A predictive model for immunotherapy in extensive-stage small cell lung cancer based on peripheral blood Authors: Jiacheng Wang, Heng Li, Lanjun Li, Zhiqing Liu, Hanting Zhao, Xiao Ma, Shaobin Qiu, Tianshun Wang, Gaofeng Li Published: 2026 ABSTRACT BackgroundExtensive-stage small cell lung cancer (ES-SCLC) has a dismal prognosis, with first-line immunochemotherapy benefiting only a minority of patients. Existing inflammatory biomarkers show inconsistent predictive utility, and accessible tools for individualized survival estimation remain lacking. This study aimed to develop and validate a nomogram incorporating the serum lactate dehydrogenase to albumin ratio (LAR) to predict overall survival (OS) and progression-free survival (PFS) in ES-SCLC patients receiving first-line immunotherapy.MethodsA retrospective cohort of 223 pathologically confirmed ES-SCLC patients treated at Yunnan Cancer Hospital (November 2021–August 2025) was analyzed. Ten inflammatory-metabolic markers [including LAR, systemic immune-inflammation index (SII), prognostic nutritional index (PNI)] and clinicopathological variables were extracted. Patients were randomly assigned in a 7:3 ratio to the training (n=156) and internal validation (n=67) cohorts. Optimal cutoff values for continuous variables were determined using the surv_cutpoint function. Independent prognostic factors were identified using univariate/multivariate Cox regression, and nomograms for OS and PFS were constructed. Model performance was evaluated via concordance index (C-index), calibration curves, time-dependent receiver operating characteristic (TimeROC), and decision curve analysis (DCA). Risk stratification was performed using median model-derived scores.ResultsMedian follow-up was 22 months. Multivariate analysis identified six independent OS predictors: LAR >6.27 [hazard ratio (HR) =2.794], adrenal metastasis (HR =3.725), PNI <50.5 (HR =0.622), smoking history (HR =1.596), white blood cell count (WBC) >8.30×109/L (HR=1.692), and age >58 years (HR =1.97). Eight independent PFS predictors: tumor-node-metastasis (TNM) stage IV disease (HR =1.407) LAR >6.36 (HR =1.482), SII <342.25 (HR =0.574), albumin (ALB) <42.3 g/L (HR =0.684), neutrophil count (NEUT) <2.64×109/L (HR =0.554), chlorine concentration <96 mmol/L (HR =0.613), male (HR =1.772) and liver metastasis (HR =1.762). The OS nomogram demonstrated superior discriminative ability compared with TNM staging (C-index: 0.728 vs. 0.519, P<0.001). The C-indices for the training and validation cohorts were 0.728 and 0.648 for OS, and 0.668 and 0.627 for PFS, respectively. Time-dependent area under the curve (AUC) for 12-month OS reached 0.793 (training) and 0.742 (validation). Calibration curves and DCA confirmed strong agreement and clinical utility. Low-risk patients (via median score) had significantly longer OS/PFS than high-risk patients (log-rank P<0.05).ConclusionsWe developed the first LAR-integrated nomogram for ES-SCLC immunotherapy outcomes, demonstrating robust predictive accuracy using routine peripheral blood parameters. This accessible tool enables individualized risk stratification and treatment optimization, particularly in resource-limited settings. Prospective multicenter validation is warranted to advance clinical adoption. Highlight box Key findings • This study established and validated the first nomogram incorporating the serum lactate dehydrogenase to albumin ratio (LAR) for predicting survival in patients with extensive-stage small cell lung cancer (ES-SCLC) treated with first-line immunotherapy. The model demonstrated robust discrimination [concordance index: 0.728 for overall survival (OS)], outperforming traditional tumor-node-metastasis (TNM) staging. What is known, and what is new? • Immunotherapy response varies significantly in ES-SCLC, and existing inflammatory biomarkers have demonstrated inconsistent predictive utility. • LAR was identified as a strong independent prognostic factor (hazard ratio =2.794 for OS). The model relies exclusively on routine peripheral blood parameters, offering a practical tool for real-world risk stratification without requiring complex molecular profiling. What is the implication, and what should change now? • Integrating LAR into clinical assessment allows for precise individual survival estimation and treatment intensity adjustment. Given its accessibility, this model has the potential to facilitate informed decision-making, especially in resource-limited settings. However, prospective, multicenter validation in diverse patient populations is essential to confirm these findings and drive clinical adoption. Introduction Background Extensive-stage small cell lung cancer (ES-SCLC) is a highly aggressive neuroendocrine tumor accounting for approximately 70% of SCLC cases at diagnosis (1). The 5-year survival rate is below 7% (2). ES-SCLC is characterized by rapid proliferation, early metastasis, and a poor prognosis. In recent years, immune checkpoint inhibitors (ICIs) combined with platinum-etoposide chemotherapy have emerged as the first-line standard treatment for ES-SCLC, extending median overall survival (OS) to 12–13 months (1,3). However, immunotherapy response rates vary considerably, with only about 15–20% of patients achieving long-term benefit (4,5). Recent studies highlight that SCLC exhibits transcriptional heterogeneity, but molecular subtyping remains clinically impractical (6,7). Thus, identifying accessible biomarkers for individualized prognosis is crucial. Traditional prognostic factors for SCLC include clinical stage, lactate dehydrogenase (LDH) levels, and metastatic patterns (8-10). Systemic inflammation and metabolic-nutritional imbalances also play pivotal roles in tumor progression and therapy resistance (6,11). Previous research has explored inflammatory markers such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), systemic immune-inflammation index (SII), and prognostic nutritional index (PNI) in SCLC outcomes, though their predictive efficacy remains controversial (12-14). For instance, some studies associate high NLR with poor survival, whereas others report no significant correlation (15,16). In this study, we aimed to identify predictive factors from inflammatory-metabolic markers, including the serum lactate dehydrogenase to albumin ratio (LAR), and clinicopathological variables derived from peripheral blood and clinical data. We then developed a nomogram to predict OS and progression-free survival (PFS) in patients with ES-SCLC who are receiving first-line immunotherapy. We present this article in accordance with the TRIPOD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-0918/rc). Methods Patient selection and data acquisition This retrospective study used electronic medical record data from Yunnan Cancer Hospital, and survival data were obtained through telephone follow-up surveys conducted by the hospital. It included 223 patients with pathologically confirmed ES-SCLC treated at Yunnan Cancer Hospital between November 2021 and August 2025. The study flow chart is shown in Figure 1. Inclusion criteria were as follows: (I) Pathologically confirmed ES-SCLC based on the Veterans Administration Lung Cancer Group (VALG) staging criteria; (II) Eastern Cooperative Oncology Group (ECOG) performance status of 0–2; (III) Completion of at least four cycles of first-line immunochemotherapy, according to the 2024 Chinese Society of Clinical Oncology (CSCO) guidelines; (IV) Availability of complete clinical, laboratory, and imaging data. The cutoff date for final follow-up was December 31, 2025. Patient selection flowchart Yunnan Cancer Hospital cohort. ES-SCLC, early-stage small cell lung cancer. Exclusion criteria included: (I) recurrent small cell lung cancer; (II) concomitant autoimmune disease or other malignancies; (III) active systemic inflammation or infection within one week before or after the initiation of treatment; (IV) severe dysfunction of the heart, liver, or kidneys; (V) long-term systemic corticosteroid use; (VI) treatment-related cytopenia requiring hematopoietic growth factor support; (VII) discontinuation of treatment or failure to initiate first-line immunochemotherapy after diagnosis; (VIII) fewer than four cycles of first-line immunochemotherapy. The Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 guidelines were used to define progressive disease (PD), stable disease (SD), partial response (PR), and complete response (CR) in this study. The primary endpoint was OS, defined as the time from diagnosis to death from any cause or last follow-up. The secondary endpoint was PFS, defined as the time from diagnosis to disease progression, or death. The cutoff date for data collection was December 31, 2025. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Yunnan Cancer Hospital (Nos. SLKYLX2026-173 and KYLX2026-199), and individual consent for this retrospective analysis was waived. This study extracted baseline clinical and pathological characteristics from patients meeting inclusion and exclusion criteria. Ten composite indicators were derived from laboratory parameters: NLR, PLR, direct neutrophil-to-lymphocyte ratio (dNLR), SII, lung immune prognosis index (LIPI), PNI, Gustave Roussy Immune Score (GRIm-Score), pan-immune inflammation value (PIV), LAR, and lymphocyte to monocyte ratio (LMR). The calculation methods for NLR, PLR, dNLR, SII, LIPI, PNI, GRIm-Score, PIV, LAR, and LMR are as follows: NLR = neutrophil count (NEUT, 109/L) / lymphocyte count (109/L); PLR = platelet count (109/L) / lymphocyte count (109/L); dNLR = NEUT (109/L) / [total white blood cell count (WBC, 109/L) − NEUT (109/L)]; SII = platelet count (109/L) × NEUT (109/L) / lymphocyte count (109/L). LIPI was divided into three subsets of scores: good, intermediate, and poor LIPI, based on the following cutoffs: dNLR ≤3 and LDH ≤ upper limit of normal (ULN), dNLR >3 or LDH > ULN, and dNLR >3 and LDH > ULN. PNI = serum albumin (ALB, g/L) + 5 × lymphocyte count (109/L) GRIm-Score is calculated by the following three variables: LDH (within normal range: 0; > ULN of each center, 250 U/L in our hospital: +1), ALB (≥35 g/L: 0; <35 g/L: +1), and NLR (≤6: 0; >6: +1). Patients were divided into two groups: high group (score 2 or 3) and low group (score 0 or 1). PIV = NEUT (109/L) × platelet count (109/L) × monocyte count / lymphocyte count (109/L) LAR = serum LDH concentration (U/L) / serum ALB (g/L) LMR = lymphocyte count (109/L) / monocyte count (109/L) Statistical analysis Nomogram development and construction Data analysis was performed using SPSS version 26.0 and R version 4.4.2 statistical software. The surv_cutpoint function from the survminer package in R version 4.4.2 was used to determine cut-off values for continuous variables. These continuous variables were converted into binary variables based on the cut-off values. Patients were then grouped into high- and low-level cohorts to construct the nomogram. Continuous variables were expressed as mean ± standard deviation or median (interquartile range), depending on data distribution. Intergroup comparisons were performed using Student’s t-test or Wilcoxon rank-sum test, depending on data distribution. Categorical variables were presented as frequencies (percentages), with intergroup comparisons performed using chi-square test or Fisher’s exact test. Univariate and multivariate survival analyses for OS and PFS were conducted using Cox proportional hazards models. Variables with P<0.05 in univariate analysis were entered into a multivariable Cox proportional hazards model. A backward stepwise selection procedure based on the Akaike information criterion (AIC) was then applied to identify the independent prognostic factors for OS and PFS, respectively. Independent variables with statistical significance in the multivariate analysis were selected to construct a nomogram model, predicting the 12-, 24-, and 36-month survival probabilities for OS and the 6-, 12-, and 24-month progression-free probabilities for PFS. The predictive performance was evaluated using calibration curves and receiver operating characteristic (ROC) curve analysis. Nomogram validation To evaluate the predictive accuracy of survival models, the concordance index (C-index) was estimated using 500 bootstrap resamples. The optimal model was selected by choosing the model with the highest bootstrap C-index in the validation cohort. To further evaluate model performance, we plotted time-dependent receiver operating characteristic (TimeROC) curves, calibration curves, and decision curve analysis (DCA) plots. After model construction, individual patient risk scores were calculated by applying each patient’s variables to the model. Patients were then categorized into high-risk (score > median) and low-risk (score ≤ median) groups based on the median risk score across the entire cohort. Survival probabilities for both groups were estimated using the Kaplan-Meier method, and differences between survival curves were assessed using log-rank tests. All statistical analyses employed two-tailed tests with P<0.05 as the threshold for statistical significance. All analyses were performed using the R 4.2.2 platform, with specific packages utilized for various tasks. Sample size estimation was performed using the “pmsampsize” package; baseline characteristics were described using the “comparegroups” package with automatic variable identification; Cox regression was conducted using the coxph function from the “survival” package; TimeROC curves were generated using the “survivalROC” package; nomograms were created with the “rms” package; time-dependent area under the curve (AUC) curves were estimated via the “pec” package; calibration curves were assessed using the “calibrate” function from the “rms” package; clinical decision curves were plotted using the “dcurves” package; and risk-stratified Kaplan-Meier curves were produced using the “survminer” package. Results Patient characteristics Based on the inclusion criteria of this study, 223 eligible patients were enrolled. As of December 31, 2025, the median follow-up duration was 22 months. The median PFS was 6.4 months, and the median OS was 11.67 months. At the data cutoff, 161 patients (72.20%) had died, and 9 patients were lost to follow-up, representing a loss-to-follow-up rate of 4.03%. Among enrolled patients, 91.03% were male and 8.97% were female. Furthermore, 67.71% had a history of smoking. The pre-treatment proportions of patients with liver, brain, adrenal, bone, pleural, and lung metastases were 22.87%, 19.73%, 14.35%, 17.19%, and 9.87%, respectively. Regarding immunotherapy strategies, 55.16% of patients received Serplulimab, 18.83% received tislelizumab, 16.59% received atezolizumab, and 9.42% received other immunotherapies. Detailed statistical analysis results are presented in Table 1. A total of 223 patients were randomly assigned in a 7:3 ratio to the training cohort (n=156) and the independent internal validation cohort (n=67). Tables 2,3 present patient characteristics for the two cohorts in the OS and PFS groups, respectively. The median OS in the OS group was 10.70 months for the training cohort and 12.05 months for the validation cohort. At the last follow-up, the mortality rates in the training and validation cohorts were 77.61% and 69.87%, respectively. The median PFS in the PFS cohort was 6.93 months in the training cohort and 5.53 months in the validation cohort. At the last follow-up, the progression rates in the training and validation cohorts were 91.03% and 92.54%, respectively. Table 1Clinical characteristics of the 223 ES-SCLC patientsCharacteristicValues (n=223)Sex Female20 (8.97) Male203 (91.03)Age, years <5858 (26.01) ≥58165 (73.99)The AJCC 8th TNM staging system 360 (26.91) 4163 (73.09)Smoking history No72 (32.29) Yes151 (67.71)Family history of cancer No215 (96.41) Yes8 (3.59)Hypertension No158 (70.85) Yes65 (29.15)Coronary heart disease No204 (91.48) Yes19 (8.52)Diabetes No199 (89.24) Yes24 (10.76)Bone metastasis No184 (82.51) Yes39 (17.49)Liver metastasis No172 (77.13) Yes51 (22.87)Brain metastasis No179 (80.27) Yes44 (19.73)Suprarenal metastasis No191 (85.65) Yes32 (14.35)Lung metastasis No172 (77.13) Yes51 (22.87)Pleural metastasis No201 (90.13) Yes22 (9.87)Superior vena cava syndrome No207 (92.83) Yes16 (7.17)Height, cm1.67 (1.62–1.70)Weight, kg65.00 (56.00–71.00)BMI, kg/m222.99 (21.10–25.33)Immunotherapy Serplulimab123 (55.16) Tislelizumab42 (18.83) Atezolizumab37 (16.59) Other21 (9.42)Data are presented as n (%) or median (interquartile range). AJCC, American Joint Committee on Cancer; BMI, body mass index; ES-SCLC, early-stage small cell lung cancer; TNM, tumor-node-metastasis. Clinical characteristics of the 223 ES-SCLC patients Data are presented as n (%) or median (interquartile range). AJCC, American Joint Committee on Cancer; BMI, body mass index; ES-SCLC, early-stage small cell lung cancer; TNM, tumor-node-metastasis. Table 2Comparison of baseline characteristics between OS training and validation cohorts of ES-SCLCVariablesTotal (n=223)Test (n=67)Train (n=156)StatisticPSex1.6450.2 Female20 [9]3 [4]17 [11] Male203 [91]64 [96]139 [89]Age, years0.1280.72 <5858 [26]19 [28]39 [25] ≥58165 [74]48 [72]117 [75]The AJCC 8th TNM staging system0.6930.40 360 [27]15 [22]45 [29] 4163 [73]52 [78]111 [71]Smoking history3.670.055 No72 [32]15 [22]57 [37] Yes151 [68]52 [78]99 [63]Family history of cancerFisher>0.99 No215 [96]65 [97]150 [96] Yes8 [4]2 [3]6 [4]Hypertension0.4010.53 No158 [71]45 [67]113 [72] Yes65 [29]22 [33]43 [28]Coronary heart disease0.1710.68 No204 [91]60 [90]144 [92] Yes19 [9]7 [10]12 [8]Diabetes0>0.99 No199 [89]60 [90]139 [89] Yes24 [11]7 [10]17 [11]Bone metastasis1.5310.22 No184 [83]59 [88]125 [80] Yes39 [17]8 [12]31 [20]Liver metastasis3.2420.07 No172 [77]46 [69]126 [81] Yes51 [23]21 [31]30 [19]Brain metastasis0.9960.32 No179 [80]57 [85]122 [78] Yes44 [20]10 [15]34 [22]Suprarenal metastasis0.2160.64 No191 [86]59 [88]132 [85] Yes32 [14]8 [12]24 [15]Lung metastasis0>0.99 No172 [77]52 [78]120 [77] Yes51 [23]15 [22]36 [23]Pleural metastasis0>0.99 No201 [90]60 [90]141 [90] Yes22 [10]7 [10]15 [10]Superior vena cava syndromeFisher0.26 No207 [93]60 [90]147 [94] Yes16 [7]7 [10]9 [6]Sodium concentration, mmol/L139 [137–141]139 [136–141]139 [137–141]4,6680.20Chlorine concentration, mmol/L103 [100–105]102 [100–104]104 [100.75–105]4,410.50.06Height, m1.67 [1.62–1.7]1.66 [1.62–1.7]1.67 [1.62–1.71]4,772.50.30Weight, kg65 [56–71]65 [58–72]65 [56–71]5,3680.75WBC, 109/L0>0.99 <8.3136 [61]41 [61]95 [61] ≥8.387 [39]26 [39]61 [39]NEUT, 109/L0.1280.72 <6.26165 [74]48 [72]117 [75] ≥6.2658 [26]19 [28]39 [25]LYMPH, 109/L0.330.57 <1.96152 [68]48 [72]104 [67] ≥1.9671 [32]19 [28]52 [33]MONO, 109/L0.5190.47 <0.4045 [20]16 [24]29 [19] ≥0.40178 [80]51 [76]127 [81]PLT, 109/L0.2190.64 <351184 [83]57 [85]127 [81] ≥35139 [17]10 [15]29 [19]ALB, g/L0>0.99 <40.950 [22]15 [22]35 [22] ≥40.9173 [78]52 [78]121 [78]LDH, U/L0.9290.33 <257129 [58]35 [52]94 [60] ≥25794 [42]32 [48]62 [40]BMI, kg/m222.99 [21.1–25.33]23.14 [21.38–25.78]22.85 [21.09–25.13]5,622.50.37NLR1.0270.31 <2.1574 [33]26 [39]48 [31] ≥2.15149 [67]41 [61]108 [69]PLR0.0940.76 <151.38115 [52]33 [49]82 [53] ≥151.38108 [48]34 [51]74 [47]dNLR0.60.44 <1.6383 [37]28 [42]55 [35] ≥1.63140 [63]39 [58]101 [65]SII0.4460.50 <741.66134 [60]43 [64]91 [58] ≥741.6689 [40]24 [36]65 [42]LIPI0.7530.69 095 [43]27 [40]68 [44] 1101 [45]30 [45]71 [46] 227 [12]10 [15]17 [11]PNI0.6040.44 <50.5070 [31]24 [36]46 [29] ≥50.50153 [69]43 [64]110 [71]GRIm-Score0>0.99 Low group202 [91]61 [91]141 [90] High group21 [9]6 [9]15 [10]ALI00.99 <166.7475 [34]22 [33]53 [34] ≥166.74148 [66]45 [67]103 [66]PIV0.1560.69 <311.50107 [48]34 [51]73 [47] ≥311.50116 [52]33 [49]83 [53]LAR1.8230.18 <6.27130 [58]34 [51]96 [62] ≥6.2793 [42]33 [49]60 [38]LMR0.1060.74 <3.83145 [65]42 [63]103 [66] ≥3.8378 [35]25 [37]53 [34]Immunotherapy1.090.78 Serplulimab123 [55]34 [51]89 [57] Tislelizumab42 [19]15 [22]27 [17] Atezolizumab37 [17]11 [16]26 [17] Others21 [9]7 [10]14 [9]Data are presented as n [%] or median [interquartile range]. AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; OS, overall survival; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Comparison of baseline characteristics between OS training and validation cohorts of ES-SCLC Data are presented as n [%] or median [interquartile range]. AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; OS, overall survival; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Table 3Comparison of baseline characteristics between PFS training and validation cohorts of ES-SCLCVariablesTotal (n=223)Test (n=67)Train (n=156)StatisticPSex0.5950.44 Female20 [9]4 [6]16 [10] Male203 [91]63 [94]140 [90]Age, years0.0090.92 <5021 [9]7 [10]14 [9] ≥50202 [91]60 [90]142 [91]The AJCC 8th TNM staging system0.030.86 360 [27]17 [25]43 [28] 4163 [73]50 [75]113 [72]Smoking history0.9570.33 No72 [32]18 [27]54 [35] Yes151 [68]49 [73]102 [65]Family history of cancerFisher0.24 No215 [96]63 [94]152 [97] Yes8 [4]4 [6]4 [3]Hypertension2.5530.11 No158 [71]42 [63]116 [74] Yes65 [29]25 [37]40 [26]Coronary heart disease0.1710.68 No204 [91]60 [90]144 [92] Yes19 [9]7 [10]12 [8]Diabetes0.650.42 No199 [89]62 [93]137 [88] Yes24 [11]5 [7]19 [12]Bone metastasis1.1450.28 No184 [83]52 [78]132 [85] Yes39 [17]15 [22]24 [15]Liver metastasis0.4020.53 No172 [77]54 [81]118 [76] Yes51 [23]13 [19]38 [24]Brain metastasis0.0110.92 No179 [80]53 [79]126 [81] Yes44 [20]14 [21]30 [19]Suprarenal metastasis0.7760.38 No191 [86]60 [90]131 [84] Yes32 [14]7 [10]25 [16]Lung metastasis1.2210.27 No172 [77]48 [72]124 [79] Yes51 [23]19 [28]32 [21]Pleural metastasis0.2960.59 No201 [90]62 [93]139 [89] Yes22 [10]5 [7]17 [11]Superior vena cava syndromeFisher>0.99 No207 [93]62 [93]145 [93] Yes16 [7]5 [7]11 [7]Sodium concentration, mmol/L0.4590.498 <13523 [10]5 [7]18 [12] ≥135200 [90]62 [93]138 [88]Chlorine concentration, mmol/L0.0680.79 <9620 [9]5 [7]15 [10] ≥96203 [91]62 [93]141 [90]Height, m0.2740.6 <1.70139 [62]44 [66]95 [61] ≥1.7084 [38]23 [34]61 [39]Weight, kg0.2060.65 <64100 [45]28 [42]72 [46] ≥64123 [55]39 [58]84 [54]WBC, 109/L0.3560.55 <4.8326 [12]6 [9]20 [13] ≥4.83197 [88]61 [91]136 [87]NEUT, 109/L0>0.99 <2.6420 [9]6 [9]14 [9] ≥2.64203 [91]61 [91]142 [91]LYMPH, 109/L0.4620.50 <1.96152 [68]43 [64]109 [70] ≥1.9671 [32]24 [36]47 [30]MONO, 109/L0.0010.97 <0.87195 [87]58 [87]137 [88] ≥0.8728 [13]9 [13]19 [12]PLT, 109/L2.2280.14 <17430 [13]13 [19]17 [11] ≥174193 [87]54 [81]139 [89]ALB, g/L0.0130.91 <42.377 [35]24 [36]53 [34] ≥42.3146 [65]43 [64]103 [66]LDH, U/L0>0.99 <258130 [58]39 [58]91 [58] ≥25893 [42]28 [42]65 [42]BMI, kg/m21.7810.18 <26.77189 [85]53 [79]136 [87] ≥26.7734 [15]14 [21]20 [13]NLR1.6220.20 <1.4220 [9]9 [13]11 [7] ≥1.42203 [91]58 [87]145 [93]PLR0.8470.36 <82.9928 [13]11 [16]17 [11] ≥82.99195 [87]56 [84]139 [89]dNLR2.4030.12 <1.1224 [11]11 [16]13 [8] ≥1.12199 [89]56 [84]143 [92]SII1.810.18 <342.2531 [14]13 [19]18 [12] ≥342.25192 [86]54 [81]138 [88]LIPI0.2490.88 095 [43]29 [43]66 [42] 1101 [45]31 [46]70 [45] 227 [12]7 [10]20 [13]PNI0.1990.66 <46.941 [18]14 [21]27 [17] ≥46.9182 [82]53 [79]129 [83]GRIm-Score0.0090.92 Low group202 [91]60 [90]142 [91] High group21 [9]7 [10]14 [9]ALI0>0.99 <130.5142 [19]13 [19]29 [19] ≥130.51181 [81]54 [81]127 [81]PIV1.1590.28 <502.67140 [63]38 [57]102 [65] ≥502.6783 [37]29 [43]54 [35]LAR0>0.99 <6.36137 [61]41 [61]96 [62] ≥6.3686 [39]26 [39]60 [38]LMR0.0650.80 <2.7181 [36]23 [34]58 [37] ≥2.71142 [64]44 [66]98 [63]Immunotherapy8.0250.045 Serplulimab123 [55]28 [42]95 [61] Tislelizumab42 [19]14 [21]28 [18] Atezolizumab37 [17]16 [24]21 [13] Others21 [9]9 [13]12 [8]Data are presented as n [%]. AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; PFS, progression-free survival; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Comparison of baseline characteristics between PFS training and validation cohorts of ES-SCLC Data are presented as n [%]. AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; PFS, progression-free survival; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Survival analysis The independent predictors for OS and PFS listed below were derived from the final multivariable models after the stepwise selection process. Prognosis analysis of OS in ES-SCLC patients: univariate analysis revealed that 10 pretreatment parameters were negatively correlated with OS (as shown in Table 4), including SII, NLR, age, WBC, liver metastasis, smoking history, LDH, LIPI, LAR, and adrenal metastasis. Three pretreatment parameters were positively correlated with OS: serum ALB, PNI, and LMR. Multivariate analysis identified six independent risk factors for OS: baseline LAR >6.27 (HR =2.794, P<0.001), baseline adrenal metastases (HR =3.725, P<0.001), PNI <50.5 (HR =0.622, P=0.03), smoking history (HR =1.596, P=0.03), baseline WBC >8.30×109/L (HR =1.692, P=0.01), and age >58 years (HR =1.97, P=0.007). The prognostic analysis of PFS in ES-SCLC patients is presented in Table 5. Univariate analysis revealed seven pretreatment parameters positively correlated with PFS: LIPI score, LDH level, immunotherapy status, LAR status, tumor-node-metastasis (TNM) stage, liver metastasis, and sex. Six pretreatment parameters were negatively correlated with PFS: NEUT, serum chloride concentration, SII, PNI, advanced lung cancer inflammation index (ALI), and serum ALB. Multivariate analysis ultimately identified eight independent risk factors for PFS: male sex (HR =1.772, P=0.08), baseline liver metastasis (HR =1.762, P=0.007), LAR >6.36 (HR =1.482, P=0.04), TNM stage VIII (HR =1.407, P=0.11), ALB <42.30 g/L (HR =0.684, P=0.051), serum chloride <96 mmol/L (HR =0.631, P=0.09), SII <342.25 (HR =0.574, P=0.056), and NEUT <2.64×109/L (HR =0.554, P=0.057). Table 4Cox regression analysis of clinical characteristics associated with overall survival in the training cohort of ES-SCLCCharacteristicsUnivariate analysisMultivariate analysisHR95% CIPHR95% CIPImmunotherapy SerplulimabRef. Tislelizumab1.3750.779–2.4270.27 Atezolizumab1.0060.613–1.6490.98 Others1.0370.543–1.9830.91LMR0.5780.379–0.8810.01LAR2.4451.668–3.585<0.0012.7941.866–4.184<0.001LYMPH1.430.966–2.1170.07PIV1.3590.929–1.9870.11ALI0.7230.49–1.0680.10GRIm-Score1.5940.891–2.8520.12NEUT1.4340.943–2.1810.09PNI0.5480.366–0.8190.0030.6220.409–0.9460.03LIPI 0Ref. 11.4910.992–2.240.055 22.3571.303–4.2640.005SII1.4861.018–2.1690.04dNLR1.470.978–2.2080.06PLR1.3780.945–2.010.10NLR1.6371.06–2.5290.03ALB0.4530.291–0.704<0.001BMI1.0180.96–1.080.55LDH2.1981.504–3.211<0.001PLT1.3330.832–2.1370.23MONO1.3110.796–2.160.29Diabetes0.9510.494–1.8310.88WBC1.6741.141–2.4560.0081.6921.134–2.5260.01Height1.0480.065–17.0190.97Weight1.0040.986–1.0210.68Chlorine concentration1.0130.965–1.0630.60Sodium concentration1.0180.966–1.0720.51Bone metastasis1.1770.765–1.8110.46Superior vena cava syndrome1.7670.889–3.5120.10Lung metastasis1.2750.827–1.9640.27Pleural metastasis1.3550.771–2.3820.29Brain metastasis1.110.702–1.7540.65Suprarenal metastasis2.5011.498–4.175<0.0013.7252.17–6.395<0.001Liver metastasis1.7031.087–2.6680.02Age1.6491.039–2.6170.031.971.201–3.2320.007Coronary heart disease0.6860.3–1.5680.37Hypertension1.3420.89–2.0240.16Smoking history1.7571.168–2.6440.0071.5961.043–2.4450.03Family history of cancer0.5770.183–1.8220.35The AJCC 8th TNM staging system1.5050.967–2.3430.07Sex1.860.905–3.8260.09AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; CI, confidence interval; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; HR, hazard ratio; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Cox regression analysis of clinical characteristics associated with overall survival in the training cohort of ES-SCLC AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; CI, confidence interval; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; HR, hazard ratio; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Table 5Cox regression analysis of clinical characteristics associated with PFS in the training cohort of ES-SCLCCharacteristicsUnivariate analysisMultivariate analysisHR95% CIPHR95% CIPThe AJCC 8th TNM staging system1.8021.227–2.6470.0031.4070.924–2.1430.11Immunotherapy SerplulimabRef. Tislelizumab1.6211.031–2.5490.04 Atezolizumab0.9920.601–1.640.98 Others1.1020.601–2.0220.75PIV1.1450.811–1.6180.44LMR0.8310.59–1.1720.29LAR1.661.17–2.3570.0051.4821.011–2.1710.044Height1.3370.952–1.8780.09Weight1.310.938–1.8310.11ALI0.6330.418–0.9590.03GRIm-Score1.220.686–2.1690.50PNI0.6020.393–0.9220.02LIPI 0Ref. 11.5571.085–2.2340.02 21.5580.906–2.6790.11SII0.5730.343–0.9570.030.5740.325–1.0140.056ALB0.6450.454–0.9150.010.6840.467–1.0010.051dNLR0.6180.332–1.150.13PLR0.6440.381–1.0890.10NLR0.7080.36–1.3940.32BMI0.7140.429–1.1870.19LDH1.5741.115–2.2220.01WBC0.6530.4–1.0660.09PLT0.7330.439–1.2230.23MONO1.4330.88–2.3320.15LYMPH1.3020.907–1.8680.15NEUT0.470.269–0.8210.0080.5540.301–1.0190.057Lung metastasis1.050.7–1.5760.81Sodium concentration0.9850.58–1.6750.96Chlorine concentration0.5660.329–0.9740.040.6130.346–1.0860.09Suprarenal metastasis1.0650.674–1.6820.79Superior vena cava syndrome1.4720.791–2.7390.22Pleural metastasis1.0850.634–1.8580.76Diabetes0.8740.524–1.4570.60Brain metastasis1.2950.854–1.9620.22Bone metastasis0.990.637–1.5380.96Liver metastasis1.951.331–2.856<0.0011.7621.164–2.6670.007Coronary heart disease0.8220.443–1.5240.53Hypertension1.3490.924–1.970.12Family history of cancer0.6340.201–1.9970.44Smoking history1.230.865–1.7470.25Age1.8920.988–3.6230.054Sex2.3261.249–4.3330.0081.7720.925–3.3950.08AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; CI, confidence interval; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; HR, hazard ratio; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; PFS, progression-free survival; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Cox regression analysis of clinical characteristics associated with PFS in the training cohort of ES-SCLC AJCC, American Joint Committee on Cancer; ALB, albumin; ALI, advanced lung cancer inflammation index; BMI, body mass index; CI, confidence interval; dNLR, direct neutrophil-to-lymphocyte ratio; ES-SCLC, early-stage small cell lung cancer; GRIm-Score, Gustave Roussy Immune Score; HR, hazard ratio; LAR, lactate dehydrogenase to albumin ratio; LDH, lactate dehydrogenase; LIPI, lung immune prognosis index; LMR, lymphocyte to monocyte ratio; LYMPH, lymphocyte count; MONO, monocyte count; NEUT, neutrophil count; NLR, neutrophil-to-lymphocyte ratio; PFS, progression-free survival; PIV, pan-immune inflammation value; PLR, platelet-to-lymphocyte ratio; PLT, platelet count; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis; WBC, white blood cell count. Construction and evaluation of nomogram Nomogram construction and validation Using data from patients diagnosed with ES-SCLC, we constructed nomograms to predict OS (Figure 2) and PFS (Figure 3). The results are reported as the probabilities of OS at 12, 24, and 36 months and PFS at 6, 12, and 24 months. In the training cohort, the nomogram C-indices for OS and PFS were 0.728 [95% confidence interval (CI): 0.681–0.775] and 0.668 (95% CI: 0.623–0.713), respectively. In the validation cohort, the nomogram C-indices for OS and PFS were 0.648 (95% CI: 0.556–0.740) and 0.627 (95% CI: 0.545–0.709), respectively. The calibration curves of the nomograms are summarized in Figure 4 (for OS) and Figure 5 (for PFS). Calibration curves for the training cohort (Figures 4A-4C,5A-5C) and validation cohort (Figures 4D-4F,5D-5F) demonstrated good agreement between model predictions and observed values for 12-, 24-, and 36-month OS and 6-, 12-, and 24-month PFS probabilities. The C-index of the OS nomogram [0.728 (95% CI: 0.681–0.775)] was significantly higher than that of the American Joint Committee on Cancer (AJCC) 8th edition TNM staging system [0.519 (95% CI: 0.468–0.569)], indicating that our proposed nomogram possesses strong prognostic capability. The TimeROC curves of the OS and PFS nomograms are illustrated in Figures 6,7, respectively. In the training cohort, the TimeROC curves displayed AUC values of 0.793, 0.822, and 0.746 for predicting 12-, 24-, and 36-month OS, respectively (Figure 6A). The AUC values for predicting 6-, 12-, and 24-month PFS were 0.748, 0.834, and 0.799 (Figure 7A). In the validation cohort, the AUC values for predicting 12-, 24-, and 36-month OS were 0.742, 0.875, and 0.844, respectively (Figure 6B). The AUC values for predicting 6-, 12-, and 24-month PFS were 0.701, 0.875, and 0.860, respectively (Figure 7B). The AUC results demonstrated the nomogram’s high predictive value. DCA accurately assessed the model’s clinical utility and net benefit. Findings indicate the nomogram model exhibits significant clinical value in predicting OS (Figure 8A-8F) and PFS (Figure 9A-9F) for ES-SCLC patients. Nomogram predicted 12-, 24-, and 36-month OS for ES-SCLC patients. ALB, albumin; ES-SCLC, early-stage small cell lung cancer; LAR, lactate dehydrogenase to albumin ratio; OS, overall survival; PNI, prognostic nutritional index; WBC, white blood cell count. Nomogram predicted 6-, 12-, and 24-month PFS for ES-SCLC patients. ALB, albumin; ES-SCLC, early-stage small cell lung cancer; LAR, lactate dehydrogenase to albumin ratio; NEUT, neutrophil count; PFS, progression-free survival; SII, systemic immune-inflammation index; TNM, tumor-node-metastasis. Calibration curves for predicting 12-, 24-, and 36-month OS of ES-SCLC patients in the training cohort (A-C) and the validation cohort (D-F). The x-axis denotes survival predicted by the nomogram and the y-axis represents actual OS as measured by Kaplan-Meier analysis. ES-SCLC, early-stage small cell lung cancer; OS, overall survival. Calibration curves for predicting 6-, 12-, and 24-month PFS of ES-SCLC patients in the training cohort (A-C) and the validation cohort (D-F). The x-axis denotes survival predicted by the nomogram and the y-axis represents actual PFS as measured by Kaplan-Meier analysis. ES-SCLC, early-stage small cell lung cancer; PFS, progression-free survival. Time-dependent ROC curves for OS at 12-, 24-, and 36-month in the training cohort (A) and the validation cohort (B). AUC, area under the curve; CI, confidence interval; OS, overall survival; ROC, receiver operating characteristic. Time-dependent ROC curves of PFS at 6-, 12-, and 24-month in the training cohort (A) and the validation cohort (B). AUC, area under the curve; CI, confidence interval; PFS, progression-free survival; ROC, receiver operating characteristic. The DCA of the OS nomogram at (A,D) 12-, (B,E) 24-, and (C,F) 36-month in the training cohort (A-C) and the validation cohort (D-F). DCA, decision curve analysis; OS, overall survival. The DCA of the PFS nomogram at (A,D) 6, (B,E) 12, and (C,F) 24 months in the training cohort (A-C) and the validation cohort (D-F). DCA, decision curve analysis; PFS, progression-free survival. Risk stratification predictive model development Using the median risk score derived from the developed OS model, ES-SCLC patients were stratified into high- and low-risk groups. Kaplan-Meier survival curves for both training and validation cohorts demonstrated a significantly longer median OS in the low-risk group than in the high-risk group, with statistically significant differences in both datasets (P<0.05, Figure 10). Median risk scores derived from the developed PFS model were used to similarly stratify ES-SCLC patients into high- and low-risk groups. Kaplan-Meier survival curves for both training and validation cohorts demonstrated a significantly longer median PFS in the low-risk group than in the high-risk group, with statistically significant differences observed in both datasets (P<0.05, Figure 11). Survival analysis of patients in different risk groups within the training cohort (A) and validation cohort (B) of the OS model. OS, overall survival. Survival analysis of patients in different risk groups within the training cohort (A) and validation cohort (B) in the PFS model. PFS, progression-free survival. Discussion LAR is a sensitive biomarker that reflects metabolic reprogramming and chronic inflammation (17). Recent studies have shown that elevated LAR is significantly associated with immunotherapy resistance in solid tumors such as gastric cancer (18) and colorectal cancer (19), and its predictive efficacy is superior to that of a single indicator. In the field of non-small cell lung cancer (NSCLC), LAR has been shown to independently predict the survival benefit of programmed cell death protein 1 (PD-1) inhibitors combined with chemotherapy (17,20,21), but its application in small cell lung cancer is still in its early stages. Although predictive tools for NSCLC have been developed [e.g., Shao et al. (17)], specific models for ES-SCLC remain scarce. We therefore developed and validated two nomograms that incorporate the inflammatory marker LAR in combination with other clinical variables to predict OS and PFS in this cohort. This is the first real-world study to use LAR to predict the efficacy of first-line immunotherapy in ES-SCLC. Both the OS and PFS nomograms included LAR as a key determinant. The OS model comprised six factors (including adrenal metastasis and PNI), whereas the PFS model consisted of eight factors (including liver metastasis and SII). Elevated lactate levels establish an immunosuppressive microenvironment by inhibiting dendritic cell maturation and promoting regulatory T cell infiltration (22,23), while low ALB levels diminish the response to immunotherapy (24). LAR thresholds (>6.27 for OS and >6.36 for PFS) were identified as independent risk factors, demonstrating the prognostic superiority of this marker over conventional metrics such as NLR (12) in patients with ES-SCLC. Furthermore, both the SII and the PNI have been recognized as readily available and promising prognostic markers for patients with SCLC undergoing immunotherapy, reflecting systemic inflammatory status and nutritional-immunological balance, respectively (25-27). Consistent with their established roles, our study further evaluated their prognostic value in the context of first-line chemoimmunotherapy for ES-SCLC. We demonstrated that an SII ≥342.25 was associated with shorter PFS (HR =1.482), while a PNI ≥50.50 was associated with improved OS (HR =0.622). Reinforcing the concept that specific metastatic patterns critically influence immunotherapy efficacy in ES-SCLC, our study substantiates that baseline liver and adrenal metastases are robust independent predictors of poor prognosis. These findings are corroborated by a growing body of evidence, especially regarding liver metastases, which have been consistently associated with inferior survival outcomes in patients receiving first-line chemoimmunotherapy (28-30). DCA further validated the clinical utility of our nomogram, demonstrating a favorable net benefit across a wide range of threshold probabilities (>10%). Complemented by a robust AUC of 0.793 for predicting 12-month OS, these findings underscore the model’s capacity to effectively stratify patients and facilitate clinical decision-making in ES-SCLC. All indicators were derived from routine blood tests, allowing for rapid stratification in resource-constrained environments, which aligns with Iliesiu et al. (31)’s focus on easily accessible inflammatory biomarkers. DCA demonstrated the model’s robust net benefit for predicting 12-month survival and 6-month progression, underscoring its utility in short-term clinical decision-making. While the range of threshold probabilities for long-term outcomes was relatively narrow, this reflects the inherent complexity of predicting long-term survival in ES-SCLC (32,33). Interestingly, we found that the optimal cutoff values for LAR (>6.27/6.36), PNI (<50.5), and SII (<342.25) in our cohort differ from those reported in other recent studies. For instance, the optimal LAR cutoff for ES-SCLC prognosis was identified as 6.537 in a 2026 study using ROC analysis (34), differing from our values (>6.27 for OS, >6.36 for PFS). Similarly, PNI cutoffs in SCLC patients receiving immunotherapy have been reported as 49.43 (35) and 50.60 (36) in studies from 2025. The SII threshold in our cohort (<342.25) is notably distinct from values reported in other lung cancer settings, such as 760.6 for limited-stage SCLC (27) and 1,172 for advanced NSCLC (37) receiving chemoimmunotherapy. This heterogeneity, stemming from cohort and methodological differences, is a common challenge in biomarker research (38). Crucially, this variability in specific numerical thresholds does not invalidate the generalizable core finding: the strong prognostic concept of these readily available indicators for metabolic-inflammatory and nutritional-immunological states. To translate this into a robust clinical tool, we constructed a multivariate nomogram that integrates these markers into a composite, continuous risk score. This approach, aligning with recent efforts in ES-SCLC prognostication (39), mitigates dependency on any single, population-specific cutoff and provides a more stable prognostic estimate. Future efforts should focus on the external validation and calibration of this nomogram in multi-center cohorts to confirm its broad applicability. Advancing standardized, data-driven methodologies for risk stratification and exploring the value of dynamic marker monitoring will further enhance the clinical utility of such prognostic models (40). Current immunotherapy prediction models are evolving towards the integration of multimodal biomarkers, such as Han et al. (41), who used Delta radiomics to capture the spatiotemporal heterogeneity of tumors. This study demonstrates that peripheral blood-based biomarkers (such as LAR) can serve as a valuable adjunct to multimodal biomarkers. Future research can investigate integrative models of radiomics and blood biomarkers to improve predictive accuracy (42,43). There are several limitations in this study. First, it is retrospective, and prospective studies are required to validate the nomogram for ES-SCLC prognosis. Second, this study’s requirement for completing at least four cycles of therapy may limit the immediate generalizability (external validity) of our model, as it excludes the clinically significant subgroup of patients with early treatment failure. Future validation in broader, unselected real-world cohorts is essential to confirm the model’s utility across the full clinical spectrum. Third, it lacks certain laboratory data and detailed clinical history, which may introduce bias. Third, multicenter validation is needed to assess and compare the effectiveness of immunotherapy drugs. Finally, the determination and generalizability of optimal cutoff values for continuous inflammatory-metabolic markers such as LAR, PNI, and SII warrant careful consideration. The cutoff values identified in our study (e.g., LAR >6.27 for OS) are data-driven and specific to our cohort’s distribution, statistical method (surv_cutpointfunction), and clinical context. Variability in reported cutoffs across different studies is a recognized challenge, often stemming from differences in patient demographics, disease stages, sample sizes, and the statistical techniques used for dichotomization (e.g., X-tile, ROC curve maximization, or median split). This variability highlights that our specific cutoff points may not be directly portable to all populations without validation or recalibration. It underscores the importance of external validation studies to assess the transportability of our model and its associated thresholds. For broader clinical application, future research should aim to establish standardized, clinically validated cutoff values for these promising biomarkers, possibly through large, multicenter consortium efforts or meta-analyses. In summary, we integrated LAR into the ES-SCLC prediction system, demonstrating superior predictive performance compared to traditional inflammatory biomarkers, and validated it using Bootstrap and TimeROC curves for robustness. The nomogram provides individualized survival estimates, assisting clinicians in assessing treatment intensity based on factors such as liver metastasis and LAR scores. Our nomograms predict OS and PFS better than TNM staging and are cost-effective and accessible for screening immunotherapy candidates. Prospective validation in broader, real-world cohorts is warranted to confirm its generalizability and clinical utility. Future studies could further enhance its role in personalized treatment for ES-SCLC and pursue the standardization of key biomarker cutoff values to improve cross-study comparability. Conclusions A prognostic nomogram for ES-SCLC was created and validated by integrating biomarkers such as the LAR and clinical features. The nomogram demonstrated strong predictive accuracy in forecasting survival outcomes following immunotherapy. Supplementary The article’s supplementary files as Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-0918/prf