Lack of sensitivity of diagnostic Cushing-scores in Germany: a multicenter validation

Abstract

Abstract Objective Endogenous Cushing's syndrome (CS) is a severe condition, often diagnosed at a late stage. To reduce mortality, early diagnosis plays an important role. Two screening tools for early identification of patients with CS have been developed in multicentric cohorts, but have not yet been validated in cohorts with different geographic backgrounds. Design We validated the Spanish score published by Leon-Justel et al. in 2016 and the Italian score by Parasiliti-Caprino et al. published in 2021 in our cohort. Methods In the multicentric German Cushing registry, patients with confirmed and expected but ruled out Cushing's syndrome are prospectively diagnosed and followed up. We validated both scores in a cohort of 458 subjects: 176 patients with confirmed CS and 282 patients with suspected, but finally excluded CS. Results Using the Spanish score, 17.5% of our patients with proven CS biochemical screening would not have been recommended. This concerned patients with pituitary CS (22%) and with adrenal CS (10%). On the contrary, only 14% of patients without CS would have received a recommendation for biochemical screening. Using the Italian score, 29% of patients with proven CS were classified into the low-risk classes not recommended for biochemical screening. This mostly affected patients with adrenal (31%) and pituitary CS (30%). About 12% of subjects without CS would have received a biochemical screening recommendation. Conclusions Both scores had limited sensitivity and high specificity in a German validation cohort. Further research is necessary to develop a screening score, which is effective in different healthcare systems and ethnicities.