Microcirculation Inflammation Associates With Outcome in Renal Transplant Patients With De Novo Donor-Specific Antibodies

American Journal of Transplantation 2013; 13: 485 492 Wiley Periodicals Inc. Brief Communication C Copyright 2012 The American Society of Transplantation and the American Society of Transplant Surgeons doi: 10.1111/j.1600-6143.2012.04325.x Microcirculation Inflammation Associates With Outcome in Renal Transplant Patients With De Novo Donor-Specific Antibodies H. de Kort a, M. Willicombe b, P. Brookes c, K. M. Dominy a, E. Santos-Nunez c, J. W. Galliford b,k.chan b, D. Taube b, A. G. McLean b, H. T. Cook a and C. Roufosse a, a Department of Histopathology, Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK b Imperial College Kidney and Transplant Centre, Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK c Histocompatibility and Immunogenetics Laboratory, Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK Corresponding author: Candice Roufosse, Candice.Roufosse@imperial.nhs.uk In renal transplant patients with de novo donorspecific antibodies (dndsa) we studied the value of microcirculation inflammation (MI; defined by the addition of glomerulitis (g) and peritubular capillaritis (ptc) scores) to assess long-term graft survival in a retrospective cohort study. Out of all transplant patients with standard immunological risk (n = 638), 79 (12.4%) developed dndsa and 58/79 (73%) had an indication biopsy at or after dndsa development. Based on the MI score on that indication biopsy patients were categorized, MI0 (n = 26), MI1 + 2(n= 21) and MI 3 (n = 11). The MI groups did not differ significantly pretransplantation, whereas posttransplantation higher MI scores developed more anti-hla class I + II DSA (p = 0.011), showed more TCMR (p < 0.001) and showed a trend to C4d-positive staining (p = 0.059). Four-year graft survival estimates from time of indication biopsy were MI0 96.1%, MI1 + 2 76.1% and MI 3 17.1%; resulting in a 24-fold increased risk of graft failure in the MI 3 compared to the MI0 group (p = 0.003; 95% CI [3.0 196.0]). When adjusted for C4d, MI 3 still had a 21-fold increased risk of graft failure (p = 0.005; 95% CI [2.5 180.0]), while C4d positivity on indication biopsy lost significance. In renal transplant patients with de novo DSA, microcirculation inflammation, defined by g + ptc, associates with graft survival. Key words: Antibody-mediated rejection, C4d, donorspecific antibodies, histopathology, microcirculation inflammation, renal transplantation Abbreviations: AMR, antibody-mediated rejection; CI, confidence interval; DNS, data not shown; DSA, donorspecific antibodies; dndsa, de novo donor-specific antibodies; g, glomerulitis; HLA, human leukocyte antigen; IF/TA, interstitial inflammation/tubular atrophy; IVIg, intravenous immunoglobulin; IQR, interquartile range; MFI, mean fluorescence index; MI, microcirculation inflammation; MIC, major histocompatibility complex class I chain-related gene; MMF, mycophenolate mofetil; P, prednisolone; ptc, peritubular capillaritis; SD, standard deviation; Tac, tacrolimus; TCMR, T cell mediated rejection; TG, transplant glomerulopathy. Received 31 May 2012, revised 11 September 2012 and accepted for publication 27 September 2012 Introduction Antibody involvement in renal allograft rejection was first described in the 1960s and, after the discovery of C4d as a biomarker in 1991, criteria for antibody-mediated rejection (AMR) were developed (1). The Banff consensus for allograft pathology defines AMR as a triad of circulating donor-specific antibodies (DSA), C4d deposition in peritubular capillaries and specific histological parameters. Both acute and chronic AMR are associated with inferior graft outcome (2 4). The diagnostic criteria for AMR are currently undergoing review (5), with the increasing recognition of C4d-negative AMR based on transcript analysis studies (6) and on observations in sensitized patients with preformed DSA (7). In these patients, microcirculation inflammation (MI), defined by the combination of glomerulitis (g) and peritubular capillaritis (ptc) scores, is associated with worse outcome (7). In the setting of de novo DSA (dndsa), MI has also been noted (8). On a population scale, de novo antibodies targeted at the transplanted donor antigens are associated with worse outcome (9) but not all patients with dndsa develop detrimental damage to the graft (10,11). Our aim was to define the incidence of C4d-positive AMR and C4d-negative AMR in the first indication biopsy up to 1 month before or anytime after dndsa developed in a group of patients with dndsa, based on microcirculation inflammation (MI) and C4d scores. The value of the MI score was compared to C4d staining to establish a correlation with outcome in renal transplant patients with dndsa. 485

de Kort et al. Table 1: Clinical demographics comparing all transplant recipients with standard immunological risk to those that developed de novo DSA All De novo DSA De novo DSA with n = 559 n = 79 biopsy n = 58 Recipient Age (yr; median [IQR]) 50 [39 59] 45 [34 55] 45 [34 55] Gender (% male) 68.2 64.6 65.5 Ethnicity (% cauc/asian/afrocar/other) 48/33/14/5 38/43/14/5 40/43/14/3 Retransplanted (%) 8.6 10.1 12.1 Transplant related Total HLA mismatches 3.1 (±1.6) 3.6 (±1.3) 3.9 (±1.3) HLA-class 1 MM (mean ±SD) 2.3 (±1.2) 2.4 (±1.2) 2.8 (±1.0) HLA-class 2 MM (mean ±SD) 0.9 (±0.7) 1.2 (±0.6) 1.1 (±0.6) DDRT/LURT/LRRT (%) 52/19/29 52/20/28 52/22/26 Induction therapy (% Campath) 79.6 91.1 89.7 FU graft survival (mnth; median [IQR]) 32 [20 50] 33 [19 50] 33 [20 50] FU patient survival (mnth; median [IQR]) 34 [22 51] 39 [25 53] 40 [25 53] The category defined with all are all renal transplant patients with standard immunological risk (n = 559) in the period 2005 2010 excluding the patients that developed de novo DSA (n = 79). The de novo DSA with biopsy (n = 58) are a sample from the total 79 patients who developed de novo DSA. Afrocar = Afro-Caribbean; cauc = Caucasian; DDRT = deceased donor renal transplant; DSA = donor-specific antibody; FU = follow-up; HLA = human leukocyte antigen; IQR = interquartile range; LRRT = live-related renal transplant; LURT = live unrelated renal transplant; MM = mismatch; mnth = month; SD = standard deviation; yr = year. Material and Methods Study design All transplant recipients with standard immunological risk (n = 638) receiving a crossmatch negative kidney transplant at the Imperial College Kidney and Transplant Center between November 2005 and January 2011 were included in this retrospective cohort study. Excluded were dual-organ transplantation, ABO-incompatible patients, or patients with DSA pretransplantation. Seventy-nine (12.4%) patients developed dndsa. This was defined as (1) pretransplantation either no antibodies, only nondonor-specific HLA-antibodies or DSA with an MFI<300 assessed by Luminex single antigen beads; (2) posttransplantation any measurement of a DSA >500 MFI or MFI 300 500 in two independent serum samples. The first indication biopsy within a month before dndsa detection or the first thereafter was used for analyses and will be referred to as: index biopsy (n = 58). Twentyone dndsa patients were excluded from analyses, because no indication biopsies were available at or after dndsa development. Immunosuppressive therapy Patients received induction therapy of alemtuzumab (Campath-1H, Millennium Pharmaceuticals) 30 mg IV or daclizumab (Zenapax, Roche) 2 2 mg/kg IV, both with methyl-prednisolone (500 mg) IV preoperatively, followed by prednisolone (P) 1 mg/kg/day (maximum 60 mg) for 3 days, then 0.5 mg/kg/day (maximum 30 mg) for 4 days and then discontinued after day 7. Alemtuzumab treated patients received maintenance mono-therapy consisting of low-dose tacrolimus ([Tac] mean trough level 5 12 ng/ml). Daclizumab treated patients received maintenance therapy consisting of mycophenolate mofetil (MMF) and low-dose Tac (mean trough level 5 8 ng/ml) (12). First acute rejection episodes either histologically confirmed or clinically suspected as T cell-mediated rejection (TCMR) were treated with P + MMF (n = 10), P alone (n = 3), alemtuzumab (n = 3), P + IVIg (n = 2), or IVIg alone (n = 1). When recognized as AMR they were treated with plasma exchange + IVIg (n = 31) with rituximab (1 patient) or eculizumab (2 patients). DSA assessment Before transplantation all donor-recipient pairs had a negative T- and B cell complement-dependent cytotoxicity crossmatch and a negative T cell flow cytometric crossmatch. DSA were assessed using LABScreen mixed beads (One lambda, Canoga Park, CA, USA). When positive, the specificity of their anti-hla antibody was identified using LABScreen single antigen beads. Patients were typed for HLA A, B, Cw, DR and DQ antigens. HLA-DP directed DSA were not assessed. Histopathology Tissue was obtained from the Imperial College Healthcare Tissue Bank, which has ethics approval to both collect human tissue and release material to researchers (MREC 07/MRE09/54). All biopsies from patients with dndsa were graded according to the current Banff classification (13) by C.R. Definition: g describes the percentage of glomeruli with occlusion of the endocapillary space by mononuclear cells and endothelial swelling (range 0 3) (14); ptc describes the number of infiltrating monocytes and/or neutrophils in the most affected capillary when >10% of peritubular capillaries is affected (range 0 3) (13); MI is the sum of g + ptc (range 0 6). C4d (BI-RC4D, Biomedica, Austria) peritubular capillary staining was performed on paraffin-embedded sections and classified as negative/minimal or focal/diffuse (13). TCMR and suspicious for TCMR were defined according to Banff criteria (13). C4d-positive acute AMR was defined by C4d-positive staining with g, ptc or thrombosis, but not acute tubular injury alone. C4dnegative (suspicious for) acute AMR was defined as g, ptc or thrombosis, with no C4d staining. Transplant glomerulopathy (TG) was defined as double contours on light microscopy; with no or small amounts of immunecomplex deposition on immunofluorescence and/or electron microscopy; and absence of hepatitis C infection and of clinical features of thrombotic microangiopathy. Statistical analyses For data the median and interquartile range [IQR] were expressed. MI scores (range 0 6) were categorized into three groups; MI0 (n = 26), MI1 + 2(n= 21), and MI 3(n= 11). Clinical characteristics for the three groups were analyzed with a Fisher exact test for count data, or with oneway ANOVA for continuous variables. Graft failure was defined as resuming dialysis and censored for patient death with a functioning graft. Outcome and cumulative hazard functions were estimated with exact Cox regression 486 American Journal of Transplantation 2013; 13: 485 492

Microcirculation and De Novo DSA Figure 1: Overall death-censored graft survival stratified by total standard immunological risk population (n = 638), total population minus patients with dndsa (n = 559), all de novo DSA patients (n = 79) and de novo DSA patients with an indication biopsy from 1 month before or after dndsa developed (n = 58). analyses, adjusting for time from transplant to index biopsy. The Kaplan Meier product limit method was used to estimate the kidney allograft survival times. In addition, ROC analysis using a logistic model was performed, defining outcome as graft failure within the follow up. Significance was set at p 0.05. Statistical calculations were performed using SPSS 16.0 (SPSS, Chicago, IL, USA) and STATA 11.0 (Statacorp, College Station, TX, USA). Results All renal transplant recipients with standard immunological risk between 2005 and 2010 (n = 638) were analyzed at least once after transplantation for the formation of anti- HLA antibodies by Luminex. Seventy-nine (12.4%) patients were found to have dndsa at a median time of 3.8 [IQR 0.6 10.8] months after transplantation; with 24 patients displaying only anti-hla class I DSA, 33 only anti-hla class II DSA and 22 patients having both. Of these 79 patients, 58 (73.4%) had an index biopsy. These 58 patients received a total of 255 biopsies, reaching an average of 4.4 biopsies per patient. The patient demographics are described in Table 1, comparing the total renal transplant population minus patients with dndsa (n = 559), to the full group with dndsa (n = 79) and to the group with dndsa and an index biopsy (n = 58). Graft survival at 4 years after transplantation was estimated for the four groups: total standard immunological risk population 88.2% (n = 638), total population minus patients with dndsa 90.6% (n = 559), patients with dndsa 73.2% American Journal of Transplantation 2013; 13: 485 492 487

de Kort et al. Table 2: Comparison of microcirculation inflammation score groups for demographics, immunological risk factors and rejection characteristics MI 0 MI 1 + 2 MI 3 n = 26 n = 21 n = 11 p-value Donor Age (yr; mean ± SD) 44 [38 57] 51 [42 56] 54 [47 68] 0.22 Gender (% male) 67 50 25 0.13 Recipient Age (yr; mean ± SD) 46 [33 54] 44 [35 56] 42 [32 60] 0.98 Gender (% male) 62 57 91 0.14 Ethnicity (%) cauc/asian/afrocar/other 31/50/19/0 38/43/14/5 64/27/0/9 0.27 Total HLA mismatches 4.0 (±1.0) 3.7 (±1.5) 4.0 (±1.3) 0.61 HLA-class I MM (mean ± SD) 2.8 (±1.0) 2.7 (±1.0) 2.8 (±1.0) 0.90 HLA-class II MM (mean ± SD) 1.2 (±0.6) 1.0 (±0.7) 1.2 (±1.2) 0.72 Pretransplantation Retransplantation (% yes) 8 10 27 0.24 Preemptive (% yes) 8 14 9 0.85 Donor type (% DDRT) 62 33 64 0.12 Induction therapy (% Campath) 96 86 82 0.27 HLA-antibodies (non-dsa) (% yes) 35 19 18 0.45 Posttransplantation Time to DSA (mnth; median [IQR]) 2.9 [0.5 6.3] 4.7 [0.6 11] 5.2 [3.7 7.7] 0.46 Time to Bx (mnth; median [IQR]) 4.1 [0.8 14] 6.0 [0.6 19] 8.8 [5.6 27] 0.65 Total follow-up (mnth; median [IQR]) 31 [20 50] 45 [34 52] 48 [28 52] 0.18 Anti-HLA class I&II DSA (% yes) 31 19 73 0.011 Anti-HLA class I DSA (% yes) 69 43 91 0.20 Anti-HLA class II DSA (% yes) 62 76 82 0.42 MFI >1000 (% yes) 62 62 91 0.18 TCMR (% type 1A) 0 33 73 <0.001 C4d staining (% focal/diffuse) 23 29 64 0.059 Analysis of the three microcirculation inflammation groups as determined on their index biopsy using 1-way ANOVA and Fisher exact test. Afrocar = Afro-Caribbean; cauc = Caucasian; CI = confidence interval; DDRT = deceased donor renal transplant; DSA = donor-specific antibody; FU = follow-up; HLA = human leukocyte antigen; IQR = interquartile range; MFI = mean fluorescence index of the donorspecific HLA-antibody at the first time DSA were detected; MI = microcirculation inflammation; MM = mismatch, mnth = month; SD = standard deviation; TCMR = T cell mediated rejection; yr = year. (n = 79) and patients with dndsa and with index biopsy 70.7% (n = 58) (Figure 1), showing reduced graft survival estimates in the recipients with dndsa. There were no significant differences between the estimated graft survival rates of all dndsa patients and those 58 patients with a biopsy. Further results described below focus on the 58 dndsa patients with an index biopsy. Clinical characteristics The dndsa patients (n = 58) were divided into three categories based on MI scores assessed on the index biopsy; MI0, MI1 + 2andMI 3. Eighteen (31%) recipients had a positive g score; 31 (53%) had a positive ptc score; 17 (29%) biopsies showed both g and ptc positivity. An MI score of 1 was mostly represented by a solitary ptc1 score (91%). In Table 2 the three groups were compared and analyses showed significant differences between the three groups for posttransplantation factors; development of both donor-specific HLA antibodies class I and II (p = 0.011), TCMR on index biopsy (p>0.001) and a trend for C4d staining (p = 0.059). Graft loss occurred in 13/58 (22%) patients in this cohort, and was attributed to AMR in 5 patients, whereas in 7 patients AMR was present but in combination with other factors, and 1 was not related to AMR (Table 3). Patient survival did not differ between the three MI groups with 4-year patient survival estimates of MI0 91%, MI 1 + 2 91% and MI 3 100% (p = 0.468; data not shown [DNS]). Microcirculation inflammation is associated with graft survival in patients with de novo DSA The three MI groups were compared for graft survival from time of index biopsy (Figure 2A) showing a profound decline in graft survival for the patients who had a high MI score (MI 3) on the index biopsy. The risk of renal graft failure in the group showing an MI 3 on their index biopsy was 24 times the risk of renal graft failure in the group with an MI0 score (p = 0.003; 95% confidence interval [CI] [3.0 196.0]). Graft survival analysis comparing C4d-positive to C4d-negative staining on the index biopsy (Figure 2B) showed a much less pronounced distinction that did not quite reach statistical significance. Patients with a C4d-positive score had a 2.9 fold increased risk 488 American Journal of Transplantation 2013; 13: 485 492

Table 3: All 13 graft failures within the cohort of 58 de novo DSA patients with index biopsy At index biopsy On subsequent biopsies Mnth HLA Max Max Max Mnth MI DSA class Mnth C4d MI cg C4d MI cg Mnth graft group positive DSA biopsy (0 3) (0 6) (0 3) (0 3) (0 6) (0 3) TG failure Microcirculation and De Novo DSA 1 MI0 0.3 1 + 2 0.7 0 0 0 3 1 0 2.9 Ureteric complications 2 MI1 + 2 3.3 1 + 2 3.7 0 2 0 3 5 3 14.1 17.1 acute AMR + TG + renal artery stenosis 3 MI1 + 2 2.9 1 + 2 7.5 1 1 0 1 2 0 39.6 Pyelonephritis + TMA 4 MI1 + 2 17.7 1 + 2 17.5 0 1 0 0 2 0 23.7 C4d-negative acute AMR + TCMR + pyelonephritis 5 MI1 + 2 5.7 2 5.3 3 1 0 3 2 0 12.9 Acute AMR (reduced IS) + OI 6 MI 3 9.4 1 + 2 8.8 0 4 2 2 5 3 8.8 20.1 TG + TCMR 7 MI 3 36.0 2 37.0 1 5 3 2 5 3 37.0 45.2 acute AMR + TG 8 MI 3 4.0 1 + 2 27.4 3 3 3 3 6 3 27.4 38.4 TG 9 MI 3 5.1 1 + 2 8.6 3 4 1 3 5 3 8.6 17.4 TG 10 MI 3 17.4 1 + 2 25.9 2 4 3 1 5 3 25.9 49.0 TG + TCMR 11 MI 3 0.2 1 + 2 0.2 3 3 0 2 3 0 3.0 AcuteAMR + pyelonephritis 12 MI 3 5.2 1 + 2 4.9 3 5 0 3 4 3 14.8 35.5 TG + TCMR + pyelonephritis 13 MI 3 6.1 1 + 2 9.1 3 3 0 3 5 3 15.9 18.5 TG + TCMR + recurrent GN Findings on indication biopsy taken within 1 month before or after de novo DSA development and findings on subsequent biopsies are described with reason of failure. AMR = antibody-mediated rejection; cg = allograft glomerulopathy Banff score; DSA = donor-specific antibody; FU = follow-up; GN = glomerulonephritis; IS = immunosuppression; MI = microcirculation inflammation; mnth = month; OI = opportunistic fungal infection; TCMR = T cell mediated rejection; TMA = thrombotic microangiopathy; TG = transplant glomerulopathy. of graft failure (p = 0.066; 95% CI [0.9 9.3]). Albeit both parameters seemed to be powerful risk factors for graft survival, when the MI score was C4d adjusted, patients with an MI 3 still had 21-fold increased risk of graft failure (p = 0.005; 95% CI [2.5 180.0]), whereas a C4dpositive score lost significance with a 1.4-fold increased risk of allograft failure (p = 0.570; 95% CI [0.4 5.0]). On a separate analysis, the area under the ROC was poor for C4d-positive staining (0.64) whereas it was good for MI score (0.85). Our model showed that MI score is a significantly better predictor of graft loss (p = 0.006, Figure 3). When only diffuse C4d-staining (n = 10) was considered positive, C4d-positivity did associate with a worse survival estimate (p = 0.001; DNS) and in multivariate analysis retained significance with a fourfold increased risk of graft failure, independent of the MI score (p = 0.044; 95% CI [1.04 15.7]). The area under the ROC was again poor (0.69; DNS). Histological correlates Based on findings in the index biopsy, in the MI0 group (n = 26), 24/26 were rejection free, 2 were suspicious for TCMR and 6/26 were C4d-positive (4 focal, 2 diffuse). In the MI1 + 2 group (n = 21), 12/21 were TCMR free, 7 had TCMR (4 grade I, 3 grade II) and 2 were suspicious for TCMR. By definition all 21 patients had MI + DSA so were at least suspicious for AMR, depending on C4d status. Six of the 21 were C4d-positive (acute AMR; 3 focal, 3 diffuse) and 1/6 also had TG; 15/21 were C4d-negative, none with TG.IntheMI 3 group (n = 11), 1/11 was TCMR free, 8 had TCMR (2 grade I, 6 grade II) and 2 were suspicious for TCMR. Seven of the 11 were C4d-positive (acute AMR; 2 focal, 5 diffuse), and 3/7 also had TG. There were 4/11 C4d-negative cases (suspicious for acute AMR), and 3/4 also had TG (Table 3, Figure 4). Total incidence of (C4dpositive) acute AMR was 22.4% (13/58 patients) and total incidence of (C4d-negative) suspicious for AMR was 33% (19/58 patients). TG was present in 7/58 (12%) index biopsies, only in the MI positive groups. Including follow-up biopsies, 16/58 patients developed TG; 1 (4%), 6 (29%) and 9 (82%) in the MI0, MI1 + 2andMI 3 group, respectively. When graft survival analysis was adjusted for having TG on the index biopsy, patients with an MI 3 still had a 16-fold increased risk of graft failure compared to patients with an MI0 score (p = 0.014; 95% CI [1.8 149.3]). Patients with TG on the indexbiopsyhada9.6timesincreasedriskofgraftfailure (p = 0.001; 95% CI [2.7 35.9]) for which significance was lost after multivariate analysis including MI and C4d score. Discussion Formation of anti-hla antibodies after transplantation, and particularly DSA, is associated with acute rejection, chronic rejection and graft loss (10,15 19). We set out to study patients with de novo DSA for biopsy features that distinguish those who progress to graft failure from those who retain graft function. We observed that patients with severe microcirculation inflammation (MI 3) on the first indication biopsy taken within 1 month or after dndsa occur, are at a American Journal of Transplantation 2013; 13: 485 492 489

de Kort et al. Figure 2: Death-censored graft survival from time of index biopsy comparing (A) the 3 microcirculation inflammation groups, and (B) C4d-positive versus C4d-negative staining. (A) Microcirculation inflammation (MI) is defined by the addition of the glomerulitis and peritubular capillaritis score and separated into three categories based on the MI score on the first indication biopsy around or after de novo DSA. (B) C4d staining of peritubular capillaries scored on the first indication biopsy around or after de novo DSA. significantly increased risk for graft failure. The MI score is better at discriminating graft failure from survival than C4d staining in this cohort of patients with dndsa. We found that, in the medium term at least, patients without microcirculation inflammation (MI0) on their index biopsy behaved like the standard immunological risk patients who did not develop dndsa (4-year graft survival rates after transplantation, 91.7% vs. 90.6%, respectively). In two prospective studies on indication biopsies, late biopsies (>1-year posttransplant) often coincided with de novo anti-hla class II DSA (with or without anti-hla class I DSA), MI, microcirculation damage (TG and peritubular capillary basement membrane multilayering) and graft loss (8,20). In a study on protocol biopsies of patients with preformed DSA, MI and anti-hla class II DSA were correlated with bad outcome independent of C4d (7). In a study of standard immunological risk patients, approximately 15% developed dndsa late after transplantation predicted by nonadherence and HLA-DRb1 mismatch (11). Significantly more ptc was noted in 6-month protocol biopsies in patients with dndsa compared to those without (11). In this retrospective study, we find that anti-hla class II DSA development is associated with worse survival and although there was a trend of later biopsy time with increasing MI score, this was not significant. Our aim was to distinguish within the dndsa patients those who would progress to graft failure from those who would retain graft function. By using the MI score, shown to correlate with DSA, we found a way to do so. Our transplant center employs a steroid-sparing monoimmunosuppressive regimen, consisting of alemtuzumab induction with tacrolimus maintenance therapy in the majority of renal transplant recipients with standard immunological risk. The 13% incidence of dndsa formation 490 American Journal of Transplantation 2013; 13: 485 492

Microcirculation and De Novo DSA Figure 3: ROC analysis comparing C4d and MI score. Sensitivity (Y-as) and 1-specificity (X-as) are displayed for dichotomized C4d staining (discontinuous line; negative/minimal staining vs. focal/diffuse staining) and categorized MI score (solid line; MI0, MI1 + 2, MI 3). Reference line is depicted in gray. ROC analysis was performed using a logistic model, defining outcome as graft loss within the follow-up. in our population is similar to other centers (11) Time to development of dndsa is shorter than some reports and similar to others (8,11,20 22). One group, using a steroid-based regimen, describes an incidence of dndsa in 27% with 90% of protocolized samples showing dndsa development at the 1 or 6 month posttransplantation sample (22). Currently a US-based NIH-supported trial is underway (CTOT-02) which has as a first outcome parameter the screening for incidence and timing of alloantibody development. Future results of this trial might shed light on AMR and DSA incidence in different induction protocols. The diagnostic criteria for AMR are currently being revised to include C4d-negative cases (5 7,23,24). In a recent study in mostly nonpresensitized renal transplant patients, MI>0 in late biopsies indicated the presence of DSA and predicted inferior graft survival (25). The cut-off for defining ptc as positive was higher in this study than in ours, making direct comparison difficult. Using the Banff criteria for ptc (13,14), we found that patients with dndsa and an MI 3 did poorly, even if C4d was negative. The results in patients with MI scores of 1 or 2 were less clear cut. We also noted that in patients with dndsa and MI, C4d staining could fluctuate between positivity and negativity on follow-up biopsies, as noted in presensitized patients by others (7). In this respect, cumulative C4d scores may prove a more useful prognostic indicator than individual Figure 4: Flowchart depicting attribution of graft failures. Microcirculation inflammation and C4d peritubular capillary staining scores are based on (immuno-) histological parameters in the index biopsy. These scores are shown in relation to the attributed causes for graft failure. Gray boxes indicate C4d+ at the time of index biopsy or on a later biopsy. aamr = acute antibody-mediated rejection; GN = glomerulonephritis; MI = microcirculation inflammation; OI = opportunistic fungal infection; IS = immunosuppressive drugs; PN = pyelonephritis; RAS = renal artery stenosis; TCMR = T cell mediated rejection; TG = transplant glomerulopathy; TMA = thrombotic microangiopathy. measurements. We also confirm the existence of rare C4dpositive cases (n = 6/26) with stable clinical features and the persistent absence of MI on biopsy (MI = 0), at least in the medium term. We do not know yet whether these cases represent a form of accommodation or of indolent chronic rejection, which only time will tell. MI0 patients had good outcomes, whether C4d was positive or not. These patients are unlikely to be responders to treatment as only 6/26 had received augmented immunosuppression. A diagnostic definition of C4d-negative AMR could be established by defining a threshold for MI with the use of long-term survival data. However, a consensus is necessary on how to define the MI score, before such diagnostic criteria can be agreed upon. We acknowledge the limitations of this retrospective study. All renal transplant patients at our center had at least one test for DSA, but the timing of DSA sampling was not standardized. There was no established protocol for biopsy timing either, so the timing to detection of DSA and the timing of the index biopsy were biased. Some cases of dndsa could have been missed as HLA-DP specificities are unknown and no tests were employed for non- HLA DSA such as major histocompatibility complex class I chain-related gene A (26). The DSA described are however de novo, and we included only indication biopsies. American Journal of Transplantation 2013; 13: 485 492 491

de Kort et al. In patients who develop de novo DSA, the microcirculation inflammation score, defined by glomerulitis plus peritubular capillaritis, is associated with inferior graft survival. Much of this can be attributed to clinical episodes of AMR according to established criteria, but not the C4d-negative cases. Whether treating early dndsa development or B cell immunity may improve outcome in these patients still needs to be established. Patient-tailored administration of augmented immunosuppression is crucial hence the importance of continued biopsy and DSA screening in standard immunological risk patients. More research is necessary to determine if the MI score as a distinctive parameter for outcome holds in other renal transplant units and in comparison with other clinical and serological features. Acknowledgments The authors would like to acknowledge the European Renal Association European Dialysis and Transplant Association (ERA/EDTA) for the awarded long-term fellowship. We are grateful for support from the NIHR Biomedical Research Centre funding scheme. The authors acknowledge the work contributed by the transplant clinic staff, the Leslie Brent laboratory, the histopathology laboratory staff, and the histocompatibility and immunogenetics laboratory staff. Part of this material was presented at the British Transplant Society 15th Annual Congress February 2012 and at the American Transplant Congress June 2012. Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation. References 1. Haas M. The significance of C4d staining with minimal histologic abnormalities. Curr Opin Organ Transplant 2010; 15: 21 27. 2. Haririan A, Kiangkitiwan B, Kukuruga D, et al. The impact of C4d pattern and donor-specific antibody on graft survival in recipients requiring indication renal allograft biopsy. Am J Transplant 2009; 9: 2758 2767. 3. Worthington JE, McEwen A, McWilliam LJ, Picton ML, Martin S. Association between C4d staining in renal transplant biopsies, production of donor-specific HLA antibodies, and graft outcome. Transplantation 2007; 83: 398 403. 4. Gaston RS, Cecka JM, Kasiske BL, et al. Evidence for antibodymediated injury as a major determinant of late kidney allograft failure. Transplantation 2010; 90: 68 74. 5. Haas M. Pathologic features of antibody-mediated rejection in renal allografts: An expanding spectrum. Curr Opin Nephrol Hypertens 2012; 21: 264 271. 6. Sis B, Jhangri GS, Bunnag S, Allanach K, Kaplan B, Halloran PF. Endothelial gene expression in kidney transplants with alloantibody indicates antibody-mediated damage despite lack of C4d staining. Am J Transplant 2009; 9: 2312 2323. 7. Loupy A, Hill GS, Suberbielle C, et al. Significance of C4d banff scores in early protocol biopsies of kidney transplant recipients with preformed donor-specific antibodies (DSA). Am J Transplant 2011; 11: 56 65. 8. Hidalgo LG, Campbell PM, Sis B, et al. De novo donor-specific antibody at the time of kidney transplant biopsy associates with microvascular pathology and late graft failure. Am J Transplant 2009; 9: 2532 2541. 9. Terasaki PI. Humoral theory of transplantation. Am J Transplant 2003; 3: 665 673. 10. Hourmant M, Cesbron-Gautier A, Terasaki PI, et al. Frequency and clinical implications of development of donor-specific and nondonor-specific HLA antibodies after kidney transplantation. J Am Soc Nephrol 2005; 16: 2804 2812. 11. Wiebe C, Gibson IW, Blydt-Hansen TD, et al. Evolution and clinical pathologic correlations of de novo donor-specific hla antibody post kidney transplant. Am J Transplant 2012; 12: 1157 1167. 12. Willicombe M, Brookes P, Santos-Nunez E, et al. Outcome of patients with preformed donor-specific antibodies following alemtuzumab induction and tacrolimus monotherapy. Am J Transplant 2011; 11: 470 477. 13. Solez K, Colvin RB, Racusen LC, et al. Banff 07 classification of renal allograft pathology: Updates and future directions. Am J Transplant 2008; 8: 753 760. 14. Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713 723. 15. Zhang Q, Liang LW, Gjertson DW, et al. Development of posttransplant antidonor HLA antibodies is associated with acute humoral rejection and early graft dysfunction. Transplantation 2005; 79: 591 598. 16. Crespo M, Pascual M, Tolkoff-Rubin N, et al. Acute humoral rejection in renal allograft recipients: I. Incidence, serology and clinical characteristics. Transplantation 2001; 71: 652 658. 17. Mauiyyedi S, Pelle PD, Saidman SL, et al. Chronic humoral rejection: Identification of antibody-mediated chronic renal allograft rejection by C4d deposits in peritubular capillaries. J Am Soc Nephrol 2001; 12: 574 582. 18. Mao Q, Terasaki PI, Cai J, et al. Extremely high association between appearance of hla antibodies and failure of kidney grafts in a five-year longitudinal study. Am J Transplant 2007; 7: 864 871. 19. Terasaki PI, Ozawa M. Predicting kidney graft failure by HLA antibodies: A prospective trial. Am J Transplant 2004; 4: 438 443. 20. Einecke G, Sis B, Reeve J, et al. Antibody-mediated microcirculation injury is the major cause of late kidney transplant failure. Am J Transplant 2009; 9: 2520 2531. 21. Everly MJ, Rebellato L, Ozawa M, et al. Epidemiology of de novo donor specific anti-hla antibodies in a consecutive series of lowrisk renal transplant subjects with 5 10 years follow-up. ATC Congress 2012 Abstracts. Am J Transplant 2012; 12(Suppl S3): 122. 22. Cooper JE, Gralla J, Cagle L, Goldberg R, Chan L, Wiseman AC. Inferior kidney allograft outcomes in patients with de novo donorspecific antibodies are due to acute rejection episodes. Transplantation 2011; 91: 1103 1109. 23. Mengel M, Sis B, Haas M, et al. Banff 2011 Meeting Report: New concepts in antibody-mediated rejection. Am J Transplant 2012; 12: 563 70. 24. Loupy A, Suberbielle-Boissel C, Hill GS, et al. Outcome of subclinical antibody-mediated rejection in kidney transplant recipients with preformed donor-specific antibodies. Am J Transplant 2009; 9: 2561 70. 25. Sis B, Jhangri GS, Riopel J, et al. A new diagnostic algorithm for antibody-mediated microcirculation inflammation in kidney transplants. Am J Transplant 2012; 12: 1168 79. 26. Zhang Q, Reed EF. Non-MHC antigenic targets of the humoral immune response in transplantation. Curr Opin Immunol 2010; 22: 682 8. 492 American Journal of Transplantation 2013; 13: 485 492