Characteristics, risk factors and a risk prediction model of tocilizumab-induced hypofibrinogenemia: a retrospective real-world study of inpatients.

Abstract

Objective: The occurrence of hypofibrinogenemia after tocilizumab treatment has attracted increasing attention, which may cause bleeding and even life-threatening. This study aims to explore the risk factors for tocilizumab-induced hypofibrinogenemia (T-HFIB) and construct a risk prediction model.

Methods: A total of 221 inpatients that received tocilizumab from 2015 to 2023 were retrospectively collected and divided into T-HFIB group or control group. The risk factors for T-HFIB were obtained by logistic regression equation and used to establish the nomogram.

Results: T-HFIB was observed in 121 of 221 patients (54.75%). Multifactorial logistic regression analysis revealed that infection (OR = 2.002, 95%CI:1.018 ~ 3.935), COVID-19 (OR = 3.752, 95%CI:1.264 ~ 11.139), CAR-T therapy (OR = 4.409, 95%CI:2.017 ~ 0.894), and concomitant glucocorticoids (OR = 5.303, 95%CI:0.227 ~ 0.894) were identified as independent risk factors for T-HFIB, while high baseline fibrinogen level (OR = 0.813, 95%CI:0.670 ~ 0.988) and concomitant antirheumatic drugs (OR = 0.451, 95%CI:0.227 ~ 0.894) were identified as protective factors. A nomogram was established, and area under the curve (AUC) of prediction model was 0.772 (95%CI:0.709 ~ 0.836). Calibration curve showed a good prediction accuracy for the occurrence of T-HFIB.

Conclusion: The infection, COVID-19, CAR-T therapy, and concomitant glucocorticoids were independent risk factors for T-HFIB, while high baseline fibrinogen and concomitant antirheumatic drugs were protective factors. This nomogram can help early identify the patients at potential high risk of developing T-HFIB.