Neutrophil percentage-to-albumin ratio and late gadolinium enhancement in clinically suspected myocarditis: a multicentre retrospective study

Abstract Background Late gadolinium enhancement (LGE) on cardiac magnetic resonance (CMR) imaging identifies myocardial involvement in myocarditis and carries prognostic weight. The neutrophil percentage-to-albumin ratio (NPAR) combines neutrophil predominance with serum albumin, a negative acute-phase reactant. Whether NPAR is associated with LGE, and whether it adds information to its individual components and to routinely available clinical data, is unknown. Methods This retrospective observational study included 1,055 adults with clinically suspected myocarditis who underwent CMR at seven tertiary cardiovascular centres in Türkiye between October 2022 and September 2025. All studies were reviewed centrally by three cardiovascular radiologists blinded to clinical and laboratory data (Fleiss' κ = 0.83). Variables significant on univariable analysis (p < 0.10) were entered into multivariable logistic regression models, the shape of the association was examined using restricted cubic splines, and the incremental value of NPAR over the clinical and laboratory model was assessed using discrimination and reclassification metrics. Results LGE was present in 285 patients (27.0%). NPAR was higher in the LGE-positive group [1.65 (1.40–1.92) vs. 1.29 (1.02–1.55); p < 0.001]. In the multivariable model, NPAR remained independently associated with LGE positivity (adjusted odds ratio per 1 standard deviation 4.381, 95% confidence interval 3.248–5.907; p < 0.001), and the Nagelkerke R² rose from 0.157 to 0.387. Restricted cubic spline analysis showed a positive association across the observed range with evidence of non-linearity (p = 0.013). Adding NPAR increased the area under the curve of the model from 0.728 (0.687–0.768) to 0.839 (0.806–0.872) (difference 0.111; p < 0.001), with an integrated discrimination improvement of 0.194 (0.133–0.251) and a continuous net reclassification improvement of 0.843 (0.674–1.011), both p < 0.001. As a single marker, NPAR yielded an area under the curve of 0.750 (0.718–0.782) at a cut-off of 1.43, discriminating LGE better than either constituent (both p < 0.0001). Conclusions In clinically suspected myocarditis, NPAR was independently associated with CMR-defined LGE positivity, discriminated LGE better than either of its own components, and provided significant incremental discrimination and reclassification over a clinical and laboratory model. NPAR may serve as a low-cost adjunctive marker where access to CMR is constrained.