View and Update

Article

Clinicopathologic Characteristics and Outcomes of Renal Thrombotic Microangiopathy in Anti-Neutrophil Cytoplasmic Autoantibody-Associated Glomerulonephritis

Su-Fang Chen,*†‡ Huan Wang,*†‡ Yi-Min Huang,*†‡ Zhi-Ying Li,*†‡ Su-Xia Wang,*†‡ Feng Yu,*†‡ Ming-Hui Zhao,*†‡§ and Min Chen*†‡ *Renal Division, Department of Medicine, Peking Abstract University, First Background and objectives Thrombotic microangiopathy (TMA) in ANCA-associated vasculitis (AAV) has been Hospital, Beijing, † mainly reported in isolated case reports. The aim of this study was to analyze clinical and pathologic China; Key Laboratory of Renal characteristics and prognosis of patients with renal TMA in ANCA-associated GN in a large cohort of Chinese Disease, Ministry of patients. Health of China, Beijing, China; ‡Key Design, setting, participants, & measurements Clinical and renal histopathologic data of 220 patients with Laboratory of Chronic biopsy-proven ANCA-associated GN from 1996 to 2013 were retrospectively analyzed. Patients were followed Kidney Disease – Prevention and up for a median period of 32 (interquartile range [IQR], 12 65) months, and outcomes of patients were ana- Treatment, Peking lyzed. University, Ministry of Education, Beijing, § Results Among the 220 patients with ANCA-associated GN, 30 were identified having concomitant renal TMA China; and Peking- TsinghuaCenterfor by pathologic evaluation. Compared with the non-TMA group, patients with renal TMA presented with Life Sciences, Beijing, more severe renal injury, as evidenced clinically by a higher level of serum creatinine at diagnosis (5.0 [IQR, China 3.5–9.0] versus 3.2 [IQR, 1.7–6.8] mg/dl; P=0.02) and pathologically by a higher percentage of cellular crescents (15.0% [IQR, 6.9%–34.9%] versus 6.9% [IQR, 0%–21.1%]; P=0.04) and more severe interstitial in- Correspondence: filtration (2 [IQR, 2–2] versus 2 [IQR, 1–2]; P=0.03) in renal biopsies. Furthermore, multivariate analysis Dr. Min Chen, Renal showed that renal TMA was independently associated with mortality of patients with AAV after adjusting for Division, Department of Medicine, Peking age, sex, initial serum creatinine, tubular atrophy, and interstitial fibrosis (hazard ratio, 1.92; 95% confidence P fi University, First interval, 1.08 to 3.41; =0.03) or for age, sex, the histopathologic classi cation scheme proposed by Berden Hospital; Peking et al. (JAmSocNephrol21: 1628–1636, 2010), tubular atrophy, and interstitial fibrosis (hazard ratio, 1.95; 95% University Institute of confidence interval, 1.07 to 3.55; P=0.03). Nephrology; Key Laboratory of Renal Disease, Ministry of Conclusions Renal TMA in ANCA-associated GN is not rare and presents with more severe renal injury. Renal Health of China; Key TMA is independently associated with all-cause mortality in patients with AAV. Laboratory of Chronic Clin J Am Soc Nephrol 10: 750–758, 2015. doi: 10.2215/CJN.07910814 Kidney Disease Prevention and Treatment (Peking University), Ministry Introduction findings in the prognostication of patients at the of Education, Beijing, China.No.8,Xishiku ANCA-associated vasculitis (AAV) is a group of sys- time of diagnosis. Street, Xicheng temic autoimmune diseases characterized by pauci- Thrombotic microangiopathy (TMA) comprises a District, Beijing immune necrotizing small-vessel vasculitis and circulating group of clinical and pathologic syndromes that share 100034, China. autoantibodies against neutrophil cytoplasmic constit- a similar pathologic process, characterized by endothe- E-mail: chenmin74@ sina.com uents, especially proteinase 3 and myeloperoxidase. lial and blood cell damage and thrombotic microvas- AAV comprises granulomatosis with polyangiitis (GPA), cular occlusions. TMA comprises a spectrum of distinct microscopic polyangiitis (MPA), eosinophilic granulo- disorders, including typical and atypical hemolytic matosis with polyangiitis (EGPA), and renal-limited uremic syndrome (HUS), congenital and acquired throm- vasculitis (RLV) (1). The kidney is one of the most vul- botic thrombocytopenic purpura, malignant hyper- nerable organs, often presenting with rapidly progres- tension, pregnancy, organ transplantation, drugs, or sive GN. The diagnostic and prognostic value of the systemic autoimmune diseases. Renal involvement is renal biopsy in ANCA-associated GN is widely recog- common in TMA. nized. Moreover, the histopathologic classification To our knowledge, TMA in ANCA-associated GN of ANCA-associated GN proposed by Berden et al. has only been reported in isolated case reports (3–9), (2) has greatly prioritized the role of renal biopsy with two patients showing pathologic features of

renal TMA. The clinicopathologic characteristics, espe- may also be seen in the afferent glomeruli, small arterioles, cially the prognostic value of pathologic findings of renal and/or arteries. Chronic changes were mucoid changes and TMA, in ANCA-associated GN are far from clear. The aim onion skin lesions of arterioles and/or arteries (11). Known of this study was to analyze clinical and pathologic char- causes of renal TMA, including SLE, anti-phospholipid syn- acteristics and the prognosis of patients with renal TMA in drome, scleroderma, pregnancy-associated TMA, malignant ANCA-associated GN in a large cohort of Chinese patients. hypertension, transplantation-associated TMA, dissemi- nated intravascular coagulation, drug-mediated TMA, and TMA associated with various infections, including HIV, hep- Materials and Methods atitis B virus, and hepatitis C virus, were further excluded. Patients Informed consent for renal biopsy was obtained from each A total of 220 consecutive patients with ANCA-associated patient. The research was in compliance with the Declara- GN who received renal biopsy, diagnosed in the De- tion of Helsinki and was approved by the local ethical compartment of Nephrology, Peking University First Hos- mittees of Peking University First Hospital. pital from 1996 to 2013, were analyzed retrospectively. Follow-up was performed in outpatient clinics specific Renal biopsy was performed at the time of diagnosis and for AAV. The primary end point was defined as death, and before the initiation of immunosuppressive therapy. All the secondary end point was defined as ESRD. The com- patients met the criteria of the 2012 Chapel Hill Consensus bined end point was defined as a composite outcome of fi Conference de nition for AAV (1). Patients with comorbid death or ESRD. renal disease or secondary vasculitis, such as membra- nous glomerulonephropathy, anti-glomerular basement membrane disease, drug-induced vasculitis, or lupus ne- Detection of ANCA ANCA tests were performed by both indirect immuno- phritis, were excluded. Patients with EGPA were also ex- fluorescence assay and antigen-specific ELISA for all patients cluded because EGPA is increasingly considered a distinct at the time of presentation and before immunosuppressive type of AAV with different manifestations and outcomes treatment was instituted, according to the manufacturer’s compared with GPA, MPA, and RLV (10). The details of instruction (Euroimmun, Lübeck, Germany). the recruitment process are shown in Figure 1. Renal TMA was defined as interlobular artery and arteriole and glomerular capillary lesions, including endo- Detection of Disintegrin and Metalloproteinase with a thelial cell swelling, lumen narrowing, or obliteration and Thrombospondin Type 1 Motif, Member 13 Activity thrombi formation by light microscopy. Swelling of glo- The Disintegrin and Metalloproteinase with a Thrombo- merular endothelial cells, detachment from the glomerular spondin Type 1 Motif, Member 13 (ADAMTS-13) activity basement membrane, and widening of the subendothelial assay was assessed using a residual collagen-binding assay, fi spacewereidentified by electron microscopy (11). The le- slightly modi ed as previously described (12) (Supplemental sions were further divided into acute and chronic changes. Material). Data are reported as the percentage of collagen- The acute lesion was characterized by swelling of the en- binding activity remaining after dialysis compared with the ’ dothelial cells and subendothelial space; fibrin thrombi collagen binding activity in the individual s baseline sample. One hundred percent minus the residual collagen-binding activity was arbitrarily regarded as the ADAMTS-13 activity.

Renal Histopathology Biopsies were separately scored by two pathologists blinded to the clinical data, according to the previously standardized protocol for scoring renal biopsies of patients with AAV (13–15). In short, each glomerulus was scored separately on the presence of crescents (cellular/fibrous), glomerulosclerosis (local/segmental/global), fibrinoid necrosis, and a number of other lesions. Interstitial and tubular lesions were scored semiquantitatively on the basis of the percentage of the tubulointerstitial compartment that was affected: tubular atrophy(– for 0%, + for 1%– 50%, and ++ for.50%), interstitial infiltrates (– for 0%, + for 1%–20%, ++ for 21% –50%, and +++ for .50%), and interstitial fibrosis (– for 0%, + for 1%–50%, and ++ for .50%). Each biopsy was further classified as sclerotic, focal, crescentic, or mixed category, according to the histopathologic classification system of ANCA-associated GN proposed by Berden et al. (2).

Treatment and Response Figure 1. | Flowchart for inclusion/exclusion process. EGPA, eo- The treatment protocols have been described previously sinophilic granulomatosis with polyangiitis; TMA, thrombotic mi- (16,17). For a detailed description, see the Supplemental croangiopathy. Material. Brieﬂy, the induction therapy included corticosteroids 752 Clinical Journal of the American Society of Nephrology

in combination with cyclophosphamide. Patients with ARF Statistical Analyses or pulmonary hemorrhage received three pulses of intrave- The t test, nonparametric test, and chi-squared test were nous methylprednisolone before the standard induction performed as appropriate. Kaplan–Meier curves were therapy. Patients with severe pulmonary hemorrhage ad- used to analyze the outcomes of patients. All of the clini- ditionally received plasma exchanges. For maintenance copathologic parameters and treatment regimens listed in therapy, intravenous cyclophosphamide every 3 months Tables 1–3 were assessed as candidate predictors in the or daily oral azathioprine was given, with a duration of univariate survival analysis. If the P value was ,0.05, at least 2 years. The response to the immunosuppressive this predictor was allowed to be included in multivariable treatment was defined as previously described (18) (de- Cox regression models. Tubular atrophy and interstitial tailed in the Supplemental Material). The renal response to fibrosis were also included in the multivariable models treatment, evaluated at 6 months after initiation of immu- because they were potential histopathologic predictors ac- nosuppressive therapy, was judged according to the fol- cording to previous study (22). Because of the close corre- lowing criteria: (1) complete recovery of renal function lation between initial serum creatinine and the sequence of was indicated by normalization of renal function and res- histopathologic categories proposed by Berden et al. olution of hematuria; (2) partial recovery of renal function (r=0.53, P,0.001), the two parameters were separately in- was indicated by stabilization or improvement of renal cluded in the multivariate analysis using models A and B, function, with serum creatinine $1.5 mg/dl but dialysis respectively. Interaction effect of age and renal TMA on independent; and (3) treatment failure was indicated by all-cause mortality was also investigated. Differences were progressive decline in kidney function with persistence of considered significant if the P value was ,0.05. Analysis active urinary sediment despite immunosuppressive ther- was performed with SPSS version 11.0 statistical software apy (19–21). package (SPSS, Chicago, IL).

Table 1. Comparison of clinicopathologic parameters between ANCA-associated GN patients with and without renal TMA

ANCA-Associated GN ANCA-Associated GN Parameter P Value with Renal TMA (n=30) without Renal TMA (n=190)

Age (y) 63.9611.3 57.0614.6 0.004 Sex (male/female) 15/15 97/93 0.92 MPO-ANCA/PR3-ANCA 28/2 170/20 0.58 Fever 21 (70.0) 101 (53.2) 0.09 Fatigue 20 (66.7) 126 (66.3) 0.97 Weight loss 18 (60.0) 106 (55.8) 0.67 Muscle pain 9 (30.0) 57 (30.0) .0.99 Arthralgia 8 (26.7) 71 (37.4) 0.26 Skin rash 4 (13.3) 31 (16.3) 0.88 Pulmonary involvement 17 (56.7) 112 (59.3) 0.79 Ophthalmic involvement 9 (30.0) 40 (21.1) 0.27 ENT involvement 13 (43.3) 91 (47.9) 0.64 Nervous system 6 (20.0) 33 (17.4) 0.73 Hematuria 30 (100) 182 (95.8) 0.54 Urinary protein (g/24 h)a 1.24 (0.73–2.59) 1.73 (0.90–3.05) 0.24 Initial Scr (mg/dl)a 5.0 (3.5–9.0) 3.2 (1.7–6.8) 0.02 eGFR (ml/min per 1.73 m2)a,b 11.4 (5.2–14.4) 17.3 (7.2–39.4) 0.01 Dialysis-dependent at presentation 15 (50.0) 58 (30.7) 0.04 Hemoglobin (g/dl) 8.361.8 9.262.1 0.03 Platelet (3103/mm3)296.46135.7 287.56136.4 0.74 CRP (mg/L)a 23.6 (14.5–96.1) 18.0 (4.9–67.6) 0.28 ESR (mm/1 h) 104.2639.9 76.7639.5 ,0.001 BVAS 21.564.6 20.665.8 0.38 Normal glomeruli (%) 26.1623.1 34.9631.2 0.07 Crescent (%) 54.8624.2 56.8633.5 0.69 Cellular crescentsa 15.0 (6.9–34.9) 6.9 (0–21.1) 0.04 Fibrous crescentsa 7.3 (0–18.3) 3.6 (0–18.2) 0.65 Glomerular sclerosisa (%) 0 (0–19.0) 0 (0–5.0) 0.20 Fibrinoid necrosisa (%) 3.4 (0–9.6) 0 (0–4.6) 0.02

MPO, myeloperoxidase; PR3, proteinase 3; Scr, serum creatinine; ENT, ear, nose, and throat; CRP, C-reaction protein; ESR, erythrocyte sedimentation rate; BVAS, Birmingham Vasculitis Activity Scores; TMA, thrombotic microangiopathy. aValues are median (interquartile range). beGFR (ml/min per 1.73 m2)=1753(plasma creatinine)21.2343age20.17930.79 (if female) (41). Clin J Am Soc Nephrol 10: 750–758, May, 2015 Renal TMA in AAV, Chen et al. 753

Table 2. Comparison of treatment data between ANCA-associated GN patients with and without renal TMA

ANCA-Associated GN ANCA-Associated GN Treatment P Value with Renal TMA (n=30) without Renal TMA (n=190)

Induction therapy 30 190 PE 3 (10.0) 14 (7.6) 0.93 MP pulse 25 (83.3) 123 (66.5) 0.07 CTX 27 (90.0) 177 (93.2) 0.81 iv 24 (88.9) 145 (81.9) 0.54 po 3 (11.1) 32 (18.3) Maintenance therapy 16 121 CTX 6 (37.5) 48 (39.7) 0.87 AZA 10 (62.5) 73 (60.3) Treatment response Remission 27 (90.0) 178 (93.7) 0.72 Treatment failure 3 (10.0) 12 (6.3) Renal response Complete recovery 4 (13.3) 54 (28.4) 0.05 Partial recovery 16 (53.3) 102 (53.7) Treatment failure 10 (33.3) 34 (17.9)

Data are given as n (%). TMA, thrombotic microangiopathy; PE, plasma exchange; MP, methylprednisolone; CTX, cyclophosphamide; iv, intravenous; po, per os; AZA, azathioprine.

Results mg/dl; P=0.02) and a significantly higher proportion of General Data of Patients patients was dialysis dependent at diagnosis (50.0% versus Among the 220 patients with ANCA-associated GN 30.7%, P=0.04). Additionally, patients with TMA present- enrolled in this study, 30 (13.6%) were identified having ed with significantly lower levels of hemoglobin and in- concomitant renal TMA by pathologic evaluation (Figure creased levels of erythrocyte sedimentation rate than those 2). Renal TMA in ANCA-associated GN was equally dis- without TMA (8.361.8 versus 9.262.1 g/dl, P=0.03; tributed in time in our cohort during the study period, 104.2639.9 versus 76.7639.5mm/1h,P,0.001, respec- with a prevalence of five of 43 (11.6%), 13 of 86 (15.1%), tively) (Table 1). and 12 of 91 (13.2%) in the period between 1996 and 2001, 2002 and 2007, and 2008 and 2013, respectively. Renal Histopathology Among the 30 AAV patients with renal TMA, 15 (50.0%) An average of 24.7611.7 glomeruli were available for were men and 15 (50.0%) were women, with an average evaluation in the 220 renal biopsies. Parameters of renal 6 – age of 63.9 11.3 (range, 20 81) years at diagnosis. Ac- histopathology of patients with and without renal TMA cording to the 2012 Chapel Hill Consensus Conference are listed in Figure 3 and Table 1. fi de nitions (1), 24 of 30 (80.0%), four of 30 (13.3%), and Compared with the 190 patients without renal TMA in fi two of 30 (6.7%) were classi ed as MPA, GPA, and RLV, renal histopathology, patients with renal TMA had a sig- respectively. None of the 30 patients with TMA had con- nificantly higher percentage of cellular crescent formation comitant infections with HIV, hepatitis B virus, or hepa- (15.0% [IQR, 6.9%–34.9%] versus 6.9% [IQR, 0%–21.1%]; titis C virus. According to the pathologic features P=0.04) and a significantly higher percentage of fibrinoid previously described, seven patients presented with solely necrosis (3.4% [IQR, 0%–9.6%] versus 0% [IQR, 0%–4.6%]; acute lesions, seven patients presented with solely chronic P=0.02) (Table 1). There was a significant difference in the lesions, and 16 patients presented with both acute and classification scheme proposed by Berden et al. between chronic lesions. patients with and without renal TMA (P=0.03) (Figure 3), with the proportion of focal category being much lower in Serum ADAMTS-13 Activity the TMA group. Interstitial infiltrates were significantly None of the patients with renal TMA in our study more severe in patients with renal TMA than in those presented with a deficiency of serum ADAMTS-13 activity, without (2 [IQR, 2–2] versus 2 [IQR, 1–2]; P=0.03), whereas with a median activity of 98.6% (range, 97.1%–99.6%). the severity of interstitial fibrosis and tubular atrophy were comparable (1 [IQR, 0–1] versus 1 [IQR, 0–1]; P=0.42; 1 [IQR, Clinical and Laboratory Parameters 1–2] versus 1 [IQR, 1–1]; P=0.16, respectively) (Figure 3). The clinical and laboratory features of patients in the two groups are listed in Table 1. Patients with renal TMA were Treatment and Outcome significantly older at diagnosis than those without renal Therapy of patients with renal TMA and nonrenal TMA TMA (63.9611.3 versus 57.0614.6 years, P=0.004). Com- was comparable, with the exception of patients receiving pared with patients without TMA, patients with TMA had methylprednisolone pulse therapy, the proportion of significantly higher levels of initial serum creatinine (5.0 which was marginally higher in the renal TMA group (interquartile range (IQR), 3.5–9.0] versus 3.2 [IQR, 1.7–6.8] (P=0.07) (Table 2). After the aforementioned induction 754 Clinical Journal of the American Society of Nephrology

Table 3. Univariate analysis of patients’ survival in ANCA-associated GN

Predictor Hazard Ratio 95% Conﬁdence Interval P Value

Age(per10y) 1.60 1.29to1.98 ,0.001 Sex (male versus female) 1.97 1.21 to 3.18 0.01 MPA/GPA/RLV 0.15 MPA 2.10 0.76 to 5.82 0.15 GPA 1.38 0.46 to 4.10 0.57 RLV Reference group ANCA specificity in ELISA 2.20 0.79 to 6.09 0.09 (anti-MPO versus anti-PR3) BVAS 1.02 0.98 to 1.06 0.46 Initial serum creatinine (per mg/dl) 1.06 1.01 to 1.12 0.03 Total crescents (%) 0.95 0.49 to 1.85 0.87 Berden classification 0.02 Focal Reference group Cresentic 1.25 0.69 to 2.28 0.46 Mixed 1.74 0.94 to 3.20 0.08 Sclerotic 8.71 1.97 to 38.45 0.004 Renal TMA 2.41 1.39 to 4.16 0.004 Tubular atrophy 0.18 0% Reference group 1%–50% 0.89 0.35 to 2.25 0.80 .50% 1.43 0.54 to 3.77 0.47 Interstitial infiltrates 0.42 0% Reference group 1%–20% 1.58 0.46 to 5.40 0.47 21%–50% 1.78 0.55 to 5.78 0.34 .50% 2.62 0.73 to 9.46 0.14 Interstitial fibrosis 0.26 0% Reference group 1%–50% 1.01 0.60 to 1.72 0.96 .50% 1.77 0.83 to 3.78 0.14 Cumulative CTX dose (per gram) 0.19 0.01 to 3.85 0.28

MPA, microscopic polyangiitis; GPA, granulomatosis with polyangiitis; RLV, renal-limited vasculitis; BVAS, Birmingham Vasculitis Activity Scores; TMA, thrombotic microangiopathy.

therapy, 20 out of 30 (66.7%) patients with renal TMA compared the prognosis of patients with renal TMA achieved recovery of renal function, of which four patients among different categories (solely acute versus acute achieved complete recovery and 16 patients achieved partial and chronic versus solely chronic). The Kaplan–Meier sur- recovery; 10 of 30 (33.3%) patients had treatment failure. vival analysis revealed that patients with acute TMA le- The proportion of recovery of renal function (complete or sions only had a relatively favorable renal outcome, partial) was marginally higher in patients without renal whereas patients with solely chronic TMA lesions had TMA than in patients with renal TMA (82.1% versus 66.7%, the highest risk for developing ESRD (P=0.01) (Supple- P=0.05). mental Figure 1). Patients were followed-up on for a median period of 32 Univariate survival analysis of long-term prognosis in (IQR, 12–65) months. Six out of 30 (20.0%) patients with all patients with ANCA-associated GN showed that renal renal TMA and 57 out of 190 (30.0%) patients without TMA was associated with all-cause mortality (P=0.004). renal TMA experienced a disease relapse, respectively, Besides TMA, variables including age, sex, initial serum but there was no significant difference between the two creatinine, and the classification scheme proposed by groups (P=0.60). In the TMA group, 17 (56.7%) patients Berden et al. were predictors of death in univariate anal- died, and 10 (33.3%) patients reached ESRD. In the non- ysis (Table 3). Multivariate analysis revealed that renal TMA group, 55 (28.9%) patients died, and 56 (29.5%) pa- TMA was still an independent risk factor for patients’ tients reached ESRD. Main causes of death are listed in survival after adjusting for age, sex, initial serum creat- Supplemental Table 1. The long-term survival was signif- inine, interstitial fibrosis, and tubular atrophy (hazard icantly poorer in the TMA group (P=0.001) (Figure 4A), ratio, 1.92; 95% confidence interval, 1.08 to 3.41; whereas there was no significant difference in renal out- P=0.03) or for age, sex, the histopathologic classification come between the two groups (P=0.29) (Figure 4B). Addi- scheme proposed by Berden et al., tubular atrophy, and tionally, when comparing the combined end points (i.e., interstitial fibrosis (hazard ratio, 1.95; 95% confidence death, ESRD), patients with renal TMA also had signifi- interval, 1.07 to 3.55; P=0.03) (Table 4). Beceause patients cantly poorer outcomes (P=0.03) (Figure 4C). We further with renal TMA were significantly older than those Clin J Am Soc Nephrol 10: 750–758, May, 2015 Renal TMA in AAV, Chen et al. 755

of patients with renal biopsy-proven TMA in ANCA- associated GN in a large cohort. Our study found that renal TMA in ANCA-associated GN is not rare, with a prevalence of 13.6% (30 of 220 patients). In the literature, epidemiologic series reporting the prevalence of renal TMA in total renal biopsies were insufficient. Recently, a large cohort study in our center reported that adult renal TMA accounted for 1.4% (109 of 7589 patients) of the total biopsied patients in the same Figure 2. | Representative renal TMA lesions in ANCA-associated period (23). However, the prevalence in other countries is GN. (A) Thrombus in glomerular endocapillary (Masson’s trichrome unclear, especially for renal TMA in ANCA-associated 3200). (B) The thickened arteriole with mucoid intima edema and fibrinoid necrosis; arteriolar lumen was severely narrowed (Masson’s GN.Asreportedinpreviouscasereports,patientswith trichrome 3200). TMA, thrombotic microangiopathy. concomitant renal TMA in ANCA-associated GN presented with more severe renal injury (3–9). Patients with renal TMA in our cohort also had a higher level of initial without renal TMA, we also investigated the effect of age serum creatinine, higher percentage of cellular crescents and renal TMA on all-cause mortality in the multivariate and fibrinoid necrosis, and more severe interstitial infil- regression analysis, but no significant interaction was tration compared with patients without TMA, which, by a observed (P=0.64). large cohort study, further extends previous findings. Meanwhile, no significant differences were observed in chronic lesions, including fibrous crescents, glomerular Discussion sclerosis, interstitial fibrosis, and tubular atrophy between TMA comprises a group of disorders, including HUS, patients with and without renal TMA. Therefore, patients thrombocytopenic purpura, and disease-associated TMA. with renal TMA had more acute renal diseases. Further- Renal TMA in AAV has mainly been reported in isolated more, renal TMA was found to be independently associated case reports (3–9). Our study is, to our knowledge, the first with all-cause mortality of patients with AAV. It might sug- to assess clinical and pathologic features and the prognosis gest that patients with TMA in AAV should receive more

Figure 3. | Comparison of tubulointerstitial lesions between patients with and without renal TMA. Tubulointerstitial lesions were scored semiquantitatively on the basis of the percentage of the affected compartment: interstitial inﬁltrates (2 for 0%, + for 1%–20%, ++ for 21%–50%, and +++ for .50%), interstitial ﬁbrosis (2 for 0%, + for 1%–50%, and ++ for .50%), and tubular atrophy (2 for 0%, + for 1%–50%, and ++ for .50%). TMA, thrombotic microangiopathy. 756 Clinical Journal of the American Society of Nephrology

The pathogenesis of renal TMA in ANCA-associated GN and why only a small subset of patients with ANCA GN developed TMA is far from clear. However, we observed that greater than one in eight patients with AAV combined with TMA, which suggests that, rather than a casual phe- nomenon, these two entities may share a common patho- physiologic process. Theoretically, there are several potential explanations. First, considering the pathologic features of TMA, endothelial damage has long been regarded as an important disease mechanism in TMAs (24). On the other hand, ANCA-mediated activation of neutro- phils that results in endothelial injury is the basic patho- physiologic mechanism involved in AAV (25). It seems that TMA and AAV share the same target cells, namely, endothelial cells, which may contribute to the development of renal TMA in ANCA-associated GN. Second, the complement system might be another potential con- tributor. Although AAV has traditionally been characterized as pauci-immune, and decreased levels of circulating C3 are uncommon in AAV, recent studies have dem- onstrated that activation of the alternative complement pathway plays a critical role in the pathogenesis of AAV (26–30). Patients with active AAV have elevated levels of circulating C3a, C5a, soluble C5b-9, and Bb, which suggests activation of the complement system via the alternative pathway (30). As for TMA, a variety of hereditary and acquired disorders, which contribute to the loss of alternative pathway regulation on endothelial cells and on the surface of platelets, have been documented and predispose patients to TMA susceptibility (31–34). Indeed, it has been suggested that most TMAs are characterized by misdirected complement activation affecting endothelial cell and platelet integrity (34). Therefore, we speculate that systemic activation of complement, especially the alternative complement pathway, might play an important role in the development of renal TMA in AAV. Animal studies in AAV have suggested a potential therapeutic role of anti-C5 inter- vention (27,28,35), and clinical trials with CCX168, a small molecule antagonist of the human C5aR, are ongoing (36). Moreover, eculizumab, a human C5 inhibitor, has been found to be remarkably efficient for the treatment of atypical Figure 4. | Comparison of prognosis between ANCA patients with haemolytic uraemic syndrome (37). Whether anti-C5 ther- and without renal TMA (Kaplan–Meier analysis). (A) Comparison of apy is beneficial in the treatment of patients with TMA in patients’ survival. (B) Comparison of renal survival. (C) Comparison of composite outcomes. TMA, thrombotic microangiopathy. AAV could be of interest to study in the future. Third, neutrophil extracellular traps (NETs) have been shown to par- ticipate in the pathogenesis of AAV (38,39). Interestingly, a intensive immunosuppressive therapy, such as plasma ex- recent study suggested the formation of NETs as a second change, as described in previous case reports (3–8). hit that precipitates acute disease in patients at risk for TMA In this study, TMA was associated with mortality, rather (40). Therefore, NETs might be relevant to the occurrence of than ESRD. Considering that TMA is a pathologic process TMA in AAV. Finally, severe lesion of AAV might act as with multiple organs involved, we speculate that the another potential trigger for mediating a secondary TMA extrarenal disorder of TMA is a possible explanation for reaction. Collectively, we speculate that the occurrence of patients’ poor long-term survival. In our cohort, the TMA in AAV is a multifactorial process. The exact mecha- higher proportion of crescents and the more severe inter- nism is of great interest for further investigation. stitial inflammation observed in patients with renal TMA, Our study has several limitations. Because it is a ret- compared with those without TMA, might be indicative of rospective and observational study, it is difficult to sort out overall cellular activation. However, because this study whether the TMA is secondary to more severe disease or is a retrospective one, direct evidence of extrarenal micro- whether the TMA is a causal factor of the more severe vascular lesions was not available in some patients, espe- disease caused by a separate primary TMA process. In pre- cially during follow-up; the contribution of extrarenal vious case reports (4–8),itwassuggestedthatTMAin TMA to the mortality in patients with AAV needs to be AAV is more likely to be a disorder secondary to ANCA- confirmed by prospective studies. associated GN. However, this contention needs to be Clin J Am Soc Nephrol 10: 750–758, May, 2015 Renal TMA in AAV, Chen et al. 757

Table 4. Multivariate analysis of patients’ survival in ANCA-associated GN

Predictor Hazard Ratio 95% Confidence Interval P Value Model A Age(per10y) 1.68 1.34to2.12 ,0.001 Sex (male versus female) 2.08 1.26 to 3.43 0.004 Initial serum creatinine (per mg/dl) 1.05 0.99 to 1.11 0.11 Tubular atrophy 0.82 0% Reference group 1%–50% 0.76 0.29 to 2.03 0.59 .50% 0.87 0.29 to 2.62 0.80 Interstitial fibrosis 0.14 0% Reference group 1%–50% 1.27 0.72 to 2.22 0.41 .50% 2.61 1.01 to 6.71 ,0.05 Renal TMA 1.92 1.08 to 3.41 0.03 Model B Age(per10y) 1.72 1.36to2.17 ,0.001 Sex (male versus female) 2.12 1.29 to 3.48 0.003 Berden classification 0.02 Focal Reference group Cresentic 1.31 0.68 to 2.52 0.42 Mixed 1.62 0.78 to 3.39 0.20 Sclerotic 13.52 2.55 to 71.78 0.002 Tubular atrophy 0.88 0% Reference group 1%–50% 0.78 0.29 to 2.10 0.62 .50% 0.75 0.23 to 2.42 0.63 Interstitial fibrosis 0.10 0% Reference group 1%–50% 1.16 0.66 to 2.02 0.62 .50% 2.88 1.08 to 7.69 0.04 Renal TMA 1.95 1.07 to 3.55 0.03

TMA, thrombotic microangiopathy. conﬁrmed in future studies. Furthermore, limited by the revised International Chapel Hill Consensus Conference Nomen- relatively small sample size in each subgroup of TMA clature of Vasculitides. Arthritis Rheum 65: 1–11, 2013 2. Berden AE, Ferrario F, Hagen EC, Jayne DR, Jennette JC, Joh K, (i.e., solely acute lesions, acute and chronic lesions, solely Neumann I, Noe¨l LH, Pusey CD, Waldherr R, Bruijn JA, Bajema chronic lesions), the results of renal outcomes of patients IM: Histopathologic classiﬁcation of ANCA-associated glomer- with renal TMA among different categories needs to be ulonephritis. J Am Soc Nephrol 21: 1628–1636, 2010 validated in larger cohorts. 3. Hirsch DJ, Jindal KK, Trillo AA: Antineutrophil cytoplasmic In conclusion, renal TMA in ANCA-associated GN is not antibody-positive crescentic glomerulonephritis and thrombotic microangiopathy. Am J Kidney Dis 26: 385–386, 1995 rare, and it presents with more severe renal injury. Renal 4. Lim HE, Jo SK, Kim SW, Choi HK, Suh IB, Yoon SY, Moon JS, Won TMA is independently associated with all-cause mortality NH, Kwon YJ, Pyo HJ: A case of Wegener’s granulomatosis of patients with AAV. complicated by diffuse pulmonary hemorrhage and thrombotic thrombocytopenic purpura. Korean J Intern Med 13: 68–71, 1998 Acknowledgments 5. Stefanidis I, Helmchen U, Schmitt H, Maurin N, Sieberth HG: We are very grateful to Professor Peter Heeringa for critically Coincidence of haemolytic uraemic syndrome and c-ANCA- reading this manuscript and Professor Xue-Ying Li for the advice on associated rapidly progressive glomerulonephritis. Nephrol Dial statistical analysis. Transplant 13: 1818–1821, 1998 6. Agrawal V, Vaidya CK, Ye J, Freeman J, McKiernan C, Blier PR, This study was supported by a grant of Chinese 973 project (no. Andrzejewski C Jr, Germain M, Braden GL: Concomitant thrombotic 2012CB517700)andfour grants of the National Natural Science Fund thrombocytopenic purpura and ANCA-associated vasculitis in an (nos. 81425008, 81370829, 81321064, and 8140040085). adolescent. Pediatr Nephrol 26: 1317–1320, 2011 7. Nagai K, Kotani T, Takeuchi T, Shoda T, Hata-Kobayashi A, Disclosures Wakura D, Kagitani M, Makino S, Hanafusa T: Successful treat- None. ment of thrombotic thrombocytopenic purpura with repeated plasma exchange in a patient with microscopic polyangitis. Mod Rheumatol 18: 643–646, 2008 References 8. Yamauchi Y, Nagatoya K,Okuno A, Fujii N, Inoue T: Successful 1. Jennette JC, Falk RJ, Bacon PA, Basu N, Cid MC, Ferrario F, Flores- treatment for thrombotic thrombocytopenic purpura compli- Suarez LF, Gross WL, Guillevin L, Hagen EC, Hoffman GS, Jayne DR, cated with myeloperoxidase anti-neutrophil cytoplasmic auto- Kallenberg CG, Lamprecht P, Langford CA, Luqmani RA, Mahr antibody-associated vasculitis. NDT Plus 3: 279–281, 2010 AD, Matteson EL, Merkel PA, Ozen S, Pusey CD, Rasmussen N, Rees 9. Asamiya Y, Moriyama T, Takano M, Iwasaki C, Kimura K, Ando Y, AJ, Scott DG, Specks U, Stone JH, Takahashi K, Watts RA: 2012 Aoki A, Kikuchi K, Takei T, Uchida K, Nitta K: Successful 758 Clinical Journal of the American Society of Nephrology

treatment with rituximab in a patient with TTP secondary to severe anti-neutrophil cytoplasmic autoantibodies. Am J Pathol 170: ANCA-associated vasculitis. Intern Med 49: 1587–1591, 2010 52–64, 2007 10. Abril A, Calamia KT, Cohen MD: The Churg Strauss syndrome 27. Huugen D, van Esch A, Xiao H, Peutz-Kootstra CJ, Buurman WA, (allergic granulomatous angiitis): Review and update. Semin Tervaert JW, Jennette JC, Heeringa P: Inhibition of complement Arthritis Rheum 33: 106–114, 2003 factor C5 protects against anti-myeloperoxidase antibody- 11. Zoltan GL, Fred GS: Hemolytic uremic syndrome, thrombotic mediated glomerulonephritis in mice. Kidney Int 71: 646–654, 2007 thrombocytopenic purpura, and other thrombotic micro- 28. Schreiber A, Xiao H, Jennette JC, Schneider W, Luft FC, Kettritz R: angiopathies. In: Hepinstall’s Pathology of the Kidney, 6th Ed., C5a receptor mediates neutrophil activation and ANCA-induced edited by Jennette JC, Olson JL, Schwartz MM, Silva FG, Phila- glomerulonephritis. J Am Soc Nephrol 20: 289–298, 2009 delphia, Lippincott Williams & Wilkins, 2007, pp 711–717 29. Xing GQ, Chen M, Liu G, Heeringa P, Zhang JJ, Zheng X, E J, 12. Rick ME, Moll S, Taylor MA, Krizek DM, White GC 2nd, Aronson DL: Kallenberg CG, Zhao MH: Complement activation is involved in renal Clinical use of a rapid collagen binding assay for von Willebrand damage in human antineutrophil cytoplasmic autoantibody associ- factor cleaving protease in patients with thrombotic thrombocyto- ated pauci-immune vasculitis. J Clin Immunol 29: 282–291, 2009 penic purpura. Thromb Haemost 88: 598–604, 2002 30. Gou SJ, Yuan J, Chen M, Yu F, Zhao MH: Circulating complement 13. Bajema IM, Hagen EC, Hansen BE, Hermans J, Noe¨l LH, activation in patients with anti-neutrophil cytoplasmic antibody- Waldherr R, Ferrario F, van der Woude FJ, Bruijn JA: The renal associated vasculitis. Kidney Int 83: 129–137, 2013 histopathology in systemic vasculitis: An international survey 31. Noris M, Remuzzi G: Atypical hemolytic-uremic syndrome. study of inter- and intra-observer agreement. Nephrol Dial N Engl J Med 361: 1676–1687, 2009 Transplant 11: 1989–1995, 1996 32. Zipfel PF, Heinen S, Skerka C: Thrombotic microangiopathies: 14. Bajema IM, Hagen EC, Hermans J, Noe¨l LH, Waldherr R, Ferrario New insights and new challenges. Curr Opin Nephrol Hypertens F, Van Der Woude FJ, Bruijn JA: Kidney biopsy as a predictor for 19: 372–378, 2010 renal outcome in ANCA-associated necrotizing glomerulone- 33. George JN, Nester CM: Syndromes of thrombotic micro- phritis. Kidney Int 56: 1751–1758, 1999 angiopathy. NEnglJMed371: 1847–1848, 2014 15. Hauer HA, Bajema IM, van Houwelingen HC, Ferrario F, Noe¨l 34. Meri S: Complement activation in diseases presenting with throm- LH, Waldherr R, Jayne DR, Rasmussen N, Bruijn JA, Hagen EC; botic microangiopathy. Eur J Intern Med 24: 496–502, 2013 European Vasculitis Study Group (EUVAS): Renal histology in 35. Xiao H, Dairaghi DJ, Powers JP,Ertl LS, Baumgart T, Wang Y,Seitz LC, ANCA-associated vasculitis: Differences between diagnostic and PenfoldME,GanL,HuP,LuB,GerardNP,GerardC,SchallTJ,Jaen serologic subgroups. Kidney Int 61: 80–89, 2002 JC, Falk RJ, Jennette JC: C5a receptor (CD88) blockade protects 16. Li ZY,Chang DY, Zhao MH, Chen M: Predictors of treatment against MPO-ANCA GN. JAmSocNephrol25: 225–231, 2014 resistance and relapse in antineutrophil cytoplasmic antibody- 36. ClinicalTrials.gov: A Study to Evaluate the Safety and Efficacy of associated vasculitis: A study of 439 cases in a single Chinese CCX168 in Subjects With ANCA-Associated Vasculitis, 2014. center. Arthritis Rheumatol 66: 1920–1926, 2014 Available at: http://clinicaltrials.gov/ct2/show/NCT01363388. 17. Jennette JC: Rapidly progressive crescentic glomerulonephritis. Accessed September 27, 2014 Kidney Int 63: 1164–1177, 2003 37. Legendre CM, Licht C, Muus P, Greenbaum LA, Babu S, 18. Hellmich B, Flossmann O, Gross WL, Bacon P, Cohen-Tervaert Bedrosian C, Bingham C, Cohen DJ, Delmas Y, Douglas K, Eitner JW, Guillevin L, Jayne D, Mahr A, Merkel PA, Raspe H, Scott DG, F, Feldkamp T, Fouque D, Furman RR, Gaber O, Herthelius M, Witter J, Yazici H, Luqmani RA: EULAR recommendations for Hourmant M, Karpman D, Lebranchu Y, Mariat C, Menne J, conducting clinical studies and/or clinical trials in systemic Moulin B, Nu¨rnberger J, Ogawa M, Remuzzi G, Richard T, vasculitis: Focus on anti-neutrophil cytoplasm antibody-associated Sberro-Soussan R, Severino B, Sheerin NS, Trivelli A, vasculitis. Ann Rheum Dis 66: 605–617, 2007 Zimmerhackl LB, Goodship T, Loirat C: Terminal complement 19. Hogan SL, Falk RJ, Chin H, Cai J, Jennette CE, Jennette JC, inhibitor eculizumab in atypical hemolytic-uremic syndrome. Nachman PH: Predictors of relapse and treatment resistance in N Engl J Med 368: 2169–2181, 2013 antineutrophil cytoplasmic antibody-associated small-vessel 38. Kessenbrock K, Krumbholz M, Scho¨nermarck U, Back W, Gross vasculitis. Ann Intern Med 143: 621–631, 2005 WL, Werb Z, Gro¨ne HJ, Brinkmann V, Jenne DE: Netting neu- 20. Hogan SL, Nachman PH, Wilkman AS, Jennette JC, Falk RJ: trophils in autoimmune small-vessel vasculitis. Nat Med 15: Prognostic markers in patients with antineutrophil cytoplasmic 623–625, 2009 autoantibody-associated microscopic polyangiitis and glomer- 39. Nakazawa D, Shida H, Tomaru U, Yoshida M, Nishio S, Atsumi T, ulonephritis. J Am Soc Nephrol 7: 23–32, 1996 Ishizu A: Enhanced formation and disordered regulation of NETs 21. Nachman PH, Hogan SL, Jennette JC, Falk RJ: Treatment response in myeloperoxidase-ANCA-associated microscopic polyangiitis. and relapse in antineutrophil cytoplasmic autoantibody-associated J Am Soc Nephrol 25: 990–997, 2014 microscopic polyangiitis and glomerulonephritis. JAmSoc 40. Fuchs TA, Kremer Hovinga JA, Schatzberg D, Wagner DD, Nephrol 7: 33–39, 1996 La¨mmle B: Circulating DNA and myeloperoxidase indicate dis- 22. Hauer HA, Bajema IM, Van Houwelingen HC, Ferrario F, Noe¨l ease activity in patients with thrombotic microangiopathies. LH, Waldherr R, Jayne DR, Rasmussen N, Bruijn JA, Hagen EC; Blood 120: 1157–1164, 2012 European Vasculitis Study Group (EUVAS): Determinants of 41. Ma YC, Zuo L, Chen JH, Luo Q, Yu XQ, Li Y, Xu JS, Huang SM, outcome in ANCA-associated glomerulonephritis: A prospective Wang LN, Huang W, Wang M, Xu GB, Wang HY: Modified glo- clinico-histopathological analysis of 96 patients. Kidney Int 62: merular filtration rate estimating equation for Chinese patients 1732–1742, 2002 with chronic kidney disease. J Am Soc Nephrol 17: 2937–2944, 23. Yu XJ, Yu F, Song D, Wang SX, Song Y, Liu G, Zhao MH: Clinical 2006 and renal biopsy findings predicting outcome in renal thrombotic microangiopathy: A large cohort study from a single institute in Received: August 8, 2014 Accepted: January 21, 2015 China. ScientificWorldJournal 2014: 680502, 2014 24. Rosove MH: Thrombotic microangiopathies. Semin Arthritis Rheum 43: 797–805, 2014 Published online ahead of print. Publication date available at www. 25. van Timmeren MM, Heeringa P: Pathogenesis of ANCA-associated cjasn.org. vasculitis: Recent insights from animal models. Curr Opin Rheumatol 24: 8–14, 2012 This article contains supplemental material online at http://cjasn. 26. Xiao H, Schreiber A, Heeringa P, Falk RJ, Jennette JC: Alternative asnjournals.org/lookup/suppl/doi:10.2215/CJN.07910814/-/ complement pathway in the pathogenesis of disease mediated by DCSupplemental. NDT Plus (2010) 3: 279–281 doi: 10.1093/ndtplus/sfq013 Advance Access publication 28 February 2010 Case Report

Successful treatment for thrombotic thrombocytopenic purpura complicated with myeloperoxidase anti-neutrophil cytoplasmic autoantibody-associated vasculitis

Yoko Yamauchi1,2, Katsuyuki Nagatoya2, Ayako Okuno1, Naohiko Fujii1 and Toru Inoue2

1Department of Nephrology, Hyogo Prefectural Nishinomiya Hospital, Hyogo, Japan and 2Department of Nephrology, Osaka Medical

College, Osaka, Japan Downloaded from Correspondence and offprint requests to: Toru Inoue; E-mail: [email protected]

Abstract reveal any abnormality or renal dysfunction, and the pa-

Thrombotic thrombocytopenic purpura (TTP) complicated tient had no familial history of renal disease. Clinical ex- http://ckj.oxfordjournals.org/ with myeloperoxidase anti-neutrophil cytoplasmic autoanti- amination revealed a regular pulse of 76 bpm, a blood body (MPO-ANCA)-associated vasculitis is rare and gener- pressure of 142/78 mmHg and a temperature of 38.5°C. ally has a serious prognosis. We report a case wherein TTP Pretibial pitting oedema and purpura were observed in was successfully treated with repeated plasma exchange both legs. Neurological examination revealed sensory dis- (PE) and MPO-ANCA-associated vasculitis with corticos- turbance of the lower limbs. Laboratory examination re- teroids. The renal function consequently improved such that vealed the following values: haemoglobin (Hb) level, 9 haemodialysis could be discontinued and the patient was 81 g/L; white blood cell count, 19 300 × 10 /L; platelet count, 340 × 109/L; serum creatinine level, 249.29 μmol/L; discharged without any significant complications. at Beijing Medical University on May 14, 2014 serum urea nitrogen level, 14.99 mmol/L; C-reactive protein Keywords: myeloperoxidase anti-neutrophil cytoplasmic autoantibody (CRP) level, 170 mg/L; and MPO-ANCA level, 95 EU. No (MPO-ANCA)-associated vasculitis; plasma exchange (PE); thrombotic other major autoantibodies were detected. Urinalysis re- thrombocytopenic purpura (TTP) vealed the presence of protein (2+) and blood (3+), and the sediment contained >100 red blood cells (RBCs) per high-power field. The urinary protein excretion Background was 0.8 g/day. Computed tomography (CT) of the chest revealed bilateral pleural effusion and ground-glass opacity. Thrombotic thrombocytopenic purpura (TTP) is a multi- Ultrasonography and abdominal CT revealed normal-sized system disorder that is characterized by thrombotic micro- kidneys. On the basis of the clinical course and findings, ac- angiopathy and generally has a serious prognosis. TTP can tive nephritis was suspected and renal biopsy was per- develop idiopathically and also be caused by drugs and au- formed. Light microscopy revealed 10 glomeruli—one toimmune diseases. TTP-associated autoimmune diseases with global sclerosis and two with cellular crescent forma- primarily include systemic lupus erythematosus, systemic tion. The remaining glomeruli exhibited segmental prolif- sclerosis, myositis and, in rare cases, forms of anti-neutro- eration of mesangial cells and a moderate increase of the phil cytoplasmic autoantibody (ANCA)-associated vascu- mesangial matrix. The renal interstitium exhibited inflam- litis, including microscopic polyangiitis. matory cell infiltration and focal atrophic tubules. Severe We report a case of successful treatment of both TTP vasculitis and necrotic lesions were observed in small and myeloperoxidase (MPO)-ANCA-associated vasculitis. vessels. Immunofluorescence staining (immunoglobulin A few case reports on TTP complicated with MPO- [Ig]G, IgA, IgM, C3 and C1q) yielded negative results. ANCA-associated vasculitis have been published. On the The histology was compatible with that of ANCA-related basis of these cases, we discuss the mechanism underlying glomerulopathy. the development of TTP with MPO-ANCA-associated The patient’s clinical course is shown in Figure 1. She vasculitis and the available treatment for the same. had multiple mononeuritis, purpura and haemorrhage of the digestive tract. Renal biopsy revealed extracapillary Case report proliferative glomerulonephritis compatible with rapidly progressive glomerulonephritis. Laboratory data revealed A 59-year-old woman was referred to our hospital for fever elevated MPO-ANCA levels. On the basis of the above and leg oedema. Urinalysis performed previously did not findings, the condition was diagnosed as microscopic

Fig. 1. Clinical and therapeutic course (MPSL, methylprednisolone pulse therapy; PSL, prednisolone; HD, haemodialysis; PE, plasma exchange; RCC, red cell concentrate; Hb, haemoglobin; Plt, platelet; Cr, creatinine; MPO-ANCA, myeloperoxidase anti-neutrophil cytoplasmic autoantibody; BVAS, http://ckj.oxfordjournals.org/ Birmingham vasculitis activity score; CRP, C-reactive protein; ADAMTS13, activity of disintegrin and metalloproteinases with thrombospondin type 1 motif 13). polyangiitis with a Birmingham vasculitis activity score Discussion (BVAS) of 39. On Day 20 of hospitalization, severe haemorrhage of the digestive tract was noted. Therefore, ste- TTP with ANCA-associated vasculitis was initially de- roid pulse therapy with methylprednisolone at a dose of scribed in 1995 [1]. In 2008, Nagai et al. [2] reported at Beijing Medical University on May 14, 2014 0.5 g/day for 3 days was initiated according to the guide- one case and discussed eight other cases reported previous- lines of the Japanese Society of Nephrology, and prednis- ly in the English or Japanese literature. They found that all olone (PSL) at a dose of 50 mg/day was subsequently the concerned patients were middle-aged women. All pa- administered intravenously. On hospitalization Day 22, tients developed TTP during the active phase of vasculitis. haemodialysis was initiated because of progressive renal No relationship was noted between the outcomes and the dysfunction. Immediately after steroid therapy was initiat- degree of angiitis or the ANCA titre. Furthermore, patients ed, the patient’s condition, including fever and general fa- with a platelet count of <50 × 109/L had a poor prognosis, tigue, improved and the serum CRP and MPO-ANCA whereas the proposed severity scoring index predicting the values gradually decreased to <0.2 mg/dL and <10 EU, re- survival of TTP patients does not include the platelet count spectively. However, on hospitalization Day 28, haemoly- [3]. In the present case, the patient was a 59-year-old wom- tic anaemia (Hb level, 64 g/L) with RBC fragmentation an who developed TTP during the active phase of MPO- and thrombocytopenia (61 × 109/L) developed. At this ANCA-associated vasculitis. time, the lactate dehydrogenase level was 680 U/L and TTP may be congenital, as in the case of Upshaw– the haptoglobin level reduced to <1 μmol/L. We suspected Schulman syndrome, or acquired, as in the case of autoim- TTP and initiated plasma exchange (PE) with 3600 mL of mune diseases, infection, malignancy, pregnancy, drugs in- fresh-frozen plasma as the replacement fluid. After a PE cluding cancer chemotherapy and idiopathic type. Anti- session, we confirmed a 41% reduction in the activity of ADAMTS13 IgG autoantibodies, which reduce disintegrin and metalloproteinases with thrombospondin ADAMTS13 activity, have been detected in TTP patients. type 1 motif 13 (ADAMTS13). Anti-ADAMTS13 IgG ADAMTS13 deficiency is responsible for most cases of autoantibodies were not detected. A few days later, psy- acquired TTP. Therefore, PE was performed to supplement chosis (delirium) and pyrexia developed. The diagnosis ADAMTS13 and to eliminate anti-ADAMTS13 IgG auto- of TTP was confirmed, and 17 consecutive PE sessions antibodies and very high-molecular-weight von Willeb- were conducted. Thereafter, the clinical and laboratory rand factor multimers. In TTP secondary to autoimmune findings gradually improved. Steroid therapy proved effec- diseases, ADAMTS13 activity can vary from normal to tive against microscopic polyangiitis, and the disease activ- markedly reduced. However, in some cases, anti- ity reduced such that haemodialysis was not warranted, ADAMTS13 IgG antibodies are not detected, as in the even after the PSL dose was tapered. Finally, the patient present case. Thus, factors other than decreased was discharged without any significant complications ADAMTS13 activity may contribute to the pathogenesis and eventually resumed work. of TTP secondary to autoimmune diseases [4]. We per- Successful treatment for TTP complicated with MPO-ANCA-associated vasculitis 281 Table 1. Profiles of patients with TTP complicated with ANCA-associated vasculitis, requiring PE and haemodialysis

Platelet Age in ADAMTS13 Angiitis-related count Author and year Ref. no. years/sex Angiitis Therapy activity (%) antibody (EU) (×109/L) Outcome

Hirsch et al. 1995 1 66/F MPA PSL, MPSL, CPA, PE ND MPO-ANCA (340) 9.8 Alive Lim et al. 1998 66/F WG PSL, MPSL, CPA, PE ND PR3-ANCA (640) 2.8 Alive Yamasaki et al. 2001 56/F PN MPSL, IVCY, PI, PE ND MPO-ANCA (201) 1.5 Dead Fujisaki et al. 2005 70/F PN PSL, MPSL, CPA, PE 7 MPO-ANCA (21) ND Dead Nagai et al. 2008 2 77/F MPA PSL, MPSL, PI, PE 27 MPO-ANCA (238) 4.8 Alive Present case 59/F MPA PSL, MPSL, PI, PE 41 MPO-ANCA (95) 6.1 Alive

MPA, microscopic polyangiitis; WG, Wegener granulomatosis; PN, polyarteritis nodosa; PSL, prednisolone; MPSL, methylprednisolone pulse therapy; CPA, cyclophosphamide; IVCY, intravenous pulse cyclophosphamide; PI, plasma infusion; PE, plasma exchange; ND, not described. formed PE in the present case to eliminate some unknown emphasize that TTP should be considered in the differential factor(s) inhibiting ADAMTS13. diagnosis in cases of thrombocytopenia with vasculitis. Ear- The relationship between TTP and ANCA-associated ly appropriate treatment of TTP can improve the patient mor- Downloaded from vasculitis remains unclear. Cases of TTP complicated with bidity and mortality. ANCA-associated vasculitis, requiring PE and haemodialysis, have been reported previously (Table 1). In all these Conflict of interest statement. None declared. cases, TTP developed rather early after glucocorticoid therapy was initiated. Given that TTP is characterized by thrombotic microangiopathy, endothelial damage due to References http://ckj.oxfordjournals.org/ ANCA-associated vasculitis and/or the process of endothe- 1. Hirsch DJ, Jindal KK, Trillo AA. Antineutrophil cytoplasmic anti- lial healing after glucocorticoid therapy may promote the body-positive crescentic glomerulonephritis and thrombotic microan- development of TTP. Therefore, careful observation is nec- giopathy. Am J Kidney Dis 1995; 26: 385–386 essary when glucocorticoid therapy is initiated for various 2. Nagai K, Kotani T, Takeuchi T et al. Successful treatment of throm- diseases, including ANCA-associated vasculitis. In two of botic thrombocytopenic purpura with repeated plasma exchange in a the six described cases, haemodialysis was discontinued; patient with microscopic polyangitis. Mod Rheumatol 2008; 18: 643–646 in the present case, the patient was discharged. The degree 3. Wyllie BF, Garg AX, Macnab J et al. Thrombotic thrombocytopenic of renal dysfunction in the early phase seems to reflect the purpura/haemolytic uraemic syndrome: a new index predicting res- at Beijing Medical University on May 14, 2014 renal outcome. Of the six reported cases, five involved ponse to plasma exchange. Br J Haematol 2006; 132: 204–209 Asian patients and the sixth was a Canadian patient. The 4. Sato T, Hanaoka R, Ohshima M et al. Analyses of ADAMTS13 ac- prevalence of MPO-ANCA-associated vasculitis, which is tivity and its inhibitor in patients with thrombotic thrombocytopenic higher than that of Wegener granulomatosis in Asia, may purpura secondary to connective tissue diseases: observations in a single hospital. Clin Exp Rheumatol 2006; 24: 454–455 be attributed to the bias. 5. Howard MA, Williams LA, Terrell DR et al. Complications of plas- The risk of PE should be recognized. In a 9-year cohort ma exchange in patients treated for clinically suspected thrombotic study on 206 consecutive patients treated for TTP [5], 26% thrombocytopenic purpura-hemolytic uremic syndrome. Transfusion of the patients had major PE-associated complications, in- 2006; 46: 154–156 cluding systemic infection, venous thrombosis and hypoten- 6. von Baeyer H. Plasmapheresis in thrombotic microangiopathy-asso- sion warranting dopamine treatment, and 2% of the patients ciated syndromes: review of outcome data derived from clinical trials and open studies. Ther Apher 2002; 6: 320–328 died of such complications (one died of haemorrhage owing 7. Remuzzi G. HUS and TTP: variable expression of a single entity. to the insertion of a central venous catheter and one died of Kidney Int 1987; 32: 292–308 catheter-related sepsis). However, the mortality rate of TTP 8. Amorosi EL, Ultmann JE. Thrombotic thrombocytopenic purpura: is currently only 12–14% in PE-treated patients [3,6] but ap- report of 16 cases and review of the literature. Medicine 1966; 45: proximately 90% without PE treatment [6–8]. The above 139–159 9. Moake JL. Thrombotic microangiopathies. N Engl J Med 2002; 347: findings indicate that the potential risk of TTP exceeds that – of PE treatment. PE should be initiated even if the diagnosis 589 600 10. George JN. How I treat patients with thrombotic thrombocytopenic of TTP is not confirmed [9,10]. In our case, PE was initiated purpura-hemolytic uremic syndrome. Blood 2000; 96: 1223–1229 immediately after TTP was suspected. After 17 PE sessions, the clinical and laboratory findings gradually improved. We Received for publication: 27.1.10; Accepted in revised form: 2.2.10