Prediction of Severity and Mortality in Acquired Thrombotic Thrombocytopenic Purpura (aTTP). Utility of Clinical-biological Scores

C. Pascual-Izquierdo1,2*, A. Domingo-González3,4 1Servicio de Hematología y Hemoterapia. Hospital General Universitario Gregorio Marañón, Madrid, Spain 2Instituto de Investigación Sanitaria Gregorio Marañón, Madrid, Spain 3Servicio de Hematología y Hemoterapia. Hospital Universitario Virgen del Rocío, Sevilla, Spain 4Instituto de Investigación Sanitaria Virgen del Rocío, Sevilla, Spain *Correspondence should be addressed to Cristina Pascual Izquierdo; crisizquierdo3@yahoo.es

autoantibodies to ADAMTS13, which is the most common form of the disease, or congenitally via recessively inherited biallelic mutations of the ADAMTS13 gene. The latter is rare and represents 2% of all aTTP cases [1]. Independently of its origin, the lack of ADAMTS13 activity (less than 10%) is the only biologic marker specific for aTTP and it is a crucial factor in the diagnosis [1,3].
The presence of severe thrombocytopenia, microangiopathic hemolytic anemia, and variable organ ischemia remains the cornerstone of clinical diagnosis. However, none of these symptoms is exclusive to aTTP and they may be present in other pathologies such as hemolytic uremic syndrome (HUS), other TMA syndromes, and Evans syndrome [1,10]. For this reason, the presence of a severe deficiency of ADAMTS13 (<10% ADAMTS13 activity), although not required to start treatment, has become necessary to confirm the diagnosis [1,3].
The first-line therapy for acute aTTP is based on daily therapeutic plasma exchange (TPE) for ADAMTS13 deficiency in conjunction with steroids as an adjunct treatment due to the autoimmune nature of aTTP. The use of a humanized anti-CD20 monoclonal antibody (rituximab) as part of the frontline treatment has been proposed lately, due to the high response rates, the shorter hospitalization periods, and the fewer relapses reported in some studies [3,11].
Despite appropriate therapeutic management, aTTP still has a mortality rate of 10-20% and its clinical course is characterized by a tendency to relapse in up to 50% of cases [1]. Early recognition of patients at higher risk of mortality is of notable importance because they can benefit from emerging agents like caplacizumab, an anti-von Willebrand (vW) nanobody that blocks platelet aggregation into the small vessels [12].
Most relapses occur during the first or second year, but some of them can occur as late as 10 or 20 years after the first episode of aTTP [1]. Apart from the known short-term risk, over the last years, relapses have been associated with deficits in health-related quality of life such as chronic neurocognitive disability and arterial hypertension [2]. Relapsing TTP is related to recurrent or persistent severe ADAMTS13 deficiency, which suggests that monitoring ADAMTS13 during follow-up might enable a prompt diagnosis. However, as a severe ADAMTS13 deficiency is not always followed by the development of the clinical syndrome, the use of preemptive treatment should be individualized [1,13].
It is becoming essential to understand the factors associated with a fatal outcome to better tailor treatment. In this commentary, we will review the main TTP prognostic scores, their history and application, and the differences between first and relapsed episodes in terms of prognosis.

aTTP Prognostic Scores: A Journey through History
Since the first description of aPTT case by Eli Moschcowitz in 1924 [1], a tremendous effort has been made to deepen our knowledge of this disease and improve its dismal prognosis. The introduction of empirical treatment with plasma infusions and exchange in 1977 brought about a dramatic change in the prognosis of this disease and enabled the scientific community to describe the clinical course and outcomes of TTP survivors.
In the 1980s, the first cases of TTP relapses were described and in 1987 the first clinical severity score for TTP, later known as the Rose and Eldor score [14,15], was developed.

Rose and Eldor Score 1987
The Rose and Eldor score [14] were the result of the analysis of a retrospective series of 38 patients diagnosed and treated with TTP in 15 hospitals in Israel and the United States between 1977 and 1985. In this study, 26 patients had a first episode of TTP, while 12 patients had relapsing TTP. Ninety-five percent of patients had thrombocytopenia, 86% anemia, 37% fever, 73% renal involvement, and 94% neurologic manifestations. Treatment included plasma infusions (92% of the patients), plasma exchange (76%), steroids (95%), platelet inhibitor drugs (87%), platelet transfusions (36%), splenectomy (31%), and vincristine (8%). The mortality rate in the study was 21%.
A scoring system was developed based on 4 clinical and laboratory parameters: neurologic findings, renal function impairment, platelet count, and hemoglobin value at presentation ( Table 1).
The historical importance of the Rose and Eldor score is indisputable, since it constitutes the first scoring TTP system and these authors suggested for the first time that relapsing thrombotic thrombocytopenic purpura, although not a benign disorder, was a milder form of thrombotic thrombocytopenic purpura. However, it had several limitations such as the misidentification of other primary TMA as TTP episodes, the lack of consensus on terminology regarding concepts such as relapsed or refractory TTP, and the heterogeneity of treatment.

Canadian Apheresis Model Predicting 6-month Mortality (Wyllie et al. [16])
From 1987 on, the Rose and Eldor index of TTP severity was occasionally used to predict the response of TTP to plasma exchange, however, as discussed above, it had several limitations. In fact, it did not predict outcomes when applied in another case series [17]. As a result, in 2005 a new model predicting 6-month mortality was proposed by the Canadian Apheresis Group [16]. The goal of this study was to identify those patients who would benefit from higher treatment intensity, but no comparisons between first episodes and relapsing episodes of TTP were included.
In this registry, 171 patients with aTTP previously included in large randomized trials were chosen from 19 major medical centers across Canada over 22 years. Two cohorts with 2 different treatments (plasma exchange vs plasma infusion) were selected. Characteristics of patients from these cohorts were compared and no significant differences were found. A multivariable logistic model of 6-month survival was developed, containing age, hemoglobin, and the presence of fever at presentation ( Table 2). This model predicted 13.4% of outcome variance. Predictive scores of 0, 2, 4 and 6 correlated with 6-month mortality rates of 12.5%, 14.0%, 31.3% and 61.5% respectively. The modified Rose index (neurological abnormalities were modified as present or absent) was applied to the data set and no significant difference in scores was noted in terms of survival.
This study had several advantages in comparison to the Rose and Eldor score such as the statistical analysis performed to select the clinical and laboratory parameters and the careful data collection within large randomized trials. Moreover, treatment was specific and homogeneous. However, some limitations of this model need to be highlighted. On one hand, a common limitation with the Rose and Eldor score was the lack of a specific diagnosis of TTP, since ADAMTS13 activity was not included as part of the initial work-up. Besides, consensus on TTP terminology was still lacking.

French TMA Reference Center Score (FTRCC) 2012
Many advances regarding diagnosis of TTP have taken place since the Canadian prognostic model was published. However, despite this significant progress, death still occurred in up to 10-20% of patients. It therefore became essential to understand the factors associated with a fatal outcome. As a result of the many efforts made in this field, in 2012 the French TMA Reference Center published a prognostic score composed of original independent risk factors associated with 30-day mortality in patients diagnosed with TTP [18].
This retrospective analysis registered 281 patients diagnosed with TTP with an acquired severe ADAMTS13 deficiency over 10 years. The features of patients who died during the 30 days starting from treatment initiation were Table 1: Rose and Eldor score [14].

Mortality in TTP Score (MITS) 2016
In 2016 a comprehensive prognostic risk predictive score of all-cause mortality in hospitalized patients with TTP was published by Ruchika Goel et al, from Johns Hopkins University, Baltimore (United States) [19]. This registry included 8,203 patients diagnosed with TTP from 2007 to 2012 who received TPE treatment. Based on the results of univariate analysis, the significant variables were added in a stepwise manner in a multivariable model. This gave place to the Mortality in TTP Score (MITS), taking values from 1 to 3 for each variable (Table 4). This mortality score was thought to be applicable at each point of the treatment course including remission, relapse, or refractory disease.
This study provided the largest cohort of mortality in TTP patients to date with an estimated >1000 TTP-related deaths identified over 1100 hospitals nationwide. Among its limitations are the lack of information regarding ADAMTS13 data or particular treatments, and the exclusion of individuals who died before TPE initiation.

Application of the French TMA Reference Center Score and the Mortality in TTP Score in de novo and Relapsed Episodes of aTTP at a Tertiary Care Facility in Spain 2020
The above-mentioned study, conducted by the authors of this commentary (Domingo-González A. et al., [20]), was aimed to give a better characterization of relapses, comparing the prognosis of TTP relapses versus first episodes according to the French TMA Reference Center Score [18] and the Mortality in TTP Score [19]. Moreover, it compared laboratory and clinical data at presentation of de novo versus relapsed episodes of TTP, along with their response to treatment [20].
A total of 29 episodes of acquired TTP (16 de novo and 13 relapses) were analyzed. All patients were diagnosed and treated in a Spanish single tertiary care center between May 2008 and May 2020. The severe deficit of ADAMTS13 activity, defined as <10%, was an indispensable requirement, so other TMAs were excluded from the registry. Disease severity at presentation was evaluated by the French TMA Reference Center experience predictive model for death and the Mortality in TTP Score. Response to treatment was described according to Scully et al. [21]. Treatment protocol consisted of daily TPE and corticosteroids as first-line treatment. Rituximab was restricted to refractory or exacerbated cases up to 2017 when it was included as front-line therapy in relapsing TTP or first episodes of TTP with severe CNS involvement. In cases of refractory/ exacerbated TTP, the attending physicians were free to use any adjuvant treatment at their discretion.
Five out of 16 (31%) first episodes scored >2 in the French TMA Reference Center Score, whereas every relapsed episode scored ≤ 2 (p=0.027). The median number of points according to this score was 2 (r: 1-3) for the first group and 1 (r: 0-1) for the latter (p=0.006).
On the other hand, the worse short-term prognosis of de novo TTP episodes was sustained by the MITS. Four out of 16 (25%) de novo episodes scored >3 in MITS, compared to 0% of TTP relapses (p=0.052). The median number of points scored also reflected a higher risk of short-term mortality in first TTP episodes (1 r:1-4 vs 0 r:1-2, p=0.038) ( Table 5).
J Clin Haematol. 2021 Volume 2, Issue 2 69 The worse prognosis of first episodes compared to relapses may have been related to 2 clinical prognostic factors: older age and CNS involvement. Although, no significant difference in median age at presentation between first episodes and relapses was found, 6 out of 16 patients (37.5%) from the first group were older than 60 years whereas all recurrent TTP episodes occurred in patients below this age (p=0.013). As for CNS involvement, the prevalence was higher in first episodes compared to relapses (p=0.001) with no cases of severe neurological abnormalities identified in the last group.
Regarding clinical and laboratory findings, the results were in line with previous studies [22][23][24][25]. It is worth noting that bruising was by large the most common finding in relapses and it was found statistically significant compared to first episodes. Platelet count at presentation was found to be higher in recurrences than in the first disease episode (p=0.016) and ADAMTS13 activity <5% was more frequent in the last group (p=0.016) ( Table 6). Although this could be related to a faster medical evaluation, this study was not able to prove that the time that elapsed between the disease onset and the initiation of treatment was shorter in relapsed cases. Respecting treatment outcomes, there was no significant difference in the rate of refractoriness or exacerbations or in the number of days spent in the hospital (Table 7). Continuous variables with normally and nor normally distributed parameters were analyzed with the unpaired t-test and Mann-Whitney U-test respectively while categorical variables were analyzed with Chi-square. Calculations were carried out using SPSS version 12. ADAMTS13 activity at presentation was not available for the first episode in n = 2 patients, for relapses in n = 5 patients. ADAMTS13 normal range was 40-130%. Continuous variables with normally and nor normally distributed parameters were analyzed with the unpaired t-test and Mann-Whitney U-test respectively while categorical variables were analyzed with Chi-square. Calculations were carried out using SPSS version 12.

Conclusions
Despite the undeniable improvement in the prognosis of aTTP over the last few decades, aTTP is still related to significant mortality and relapsing tendency, with deficits in health-related quality in the long term. As a result, many prognostic scores have been developed to assist physicians in treating this life-threatening disease. The characterization of the pathophysiology of the disease and the great effort made by the scientific community to achieve consensus in terminology has enabled a continuous amelioration of these scores. In this commentary, we reviewed the main TTP prognosis scores and highlighted their contribution to the field. We also set out the results of the clinical application of the French TMA Reference Center Score and the Mortality in TTP Score in a tertiary hospital in Spain, to offer an up-dated characterization in terms of prognosis of first and relapsed episodes of TTP in our population.
In summary, the results presented are in line with previous studies and show that relapsed TTP episodes have a lower probability of short-term mortality compared to first TTP episodes. Nationwide studies are needed for prospective validation and to reevaluate these results considering new therapeutic strategies such as the addition of caplacizumab to standard treatment.

Funding
None.