The utility of ADAMTS13 in differentiating TTP from other acute thrombotic microangiopathies: results from the UK TTP Registry

research paper The utility of ADAMTS13 in differentiating TTP from other acute thrombotic microangiopathies: results from the UK TTP Registry Sevda Hassan, 1 John-Paul Westwood, 2 Debra Ellis, 2 Chris Laing, 3 Siobhan Mc Guckin, 2 Sylvia Benjamin 4 and Marie Scully 2 1 Department of Nephrology, The Royal London Hospital, 2 Department of Haematology, University College London Hospital, 3 UCL Centre for Nephrology, Royal Free Hospital, London, and 4 NHSBT Therapeutic Apheresis Services (TAS) and Oxford University Hospitals Trust, Oxford, UK Received 2 April 2015; accepted for publication 22 July 2015 Correspondence: Dr John-Paul Westwood, Department of Haematology, University College London Hospital, 1st Floor Central, 250 Euston Road, London NW1 2PQ, UK. E-mail: j.westwood@ucl.ac.uk Summary Thrombotic microangiopathies (TMAs) are frequently difficult to differentiate clinically, and measurement of ADAMTS13 (a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13) remains vital in thrombotic thrombocytopenic purpura (TTP) diagnosis. We retrospectively reviewed cases referred for ADAMTS13 testing, using UK TTP Registry screening data. Of a total 810 cases, 350 were confirmed as TTP. The 460 non-ttp cases comprised secondary TMAs (2457%) and haemolytic uraemic syndrome (HUS) (2717% ahus, 283% Shiga-like toxin-producing E. coli [STEC]-HUS); the remainder were TMAs with no clear association, not TMAs, or had no confirmed diagnosis. ADAMTS13 levels were significantly lower in TTP than STEC-HUS, ahus and other TMAs. TTP patients had significantly lower platelet count (15 9 10 9 /l; range 0 96) than ahus (57 9 10 9 /l; range 13 145, P < 00001) or STEC-HUS (35 9 10 9 /l; range 14 106, P < 00001); they also had lower creatinine levels (92 lmol/l; range 43 374) than ahus (255 lmol/l; range 23 941, P < 00001) and STEC-HUS (324 lmol/l; range 117 639, P < 00001). However, 12/34 (353%) ahus patients had a platelet count <30 9 10 9 /l and 26/150 (173%) of TTP patients had a platelet count >30 9 10 9 /l; 23/ 150 (153%) of TTP patients had a creatinine level >150 lmol/l. This study highlights the wide variety of TMA presentations, and confirms the utility of ADAMTS13 testing in TTP diagnosis. Keywords: thrombotic microangiopathies, thrombotic thrombocytopenic purpura, haemolytic uraemic syndrome, ADAMTS13. The term thrombotic microangiopathies (TMAs) describes a group of disorders characterized by thrombocytopenia, microangiopathic haemolytic anaemia (MAHA) and microvascular thrombosis. Of these, thrombotic thrombocytopenic purpura (TTP) and haemolytic uraemic syndrome (HUS) are two of the best characterized, and typically target the brain and heart in the case of TTP, and the kidney in HUS. Greater understanding of the pathophysiology of these disorders has lead to changes in definitions, diagnostic criteria and treatment. Thrombotic thrombocytopenic purpura is most commonly caused by antibody-mediated depletion of the metalloprotease ADAMTS13 (a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13) (Furlan et al, 1998; Tsai & Lian, 1998). HUS is frequently caused by the action of Shiga-like toxin on the renovascular First published online 11 September 2015 doi: 10.1111/bjh.13654 endothelium, a sub-type known as STEC (Shiga-like toxinproducing E. coli)-hus; a more rare form, known as atypical HUS (ahus), is a disorder driven by defective complement regulation. Despite advances in our understanding of these disorders, differentiating between them clinically can be challenging, and frequently other more common TMAs need to be excluded. This is particularly important given that treatments for these disorders differ, and not all benefit from plasma exchange. Effective therapies exist for TTP (immunosuppression) and ahus (complement C5 inhibition with eculizumab), and early definitive treatment is associated with improved outcomes in both disorders (Scully & Goodship, 2014). Differentiating between STEC-HUS and ahus is also important, given that STEC-HUS often responds to supportive therapy alone. ª 2015 John Wiley & Sons Ltd

Utility of ADAMTS13 in differentiating TTP from other TMAs Currently, demonstration of low ADAMTS13 levels (<10%) remains the only definitive method for diagnosis of TTP, but the availability of this test is limited. Atypical HUS can be diagnosed by screening for mutations in genes encoding complement regulatory proteins, but this is timeconsuming and mutations are only detected in up to 60% of cases (Kavanagh & Goodship, 2010). Lack of rapid diagnostic methods and the need to instigate treatment early has resulted in a reliance on clinical criteria alone, pending a conclusive diagnosis. In the case of differentiating TTP and ahus, it has been suggested that platelet counts (>30 9 10 9 /l in ahus and <30 9 10 9 /l in TTP) and creatinine (>150 lmol/l in ahus) are useful (Coppo et al, 2010; Zuber et al, 2012). As our understanding of the phenotype-genotype relationship of ahus expands, overlap between these groups is evident. Historically, the presence or absence of bloody diarrhoea was thought to be central to the diagnosis of HUS, but this is most relevant in STEC-HUS and less common in ahus. The UK TTP registry collects all cases of TTP in the UK. However, data is also gathered via referrals for ADAMTS13 testing in patients with a presumed diagnosis of ahus, in whom testing is a requirement prior to treatment with eculizumab. Furthermore, other conditions presenting as acute TMAs may require ADAMTS13 testing, where TTP is not the final diagnosis based on ADAMTS13 levels or further investigations. The aim of this study was to perform a cross-sectional analysis of patients referred for ADAMTS13 testing between 2009 and 2013, including TTP and non-ttp cases, to enable the generation of summary data and identify the underlying final diagnosis for each case. This allowed comparison of ADAMTS13 levels in TTP, HUS (ahus and STEC-HUS) and other TMAs, with the aim of better understanding how ADAMTS13 levels may help differentiate these cases. Finally, where available, a comparison of baseline laboratory characteristics in TTP, STEC-HUS and ahus was performed; it was hoped this would help determine the utility of these clinical parameters in differentiating these disorders. Methods Patients A retrospective review was performed of cases referred for ADAMTS13 testing over a 5-year period, between January 2009 and December 2013, using data obtained from the UK TTP Registry (Multicentre Research Ethics Committee [MREC] reference 08/H0810/54). The Registry was set up in January 2009 with the primary aim of capturing all cases of TTP presenting in the UK, and prospectively collects clinical and biological data. Screening data, including final diagnosis, is collected on non-ttp cases where ADAMTS13 testing has been requested; these include patients with HUS (ahus and STEC-HUS) and other TMAs. Classification As part of this review, Registry data was used to classify non- TTP cases into 6 categories including: ahus, STEC-HUS, secondary TMA, TMA other, non-tma and no diagnosis. The secondary TMA category was used for patients with a TMA having a recorded association (e.g. cancer, autoimmune disease). Those patients recorded as having a TMA but with no clear association, were placed into the TMA other category. All remaining unclassified patients were placed into the non-tma or no diagnosis category. Where the final diagnosis had not been recorded, the referring centre was contacted to obtain this where possible. In a proportion of non-ttp cases, baseline laboratory parameters at presentation were available, although this was not a specific requirement of the Registry. Laboratory parameters included Hb, platelet count and creatinine level. ADAMTS13 results were recorded for all patients. TTP cases were defined as those having an ADAMTS13 activity <10% and/or the presence of anti-adamts13 IgG antibodies. Diagnosis of TMA included a specific blood film characteristic of microangiopathic haemolytic anaemia (polychromasia, schistocytes, anaemia), with thrombocytopenia (platelet count <150 9 10 9 /l), reticulocytosis and a raised lactate dehydrogenase level (Scully et al, 2012). ADAMTS13 assays ADAMTS13 activity was analysed by modification of the Fluorescence Resonance Energy Transfer (FRETS) assay (normal range: 60 123%) (Kokame et al, 2005). This was undertaken in all cases captured in the registry, using citrated samples taken prior to any plasma therapy. In those cases with ADAMTS13 <30%, anti-adamts13 immunoglobulin G (lgg) levels were analysed using an enzyme-linked immunosorbent assay technique (Chow et al, 2007; Scully et al, 2007). Statistical Analysis Given laboratory data measured (ADAMTS13, platelet count and creatinine) was not normally distributed, non-parametric methods were used for analysis. The Mann Whitney U-test was used to compare parameters across groups, with a P value of <005 being considered statistically significant. Results Patient demographics and subgroups Over the 5-year period (2009 2013) there were a total of 810 samples referred for ADAMTS13 testing, which resulted in final diagnoses of 350 TTP cases and 460 non-ttp cases. Of the 350 TTP cases, 225 were female (643%) and 108 were male (309%), with 48% not recorded; median age of these ª 2015 John Wiley & Sons Ltd 831

S. Hassan et al patients was 44 years (range birth to 89 years). Of the 460 non-ttp cases, 230 were female (50%) and 146 male (317%), with 84 (183%) not recorded; the median patient age was 45 years (range birth to 90 years). Figure 1 shows the subgroups of the 460 non-ttp cases. Of these, the largest subgroup was secondary TMAs, comprising 113 cases (2457%). There were 138 HUS cases, comprising 125 ahus (2717%) and 13 STEC-HUS (283%). The TMA other group consisted of 31 cases (674%). Twentytwo cases (478%) were in the non-tma category and no diagnosis was available in 152 cases (3391%). HUS: ahus & STEC-HUS Of the 138 ahus/stec-hus cases in total, the median age of patients was 40 years (range birth to 87 years); 69 (50%) were female and 51 (37%) male, with 18 (13%) not recorded. Of the 125 ahus cases, 34 had further laboratory/ clinical data available. The most common presentation was with gastrointestinal symptoms, affecting 10/34 (294%) patients. Unlike the STEC-HUS cases, ahus patients who presented with diarrhoea were not found to have positive stool cultures. 8/34 (235%) of patients presented with renal impairment alone (no other features at presentation, other than MAHA and thrombocytopenia). 10/34 (294%) patients had neurological symptoms at presentation, including seizures (4/34), headache (3/34), focal deficits (3/34) and flitting neurology (1/34). Within the STEC-HUS subgroup (n = 13), the underlying microbial infection was identified in all but two cases. The Fig 1. Non-TTP cases (n = 460) divided into six subgroups. TTP, thrombocytopenic purpura; ahus, atypical haemolytic uraemic syndrome; STEC-HUS, Shiga-like toxin-producing E. coli haemolytic uraemic syndrome; TMA, Thrombotic microangiopathy bacteria identified were Escherichia Coli (n = 9), Streptococcus (n = 1) and Clostridium Difficile (n = 1). Of the remaining two cases, one had features consistent with infective colitis confirmed on sigmoidoscopy; the other was given a diagnosis of STEC-HUS but no further information was available. Secondary TMA and Non-TMA subgroups The underlying diagnoses varied in the Secondary TMA subgroup. Of these 113 cases, the largest groups were patients with cancer (23%; both solid organ and haematological) and liver disease (274%). Pregnancy-related disorders (including HELLP syndrome [haemolysis, elevated liver enzymes, low platelet count], acute fatty liver, premature rupture of membranes) comprised 142% of cases, autoimmune disorders 106% (including lupus, Sjogren syndrome, mixed connective tissue disorders and scleroderma), transplant patients (including bone marrow transplant) 115% and sepsis 796%. Of the 22 cases in the non-tma subgroup, 10 patients had ITP; the others included vitamin B12 deficiency, sickle cell disease and autoimmune haemolysis. Laboratory parameters: TTP vs other groups Laboratory parameters (ADAMTS13 levels, platelet count and creatinine) are shown in Table I. ADAMTS13 levels. ADAMTS13 levels for each group are shown in Fig 2. Patients with TTP had a significantly lower median ADAMTS13 levels (5%, range 0 11) than patients with ahus (665%, range 12 119, P < 00001), HUS (56%, range 11 82, P < 00001), or MAHA/TMA (51%, range 5 105, P < 00001). There was no significant difference in ADAMTS13 levels between patients in the ahus, HUS or MAHA/TMA groups. Platelet count. Patients with TTP had a median platelet count of 15 9 10 9 /l (range 0 96), which was significantly lower than patients with ahus (57 9 10 9 /l; range 13 145, P < 00001) or HUS (35 9 10 9 /l; range 14 106, P < 00001) (Fig. 3A). There was no significant difference between the ahus and HUS groups. However, 12/34 (353%) patients in the ahus group had a platelet count <30 9 10 9 /l and Table I. Laboratory parameters by patient group. TTP ahus STEC-HUS Secondary TMA Median ADAMTS13 (%) (range) 5 (0 11) 665 (12 119) 56 (11 82) 65 (12 130) Median platelet count (9 10 9 /l) (range) 15 (0 96) 57 (13 145) 35 (14 106) N/A Median creatinine level (lmol/l) (range) 92 (43 374) 255 (23 941) 324 (117 639) N/A TTP, thrombocytopenic purpura; ahus, atypical haemolytic uraemic syndrome; STEC-HUS, Shiga-like toxin-producing E. coli haemolytic uraemic syndrome; TMA, Thrombotic microangiopathy; ADAMTS13, a disintegrin and metalloproteinase with a thrombospondin type 1motif, member 13. 832 ª 2015 John Wiley & Sons Ltd

Utility of ADAMTS13 in differentiating TTP from other TMAs 26/150 (173%) of TTP patients had a platelet count >30 9 10 9 /l. Creatinine level. Patients with TTP had a significantly lower median creatinine level (92 lmol/l; range 43 374) than patients with ahus (255 lmol/l; range 23 941, P < 00001) and patients with HUS (324 lmol/l; range 117 639, P < 00001) (Fig 3B). However, 23/150 (153%) of TTP patients had a creatinine level >150 lmol/l. Discussion This review, encompassing a large number of TMA cases referred for ADAMTS13 testing, has not only demonstrated the variety of disorders for which ADAMTS13 testing is Fig 2. ADAMTS13 (%) levels in each group. TTP patients had levels that were significantly lower than patients with ahus, STEC-HUS or secondary TMA (P < 00001). TTP, thrombocytopenic purpura; ahus, atypical haemolytic uraemic syndrome; STEC-HUS, Shiga-like toxin-producing E. coli haemolytic uraemic syndrome; TMA, Thrombotic microangiopathy; ADAMTS13, a disintegrin and metalloproteinase with a thrombospondin type 1motif, member 13. requested, but also allowed comparison of laboratory features of TTP and HUS/aHUS. These are rare disorders and this study has highlighted the utility of the UK TTP Registry in gathering information across many centres nationally, to help further understanding of the diagnostic issues relating to TMAs. Overall, more cases referred were non-ttp than TTP (460 vs. 350). Unlike TTP, which demonstrates a typical female preponderance (Scully & Goodship, 2014), the non-ttp cases had an equal male: female presentation. However, the age distribution was comparable with the confirmed TTP group. Although approximately a quarter of non-ttp referrals were for patients with a presumed diagnosis of ahus [where ADAMTS13 testing is now mandatory prior to starting Eculizumab therapy (Scully & Goodship, 2014)], the majority of patients in the non-ttp category would have had an ADAMTS13 test sent on the basis that TTP was a possible differential. This highlights the fact that, at least for the cases in this study, clinical parameters alone were not felt to sufficiently rule out a TTP diagnosis. Atypical HUS is usually associated with oliguric/anuric renal failure, but not all cases have such acute renal impairment (Loirat & Fremeaux-Bacchi, 2011). In contrast, undiagnosed congenital TTP can present with significant renal impairment (Scully et al, 2012). ADAMTS13 activity was significantly higher in HUS than in TTP cases, and those with the lowest levels (e.g. 16%) were not associated with anti- ADAMTS13 antibodies. Similarly, in other TMAs, the ADAMTS13 activity helped to differentiate these disorders from acute TTP. Haemolytic uraemic syndrome may have an infective precipitant and requires supportive care. However, there were still STEC-HUS positive cases for which ADAMTS13 testing was requested, suggesting that even these may be associated with a diagnostic dilemma, in particular, delay in available confirmation of the infective trigger. Atypical HUS, which is complement-mediated, has improved outcomes with the complement inhibitor, Eculizumab. It has been suggested Fig 3. (A) Platelet count (9 10 9 /l) by group. Patients with TTP had significantly lower platelet counts than those with ahus or HUS (P < 00001). (B) Creatinine level (lmol/l) by group. Platelet counts were significantly lower in the TTP group when compared with patients with ahus or HUS (P < 00001). TTP, thrombocytopenic purpura; ahus, atypical haemolytic uraemic syndrome; STEC-HUS, Shiga-like toxin-producing E. coli haemolytic uraemic syndrome. ª 2015 John Wiley & Sons Ltd 833

S. Hassan et al that a platelet count of >30 9 10 9 /l and creatinine of >150 lmol/l is satisfactory to exclude TTP (Coppo et al, 2010). While access to ADAMTS13 assays may be difficult, we have shown that within the TTP confirmed and non-ttp cases, there is significant overlap using routine laboratory parameters. Indeed, for approximately a fifth of all TTP cases, using routine laboratory parameters would have failed to identify the correct diagnosis, which could have an impact on therapy. Similarly, of the cases referred for ADAMTS13 testing for suspected ahus, more than a third had platelet counts less than 30 9 10 9 /l and a fifth had creatinine levels less than 150 lmol/l. ADAMTS13 results are required preconfirmation of complement inhibitor therapy for ahus in the UK (National Institute for Health and Care Excellence [NICE] 2015), to ensure TTP and, specifically, congenital TTP are not misdiagnosed. The most common conditions identified in patients presenting with a TMA who did not have TTP were cancer, liver disease and pregnancy-associated TMA. The utility of measuring ADAMTS13 levels is to determine if plasma exchange should be instigated or continued in such scenarios; this is best exemplified in the case of cancer-associated TMA, where plasma exchange has not been found to be useful (Fontana et al, 2001). Continuing plasma exchange is not without complications, including large volumes of plasma and central venous access insertion (Som et al, 2012). It was unexpected that a quarter of all cases within this group were due to cancer. It is unknown if this was the initial presentation or progressive disease associated with a TMA. Furthermore, the acute diagnostic dilemma occasionally heralds the requirement to undertake specialist testing, but subsequently, more common diagnoses were confirmed, e.g. ITP or anaemia associated with vitamin B12 deficiency. Vitamin B12 deficiency has been associated with presentation of thrombocytopenia and MAHA (Noel et al, 2013). A limitation of this study is that the UK TTP registry only captures the complete data set from TTP cases. Samples referred that are not TTP are not required to provide extended information. Therefore, presenting blood counts and final diagnosis were not available for all non-ttp cases. There is the potential that the final diagnosis differs from the initial diagnosis considered at the time of sending the ADAMTS13 sample this study was not designed to capture these cases, but based on the numbers in the non-tma group, in at least 5% of referrals the patient did not have a TMA (this number may well be larger than this, given the numbers in the No diagnosis category). In conclusion, ADAMTS13 assays are very useful to differentiate TTP from other associated TMAs. From this dataset, the number of samples requested suggests that, within the acute clinical situation, diagnoses can be challenging but that the assays usually provide a clear differentiation of other diagnoses from TTP. However ADAMTS13 testing is not routinely available in all centres, and urgent assays are often not performed out of hours. Accuracy of assays and availability of advice regarding interpretation is vital to enable appropriate use of this test. Although routine laboratory parameters (e.g. platelets, creatinine) can help differentiate TTP from other TMAs, we have demonstrated that there is approximately a 20% overlap using laboratory parameters alone, which could have significant therapeutic consequences. The mortality is similar in both instances; 10 20% TTP and 25% atypical HUS (Scully & Goodship, 2014). Therefore, prompt identification, treatment and referral are essential and highlights the importance of using a number of methods, including ADAMTS13 assays, to confirm the diagnosis. Acknowledgements SH designed the study, collected and analysed data and wrote the manuscript. JW analysed data and revised the manuscript. DE and SM collected data. CL and SB revised and approved the manuscript. MS designed the study, wrote and revised the manuscript. Conflicts of interest This project was supported by an unrestricted educational grant from Alexion. References Chow, K.V., Carroll, R., Branley, P., Nicholls, K., Becker, G. & Hogan, C. (2007) Anti-CD20 antibody in thrombotic thrombocytopenic purpura refractory to plasma exchange. Internal Medicine Journal, 37, 329 332. Coppo, P., Schwarzinger, M., Buffet, M., Wynckel, A., Clabault, K., Presne, C., Poullin, P., Malot, S., Vanhille, P., Azoulay, E., Galicier, L., Lemiale, V., Mira, J.P., Ridel, C., Rondeau, E., Pourrat, J., Girault, S., Bordessoule, D., Saheb, S., Ramakers, M., Hamidou, M., Vernant, J.P., Guidet, B., Wolf, M. & Veyradier, A. & French Reference Center for Thrombotic Microangiopathies. (2010) Predictive features of severe acquired ADAMTS13 deficiency in idiopathic thrombotic microangiopathies: the French TMA reference center experience. PLoS ONE, 5, e10208. Fontana, S., Gerritsen, H.E., Kremer Hovinga, J., Furlan, M. & Lammle, B. (2001) Microangiopathic haemolytic anaemia in metastasizing malignant tumours is not associated with a severe deficiency of the von Willebrand factorcleaving protease. British Journal of Haematology, 113, 100 102. Furlan, M., Robles, R., Solenthaler, M. & Lammle, B. (1998) Acquired deficiency of von Willebrand factor-cleaving protease in a patient with thrombotic thrombocytopenic purpura. Blood, 91, 2839 2846. Kavanagh, D. & Goodship, T. (2010) Genetics and complement in atypical HUS. Pediatric Nephrology(Berlin, Germany), 25, 2431 2442. Kokame, K., Nobe, Y., Kokubo, Y., Okayama, A. & Miyata, T. (2005) FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. British Journal of Haematology, 129, 93 100. Loirat, C. & Fremeaux-Bacchi, V. (2011) Atypical hemolytic uremic syndrome. Orphanet journal of rare diseases, 6, 60. NICE (2015) Eculizumab for Treating Atypical Haemolytic Uraemic Syndrome. NICE Highly Specialised Technology Guidance (HST1). 834 ª 2015 John Wiley & Sons Ltd

Utility of ADAMTS13 in differentiating TTP from other TMAs National Institute for Health and Care Excellence, London. Available at: http://www.nice.org.uk/guidance/hst1. Noel, N., Maigne, G., Tertian, G., Anguel, N., Monnet, X., Michot, J.M., Goujard, C. & Lambotte, O. (2013) Hemolysis and schistocytosis in the emergency department: consider pseudothrombotic microangiopathy related to vitamin B12 deficiency. QJM, 106, 1017 1022. Scully, M. & Goodship, T. (2014) How I treat thrombotic thrombocytopenic purpura and atypical haemolytic uraemic syndrome. British Journal of Haematology, 164, 759 766. Scully, M., Cohen, H., Cavenagh, J., Benjamin, S., Starke, R., Killick, S., Mackie, I. & Machin, S.J. (2007) Remission in acute refractory and relapsing thrombotic thrombocytopenic purpura following rituximab is associated with a reduction in IgG antibodies to ADAMTS-13. British Journal of Haematology, 136, 451 461. Scully, M., Hunt, B.J., Benjamin, S., Liesner, R., Rose, P., Peyvandi, F., Cheung, B. & Machin, S.J. (2012) Guidelines on the diagnosis and management of thrombotic thrombocytopenic purpura and other thrombotic microangiopathies. British Journal of Haematology, 158, 323 335. Som, S., Deford, C.C., Kaiser, M.L., Terrell, D.R., Kremer Hovinga, J.A., Lammle, B., George, J.N. & Vesely, S.K. (2012) Decreasing frequency of plasma exchange complications in patients treated for thrombotic thrombocytopenic purpurahemolytic uremic syndrome, 1996 to 2011. Transfusion, 52, 2525 2532; quiz 2524. Tsai, H.M. & Lian, E.C. (1998) Antibodies to von Willebrand factor-cleaving protease in acute thrombotic thrombocytopenic purpura. New England Journal of Medicine, 339, 1585 1594. Zuber, J., Fakhouri, F., Roumenina, L.T., Loirat, C. & Fremeaux-Bacchi, V. (2012) Use of eculizumab for atypical haemolytic uraemic syndrome and C3 glomerulopathies. Nature Reviews Nephrology, 8, 643 657. ª 2015 John Wiley & Sons Ltd 835