Prognostic value of laboratory biomarkers for mortality risk stratification in thrombotic thrombocytopenic purpura

TL;DR

This study identifies lab biomarkers that predict short-term mortality in thrombotic thrombocytopenic purpura patients and builds a risk model to guide treatment.

Contribution

A novel risk stratification model using cTnI, LDH, BUN, and IBIL levels to predict mortality in TTP patients is developed and validated.

Findings

Elevated cTnI, LDH, BUN, and IBIL levels are strongly associated with increased mortality risk in TTP patients.

A 0–4 point risk model accurately stratifies mortality risk, with scores ≥3 indicating high mortality.

The model showed strong stability through internal and external validation.

Abstract

This study aimed to explore the relationship between laboratory indicators and short-term mortality risk in patients with thrombotic thrombocytopenic purpura (TTP) and to construct a risk stratification model. We retrospectively analyzed the clinical data of 106 patients with TTP admitted to Tongji Hospital between June 2016 and February 2025. Patients were grouped by 28-day survival: death (n = 45) and survival (n = 61). Prognosis-related indicators were identified using receiver operating characteristic (ROC) curves and logistic regression analyses. A mortality risk stratification model was established. To validate the stability of the model, 30 external cases of TTP (January 2022–December 2024) were collected. The levels of cardiac troponin I (cTnI), lactate dehydrogenase (LDH), blood urea nitrogen (BUN), and indirect bilirubin (IBIL) were significantly higher in the death group compared to those in the survival group. ROC analysis identified optimal mortality risk cutoffs: cTnI at 353.1 pg/mL (odds ratio [OR] = 4.778), LDH at 992 U/L (OR = 2.842), BUN at 10.9 mmol/L (OR = 5.527), and IBIL at 32.3 µmol/L (OR = 2.995). A subsequent 0–4 point risk stratification model demonstrated increasing mortality rates of 21.3% (0–1 point), 39.1% (2 points), 60.9% (3 points), and 92.3% (4 points). Each 1-point increase in the risk score was associated with a 2.324-fold rise in mortality risk (95% confidence interval 1.597–3.383). Internal and external validations confirmed the model’s stability and strong prognostic performance in patients with TTP. The risk model incorporating cTnI, LDH, BUN, and IBIL levels effectively predicted short-term mortality in patients with TTP. Patients with a score of 3 or higher required close monitoring and aggressive therapeutic intervention to mitigate mortality risk.

The online version contains supplementary material available at 10.1007/s00277-025-06584-8.

Figures