Heyang Zhang, Guotian Ruan, Zelin Chai, Chong Li, Jinyu Shi, Chenan Liu, Shutian Zhang, Peng Li, Shengtao Zhu, Hanping Shi
ALNCR is a simple inflammation-based prognostic index that outperforms conventional inflammatory markers for OS prediction. Integrating ALNCR with TNM stage improves risk stratification and prognostic accuracy, supporting clinical implementation via a nomogram and web-based calculator.
BACKGROUND: Systemic inflammation is associated with cancer prognosis, but commonly used inflammatory indices show limited discrimination across heterogeneous malignancies. We developed and validated a novel inflammatory burden index, ALNCR, and tested whether integrating ALNCR with TNM stage improves overall survival (OS) prediction.
METHODS: We analysed 6374 adults with cancer from the multicentre INSCOC cohort (2013-2021), randomly split into a training cohort (n = 4464) and internal validation cohort (n = 1910), and an external validation cohort from Fujian Cancer Hospital (n = 759). Female proportions were 39.0%, 41.2% and 32.0%, and median ages were 60, 60 and 58 years, respectively; Stage IV disease accounted for 45.4%, 45.8% and 64.8%. ALNCR was defined as (albumin × lymphocyte) / (neutrophil × C-reactive protein). Discrimination was assessed using C-index and time-dependent AUC, with comparisons to NLR, PLR, CAR, SII and PNI. A combined model (ALNCR.TNM) used ALNCR (cutoff 3.21) and TNM stage (I/II vs. III/IV); incremental value was quantified by net reclassification improvement (NRI) and integrated discrimination improvement (IDI).
RESULTS: ALNCR showed the highest discrimination for OS among evaluated indices (C-index 0.645 [95% CI 0.631-0.658] in training; 0.666 [0.645-0.686] in internal validation; and 0.657 [0.624-0.690] in external validation). High ALNCR (≥ 3.21) was associated with longer OS than low ALNCR (median OS not reached vs. 22.3 months; log-rank p < 0.001). After multivariable adjustment, high ALNCR remained associated with lower mortality (HR 0.56 [0.51-0.61], 0.50 [0.43-0.58] and 0.53 [0.42-0.67] across the three cohorts; all p < 0.001). Adding ALNCR to TNM improved prediction over TNM alone (ΔC-index +0.021; NRI 0.067, p = 0.010; IDI 0.004, p < 0.001) and stratified patients into four risk groups (p < 0.0001), with consistent associations across major cancer types and clinical strata.
CONCLUSIONS: ALNCR is a simple inflammation-based prognostic index that outperforms conventional inflammatory markers for OS prediction. Integrating ALNCR with TNM stage improves risk stratification and prognostic accuracy, supporting clinical implementation via a nomogram and web-based calculator.