European Society of Organ Transplantation (ESOT) Congress 2017 - Our abstracts

​Sunday September 24, 2017

1/ MOLECULAR CORRELATES OF ENDOTHELIAL MTOR ACTIVATION IN HEART TRANSPLANT RECIPIENTS
Session: “ Basic and translational immunology” (17:45 to 18:45)
Location: Room 118 + 119 – 18:29 

M Racapé, A Loupy, J Reeve, J Venner , R Guillemain, L Hidalgo, C Lefaucher, X Jouven, P Bruneval, J Duaong Van Huyen, P Halloran 
Background
The detection of phosphorylated effectors of the mTOR pathway such as phosphorylated-S6RP in endothelial cells by immunohistochemistry (IHC) has been associated with Antibody-Mediated allograft Rejection (AMR). The aim of this study was to evaluate the molecular phenotype related to the endothelial detection of pS6RP in heart transplant recipients.
Methods
This case-control study included 41 heart transplant patients from four French referral centers with biopsy proven antibody-mediated rejection (pAMR+) and a matched control group of 30 patients without rejection (pAMR0) based on the updated ISHLT classification. From these patients, 94 endomyocardial biopsies (EMB) had adequate material for microarray analysis and endothelial expression analysis of pS6RP by IHC. We also determined the allograft gene expression profile using the ABMR molecular score in addition to pathogenesis-based transcripts reflecting endothelial activation (DSAST and ENDAT), macrophage burden (QCMAT), gamma-interferon response (GRIT) and NK-cell burden (NKB) (http://atagc.med.ualberta.ca).
Results
Among the 94 EMBs included in the main analyses, 50 were pAMR+ (53.2%) and 44 (46.8%) were pAMR0 normal EMBs. Endothelial expression of pS6RP was observed in 27/50 (54%) of pAMR+ biopsies and 12 out of 44 normal biopsies (27.3%, Fischer's exact: p=0.012). As compared with biopsies without pS6RP labeling, biopsies with pS6RP staining showed increased expression of DSAST (Mann-Whitney: p<0.0001), ENDAT (p=0.0009), QCMAT (p=0.0046), NKB (p=0.0001), GRIT (p=0.0008) and increased ABMR molecular score reflecting AMR injury (p=0.0001).
Conclusion
Endothelial activation of mTOR pathway is associated with AMR and increased expression in transcripts reflecting endothelial activation, macrophage burden, microcirculation and NK burden. Our results suggest the importance of the mTOR pathway activation in AMR injury and the potential interest of using mTOR inhibitors in this setting.

Monday September 25, 2017

2/ SPECIFIC GENE EXPRESSION SIGNATURE OF COMPLEMENT-ACTIVATING DONOR SPECIFIC ANTI-HLA ANTIBODY-MEDIATED REJECTION IN KIDNEY ALLOGRAFTS
Session: “Best abstract challenge” (11:10 to 12:40)
Location: Room 117 – 12:30 

C Lefaucheur, D Viglietti, O Aubert, A Zeevi, P Halloran, A Loupy
Background
We investigated whether the complement-binding capacity of circulating donor-specific anti-HLA antibodies (DSA) is associated with specific gene expression signature in the kidney allograft.
Methods
We prospectively enrolled 931 kidney recipients transplanted between 2011 and 2014, with systematic screening for circulating DSA in the first year post-transplantation. We assessed DSA specificity, MFI, C1q-binding capacity and IgG1-4 subclasses using SAB. All patients underwent allograft biopsy at the time of post-transplant DSA detection to assess gene expression using microarray. We compared gene expression according to DSA C1q-binding status.
Results
We identified 157 (17%) patients with DSA, 44 (28%) with C1q-binding DSA, and 113 (72%) with non-C1q-binding DSA. Patients with C1q-binding DSA showed higher MFI levels (P<0.001) and greater prevalence of IgG1 (P<0.001) and IgG3 (P<0.001) subclasses than patients with non-C1q-binding DSAs. Among the 9954 inter-quartile range filtered transcripts that were most significantly expressed in the C1q-binding DSA patients, those most associated with C1q-binding DSAs were composed primarily of NK, endothelial, interferon gamma, macrophage, and effector T cell genes. We defined a discriminative gene set for C1q-binding DSA status (CXCL11, CCL4, MS4A6A, GBP1, MS4A7) that outperformed conventional histology to predict DSA C1q-binding status (AUC of 0.85 vs. 0.76, P=0.006). The 5-gene set was associated with the C1q-binding capacity of DSA independently of MFI level and IgG subclass composition. The integration of the gene set to conventional histology in unsupervised hierarchical clustering and principal component analysis allowed identifying a distinct pattern of allograft injury reflecting DSA C1q-binding capacity.
Conclusion
We identified a gene expression signature for kidney allograft injury related to the complement-binding capacity of circulating DSA that outperformed conventional histology.  


3/ CHARACTERISTICS AND OUTCOMES OF STREPTOCOCCUS PNEUMONIAE INVASIVE INFECTION IN LUNG TRANSPLANT RECIPIENTS
Session : « Improving lung transplant outcome » (11:10 to 12:40) 
Location : Room 120+121 – 11 :52 

C Picard, L Azuar-Beaumont, E Farfour, A Monem Hamid, G Trebbia, F Parquin, E Sage, A Roux
Background
Lung transplant recipients (LTR) have an increased risk of streptococcus pneumoniae (Sp) invasive infection (SPII).
Methods
We performed a retrospective analysis of any consecutive patient with at least one Sp isolate in bacteriological sample and/or urinary antigen in 538 LTR followed in a French center between 2010 and 2016. Patients with asymptomatic colonization, bronchitis and SPII are described.
Results
83 isolates were obtained in 60 patients (11%) after a median time of 7.5 months post LT (IQR, 3.6; 12.4). Among them, 21 patients remained asymptomatic, 12 patients had bronchitis and 27 patients had at least one episode of SPII, mostly pneumonia. Sp was isolated in the immediate post-operative course in 12 patients (20%). Strains were susceptible to penicillin G and A in respectively 68 and 89% of the cases. Patients with isolates had another potential factor of colonization in 48% of the cases, such as the presence of bronchial stent or tracheostomy (23%), current tobacco use (15%) and pulse steroids in the previous 2 months (15%). Biopsy proven acute cellular rejection occurred simultaneously in 20% the patients. SPII incidence was 1.72 cases/100 patient-years. 80% of the cases of SPII occurred in the first 3 years post LT. Co-infection with various respiratory virus was documented in 29% of the cases. Co-infection with chronic colonizing bacteria, such as Pseudomonas or staphylococcus was present in 35%. None of the patients with SPII had received previous combined pneumococcal vaccination and only 6 (22%) of them were under azithromycin at the time of SPII. 19% of the patients required ICU admission, due to shock. None of them died.
Conclusion
Colonization of the lung allograft with Sp may occur very early and is frequently associated with bronchial stent, tobacco exposure or previous pulse steroids. SPII occurs in half of the cases. As previously described in the immunocompetent host, viral co-infection may be critical in the occurrence of SPII. 


4/ IMPACT OF RESPIRATORY SYNCYTIAL VIRUS INFECTION ON LONG-TERM OUTCOMES IN LUNG TRANSPLANT RECIPIENTS
Session : « Improving lung transplant outcome » (11:10 to 12:40) 
Location : Room 120+121 – 11 :59 

Lafoeste H, Picard C, Azuar-Beaumont L, Farfour E, De Miranda S, Douvry B, Monem Hamid A, Carlier N, Grenet D, Parquin F, Sage E, Roux A
Background
Acute respiratory virus infections have been associated with lung function decline at time of infection in lung recipients (LTR). An increased risk of acute cellular rejection (ACR) and chronic lung allograft dysfunction (CLAD) has also been reported following viral infection. We report on the lung function at time of infection and 6 months after respiratory syncytial virus (RSV) infection.
Methods
We performed a retrospective analysis of all consecutive patients with at least one positive PCR for RSV on a respiratory sample in 439 LTR followed between 2013 and 2016. At the time of infection, decliners, defined as patients with a drop of FEV1 > 10% from baseline, were compared to non-decliners. At 6 months, the occurrence of ACR, anti-HLA DSA and the occurrence/evolution of bronchiolitis obliterans syndrome (BOS) were assessed.
Results
RSV infection occurred in 42 patients (9.7%; incidence 3.95/100 patient-years) at a median time post LT of 760 days (IQR, 272; 1432). 30 (71%) were decliners and 12 (29%) non-decliners. Decliners had more frequently a past history of similar episode than non-decliners (respectively 16/30, vs 2/12; p=0.036) and had more frequently pre-existing BOS stage 1 or more (8/30, vs 0/12; p=0.08). The diagnosis, the immunosuppression, co infection with bacteria or other viruses and concomitant ACR did not differ between the two groups. At 6 months, ACR occurred in 1 patient in each group (p=0.48), and de novo or aggravating BOS stage 0p or more occurred equally in both groups (10/30, vs 4/12; p=0.99). 7/30 (23%) of the decliners had de novo or increasing DSA (vs 1 non-decliners; p=0.4), and 2 of them experienced antibody-mediated rejection in the following 6 months.
Conclusion
RSV infection in LTR is frequently associated with an acute drop in FEV1, especially in case of pre-existing BOS and past history of similar episodes. De novo or aggravating BOS occurs in 1/3 of patients in the following 6 months independently of the occurrence of decline at the time of infection.


5/ DONOR SPECIFIC HLA ANTIBODY-MEDIATED COMPLEMENT ACTIVATION IS A SIGNIFICANT INDICATOR OF ANTIBODY- MEDIATED REJECTION AND POOR LONG-TERM GRAFT OUTCOME DURING LUNG TRANSPLANTATION
Session : « Improving lung transplant outcome » (11:10 to 12:40) 
Location : Room 120+121 – 12 :06 

Roux A, Thomas K, Suberbielle-Boissel C, Sage E, Azuar-Beaumont L, Parquin F, Monem Hamid A, Reed E
Background
It has been demonstrated in heart and kidney transplantation that anti-HLA DSA mediate damage to the allograft via multiple mechanisms, including complement-dependent and independent actions. The transplant community is now in pursuit of new tools or technologies that will allow for stratification of DSA+ patients for antibody mediated rejection (AMR). C3d assay allow direct measure of DSA activation of complement.
Methods
According to routine DSA detection strategy (Single Antigen One Lambda), 105 patients within our Foch Lung transplant cohort (sept 2008-dec 2013) were considered as DSA positive patients and further tested using the Immucor® LSA Lifecodes bead-based assays for single antigen and C3d testing. AMR diagnosis associated graft failure and DSA presence and histological lesion compatible with AMR and/or C4d positive staining. DSA positivity was defined by BCM>500 (Background corrected MFI). C3d ratio was calculated as BCM of patients beads/BCM of negative control
Results
Among 105 patients, 25 were considered DSA+AMR- and 15 DSA+AMR+. Comparison of donor specific beads BCM and donor specific beads C3d ratio between DSA+AMR+ and DSA+AMR- patients show a significantly higher value in DSA+AMR+ patients. BCM and C3d ratio of donor specific beads show strong correlation in DSA+AMR+ patients (R=0.6). Within DSA+AMR- group the correlation was weaker (R=0.09). Within our study population, immunodominant DSA C3d ratio>4 was significantly associated with graft loss.
Conclusion
DSA capacity to bind C3d is clearly associated with AMR and graft loss in lung transplantation. These results need to be validated in larger multicentric sutdy.


6/ INFLAMMATION IN SCARED AREAS IN THE KIDNEY ALLOGRAFT: MAJOR IMPACT OF T CELL-MEDIATED REJECTION AND UNDER-IMMUNOSUPPRESSION
Session : « Kidney transplant pathology» (16:00 to 17:00) 
Location : Room 122 – 16 :20 

C Gosset, D Viglietti, M Rabant, A Loupy, C Lefaucheur
The important role of inflammation in scared areas in the kidney allograft (i-IF/TA) on transplant outcomes has been acknowledged by the 2015 Banff report, which has highlighted the need to clarify its significance. We investigated the determinants of i-IF/TA in a population-based study.
This observational, prospective cohort study included 1539 consecutive kidney recipients transplanted at Paris between 2004 and 2010, with systematic assessment of inflammation in allograft scarred area using the i-IF/TA Banff score on biopsies performed at 1 year post-transplantation. We considered donor and recipient baseline characteristics, transplant characteristics, immunosuppressive regimens, immunological parameters (HLA matching, anti-HLA DSA), infectious diseases (CMV, pyelonephritis, BK virus) and all the biopsy-provren diagnoses (TCMR, AMR, BK virus-associated nephropathy, CNI toxicity, acute tubular necrosis), recorded at the time of transplantation and in the first year after transplantation.
We identified 893 patients with IF/TA at 1 year post-transplant, among whom 194 (22%) showed severe i-IF/TA (≥2). The independent determinants for severe i-IF/TA were TCMR (OR, 2.7; 95%CI, 1.7-4.1), BK virus-associated nephropathy (OR, 3.5; 95%CI, 1.5-8.2), recipient age (OR, 0.9 95%CI, 0.9-0.9), CNI therapy (OR, 0.5; 95%CI, 0.2-0.9) and steroid therapy (OR, 0.5; 95%CI, 0.2-0.9). Patients with i-IF/TA showed a specific injury phenotype including increased inflammation in non-scarred areas (P<0.001), total inflammation (P<0.001) and tubulitis (P<0.001). Patients with persisting severe i-IF/TA after TCMR treatment showed an accelerated progression of fibrosis over time (P=0.01) and a decreased allograft survival (P=0.003) as compared to those without persisting inflammation after TCMR treatment.
Inflammation in scared areas of the kidney allograft is mainly determined by previous TCMR and the level of immunosuppression, and is associated with poor kidney transplant outcomes. 


7/ DECIPHERING THE SPECIFIC CAUSES OF KIDNEY ALLOGRAFT LOSS. A POPULATION BASED STUDY
Session : « Outcomes in kidney transplantation» (16 :00 to 17 :00) 
Location : Room 116 – 16 :28 

C Loheac, O Aubert, N Kamar, JP Duong Van Huyen, M Delahousse, D Glotz, C Legendre, C Lefaucheur, A Loupy. 
Background: 
Understanding the specific causes of kidney allograft loss are mandatory to improve the longevity of allografts. However, the current literature is limited by the low level of phenotyping, small series with selected populations or data from registries that lack precision diagnoses. 
Methods: 
We conducted a multicentric prospective study including unselected kidney transplant recipients from 4 referral centers transplanted between 2004 and 2014. Donor, recipient, transplant parameters, clinical and biological post-transplant parameters as well as circulating anti-HLA DSA and allograft biopsies performed post-transplant were included. The main outcome was long-term kidney allograft survival and specific causes of graft loss.
Results: 
4,783 kidney recipients were included and 9,959 kidney allograft biopsies were analyzed. During a median follow-up post-transplant of 4.51 years, 732 graft losses occurred. After inclusion of biopsy, clinical, biological, and anti-HLA DSA data, a primary cause of allograft loss was identified in 95% of cases. The causes were: immune related (31.7% of antibody-mediated rejection (ABMR), 4.8% of T-cell mediated rejection), surgical related (25.6% of thrombosis, 2,5% of urinary disease), medical related (tumoral, infectious or cardiac intercurrent disease (14,6%)), recurrence of primary disease (7.1%), virus related (BK or CMV associated nephropathy, 4.78%) and calcineurin inhibitor nephrotoxicity related (1.1%). ABMR was associated with the worse allograft survival (55% at 9 years). 
Conclusion: 
In this multicentric extensively phenotyped population of kidney transplant recipients, we identify the contemporary picture of specific causes of kidney allograft loss. Such effort highlights the current priorities to detect such complications to improve long-term allograft outcomes.   


8/ A MULTIDIMENSIONAL PROGNOSTIC SCORE AND NOMOGRAM TO PREDICT KIDNEY TRANSPLANT SURVIVAL: THE INTEGRATIVE BOX (IBOX) SYSTEM
Session : « Outcomes in kidney transplantation» (16 :00 to 17 :00) 
Location : Room 116 – 16 :49 

A Loupy, O Aubert, B Orandi, A Jackson, M Naessens, N Kamar, O Thaunat, E Morelon, M Delahousse, D Viglietti, C Legendre, D Glotz, X Jouven, RA Montgomery, M Stegall, DL Segev, C Lefaucheur.
Background : 
The transplant field currently lacks of robust risk stratification models. The goal of this study was to generate an integrative scoring system that predicts kidney allograft loss.   
Methods : 
This international study involved 8 European and 2 North American centers including 4,806 kidney recipients transplanted between 2002 to 2014 and integrated the full spectrum of allograft parameters. The development cohort included 1540 patients. The prognostic ability of 80 parameters was evaluated using univariate and multivariate Cox regression analyses. Performance assessment and internal validation of the final model were done with Harrell’s C-index, calibration plot and bootstrap sample procedures. Using the final model, a prognostic nomogram and a score were developed.
Results : 
Expanded criteria donor, allograft scarring (IFTA), allograft injury (g and ptc), allograft function (eGFR and proteinuria at 1 year post transplant), anti-HLA DSA status and level (MFI) were independently associated with 7-year post-transplant allograft survival. The model exhibited good calibration, excellent discrimination (C-statistic=0.80; 95%CI=0.76-0.84) and a nomogram for individual graft survival prediction was generated. Time dependent ROC curves, Random survival forests survival analyzes confirmed the robustness of the iBox score. The performance of the score was confirmed in the external validation cohort from Europe (n=2130, C-statistic=0.81; 95%CI=0.76-0.86) and North America (n=1136, C-statistic=0.83; 95%CI=0.77-0.89). The capacity of the score to predict graft failure remained high in 1118 patients evaluated beyond 1-year post transplant (C-statistic=0.80; 95%CI=0.76-0.85).
Conclusion : 
The iBox prognostic score and nomogram precisely predict the individual long-term graft survival probability. The score demonstrates high performance, exportability across centers worldwide and adaptability beyond the first year post transplant evaluation. 


9/ IDENTIFYING THE SPECIFIC CAUSES AND THE DETERMINANTS OF OUTCOME IN KIDNEY RECIPIENTS WITH TRANSPLANT GLOMERULOPATHY: A MULTICENTER STUDY.
Session : « Kidney transplant pathhology» (16 :00 to 17 :00) 
Location : Room 122 – 16 :28 

O Aubert, S Higgins, P Campbell, C Loheac, D Viglietti, D Glotz, Ch Legendre,B Sis, C Lefaucheur, JP Duong and A Loupy.
Background: 
Understanding the specific causes of Transplant Glomerulopathy (TG) and its long-term consequences is lacking.
Methods: 
This study includes all kidney allograft biopsies performed between 2004 and 2014 in three French referral centers and one Canadian center showing TG. All TG cases were extensively phenotyped and systematically assessed using light microscopy, immunohistochemistry, immunofluorescence, together with circulating anti-HLA-DSA at the time of biopsy. 
Results: 
Among the 8207 post-transplant allograft biopsies performed during the inclusion period, 559 (6.8%) had double contours and corresponded to 392 patients. Three overlapping etiologies accounted for 467 (84%) cases. 417 biopsies showed alloantibodies-mediated injury (75%), 91 biopsies showed TMA (16%), 65 showed MPGN (12%), while 92 cases (17%) remained equivocal with no specific lesions identified by the pathologist (Figure 1). The median time to first cg lesion occurrence after transplantation was 32.7 months (IQR: 12.0–77.9). Kidney allograft survival after TG diagnosis was 59% at 5 years and 41% at 8 years (Figure 2). After adjusting for donor, recipient and transplant characteristics, immunological and histological parameters, we identified the following independent factors associated with long-term allograft survival in patients with TG: eGFR (HR:0.96; p<0.0001) and proteinuria level (HR:2.17; p<0.0001) at the time of biopsy, deceased donor (HR:1.64; p=0.0139), delay between transplantation and TG diagnosis (HR:1.34p<0.0001) and endarteritis Banff scores (HR:1.82; p=0.0018).
Conclusion: 
Using a large cohort of kidney recipients with a diagnosis of TG, we identify three overlapping pathways in TG: ABMR, TMA and MPGN. The identification of the main independent determinants of TG prognosis may help improving risk stratification and define specific causes and disease process in patients with TG.


10/ Accelerated Kidney Allograft Arteriosclerosis: Major Role of Donor-Specific Anti-HLA Antibody Strength And Complement-Activating Capacity
Session : « ABMR in kidney transplantation» (17:05 to 18:05) 
Location : Room 122 – 17:33 

D. Viglietti, A. Loupy, O. Aubert, A. Zeevi, C. Lefaucheur
Background: 
The role of circulating anti-HLA DSAs in the development of accelerated arteriosclerosis has been recently reported in kidney recipients. This study investigated the characteristics of DSA that are associated with the severity of allograft arteriosclerosis.
Methods: 
We included 744 kidney transplantation performed between 2004 and 2010 at Paris, France, with systematic assessment of injury phenotype and arteriosclerotic lesions using the vascular fibrous intimal thickening (cv) Banff score on allograft biopsies performed at one year after transplantation. We assessed circulating DSAs and their characteristics (specificity, HLA class, mean fluorescence intensity [MFI] and C1q-binding) at 6 months after transplantation.
Results: 
We identified 281 patients with cv0 score, 213 patients with cv1 score, 189 patients with cv2 score and 61 patients with cv3 score. The distribution of DSAs according to cv score was: 47/281 (17%) in cv0 patients, 39/213 (18%) in cv1 patients, 63/189 (33%) in cv2 patients and 28/61 (46%) in cv3 patients. DSA MFI level was positively correlated with the severity of arteriosclerosis (rho=0.23, p=0.002), with a mean MFI of 3204.0±3725.2 in cv0 patients, 3760±3598 in cv1 patients, 4892±4676 in cv2 patients and 5541±3892 in cv3 patients. C1q-binding DSA prevalence increased with the severity of allograft arteriosclerosis: 8/281 (3%) in cv0 patients, 6/213 (3%) in cv1 patients, 25/189 (13%) in cv2 patients and 9/61 (15%) in cv3 patients (p<0.001). Patients with C1q-binding DSA had a higher cv score compared with patients with non-C1q-binding DSA (1.7±1.0 versus 1.3±1.1, respectively, p=0.01). C1q-binding DSAs were associated with increased microvascular inflammation (p<0.001) and C4d deposition in peritubular capillaries or arteries (p<0.001).
Conclusion: 
There is a biological gradient between DSA level and the severity of allograft arteriosclerosis. Complement-activating DSAs are associated with an increased severity of arteriosclerosis and complement deposition in allograft.   


11/ GENE EXPRESSION SIGNATURE OF SUBCLINICAL ALLOGRAFT KIDNEY ABMR COMPARED TO CLINICAL ABMR.  
Session : « ABMR in kidney transplantation» (17:05 to 18:05) 
Location : Room 122 – 17:47 

O Aubert, C Lefaucheur, S Higgins, L Hidalgo, J-P Duong van Huyen, D Viglietti, X Jouven, D Glotz, C Legendre, P Halloran and A Loupy.
Background: 
Antibody-mediated rejection occurs in clinically stable patients (sABMR) or in patients with allograft dysfunction (ABMR), but only little is known about the similitudes or discrepancies between the molecular landscape of sABMR and ABMR. 
Methods: 
We used an integrative analysis strategy comprising a systematic assessment of clinical-biological parameters, transplant characteristics, histopathology, immunohistochemistry, type of treatment, circulating anti-HLA DSA assessment and gene expression.
Results: 
Among the 131 patients with ABMR, 99 had a clinical ABMR while 32 were sABMR. Clinical ABMR displayed a worse kidney allograft function (mean eGFR: 34.45 vs 48.24 mL/min/1.73m2, p<0.001) and an increased proteinuria rate (mean: 0.54 vs 0.31 g/g creatinine, p=0.041) compared to sABMR. There was no difference for the highest anti-HLA DSA MFI at the time of ABMR between the two groups (Clinical ABMR:median MFI: 2779 (IQR: 1028-7573); sABMR: median MFI: 1842 (IQR: 1191-5391); p=0.356). Allograft gene expression showed that patients with clinical ABMR exhibited significantly more injury-repair response transcripts (p<0.001), more macrophage associated transcripts (p=0.003) and a trend towards more IFNG production and inducing transcripts (p=0.066). We compared the top 10 ABMR related transcripts with their corresponding fold change and p value (t-test) compared the sABMR. Five transcripts were AKI associated (LCN2, LTF, SERPINA3, SLPI and PTX3) and two were macrophages associated (CD163 and MSR1). A principal component analysis integrating the histological and molecular parameters identified a distinct histo-molecular allograft rejection phenotype in patients with clinical ABMR vs sABMR.
Conclusion: 
cABMR is associated with a distinct histo-molecular phenotype of kidney allograft rejection mainly driven by AKI and macrophages burden compared to sABMR, which could explain the difference in terms of allograft survival. 


12/ CIRCULATING DONOR-SPECIFIC ANTI-HLA ANTIBODIES ACCELERATE THE PROGRESSION OF INTERSTITIAL FIBROSIS IN KIDNEY ALLOGRAFTS
Session : « ABMR in kidney transplantation» (17:05 to 18:05) 
Location : Room 122 – 17:54 

C Gosset, D Viglietti, M Rabant, A Loupy, C Lefaucheur
Background: 
Addressing the causes of accelerated ageing of kidney allografts represents an important challenge to improve long-term transplant outcomes. We investigated the role of donor-specific anti-HLA antibodies (DSA) in the progression of kidney allograft interstitial fibrosis.
Methods: 
We prospectively enrolled 913 kidney recipients transplanted between 2004 and 2010. All patients were assessed for allograft interstitial fibrosis on biopsies performed at Day 0 and at 1 year after transplantation using the IF/TA Banff grade. We also integrated all the “for cause” biopsies performed in the first year post-transplant (N=1035) and after the first year (N=784, median time of biopsies 18.4 months; IQR, 13.3-40.4). All patients were screened for DSA by SAB at the time of transplantation (Day 0) and within the first year post-transplantation. The progression of IF/TA within the first year post-transplantation was evaluated by the difference between the 1-year and Day-0 IF/TA grades (∆IF/TA). The progression of IF/TA over the long term was modelled using mixed-effect models.
Results: 
The distribution of IF/TA on pre-implantatory biopsies (N=913) was: 726 (80%) IF/TA0, 145 (15%) IF/TA1, 36 (4%) IF/TA2 and 6 (1%) IF/TA3 as compared to 325 (35%), 263 (29%), 173 (19%), and 152 (17%) on 1-year biopsies (N=913) (P<0.001). Over the first year, 507 (56%) patients presented progression of IF/TA (∆IF/TA>0). Patients with Day-0 DSA (N=198) showed increased progression of fibrosis within the first year (∆IF/TA: 1.08±1.15) as compared to those without Day-0 DSA (N=715, 0.86±1.12) (P=0.016). Patients with post-transplant DSA (preformed or de novo) (N=236) exhibited accelerated progression of IF/TA as compared to patients without post-transplant DSA (N=677) (P=0.0078) when integrating the biopsies performed at 1-year post-transplant and beyond.
Conclusion: 
Pre-transplant circulating DSA increase premature allograft fibrosis and post-transplant DSA accelerate the progression of allograft fibrosis over the long term.


13/ DETERMINANTS OF SEVERE FIBROSIS IN KIDNEY ALLOGRAFTS: MAJOR IMPACT OF CIRCULATING DONOR-SPECIFIC ANTI-HLA ANTIBODIES
Session : « ABMR … again !» (17:05 to 18:05) 
Location : Room 124 – 17:53 

C Gosset, D Viglietti, M Rabant, A Loupy, C Lefaucheur
Background: 
We investigated the independent contribution of circulating donor-specific anti-HLA antibodies (DSA) in the development of severe kidney allograft fibrosis with integration of traditional risk factors for allograft fibrosis.
Methods: 
We prospectively enrolled 1539 consecutive kidney recipients transplanted between 2004 and 2010, with systematic assessment of allograft fibrosis using the IF/TA Banff score on biopsies performed at 1-year post-transplantation. We considered all of the traditional determinants of allograft fibrosis reported in the literature, recorded at the time of transplantation and in the first year after transplantation. We also integrated DSA assessment and all the histologic diagnoses (“for cause” biopsies; N=1804) performed in the first year after transplantation.
Results: 
We identified 498 (32%) patients with severe IF/TA (Banff grade≥2). DSA were associated with severe IF/TA at 1-year post transplant (adjusted OR, 1.53; 95%CI, 1.16-2.01; P=0.002), independently of the traditional determinants, including: T cell-mediated rejection, antibody-mediated rejection, BK virus-associated nephropathy, calcineurin inhibitor toxicity, initial disease recurrence, pyelonephritis, acute tubular necrosis, donor and recipient baseline parameters, and transplant characteristics. DSA remained associated with severe IF/TA even in patients without episode of antibody-mediated rejection (OR, 1.47; 95%CI, 1.10-1.96; P=0.008). Among the modifiable risk factors for severe IF/TA, DSA were found to be the first contributor, being involved in 11% of cases while T cell-mediated rejection, calcineurin inhibitor toxicity, acute tubular necrosis, pyelonephritis and BK virus-associated nephropathy were involved in 9%, 8%, 6%, 5%, and 4% of cases, respectively.
Conclusion: 
Circulating DSA are major contributor to severe allograft interstitial fibrosis independent of traditional risk factors and of antibody-mediated rejection. 


14/ SECOND-LINE THERAPY AFTER STANDARD OF CARE IN ANTIBODY-MEDIATED REJECTION: A PROSPECTIVE STUDY
Session : « ABMR … again !» (17:05 to 18:05) 
Location : Room 124 – 17:57 

D Viglietti, A Loupy, O Aubert, C Legendre, D Glotz, C Lefaucheur
Background: 
There is a wide heterogeneity in antibody-mediated rejection (AMR) patients’ prognosis after standard of care (SOC) including plasma exchange (PE) and intravenous immunoglobulin (IVIG). We investigated whether a composite prognostic score might identify patients at poor prognosis after AMR SOC therapy eligible for second-line intervention.
Methods: 
We prospectively included kidney recipients diagnosed with active AMR (2012-2014) who received standardized SOC therapy (PE x4 and IVIG 2 g/kg repeated every 3 weeks x3). Patients were stratified according to their risk of graft loss after SOC based on an AMR prognostic score. High-risk patients received second-line treatment with complement-targeting agent (C5-inhibitor Eculizumab or C1-inhibitor Berinert) and high-dose IVIG for 6 months. The prognostic score was built in a prospective cohort of 284 kidney recipients with biopsy-proven active AMR receiving the standardized SOC therapy (abstract N° 3720943).
Results: 
We enrolled 83 kidney recipients with active AMR diagnosed at a median time of 4.3 months post-transplantation. Patients received the SOC therapy, after which they were stratified in 3 risk groups defined according to the prognostic score. 15 (18%), 11 (13%) and 57 (69%) patients were stratified in the high-risk, intermediate-risk and low-risk groups, respectively. Predicted 3-year graft survivals after AMR were 94% (95%CI, 89-96), 64% (95%CI, 43-79) and 32% (95%CI, 17-48), respectively. The characteristics of high-risk patients were: GFR of 27.4±8.9 mL/min, g+ptc Banff score of 4.1±1.2 and DSA MFImax of 14482±485 after SOC therapy, and 37.1±10.5 mL/min, 3.5±1.4 and 10319±646 before SOC therapy. High-risk patients receiving second-line therapy showed a 3-year graft survival of 84%.
Conclusion: 
Risk stratification for kidney graft loss by a composite prognostic score based on clinical, histological and immunological parameters allows to identify high-risk patients after SOC treatment of AMR who may benefit from second-line strategies.

Tuesday September 26, 2017

15/ COMPLEMENT-ACTIVATING ANTI-HLA ANTIBODIES: IDENTIFICATION OF SPECIFIC HISTO-MOLECULAR PHENOTYPE OF REJECTION FOR COMPLEMENT-TARGETING THERAPY
Session : « ABMR : from diagnosis to treatment » (11:10 to 12:40) 
Location : Room 123 – 11:31 

C Lefaucheur, B Sis, D Viglietti, O Aubert, D Glotz, C Legendre, A Zeevi, P Halloran, A Loupy
Background: 
We investigated whether circulating complement-binding donor-specific anti-HLA antibodies (DSA) induce specific rejection phenotype and influence response to complement-targeting treatment.
Methods: 
We prospectively enrolled 931 kidney recipients transplanted between 2011 and 2014, with systematic screening for circulating DSA in the first year post-transplantation. All patients underwent allograft biopsy at the time post-transplant DSA detection. The allograft rejection phenotypes were assessed by histopathology, immunochemistry and allograft gene expression (microarray). A model of fully MHC-mismatched male CBA (H-2k) kidneys transplanted into B6.RAG1-/- (H-2b) immunodeficient mice with adoptive transfer of complement and non-complement-activating DSA was studied. The effect of complement inhibition therapy (Eculizumab) on allograft injury phenotype was assessed in two prospective studies (N=116).
Results: 
The histo-molecular phenotype of C1q-binding DSA allograft rejection (N=44) was characterized by increased microvascular infiltration by NK cells (P<0.001), monocyte/macrophages (P<0.001), greater prevalence of complement deposition (P<0.001), and selective changes in gene expression including interferon-gamma and endothelial activation (CXCL11, CCL4, MS4A6A, MS4A7, GBP1; P<0.01) as compared with patients with non-C1q-binding DSA (N=113). This phenotype was distinct from that of patients with non-C1q-binding DSA and without DSA in unsupervised principal component analysis. Mice receiving complement-binding DSA reproduced the human complement-activating antibody-mediated histo-molecular rejection phenotype. Eculizumab specifically abrogated the histo-molecular phenotype induced by C1q-binding DSAs and showed no effect on allograft injury in patients with non-C1q-binding DSA.
Conclusion: 
Circulating complement-binding DSA induce a specific histo-molecular phenotype of kidney allograft rejection that can be reversed by complement inhibition.  


16/ COMPOSITE PROGNOSTIC SCORE IMPROVES CLINICAL BENEFIT IN KIDNEY RECIPIENTS RECEIVING STANDARD OF CARE THERAPY FOR ANTIBODY-MEDIATED REJECTION
Session : « ABMR : from diagnosis to treatment » (11:10 to 12:40) 
Location : Room 123 – 12:13 

D Viglietti, A Loupy, O Aubert, C Legendre, D Glotz, C Lefaucheur
Background: 
There is a substantial heterogeneity in antibody-mediated rejection (AMR) patients’ prognosis after standard of care (SOC) including plasma exchange (PE) and intravenous immunoglobulin (IVIG). We investigated whether the use of a prognostic score in kidney recipients receiving AMR SOC therapy improves clinical-decision making.
Methods: 
We prospectively enrolled 2666 kidney recipients transplanted between 2004 and 2012 and included those diagnosed with active AMR who received standardized treatment with PE (x4) and high-dose IVIG (2 g/kg every 3 weeks x3). Patients were assessed at diagnosis and 3 months post-treatment for clinical and histological characteristics, and donor-specific anti-HLA antibodies (DSA) by SAB. An AMR prognostic score was derived from the multivariate Cox model for allograft loss. The net clinical benefit of the prognostic score was assessed by decision curve analysis.
Results: 
We included 284 patients with biopsy-proven active AMR who received SOC treatment. The independent predictors of graft loss were: GFR (HR, 0.93; 95%CI, 0.90-0.95; P<0.001) and presence of IF/TA (HR, 2.44; 1.36-4.37; P=0.003) at AMR diagnosis, and change in GFR (HR, 0.24; 95%CI, 0.16-0.35; P<0.001), ptc Banff score (HR, 1.50; 95%CI, 1.16-1.93; P=0.002) and DSA level after treatment (HR,1.30; 95%CI, 1.11-1.52; P=0.001). The AMR prognostic score showed good discrimination (C-statistic, 0.84). Decision-making after AMR SOC based on the prognostic score provided greater net clinical benefit than considering patients on the same risk level. The initiation of a second-line therapy based on the prognostic score would lead to treat 11 patients who will lose their graft in the absence of clinical intervention per 100 patients receiving SOC while not treating patients who will not lose their graft.
Conclusion: 
The use of an accurate composite prognostic score based on clinical, histological and immunological parameters in kidney recipients receiving SOC therapy for AMR improved clinical decision-making.  


17/ LONG-TERM OUTCOME OF ANTIBODY-MEDIATED REJECTION DUE TO PRE-EXISTING VS. DE NOVO DSA IN KIDNEY ALLOGRAFT RECIPIENTS.
Session : « ABMR : from diagnosis to treatment » (11:10 to 12:40) 
Location : Room 123 – 12:27

O Aubert, C Lefaucheur, S Higgins, L Hidalgo, J-P Duong van Huyen, D Viglietti, X Jouven, D Glotz, C Legendre, P Halloran and A Loupy.
Background: 
Antibody-mediated rejection (ABMR) can occur in patients with pre-existing anti-HLA donor-specific antibodies (DSA) or in patients who develop de novo DSA. However, how these processes compare in terms of kidney allograft outcome has not been addressed. 
Methods: 
From a cohort of 771 kidney biopsies, we included all patients with a diagnosis of ABMR. We used an integrative analysis strategy comprising a systematic assessment of clinical-biological parameters, transplant characteristics, histopathology, immunohistochemistry, type of treatment and circulating DSA assessment.
Results: 
Among the 205 patients with ABMR, 103 were related to pre-existing DSA while 102 were related to de novo DSA. ABMR due to de novo DSA displayed increased proteinuria and transplant glomerulopathy lesions, lower glomerulitis, but similar peritubular-capillaritis Banff scores and C4d-deposition than patients with ABMR due to pre-existing DSA. Graft survival was superior in patients with pre-existing DSA ABMR compared to patients with de novo DSA ABMR (graft survival at 8 years post ABMR of 63% vs 35% respectively, p<0.001). We identified de novo DSA ABMR (HR=1.82 compared to pre-existing DSA ABMR); low (<30mL/min/1.73m2) eGFR at diagnosis (HR=3.27; p<0.001); ≥0.30g/g proteinuria/creatinine ratio (HR=2.44; p<0.001); and presence of cg-lesions (HR=2.25; p=0.002) as the main determinants of allograft loss independently of type of treatment, time of diagnosis and degree of allograft injury and atrophy scarring at the time of diagnosis.
Conclusion: 
We found that these diseases have distinct prognosis with an acceptable and superior allograft survival in patients with preexisting/persisting DSA related ABMR. This supports the transplantation of highly-sensitized patients but also encouraging efforts to monitor patients for de novo DSA and avoidance of minimization strategies. 


18/ ABSENCE OF INDEPENDENT AND ADDITIONAL PREDICTIVE ABILITY OF PREIMPLANTATION KIDNEY ALLOGRAFT BIOPSIES FOR LONG-TERM OUTCOME: POPULATION BASED STUDY
Session : « Deceased donor risk and kidney allocation» (17:05 to 18:35) 
Location : Room 116 – 17:47

O Aubert, D Viglietti, C Loheac, M Rabant, C Gosset, J-P Duong van Huyen, D Glotz, C Legendre, C Lefaucheur and A Loupy.
Background: 
A significant number of kidneys are discarded worldwide. The mean cause is the result of the preimplantation biopsy without clear evidence that its results are associated with long-term allograft survival. 
Methods: 
We included patients who underwent kidney transplantations from a deceased donor in 2 French referral centers between 2004 and 2011 where preimplantation are routinely performed. All the patients with preimplantation biopsy were included. A systematic assessment of donor, recipient, and transplant clinical characteristics, a preimplantation biopsy and an evaluation of baseline circulating donor-specific anti-HLA antibody (DSA) levels were performed. 
Results: 
A total of 882 patients were included in the study. A total of 352/882 (40%) transplantations were performed using ECD kidneys and a total of 143/882 (16%) had an anti-HLA DSA at the day of transplantation. The mean follow-up time after transplantation was 6.56 ± 2.37 years. After adjusting for donor, recipient, and transplant characteristics as well as for preimplantation biopsy findings (including the atrophy-fibrosis, percentage of sclerotic glomeruli, arteriosclerosis and arteriolar hyalinosis scores) and baseline immunological parameters, we identified the KDRI score (HR=2.17; p=0.002) and the presence of circulating anti-HLA DSA on the day of transplantation (HR=2.89p<0.0001) as the main independent determinants of long-term allograft loss. None of the preimplantation biopsy findings showed independent association with the kidney allograft survival. 
Conclusion: 
Preimplantation biopsy assessment does not provide independent and additional predictive ability for long-term allograft outcome at a population level in deceased donor program. The current practice of discarding kidneys based on preimplantation biopsy findings may not be optimal for decision-making and is a barrier to the decrease in the rate of discarded kidneys.  


19/ GENE EXPRESSION PROFILING FOR THE IDENTIFICATION AND CLASSIFICATION OF ANTIBODY-MEDIATED HEART REJECTION 
Session : « Challenges in heart transplantation» (17:05 to 18:35) 
Location : Room 124 – 17:05

A Loupy, JP Duong Van Huyen, L Hidalgo, J Reeve, M Racapé, O Aubert, D Viglietti, P Bruneval, C Lefaucheur, PF Halloran 
Background: 
The specific effects of anti-HLA antibodies on heart allograft injury have not been addressed at a population level.
Methods: 
We prospectively monitored 617 heart recipients referred from 4 French heart transplant centers (2006 - 2011) for antibody-mediated rejection (AMR). We compared AMR patients (n=50) to a matched control group of 50 patients without AMR. We characterized all patients using histopathology (ISHLT 2013), immunostaining, gene expression assessments of allografts (microarray) and circulating anti-HLA DSA at the time of biopsy. We studied an external validation cohort of 98 heart recipients transplanted in Edmonton (Canada) including 27 pAMR cases and 71 controls.
Results: 
240 heart transplant EMB were assessed. AMR showed a distinct pattern of injury characterized by microcirculation inflammation by monocytes/macrophages and NK cells, and very selective changes in endothelial/angiogenesis and NK cell transcripts. The AMR selective gene sets discriminated patients with AMR from those without and included NK (AUC=0.87), endothelial activation (AUC=0.80), macrophage (AUC=0.86) and IFNG transcripts (AUC=0.84, p<0.0001 for all comparisons). These 4 gene sets showed increased expression with increasing AMR ISHLT grades (p<0.001), association with DSA levels and chronic allograft vasculopathy. The unsupervised PCA analysis demonstrated a high proportion of molecular inactive pAMRI+ compared with a significant molecular overlap between pAMR1H+ and pAMR2-3 reclassifying 25% of AMR cases. The molecular architecture and selective AMR transcripts, and the gene set discrimination capacity for AMR were highly conserved in the external validation cohort.
Conclusion: 
Antibody-mediated heart rejection is mainly driven by NK burden, endothelial activation, macrophage burden and IFNG effects. Molecular intragraft measurements for these specific pathogenesis-based transcripts classify AMR with great accuracy and correlate with the degree of injury and disease activity. 


20/ A NEW DIAGNOSTIC SCORE OF INTRAVASCULAR ACTIVATED MONONUCLEAR CELLS IN ANTIBODY-MEDIATED REJECTION IN HEART TRANSPLANTATION 
Session : « Challenges in heart transplantation» (17:05 to 18:35) 
Location : Room 124 – 17:12

A Loupy, X Jouven, M Racapé 
Background
The ISHLT classification defines Intravascular Activated Mononuclear Cells (IAMC) as one of the histopathologic features of Antibody-Mediated Rejection (pAMR) in heart transplantation. However, no accurate grading of IAMC correlating with pAMR diagnosis has been proposed. The aim of this study was to develop a score to grade the extent and the pattern of the IAMC in endomyocardial biopsies (EMB) with AMR.
Methods
This case-control study included heart transplant patients from five French referral centers with biopsy-proven AMR (pAMR1-3) (n=64) and a matched control group of 44 patients without rejection (pAMR0). IAMC on EMBs was graded blind of pAMR grades by two skilled pathologists according to the percentage of the area with IAMC in capillaries and to the maximum number of IAMC in the most affected capillary on a 0 to 3 scale and a positivity defined by a grade ≥ 1. The score was compared to the gene expression profile in EMBs by microarray using the ABMR molecular score and pathogenesis-based transcripts reflecting endothelial activation (ENDAT), DSA (DSAST), macrophage burden (QCMAT), gamma-interferon response (GRIT) and NK-cell burden (NKB) (http://atagc.med.ualberta.ca).
Results
100% of control EMBs were graded as IAMC score 0. All pAMR1(I+) EMBs and none of the pAMR1(H+) and pAMR2-3 were graded as IAMC score 0. The highest IAMC score 3 was mainly distributed in pAMR2-3 (Fischer's exact=0.000). Increase in the IAMC score was associated with an increase in the proportion of C4d and CD68 (macrophage marker) positive EMBs, and a higher proportion of DSA positive at EMB. It was also associated with an increase in the expression of ENDAT, DSAST, GRIT, NKB and QCMAT transcripts (All Kruskal-Wallis<0.001).
Conclusion
The IAMC score is associated with molecular activation in grafted myocardial tissue. The IAMC score could help pathologists for AMR diagnosis, emphasizing the value of IAMC in AMR detection.


21/ DETERMINANTS AND OUTCOMES OF CARDIAC ALLOGRAFT VASCULOPATHY / MAJOR ROLE OF DONOR-SPECIFIC ANTIBODY 
Session : « Challenges in heart transplantation» (17:05 to 18:35) 
Location : Room 124 – 17:47

M Racapé, G Bonnet, MC Bories, S Varnous P. Rouvier, R. Guillemain, P. Bruneval, JL Taupin, C Lefaucheur, A. Loupy

Background
The role of circulating DSA in addition to traditional cardiovascular risk factors in the development of CAV has not been demonstrated.
Methods
This observational, prospective cohort study included 723 heart transplant patients from 2 centers between 2004 and 2011. Participants were screened for traditional cardiovascular risk factors, circulating anti-HLA antibodies (specificity, HLA class, strength). All patients underwent prospective heart allograft biopsies and angiogram with assessment of CAV lesions. We assessed the independent determinants of CAV at 3 years after transplantation.
Results
A total of 145 patients (20.1%) had circulating DSA at transplantation. 170 patients (23.5%) experienced acute rejection in the first 3 years post-transplant with 128 cases (17.7%) of grade 2R ACR and 83 cases (11.5%) of antibody-mediated rejection (ISHLT pAMR+). At 3 years post transplant, 29.7% of patients had CAV (20.8%, 7.4% and 1.5% with CAV scores of 1, 2 and 3 respectively). After adjusting on traditional risk factors (recipient age, primary heart disease, gender, hypertension, tobacco use, dyslipidemia, diabetes mellitus, BMI), donor factors (age, gender, BMI, cause of death), transplant characteristics (cold ischemia time, center, emergency heart transplantation), immunological parameters (circulating DSA at transplantation and acute rejection), and CMV disease occurence, the independent determinants of CAV at 3 years were: donor age (RR=1.05; 95%CI=1.03-1.08), recipient dyslipidemia (RR=2.1; 95%CI=1.02-4.29), presence of circulating DSA at transplantation (RR=2.45 95%CI=1.45-4.12). Occurrence of pAMR2 post-transplant was also a strong and independent factor associated with CAV at 3 years post-transplant (RR=3.51, IC95%=1.84-6.69). This group showed decreased patient survival (HR=1.8 p=0.01).
Conclusion
Circulating DSA are major determinants of severe arteriosclerosis, independent of traditional cardiovascular risk factors. Antibody-mediated CAV is associated with reduced patient survival. ​