Pulmonary [18F]FDG uptake demonstrates phenotypic PET behaviour in non-specific interstitial pneumonitis and refines risk stratification

Rationale Effective prognostic biomarkers in Non-specific Interstitial Pneumonitis (NSIP) are limited. Objective To investigate the potential of [ 18 F]FDG PET/CT to predict mortality in NSIP and to assess its relationship with physiological indices and the Interstitial Lung Disease – Gender-Age-Physiology (ILD-GAP) score. Methods Ninety-six patients with a multidisciplinary team-confirmed diagnosis of NSIP (40 male, 56 female; mean age 59.8 ± 9.8 years) underwent baseline [ 18 F]FDG PET/CT and 83 patients completed full pulmonary function testing. Pulmonary uptake of [ 18 F]FDG was quantified using region-of-interest analysis as the maximum uptake in diseased lung (SUV max ) and the background uptake in visually normal lung (SUV min ). A target-to-background (TBR) ratio was calculated as SUV max /SUV min . Kaplan-Meier survival analysis examined associations between the [ 18 F]FDG PET/CT metrics, the pulmonary function tests and ILD-GAP scores. Stepwise forward Wald multivariable Cox regression assessed independence of the significant [ 18 F]FDG PET/CT parameters from the ILD-GAP index. The modified ILD-GAP (mGAP) models were generated by incorporating the [ 18 F]FDG PET/CT data into the ILD-GAP index. Results 65 patients died during a mean follow up of 83.2 ± 48.2 months. Mean ± SD SUV max, SUV min and TBR were 3.2 ± 1.2, 0.6 ± 0.3 and 6.0 ± 2.4, respectively. Mortality was associated with higher SUV max (median cutoff ≥ 2.9, p = 0.038) and higher SUV min (optimised cutoff ≥ 0.675, p < 0.001), whereas TBR was not prognostic ( p = 0.48). Multivariable Cox regression confirmed median SUV max as independent of ILD-GAP index ( p = 0.031), although optimised SUV min was not independent. Incorporating either parameter into a modified ILD-GAP model improved survival stratification compared with ILD-GAP index alone ( p < 0.01, log-rank test). Conclusion High pulmonary [ 18 F]FDG PET/CT uptake predicts increased mortality in NSIP. Median SUV max is an independent predictor of mortality and adding [ 18 F]FDG PET/CT parameters to the ILD-GAP index enhances prognostic discrimination. These findings support a disease-specific interpretation of [ 18 F]FDG PET/CT metrics in ILD, where SUV max and SUV min are dominant prognostic markers in diffuse NSIP.