained from hospital records and telephone follow-up, while progression events were confirmed based on radiological reports and clinical documentation according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 (17). Treatment responses were categorized as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). Investigators adjudicating outcomes were blinded to baseline pathological scores and molecular markers.

Predictive model development and validation

Patients from Union Hospital, Tongji Medical College, Huazhong University of Science and Technology and Xiangyang No. 1 People’s Hospital were combined to form the internal dataset for model development, while patients from the 900th Hospital of Joint Logistics Support Force were used as an independent external validation cohort. The internal dataset was randomly divided in a 7:3 ratio into a derivation cohort and an internal validation cohort.

A prognostic nomogram was developed to estimate individualized 1-year OS probability by integrating demographic, clinicopathological, molecular, and treatment-related variables. Candidate predictors were screened using a two-step variable selection procedure. First, the Boruta algorithm was applied as an exploratory feature screening method to identify potentially relevant variables and exclude irrelevant features. Variables retained after Boruta screening were then entered into the least absolute shrinkage and selection operator (LASSO) regression to reduce dimensionality further, minimize multicollinearity, and select the most informative predictors. Variables with non-zero coefficients in the LASSO model were subsequently included in the final multivariable Cox proportional hazards model. The final Cox model was developed in the internal derivation cohort, and the regression coefficients were used to construct the nomogram and calculate an individual risk score. The final nomogram included pathological score, therapy, clinical T stage, clinical N stage, and TNM clinical stage. The predicted 1-year OS probability for each patient was estimated from the baseline survival function of the final model. Sample size adequacy was assessed using the events-per-variable (EPV) principle to reduce the risk of overfitting (18). An EPV of at least 10 was considered acceptable. In the derivation cohort, 215 death events occurred within 1 year, and 5 predictors were retained in the final model, yielding an EPV of 43, which indicated an adequate sample size for model development.

Model performance was assessed in the derivation cohort, internal validation cohort, and external validation cohort. Discrimination was evaluated using time-dependent receiver operating characteristic (ROC) analysis and the area under the curve (AUC). A model achieving a time-dependent AUC of 0.80 or higher was considered to have sufficient discriminative power to be clinically meaningful, based on widely accepted criteria in medical risk prediction (19). Calibration was assessed using calibration curves for 1-year OS. Clinical usefulness was evaluated by decision curve analysis (DCA).

To determine whether the nomogram provided incremental prognostic value beyond conventional staging systems, its performance was directly compared with that of the American Joint Committee on Cancer (AJCC) TNM and VALG staging systems. Time-dependent ROC curves were generated for the nomogram, TNM stage, and VALG stage in the derivation, internal validation, and external validation cohorts, and the corresponding AUC values for 1-year OS prediction were compared.

Statistical analysis

All statistical analyses were performed using R software (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were expressed as mean ± standard deviation (SD) or median [interquartile range (IQR)], depending on data distribution. Between-group comparisons were performed using Student’s t-test for continuous variables and Pearson’s Chi-squared or Fisher’s exact tests for categorical variables. Missing data were imputed using the random forest algorithm implemented in the “mice” package of R. The Boruta algorithm ranked variable importance by comparing Z-scores with corresponding shadow features, while cross-validation in the LASSO procedure determined the optimal regularization parameter (λ) (20). The construction of the model conformed to the EPV principle, ensuring a minimum of ten outcome events per predictor included in the final model. Predictive performance was assessed using the area under the ROC curve for discrimination, with calibration curves evaluating agreement between predicted and observed outcomes. DCA was used to quantify the clinical net benefit of the final model (21,22).

Results

Patient selection and overall cohort characteristics

This study primarily examined non-surgical treatment options, including CT, immunotherapy, targeted therapy, and radiotherapy. According to the VALG classification, the 728 enrolled patients were categorized as ES-SCLC (n=485) and limited-stage SCLC (LS-SCLC) (n=243) (Figure 1). Among all participants, 628 patients (86.3%) were male, and 405 (55.6%) were younger than 65 years. Sex, age, body mass index (BMI), smoking status, alcohol use, and the grade of Ki-67, CgA, Syn2, and CD56 were comparable between the two groups. However, the mean pathological score was significantly higher in ES-SCLC patients than in LS-SCLC patients (4.58±1.35 vs. 4.17±1.63, P<0.001; Table 1). Treatment modality differed significantly by stage (P=0.005; Table 2). Chemo-immunotherapy (CIT) was the most common modality (36.8%), followed by CT (29.0%), radio-chemo-immunotherapy (RCIT) (22.3%), and radio-chemotherapy (RCT) (11.9%). CIT was more frequent in ES-SCLC, whereas RCT was more common in LS-SCLC. Etoposide plus platinum was the most frequently used CT regimen, and PD-1/PD-L1 inhibitors were the major immunotherapy agents. Anlotinib constituted 78.1% of targeted therapies (Table 2).

Figure 1 Flowchart of patient selection. 900th, 900th Hospital of Joint Logistics Support Force; DCA, decision curve analysis; LASSO, least absolute shrinkage and selection operator; ROC, receiver operating characteristic; SCLC, small cell lung cancer; WUHU, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology; XYPH, Xiangyang No. 1 People’s Hospital.

Treatment response and preliminary survival analysis

Treatment responses differed notably between ES-SCLC and LS-SCLC (P<0.001). LS-SCLC patients achieved a higher CR rate (3.7% vs. 0.5%), whereas ES-SCLC patients exhibited a significantly higher rate of PD (25.6% vs. 16.1%; P<0.001; Table 2). The majority of patients were diagnosed at an advanced stage, predominantly stage IV, consistent with the aggressive nature of SCLC. Overall, the objective response rate remained modest across the cohort, reflecting the rapid progression and intrinsic resistance characteristic of SCLC. Notably, LS-SCLC patients demonstrated superior early treatment responses compared with ES-SCLC patients, aligning with their generally lower disease burden and greater treatment accessibility. Preliminary survival analysis suggested a trend toward longer PFS and OS in LS-SCLC relative to ES-SCLC patients (P<0.001; Table 2). These findings collectively highlight stage-dependent differences in short-term treatment efficacy and provide a rationale for the subsequent evaluation of prognostic factors and treatment modalities.

Impact of treatment on prognosis

Kaplan-Meier analyses were performed according to CT regimen, immune checkpoint inhibitor class, and treatment modality. Among CT regimens, platinum-based etoposide was associated with longer PFS than irinotecan plus platinum (P=0.03), while no significant difference was observed versus paclitaxel plus platinum (P=0.37) (Figure 2A). OS did not differ significantly among CT regimens (Figure S1A). PD-1 inhibitors were associated with longer PFS than PD-L1 inhibitors (P=0.04; Figure 2B), whereas OS was comparable between the two groups (P=0.51; Figure S1B). Overall, treatment modality significantly impacted OS, but not PFS (Figure 2C, Figure S1C). Patients receiving CT alone had the shortest OS [median 41.3 weeks, 95% confidence interval (CI): 34.7–45.6], whereas those receiving RCIT achieved the longest OS (median 92.3 weeks, 95% CI: 77.3–102.9; P<0.001). No significant OS difference was observed between patients treated with chemo-immunotherapy and those receiving chemoradiotherapy (P=0.41). Overall, these findings suggest that treatment modality had a greater influence on OS than on PFS, and that combined-modality treatment may confer a survival advantage in SCLC.

Figure 2 Kaplan-Meier curves of treatment modalities. (A) Kaplan-Meier curves of PFS by first-line platinum-based CT regimen. (B) Kaplan-Meier curves of PFS for patients treated with PD-1 vs. PD-L1 inhibitors. (C) Kaplan-Meier curves of OS by treatment modality. (D) Heatmap of molecular pathology markers by VALG stage (LS-SCLC vs. ES-SCLC). CD56, cluster of differentiation 56; CgA, chromogranin A; CIT, chemo-immunotherapy; CT, chemotherapy; ES-SCLC, extensive-stage small-cell lung cancer; Ki-67, Ki-67 proliferation index; LS-SCLC, limited-stage small-cell lung cancer; OS, overall survival; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1; PFS, progression-free survival; RCIT, radio-chemo-immunotherapy; RCT, radio-chemotherapy; Syn2, Synaptophysin II; VALG, Veterans Administration Lung Study Group.

Relationship between molecular pathology and prognosis

SCLC is characterized by neuroendocrine differentiation, which can be quantified using molecular markers such as Ki-67, CgA, Syn2, and CD56 (23). The mean pathological score, calculated by aggregating IHC expression levels of these markers, was significantly higher in ES-SCLC than in LS-SCLC patients (Figure 2D), reflecting the more aggressive biology of extensive-stage disease. Univariate Cox regression identified positivity for individual markers as potential risk factors. Multivariate analysis demonstrated that the composite pathological score remained an independent prognostic indicator [hazard ratio (HR) =4.84, 95% CI: 2.17–10.79; Figure 3], suggesting that integrating multiple neuroendocrine markers provides a more robust assessment of patient prognosis than any single marker alone. These findings highlight the value of molecular pathology in patient risk stratification and support its inclusion in predictive models for SCLC survival.

Figure 3 Univariate and multivariate analysis of OS. BMI, body mass index; CD56, cluster of differentiation 56; CgA, chromogranin A; CI, confidence interval; CIT, chemo-immunotherapy; CR, complete response; CT, chemotherapy; ES-SCLC, extensive-stage small-cell lung cancer; HR, hazard ratio; Ki-67, Ki-67 proliferation index; LS-SCLC, limited-stage small-cell lung cancer; OS, overall survival; PD, progressive disease; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1; PR, partial response; RCIT, radio-chemo-immunotherapy; RCT, radio-chemotherapy; SD, stable disease; Syn2, synaptophysin 2; TNM, tumor, node, metastasis; VALG, Veterans Administration Lung Study Group.

Screening of OS-related variables and prediction model construction

According to the study design, patients from Union Hospital, Tongji Medical College, Huazhong University of Science and Technology (WUHU) and Xiangyang No. 1 People’s Hospital (XYPH) comprised the internal dataset, which was randomly divided into a derivation cohort (70%, n=468) and an internal validation cohort (30%, n=200), whereas patients from 900th served as the external validation cohort (n=60), with no significant baseline differences (Table S1). To identify factors associated with OS, a two-step feature selection procedure was applied using the Boruta algorithm followed by LASSO regression. In Boruta analysis, Clinical TNM stage, therapy, pathological score, and clinical N stage ranked among the most important variables, whereas immunotherapy, clinical T stage, radiotherapy, VALG stage, and clinical M stage also showed substantial importance (Figure 4A). These candidate variables were then subjected to LASSO regression for further shrinkage and selection, and the final nomogram incorporated pathological score, therapy, clinical T stage, clinical N stage, and TNM clinical stage (Figure 4B-4D). Higher pathological scores and more advanced tumor burden were associated with worse survival, whereas more intensive multimodal treatment was associated with a more favorable prognosis.

Figure 4 Screening for OS-related variables. (A) Results of the Boruta algorithm variable screening. (B) Regression coefficient path diagram for LASSO. (C) Cross-validation curves for LASSO regression. (D) Nomogram based on pathological score, therapy, and clinical T stage, clinical N stage, and TNM clinical stage. BMI, body mass index; CD56, cluster of differentiation 56; CgA, chromogranin A; CIT, chemo-immunotherapy; CT, chemotherapy; Ki-67, Ki-67 proliferation index; LASSO, least absolute shrinkage and selection operator; OS, overall survival; RCIT, radio-chemo-immunotherapy; RCT, radio-chemotherapy; Syn2, synaptophysin 2; TNM, tumor, node, metastasis; VALG, Veterans Administration Lung Study Group.

Validation and comparative performance of the nomogram

The nomogram was evaluated in the derivation, internal validation, and external validation cohorts. For 1-year OS prediction, the nomogram demonstrated better discrimination than the TNM and VALG staging systems in all cohorts (Figure 5A). The AUCs of the nomogram were 0.801 (95% CI: 0.758–0.844) in the derivation cohort, 0.800 (95% CI: 0.734–0.866) in the internal validation cohort, and 0.776 (95% CI: 0.646–0.906) in the external validation cohort, all of which were higher than those of the TNM and VALG staging systems.

Figure 5 Performance of the nomogram in the development and validation cohorts. (A) Time-dependent ROC curves of the nomogram for predicting 1-year OS in the internal derivation cohort, internal validation cohort, and external validation cohort. (B) Calibration curves of the nomogram for predicting 1-year OS in the internal derivation cohort, internal validation cohort, and external validation cohort. (C) DCA curves of the nomogram in the internal derivation cohort, internal validation cohort, and external validation cohort. AUC, area under the curve; CI, confidence interval; DCA, decision curve analysis; OS, overall survival; ROC, receiver operating characteristic; TNM, tumor, node, metastasis; VALG, Veterans Administration Lung Study Group.

Calibration analysis showed good agreement between predicted and observed 1-year OS in the derivation, internal validation, and external validation cohorts (Figure 5B). The corresponding Brier scores were 18.4% (95% CI: 16.7–20.0%), 18.8% (95% CI: 16.4–21.2%), and 19.8% (95% CI: 14.3–25.4%), respectively. DCA demonstrated that the nomogram yielded favorable net benefit across a broad range of threshold probabilities in all three cohorts (Figure 5C), supporting its potential clinical utility.

Discussion

In this multicenter real-world study, we developed and validated a prognostic nomogram for predicting 1-year OS in patients with SCLC by integrating pathological score, therapy, clinical T stage, clinical N stage, and TNM clinical stage. The model showed stable discrimination, good calibration, and favorable clinical utility in the derivation, internal validation, and external validation cohorts. Importantly, the nomogram consistently outperformed the conventional TNM and VALG staging systems, suggesting that prognostic assessment in SCLC may benefit from moving beyond stage-only classification toward a more integrated framework incorporating treatment-related and pathological information.

SCLC accounts for 15–20% of all lung cancer cases and is characterized by rapid growth, high malignancy, and low differentiation. Approximately 70% of patients present with extensive-stage disease at diagnosis, highlighting the need for effective non-surgical strategies. According to National Comprehensive Cancer Network (NCCN) guidelines, surgery is typically reserved for early-stage patients (I–IIA, T1–2, N0, M0), whereas most require multimodal non-surgical treatment (24).

CT remains the cornerstone of SCLC treatment, particularly for ES-SCLC (25). The standard regimen combines etoposide with platinum-based agents (cisplatin or carboplatin), typically administered over 4–6 cycles. Etoposide inhibits topoisomerase II, disrupting DNA replication and repair (26,27), while irinotecan inhibits topoisomerase I (28,29). Platinum agents induce DNA crosslinking, and taxanes stabilize microtubules to block cell division (30). These agents, however, carry side effects such as myelosuppression and neurotoxicity (31). Although meta-analyses suggest that irinotecan-platinum (IP) regimens may outperform etoposide-platinum (EP) in ES-SCLC (32), our data showed no significant survival difference (P=0.47), likely because all platinum-based combinations with etoposide were grouped under EP, instead of focusing on etoposide and cisplatin alone.

Radiotherapy is indispensable for SCLC treatment, proving effective in both post-surgical LS-SCLC and ES-SCLC cases (33,34). Phase III studies have demonstrated that concurrent chemoradiotherapy can significantly reduce the risk of local recurrence (35). In LS-SCLC, chemoradiotherapy improves local control and extends PFS and OS, while prophylactic cranial irradiation (PCI) reduces brain metastasis rates from approximately 60% to 30% and improves 3-year OS by around 5% (36). In ES-SCLC, radiotherapy is primarily used to target metastatic sites, particularly bone and brain metastases. According to Slotman et al., thoracic radiotherapy should be considered for ES-SCLC patients who respond to CT, in addition to PCI (37). Advances in radiotherapy techniques, such as image-guided radiotherapy (IGRT), intensity-modulated radiotherapy (IMRT), and stereotactic body radiotherapy (SBRT), have enhanced precision, allowing more accurate tumor targeting while sparing normal tissues. Radiotherapy not only kills tumor cells directly but also releases tumor-associated antigens, enhancing T-cell-mediated immune responses (38). This “radiotherapy-induced immune effect” highlights the potential of combining radiotherapy with immune checkpoint inhibitors as a treatment strategy.

Immunotherapy, particularly PD-1 and PD-L1 inhibitors, has emerged as a promising approach in SCLC (39-41). SCLC cells evade immunity via PD-L1 upregulation, which suppresses T-cell activity (42). Clinical trials, including IMpower133, have shown survival benefits when adding atezolizumab to CT in ES-SCLC (43). In our study, PD-1/PD-L1 inhibitors improved PFS, though OS benefits were limited. A real-world analysis by Wang et al. showed no significant difference in efficacy or prognosis between PD-1 inhibitors combined with EP CT and PD-L1 inhibitors combined with EP CT (44). Although the PD-1/PD-L1 pathway plays a vital role in immune evasion in SCLC, immune escape mechanisms are multifaceted and may involve pathways beyond PD-1/PD-L1 (45). Immune escape in SCLC is multifactorial: the PD-1/PD-L1 pathway primarily affects T-cell activity within the tumor, whereas CTLA-4 regulates T-cell priming in lymph nodes (46). Dual checkpoint inhibition has shown efficacy in recurrent SCLC patients but has yet to significantly extend OS in treatment-naïve ED-SCLC patients (47,48).

Unlike non-SCLC (NSCLC), SCLC lacks dominant driver mutations, limiting targeted therapy options. Agents such as rovalpituzumab tesirine (DLL3-targeting) and danusertib have shown activity in pretreated patients (49,50). Smoking is a known risk factor for SCLC, although no clear correlation was observed in our cohort. Molecular pathology, including markers Ki-67, CgA, synaptophysin, and CD56, provides insights into tumor biology and prognosis (51). Composite pathological scores integrating these markers were predictive of survival, whereas single markers alone were insufficient. The aggressive behavior and limited durability of treatment response in SCLC are closely associated with its complex tumor biology. A recent single-patient single-cell RNA sequencing study demonstrated neuroendocrine predominance together with an immunosuppressive tumor microenvironment, providing additional insight into the biological basis of immune evasion, treatment resistance, and poor prognosis (52).

Our study demonstrates that multimodal strategies integrating chemoradiotherapy with immunotherapy significantly improve survival compared with single-agent therapies. While platinum-based CT and immune checkpoint inhibitors remain effective, no single-agent therapy is sufficient for most patients. Tumor biology, biomarker expression, and patient-related factors necessitate individualized, multimodal approaches. In particular, recent evidence indicates that host-related characteristics, such as BMI in elderly patients with extensive-stage disease, may influence treatment tolerability, further underscoring the importance of personalized therapeutic strategies (53). Future studies should explore optimal treatment combinations and validate these strategies in diverse populations to enhance clinical outcomes.

There are several limitations in this study. First, although the model was externally validated using an independent cohort from a separate center, the external validation sample size was relatively small, which may limit the stability and precision of the performance estimates. Larger prospective multicenter studies are still needed to further confirm the generalizability and clinical utility of the model. Second, as this was a retrospective study, selection bias and residual confounding could not be fully avoided. Third, the composite pathological score was constructed as a pragmatic integrative variable based on routinely available IHC markers, and its weighting scheme warrants further optimization and prospective validation.

Conclusions

In conclusion, we developed and externally validated a multicenter nomogram for 1-year OS prediction in SCLC that integrates treatment pattern, pathological score, and clinical staging variables. The model outperformed conventional TNM and VALG staging systems and may provide a practical tool for individualized prognostic assessment, treatment planning, and follow-up management. Further prospective multicenter studies are warranted to validate its clinical utility and determine whether its use can improve risk-adapted therapeutic decision-making.

Acknowledgments

We express gratitude to the patients and their families for contributing to this study. We would also like to thank all contributing physicians and study assistants for providing clinical data.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1487/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1487/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1487/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1487/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This retrospective cohort study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by the institutional review boards and local ethics committees of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology (IRB No. 0464), Xiangyang No. 1 People’s Hospital (IRB No. XYYYE20240076), and 900th Hospital of Joint Logistics Support Force, Fujian Medical University (IRB No. 2023-097). Written informed consent was obtained from all patients for publication of this article. Clinical data were analyzed retrospectively and anonymously.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
2. Kim SY, Park HS, Chiang AC. Small Cell Lung Cancer: A Review. JAMA 2025;333:1906-17. [Crossref] [PubMed]
3. Herzog BH, Devarakonda S, Govindan R. Overcoming Chemotherapy Resistance in SCLC. J Thorac Oncol 2021;16:2002-15. [Crossref] [PubMed]
4. Mathieu L, Shah S, Pai-Scherf L, et al. FDA Approval Summary: Atezolizumab and Durvalumab in Combination with Platinum-Based Chemotherapy in Extensive Stage Small Cell Lung Cancer. Oncologist 2021;26:433-8. [Crossref] [PubMed]
5. Wang T, Huang KL, Pu D, et al. Neoadjuvant immunochemotherapy in potentially resectable small cell lung cancer: a real-world retrospective analysis. Transl Lung Cancer Res 2026;15:38. [Crossref] [PubMed]
6. Wang S, Yang L, Ci B, et al. Development and Validation of a Nomogram Prognostic Model for SCLC Patients. J Thorac Oncol 2018;13:1338-48. [Crossref] [PubMed]
7. Yao Z, Li C, Han F, et al. Development and validation of a serum inflammatory biomarker-driven machine-learning model for prognostic stratification in surgical limited-stage small cell lung cancer. Transl Lung Cancer Res 2025;14:2983-95. [Crossref] [PubMed]
8. Krpina K, Vranić S, Tomić K, et al. Small Cell Lung Carcinoma: Current Diagnosis, Biomarkers, and Treatment Options with Future Perspectives. Biomedicines 2023;11:1982. [Crossref] [PubMed]
9. Raso MG, Bota-Rabassedas N, Wistuba II. Pathology and Classification of SCLC. Cancers (Basel) 2021;13:820. [Crossref] [PubMed]
10. Pei R, Zhang L, Xie C, et al. Prognostic value of Ki-67 expression in patients with extensive-stage small cell lung cancer. Future Oncol 2017;13:1247-52. [Crossref] [PubMed]
11. Zhang X, He M, Zheng G, et al. The impact of stratified management of Ki-67 on the prognosis of small-cell lung cancer. BMC Cancer 2025;25:1197. [Crossref] [PubMed]
12. Gkika E, Benndorf M, Oerther B, et al. Immunohistochemistry and Radiomic Features for Survival Prediction in Small Cell Lung Cancer. Front Oncol 2020;10:1161. [Crossref] [PubMed]
13. He Y, Zhao L, Tang X, et al. Prognostic implications of synaptophysin, CD56, thyroid transcription factor-1, and Ki-67 in pulmonary high-grade neuroendocrine carcinomas. Ann Diagn Pathol 2024;68:152239. [Crossref] [PubMed]
14. World Medical Association. World medical association declaration of helsinki. JAMA 2013;310:2191.
15. Moons KG, Altman DG, Reitsma JB, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med 2015;162:W1-73. [Crossref] [PubMed]
16. Wang Q, Gümüş ZH, Colarossi C, et al. SCLC: Epidemiology, Risk Factors, Genetic Susceptibility, Molecular Pathology, Screening, and Early Detection. J Thorac Oncol 2023;18:31-46. [Crossref] [PubMed]
17. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228-47. [Crossref] [PubMed]
18. Austin PC, Steyerberg EW. Events per variable (EPV) and the relative performance of different strategies for estimating the out-of-sample validity of logistic regression models. Stat Methods Med Res 2017;26:796-808. [Crossref] [PubMed]
19. Çorbacıoğlu ŞK, Aksel G. Receiver operating characteristic curve analysis in diagnostic accuracy studies: A guide to interpreting the area under the curve value. Turk J Emerg Med 2023;23:195-8. [Crossref] [PubMed]
20. How to develop a more accurate risk prediction model when there are few events. BMJ 2016;353:i3235.
21. Royston P, Sauerbrei W. A new measure of prognostic separation in survival data. Stat Med 2004;23:723-48. [Crossref] [PubMed]
22. Huang Y, Li W, Macheret F, et al. A tutorial on calibration measurements and calibration models for clinical prediction models. J Am Med Inform Assoc 2020;27:621-33. [Crossref] [PubMed]
23. Sakakibara R, Kobayashi M, Takahashi N, et al. Insulinoma-associated Protein 1 (INSM1) Is a Better Marker for the Diagnosis and Prognosis Estimation of Small Cell Lung Carcinoma Than Neuroendocrine Phenotype Markers Such as Chromogranin A, Synaptophysin, and CD56. Am J Surg Pathol 2020;44:757-64. [Crossref] [PubMed]
24. Megyesfalvi Z, Gay CM, Popper H, et al. Clinical insights into small cell lung cancer: Tumor heterogeneity, diagnosis, therapy, and future directions. CA Cancer J Clin 2023;73:620-52. [Crossref] [PubMed]
25. Ganti AKP, Loo BW, Bassetti M, et al. Small Cell Lung Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2021;19:1441-64. [Crossref] [PubMed]
26. O'Dwyer PJ, Leyland-Jones B, Alonso MT, et al. Etoposide (VP-16-213). Current status of an active anticancer drug. N Engl J Med 1985;312:692-700.
27. Le TT, Wang MD. Topoisomerase II and etoposide - a tangled tale. Nat Chem Biol 2023;19:546-7. [Crossref] [PubMed]
28. Bailly C. Irinotecan: 25 years of cancer treatment. Pharmacol Res 2019;148:104398. [Crossref] [PubMed]
29. Talukdar A, Kundu B, Sarkar D, et al. Topoisomerase I inhibitors: Challenges, progress and the road ahead. Eur J Med Chem 2022;236:114304. [Crossref] [PubMed]
30. Zhang C, Xu C, Gao X, et al. Platinum-based drugs for cancer therapy and anti-tumor strategies. Theranostics 2022;12:2115-32. [Crossref] [PubMed]
31. Zhu L, Chen L. Progress in research on paclitaxel and tumor immunotherapy. Cell Mol Biol Lett 2019;24:40. [Crossref] [PubMed]
32. Han D, Wang G, Sun L, et al. Comparison of irinotecan/platinum versus etoposide/platinum chemotherapy for extensive-stage small cell lung cancer: A meta-analysis. Eur J Cancer Care (Engl) 2017;
33. Merie R, Gee H, Hau E, et al. An Overview of the Role of Radiotherapy in the Treatment of Small Cell Lung Cancer - A Mainstay of Treatment or a Modality in Decline? Clin Oncol (R Coll Radiol) 2022;34:741-52. [Crossref] [PubMed]
34. Zhang C, Zhao G, Wu H, et al. Application of postoperative adjuvant radiotherapy in limited-stage small cell lung cancer: A systematic review and meta-analysis. Radiother Oncol 2024;193:110123. [Crossref] [PubMed]
35. Takada M, Fukuoka M, Kawahara M, et al. Phase III study of concurrent versus sequential thoracic radiotherapy in combination with cisplatin and etoposide for limited-stage small-cell lung cancer: results of the Japan Clinical Oncology Group Study 9104. J Clin Oncol 2002;20:3054-60. [Crossref] [PubMed]
36. Suwinski R. Prophylactic cranial irradiation in SCLC. Transl Lung Cancer Res 2021;10:2071-8. [Crossref] [PubMed]
37. Slotman BJ, van Tinteren H, Praag JO, et al. Use of thoracic radiotherapy for extensive stage small-cell lung cancer: a phase 3 randomised controlled trial. Lancet 2015;385:36-42. [Crossref] [PubMed]
38. Demaria S, Coleman CN, Formenti SC. Radiotherapy: Changing the Game in Immunotherapy. Trends Cancer 2016;2:286-94. [Crossref] [PubMed]
39. Goldman JW, Dvorkin M, Chen Y, et al. Durvalumab, with or without tremelimumab, plus platinum-etoposide versus platinum-etoposide alone in first-line treatment of extensive-stage small-cell lung cancer (CASPIAN): updated results from a randomised, controlled, open-label, phase 3 trial. Lancet Oncol 2021;22:51-65. [Crossref] [PubMed]
40. Horn L, Mansfield AS, Szczęsna A, et al. First-Line Atezolizumab plus Chemotherapy in Extensive-Stage Small-Cell Lung Cancer. N Engl J Med 2018;379:2220-9. [Crossref] [PubMed]
41. Ready N, Farago AF, de Braud F, et al. Third-Line Nivolumab Monotherapy in Recurrent SCLC: CheckMate 032. J Thorac Oncol 2019;14:237-44. [Crossref] [PubMed]
42. Carvajal-Hausdorf D, Altan M, Velcheti V, et al. Expression and clinical significance of PD-L1, B7-H3, B7-H4 and TILs in human small cell lung Cancer (SCLC). J Immunother Cancer 2019;7:65. [Crossref] [PubMed]
43. Liu SV, Reck M, Mansfield AS, et al. Updated Overall Survival and PD-L1 Subgroup Analysis of Patients With Extensive-Stage Small-Cell Lung Cancer Treated With Atezolizumab, Carboplatin, and Etoposide (IMpower133). J Clin Oncol 2021;39:619-30. [Crossref] [PubMed]
44. Wang Y, Li L, Hu J, et al. Comparison of efficacy and safety between PD-1 inhibitors and PD-L1 inhibitors plus platinum-etoposide as first-line treatment for extensive-stage small-cell lung cancer: a multicenter, real-world analysis. BMC Cancer 2023;23:1196. [Crossref] [PubMed]
45. Remon J, Passiglia F, Ahn MJ, et al. Immune Checkpoint Inhibitors in Thoracic Malignancies: Review of the Existing Evidence by an IASLC Expert Panel and Recommendations. J Thorac Oncol 2020;15:914-47. [Crossref] [PubMed]
46. Rowshanravan B, Halliday N, Sansom DM. CTLA-4: a moving target in immunotherapy. Blood 2018;131:58-67. [Crossref] [PubMed]
47. Antonia SJ, López-Martin JA, Bendell J, et al. Nivolumab alone and nivolumab plus ipilimumab in recurrent small-cell lung cancer (CheckMate 032): a multicentre, open-label, phase 1/2 trial. Lancet Oncol 2016;17:883-95. [Crossref] [PubMed]
48. Owonikoko TK, Park K, Govindan R, et al. Nivolumab and Ipilimumab as Maintenance Therapy in Extensive-Disease Small-Cell Lung Cancer: CheckMate 451. J Clin Oncol 2021;39:1349-59. [Crossref] [PubMed]
49. Rudin CM, Pietanza MC, Bauer TM, et al. Rovalpituzumab tesirine, a DLL3-targeted antibody-drug conjugate, in recurrent small-cell lung cancer: a first-in-human, first-in-class, open-label, phase 1 study. Lancet Oncol 2017;18:42-51. [Crossref] [PubMed]
50. Schöffski P, Besse B, Gauler T, et al. Efficacy and safety of biweekly i.v. administrations of the Aurora kinase inhibitor danusertib hydrochloride in independent cohorts of patients with advanced or metastatic breast, ovarian, colorectal, pancreatic, small-cell and non-small-cell lung cancer: a multi-tumour, multi-institutional phase II study. Ann Oncol 2015;26:598-607.
51. Baykara Y, Xiao Y, Yang D, et al. Utility of secretagogin as a marker for the diagnosis of lung neuroendocrine carcinoma. Virchows Arch 2022;481:31-9. [Crossref] [PubMed]
52. Wang Y, He S, Wang R, et al. Single-patient single-cell RNA sequencing reveals neuroendocrine predominance and immunosuppression in small-cell lung cancer. Transl Cancer Res 2026;15:46. [Crossref] [PubMed]
53. Terashima Y, Takeuchi S, Ishibashi Y, et al. Retrospective study on the tolerability using a four-cycle completion in elderly patients with extensive-stage small-cell lung cancer: a body mass index-based analysis. Transl Lung Cancer Res 2026;15:8. [Crossref] [PubMed]