A First-Visit Clinical Score to Distinguish Immunoglobulin Light-Chain from Wild-Type Transthyretin Cardiac Amyloidosis: Derivation and Internal Validation

Background: Delayed diagnosis of immunoglobulin light-chain cardiac amyloidosis (AL-CA) worsens prognosis, partly because AL-CA and wild-type transthyretin cardiac amyloidosis (ATTRwt-CA) are difficult to distinguish at the initial cardiology visit. We developed and validated a practical score to triage AL-CA from ATTRwt-CA. Methods: We retrospectively analyzed 541 consecutive patients with cardiac amyloidosis (91 AL-CA, 450 ATTRwt-CA) at our institution's amyloidosis center, randomly divided into derivation (n=378) and validation (n=163) cohorts. Candidate first-visit variables underwent LASSO selection after multiple imputation, with estimates pooled by Rubin's rules. Performance was assessed by bootstrap validation, calibration, and decision curve analysis. The model was converted into a simplified point-based score. Results: Seven variables were retained: age, absence of atrial fibrillation, absence of carpal tunnel syndrome, low serum albumin, proteinuria ([≥]1+), low voltage, and thinner left ventricular posterior wall thickness. Areas under the curve (AUC) were 0.97 in the derivation cohort and 0.99 in the validation cohort; the bootstrap optimism-corrected AUC was 0.96. The model provided net benefit across threshold probabilities of 0.01-0.50. Risk tiering categorized the validation cohort into low (0-5 points; AL-CA 0%), intermediate (6?9 points; 26%), and high (10-13 points; 95%) probability groups. Discrimination remained high for AL-CA versus monoclonal protein-positive ATTRwt-CA at a cutoff of [≥]8 points (AUC 0.90). Adding this score to monoclonal protein testing significantly improved diagnostic performance compared with monoclonal protein testing alone. Conclusions: This simple first-visit score may help cardiologists rapidly triage suspected AL-CA and prioritize urgent hematology referral and subtype-specific confirmatory testing.