Thrombotic microangiopathies: First report of 294 cases from a single institution experience in Argentina

Abstract Introduction Introduction: Thrombotic microangiopathies (TMAs) are rare disorders associated with fatal outcomes if left uncared for. However, healthcare problems in developing countries tend to limit medical assistance to patients. Methods Methods: We prospectively studied an Argentine cohort of 294 consecutive patients from 2013 to 2016. Patients’ subcategory classification relied on clinical symptoms and presence or absence of trigger events associated with TMA. Results Main suspected disorders were the primary TMAs known as thrombotic thrombocytopenic purpura (TTP) (n = 72/294, 24%) and atypical haemolytic uraemic syndrome (aHUS) (n = 94/294, 32%). In acute phase, demographic parameters for acquired TTP (aTTP) (n = 28) and aHUS (n = 47) showed that both groups were characterised by a young median age (37 and 25 years, respectively) and female predominance (60% and 86%). Median of a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 activity was significantly lower in aTTP than in aHUS group (1.4% vs 83%) and was associated with a more severe thrombocytopenia (15 × 109 vs 53 × 109/L). Creatinine (Cr) and urea (Ur) were significantly increased in aHUS compared to aTTP subjects (Cr: 3.7 vs 0.7 mg/dL, Ur: 118 vs 33 mg/dL). Gastrointestinal and neurological symptoms were more frequent in aHUS and aTTP, respectively. Conclusion The first description of a TMA cohort in Argentina revealed similar clinical presentations to those of other countries.

The list of entities belonging to TMA is extensive, and classification remains a challenge while some pathophysiological mechanisms associated with the syndromes are still unclear [3]. Haemolytic uraemic syndrome (HUS) is one of the two main TMA disorders investigated together with thrombotic thrombocytopenic purpura (TTP) in the past few decades. Although the two diseases have overlapping clinical features, HUS and TTP are pathophysiologically distinct entities [4]. TTP is characterised by severe acquired (aTTP) or congenital (cTTP) deficiency of a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS13). Both forms lead to persistent ultra large von Willebrand factor multimers, platelet activation and microvascular thrombosis [5]. The triad of acute MAHA, thrombocytopenia and acute kidney injury defines HUS. This syndrome is characterised by the typical form, secondary to an infection by Shiga toxin (Stx)-producing Escherichia coli (STEC) and the extremely rare atypical form (aHUS), caused by the dysregulation of the alternative pathway of the complement, leading to its activation. Elucidated pathological mechanisms of aHUS involve genetic or acquired abnormalities identified in complement components or coagulation-related factors [6]. TMA

Definition of syndromes
Patients who manifested a non-immune MAHA (haemoglobin < 12 g/dL with schistocytes, elevated LDH), thrombocytopenia (<150x10 9 /L platelets count or a decrease of 25% from baseline) with or without organ injury of variable severity were defined as suspected TMA (Figure 1). Some patients were identified as TMA after performing a renal biopsy. The first step was to measure ADAMTS13 activity. When activity was less than 10%, TTP was confirmed [7]. This diagnostic criterion identifies almost all patients but it was shown that the 10% cutoff level can be problematic in some cases where normal ADAMTS13 activity can be associated with high levels of ADAMTS13 autoantibodies [1].

Determination of ADAMTS13 parameters in plasma by immunoassay
Measurement of ADAMTS13 activity in human plasma was performed using Technozym ADAMTS13 Activity ELISA (Technoclone, Vienna, Austria). The anti-ADAMTS13 IgG antibodies assay used Technozym ADAMTS13 Inh ELISA (Technoclone, Vienna, Austria). Patient samples were negative when value was below 12 U/mL, positive when value was above 15 U/mL and considered borderline when value was between 12 and 15 U/mL. Patients with severe deficiency of ADAMTS13 activity but absence of IgG were tested for IgA, IgM or neutralising activity of anti-ADAMTS13 antibodies (data not shown).
(HIV, HCV, tuberculosis) were identified in six patients. Six other cases were associated with various events such as pancreatitis (n = 1), uterine fibroids (n = 2), mild pregnancy complication (n = 2) or a case of premature neonate with a significant patent ductus arteriosus. Confirmation of TMA by renal biopsy only, with incomplete pattern of clinical manifestations, was categorised as other TMA (n = 11, 4%). Twenty women (7%), diagnosed with preeclampsia or HELLP syndrome, were studied for ADAMTS13 parameters to discard TTP diagnosis. Finally, eight patients could not be classified due to lack of clinical data from the acute episode or absence of suitable blood samples for ADAMTS13 testing.

Demographic and laboratory parameters in TTP versus aHUS patients
Median age was higher in aTTP patients compared to cTTP (P = .06) and aHUS (P < .05) subjects (Table 1). Within the cTTP group, three experienced a first episode during pregnancy, five during childhood and two were diagnosed in adulthood. Acquired TTP and aHUS cohorts were Abbreviations: ADAMTS13, a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13; aHUS, atypical haemolytic uraemic syndrome; aTTP, acquired thrombotic thrombocytopenic purpura; cTTP, congenital thrombotic thrombocytopenic purpura.
composed of a majority of females whereas sex distribution was similar in cTTP group. Median of ADAMTS13 activity was significantly lower in TTP than in aHUS patients (P < .0001). The presence of IgG anti-ADAMTS13 confirmed aTTP diagnostic (Table 1). Patients with aTTP presented a significant (P < .0001) and more severe thrombocytopenia compared with aHUS ( Figure 5A). Low haemoglobin and haematocrit levels are characteristics of both TMA syndromes, and no significant difference was observed between aTTP and aHUS ( Figures 5B and 5C).
While no significant variation in LDH levels was observed between the two groups ( Figure 5D), creatinine and urea levels were significantly higher (P < .0001) in aHUS than in aTTP ( Figures 5E and 5F). Clinical parameters in cTTP patients recorded during first episode were similar to parameters observed in aTTP cohort independently of age at onset (data not shown).

Clinical presentation of extrarenal symptoms and supportive treatments in TTP and aHUS patients
Renal failure is part of aHUS diagnosis but extrarenal manifestations can be commonly observed in TMA disorders. Our cohort showed that aHUS patients were more likely to present gastrointestinal symptoms (40%) than aTTP subjects (4%) (P < .0005) ( Table 2). Both groups experienced alterations in neurological function; however it was shown to be significantly more frequent in aTTP (57%) than in aHUS (28%) (P < .05). Patients suffering from TMA require supportive care earlier.
Atypical HUS patients rationally presented a higher rate of dialysis than the aTTP group (62% vs 4%, P < .0001). In contrast, corticosteroids seemed to be more frequently administered to patients with aTTP diagnosis (75%) than patients with aHUS clinical manifestations (32%), but this difference was not significant. In the first 48 hours of the acute phase, replacement therapy was applied as frequently to F I G U R E 5 Laboratory parameters analysis in aTTP versus aHUS patients during acute phase ***P < .05, Mann-Whitney U test. Abbreviations: aHUS, atypical haemolytic uraemic syndrome; aTTP, acquired thrombotic thrombocytopenic purpura; LDH, lactate dehydrogenase. the aTTP cohort as to aHUS individuals ( Table 2

DISCUSSION
TMAs are rare disorders and life-threatening entities whose diagnosis is challenging for physicians. Most reports on TMAs have been published mainly in Western Europe [8][9][10] and in the United States [11]. More recently, data from the Asian-Pacific region have started to emerge, including cohort studies from Australia [12], China [13] and Korea [14]. In Latin America, TMA research has mostly focused on STEC-HUS, described as endemic in Argentina and Uruguay [15].
Our study described the frequency of TMA in an Argentine consecutive cohort for the first time. Compared to a similar study [16], TTP and aHUS were the most frequently suspected entities. A recent retrospective research on a large consecutive cohort of TMA patients showed that less than 10% were TTP, aHUS or STEC-HUS cases [17].
The authors concluded that divergences in TMA frequency and incidence among studies are likely due to the criteria initially used for patient selection. Compared to studies in the literature, our limited access to complement screening for aHUS diagnosis reduced the scope of our results. However, we were able to confirm TTP diagnosis in 46% of suspected cases in our Argentine cohort. From our results we estimated the annual incidence of aTTP in Argentina to be seven cases, corresponding to 0.2 cases per million people per year. Even if aTTP is more frequent than the inherited form, it is noteworthy to mention that prevalence of cTTP in our Argentine cohort (n = 10/39, 26%) was higher than in the literature. This could be partially explained by the proportion of patients excluded from the analysis (n = 20/59), twelve of whom had been previously diagnosed with aTTP.
Regarding demographic characteristics, both aTTP and aHUS groups were characterised by a predominance of females, as observed previously [18,19]. Our subject population of suspected TMA was young (median age 31 years), 25% being paediatric patients. In a similar cohort size, median age was slightly higher (34 years) in a Malaysian study (n = 243) with 30% of cases under 24 years [20]. In a German cohort, where the population of TMA paediatric cases was under-represented (n = 13/232, 6%), the mean age at enrolment was 53 years [21]. Our median age for adult population only was 36 years, which supports literature demonstrating that TMA is more frequent in young adults [1]. The median age difference observed between aTTP and aHUS patients in our cohort reflects that the incidence of aTTP is higher in adults than in children, while aHUS primarily affects children and young adults. We should not discard the possibility that the average lower age of our cohort could be the effect of a referral bias due to inequity in accessing quality healthcare by elderly population in Argentina. This same selection bias, in addition to the difficulties of accessing mid or high-complexity healthcare centres for TMA diagnosis could have resulted in collecting and evaluating a higher percentage of severe cases compared to other cohorts worldwide.
The commercial chromogenic assay Technozym ADAMTS13 Activity ELISA (Chr-VWF73) is one of the most popular clinical laboratory methods for ADAMTS13 activity. However lack of standardisation in methodologies and sources of discrepancy between them can be challenging if carried out by non-expert laboratories. Our reference centre has accumulated experience of studying TTP patients over the past 20 years [22,23]. This expertise is important in the management of cases that might present divergence between laboratory results and clinical manifestations, as observed by George [24]. Samples of patients that present overlapping results and show inconsistency with clinical manifestations are rare but may occur [25,26]. A recent consensus on clinical utility of assays recommends repeating ADAMTS13 testing in TMA patients with an indeterminate range of 10-20% of activity [27]. In our cohort, 10% (n = 3) of our confirmed aTTP patients presented a normal ADAMTS13 activity (between 10% and 30%). Repeating the assay using a new sample from the patients was challenging as subjects would usually receive replacement therapy at the earliest.
Additional methods in the laboratory were then useful to support a diagnosis including evaluation of ultra large VWF multimers and collagen binding assay. For aHUS diagnosis, no complement biomarker alone has proved its utility, but various serological laboratory tests are of interest if examined altogether [28]. However, the lack of resources encountered by developing countries is limiting their capacity to manage rare syndromes like TMA properly, as exposed by a recent publication from India [29]. Lack of clinical or laboratory data can be a limiting factor when studying TMA. In the past few years, the PLASMIC score was validated in the literature as a useful tool to predict TTP early [30][31][32]. This prognostic score as well as others was developed with the objective of reducing the time from diagnosis to treatment [33]. The use of a prognostic score is of great interest in Argentina as resources for TMA diagnosis can be limited. Moreover, our results clearly showed that the two parameters described as useful and reliable to predict severe ADAMTS13 deficiency, platelet count and serum creatinine [34], were indeed both statistically significant when comparing patients from TTP and aHUS groups. This study provides the first report of epidemiological information about these rare diseases in Argentina. Accumulating data and knowledge about all TMA disorders aims to offer healthcare professionals the possibility to improve management, treatment and prevention of affected patients.

ACKNOWLEDGEMENTS
We would like to thank the patients and their families, the physicians and the institutions who contributed to this study. This work was supported by grants from Alexion-Argentina and from CONICET-MINCyT (PIP 2017 #11220170100338). Célia Dos Santos and Sabrina Rotondo received a graduate studentship award from Fundación René Baron.
Lastly, we would like to pay our gratitude and our respects to our former director and colleague Dr. María Angela Lazzari, who passed away in August 2018, and whose contribution to the development of research in our Department was most significant and will always be remembered.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.