Development of a Suspicion Index to aid diagnosis of Niemann-Pick disease type C

Abstract

Niemann-Pick disease type C (NP-C) is a rare, autosomal recessive lysosomal lipid storage disorder that is invariably fatal. NP-C diagnosis can be delayed for years due to heterogeneous presentation; adult-onset NP-C can be particularly difficult to diagnose. We developed a Suspicion Index tool, ranking specific symptoms within and across domains, including family members who have NP-C, to provide a risk prediction score to identify patients who should undergo testing for NP-C. A retrospective chart review was performed in 5 centers in Europe and 2 in Australia (n = 216). Three patient types were selected: classic or variant filipin staining NP-C cases (n = 71), NP-C noncases (confirmed negative by filipin staining; n = 64), or controls with at least 1 characteristic symptom of NP-C (n = 81). NP-C signs and symptoms were categorized into visceral, neurologic, or psychiatric domains. Logistic regression was performed on individual signs and symptoms within and across domains, and regression coefficients were used to develop prediction scores for NP-C. Internal validation was performed with the bootstrap resampling method. The Suspicion Index tool has good discriminatory performance with cutpoints for grading suspicion of NP-C. Neonatal jaundice/cholestasis, splenomegaly, vertical supranuclear gaze palsy, cataplexy, and cognitive decline/dementia were strong predictors of NP-C, as well as symptoms occurring in multiple domains in individual patients, and also parents/siblings or cousins with NP-C. The Suspicion Index tool is a screening tool that can help identify patients who may warrant further investigation for NP-C. A score ≥70 indicates that patients should be referred for testing for NP-C.