Subclinical ocular alterations in Graves' disease: The FUMO score, a new tool to predict Graves' orbitopathy progression

Introduction This observational investigation aimed to assess the accuracy of a novel score in predicting the onset of overt Graves' orbitopathy (GO) in patients with Graves' disease (GD). Materials and methods A total of 156 consecutive patients with GD without GO were enrolled. As control group, 45 patients with non-autoimmune hyperthyroidism were included. At baseline, an ophthalmological evaluation was performed, including (1) visual function tests and (2) orbital ultrasound. After 24 months, the occurrence of GO was assessed in all patients. Results At baseline, a score from 0 to 3 and a score from 0 to 5 were assigned based on the results of visual function tests and the results of orbital ultrasound, respectively. The scores were combined into an overall FUnctional and MOrphological risk score (FUMO score) (0-8 points), classifying patients as low risk (score 0-2) or medium-high risk (score 3-8). After 24 months, the 2 risk groups were compared for differences in (1) presence/absence of GO and (2) GO activity and severity. Patients in the medium-high risk group developed overt GO more frequently than those in the low-risk group. Additionally, GO was more frequently active and moderate-to-severe in medium-high risk patients. Multiple logistic regression showed that TRAb levels and FUMO were the strongest independent predictors of GO, with higher FT3 and smoking habit levels also conferring increased risk. The model demonstrated good calibration and discrimination (AUC = 0.84; P < .0001), with high positive (73%) and negative (72%) predictive value. Conclusion Subclinical ocular alterations can predict the progression of GO in patients with GD. The FUMO score, particularly when combined with TRAb and FT3 levels, reliably identifies patients at risk of developing overt GO, supporting its use for early risk stratification.