#NephJC Chat
Tuesday, April 29, 2025, 9 pm Eastern on Bluesky
N Engl J Med. 2025 Feb 20;392(8):763-776.doi: 10.1056/NEJMoa2408835. Epub 2024 Oct 24.
Microvascular Inflammation of Kidney Allografts and Clinical Outcomes
Marta Sablik, Aurélie Sannier, Marc Raynaud, Valentin Goutaudier, Gillian Divard, Brad C Astor, Patricia Weng, Jodi Smith, Rouba Garro, Bradley A Warady, Rima S Zahr, Katherine Twombley, Vikas R Dharnidharka, Raja S Dandamudi, Marc Fila, Edmund Huang, Anne-Laure Sellier-Leclerc, Burkhard Tönshoff, Marion Rabant, Jérôme Verine, Arnaud Del Bello, Thierry Berney, Olivia Boyer, Rusan Ali Catar, Richard Danger, Magali Giral, Daniel Yoo, François R Girardin, Alaa Alsadi, Pierre-Antoine Gourraud, Emmanuel Morelon, Moglie Le Quintrec, Mélanie Try, Jean Villard, Weixiong Zhong, Oriol Bestard, Klemens Budde, Bertrand Chauveau, Lionel Couzi, Sophie Brouard, Julien Hogan, Christophe Legendre, Dany Anglicheau, Olivier Aubert, Nassim Kamar, Carmen Lefaucheur, Alexandre Loupy
PMID: 39450752
Introduction
Antibody-mediated rejection (AMR), depending upon your definition, complicates 5 - 40% of kidney transplants, and reduces graft survival. Early AMR can be especially problematic, and at least one study showed decreased 5-year graft survival in allografts with AMR of 76% versus 88% in matched controls without AMR (Orandi et al, Am J Transplant 2015).
Antibody-mediated rejection is characterized by a triad of donor-specific antibodies (DSAs), C4d staining along peritubular capillaries, and microvascular inflammation (MVI). C4d staining represents complement degradation products due to activation on endothelium, and is highly associated with the presence of DSAs, but they can also be identified separately. Donor-specific antibodies are not always due to antibodies against HLA antigens and instead may have other targets, leading to “false negative” DSA testing. Microvascular inflammation consists of infiltration of inflammatory cells within glomerular capillaries (glomerulitis) or within peritubular capillaries (peritubular capillaritis). C4d can be negative within peritubular capillaries in the setting of AMR (Haas et al, Am J Transplant. 2014) or there can be a lack of donor specific antibodies, and therefore, one of the two components (DSA+ or C4d+) are required for a diagnosis of antibody-mediated rejection, in the presence of microvascular inflammation. These entities may not be equal, as patients who are DSA+ show inferior graft survival compared to those who are DSA-negative (Callemeyn et al, J Am Soc Nephrol 2020; Senev et al, Am J Transplant 2019).
The third possibility is the presence of isolated microvascular inflammation in the absence of donor-specific antibodies and C4d staining throughout peritubular capillaries (MVI+ with DSA- C4d-). This provisional entity creates a diagnostic dilemma, which was previously classified as “No AMR” in 2019, but has subsequently been given its own category in the 2022 Banff criteria. The Banff Consortium called for future research to determine the significance of microvascular inflammation in the absence of DSAs or C4d positivity, the focus of this study.
Figure 2. Flowchart of the Banff 2022 Classification, from Naesens et al, Am J Transplant, 2024
Early work (Gupta et al, Kidney Int. 2016; de Kort et al, Am J Transplant 2013) demonstrated that a higher MVI score not only correlates with a histologic diagnosis of AMR, independent of C4d staining, but also associates with increased expression of pathogenesis-based gene transcripts, including those related to cytotoxic T cells, natural killer cells, and antibody-mediated injury pathways. MVI was found to be required for an AMR signature, which was not present in patients with positive DSAs and C4d positivity in the absence of MVI (Harmaceck et al, Am J Transplant 2024). These findings reinforce that MVI is not merely a histologic observation but a potent, independent predictor of graft dysfunction and loss.
More recently, as summarized in NephJC's Top Stories of 2024 and discussed during Freely filtered, that MVI even when solo, carries a clinically meaningful risk of graft dysfunction, recurrent AMR, and development of transplant glomerulopathy. Transplant glomerulopathy itself has a severe impact on allograft survival, with >70% graft loss within 5 years (Patri et al, Kidney Int 2016). This further underscores the urgent need to refine and standardize MVI assessment in kidney transplant.
Given this evolving understanding, Sablik et al, NEJM 2025 conducted a large cohort study to better characterize the prognostic significance of isolated MVI in kidney allografts. By reclassifying biopsies and analyzing graft outcomes, the study aimed to determine whether isolated MVI represents a distinct clinical entity with implications for graft survival, disease recurrence, and chronic kidney injury.
The Study
Methods
Study Design: This was a large, multicenter, population-based retrospective cohort study including 6798 kidney transplant recipients from more than 30 centers in Europe and North America who underwent allograft biopsy between 2004 and 2023. Biopsies were re-classified according to the Banff 2022 criteria, which include the new categories of ‘Probable AMR’ and ‘MVI DSA-C4d-‘.
Data sources: The study used data from five cohorts, including: The Paris Transplant Group, the University of Wisconsin–Madison, NIH Pediatric Transplant cohort, and two clinical trials: KTD-Innov and EU-TRAIN study. Data included clinical, demographic, immunologic, and histopathological variables.
Diagnostic Classification and Reclassification of Biopsy Specimen: Biopsies were reviewed using the Banff classifications per 2019 and 2022 criteria, using a validated computer-based diagnostic automation system, assigning diagnostic labels across categories, including non-rejection, T-cell–mediated rejection (TCMR), antibody-mediated rejection (AMR), mixed rejection, and focusing on two novel phenotypes: MVI (microvascular inflammation) DSA-negative/C4d-negative, and probable AMR. This re-classification included 641 biopsies that would be under the category of “no rejection” by the previous 2019 Banff criteria.
Outcome: Microvascular inflammation was defined according to the 2022 Banff classification, and graft loss as the return to dialysis or preemptive retransplantation. Secondary outcomes were new or recurrent AMR and transplant glomerulopathy.
Table S4. Criteria Leading to Diagnoses According to the Banff 2022 Classification, from Sablik et al, NEJM, 2025
Statistical Methods: The authors generated a Sankey diagram to visually illustrate diagnostic reclassification between the 2019 and 2022 Banff criteria, highlighting shifts in biopsy interpretations across the two classification systems. For outcome analysis, Kaplan–Meier survival curves and Fine–Gray competing-risk models were employed to assess graft loss, recurrent antibody-mediated rejection (AMR), and transplant glomerulopathy. Initial evaluations were conducted using univariable Cox proportional hazards models. Additionally, subgroup analyses stratified outcomes based on the severity of glomerulitis and peritubular capillaritis (g + ptc scores), the indication for biopsy (protocol vs. for-cause), and the presence of donor-specific antibodies (DSA).
Funding Sources: The study was supported by the nonprofit organization OrganX. The KTD-Innov study was supported by the French government through a grant from the Investissements d’avenir program of the French National Research Agency. The EU-TRAIN study was funded by a grant from the European Union Horizon 2020 Research and Innovation Program. The AI-Care project was funded by a grant from the European Union 2023 European Research Council Consolidator Grant program.
Results
This multicenter study included a diverse cohort of 6798 kidney transplant recipients from 16,293 allograft biopsy specimens collected over nearly two decades. Biopsies were performed for clinical indications or protocol surveillance, with a median biopsy timing of 8.1 months post-transplantation.
Figure S1. Study flowchart, from Sablik et al, NEJM, 2025
Overall, the patients' mean age was 44.6 years, 38.6% female, 78.6% of transplants were from deceased donors, 25.1% of patients had donor-specific antibodies (DSA), and 12.7% were retransplanted patients. Common causes of end-stage kidney disease were glomerulonephritis (22.5%), autosomal dominant polycystic kidney disease (14%), and diabetes (12.3%). These characteristics reflect a clinically diverse population with broad transplant exposures.
Adapted from table 1. Characteristics of Kidney-Transplant Recipients, Donors, and Biopsy Specimens, from Sablik et al, NEJM, 2025
Notably, 788 biopsies (4.8%) were newly categorized as MVI phenotypes under the 2022 Banff Classification, of which 641 were previously considered non-rejection under the 2019 criteria.
Figure 1. Reclassification of Banff 2019 → Banff 2022 diagnoses, from Sablik et al, NEJM, 2025. Note the change in the patients diagnosed with “no rejection” to those with the new categories of “Probable AMR” and “MVI, DSA negative and C4d negative”.
In terms of histologic findings in these MVI phenotypes included glomerulitis (up to 84.1%) and peritubular capillaritis (up to 85.8%), with chronic injury markers (fibrosis, atrophy) equally prevalent across rejection and non-rejection categories.
Figure S7. Cumulative Incidence of Development or Progression of Transplant Glomerulopathy Stratified on g and ptc Score. The greater the degree of MVI, the greater the odds of developing transplant glomerulopathy, from Sablik et al, NEJM, 2025
Among 4374 patients analyzed for graft survival over a median 5-year follow-up, 11.2% experienced graft loss. Those reclassified with MVI DSA−/C4d− had a 2.1-fold higher risk for graft loss compared with non-rejection diagnoses.
Figure 2. Graft survival across rejection phenotypes, from Sablik et al, NEJM, 2025
Patients with probable AMR had a 1.3-fold increased risk by year 5, which rose to 1.7 beyond five years. The highest risk was observed in classic AMR patients, with a hazard ratio of 2.7.
Interestingly, the “MVI DSA-C4d-“category had a worse impact on prognosis compared to “Probable AMR”. Ultimately, this highlights that “MVI DSA-C4d-“ should not be seen as “no rejection”, as from the prior Banff classification.
Of 5235 patients without previous rejection, 8.1% developed new or recurrent AMR during follow-up. MVI DSA−/C4d− and probable AMR groups exhibited intermediate AMR incidence, between non-MVI and active AMR patients. Their subdistribution hazard ratios were similar (around 0.7; 95% CI, 0.4 to 1.2), suggesting overlapping clinical behavior and progression potential between the two new categories.
Figure 3A. Cumulative incidence of new/recurrent AMR among MVI subtypes, from Sablik et al, NEJM, 2025
Transplant glomerulopathy developed in 6.7% of the 5235 recipients assessed. The cumulative risk of glomerulopathy in the MVI DSA−/C4d− group paralleled that of the probable AMR group and was consistently lower than in the active AMR group. Stratified analyses also showed that higher Banff g and ptc scores were associated with increased risk for both AMR and glomerulopathy progression, reinforcing their prognostic value.
Discussion
This study reclassified over 16,000 kidney allograft biopsies using the 2022 Banff criteria, making a major advancement in the understanding of microvascular inflammation (MVI) phenotypes in kidney transplantation. The development of two new diagnostic categories: MVI, DSA-negative and C4d-negative, and probable antibody-mediated rejection, demonstrates how histopathologic frameworks can evolve to uncover the relevant injury patterns that were previously miscategorized; a finding also reflected by Nankivell et al, Am J Transplant 2025.
Prior cohorts, including Nankivell et al, Transplantation 2025, established that the late appearance of MVI is independently associated with increased risks of graft failure and progression to transplant glomerulopathy. Similarly, other studies (Coemans et al, Transplant Int 2021, Parajuli et al, Transplantation 2019) consistently suggest that persistent inflammatory patterns significantly impacted long-term graft outcomes.
New approaches have revealed that cg+MVI DSA-/C4d- have a distinct immune profile, showing a T-cell dominant mechanism rather than purely antibody-mediated injury, challenging existing AMR classifications (Cristoferi et al, Kidney Int. 2024).
Often, at the time of biopsy diagnosis, DSA testing results are often not available. The presence of MVI in the absence of C4d without knowledge of DSA status is a challenging clinical scenario for the nephrologist. These data suggest that MVI on biopsy alone, even if DSAs end up being negative, is significant and may warrant treatment.
Strengths
Large multicenter cohort.
Automation of Banff classification for more diagnostic consistency.
Application of risk models to evaluate graft outcomes.
Limitations
Lack of multivariable adjustment.
Potential selection bias, as biopsies were performed for clinical indication.
Missing treatment data on how strategies influence outcomes.
Conclusion
In conclusion, these findings provide important insights into the prognostic significance of microvascular inflammation in kidney transplantation. However, the variability in study designs, MVI definitions, and outcome measures across cohorts emphasizes the urgent need for standardized diagnostic criteria and management strategies. Optimal management strategies and response to treatment of MVI alone is yet to be determined.
Future research integrating histology, molecular diagnostics, and clinical decision-making will be essential to refine the role of MVI in guiding personalized transplant care.