Clinicopathological Analyses of Chronic Renal Allograft Arteriopathy after Kidney Transplantation

Chronic renal allograft arteriopathy (CRA) is a morphologic pattern of chronic kidney allograft injury [1]. The pathological features of CRA described in the Banff 1997 classification are as follows: disruption of the elastica, inflammatory cells in the expanded intima, proliferation of myofibroblasts in the expanded intima, and the formation of a second “neo-intima” [1]. CRA represents chronic T-cell-mediated rejection (TCMR) and/or chronic active antibody-mediated rejection (AMR) according to the Banff 2019 classification [2]. A few reports have described CRA [3‒7]. Here, we present the results of our clinicopathological analyses of CRA cases after kidney transplantation and attempt to clarify the underlying mechanisms and prognostic significance.

Two hundred and twenty-seven kidney transplantations and 1,163 allograft biopsies were performed between January 2010 and December 2020 at the Toda Chuo General Hospital. CRA was diagnosed in 34 renal allograft biopsy specimens (BSs) obtained from 27 renal transplant patients who were followed up at our institute between January 2010 and December 2020. Data for the 34 BSs and 27 patients obtained from the medical records were retrospectively reviewed. The patients received a triple immunosuppressive protocol consisting of tacrolimus, mycophenolate mofetil, mizoribine, and methylprednisolone (Table 1). Basiliximab and rituximab (RIT) were also used in some cases (Table 1). In particular, RIT was used for ABO incompatible kidney transplantation and positive cases of anti-donor antibody before transplantation.

Renal allograft biopsy was performed as part of the diagnostic workup for allograft dysfunction and proteinuria, or as a protocol biopsy. BS was examined using light, electron, and immunofluorescence microscopy. BS was diagnosed and scored according to the Banff 2019 classification [2]. CRA was defined based on the quantitative criteria for vascular fibrosis intimal thickening detected by light microscopy, as mentioned in the Banff 2019 classification (Banff cv score) (Fig. 1) [2]. Moreover, we defined CRA as a cv ≥1 with inflammatory cell infiltration of the fibrotic intima. We used the peritubular capillary basement membrane (PTCBM) score employed for thickening of the PTCBM, which is evaluated by light rather than electron microscopy of PTCBM multilayering [8].

Patient sera collected during the peri-biopsy period were screened for anti-human leukocyte antigen (HLA) class 1 and 2 antibodies using plastic beads coated with HLA (Luminex Technology, One Lambda, Veritas, Tokyo, Japan). We defined anti-donor-specific HLA antibody (DSA) as being present when the mean fluorescence intensity determined by Luminex Technology was ≥800. Regarding the renal allograft function after allograft biopsy, an increase in serum creatinine level of 20% or more was regarded as deterioration of renal allograft function, a decrease of 20% or more was regarded as improvement in renal allograft function, and a change of less than 20% was regarded as stable renal allograft function.

All patients provided informed consent for biopsy and blood sample collection. The study was approved by the Ethics Committee of Toda Chuo General Hospital (No. 0488).

Categorical variables were analyzed using the χ² test. The mean values in the two groups were compared using analysis of variance and the χ² test if the data were not normally distributed. Statistical significance was set at p < 0.05.

The clinical characteristics of the 27 patients with CRA diagnosed via biopsy are shown in Table 1. Sixteen patients (59%) experienced one or more biopsy-proven rejection episodes prior to the study.

The biopsy-proven rejection episodes were seen on an average of 928 days (42–3,392 days) before CRA diagnosis. Prior biopsies of 3 patients showed active AMR alone, five showed acute TCMR alone, three showed borderline changes alone, and three showed chronic active AMR alone. Combined chronic active AMR and acute TCMR were noted in 1 patient. CRA was diagnosed at a median of 1,003 days post-transplantation, with a mean serum creatinine of 1.56 mg/dL.

The histopathology of 34 allograft BSs with CRA is shown in Table 2. Of the 34 CRA biopsies, 22 were mild (cv1 in Banff 2019 classification), seven were moderate (cv2), and five were severe (cv3). The relationships between the Banff cv score and other Banff scores are also shown in Table 2, wherein the Banff ptc, ci, and ct scores are high.

Of the 34 renal allograft biopsies, 29 had serum samples available for anti-HLA antibody analysis during the peri-biopsy period. Anti-HLA antibodies were detected in 23 (79%) samples. Of the 23 detections of anti-HLA antibodies, 14 (48%) were DSA, including class 1 antibody alone in one (4%), class 2 antibody alone in nine (31%), and both class 1 and 2 in three (10%). De novo DSA was detected six times (21%), and all were class 2 DSAs.

We classified these 34 CRA BSs based on their histopathological features (Table 3). The cv alone (isolated cv lesion) was seen in 11 BSs (32%), the isolated v lesion was seen in five BSs (15%), and cv + active AMR, which included g (Banff g score) + ptc (Banff ptc score) ≥2 and g ≥1 + anti-HLA antibody positivity, was seen in two BSs (6%). The cv + chronic active AMR, which included cg (Banff “g” score) ≥1 and/or PTCBM ≥1 + anti-HLA antibody positivity, was seen in eight BSs (24%). The cv + acute TCMR, which includes i (Banff i score) + t (Banff t score) ≥2 and t ≥1, was observed in six BSs (18%). In total, the cv + AMR was seen in 12 BSs (35%), and the cv + TCMR was seen in eight BSs (24%), including two BSs with both AMR and TCMR.

Regarding the prognosis of the patients with CRA, 3 cases (11%) of graft loss occurred during the observation period. Of the graft loss cases, one was a cv + isolated v lesion, one was a cv + chronic active AMR, and one was a cv + acute TCMR (Table 3). Deterioration of renal allograft function after biopsy occurred in 7 patients (26%) during the observation period. The relationship between graft function and Banff scores is present in Table 4. Patients with deterioration of renal allograft function and graft loss showed significantly high Banff v scores, but no statistically significant relationships with other Banff scores were noted. As for the relationship between graft function and clinical characteristics, there was no significant difference in patient background between the deterioration or graft loss and stable or improvement groups.

While there are few reports on CRA after kidney transplantation, the clinical and pathological aspects of CRA have not yet been fully discussed [3‒7]. In our study, 59% of all patients with CRA had a previous history of rejection episodes, including AMR alone in 22% of the patients, TCMR (including borderline changes) alone in 33%, and combined AMR and TCMR in 4%. Thus, rejection episodes may be a risk factor for CRA.

In our study, of the 34 BSs showing evidence of CRA, the CRA was mild (cv1) in 22 BSs, moderate (cv2) in 7, and severe (cv3) in 5. The presence of interstitial fibrosis and tubular atrophy (IF/TA), peritubular capillaritis, and arteriolar hyaline thickening was associated with the presence of CRA. Furthermore, tubulitis, interstitial inflammation, glomerulitis, glomerular double contours, intimal arteritis (IA), PTCBM multilayering, C4d deposition in the peritubular capillary (PTC), and calcineurin inhibitor arteriolopathy were weakly associated with CRA. Hill et al. [5] reported that accelerated arteriosclerosis (AS), which is provoked by DSA, was significantly associated with peritubular capillaritis, C4d deposition in PTC, IF/TA, glomerulitis, tubulitis, and interstitial inflammation. The strong association between IF/TA and peritubular capillaritis was consistent with the results of our study; however, the strong associations with glomerulitis, tubulitis, interstitial inflammation, and C4d deposition in PTC differed from the results of the present study [5]. However, in the DSA-negative cases in Hill’s study, accelerated AS was not associated with peritubular capillaritis, C4d deposition in PTC, IF/TA, glomerulitis, tubulitis, or interstitial inflammation. Based on these findings and our results, we speculate that CRA may be caused by both DSA and T-cell-mediated immunity.

In our study, anti-HLA antibodies were detected in 23 of the 29 (79%) samples obtained during the peri-biopsy period, for which the results of HLA analysis were available. Of the 23 samples with anti-HLA antibodies, 14 (48%) contained DSA. In our study, CRA may have been provoked by anti-HLA antibodies in many, but not all, cases.

We then classified 34 BSs based on the overall histopathological features: cv alone (isolated cv) was observed in 11 BSs (32%), cv + isolated v lesion in five BSs (15%), cv + AMR in 12 BSs (35%), cv + TCMR (involving borderline changes) in eight BSs (24%), and cv+ in AMR and TCMR in two BSs (6%). From these results, CRA was related to AMR in 30–40% of the cases and to TCMR in 20–30% of cases. Therefore, we speculate that both cellular and humoral immunity play important roles in the pathogenesis of CRA after renal transplantation. Endarteritis (IA), which is defined as “v” in the Banff 1997 classification, has been discussed previously, and it was concluded that both cellular and humoral immunity play roles in this type of vascular rejection; in some cases, IA may be provoked by anti-donor antibodies, while in others it may be provoked by T-cells and natural killer cells [1, 9, 10]. This theory regarding the pathogenesis of IA may also apply to CRA. Thus, we suggest that while CRA may be caused by IA in many cases, TCMR may be involved in other cases. Indeed, in the Banff 2019 classification, CRA appears under both chronic active AMR and chronic active TCMR [2].

Regarding the prognosis of patients with CRA, graft loss occurred in 3 of 23 patients (11%) during the observation period. Of the remaining patients with functioning grafts, deterioration of renal allograft function after biopsy was seen in seven cases (26%), while stable or improved renal allograft function was seen in the other 17 (63%). In our study, the immunosuppressive protocol consisted mainly of tacrolimus, mycophenolate mofetil, and methylprednisolone, with additional use of basiliximab or RIT in some cases. Our study results showed that the prognosis of grafts exhibiting evidence of CRA was not poor under the present immunosuppressive protocol.

Loupy et al. [7] described the prognosis of patients with CRA. They reported that circulating DSA was significantly associated with severe allograft AS, that patients with antibody-associated severe AS had decreased allograft survival and increased mortality, and that more circulating DSA was significantly associated with the occurrence of major adverse cardiovascular events, independent of traditional risk factors. They concluded that circulating antibodies are major determinants of severe AS and major adverse cardiovascular events, independent of traditional cardiovascular risk factors. In our study, no cardiovascular disease was the cause of death.

We studied the relationship between graft function and Banff scores and found no statistically significant relationship with Banff scores without v scores, which were significantly higher in patients whose renal allograft function showed deterioration (p = 0.0420). The v score showed a strong relationship with renal allograft function. Therefore, we speculate that IA plays an important role in the prognosis of patients with CRA.

This study had some limitations. First, the chronic Banff scores, such as the Banff ti score, i-IFTA score, and t-IFTA, could not be examined because some BSs had been diagnosed at times without such Banff scores. Second, this study was retrospective in nature, with a small sample size and shorter follow-up period. Finally, among the 27 patients with CRA, 19 underwent follow-up biopsies, and it was only in seven (37%) cases that CRA was present again in the BS. It is possible that the same arteries were not been obtained in the past. Hence, the difficulty of diagnosing arterial lesions is keenly realized.

In conclusion, our study results suggest that AMR contributes to CRA in 30–40% of cases, TCMR in 20–30% of cases, isolated v lesions in 15% of cases, and cv lesions alone in 30%. The IA was a prognostic factor in CRA.

We thank Shigeru Horita and Hideki Nakayama (Division of Pathology, Kidney Center, Tokyo Women’s Medical University) and Miyuki Furusawa (Division of Transplant Immunology, Department of Urology, Tokyo Women’s Medical University) for technical support.

This study was conducted in accordance with the Declaration of Helsinki of the World Medical Association Declaration Helsinki. All patients provided informed consent for biopsy and collection of blood samples. The study was approved by the Ethics Committee of Toda Chuo General Hospital. An ethics statement is not applicable because this study is based exclusively on the published literature. This study protocol was reviewed and approved by Toda Chuo General Hospital, approval number 0488.

Tomokazu Shimizu, Kazuya Omoto, Masashi Inui, Taiji Nozaki, Toshio Takagi, and Hideki Ishida have no conflicts of interest to declare.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Tomokazu Shimizu provided the conception and design of the study, acquisition of data, analysis and interpretation of data, and participated in the writing and approval of the article; Kazuya Omoto, Masashi Inui, Taiji Nozaki, and Toshio Takagi were responsible for the article critically for important intellectual content; and Hideki Ishida provided the revised the article critically for important intellectual content and gave final approval of the version to be submitted.

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.