PB1327 - Searching for a reliably predictive PLASMIC score: Results of a multicenter study

Tuesday, June 25, 2024

13:45 – 14:45 ICT

Room: Exhibition Hall

Presenting Author(s)

MA

Mouhamed Yazan Abou-Ismail

Assistant Professor
Division of Hematology and Hematologic Malignancies, Department of Internal Medicine, University of Utah
Salt Lake City, Utah, United States

Co-Author(s)

AK

Ashley Khouri

Medical Student
University of Utah School of Medicine
Salt Lake City, Utah, United States
HL

Ho Jun Lee

Resident
Case Western Reserve University/University Hospitals Cleveland Medical Center
Cleveland, Ohio, United States
SK

Sanjana Kalvehalli Kashinath

Assistant Professor
Temple University Hospital
Philadelphia, Pennsylvania, United States
MS

Mukul Singal

Assistant Professor
Indiana Hemophilia and Thrombosis Center
Indianapolis, Indiana, United States
CP

Cecilia Peterson

Medical Student
University of Utah School of Medicine
Salt Lake City, Utah, United States
CM

Carley Mitchell

Fellow
Memorial Sloan Kettering Cancer Center
New York, New York, United States
CZ

Chong Zhang

Biostatistician III
Division Of Epidemiology, Department of Internal Medicine
Salt Lake City, Utah, United States
JG

Jeffrey Gilreath

PharmD
Division of Hematology and Hematologic Malignancies, Department of Internal Medicine, University of Utah
Salt Lake City, Utah, United States
PK

Peter Kouides

Professor
University of Rochester School of Medicine
Rochester, New York, United States
LN

Lalitha Nayak

Associate Professor
Case Western Reserve University/University Hospitals Cleveland Medical Center
Cleveland, Ohio, United States
ML

Ming Y. Lim

Associate Professor
Division of Hematology and Hematologic Malignancies, Department of Internal Medicine, University of Utah
Salt Lake City, Utah, United States

Background: The PLASMIC score is a validated clinical tool for predicting ADAMTS13 deficiency in suspected thrombotic thrombocytopenic purpura (TTP). Some of its components, such as mean corpuscular volume (MCV) may be impacted by common conditions such as iron deficiency or liver disease, which can diminish its accuracy or complicate clinical interpretation.

Aims: 1- To validate the PLASMIC score across three different hospitals (University of Utah, Case Western/University Hospitals, Rochester Regional Health).
2- To evaluate the contribution of PLASMIC score modifications in improving its performance using the area under the receiver operating characteristic (AUROC).

Methods: We identified all ADAMTS13 testing performed between 2012-2022. TTP was defined as an ADAMTS13 activity < 10%, or 10%-20% with clinical diagnosis, and controls as ADAMTS13 >20%. For the PLASMIC score and its variations, we calculated AUROC with 95% confidence intervals estimated from bootstrap with 2000 replications. Multivariable logistic regressions were used to assess each component’s predictive ability. Variations were explored by excluding 'A' (PLSMIC), 'S' (PLAMIC), 'M' (PLASIC), and introducing LDH, dichotomized at an optimal threshold which maximizes Youden’s J index. All analyses were conducted in Rv4.2[2], all tests were two-sided.

Results: We included 80 TTP patients and 402 controls. All PLASMIC components were independent predictors of TTP except for MCV, which showed no significant association (Table 1). LDH dichotomized at 726.5 U/L (threshold selected by maximizing Youden index) was an independent predictor of TTP (OR = 3.2, 95% CI: 1.7-6.1). Excluding MCV slightly improved the AUROC. The addition of LDH further improved the score compared to PLASMIC (AUROC=0.883 vs 0.854) (Table 2).

Conclusion(s): • This is the largest multi-center study of PLASMIC score modification, where the original score performed well.
• MCV had no significant association with TTP, while LDH was independently predictive of TTP.
• Substituting LDH for MCV improved the score’s performance.