Development and validation of a nomogram for predicting immunotherapy outcomes in lung cancer patients using clinical and blood biomarkers

BMC Cancer 2025 AI 6 Explanations View Original
Original Paper (PDF)

Unable to display PDF. Download it here or view on PMC.

Plain-English Explanations
Pages 1-2
Predicting Who Benefits from Immunotherapy in Lung Cancer

The Problem with Immunotherapy Response: Immune checkpoint inhibitors (ICIs) have transformed lung cancer treatment, but only a subset of patients respond favorably. Without reliable predictors, clinicians cannot easily identify who will benefit or experience adverse events.

Study Goal: Researchers aimed to build a practical nomogram - a visual scoring tool - that combines routine clinical factors and blood biomarkers to predict both overall survival (OS) and progression-free survival (PFS) in lung cancer patients receiving immunotherapy.

Patient Population: The study enrolled 436 lung cancer patients across multiple centers, divided into a training cohort of 306 patients and a validation cohort of 130 patients, providing a robust dataset for model development and testing.

Significance: A validated nomogram translates complex statistical models into a simple visual tool clinicians can use at the bedside, making individualized risk stratification accessible without specialized software.

TL;DR: This study developed a nomogram combining blood tests and clinical variables to predict how long lung cancer patients will survive and remain progression-free on immunotherapy.
Pages 2-3
How the Nomogram Was Built

Biomarker Selection: From dozens of candidate variables, researchers used LASSO (Least Absolute Shrinkage and Selection Operator) regression to identify the most predictive factors while avoiding overfitting - a critical step when many variables are tested simultaneously.

Multivariable Cox Regression: Selected variables were entered into Cox proportional hazards models to quantify each factor's independent contribution to OS and PFS, producing hazard ratios with 95% confidence intervals.

Nomogram Construction: The finalized predictors were graphically combined into a nomogram where each factor is assigned a point score, and the total score maps to predicted 1-year and 2-year survival probabilities.

Validation: Model performance was assessed using the C-index (concordance index, measuring discrimination), calibration curves (measuring accuracy of probability estimates), and decision curve analysis (measuring clinical utility) in both training and external validation cohorts.

TL;DR: LASSO regression identified key predictors, which were then combined into a Cox regression-based nomogram validated for discrimination and calibration.
Pages 3-4
Key Predictors in the Nomogram

Neutrophil-to-Lymphocyte Ratio (NLR): A high NLR reflects an imbalance between inflammatory neutrophils and immune-active lymphocytes. Elevated NLR was associated with worse outcomes, consistent with its role as a marker of systemic inflammation and immunosuppression.

Surgical History: Patients who had prior surgical resection of their tumor showed better survival on immunotherapy, likely because surgery reduced tumor burden and altered the immune microenvironment favorably.

Disease Staging and Treatment Lines: More advanced disease stage and later lines of therapy (meaning the patient had already failed prior treatments) were associated with poorer immunotherapy outcomes.

Liver Metastases: The presence of liver metastases emerged as an independent negative predictor, consistent with the known immunosuppressive microenvironment of the liver and reduced efficacy of ICIs in hepatic metastatic disease.

TL;DR: NLR, surgical history, stage, treatment line, and liver metastases were the five key factors incorporated into the predictive nomogram.
Pages 4-5
Model Performance and Validation

Overall Survival Discrimination: The nomogram achieved a C-index of 0.709 for OS in the training cohort and 0.655 in the external validation cohort, indicating moderate-to-good ability to rank patients by survival risk.

Progression-Free Survival Discrimination: For PFS, C-index values were 0.730 (training) and 0.694 (validation), suggesting the model predicts time-to-progression somewhat more accurately than overall death.

Calibration: Calibration curves showed close agreement between nomogram-predicted probabilities and observed outcomes, confirming the model does not systematically over- or underestimate risk.

Clinical Utility: Decision curve analysis demonstrated net benefit of the nomogram over default strategies (treating all or no patients) across a broad range of threshold probabilities, supporting its practical clinical value.

TL;DR: The nomogram achieved C-index values of 0.655-0.730 across cohorts with good calibration, confirming reliable individualized survival prediction.
Pages 5-6
Clinical Applications of the Nomogram

Risk Stratification: By categorizing patients into low, intermediate, and high risk groups based on nomogram scores, oncologists can tailor treatment intensity, monitoring frequency, and supportive care accordingly.

Treatment Selection: Patients with high nomogram risk scores (e.g., high NLR, liver metastases, late treatment line) may be redirected toward alternative treatment strategies rather than continuing ineffective immunotherapy.

Clinical Trial Design: The nomogram can help stratify patients during randomization in future clinical trials, ensuring balanced distribution of high-risk and low-risk patients across treatment arms.

Accessibility: Because the nomogram uses only NLR (a routine complete blood count calculation) and clinical variables readily available in medical records, it can be applied in community oncology settings without specialized laboratory testing.

TL;DR: The nomogram enables practical risk stratification using routine blood tests and clinical data, supporting individualized treatment decisions at point of care.
Pages 6-7
Limitations and Future Research Directions

Retrospective Design: As a retrospective study, treatment selection and follow-up practices may not have been uniform, introducing potential confounders that a prospective trial would control for more rigorously.

Biomarker Scope: The nomogram relies primarily on NLR and clinical variables; incorporating newer biomarkers such as tumor mutational burden (TMB), PD-L1 expression level, or circulating tumor DNA could further improve predictive accuracy.

External Generalizability: The validation cohort, while external, came from similar institutional settings. Validation in diverse geographic and ethnic populations is needed to confirm applicability across broader patient populations.

Prospective Validation: Future studies should prospectively test whether using the nomogram to guide treatment selection actually improves patient outcomes compared to standard clinical decision-making, which is the ultimate test of any predictive model.

TL;DR: Prospective validation in diverse populations and integration of molecular biomarkers are the key next steps to strengthen this nomogram's clinical impact.
Citation: Open Access, 2025. Available at: PMC12372320.