To evaluate the role of a therapeutic regimen with plasma exchange, intravenous immunoglobulins and rituximab in chronic-active antibody-mediated rejection (cAMR) settings.

We compared 21 kidney transplant recipients (KTRs) with a diagnosis of cAMR in a retrospective case-control analysis: nine KTRs treated with plasmapheresis, intravenous immunoglobulins and rituximab (PE-IVIG-RTX group) vs 12 patients (control group) not treated with antibody-targeted therapies. We examined kidney survival and functional outcomes 24 mo after diagnosis. Histological features and donor-specific antibody (DSA) characteristics (MFI and C1q-fixing ability) were also investigated.

No difference in graft survival between the two groups was noted: three out of nine patients in the PE-IVIG-RTX group (33.3%) and 4/12 in the control group (33.3%) experienced loss of allograft function at a median time after diagnosis of 14 mo (min 12-max 18) and 15 mo (min 7-max 22), respectively. Kidney functional tests and proteinuria 24 mo after cAMR diagnosis were also similar in both groups. Only microvascular inflammation (glomerulitis + peritubular capillaritis score) was significantly reduced after PE-IVIG-RTX in seven out of eight patients (87.5%) in the PE-IVIG-RTX group (median score 3 in pre-treatment biopsy vs 1.5 in post-treatment biopsy; P = 0.047), without any impact on kidney survival and/or DSA characteristics. No functional or histological parameter at diagnosis was predictive of clinical outcome.

Our data showed no difference in the two year post-treatment outcome of kidney grafts treated with PE-IVIG-RTX for cAMR diagnosis, however there were notable improvements in microvascular inflammation in post-therapy protocol biopsies. Further studies, especially involving innovative therapeutic approaches, are required to improve the management and long-term results of this severe condition.

Chronic-active antibody-mediated rejection (cAMR) due to de novo or pre-formed donor specific antibody (DSA) is currently considered the main cause of long-term allograft losses[1,2].

From the first pilot test with intravenous immunoglobulins (IVIG) and rituximab (RTX) reported by Billing et al[3], based on the aim of reducing or eliminating DSA, some authors antagonized their detrimental effects on the graft and proposed different therapeutic regimens for cAMR treatment. All of these protocols were derived from previous experience using acute antibody-mediated rejection and desensitization protocols, and mainly consisted of steroids, plasma exchange (PE), IVIG and RTX in various modalities[4-7]. More recently, bortezomib and eculizumab were also proposed[8-10].

Specifically, an antibody-directed treatment combining high-dose IVIG and RTX showed beneficial effects [reduction in allograft losses and/or stabilization of glomerular filtration rate (GFR)] in some patients with cAMR[3-5,11], but these positive results have now been partially questioned[12-15].

The role of functional and histological parameters (i.e., GFR proteinuria at diagnosis, microvascular inflammation) in predicting response to antibody-targeted therapy has also been evaluated[6,16].

In spite of the aforementioned studies, the question of when these protocols should be adopted (in all patients or in only specific histopathological and functional settings) is still open.

In our Transplantation Center, we adopted a therapeutic protocol from 2011 that includes PE, IVIG and RTX in patients with a diagnosis of cAMR. In this paper, we compare, in a retrospective case-control analysis, nine patients treated with a combination of PE, IVIG and RTX (PE-IVIG-RTX group) for cAMR with a historical cohort of 12 kidney transplant recipients (KTRs) (control group). These control patients displayed similar histological and clinical profiles to the experimental patients, however they were not treated with antibody-targeted therapies. The primary outcome of our analysis was the difference in graft survival at 12 and 24 mo following diagnosis. Renal functional tests (including proteinuria), changes in histological features and/or DSAs-MFI, and C1q-binding ability were considered as secondary endpoints.

Twenty-one adult KTRs with a diagnosis of cAMR according to the BANFF 2015 criteria (see Histology section) were included in this retrospective study. These 21 patients included nine with a consecutive diagnosis of cAMR from January 1, 2011 to December 31, 2014 who were treated with PE, IVIG and RTX (PE-IVIG-RTX group), and 12 KTRs with the same consecutive diagnosis performed in the period between January 2009 and December 2012 (control group). In that early period, antibody-targeted therapies were not currently adopted, or patients did not give their consent to these therapies.

At the time of diagnosis, patients were treated with a CNI-based immunosuppression (28.6% Cyclosporine A, 71.4% Tacrolimus, equally distributed into two groups), with Mycophenolate Mofetil/Mycophenolic Acid (77.8% in the PE-IVIG-RTX group and 66.7% in the control group) or an mTOR inhibitor drug (11.1% in the PE-IVIG-RTX group and 37.3% in the control group). Azathioprine was used only in one patient in the PE-IVIG-RTX group, and 77.8% of patients in the PE-IVIG-RTX group vs 66.7% in the control group were treated with steroids, respectively.

After cAMR diagnosis, maintenance therapy was reinforced in both groups by either introducing MMF and/or steroids, (with contemporary suspension of the mTOR inhibitor drug, if used) or switching from Cyclosporine A to Tacrolimus.

The PE-IVIG-RTX schedule was defined as follows: (1) Four or five PE (one plasma volume removal and 5% Albumin or plasma infusion) sessions in the first two weeks, (2) subsequent high-dose 2 g/kg IVIG (in one or two days), and (3) intravenous RTX (375 mg/m², one dose) after IVIG. Three patients in both groups also received steroid boluses after diagnosis (4 mg/kg methylprednisolone, tapered in five to seven days with a total steroid dose of about 1.5 g). One patient in the PE-IVIG-RTX group received a second RTX dose (375 mg/m²) because of a concomitant diagnosis of membranous nephropathy.

Renal function was measured by serum creatinine (sCr) and GFR (estimated using the Cockroft-Gault formula). Patients were also tested repeatedly pre-transplantation for anti-HLA antibodies using the panel reactive lymphocytotoxicity assay, and maximum values from this assay were considered for our analysis.

We obtained an informed consent about potential complications and adverse events from all treated patients.

All biopsies were performed for cause, i.e., in case of a significant and/or unexplained increase of serum creatinine > 25% from baseline, proteinuria, or both. Biopsies were reviewed according to the Banff 2015 classification[17], and only patients with a diagnosis of cAMR meeting all the requested criteria were included in this study. These criteria are as follows: (1) Histologic evidence of chronic tissue injury (transplant glomerulopathy - expressed by a cg score > 0, and/or severe peritubular capillary basement membrane multilayering, and/or arterial intimal fibrosis of new onset; (2) evidence of antibody-endothelium interaction [C4d > 0 in paraffin sections of peritubular capillaries and/or microvascular inflammation (MVI) – expressed by a g + ptc score ≥ 2, considering that in the presence of acute TCMR, borderline infiltrate, or infection, g must be ≥ 1]; and (3) serologic evidence of DSAs. We also evaluated a chronicity score (ci + ct), as reported by other authors[18].

In the PE-IVIG-RTX group, we also performed a protocol kidney biopsy at a median time of ten months after therapy (as discussed below in the Results section) in order to assess histopathological improvement when present.

Sera were evaluated twice, at both the time of biopsy and after 12 mo. As discussed in our previous paper[19], we tested all sera with a Luminex platform and commercially-available SAB kits (LABScreen One Lambda, Canoga Park, CA, United States) in order to identify HLA Classes I and II IgG DSA. Sera were also studied with the C1qScreen (One Lambda) to assess DSA complement-fixing ability. The cut-off was set at the normalized MFI value of 1000 for both tests.

In this study, we performed retrospective case-control analysis to study the mid-term clinical outcomes (24 mo) in 21 KTRs with a diagnosis of cAMR. We compared nine patients treated with PE, IVIG and RTX with a historical cohort of 12 patients who featured similar clinical and histological characteristics yet did not receive these antibody-targeted therapies.

Our data showed no clinical improvement after therapy with PE-IVIG-RTX, either in graft survival or in renal functional tests. In addition, proteinuria values were not influenced by the treatment.

On the contrary, upon evaluating histological features in protocol biopsies after PE-IVIG-RTX, microvascular inflammation (estimated by g + ptc score) was found to improve after PE-IVIG-RTX treatment. These data are quite similar to what was observed in the RITUX-ERAH trial in patients with acute AMR who were treated with PE, IVIG and steroids, either in association or not in association with RTX[18]. In Muller’s paper[15], patients treated for cAMR with only Rituximab improved in g + ptc score after one year. Despite different histological settings (acute AMR in tge RITUX-ERAH trial vs cAMR in our study and in Muller et al[15]) and different follow-ups (12 mo in the RITUX-ERAH trial and in Muller et al[15] vs 24 mo in our study), the evidence for an improvement in renal histology was not supported by an amelioration in kidney survival at a mid-term follow-up.

As for DSA, a lowering effect was not obtained in all patients (the median value was unchanged after treatment). These data may suggest that, in the context of chronic antibody production, the B cell target for PE-IVIG-RTX may elude the RTX effect and is likely represented by CD20-negative cells, as previously reported by other authors[12,20]. In two patients, we observed no DSA detection after treatment, although this was in association with highly different functional data (stabilization of GFR in one patient, graft failure in the other one).

No significant difference was noted in pre- and post-treatment C1q-fixing ability, or in DSA fixing complement ability at diagnosis. In addition, the clinical outcomes were similar at 24 mo. Our analysis is underpowered for the evaluation of DSA C1q-fixing ability as a marker of severe cAMR, which was positively reported in a larger cohort study[21]; however, we have recently observed in 35 KTRs with a transplant glomerulopathy diagnosis and de novo DSA (dnDSA) that a higher percentage of patients with dnDSA-associated transplant glomerulopathy was C1q-negative, and that the presence of C1q-fixing dnDSA did not significantly correlate with graft outcome19.

We are aware that the lack of difference in the immunodominant DSA-MFIs before and after treatment may be due to technical limitations related to the “prozone” effect[22]. However, it is remarkable that the MFI titer in three patients increased after treatment and that in 6/9 it remained higher than 3000, a threshold value considered by several centers.

We also evaluated functional, histological and immunological parameters at diagnosis to detect potential risk factors for allograft loss. In the control group, we found a trend towards a higher DSA-MFI titer, C1q-fixing DSA positivity, a higher sCr, and a lower GFR. On the contrary, the histological findings at diagnosis showed no significant difference between failed and unfailed grafts at 24 mo in both groups.

Based of our analysis, we are unable to define any characteristics at diagnosis that influence prognosis. The goal of any study on this topic should be to identify a certain population who would benefit from therapy (in this case Rituximab associated with PE and IVIG). Unfortunately, no study has fulfilled this scope to the best of our knowledge[15,16]. The search for characteristics that label the population that would benefit from these therapies is even more important when we consider the significant risk associated with these therapies. In our study, we noted a significant increase in the bacterial infection rate in the PE-IVIG-RTX group (OR: 4, 1.7-9.3).

Upon comparing our results to the literature data, Bachelet et al[13] also reported no improvements in graft survival or renal functional tests in 21 patients with cAMR-associated severe transplant glomerulopathy who received IVIG and two doses of RTX. Similar outcomes (no differences in eGFR decline, increase of proteinuria, Banff scores at one year, or MFI of the immunodominant DSA) were also shown in a very recent randomized clinical trial evaluating efficacy and safety of IVIG combined with RTX in 25 patients with cAMR[14].

All these data are in contrast with previous evidence from Billing’s paper, showing a GFR improvement or stabilization at 12 mo in four out of six pediatric patients who were IVIG and RTX treated[3]. A subsequent analysis of 20 pediatric patients, published by the same author, reported a lower median GFR loss in the 24 mo follow- up after IVIG and RTX, compared with GFR loss in the 6 mo prior to treatment[23]. When also excluding differences between pediatric and adult KTRs, and the absence of a control group in the two studies by Billing et al[23], it is clear that a minor GFR-worsening might not result from a therapeutic effect, but instead represent the natural history of the disease and its early diagnosis.

In a retrospective analysis, Redfield et al[2] examined 123 patients with severe cAMR; Kaplan–Meier survival showed an association of steroids/IVIG (together or in combination with rituximab and/or Thymoglobulin) with better graft survival. However, the association between the addition of rituximab or Thymoglobulin to steroids/IVIG with better graft survival did not reach statistical significance.

We acknowledge the limitations of our study, which include the low numerosity, the retrospective design, and the absence of protocol biopsies in the control group. Nonetheless, a low number of treated subjects, the absence of a control group, and retrospective analysis can be found in most studies that involve treatment of this clinical condition[2,4,5,13]. Moreover, we also recognize that three patients in both groups were also treated with steroid boluses in low doses. This observation may be considered as a bias in interpretation due to a possible “positive” effect in the control group, however this may also be seen as a negligible aspect since two out of two of these patients lost their graft.

We recognize that protocol biopsies could have enlightened the question as to whether early lesions could be a marker for a better response to treatment. The absence of protocol biopsies in the control group precludes an adequate histological comparison between the populations. We are therefore able to compare the histopathological findings inside the treatment group, but we are unable to evaluate the progression of the chronic lesions in the control group. However, protocol biopsies are not a current practice for some centers, and cAMR is often diagnosed only after appearance of clinical abnormalities that trigger biopsy indication.

Regarding microvascular inflammation lesions, which are considered to be crucial for disease progression[6,24], we found a reduction in g + ptc score after treatment with PE-IVIG-RTX. One could speculate that if the amelioration of these lesions have a significant clinical impact, it could potentially be noted in a longer follow-up.

In conclusion, no guidelines about the therapeutic management of cAMR is currently available. Our data, along with the results of other groups[12-15], suggest the lack of a prompt and marked effect of a therapeutic protocol with PE, IVIG and RTX, despite good histological improvement (reduction in microvascular inflammation) in the majority of treated patients. It is possible that this treatment could have greater efficacy with a longer follow-up, or in a subset of patients not yet identified, as suggested by other authors[15,16]. Further prospective studies, especially involving innovative therapeutic approaches, are required to improve both the management and long-term results of this severe condition.