- Top of page
- Materials and Methods
Subclinical antibody-mediated rejection (AMR) has been described in renal allograft recipients with stable serum creatinine (SCr), however whether this leads to development of chronic allograft nephropathy (CAN) remains unknown.
We retrospectively reviewed data from 83 patients who received HLA-incompatible renal allografts following desensitization to remove donor-specific antibodies (DSA). Ten patients had an allograft biopsy showing subclinical AMR [stable SCr, neutrophil margination in peritubular capillaries (PTC), diffuse PTC C4d, positive DSA] during the first year post-transplantation; 3 patients were treated with plasmapheresis and intravenous immunoglobulin. Three patients had a subsequent rise in SCr and an associated biopsy with AMR; 5 others showed diagnostic or possible subclinical AMR on a later protocol biopsy. One graft was lost, while remaining patients have normal or mildly elevatedSCr 8–45 months post-transplantation. However, the mean increase in CAN score (cg + ci + ct + cv) from those biopsies showing subclinical AMR to follow-up biopsies 335 ± 248 (SD) days later was significantly greater (3.5 ± 2.5 versus 1.0 ± 2.0, p = 0.01) than that in 24 recipients of HLA-incompatible grafts with no AMR over a similar interval (360 ± 117 days), suggesting that subclinical AMR may contribute to development of CAN.
- Top of page
- Materials and Methods
Over the past 10–15 years, an increasing number of transplant centers worldwide have successfully expanded the potential pool of living kidney donors by performing transplants of crossmatch positive (HLA-incompatible) kidneys into recipients who are preconditioned to remove donor-specific antibodies (DSA) (1–7). A potential risk of such procedures, however, is the continued presence or reappearance of such antibodies with resulting antibody-mediated rejection (AMR) of the graft. Although successful strategies have been developed to treat acute episodes of AMR (6–10), humoral alloreactivity in the early post-transplant period adversely impacts long-term renal allograft survival (11) and very likely contributes to the development of chronic rejection (12,13).
Patients receiving HLA-incompatible renal allografts often undergo one or more protocol biopsies of their grafts during the first year post-transplantation to detect evidence of subclinical rejection. It has been well documented in studies of crossmatch negative renal allografts that acute cellular rejection [ACR; Banff '97 (14) grade 1A or greater] may be detected on protocol biopsies of stably functioning grafts. In patients receiving baseline immunosuppressive regimens consisting of cyclosporine, azathioprine and corticosteroids, such subclinical ACR, if untreated, is associated with an increased likelihood of developing chronic allograft nephropathy (CAN) as detected on subsequent protocol biopsies (15,16). Subclinical AMR, meeting established pathologic and serologic criteria for AMR (17) without associated graft dysfunction or concurrent ACR, has also been described in a small but significant subset of protocol biopsies of HLA-incompatible grafts (7,18). Gloor et al. (7) reported 4 such patients diagnosed with subclinical AMR during the first week posttransplantation; each responded to treatment with methyprednisolone, plasmapheresis (PP) and low-dose intravenous immunoglobulin (IVIG) with resolution of histologic abnormalities. All 4 patients had stable graft function 8–24 months later, although 3 had mildly elevated serum creatinine (SCr) levels (1.3–1.7 mg/dL), and it was not stated whether follow-up biopsies showed changes of CAN (7).
In a recent study (18) that included 66 recipients of HLA-incompatible grafts, we noted that 17 of 103 protocol biopsies of stably functioning grafts taken at 1, 3, 6 or 12 months posttransplantation showed strong, diffuse peritubular capillary (PTC) staining for C4d. In the majority of these cases, there were also histologic changes [neutrophil margination in PTC and/or thrombotic microangiopathy (TMA)] suggestive of AMR and low levels of DSA at the time of the biopsy, and as such these would satisfy criteria for AMR developed at the 2001 Banff Conference on Allograft Pathology (17). In the present study, we retrospectively examined the clinical and pathologic courses of those patients with subclinical AMR, focusing primarily on the results of follow-up biopsies. The major aim of the study is to determine whether subclinical AMR is associated with subsequent development of CAN.
- Top of page
- Materials and Methods
Ten of the 83 patients (12%) had at least 1 biopsy meeting diagnostic criteria for subclinical AMR. At the time of the initial biopsy showing subclinical AMR, each of these 10 patients had a positive flow cytometric crossmatch with documented DSA (2 with antibody against HLA class I, 4 with anti-class II, 4 with anti-class I and anti-class II), although none of the patients had a positive CDC crossmatch. Five of these patients had residual DSA at the time of transplantation (3 anti-class II by flow only, 1 anti-class II by CDC, 1 anti-class I and anti-class II by CDC). In addition to pre- and postoperative PP/CMVIg treatments and the standard quadruple-drug immunosuppressive regimen (see Materials and Methods), 1 patient underwent splenectomy and 2 others received anti-CD20 antibody (Rituxan, IDEC-Genentech, San Francisco, CA) within 1 week of transplantation; each had residual DSA at the time of surgery.
For each patient, clinical data at the time of the diagnostic biopsy and findings on this biopsy are summarized in Table 1. Subclinical AMR was diagnosed at varying times posttransplantation, ranging from 8 days to approximately 1 year; the biopsy on PTD 8 was done at the time of a fascial repair in a patient with excellent and stable graft function. One additional biopsy showing subclinical AMR (Patient 6) was done because of proteinuria without associated edema or a rise in SCr; this biopsy showed moderate (g2) glomerulitis but no other glomerular pathology, although a biopsy done approximately 2 months later showed evidence of focal-segmental glomerulosclerosis (FSGS), most likely recurrent. The remaining 8 biopsies were scheduled protocol biopsies. Six biopsies showed focal, mild interstitial lymphocytic infiltrates with mild tubulitis, and were diagnosed as having borderline lesions by the Banff '97 classification. No biopsy showed more than mild interstitial inflammation (i1) or tubulitis (t1), and none showed intimal arteritis, although 8 showed glomerulitis (3 g1, 4 g2, 1 g3) and 1 very focal glomerular fibrin thrombi.
Table 1. Clinical, histologic and immunohistologic data for patients with subclinical AMR
|Patient||Biopsy PTD||SCr (mg/dL) Baseline||Biopsy||Peak1||Banff '97 grade2||C4d3||Margination score4||CAN score5||Postbiopsy treatment|
|1 (52 WF)||363||1.0||1.1||1.2||B||2+||1+||5||None|
|2 (64 BF)||97||1.2||1.2||1.5||NR||2+||1+||1||None|
|3 (48 WF)||35||0.8||0.8||1.0||B||3+||1+||0||None|
|4 (72 WF)||111||0.8||0.7||0.9||NR||2+||1+||0||None|
|5 (38 WF)||8||0.9||0.9||1.0||B||1−2+||1−2+||0||PP/IVIG x3|
|6 (46 WF)||139||1.0||1.0||1.2||NR||1−2+||1+||2||None|
|7 (44 WF)||22||1.1||1.1||1.1||NR||3+||1+||0||None|
|8 (27 WM)||87||0.8||0.7||0.8||B||1−2+||3+||1||incr. MMF, tac|
|9 (31 WM)||87||1.3||1.3||4.3||B||2+||2+||0||PP/IVIG x6|
|10 (54 WF)||271||1.2||1.3||1.3||B||2+||1+||4||PP/IVIG x4|
Following the diagnosis of subclinical AMR, 3 patients were treated with PP/CMVIg, 1 had his daily dosages of MMF and tacrolimus increased, and 6 received no additional therapy (Table 1). Still, as shown in Table 2, only 3 patients (1 treated with PP/CMVIg) subsequently developed an increase in SCr warranting a biopsy to be performed prior to the next scheduled protocol biopsy. All 3 of the latter biopsies showed neutrophil margination and diffuse C4d staining in PTC, and all 3 patients had detectable DSA at the time of the biopsy, thus meeting Banff (17) criteria for AMR, although the biopsy of Patient 2 also showed other lesions (evolving chronic transplant glomerulopathy and de novo FSGS) that could also have accounted for the increased SCr. The remaining patients maintained a stable SCr and underwent a protocol biopsy 2–12 months later; 2 of these biopsies showed concurrent AMR and ACR (Banff '97 grades 1A and 2A), 3 met 2 of 3 Banff criteria for AMR, and 2 showed only recurrent glomerular disease with no evidence of AMR or ACR (Table 2). However, compared with the diagnostic biopsy (Table 1), all but 1 follow-up biopsy showed an increase in CAN score (Table 2), with an increase of ≥3 in 5 patients. Seven of the 10 patients had at least 1 additional biopsy; and at the time of their most recent biopsies, taken 335 ± 248 (SD) days (range 49–798) after the diagnostic biopsies, CAN scores had increased by ≥3 in 7 of the 10 patients and by ≥5 in four. Corresponding to these increases in CAN score, SCr levels at most recent follow-up were at least 0.3 mg/dL higher than at the time of the diagnostic biopsy in 8/10 patients (Table 3). Notably, the 2 patients (Patients 1 and 6) whose SCr did not increase were 2 of the 3 patients whose CAN score did not increase by 3 or more. One patient (Patient 2) lost her graft; 3 biopsies from this patient (the biopsy with diagnostic changes of subclinical AMR, the next biopsy with CAN, evolving chronic transplant glomerulopathy, and persistent AMR, and the most recent biopsy taken ∼2 weeks prior to resumption of dialysis) are shown in Figure 1. None of the remaining patients had a SCr above 1.8 mg/dL at last follow-up (Table 3).
Table 2. Clinical, histologic and immunohistologic data for patients with subclinical AMR at first follow-up biopsy
|Patient||Follow-up biopsy PTD||Biopsy indication||SCr (mg/dL)||Banff '97 grade||C4d1||Margination score2||CAN score3||Diagnoses|
|1 (52 WF)||763||Protocol||1.2||NR||2+ D||1+||6||Possible AMR|
|2 (64 BF)||333||Increase SCr||2.1||B||1+ D||1+||6||CTG, FSGS, AMR|
|3 (48 WF)||94||Protocol||0.9||1A||2+ D||2+||3||AMR + ACR|
|4 (72 WF)||238||Increase SCr||1.3||B||2+ D||2+||2||AMR|
|5 (38 WF)||76||Protocol||1.0||NR||0||0||3||Recurrent MPGN|
|6 (46 WF)||188||Protocol||1.2||NR||1 + F||0||1||Recurrent FSGS|
|7 (44 WF)||85||Protocol||1.1||2A||3 + D||1+||5||AMR + ACR|
|8 (27 WM)||136||Protocol||1.2||B||1–2+ F||3+||2||Possible AMR|
|9 (31 WM)||96||Increase SCr||3.5||B||2+ D||3+||1||AMR|
|10 (54 WF)||346||Protocol||1.3||NR||1–2+ D||trace||7||Resolving AMR|
Table 3. Clinical outcomes of patients with subclinical AMR
|Patient||Posttransplant day First biopsy with subclinical AMR||Most recent follow-up||SCr (mg/dL) First biopsy with subclinical AMR||Most recent follow-up|
|1 (52 WF)||363||1361||1.1||1.2|
|2 (64 BF)||97||795||1.2||Graft Loss1|
|3 (48 WF)||35||1162||0.8||1.3|
|4. (72 WF)||111||898||0.7||1.2|
|5 (38 WF)||8||883||0.9||1.5|
|6 (46 WF)||139||596||1.0||1.0|
|7 (44 WF)||22||581||1.1||1.4|
|8 (27 WM)||87||197||0.7||1.2|
|9 (31 WM)||87||1175||1.3||1.8|
|10 (54 WF)||271||1069||1.3||1.7|
Figure 1. Histologic changes (A, C, E) and C4d staining (B, D, F) in renal allograft biopsies from Patient 2. (A,B) Biopsy first showing subclinical AMR, done on posttransplantation day (PTD) 97. There is acute glomerulitis, neutrophil and mononuclear leukocyte (ML) margination in PTC (arrows), diffuse PTC C4d staining (2+) and very focal interstitial ML infiltrate, with minimal tubulitis. (C,D) Subsequent biopsy on PTD 333, done because of elevated SCr (2.1 mg/dL). There is residual AMR with focal neutrophil margination (arrows) and 1+, diffuse C4d staining in PTC, early chronic transplant glomerulopathy with segmental duplication of glomerular basement membranes (arrowheads), tubular atrophy and interstitial fibrosis. (E,F) Most recent renal allograft biopsy, done on PTD 780, 15 days prior to re-initiation of dialysis. There is chronic transplant glomerulopathy and extensive tubular atrophy and interstitial fibrosis. Only very focal PTC C4d staining is noted (1+). (A) Hematoxylin-eosin stain, original magnification ×400. (C) Periodic acid-Schiff (PAS) stain, ×400. (E) PAS, ×200. (B, D and F) Indirect immunofluorescence with mouse monoclonal anti-human C4d and fluorescein isothiocyanate-conjugated goat anti-mouse IgG, ×400 (B, D) or ×200 (F).
Download figure to PowerPoint
To assess whether the increase in CAN score in patients with subclinical AMR was greater than in patients without AMR, we compared the above findings to those in a control group of patients consisting of all patients receiving a positive crossmatch renal allograft over the same study period who met the following criteria: (1) a biopsy was done during the first 3 months posttransplantation to provide a baseline CAN score, and there was at least 1 biopsy taken ≥6 months posttransplantation, and (2) There were no biopsies showing diagnostic changes of AMR. Twenty-four of the remaining 73 recipients of positive crossmatch renal allografts met these criteria. Of the 49 patients who did not meet these criteria, 34 had 1 or more biopsies showing AMR, and 15 did not have an adequate biopsy during the first 3 months or ≥6 months posttransplantation. There were no graft losses among the 24 control patients during the first 3.5 years posttransplantation. As was the case with 2 patients with subclinical AMR (Table 2), 2 of the control patients had a recurrence of their original disease in the allograft (both membranoproliferative glomerulonephritis).
Table 4 compares demographic features and serologic data from the 24 control patients with those from the 10 patients with subclinical AMR. There were no significant differences between these 2 groups with regard to patient demographics. In each group, half of the patients had residual DSA at the time of transplantation, although in each the majority of the latter patients had only low levels of DSA, detectable by flow cytometric but not by CDC crossmatch. None of the positive CDC crossmatches had a titer >1:4. Patients who developed subclinical AMR required a higher mean number of pretransplant PP/IVIG treatments than controls (p = 0.05; Table 4), suggesting that removal of DSA in the former patients tended to be more difficult.
Table 4. Comparison of demographic and serologic data in patients with subclincial AMR and controls
| ||Subclinical AMR||Controls||p-value|
|Number of patients||10||24|| |
|Age (years, mean ± SD)1||47.6 ± 13.9||44.2 ± 9.1||0.412|
|Age range (years)||27–72||30–63|| |
|DSA at time of transplantation|
| Negative||5 (50%)||12 (50%)|| |
| Pos. by CDC or flow||5 (50%)||12 (50%)||1.03|
| Pos. by CDC||2 (20%)||2 (8%)|| |
| Anti-class I only||0||2|| |
| Anti-class II only||4||7|| |
| Anti-class I and II||1||3|| |
|Number of PP/IVIG treatments|
| Pretransplant (mean ± SD)||4.9 ± 4.0||2.8 ± 1.9||0.052|
| Pretransplant, range||1–14||0–8|| |
| Posttransplant (mean ± SD)||5.2 ± 5.6||3.3 ± 2.1||0.162|
| Posttransplant, range||2–20||1–12|| |
Biopsy findings and SCr levels in these 2 groups of patients are summarized in Table 5; SCr values shown for the control group included only values from those 14 patients whose initial biopsy was a protocol biopsy rather than a biopsy done because of graft dysfunction. Compared to patients with subclinical AMR, a greater fraction of control patients had ACR (Banff '97 type 1A or greater) on 1 or more renal allograft biopsies including the initial biopsy within the first 3 months posttransplantation (for control patients), the first biopsy showing diagnostic changes of subclinical AMR (for patients with subclinical AMR), the most recent follow-up biopsy and any biopsy between these. There was no difference between the 2 groups in the mean interval between initial/diagnostic and most recent biopsies, or in the mean SCr level and mean CAN score (cg + ci + ct + cv) at the time of the initial/diagnostic biopsy. However, as shown in Table 5, the mean change (increase) in SCr from the time of initial/diagnostic biopsy to that of the most recent follow-up biopsy was significantly greater in the patients with subclinical AMR. The mean CAN score on the most recent follow-up biopsies was also significantly higher in patients with subclinical AMR than in the control patients. In control patients, the mean CAN score did increase significantly from the initial to the most recent follow-up biopsies (p = 0.02 by paired t-test), although only 5 (21%) of these patients showed an increase in CAN score of ≥3 and only 1 (4%) an increase of ≥5. In the 10 patients with subclinical AMR the increase in mean CAN score from the diagnostic biopsies first showing subclinical AMR to the most recent biopsies was highly significant (p = 0.001 by paired t-test). However, in contrast to the control patients and as noted above, 7 (70%) of the patients with subclinical AMR had an increase in CAN score of ≥3, and 4 (40%) had an increase of ≥5. Furthermore, the mean increase in CAN score between the biopsies was significantly greater in patients with subclinical AMR than in control patients (Table 5). Figure 2 illustrates the changes over time in SCr and CAN score in each of the 10 patients with subclinical AMR.
Table 5. Comparison of clinical and biopsy results in patients with subclincial (SC) AMR and controls
| ||SC AMR (n = 10)||Controls (n = 24)||p-value|
|Number (%) of patients with ACR||2 (20%)||17 (71%)||0.013|
|ACR Type (Banff '97)1|
| 1A||1||3|| |
| 1B||0||2|| |
| 2A||1||10|| |
| 2B||0||1|| |
| 3||0||1|| |
|Interval, diagnostic/initial—most recent biopsy (days, mean ± SD)||335 ± 248||360 ± 117||0.694|
|Interval, range||49–798||142–645|| |
| At diagnostic/initial biopsy (mean ± SD)||1.0 ± 0.2||1.2 ± 0.3||0.104|
| At diagnostic/initial biopsy, range||0.7–1.3||0.7–1.6|| |
| At most recent biopsy (mean ± SD)||1.7 ± 1.0||1.3 ± 0.3||0.154|
| At most recent biopsy, range||1.1–4.3||0.8–1.9|| |
| Difference (mean ± SD)||0.7 ± 0.9||0.1 ± 0.3||0.024|
| Difference, range||0.1–3.1||(−0.3) to 0.8|| |
| At diagnostic/initial biopsy (mean ± SD)||1.3 ± 1.8||1.9 ± 1.6||0.374|
| At diagnostic/initial biopsy, range||0–5||0–6|| |
| At most recent biopsy (mean ± SD)||4.8 ± 2.3||2.9 ± 1.8||0.024|
| At most recent biopsy, range||1–9||1–7|| |
| Difference (mean ± SD)||3.5 ± 2.5||1.0 ± 2.0||0.014|
| Difference, range||(−1) to 8||(−2) to 5|| |
Figure 2. Changes in CAN scores (solid lines, filled circles) and SCr levels (dashed lines, open circles) in all 10 recipients of HLA-incompatible renal allografts with subclinical AMR. For each patient, CAN scores (cg + ci + ct + cv) from the diagnostic biopsy first demonstrating subclinical AMR, the most recent biopsy, and any intervening biopsies, as well as SCr levels at the time of each biopsy and at most recent follow-up are plotted. Left column, top to bottom: Patients 1, 3, 5, 7 and 9. Right column, top to bottom: Patients 2, 4, 6, 8 and 10. Patient numbers are also shown in the lower right corner of each individual graph. Patient 2 started dialysis on posttransplant day 795, and as such a value for the most recent SCr is not indicated for this patient.
Download figure to PowerPoint
Table 6 shows changes in the individual indices comprising the CAN score for patients with subclinical AMR and controls. For controls, only ci showed a significant increase in mean value in the most recent biopsies compared to the initial biopsies (p = 0.04 by paired t-test); there was a trend toward an increase in ct (p = 0.06) by not in cg or cv (p = 0.33 for both). By contrast, in patients with subclinical AMR mean cg, ci and ct values were all significantly greater on the most recent biopsies than on the initial diagnostic biopsies (p-values of 0.001, 0.02 and 0.02, respectively, by paired t-tests; for cv p = 0.10). Furthermore, the mean difference in cg score (but not ci, ct or cv) between the diagnostic/initial and most recent biopsies was significantly greater in patients with subclinical AMR compared with controls (Table 6). This is a notable finding because of these chronic indices cg is felt to be the most closely associated with chronic AMR (13).
Table 6. Comparison of individual Banff chronic indices in patients with subclincial (SC) AMR and controls
| ||SC AMR (n = 10)||Controls (n = 24)||p-value (t-test)|
| At diagnostic/initial biopsy (mean ± SD)||0.1 ± 0.3||0.1 ± 0.3||0.88|
| At diagnostic/initial biopsy, range||0–1||0–1|| |
| At most recent biopsy (mean ± SD)||1.1 ± 0.7||0.1 ± 0.3||<0.001|
| At most recent biopsy, range||0–2||0–1|| |
| Difference (mean ± SD)||1.0 ± 0.7||0 ± 0.2||<0.001|
| Difference, range||0–2||0–1|| |
| At diagnostic/initial biopsy (mean ± SD)||0.4 ± 0.7||0.4 ± 0.5||0.91|
| At diagnostic/initial biopsy, range||0–2||0–1|| |
| At most recent biopsy (mean ± SD)||1.4 ± 0.8||0.8 ± 0.8||0.06|
| At most recent biopsy, range||0–3||0–2|| |
| Difference (mean ± SD)||1.0 ± 1.1||0.4 ± 0.9||0.11|
| Difference, range||(−1) to 3||(−1) to 2|| |
| At diagnostic/initial biopsy (mean ± SD)||0.5 ± 0.7||0.5 ± 0.6||1.0|
| At diagnostic/initial biopsy, range||0–2||0–2|| |
| At most recent biopsy (mean ± SD)||1.5 ± 0.7||0.9 ± 0.8||0.04|
| At most recent biopsy, range||1–3||0–2|| |
| Difference (mean ± SD)||1.0 ± 1.1||0.4 ± 0.9||0.09|
| Difference, range||0–3||(−1) to 2|| |
| At diagnostic/initial biopsy (mean ± SD)||0.3 ± 0.7||0.9 ± 0.8||0.05|
| At diagnostic/initial biopsy, range||0–2||0–2|| |
| At most recent biopsy (mean ± SD)||0.8 ± 0.6||1.1 ± 0.7||0.23|
| At most recent biopsy, range||0–2||0–2|| |
| Difference (mean ± SD)||0.5 ± 0.8||0.3 ± 1.0||0.43|
| Difference, range||(−1) to 2||(−2) to 2|| |
- Top of page
- Materials and Methods
The diagnosis of subclinical AMR has been somewhat controversial. In a report of a 2003 conference held in part to establish diagnostic criteria for AMR in solid organ allografts, clinical evidence of graft dysfunction is listed as 1 of 4 criteria required for definitive diagnosis of AMR (25). By contrast, the diagnostic criteria for AMR proposed by the Banff working group (17) do not require evidence of graft dysfunction, thus recognizing the possibility of subclinical AMR, and indeed PTC C4d staining and leukocyte margination in PTC have been identified in a small fraction of protocol biopsies of stably functioning renal allografts (26,27). Still, it has not been established whether PTC C4d deposition in such grafts, with or without accompanying capillary inflammation, is actually injurious to the graft in a similar manner to subclinical ACR (15,16). In ABO-incompatible grafts, PTC C4d is often seen on protocol biopsies in the absence of inflammation (18), and it has been suggested that C4d deposition in stably functioning ABO-incompatible grafts may be involved in graft accommodation, whereby the kidney continues to function normally despite the presence of antibodies directed against it (28–30). A potential scenario in which C4d is deposited in accommodating grafts would be if accommodation involved inhibition of the complement cascade at a site distal to C4 cleavage, such as at the level of C4b, C3 or C5 (31,32). For example, when C4b, an active cleavage product of C4, is cleaved by the complement regulatory factor I, continuation of the classical complement cascade is inhibited and C4d is left as an inactive end product (32).
Although evidence for accommodation is strongest in ABO-incompatible grafts, there is also some evidence for this in HLA-incompatible grafts. Salama et al. (33) studied 7 patients with HLA-incompatible grafts who were treated preoperatively to remove DSA, and found upregulation of Bcl-xL in glomerular endothelial cells and tubular epithelial cells in allograft biopsies from 3 of 4 patients who developed recurrence of DSA 1–10 days post-transplantation, without histologic evidence of AMR (although all 4 patients had delayed graft function). Low concentrations of anti-HLA antibody from 1 of these patients induced Bcl-xL expression in cultured human endothelial cells, which did not become activated and were resistant to complement-mediated lysis; these changes were not seen in cells exposed to saturating levels of anti-HLA antibody or to control IgGs (33). Salama et al. (33) thus proposed that low levels of anti-HLA antibody may mediate graft accommodation, in part through upregulation of Bcl-xL. Of 14 recipients of HLA-incompatible renal allografts studied by Gloor et al. (7), 2 developed clinical AMR and four subclinical AMR. While 1 of the former patients lost their graft and the other died, all 4 patients with subclinical AMR had functioning grafts 232–715 days posttransplantation, albeit with mildly elevated SCr levels (1.3–1.7 mg/dL) in 3 patients (7). While it was suggested that these patients may be undergoing graft accommodation, these workers also acknowledged that it remained unclear whether the presence of low levels of DSA and associated subclinical AMR were in fact detrimental to the graft (34).
In a recent study comparing results of 1-year protocol biopsies of HLA-incompatible, ABO-incompatible and conventional renal allografts, Gloor et al. (35) found no significant differences between the 3 groups with regard to mean tubular atrophy (ci), interstitial fibrosis (ct) and chronic vasculopathy (cv) scores, although the chronic glomerulopathy (cg) score was higher in the HLA-incompatible group than in the other 2. However, they also found that when patients from all groups were considered, mean cg, ci and cv scores on these protocol biopsies were all significantly higher in patients with 1 or more documented episodes of AMR associated with graft dysfunction than in patients without a history of AMR (35). Twenty-eight of 32 episodes of AMR were in patients with HLA- or ABO-incompatible grafts. The findings in the present study provide strong evidence that subclinical AMR is likewise associated with the development of CAN, at least in recipients of HLA-incompatible kidneys. While only 1 of our 10 such recipients who developed subclinical AMR lost their graft, and the other 9 maintained good graft function (SCr 1.0–1.8 mg/dL) 197–1361 days posttransplantation, follow-up allograft biopsies clearly indicate chronic changes in these patients that significantly exceed those seen in a control group of recipients of HLA-incompatible kidneys who did not develop AMR. Although the mean histologic follow-up in most of the 10 patients with subclinical AMR was no more than ∼1 year posttransplantation, there is clear evidence that CAN detected during the first year posttransplantation is associated with diminished long-term graft survival (36,37). Furthermore, while the effect of mild interstitial fibrosis (ci1) alone on long-term graft function is debatable, the combination of mild fibrosis with mild inflammation (i1) was found to be significantly associated with poorer graft survival (37), and indeed 7 of our 10 patients with subclinical AMR had at least mild interstitial fibrosis and inflammation on their most recent biopsy.
In that subclinical AMR appears to lead to development of CAN, a key question that remains is whether treatment of subclinical AMR can prevent or reduce this. While 3 patients in the present study were treated with PP/IVIg following the diagnosis of subclinical AMR, the number of patients in this retrospective study is clearly too small to draw any meaningful conclusions regarding the efficacy of such treatment. Particularly as transplantation involving recipients with an initially positive crossmatch becomes more commonplace, a larger and prospective study is needed to address this important issue.
In patients receiving baseline immunosuppression consisting of cyclosporine, azathioprine and prednisone, treatment of subclinical ACR during the first 6 months post-transplantation has been found to significantly reduce clinical rejection episodes occurring during the subsequent 6 months, as well as the likelihood of development of CAN in 6 months (15). This suggests that in such patients subclinical ACR can progress to clinical ACR, particularly if untreated, although findings in patients receiving baseline immunosuppression including tacrolimus and MMF [who have a lower incidence of subclinical ACR (38)] have been less convincing with respect to whether untreated subclinical ACR is associated with an increased likelihood of developing CAN (39,40). Nonetheless, our findings suggest that a significant fraction of patients with subclinical AMR will subsequently develop clinical AMR (with associated graft dysfunction). Again, the size of our patient sample is too small to determine if those patients with subclinical AMR who later develop clinical AMR are at greater risk for developing CAN and for graft loss than those who do not, and if treatment of subclinical AMR can reduce the incidence of subsequent clinical AMR, although these are also important questions that could be addressed in a larger, prospective study.
In summary, we have shown that subclinical AMR occurs in a relatively small but significant fraction of recipients of HLA-incompatible renal allografts, and that this may lead to development of CAN. Limitations of our study include the small number (10) of patients with subclinical AMR, the relatively short period (mean 335 days) of histologic follow-up after the initial diagnosis of subclinical AMR, and the inclusion of only recipients of HLA-incompatible grafts. It will therefore be important to confirm these findings in studies with larger numbers of patients (including recipients not having to undergo desensitization to remove DSA) and longer follow-up periods, and also to determine if treatment of subclinical AMR reduces CAN and improves long-term graft function in these patients.