Inflammatory cell subtypes and relation to clinical and electron microscopic findings in the diagnosis of antibody-mediated rejection in renal transplant biopsies

Aim: Renal transplant rejection develops in two different pathways: T cell mediated rejection (TCMR) and antibody mediated rejection (AMR). The diagnosis is rendered based on the Banff classification system. Various studies demonstrated the role of natural killer (NK) cells, part of the innate immune system, and macrophages in graft dysfunction and rejection pathogenesis. Electron microscopic examination poses an essential role in detecting early lesions, otherwise absent in light microscopy. Banff classification system recognizes the use of electron microscopy in determining early lesions of transplant glomerulopathy, such as cg1a, and peritubular capillary basement membrane multilamellation (ptcml). In this study, we aim to investigate the relationship between the density of CD163 expressing monocyte/macrophages and CD56 expressing NK cells in the inflammatory infiltration localised within the glomerular and peritubular capillary lumina in renal transplant biopsies diagnosed as AMR, and histological parameters and prognosis; and to evaluate electron microscopic glomerulitis, cg1a and ptcml lesions and their correlations with histopathologic and prognostic data. Materials and Methods: Intraluminal and interstitial inflammatory cell infiltration was investigated by immunohistochemistry for CD56/CD34 and CD163/CD34 dual markers in 161 renal transplant indication biopsies of 85 patients, diagnosed as AMR and/or TCMR at Istanbul Faculty of Medicine between 2013-2018. The presence and severity of glomerulitis, cg, ptcml, and related lesions were examined under electron microscopy and compared to histopathological and prognostic data. All biopsies were reevaluated and the lesion scores were updated according to the current 2019 Banff classification. Results: Fifty-three (62.3%) patients were male and 32 (37.6%), female. Sixty-nine (80.2%) transplantations were from living donors, and 83% of biopsies were performed ≥12 months posttransplant. Follow-up time was 25 months (0-93) (median). Thirty-three patients showed graft loss (48,5%). Averages of Banff lesion scores ranged between 0.16 and 1.78. DSA was positive in 61.6% and PRA was positive in 60.7% patients. No rejection was detected in 31 (19.7%) biopsies. In 126 biopsies with a diagnoses of rejection, 56 (35.7%) was diagnosed as AMR/suspicious for AMR, 38 (24.2%) as TCMR, and 32 (20.4%) as mixed rejection. CD163 immunohistochemical analysis was applied on 131 (18,6%) biopsies. Interstitial CD163 antibody positivity was evaluated as score 0 in 30 (22.9%) cases, score 1 in 58 (44.3%), score 2 in 29 (22.1%), and score 3 in 14 (10.7%). Lesion scores i, ti, ct, ci, and t-IFTA were higher in biopsies with an interstitial CD163 positivity score ≥2, compared to the group with score <2 (p<0.05). The number of CD163 positive cells in glomeruli was significantly different between no rejection and AMR groups, no rejection and mixed rejection groups, and AMR and TCMR groups (p<0.05). The numbers of total and mesengial CD163 positive cells were higher in AMR and mixed rejection groups (p<0.001). The number of CD163 positive intracapillary cells within peritubular capillary lumina per high-power field (HPF) and the maximum number of CD163 positive intraluminal cells were higher in acute AMR, compared to chronic AMR (p <0.05). The numbers of total and mesengial CD163 positive cells in glomeruli were correlated with g, ptc and cg lesions (rho >0,3 and p <0.005). CD56 immunohistochemistry was applied on 117 biopsies. Sixteen (13.4%) biopsies showed interstitial positivity for CD56. Whereas biopsies with an i score ≥2 showed more CD56 positive cells in the interstitium, the group with mm score ≥1 showed lower CD56 positivity(p<0.05). No relationship between CD163 immunohistochemistry results and the presence of interstitial CD56 positivity was found (p>0.05). Fifty-three biopsies (32.9%) of 33 patients were evaluated under electron microscopy. Glomerulitis was detected in 32 biopsies (71.1%), out of which, 26 also showed occlusion in capillary lumina. Endothelial swelling was detected in 29 (64.4%) biopsies, increased lucency of lamina rara interna in 35 (77.8%), subendothelial fluffy material in 21 (46.7%), loss of fenestration in 38 (84.4%), mesengial interposition in 22 (28.8%), ≥7 layers of multilamellation in ≥1 peritubular capillaries in 10 (22.2%), and ≥5 layers of multilamellation in ≥1 peritubular capillaries in 18 (40%). Banff lesion scores g, cg, ptc and CD163 positive glomerular cells showed significant correlation with electron microscopic findings (p<0.05). The numbers of CD163 positive mesengial, intracapillary and total macrophages per glomerule and the number of CD163 positive macrophages within peritubular capillary lumina per HPF were significantly different within groups of Banff lesion scores cg0, cg1a and ≥cg1b (p<0.05). DSA scores were correlated with the number of layers in the most severely affected peritubular capillary and the ptcml score (rho and p values, respectively: 0.658 and 0.010, 0.523 and 0.055). Banff lesion scores ah, mm and i were higher in patients with graft loss (p<0.05). Graft loss was higher in patients with an interstitial inflammation score 2-3 on CD163 immunohistochemical examination, compared to patients with a score of 0-1 (p=0.04) (Odds ratio: 7.08 [1.68-29.76]). Conclusion: To conclude, our study showed that rejection subcategories correspond to distinct histopathological groups and postulated a novel perspective for an alternative utility of the CD163 antibody. CD56 antibody positivity in tissues was lower than expected in rejection and this parameter may yield different results in other series. It was demonstrated that inflammatory reaction in all compartments may trigger a fibrotic response and result in graft loss. This should be considered in profit and loss balance in therapy decisions.