NSCLC Nomogram Adds WBC, LMR, Albumin to Prognosis
Key Takeaways
- In young adults with stage IV NSCLC treated at a single cancer center, median follow-up was 56.6 months and median overall survival was 27.2 months.
- Driver gene mutations were present in 67.7% of the cohort, and sex, liver metastasis, gene mutation status, and targeted therapy recurred across multivariable analyses as prognostic factors for overall survival.
- The final combined model also retained white blood cell count, LMR, and serum albumin, with worse survival linked to higher white blood cell count and longer survival linked to higher LMR and albumin.
- After bootstrap validation, the combined clinical-inflammatory-nutritional model showed the strongest reported discrimination and greater net benefit than the narrower models and tumor-node-metastasis staging.
- The findings came from an exploratory retrospective single-center analysis without external validation, so generalizability remains uncertain.
A retrospective cohort of young adults with stage IV NSCLC at the Fourth Hospital of Hebei Medical University included 514 patients aged 18 to 45 years with newly diagnosed, pathologically confirmed stage IV non-small cell lung cancer between January 2013 and March 2025, with follow-up through 31 October 2025. Overall survival (OS) was the primary endpoint, defined from diagnosis to death from any cause or last follow-up; the mean age was 41.0 ± 0.7 years, 238 patients were male and 276 were female, and 40 were lost to follow-up for a 92.2% follow-up rate. Inflammatory markers assessed at diagnosis were white blood cell count, neutrophil count, monocyte count, lymphocyte count, neutrophil-to-lymphocyte ratio (NLR), lymphocyte-to-monocyte ratio (LMR), platelet-to-lymphocyte ratio (PLR), systemic immune-inflammation index (SII), and systemic inflammation response index (SIRI), alongside hemoglobin, serum albumin, body mass index (BMI), and prognostic nutritional index (PNI) as nutritional indicators. The investigators built three exploratory models and compared them with tumor-node-metastasis (TNM) staging, selected cutoffs with maximally selected rank statistics, and used 1000 bootstrap resamples for internal validation. These methods positioned the analysis as exploratory prognostic model development rather than treatment comparison.
Across the cohort, median follow-up was 56.6 months and median OS was 27.2 months. In clinicopathologic multivariable analysis, female sex, absence of bone metastasis, absence of liver metastasis, presence of a driver gene mutation, and receipt of targeted therapy were independently associated with better OS. In the model-construction results, the authors report that the final combined clinical-inflammatory-nutritional model retained sex, liver metastasis, gene mutation, targeted therapy, white blood cell count, LMR, and serum albumin. White blood cell count > 8.57 × 10^9/L was associated with worse survival (HR 1.832, 95% CI 1.476-2.274; p < 0.001), whereas LMR > 3.16 (HR 0.497, 95% CI 0.400-0.617; p < 0.001) and serum albumin > 34.10 g/L (HR 0.489, 95% CI 0.359-0.665; p < 0.001) were associated with longer survival. Overall, higher white blood cell count tracked with worse prognosis, whereas higher LMR and serum albumin tracked with longer survival.
After 1000-bootstrap internal validation, the combined clinical-inflammatory-nutritional model had the highest reported discrimination, with a concordance index (C-index) of 0.711 (95% CI 0.696-0.726), compared with 0.693 (95% CI 0.678-0.708) for the clinical-inflammatory model, 0.687 (95% CI 0.672-0.702) for the clinical-nutritional model, and 0.545 (95% CI 0.527-0.563) for TNM staging. Its 1-, 2-, 3-, 4-, and 5-year area under the curve (AUC) values were 0.788, 0.761, 0.756, 0.720, and 0.704, respectively. These values exceeded those of the comparator models, and decision curve analysis showed greater net benefit across a wider range of threshold probabilities. Within this dataset, the combined model showed the strongest apparent prognostic separation.
The study was retrospective and single center, with potential selection bias, no external validation cohort, and treatment heterogeneity across more than a decade of accrual. The reported comparisons therefore reflect apparent predictive performance within the study dataset after internal validation, not treatment effectiveness, external generalizability, or evidence that changing inflammatory or nutritional markers improves outcomes. The findings should be interpreted as prognostic associations within the analyzed cohort.
Overall, the investigators reported that peripheral blood inflammatory markers and nutritional indices were associated with survival in young patients with advanced non-small cell lung cancer. They also reported that the combined model outperformed the narrower models and TNM staging on internal performance measures. Further validation in independent cohorts is still required before broader application.
Clinician Questions
Which variables were retained in the combined prognostic model for young adults with stage IV NSCLC?
In this exploratory retrospective model-development analysis of young adults with newly diagnosed stage IV NSCLC, the final clinical-inflammatory-nutritional model retained sex, liver metastasis, gene mutation, targeted therapy, white blood cell count, lymphocyte-to-monocyte ratio, and serum albumin as independent prognostic factors.
What white blood cell, LMR, and albumin cutoffs were associated with survival in young advanced NSCLC?
In young patients with advanced NSCLC at diagnosis, white blood cell count > 8.57 × 10^9/L was associated with worse prognosis, LMR > 3.16 was associated with better prognosis, and serum albumin > 34.10 g/L was associated with longer overall survival; these were prognostic associations rather than evidence that changing the markers improves outcomes.
How did the combined nomogram compare with TNM staging for overall survival prediction in young stage IV NSCLC?
After 1000-bootstrap internal validation, the combined clinical-inflammatory-nutritional model had higher reported discrimination than TNM staging in young stage IV NSCLC, with a C-index of 0.711 versus 0.545, along with higher 1- through 5-year AUCs; decision curve analysis also showed greater net benefit across a wider range of threshold probabilities, although external validation was lacking.
Which inflammatory and nutritional markers were evaluated at diagnosis in young adults with advanced NSCLC?
In young adults with advanced NSCLC, the inflammatory markers evaluated at diagnosis were white blood cell count, neutrophil count, monocyte count, lymphocyte count, NLR, LMR, PLR, SII, and SIRI, and the nutritional indicators were hemoglobin, serum albumin, BMI, and PNI.
Recommended Reading
- For more on mutation-guided NSCLC treatment: Novel Strategies for Managing EGFR-mutated mNSCLC: Expert Insights into the Latest Data and Recommendations