A key step towards setting a benchmark for tackling transplant-associated thrombotic microangiopathy

Commentary

Improvements in allogeneic haematopoietic cell transplantation (allo-HCT) techniques and supportive care strategies over the last several decades have resulted in a marked reduction in non-relapse mortality (NRM) due to fewer serious complications. (Gooley et al., 2010; D’Souza et al., 2017) Despite these advances, several complex post-HCT toxicity syndromes that share areas of overlapping pathobiology involving endothelial injury and often underlying allogeneic-mediated reactivity continue to contribute considerable morbidity and risk of mortality. (Cooke et al., 2008; Carreras & Diaz-Ricart, 2011; Panoskaltsis-Mortari et al., 2011; Dalle & Giralt, 2015; Cornell et al., 2015; Rodrigues-Diez et al., 2016; Moiseev et al., 2019; Pagliuca et al., 2019) In the modern era, perhaps the most illustrative of these syndromes is transplant-associated thrombotic microangiopathy (TA-TMA), a potentially devastating complication resulting from multifactorial endothelial damage and subsequent complement activation leading to microangiopathic haemolysis, microvascular thrombosis, and eventually end-organ damage. (Laskin et al., 2011; Jodele et al., 2014; Postalcioglu et al., 2018; Dvorak et al., 2019) Recent evidence suggests that the mechanistic link between endothelial damage and subsequent complement activation may be the formation of neutrophil extracellular traps that activate the classical and alternative complement pathways further propagating of microvascular injury. (Gloude et al., 2017).

Several obstacles have impeded the clinical study of TA-TMA. Most notably, numerous diagnostic classification systems have been developed over the years without established consensus criteria (Ho et al., 2005; Ruutu et al., 2007; Cho et al., 2010). Also, the majority of previously published data are from small single-institution studies involving heterogeneous patient populations with variable haematological conditions and HCT characteristics.(Laskin et al., 2011; Jodele et al., 2014) This has resulted in imprecise cumulative incidence estimates and an uncertain understanding of the patient and HCT characteristics associated with an increased risk of TA-TMA. Consequently, the timely and consistent diagnosis of TA-TMA and the development of highly effective treatments remain areas of truly unmet need in the field of allo-HCT. This is especially relevant given the advent of novel complement pathway inhibitors and agents aimed at restoring endothelial integrity (Sicre De Fontbrune et al., 2014; De Fontbrune et al., 2015; Yeates et al., 2017; Khaled et al., 2017). Rationally planning future studies testing novel agents requires accurate identification and understanding of several key elements of TA-TMA: its true incidence, risk factors associated with its development, and its impact on post-HCT outcomes in the modern era.

In this issue Epperla and colleagues (Epperla et al., 2020) have brought us closer to understanding each of these elements with a study utilising comprehensive registry data from the Center for International Blood and Marrow Transplant Research (CIBMTR). Among the 23 665 paediatric and adult and allo-HCT recipients, they found 661 cases of TA-TMA resulting in a 3-year cumulative incidence estimate of 3%, which is considerably lower than what has been reported in recent studies (Jodele et al., 2014; Dvorak et al., 2019). Several baseline patient and HCT characteristics were found to be associated an increased risk of TA-TMA in multivariable analyses, including: African-American race, diagnosis of acute lymphoblastic leukaemia or aplastic anaemia, baseline poorer renal function, combination of calcineurin inhibitor (CNI) with sirolimus as graft-versus-host disease (GVHD) prophylaxis, history of prior autologous HCT, and having a human leukocyte antigen (HLA)-matched unrelated or mismatched donor. In the post-HCT period, the degree of acute GVHD severity correlated with a risk of TA-TMA. Patients who were male, who received in vivo T cell depletion with anti-thymocyte globulin and/or alemtuzumab, and who had received reduced-intensity or non-myeloablative conditioning regimens had a lower risk of TA-TMA. These data generally verify what has been previously shown across various smaller retrospective and prospective studies with the exception of the intriguing finding of African-American race as a risk factor for TA-TMA, for which the authors did not propose a potential hypothesis (Dvorak et al., 2019). Interestingly, a recent analysis of complement gene mutations demonstrated that non-Whites more frequently had these variants, thereby leading to a higher risk of developing TA-TMA and having more severe clinical presentations (Jodele et al., 2016). These potential racial disparities in TA-TMA risk and severity require further inquiry. Remarkably, the Epperla et al. study further confirms previous evidence by Luft and others relating to severe GVHD and TA-TMA (Luft et al., 2011; Dietrich et al., 2013; Luft et al., 2017; Zeisbrich et al., 2017; Rachakonda et al., 2018). The convergent pathobiology of steroid refractory GVHD and TA-TMA, linked by underlying endothelial activation and injury, will be critically important to consider in the design of clinical trials aimed at treating TA-TMA through complement blockade, as patients with concomitant acute GVHD may require additional non-complement directed therapies (Gloude et al., 2017; Wall et al., 2018). Most importantly, the authors found that patients with TA-TMA after allo-HCT had a sevenfold higher risk of requiring renal replacement therapy and threefold increase in all-cause mortality compared to patients without TA-TMA, resulting in strikingly inferior overall survival driven by NRM.

In the absence of large, multicentre, prospective clinical trials, the Epperla et al. study invokes a critical question: should these data represent the new benchmark for our understanding of the true incidence of TA-TMA? This query is nuanced and speaks to the study’s major limitation: the presence of TA-TMA was recorded by the treating clinicians without specific TA-TMA diagnostic criteria and reported to the CIBMTR as a binary outcome. This may considerably underestimate the true incidence of TA-TMA. Without standard consensus criteria, it is highly plausible that local clinicians recorded only the most severe TA-TMA cases. The authors addressed the issue of potential underreporting in several ways, suggesting that underreporting was likely proportionate to the disease severity; that is, cases of mild TA-TMA were less likely to be captured. However, TA-TMA likely has a spectrum of severity, and it is important to recognise that subclinical, laboratory-detected manifestations of TA-TMA may result in clinically relevant end-organ damage, such as acute kidney injury, that are known contributors to NRM (Changsirikulchai et al., 2009; Hingorani, 2016; Li et al., 2019). Many experts in the field have argued that with the use of rigorous, prospective monitoring using widely available laboratory tests and novel blood biomarkers, TA-TMA can be detected in significantly more patients than previously appreciated (Jodele et al., 2014). Advances in the biological understanding of TA-TMA and discovery of more specific and predictive biomarkers have led to more awareness and the potential for observer bias in recent years (Khosla et al, 2018). Despite the potential for observer bias, Epperla and colleagues did not find a change in TA-TMA incidence over time during the study period (2008–2016), which again may reflect underreporting of TA-TMA due to non-standardised diagnostic criteria.

Within the inherent confines of a retrospective analysis, Epperla and colleagues went to great lengths to address the limitations of a registry-based analysis. Ultimately, this is largest TA-TMA study ever published and warrants considerable praise for its methodology and use of the rich data resources provided by the CIBMTR. Nevertheless, our field continues to strive towards defining the most accurate benchmark for future comparative studies and clinical trials in patients with TA-TMA. The critical first step is to define and validate simple, universally translatable diagnosis and severity consensus criteria through a collaborative, multidisciplinary, and multicentre expert working group. Determining the truest incidence of TA-TMA, its risk factors, and effects on outcomes will ultimately result only from a large, multicentre, prospective clinical trial with rigorous clinical and laboratory monitoring. With this, we can more soundly develop well-designed and impactful prospective studies testing the effectiveness of promising novel therapies for TA-TMA. Moreover, given the exceptionally expensive prices of these agents, prospective studies should include aims to evaluate healthcare utilisation costs in patients treated with or without these agents. In the meantime, as the HCT community strives to realise these goals, Epperla et al. have provided our field with a key step on the journey towards tackling TA-TMA.

Disclosures

Michael Scordo has served as a paid consultant for McKinsey & Company, Angiocrine Bioscience, Inc., and Omeros Corporation. He has received research funds from Angiocrine Bioscience, Inc. He has served on an ad hoc advisory board for Kite – A Gilead Company. Sergio A. Giralt has served as a paid consultant for Amgen, Actinium, Celgene, Johnson & Johnson, Takeda, Jazz Pharmaceuticals, Novartis, Kite – A Gilead Company, and Spectrum Pharmaceuticals. He has received research funding from Amgen, Actinium, Celgene, Johnson & Johnson, Takeda, and Miltenyi.