The prevalence and clinical outcomes of microangiopathic hemolytic anemia in patients with biopsy‐proven renal thrombotic microangiopathy

To the Editor: Thrombotic microangiopathy (TMA) refers to the characteristic pathologic findings of endothelial damage that include microvascular thrombosis and result from a diverse set of etiologies.1 The classic hemolytic uremic syndrome (HUS) presents with microangiopathic hemolytic anemia (MAHA) and reduced renal function, and a diagnostic tissue biopsy is often deemed unnecessary if a clinical diagnosis can be established firmly.1 On the other hand, there are reports of “renal limited TMA,” where systemic MAHA was either absent or indeterminate, making the diagnosis of TMA difficult without a kidney biopsy.2 Recent epidemiologic studies have only looked at cohorts of clinically diagnosed TMA (i.e., HUS)3 and thus our understanding of the epidemiology of renal microangiopathy, beyond the classic HUS, remains incomplete. Furthermore, the diagnostic work-up for TMA remains challenging and has yet to be standardized with the limited available society guidelines differing in the criteria for both the demonstration of MAHA and renal involvement.4 We undertook a retrospective observational study of TMA diagnosed on native kidney biopsy to: (a) evaluate the role of kidney biopsy in TMA; (b) better understand the epidemiology of biopsy-proven TMA; and (c) assess the significance of clinicopathologic findings in predicting prognosis in renal TMA. The Mayo Clinic renal pathology database (between January 1, 2000 and December 31, 2013), was queried for the terms “thrombotic” OR “microangiopathy.” In the group of patients identified from the query, the renal biopsy report was reviewed. TMA was defined by the presence of one or more of the following: glomerular or vascular thrombosis, mesangiolysis, arterial mucoid edema, glomerular basement membrane duplication by subendothelial fluff and/or vascular onion skinning (Figure S1). All patients who met the TMA definition, were diagnosed on native kidney, and received clinical care at Mayo Clinic, Rochester, were included in the study (N = 128). A detailed re-evaluation of TMA changes on kidney tissue of the study cases was performed by two renal pathologists (S.M.S. and S.H.N.) who were blinded to clinical information. A review of the electronic medical record was conducted by a nephrologist (G.B.) who was blinded to pathology findings. Definitions of clinical and histologic renal findings, MAHA, and cause of TMA are provided in Table S1. The following outcomes were identified at various time points from clinical presentation (6 months, 1 year, entire follow-up): (a) death, (b) doubling of SCr, (c) dialysis initiation. Primary outcome was defined as dialysis initiation and/or death by 1-year of follow-up. Statistical analysis methodology is detailed in Supporting Information Materials Section I. The renal presentation of TMA in our cohort varied but most commonly, increased serum creatinine (SCr) and proteinuria (>0.5 g/day) were seen (78% and 87%, respectively). The degree of proteinuria ranged widely from 59 to 30 280 mg/day (median, 2060 mg/day). Nephrotic syndrome was seen in 14%, hematuria in 64% and normal urine microscopy in 17%. The presence of MAHA markers also varied. Elevated lactate dehydrogenase was most common (75%) followed by thrombocytopenia (55%), peripheral schistocytes (43%) and low serum haptoglobin (39%). Thirty-one percent had no known systemic indicators of MAHA. In the kidney, renal thrombi and/or fragmented erythrocytes were found in 58%. TMA involved glomeruli alone in 38% (“glomerular [only]” TMA; gTMA) and involvement of arterioles or larger renal vessels by TMA was present in 62% (“[any] vascular” TMA; vTMA). The vTMA group includes patients with TMA involving vessels alone in 11% (“vascular alone” TMA), as well as those with combined glomerular and vascular involvement in 51% (“combined” TMA). A concurrent renal disease (in addition to the TMA and its sequelae) was present in 23% of patients. Detailed description of clinical and pathologic characteristics is shown in Tables S2–S4. Peripheral schistocytosis strongly associated with female gender, other markers of MAHA, histologic findings of acute TMA (e.g., renal thrombi, renal fragmented erythrocytes) and vascular location of TMA (p < .05 for all; Figure 1A,B; Tables S2 and S4). vTMA (in comparison to gTMA) also associated with female gender, and an overall pattern of a more severe clinical phenotype, including higher SCr, systemic MAHA markers and severe hypertension (p < .05 for all; Figure 1A; Tables S2 and S4). The pattern of histologic findings of acute and chronic TMA in vTMA differed from gTMA based on their respective definitions (Tables S1 and S4). In addition, renal thrombi including glomerular thrombi as well as most findings of chronic kidney disease (CKD) and acute tubular damage were more common in vTMA (Table S4). Secondary TMAs predominated in this biopsy-based study. The underlying cause of renal TMA was autoimmune disease (31%) followed by hematologic clonal disorders in 17%, drugs in 12%, hypertension alone in 9% and complement/other primary TMA in 8% (Table S6). In 14% (18/128) of the cohort, there >1 potential reason that may have contributed to the development of TMA. The most common concern was abnormal work-up of the alternative complement pathway—either variant(s) of uncertain significance on genetic testing and/or abnormalities on functional testing (56% [10/18]). The clinical and pathologic correlations of various causes of renal TMA are detailed in Tables S7 and S8. Most causes showed a mixed presence of systemic MAHA. The most significant differences were observed between autoimmune/complement (other primary)/hypertension TMA (“AI-Compl-HTN” group) versus hematologic clonal disorders/miscellaneous cause of TMA (“Heme-Other” group). Peripheral schistocytes, vTMA and renal thrombi were all more frequent in the AI-Compl-HTN group versus Heme-Other group (p < .05 for all; Figure 1C; Tables S7 and S8). Median follow-up time of our cohort was 2.6 years (0.5–4.7). At 1 year, more than half the patients (52%; Figure 1E) had started dialysis (49%) and/or had died (3%). Detailed breakdown of outcomes is provided in Figure 1D. Univariable analysis of clinical and histologic factors revealed several factors to be associated with starting dialysis (Table S9A,B). On further multivariable analyses of clinical factors, higher SCr (hazard ratio [HR] 1.24 [1.14–1.36] per unit change; p < .01) and peripheral schistocytes (HR 2.04 [1.02–4.1]; p = .04) were found to be independently associated with death and/or dialysis by 1-year (Table S10A). On multivariable analysis of histologic factors, any vascular TMA (vTMA; HR 2.2 [1.17–4.16]; p = .01 in comparison to gTMA) and tubular injury (HR 3.84 [1.7–8.7]; p < .01) both independently predicted death and/or dialysis by 1-year follow-up (Table S10B). Cumulative incidence curves (Figure 1E,F) also confirm the association of peripheral schistocytes and vTMA with worse clinical outcomes. The evaluation of biopsy-proven renal TMA makes our study design unique as this methodology allows a closer look at the subset of TMAs where a clear diagnosis of TMA could not be confirmed clinically. Thus, our study distinctively highlights the clinical spectrum of renal TMA beyond MAHA, and essentially, beyond the HUS. Our findings revealed that a substantial proportion of biopsy-proven renal TMA is “renal limited” (no peripheral schistocytes in 57%; Figure 1A). This clinical phenotype was seen across diverse causes. Most often, this presentation was observed in TMA secondary to clonal hematologic disorders, in which only 18% had peripheral schistocytes, but was also noted in all other causes of TMA, wherein peripheral schistocytosis ranged from 35% to 73% (Figure 1C; Table S7). In addition to the hematologic presentation, the histologic findings of acute TMA and the patient outcomes varied substantially between HUS and “renal limited TMA” (Figure 1B,F; Tables S4 and S10A). Furthermore, despite the lack of clear systemic MAHA in “renal limited TMA,” kidney biopsy findings enabled a diagnosis of TMA to be established. The absence of renal thrombi in a proportion of renal TMAs cases is well-known and reclassifying renal TMA to microangiopathy with or without thrombosis has been proposed.5 Our study revealed absence of renal thrombi in about half the patients with renal TMA; however, histologic location of the renal thrombi was more pertinent, as only vascular thrombi correlated with poor renal outcomes (Table S9B). Vascular histologic location of renal TMA, in fact, strongly correlated with systemic MAHA markers (Figure 1A; Table S2) and worse renal outcomes (Figure 1G; Table S10B). Our findings recommend that, rather than renal thrombi, peripheral schistocytes (MAHA) and histologic location of TMA be considered as important prognostic factors in any future reclassification of renal TMAs. Our study has several limitations, which include its retrospective nature, single center population with variable cause of TMA, and the presence of concurrent kidney disease in 23%. Since complement data was available only for a minority of cases (18% [23/128]), it is unclear how many cases attributed to secondary TMA also had underlying complement defects. It was not possible for us to analyze response to specific treatments due to the retrospective study methodology and heterogeneity of the cohort. In conclusion, our study highlights several important aspects of the complex syndrome of renal TMA and our findings strongly recommend that (1) renal TMA be considered as an important differential in AKI or CKD of unclear cause, even when clear evidence of systemic MAHA is lacking; (2) kidney biopsy-based confirmation be performed in this scenario, if safely possible; and (3) “renal limited TMA” and the histologic location of TMA be considered in future studies and reclassification of renal TMAs, as these are likely important clinicopathologic distinctions within the syndromes of TMA. Research idea and study design: Gauri Bhutani, Nelson Leung, and Samih H. Nasr. Data acquisition and data analysis: Gauri Bhutani, Nelson Leung, Samar M. Said, Anthony M. Valeri, Brad C. Astor, and Samih H. Nasr. Writing original draft: Gauri Bhutani. Writing review/editing: Samar M. Said, Anthony M. Valeri, Nelson Leung, Brad C. Astor, Mary E. Fidler, Mariam P. Alexander, Lynn D. Cornell, and Samih H. Nasr. Supervision/mentorship: Samih H. Nasr and Nelson Leung. Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual's own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work, even one in which the author was not directly involved, are appropriately investigated, and resolved, including with documentation in the literature if appropriate. The authors declare no conflict of interest. The data that supports the findings of this study are available in the supplementary material of this article. If additional supporting data are needed, please contact the corresponding author. Appendix S1 Supporting Information Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.