LODDS-based nomogram for breast cancer-specific survival in surgically treated breast cancer in young patients
Year: 2026
Type: article
Abstract: Traditional tumor staging relies on absolute lymph node counts, making it susceptible to stage migration, particularly in young breast cancer (YBC) patients with aggressive disease. This study assessed the prognostic value of the continuous log odds of positive lymph nodes (LODDS) for breast cancer-specific survival (BCSS) and developed a predictive nomogram utilizing data from 9,989 surgically treated YBC patients (aged 20-39) sourced from the Surveillance, Epidemiology, and End Results (SEER) database (2010-2015). A multivariable Cox proportional hazards model, adjusted to account for non-proportional hazards, established continuous LODDS as an independent risk factor for poorer BCSS (hazard ratio [HR] = 1.60, 95% confidence interval [CI]: 1.45-1.77, p < 0.001), without relying on data-driven dichotomization. We constructed a nomogram that integrates continuous LODDS with clinicopathological features selected via least absolute shrinkage and selection operator (LASSO). Under a 1,000-resample bootstrap optimism-correction, the nomogram exhibited strong discriminatory power (corrected concordance index [C-index] = 0.777), surpassing the traditional tumor-node-metastasis (TNM) staging model (C-index = 0.736). Moreover, the continuous LODDS-integrated model improved 5-year mortality risk reclassification (continuous net reclassification improvement [NRI] = 0.278, integrated discrimination improvement [IDI] = 0.034; both p < 0.001). The model also demonstrated high calibration accuracy (5-year Brier score = 0.0757) and presented a greater estimated net benefit in Decision Curve Analysis (DCA). In conclusion, continuous LODDS offers critical incremental prognostic information beyond traditional staging. The proposed nomogram serves as an internally validated prediction model for risk stratification in YBC patients.
Source: Biomolecules and Biomedicine
Authors: Ziqiang Wang, Zhenlin Yang, Haojie Zhang
Language: English
Subfield: Cancer Research
Domain: Life Sciences
Open Access status: diamond
OpenAlex ID: W7165803974
FWCI:
Cites:
Cited by:
Related to: