A machine learning approach to predict mortality due to immune-mediated thrombotic thrombocytopenic purpura.

Background: Mortality due to immune-mediated thrombotic thrombocytopenic purpura (iTTP) remains significant. Predicting mortality risk may potentially help individualize treatment. The French Thrombotic Microangiopathy (TMA) Reference Score has not been externally validated in the United States. Recent advances in machine learning technology can help analyze large numbers of variables with complex interactions for the development of prediction models. Objectives: To validate the French TMA Reference Score in the United States Thrombotic Microangiopathy (USTMA) iTTP database and subsequently develop a novel mortality prediction tool, the USTMA TTP Mortality Index. Methods: We analyzed variables available at the time of initial presentation, including demographics, symptoms, and laboratory findings. We developed our model using gradient boosting machine, a machine learning ensemble method based on classification trees, implemented in the R package gbm. Results: In our cohort (n = 419), the French score predicted mortality with an area under the receiver operating characteristic curve of 0.63 (95% CI: 0.50-0.77), sensitivity of 0.35, and specificity of 0.84. Our gradient boosting machine model selected 8 variables to predict acute mortality with a cross-validated area under the receiver operating characteristic curve of 0.77 (95% CI: 0.71-0.82). The 2 cutoffs corresponded to sensitivities of 0.64 and 0.50 and specificities of 0.76 and 0.87, respectively. Conclusion: The USTMA Mortality Index was acceptable for predicting mortality due to acute iTTP in the USTMA registry, but not sensitive enough to rule out death. Identifying patients at high risk of iTTP-related mortality may help individualize care and ultimately improve iTTP survival outcomes. Further studies are needed to provide external validation. Our model is one of many recent examples where machine learning models may show promise in clinical prediction tools in healthcare.

Plasma levels of S100A8/A9, histone/DNA complexes, and cell-free DNA predict adverse outcomes of immune thrombotic thrombocytopenic purpura.

BACKGROUND: Immune thrombotic thrombocytopenic purpura (iTTP) is a life-threatening blood disorder, primarily resulting from autoantibodies against ADAMTS13. Infection or inflammation often precedes acute iTTP. However, the association of inflammation and inflammatory mediators with disease severity and outcome of acute iTTP is not fully assessed. OBJECTIVES: Here, we determined plasma levels of S100A8/A9, histone/DNA complexes, citrullinated histone H3 (CitH3), and cell-free DNA (cfDNA) in a cohort of 108 acute episodes from 94 unique iTTP patients and healthy controls, and assessed the association of each of these biomarkers with the disease severity and mortality. RESULTS: All acute iTTP patients had significantly increased plasma levels of S100A8/A9 (median 84.8, interquartile range [IQR] 31.2-157.4 µg/mL), histone/DNA complexes (median 55.7, IQR 35.8-130.8 U/mL), CitH3 (median 3.8, IQR 2.2-6.4 ng/mL), and cfDNA (median 937.7, IQR 781.3-1420.0 ng/mL) on the admission blood samples when compared with healthy controls. An increased plasma level of S100A8/A9, histone/DNA complex and cfDNA was associated with organ damage, coagulopathy, and mortality in iTTP. After being adjusted for age and history of hypertension, Cox proportional hazard regression analysis demonstrated that a hazard ratio (95% confidence interval) for an elevated plasma level of S100A8/A9, histone/DNA complexes, and cfDNA was 11.5 (1.4-90.9) (P = .021), 10.3 (2.7-38.5) (P = .001), and 12.8 (3.9-42.0) (P = .014), respectively. CONCLUSION: These results indicate that inflammation or plasma inflammatory mediators such as S100A8/A9 or NETosis markers such as histone/DNA complexes and cfDNA may play a role in pathogenesis of iTTP, which may help stratify patients with a high risk of death during acute iTTP episodes.

Longitudinal assessments of plasma ADAMTS13 biomarkers predict recurrence of immune thrombotic thrombocytopenic purpura.

Immune thrombotic thrombocytopenic purpura (iTTP) is primarily caused by immunoglobulin G (IgG)-type autoantibodies that bind and inhibit plasma ADAMTS13 activity and/or accelerate its clearance from circulation. Approximately 50% of patients with iTTP who achieve initial clinical response to therapy experience recurrence (ie, exacerbation and/or relapse); however, a reliable biomarker that predicts such an event is currently lacking. The present study determines the role of longitudinal assessments of plasma ADAMTS13 biomarkers in predicting iTTP exacerbation/recurrence. Eighty-three unique iTTP patients with 97 episodes from the University of Alabama at Birmingham Medical Center between April 2006 and June 2019 were enrolled. Plasma levels of ADAMTS13 activity, antigen, and anti-ADAMTS13 IgG on admission showed no significant value in predicting iTTP exacerbation or recurrence. However, persistently low plasma ADAMTS13 activity (<10 U/dL; hazard ratio [HR], 4.4; 95% confidence interval [CI], 1.6-12.5; P = .005) or high anti-ADAMTS13 IgG (HR, 3.1; 95% CI, 1.2-7.8; P = .016) 3 to 7 days after the initiation of therapeutic plasma exchange was associated with an increased risk for exacerbation or recurrence. Furthermore, low plasma ADAMTS13 activity (<10 IU/dL; HR, 4.8; 95% CI, 1.8-12.8; P = .002) and low ADAMTS13 antigen (<25th percentile; HR, 3.3; 95% CI, 1.3-8.2; P = .01) or high anti-ADAMTS13 IgG (>75th percentile; HR, 2.6; 95% CI, 1.0-6.5; P = .047) at clinical response or remission was also predictive of exacerbation or recurrence. Our results suggest the potential need for a more aggressive approach to achieve biochemical remission (ie, normalization of plasma ADAMTS13 activity, ADAMTS13 antigen, and anti-ADAMTS13 IgG) in patients with iTTP to prevent the disease recurrence.

Clinical factors and biomarkers predict outcome in patients with immune-mediated thrombotic thrombocytopenic purpura.

Immune-mediated thrombotic thrombocytopenic purpura is characterized by severe thrombocytopenia and microangiopathic hemolytic anemia. It is primarily caused by immunoglobin G type autoantibodies against ADAMTS13, a plasma metalloprotease that cleaves von Willebrand factor. However, reliable markers predictive of patient outcomes are yet to be identified. Seventy-three unique patients with a confirmed diagnosis of immune-mediated thrombotic thrombocytopenic purpura between April 2006 and December 2017 were enrolled from the Univeristy of Alabama at Birmingham Medical Center. Clinical information, laboratory values, and a panel of special biomarkers were collected and/or determined. The results demonstrated that the biomarkers associated with endothelial injury (e.g., von Willebrand factor antigen and collagen-binding activity), acute inflammation (e.g., human neutrophil peptides 1-3 and histone/deoxyribonucleic acid complexes), and activation of the complement alternative pathway (e.g., factors Bb and iC3b) were all significantly increased in patients with acute immune-mediated thrombotic thrombocytopenic purpura compared to those in the healthy controls. Moreover, failure to normalize platelet counts within 7 days or failure to markedly reduce serum lactate dehydrogenase by day 5, low total serum protein or albumin, and high serum troponin levels were also predictive of mortality, as were the prolonged activated partial thromboplastin time, high fibrinogen, and elevated serum lactate dehydrogenase, Bb, and sC5b-9 on admission. These results may help to stratify patients for more intensive management. The findings may also provide a framework for future multicenter studies to identify valuable prognostic markers for immune-mediated thrombotic thrombocytopenic purpura.

Transfusion of Platelets Loaded With Recombinant ADAMTS13 (A Disintegrin and Metalloprotease With Thrombospondin Type 1 Repeats-13) Is Efficacious for Inhibiting Arterial Thrombosis Associated With Thrombotic Thrombocytopenic Purpura.

Objective- ADAMTS13 (a disintegrin and metalloprotease with thrombospondin type 1 repeats-13) cleaves VWF (von Willebrand factor). This process is essential for hemostasis. Severe deficiency of plasma ADAMTS13 activity, most commonly resulting from autoantibodies against ADAMTS13, causes thrombotic thrombocytopenic purpura. Therapeutic plasma exchange is the standard of care to date, which removes autoantibodies and replenishes ADAMTS13. However, such a therapy is often ineffective to raise plasma ADAMTS13 activity, and in-hospital mortality rate remains as high as 20%. Approach and Results- To overcome the inhibition by autoantibodies, we developed a novel approach by delivering rADAMTS13 (recombinant ADAMTS13 ) using platelets as vehicles. We show that both human and murine platelets can uptake rADAMTS13 ex vivo. The endocytosed rADAMTS13 within platelets remains intact, active, and is stored in α-granules. Under arterial shear (100 dyne/cm2), the rADAMTS13 in platelets is released and effectively inhibits platelet adhesion and aggregation on a collagen-coated surface in a concentration-dependent manner. Transfusion of rADAMTS13-loaded platelets into Adamts13-/- mice dramatically reduces the rate of thrombus formation in the mesenteric arterioles after FeCl3 injury. An ex vivo transfusion of rADAMTS13-loaded platelets to a reconstituted whole blood containing plasma from a patient with immune-mediated thrombotic thrombocytopenic purpura and the cellular components (eg, erythrocytes and leukocytes) from a healthy individual, as well as a fresh whole blood obtained from a patient with congenital or immune-mediated thrombotic thrombocytopenic purpura also dramatically reduces the rate of thrombus formation under arterial flow. Conclusions- Our results demonstrate that transfusion of rADAMTS13-loaded platelets may be a novel and potentially effective therapeutic approach for arterial thrombosis, associated with congenital and immune-mediated thrombotic thrombocytopenic purpura.

Therapeutic efficacy of the platelet glycoprotein Ib antagonist anfibatide in murine models of thrombotic thrombocytopenic purpura.

Thrombotic thrombocytopenic purpura (TTP), a potentially fatal blood clot disorder, is primarily caused by severe deficiency of plasma ADAMTS13 activity resulting from acquired autoantibodies. Plasma exchange is the only effective initial therapy. However, the high mortality rate and the complications associated with plasma exchange therapy remain a major concern. To address unmet clinical needs, therapeutic efficacies of anfibatide, a snake venom-derived platelet glycoprotein Ib antagonist, in murine models of spontaneous thrombocytopenia and shigatoxin-induced TTP were determined. A light scattering platelet aggregometry, microfluidic shear-based assay, and murine models of TTP were used in the study. We showed that purified anfibatide inhibits ristocetin- or botrocetin-induced human or murine platelet agglutination in the presence of von Willebrand factor in a concentration-dependent manner. Anfibatide could also dramatically inhibit the adhesion and aggregation of murine and human platelets on a collagen surface under arterial shear stress, in the presence or absence of plasma ADAMTS13 activity. Most importantly, we demonstrated that an intraperitoneal administration of anfibatide at the dose of 60 ng/g body weight twice daily mitigated spontaneous thrombocytopenia and prevented shigatoxin-induced TTP in Adamts13-/- and disease-susceptible mice (CAST/Ei strain). Thus, we conclude that anfibatide, when administered at the optimal dosage, route, and interval, is efficacious in treating spontaneous and bacterial shigatoxin-induced TTP in the murine models. Our findings may provide the basis for further development of anfibatide for the treatment of acute TTP in humans.