Real-world efficacy of atezolizumab combined with etoposide and platinum chemotherapy in extensive-stage small cell lung cancer: a retrospective cohort study

Introduction

Small-cell lung cancer (SCLC) is a highly aggressive thoracic malignancy characterized by rapid progression, early metastasis, and poor prognosis (1,2). It accounts for approximately 15% of all lung cancer cases (3). At initial diagnosis, 60% to 70% of patients present with extensive-stage disease (ES-SCLC), with a median overall survival (mOS) less than 1 year (3,4). Although platinum-based chemotherapy combined with etoposide (EP) has long been the standard of care for first-line treatment and is associated with high initial response rates, most patients experience early relapse and very limited benefit from subsequent therapies.

Recent advances in immunotherapy have reshaped the treatment landscape for ES-SCLC. Immune checkpoint inhibitors (ICIs) when combined with chemotherapy have demonstrated promising survival benefits and are now recommended as standard first-line therapy in several international guidelines (3). In the landmark IMpower133 trial, the addition of atezolizumab to carboplatin and etoposide significantly prolonged mOS from 10.3 to 12.3 months with a hazard ratio (HR) of 0.70 (5,6). However, OS curves appeared to be identical after 30 months (16% both) (7), a finding that prompted some health authorities to decline approval (e.g., NICE, UK). Similarly, the CASPIAN trial also confirmed the efficacy of durvalumab plus chemotherapy in terms of improved mOS (13 vs. 10.3 months, HR =0.73) (7). In this trial, the observed mOS benefit was durable even after 36 months (17.6% vs. 5.8%) (8).

Interestingly, these benefits for mOS were not found in other trials using anti-programmed cell-death protein-1 (PD-1) monoclonal antibodies (e.g., Keynote-604, CheckMate-451) suggesting that SCLCs may represent a class of programmed death-ligand-1 (PD-L1)-driven cancers. Notably, several recent trials from China involving PD-1 inhibitors such as serplulimab and tislelizumab have demonstrated encouraging efficacy and led to European Medicines Agency (EMA) approval. However, it should be noted that the prevalence of PD-L1 in SCLC is lower than that published for non-small-cell lung cancer (NSCLC). Therefore, the predictive role of PD-L1 expression in SCLC treatment still remains to be established. patients enrolled in randomized controlled trials (RCTs) are often highly selected, younger individuals with good performance status [Eastern Cooperative Oncology Group (ECOG) 0–1], minimal comorbidities, and low incidence of brain. This raises concerns regarding the generalizability of RCT findings to real-world clinical settings, where patient populations are more heterogeneous and frequently include older individuals with a substantial comorbid burden and poor baseline status (9).

To address this lack of information, we conducted a retrospective real-world study based on clinical data from Tianjin Chest Hospital. This study aimed to compare the clinical outcomes of atezolizumab combined with EP chemotherapy vs. those of EP chemotherapy alone in patients with ES-SCLC. Eligible patients were diagnosed between January 2019 and December 2024 and received at least four cycles of first-line treatment without any subsequent maintenance therapy with atezolizumab. The primary endpoints were OS and progression-free survival (PFS). We hypothesized that atezolizumab-based combination therapy offers survival benefits in real-world populations, particularly among patients without brain metastases. We examined whether the delayed separation of survival curves—an immunotherapy hallmark observed in clinical trials—could be reproduced in daily life (10,11). We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1049/rc).

Methods

Study design and data sources

We conducted a single-center, retrospective cohort trial at Tianjin Chest Hospital. Data were extracted from the electronic medical record system of the Department of Medical Oncology and from outpatient records and telephone follow-up documentation. Eligible patients were those who were pathologically diagnosed with ES-SCLC between January 2019 and December 2024 and who had received first-line systemic antitumor therapy at our institution. Treatment decisions were made by the attending physicians based on clinical judgment, patient preference, and drug accessibility, and not through random assignment. All data were collected, verified, and analyzed between January and March 2025.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Tianjin Chest Hospital (approval No. 2024KY-028-01), and the requirement for informed consent was waived due to the retrospective design and use of anonymized data. Patient-level data were extracted using a Python-based script (version 3.11.9) that was applied to hospital information system records, based on predefined criteria such as diagnostic codes, treatment regimens, and medication start dates. All extracted data were subsequently reviewed manually to ensure clinical consistency.

Patient selection

Patients were eligible for inclusion if they met the following criteria: (I) histologically confirmed ES-SCLC diagnosed at Tianjin Chest Hospital between January 2019 and December 2024; (II) administration of either atezolizumab plus EP chemotherapy or EP chemotherapy alone as first-line treatment; (III) completion of at least four cycles of treatment without dose reduction or early discontinuation (no subsequent maintenance therapy with anti-PD-L1 drugs); and (IV) complete follow-up information for key clinical outcomes, including survival status and disease progression. To reduce potential confounding from poor baseline functional status, only patients with ECOG performance status of 0 or 1 were included in this analysis. Of 174 screened patients, only 18 (10.3%) had ECOG 2; fewer than 10% of them completed four chemotherapy cycles, and they were therefore excluded to preserve internal validity. There was no upper age limit for inclusion; patients aged 70 years or older were eligible, in accordance with international standards for real-world and clinical trial studies in SCLC.

Meanwhile, the exclusion criteria were as follows: (I) coexisting malignancies from other organ systems or severe underlying diseases; (II) administration of ICIs other than atezolizumab or regimens not defined as study exposures; and (III) missing of significant clinical or follow-up data preventing valid outcome analysis.

Treatment data were obtained from structured fields within the electronic medical record, including medication administration records and physician orders. The initiation of first-line therapy was defined as the exposure start point. The number of treatment cycles, dose modifications, treatment interruptions, and adverse event–related discontinuations were verified through manual chart review.

Among the initially screened 247 patients, 73 were excluded based on the eligibility criteria, leaving 174 patients in the preliminary cohort. After those with insufficient treatment duration, adverse event-related discontinuation, or missing data were excluded, 95 patients were included in the final analysis (Figure 1).

Figure 1 Flowchart of patient selection. A total of 247 patients were screened, and after the application of inclusion and exclusion criteria, 95 patients were included in the final analysis. EP, etoposide and platinum; ES-SCLC, extensive-stage small cell lung cancer.

Exposure and outcome definitions

Exposure

The primary exposure was the type of first-line treatment received, which included two regimens. In the atezolizumab plus EP group, patients received atezolizumab in combination with EP (with carboplatin or cisplatin as the platinum agent). In the EP-only group, patients who received chemotherapy with EP without any ICIs.

Treatment grouping was based on medication administration records at the time of the first EP cycle initiation. Data were extracted using automated scripts and confirmed through manual validation.

Outcomes

The two primary outcomes were OS, which was considered to be the time from initiation of first-line therapy to death from any cause or last follow-up (censored at December 31, 2024), and PFS, which was considered to be the time from treatment initiation to documented disease progression, death, or last available follow-up, whichever occurred first.

The outcome data were derived from hospitalization records, outpatient follow-up, and structured telephone interviews. Disease progression was determined based on imaging reports [computed tomography (CT) or magnetic resonance imaging (MRI)] or clinical documentation consistent with Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 (12). In the absence of progression data, death was considered to be a PFS event.

Covariates and potential confounders

To address any baseline imbalances between the treatment groups, we identified a set of the following covariates to be included in the propensity score model and weighting process (13): age (continuous), sex (male/female), ECOG performance status (0 vs. 1), smoking history (years smoked and number of cigarettes per day), comorbidities (hypertension and diabetes mellitus), presence of brain metastases (yes/no), tumor burden level (categorized as high, moderate, or low based on baseline imaging), and number of metastatic sites (continuous).

Covariate data were extracted from multiple components of the electronic medical records, including admission notes, initial treatment summaries, medical orders, laboratory data, and imaging reports. Tumor burden was assessed by two radiologists based on the number of involved organs and maximum lesion diameter: high (≥3 metastatic organs or lesions >5 cm), moderate (2 organs or lesions 3–5 cm), and low (<2 organs and all lesions <3 cm). Other comorbidities such as chronic obstructive pulmonary disease (COPD) were not included in the propensity score model, primarily due to the low recorded prevalence of COPD diagnoses in our center’s patient cohort and incomplete documentation in the electronic medical record system.

Follow-up procedures and data collection

The follow-up period for all patients began on the date of first-line treatment initiation and ended at the earliest of death, disease progression, loss to follow-up, or the predefined study cutoff date (December 31, 2024).

Follow-up data were collected from three main sources: (I) inpatient medical records, which included discharge summaries, clinical notes, and imaging reports; (II) the outpatient follow-up system, which included return visits and diagnostic evaluations; and (III) structured telephone interviews conducted by trained research personnel to confirm survival status and disease progression.

Progression events were determined based on radiologic evidence (CT or MRI) or clinical documentation of disease progression according to RECIST 1.1 (12). If progression was not documented but the patient had died, the death date was considered the PFS event time.

All clinical data were initially extracted through automated Python scripts targeting structured data fields, including demographics, diagnosis, treatment timeline, comorbidities, laboratory values, and radiology reports. After extraction, two independent investigators manually reviewed each case to ensure data completeness, accurate exposure classification, and consistency in follow-up timelines. Cases with substantial missing information were excluded from the final analysis.

Adverse events (AEs) were extracted from electronic medical charts and graded according to CTCAE v5.0. Grade ≥3 events were predefined as clinically significant and are summarised separately (Table 1).

Detailed information on systemic therapy given after first progression was unavailable in 34 % of charts and was therefore not analysed; such data are being captured prospectively in the extension study.

Statistical analysis

All data preprocessing and statistical analyses were performed using Python version 3.11.9. The core libraries included Pandas, NumPy, Lifelines, Statsmodels, and Matplotlib.

To control for potential confounding, inverse probability of treatment weighting (IPTW) based on propensity scores was applied to balance the baseline covariates between groups (13). The propensity score model included age, sex, ECOG performance status, smoking history, comorbidities, tumor burden, brain metastases, and the number of metastatic sites. Balance after weighting was assessed using standardized mean differences (SMDs), with SMD <0.1 indicating acceptable covariate balance.

Survival curves for OS and PFS were estimated using the Kaplan-Meier method. Median survival times and 95% confidence intervals (CIs) are reported. Cox proportional hazards models with robust variance estimation and IPTW weights were used to estimate HRs and 95% CIs. An HR <1 indicated better outcomes in the atezolizumab plus EP group.

Given that IPTW alters the independence and homoscedasticity assumptions required for log-rank tests, we did not use the log-rank test to compare weighted survival curves (14). Instead, statistical significance was assessed with the P values from the weighted Cox regression models (Wald tests).

Subgroup analyses were conducted according to the presence or absence of brain metastases to examine the potential effect modification (15). Two-sided P values <0.05 were considered statistically significant for all analyses.Given the limited sample size (n=95) and the collinearity between baseline covariates already balanced via IPTW, adding nine additional predictors to a conventional Cox model would result in over-fitting and highly unstable estimates. Therefore, in this retrospective cohort we confined adjusted analyses to IPTW-weighted survival curves and a brain-metastasis subgroup; a full multivariable Cox analysis has been prespecified for our prospective extension study (target sample ~220).

Results

Study population and baseline characteristics

A total of 247 patients diagnosed with ES-SCLC at Tianjin Chest Hospital between January 2019 and December 2024 were initially screened. After 73 patients were excluded who did not meet the inclusion criteria—19 due to comorbid malignancies or severe underlying diseases and 54 for receiving nonstandard first-line therapy—a preliminary cohort of 174 patients was formed.

Of these, 96 patients received atezolizumab plus EP, and 78 received EP alone. After further exclusions due to insufficient treatment cycles (<4 cycles), treatment interruption or dose reduction due to adverse events, and incomplete data, the final analytical cohort consisted of 95 patients: 42 in the atezolizumab + EP group and 53 in the EP-only group (Figure 1).

Baseline clinical characteristics of the two groups are summarized in Table 2. The atezolizumab + EP group and EP-only group were generally well-matched in terms of sex (male: 76.2% vs. 75.5%), ECOG performance status (ECOG 0: 42.9% vs. 47.2%), hypertension (54.8% vs. 52.8%), and diabetes (28.6% vs. 28.3%). Patients in the combination group were slightly older (68.2±8.5 vs. 65.7±9.1 years), had a longer smoking history, and a higher average number of cigarettes per day. A greater proportion of patients in this group also presented with high tumor burden (23.8% vs. 45.3%).

To control for baseline imbalances, IPTW was applied (13). Prior to weighting, variables such as tumor burden (SMD =0.42), smoking history (SMD =0.19), and ECOG score (SMD =0.21) showed notable differences. After weighting, all SMD values were reduced to <0.1, indicating adequate covariate balance between the groups (Figure 2) (16).

Figure 2 SMD before and after IPTW adjustment. All covariates were balanced post-weighting (SMD <0.1). ECOG, Eastern Cooperative Oncology Group; IPTW, inverse probability of treatment weighting; SMD, standardized mean difference.

OS outcomes

In the IPTW-weighted cohort, patients treated with atezolizumab plus EP demonstrated longer OS. The median OS was 9.7 months in the combination group and 7.1 months in the EP-only group (14). Weighted Cox regression yielded an HR of 0.51 (95% CI: 0.30–0.94), suggesting a significant survival advantage in the immunotherapy group (Figure 3).

Figure 3 IPTW-weighted Kaplan-Meier curve for OS. The atezolizumab plus EP group showed prolonged OS as compared to the EP-only group (median 9.7 vs. 7.1 months; HR =0.51, 95% CI: 0.30–0.94). Baseline sample sizes were 42 in the atezolizumab + EP group and 53 in the EP-only group. CI, confidence interval; EP, etoposide and platinum; HR, hazard ratio; IPTW, inverse probability of treatment weighting; KM, Kaplan-Meier; OS, overall survival.

A similar trend was observed in the unweighted Kaplan-Meier analysis (Figure 4), although the difference was nonsignificant (HR =0.54; P=0.0534).

Figure 4 Unweighted Kaplan-Meier curve for OS. The immunotherapy group showed a trend toward improved survival (HR =0.54; P=0.050). Baseline sample sizes were 42 in the atezolizumab + EP group and 53 in the EP-only group. CI, confidence interval; EP, etoposide and platinum; HR, hazard ratio; KM, Kaplan-Meier; OS, overall survival.

PFS outcomes

Median PFS was similar between the two groups, at 5.8 months in the combination group and 5.7 months in the EP-only group. However, the IPTW-adjusted survival curves (Figure 5) showed a delayed separation beginning around 3–6 months after treatment initiation (3), suggesting a potential durable benefit for a subset of patients receiving immunotherapy (10,15). The HR for PFS was 0.42 (95% CI: 0.21–0.86), favoring the combination therapy.

Figure 5 IPTW-weighted Kaplan-Meier curve for PFS. A delayed separation was observed in the immunotherapy group (HR =0.42, 95% CI: 0.21–0.86). Baseline sample sizes were 42 in the atezolizumab + EP group and 53 in the EP-only group. CI, confidence interval; EP, etoposide and platinum; HR, hazard ratio; IPTW, inverse probability of treatment weighting; KM, Kaplan-Meier; PFS, progression-free survival.

In the unweighted analysis (Figure 6), the PFS curves were largely overlapping. Nevertheless, the HR remained favorable for the immunotherapy group at 0.37 (95% CI: 0.18–0.77).

Figure 6 Unweighted Kaplan-Meier curve for PFS. Although the median PFS was similar, the immunotherapy group showed a more favorable HR (HR =0.37, 95% CI: 0.18–0.77). Baseline sample sizes were 42 in the atezolizumab + EP group and 53 in the EP-only group. CI, confidence interval; EP, etoposide and platinum; HR, hazard ratio; KM, Kaplan-Meier; PFS, progression-free survival.

Safety

Treatment-related grade ≥3 adverse events are summarised in Table 1; the most frequent were leukopenia and anaemia. Grade 1–2 laboratory abnormalities were common but clinically manageable.

Although not included in the present analysis, we acknowledge that univariate and multivariate Cox regression could help identify independent prognostic factors such as ECOG status, age, and tumor burden. These analyses are planned for future studies to further support individualized treatment approaches.

As IPTW violates key assumptions of the log-rank test, no log-rank P values were reported for the weighted curves (11). All statistical comparisons were based on weighted Cox regression models.

Subgroup analysis by brain metastasis

To examine the potential treatment heterogeneity, subgroup analyses were conducted according to brain metastasis status (17).

In the brain metastasis subgroup, no statistically significant survival differences were observed, although the combination group had longer survival. Due to the small sample size and residual baseline imbalance, further interpretation should be conducted with caution, and these findings should be considered exploratory.

In the no-brain metastasis subgroup, a significant OS benefit was observed in the atezolizumab group as compared to the EP-only group, with a median OS of 9.7 and 6.8 months, respectively (HR =0.47, 95% CI: 0.23–0.97) (Figure 7); the PFS curves in this subgroup also had delayed separation, with a median PFS of 8.7 and 5.7 months, respectively (HR =0.33, 95% CI: 0.15–0.75) (Figure 8).

Figure 7 Subgroup analysis of OS in patients without brain metastases. Atezolizumab improved OS significantly (HR =0.47, 95% CI: 0.23–0.97). Baseline sample sizes were 42 in the atezolizumab + EP group and 53 in the EP-only group. CI, confidence interval; EP, etoposide and platinum; HR, hazard ratio; KM, Kaplan-Meier; OS, overall survival.

Figure 8 Subgroup analysis of PFS in patients without brain metastases. Delayed separation in survival curves favored the immunotherapy group (HR =0.33, 95% CI: 0.15–0.75). Baseline sample sizes were 42 in the atezolizumab + EP group and 53 in the EP-only group. CI, confidence interval; EP, etoposide and platinum; HR, hazard ratio; IPTW, inverse probability of treatment weighting; KM, Kaplan-Meier; PFS, progression-free survival.

These results suggest that patients without brain metastases may derive more pronounced survival benefit from immunotherapy-based combination treatment (4,14,18), whereas findings in the brain metastasis subgroup remain hypothesis-generating due to the limited statistical power.

Discussion

In this retrospective, real-world cohort study involving patients with ES-SCLC, the addition of atezolizumab to standard EP chemotherapy was associated with significantly improved OS. After the baseline covariates were balanced via inverse IPTW, the immunotherapy group achieved a median OS of 9.7 months while that of the EP-only group was 7.1 months (HR =0.51, 95% CI: 0.30–0.94). Although the median PFS was similar between the groups (5.8 vs. 5.7 months), the immunotherapy group demonstrated a delayed separation in the Kaplan-Meier curves, with an HR of 0.42 (95% CI: 0.21–0.86), suggesting potential durable benefit in this subset of patients (4,19).

Although the observed median progression-free survival (mPFS) values in our study were comparable to those reported in the IMPower133 trial (5.2 vs. 4.3 months) (5) and in the CASPIAN trial (5.1 vs. 5.4 months) (7), the mOS in our immunotherapy group was found to be lower when compared with IMPower133 and CASPIAN trials. This might be attributed to the different treatment regimens: Patients in the CASPIAN and the IMPower133 trials received four cycles of immuno-chemotherapy followed by a maintenance phase (monotherapy with durvalumab or atezolizumab, respectively) until they had unacceptable toxic effects, disease progression, or no additional benefit, whereas patients enrolled in our retrospective cohort study were selected to exclude those who received atezolizumab maintenance therapy, in order to ensure uniform exposure and improve comparability across groups. Although some patients did receive maintenance atezolizumab in clinical practice, they were not included in this analysis to maintain consistency in treatment evaluation.

Subgroup analyses further supported this observation: patients without brain metastases experienced a more pronounced improvement in both OS and PFS. These findings collectively indicate that atezolizumab-based combination therapy may offer meaningful survival benefit in a real-world clinical population.

Our findings are directionally consistent with those from the pivotal phase III RCTs such as IMpower133 and CASPIAN (6,7,20). In the IMpower133, atezolizumab combined with carboplatin and etoposide improved the median OS from 10.3 to 12.3 months, while the CASPIAN demonstrated similar improvements from durvalumab-based regimens. These trials have laid the foundation for the global adoption of immune-chemotherapy as first-line treatment for ES-SCLC.

However, it is important to highlight the differences in patient populations between clinical trials and routine practice. Participants in RCTs typically have favorable baseline characteristics, including younger age (median age 64 years in IMpower133 and 63 years in CASPIAN), ECOG performance status of 0–1, fewer or no comorbidities, and low rates of brain metastases. By contrast, our real-world cohort included patients with higher tumor burden, older age (median 68.2 years in the atezolizumab + EP group), performance status ≥1, and multiple comorbidities (11). Despite these real-world complexities, the observed benefit of immunotherapy remained robust. This supports the generalizability of trial findings and strongly reinforces the relevance of atezolizumab in more heterogeneous and clinically challenging populations (19).

A distinctive pattern observed in this study was the delayed separation of the Kaplan-Meier curves for PFS in the atezolizumab group, beginning approximately 3 to 6 months after treatment initiation. This finding was also reported in the CASPIAN trial (7) and is known to be a common observation in clinical studies with checkpoint inhibitors (10).

This phenomenon is frequently described in immuno-oncology trials and is attributed to the unique kinetics of immune activation (10). Unlike chemotherapy, which exerts rapid cytotoxic effects, ICIs require time to modulate the tumor microenvironment, recruit effector T cells, and generate a sustained antitumor response (15). Consequently, early disease progression may occur before immune-mediated tumor control is achieved in some patients.

This temporal delay can also challenge conventional response assessment tools, such as RECIST 1.1 (12), which may misclassify early immune flares or pseudoprogression as true progression. Despite these limitations, the long-term benefit among patients who surpass this initial phase is clinically meaningful. The presence of this delayed separation in our real-world cohort suggests that the immunotherapeutic mechanism of action remains preserved outside of clinical trial settings. Recognizing this pattern is essential for clinicians to avoid premature discontinuation of therapy and to correctly interpret early imaging results (21). In addition, this observation adds further wait to the proposal that mPFS might not be an ideal surrogate parameter for mOS in trials with checkpoint inhibitors. Furthermore, to date the relationship between PFS and OS has not been established in advanced NSCLC and remains to be controversial.

Subgroup analysis revealed a differential treatment effect based on the presence of brain metastases. Among patients without brain involvement, the addition of atezolizumab significantly improved both OS and PFS (15). In contrast, patients with brain metastases who have used anti-PD-1 and anti-PD-L1 monoclonal antibodies in recent years exhibited only modest, statistically nonsignificant benefit (18,20).

Several factors may contribute to this attenuated response. First, the blood-brain barrier (BBB) poses a physical and immunologic challenge to systemic therapies, potentially limiting immune-cell trafficking and drug penetration into intracranial lesions. Second, patients with brain metastases are often treated with corticosteroids to control neurologic symptoms, which can suppress T cell-mediated immune response and counteract the efficacy of checkpoint inhibition. Third, advanced intracranial disease itself represents a more aggressive or immune-evasive tumor phenotype (20), further dampening immunotherapy responsiveness (21).

Given the small sample size and potential residual confounding, these findings should be interpreted cautiously. Nevertheless, they underscore the importance of considering brain metastasis status in the evaluation of immunotherapy outcomes and highlight the need for future studies to better understand how central nervous system involvement affects treatment efficacy in patients with ES-SCLC.

The safety profile was consistent with phase III trials such as IMpower-133 and CASPIAN: severe haematologic AEs occurred in roughly one-third of patients, while no grade ≥3 immune-related events were detected (Table 1).

Limitations

This study involved several limitations that should be acknowledged. First, as a single-center retrospective analysis, it remains subject to inherent selection bias and unmeasured confounding despite the application of IPTW. Detailed information on subsequent lines of therapy was not uniformly available and thus not analyzed in this study. This may affect survival interpretation and represents a limitation. In addition, treatment selection was not randomized and may have been influenced by factors such as patient frailty or insurance status. Safety data, including immune-related adverse events, were not consistently available across the cohort, especially for outpatient care, and thus were not included in this analysis. Factors such as treatment compliance, socioeconomic status, and patient preferences—often missing in electronic medical records—may influence survival outcomes but were not captured in our dataset.

Second, the relatively small sample size, particularly in subgroup analyses, limits the statistical power in detecting modest differences and increases the potential for type II error. This is especially relevant in the brain metastasis subgroup, where conclusions remain exploratory due to the limited patient numbers.

Third, while the study focused on survival outcomes, data on immune-related adverse events and quality of life were not available. As a result, we could not evaluate the safety profile of immunotherapy or perform a risk-benefit assessment across clinical subgroups (19).

Finally, the absence of molecular and immunologic biomarkers—such as PD-L1 expression and other recent protein targets, tumor mutation burden, or circulating immune signatures/DNA methylation profiles—precluded deeper mechanistic insights or personalized treatment stratification. Future prospective, multicenter studies incorporating biomarker data and patient-reported outcomes are warranted to validate these findings and further refine treatment strategies in real-world settings.

Conclusions

This real-world study supports the clinical utility of atezolizumab combined with EP chemotherapy as a first-line treatment for ES-SCLC. It is noteworthy that despite the inclusion of an older and more comorbid population than those typically enrolled in clinical trials (3,6), the addition of immunotherapy was associated with a survival benefit consistent with findings from key RCTs.

The observation of delayed PFS curve separation reinforces the distinctive kinetic profile of ICIs and emphasizes the importance of allowing sufficient time for treatment effects to manifest. Moreover, the limited efficacy observed in patients with brain metastases emphasizes the urgent need for further investigation into central nervous system immunodynamics and optimal management strategies for this subgroup.

Taken together, our results confirm the generalizability of immune-chemotherapy to complex real-world populations and provide practical insights for clinicians treating patients with ES-SCLC outside of controlled trial environments. Future research should prioritize prospective inter center validation, biomarker-guided patient selection, and a gaining a deeper understanding of immunotherapy responsiveness across clinically relevant subgroups.

Acknowledgments

We would like to express our sincere gratitude to Professor Fang Zhou, Professor Wei Li, and Professor Wei Sun for their invaluable guidance and support throughout the preparation and writing of this manuscript.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1049/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1049/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1049/prf

Funding: This study was funded by the Tianjin Key Medical Discipline (Specialty) Construction Project (No. TIYXZDXK-018A).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1049/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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Tianjin Chest Hospital (approval No. 2024KY-028-01), and the requirement for informed consent was waived due to the retrospective design and use of anonymized data.

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. Falchero L, Guisier F, Darrason M, et al. Long-term effectiveness and treatment sequences in patients with extensive stage small cell lung cancer receiving atezolizumab plus chemotherapy: Results of the IFCT-1905 CLINATEZO real-world study. Lung Cancer 2023;185:107379. [Crossref] [PubMed]
2. 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]
3. National Comprehensive Cancer Network. Small Cell Lung Cancer (Version 2.2022). NCCN Clinical Practice Guidelines in Oncology. Available online: https://www.nccn.org/professionals/physician_gls/pdf/sclc.pdf
4. Damiano P, Stefani A, Avancini A, et al. Real-world evidence in extensive disease small cell lung cancer: The missing piece of the puzzle. Crit Rev Oncol Hematol 2025;207:104618. [Crossref] [PubMed]
5. 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]
6. 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]
7. Paz-Ares L, Dvorkin M, Chen Y, et al. Durvalumab plus platinum-etoposide versus platinum-etoposide in first-line treatment of extensive-stage small-cell lung cancer (CASPIAN): a randomised, controlled, open-label, phase 3 trial. Lancet 2019;394:1929-39. [Crossref] [PubMed]
8. Paz-Ares L, Chen Y, Reinmuth N, et al. Durvalumab, with or without tremelimumab, plus platinum-etoposide in first-line treatment of extensive-stage small-cell lung cancer: 3-year overall survival update from CASPIAN. ESMO Open 2022;7:100408. [Crossref] [PubMed]
9. Khozin S, Abernethy AP, Zhi J. Real-world evidence and clinical trials in the era of precision oncology. JCO Precis Oncol 2019;3:PO.18.00286.
10. Oh DY, Rokutanda N, Żotkiewicz M, et al. Delayed Separation of Kaplan-Meier Curves is Commonly Observed in Studies of Advanced/Metastatic Solid Tumors Treated with Anti-PD-(L)1 Therapy: Systematic Review and Meta-Analysis. Target Oncol 2025;20:45-56. [Crossref] [PubMed]
11. Kumar V, Chaudhary N, Garg M, et al. Delayed immune-related events (DIRE) after immune checkpoint blockade: A real-world analysis. Oncologist 2020;25:747-55.
12. 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]
13. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivariate Behav Res 2011;46:399-424.
14. Choi HS, Park J, Lee SJ, et al. The impact of brain metastases on the efficacy of atezolizumab in extensive-stage small cell lung cancer: A real-world data analysis. Lung Cancer 2023;190:107595.
15. Yamada K, Saito H, Suzuki H. Immune checkpoint inhibitors for patients with small cell lung cancer and brain metastases: A multicenter analysis. Thorac Cancer 2023;14:351-8.
16. Xie Y, Liu X. Analysis of survival data under the Cox model using log-rank test and its application in clinical trials. Stat Med 2005;24:3811-25.
17. Zhou Y, Zhao J, Wang S, et al. Real-world outcomes of immune checkpoint inhibitor plus chemotherapy in extensive-stage small-cell lung cancer: A multicenter retrospective study in China. Front Oncol 2023;13:1183950.
18. Kumar V, Gupta A, Sharma S. Real-world outcomes of PD-1 inhibitors in extensive-stage small cell lung cancer: A comparative analysis with clinical trial data. Cancer Treat Rev 2023;50:102-11.
19. Qu J, Kalyani FS, Shen Q, et al. Efficacy and Safety of PD-L1 Inhibitors plus Chemotherapy versus Chemotherapy Alone in First-Line Treatment of Extensive-Stage Small-Cell Lung Cancer: A Retrospective Real-World Study. J Oncol 2022;2022:3645489. [Crossref] [PubMed]
20. Wu L, Saxena S, Awad M, et al. Real-world effectiveness of immunotherapy in metastatic lung cancer: an international multicenter analysis. ESMO Open 2022;7:100648.
21. Bourla AB, Meropol NJ. Bridging the divide between clinical research and clinical care in oncology: An integrated real-world evidence generation platform. Digit Health 2021;7:20552076211059975. [Crossref] [PubMed]

(English Language Editor: J. Gray)