Differential diagnosis of thrombotic microangiopathy in nephrology

Thrombotic microangiopathy (TMA) manifests as a histological lesion of the microvasculature characterised by thickened and swollen vessel walls, detachment of endothelial cells, build-up of proteins and cell lysis material in the sub-endothelial space, and obstruction of the vascular lumen by platelet thrombi [1].

Thrombotic thrombocytopenic purpura (TTP) and haemolytic uraemic syndrome (HUS) are the two most common clinical conditions characterised by TMA lesions but have differing aetiology, pathophysiology and management strategies [1]. TTP is a systemic disorder of microvascular thromboses due to a deficiency in a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS13) activity, as a result of autoantibodies or genetic mutations. A severe reduction of ADAMTS13 activity (<10%) results in the formation of ultra-high molecular weight von Willebrand factor multimers, causing aggregation of platelets and end-organ ischaemia [2]. HUS meanwhile, is a clinical syndrome characterised by the obstruction of microvasculature (most commonly in the kidney) by platelet-fibrin thrombi despite normal ADAMTS13 activity. Shiga-toxin producing Escherichia coli (STEC) is the most common cause of HUS (STEC-HUS) [3]. History of an accompanying condition, including the presence of non-STEC infections such as Streptococcus pneumoniae or influenza virus infection, malignant hypertension, transplantation, pregnancy and child birth, or drug usage may suggest a diagnosis of secondary TMA [3]. The atypical form of HUS (aHUS) is a rare, life-threatening disease of chronic, uncontrolled complement activation that leads to TMA with severe organ damage. The differential diagnosis of aHUS requires the exclusion of TTP and STEC-HUS [1]. The rapid progression of TMA, associated with potentially irreversible damage to organs in patients with aHUS, indicates a need for urgent treatment. Historically, disease outcomes have been poor, despite previous practice to manage aHUS by intensive plasma exchanges [4]. More recently, eculizumab, a humanised monoclonal antibody that binds to the complement protein C5 preventing the formation of the membrane attack complex, has been shown to be effective in treating patients with aHUS [5‐7]. A recent report demonstrated a greater and more sustained recovery in renal function when eculizumab therapy is initiated within 7 days of aHUS onset [8].

The differential diagnosis of TMA is complex but important to inform treatment decisions. Consensus guidelines for the differential diagnosis of TMA were updated in 2015 [1], however real-world evidence on the current diagnosis and treatment practices has not been reported. We hypothesised that there are areas of uncertainty in the work-up of patients with TMA, leading to potential delays in establishing the final diagnosis, putting patient’s lives at a higher risk.

We devised a survey with the objective of understanding current practices in the diagnosis of TMA and aHUS across Europe and the Middle East and challenges when diagnosing the cause of TMA.

This survey provides a large source of information describing current practices in the differential diagnosis of TMA amongst experienced clinicians in 16 countries in Europe and the Middle East.

Clinical guidelines are an important tool to facilitate clinical decision making and promote the delivery of evidence-based best practice [9]. However, clinical guidelines for rare diseases are often difficult to formulate due to a lack of quality evidence, surveys of clinical practices, and data about patients’ opinion (https://​www.​iqwig.​de/​download/​V10-01_​Executive_​Summary_​Evidence_​for_​guidelines_​on_​rare_​diseases.​pdf). Recommendations provided within current guidelines are opinions, based on the best available scientific evidence [1, 10‐12].

Responses suggest the awareness, use, or perceived importance of published recommendations [1, 10‐12] available for the management of patients with TMA is inconsistent between countries and even within hospitals from the same country. Adult nephrologists seem to be less aware of published guidelines, however this was rated as the least challenging aspect of making an aHUS diagnosis suggesting limited clinical impact.

In accordance with recent recommendations [1, 10‐12] the triad of haemolytic anaemia, thrombocytopenia and AKI were the three main criteria indicating a diagnosis of TMA. However, the presence of schistocytes and increased lactate dehydrogenase (LDH) were not considered primary criteria in the diagnosis of TMA by many physicians. Considering the rapid progression and severity of TMA, guidelines recommend a rapid differential diagnosis to be established allowing for appropriate supportive measures to be taken with the first 48 h from admission [1]. In practice however, approximately half of clinicians state a diagnosis of TMA takes more than 3 days. Similarly, therapeutic strategies are initiated after 3 days or more in the majority of cases (57%). These data support those from a recent survey conducted by the aHUS Alliance whereby less than 55% of 233 patients with aHUS received their diagnosis within a week of presentation [13]. Determining ADAMTS13 activity levels and the absence of STEC infection in a timely manner (<24 h) is crucial in discriminating TTP from other TMA [1, 2]. By identifying ADAMTS13 activity levels early, appropriate treatment can be initiated quickly, and patients without TTP avoid unnecessary plasma exchange. Thereby patient outcomes may be improved.

Despite being part of recommended algorithms for the differential diagnosis of TMA [1, 10] only two-thirds of respondents agreed that an ADAMTS13 activity >10% rules out TTP. aHUS is a rare disease and as such awareness of the role of ADAMTS13 testing in the differential diagnosis may be low. One guideline recommends ADAMTS-13 activity <5–10% is indicative of TTP [1]. However, other reports have used <5% [11] and <10% [14] as a reference value to inform on diagnosis. It is possible that some centres may adopt ADAMTS-13 activity >5% as a lower limit to rule out a differential diagnosis of TTP, which may partially explain this result. Alternatively, there are also reports stating 10–25% of TTP patients have normal ADAMTS-13 activity [15] with activity levels varying from 33 to 100% among patients with apparent idiopathic TTP [16]. The response received in the survey may reflect that some physicians are not medically convinced that ADAMTS13 > 10% rules out TTP in 100% of cases.

The systemic nature of TMA leads to the involvement of the microvasculature of organs other than the kidney, with neurological manifestations being the most common [17]. This is reflected in 87% of respondents stating they consider neurological symptoms when considering a TMA diagnosis. However, it is important to consider that many other organs can be affected by TMA [18].

One guideline recommends a complete and detailed clinical history should be made for all TMA patients, including personal and family history [1], unless an EHEC infection is obvious. Current data suggest this is not routinely the case, as half of respondents reported that a family history would be determined by asking the patient only. A more thorough family history would be beneficial, as the majority of the reported aHUS associated mutations are dominant but with incomplete penetrance [17].

Gene mutations or polymorphisms affecting complement regulators or proteins are identified in 60–70% of aHUS patients [19, 20]. While identification of a complement mutation or anti-complement factor H (CFH) inhibitory antibody is not required to make a diagnosis of aHUS or initiate therapy, it does inform on prognosis and risk of recurrence [1, 8, 20, 21]. In this study, genetic testing is routinely performed by just under half of the clinicians managing patients with aHUS, but more commonly by paediatric nephrologists. In contrast, a survey conducted by the aHUS Alliance observed 84% of patients with aHUS have had or are awaiting to undergo genetic testing [13]. In our survey the percentage was clearly lower, in addition, a family tree was completed for only 41% of paediatric patients and 16% of adult patients. Our survey does not allow understanding of the reasons for this relatively low proportion for genetic testing. The lack of guidelines, the cost, the time taken to get results, as well as difficulty in interpreting the results, may influence the extent to which a family history is investigated. The introduction of uniform reporting, tailored to non-geneticists may increase confidence in physicians requesting genetic tests. This may in turn result in a more consistent application of genetics to inform on prognosis and the risk of recurrence, and provide information to expand our understanding on the natural course of aHUS.

There are limitations to interpreting data collected by surveys. Here, the survey was sent to email addresses available to the authors and may have led to some survey bias. Additionally, not all countries were equally represented, however despite this the survey covered a large and unprecedented amount of countries. Secondly, we cannot be sure of the precise number of recipients, as clinicians may have forwarded the survey to colleagues. However, we estimate a satisfactory response rate compared with the average for web-based surveys [22]. Based on a meta-analysis of 68 web-based surveys in 49 studies [23], we estimate 641 clinicians would have been invited for our response rate to be similar to the 39.6% reported [23]. Thirdly, the majority of responses were provided by experienced nephrologists and not physicians from other specialties. As such, our results may not be applicable to diagnostic practices of non-nephrology departments, where awareness of this ultra-rare disease is likely to be lower. Finally, the survey did not undergo a formal validation process; however, three drafts were subsequently developed and reviewed by the authors prior to circulating the final survey to a large audience.

We hypothesised that there were areas of uncertainty in the work-up of patients with TMA, leading to potential delays in establishing the final diagnosis, putting patient’s lives at a higher risk. This survey addressed many elements of the hypothesis however other aspects warrant further investigation. The current survey focused on the diagnosis, not on the management, of TMA. Real-world information on the management practices of clinicians treating patients with TMA would be of interest and a worthwhile topic for a follow-up survey in the future.

This survey evaluated current clinical practices of experienced paediatric and adult nephrologists in the diagnosis of TMA including aHUS. We describe the variable practices in the differential diagnosis of TMA amongst respondents, despite their considerable clinical experience. Our survey suggests that some countries lack national guidelines and these should be put in place where possible. Guidelines would serve to raise awareness of the availability and relevance of ADAMTS13 and STEC testing to discriminate TTP and STEC-HUS from aHUS. These clinical guidelines could potentially increase awareness of the evidence for genetic testing after the initial therapy has been started and the importance of conducting a full family history, when presented with a genetic disease as a possible diagnosis. Our results also suggest that having guidelines does not guarantee their implementation; future activities should focus more on the dissemination of knowledge and education, to encourage guideline implementation. Continuing medical education courses on diagnostic and treatment practices for TMA may be of benefit to clinicians. Finally, there is the need for clear and practical guidelines for general practitioners and primary care doctors, in order to accelerate the diagnostic process. Systematically developed guidelines will aid clinicians in diagnosing and managing TMA, enabling a more rapid and informed diagnosis and initiation of effective treatment earlier, with a consequent improvement in patient outcomes.