Skip to main content
Journal of the American Society of Nephrology : JASN logoLink to Journal of the American Society of Nephrology : JASN
editorial
. 2026 Oct 12;37(11):2370–2373. doi: 10.1681/ASN.0000001275

Microvascular Inflammation in Kidney Transplants

More Heterogeneous Than We Thought?

Alexandre Loupy 1,2,✉, Michael Mengel 3
PMCID: PMC13742995  PMID: 42862641

Antibody-mediated rejection (ABMR) remains a leading cause of kidney allograft failure, and microvascular inflammation, defined by glomerulitis (g) and peritubular capillaritis (ptc), is its histologic hallmark. Although microvascular inflammation was first characterized as an antibody-driven lesion, it is frequently encountered without circulating donor-specific anti-HLA antibodies (DSAs), and multiple studies have reported that it carries prognostic weight independent of the underlying pathomechanism.1–5 The immune drivers of microvascular inflammation are increasingly defined, implicating natural killer cells, T cells, and myeloid effectors. A recent randomized phase 2 trial of the anti-CD38 antibody felzartamab suggests that microvascular inflammation in ABMR may respond to targeted treatment.6 The clinical need to accurately identify patients with microvascular inflammation who might respond to such novel therapies, even without DSA, led to the introduction of a dedicated phenotype: microvascular inflammation/injury and DSA negative and C4d negative in the Banff classification in 2022.7

The study by Wong et al. in this issue of JASN8 retrospectively investigated a single-center cohort of 689 kidney transplant recipients transplanted between 2004 and 2021. The authors studied the first occurrence of 3 alloimmune events, T-cell–mediated rejection (TCMR), g+ptc≥2 (microvascular inflammation), and de novo DSA (dnDSA), as time-dependent covariates in Cox models for death-censored allograft loss. Assessed individually, each event carried a roughly 4-fold hazard (TCMR hazard ratio [HR], 4.07; 95% confidence interval [CI], 2.34 to 7.08; g+ptc≥2 HR, 4.33; 95% CI, 2.49 to 7.54; dnDSA HR, 4.26; 95% CI, 2.18 to 8.29). In a combined model, TCMR (HR, 2.74) and dnDSA (HR, 2.32) remained independently associated with graft loss, whereas g+ptc≥2 did not (HR, 1.77; 95% CI, 0.84 to 3.73). The authors conclude that microvascular inflammation without DSA does not independently worsen prognosis after adjustment for serial or concomitant TCMR and dnDSA events.

This conclusion runs against the weight of the published evidence from many independent cohorts that converged on the finding that microvascular inflammation, regardless of whether an anti-HLA antibody–mediated cause is detectable, carries an adverse prognosis and elevated risk of progression to transplant glomerulopathy1–5 (see Table 1).

Table 1.

Studies assessing the effect of microvascular inflammation (glomerulitis+peritubular capillaritis) on kidney allograft loss

Study Design/Setting N Microvascular Inflammation Variable Tested Outcome Independent Effect Study Summary
Einecke et al., 2009
Am J Transplant
PMID: 19843030
Single-center, late indication biopsies 234 pts g+ptc Graft loss (death-censored) Yes Microvascular inflammation independently associated with graft failure after adjusting for IFTA, anti-HLA DSA, and C4d. TCMR alone not associated with graft loss
Loupy et al., 2011
Am J Transplant
PMID: 21199348
Single-center protocol biopsies 80 pts
157 bx
g+ptc at 3 mo after transplant (highly sensitized patients) Progression to chronic ABMR Yes Microvascular inflammation independently associated with progression to chronic ABMR
Independent of C4d; class II DSA. 55% of C4d-negative biopsies had concomitant microvascular inflammation
Gupta et al., 2016
Kidney Int
PMID: 26422506
Single-center, indication+protocol biopsies 745 bx g+ptc, including DSA-negative cases Graft loss (death-censored) Yes Microvascular inflammation independently associated with graft loss after adjustment for anti-HLA DSA status (de novo); molecular correlates support alloimmune injury in HLA DSA-negative cases
Loupy et al., 20191
BMJ
International derivation+2 external validations, 10 centers in Europe and the United States 7557 pts g+ptc adjusted for HLA DSA (preexisting and de novo), TCMR, IFTA, GFR, P/C ratio, time after transplant, cg Graft loss (death-censored) Yes Microvascular inflammation retained as an independent prognostic factor alongside eGFR, proteinuria, IFTA, TCMR, C4d, cg, and anti-HLA DSA. Microvascular inflammation independent predictor of allograft failure confirmed in independent external cohorts from Europe and the United States. C-index 0.81 (derivation), 0.81 (Europe), 0.80 (the United States)
Senev et al., 2019
Am J Transplant
PMID: 30107078
Single-center 208 ABMRh ABMRh without HLA-DSA Graft loss (death-censored) No DSA-negative ABMRh not at higher risk than no-ABMRh; lesions often transient. 10-yr survival 76% (DSA-negative) versus 57% (DSA-positive). Frequent overlap with TCMR (38%)/borderline (18%)
Buxeda et al., 2023
Am J Transplant
PMID: 36710135
Single-center 220 pts Isolated microvascular inflammation, DSA-negative and C4d-negative Graft loss (death-censored) Yes Grouped survival comparison isolated microvascular inflammation 77% versus ABMR 63% versus normal 93% (P < 0.001)—intermediate phenotype with cytotoxic T-cell and NK cell infiltration
Hogan et al., 20233
Am J Transplant
International observational, 20 institutions in Europe and the United States 706 pts (pediatric) g+ptc Graft loss (death-censored) Yes Microvascular inflammation independent factor associated with allograft failure in international pediatric cohorts. Calibration and discrimination preserved to 10 yr
Klein et al., 20232
Transpl Int
Regulatory qualification analysis EMA submission Pooled cohorts+additional centers (Helsinki+Mayo Clinic) g+ptc Graft loss (death-censored) Yes Regulatory acceptance and qualification of a model in which microvascular inflammation is a retained independent covariate predicting allograft loss. EMA qualification December 2022
Nankivell et al., 2025
Am J Transplant
PMID: 39084463
Single-center, protocol+indication biopsies 1200 pts
3398 bx
g+ptc≥2 by etiology Graft loss (death-censored) Yes All microvascular inflammation etiologies (including pure TCMR) exceeded matched normals. GECE >50% HR 2.0 (95% CI, 1.4 to 2.9); basement-membrane duplication HR 1.6 (95% CI, 1.2 to 2.3) on multivariable analysis
Sablik et al., 20254
N Engl J Med
Multicenter, >30 centers in Europe and North America; Banff 2022 reclassification 6798 pts
16,293 bx
g+ptc, DSA-negative and C4d-negative Graft loss (death-censored)
TG progression
Yes Microvascular inflammation DSA−/C4d− HR 2.1 (95% CI, 1.5 to 3.1); ABMR HR 2.7 (95% CI, 2.2 to 3.3). Graft-loss HRs are stratified by reclassified diagnosis, not adjusted for concomitant parameters
Wong et al., 20268
J Am Soc Nephrol
Single-center retrospective 689 pts g+ptc≥2 (g=0 in 66%); microvascular inflammation/DSA−/C4d− subset n=40 Graft loss (death-censored) No Time-dependent Cox: g+ptc≥2 HR 1.77 (95% CI, 0.84 to 3.73) versus TCMR 2.74 and dnDSA 2.32
Raynaud et al., 20265
BMJ Med
International multicohort validation (21 centers) 12,683 pts g+ptc Graft loss (death-censored) Yes Microvascular inflammation (g+ptc) independent factor associated with allograft loss after adjustment for anti-HLA DSA (preexisting and de novo), TCMR, IFTA, GFR, proteinuria, time, cg

Superscript numbers refer to the reference list; for studies not included in the reference list, the PubMed identifier is given in the first column. Death-censored graft loss denotes return to dialysis or retransplantation, with death with a functioning graft treated as a censoring event. ABMR, antibody-mediated rejection; ABMRh, antibody-mediated rejection histology; bx, biopsies; C4d, complement split product C4d; cg, transplant glomerulopathy Banff score; CI, confidence interval; dnDSA, de novo donor-specific anti-HLA antibody; DSA, donor-specific anti-HLA antibody; EMA, European Medicines Agency; g, glomerulitis; GECE, glomerular endothelial cell enlargement; HR, hazard ratio; IFTA, interstitial fibrosis and tubular atrophy; NK, natural killer; P/C, protein-to-creatinine; ptc, peritubular capillaritis; pts, patients; TCMR, T-cell–mediated rejection; TG, transplant glomerulopathy.

The finding by Wong et al. that 88% of recipients with a g+ptc≥2 event also had preceding, concomitant, or subsequent TCMR or dnDSA is an important observation, adding to our understanding of the heterogeneous phenotype of microvascular inflammation. What a g+ptc≥2 event signifies, however, depends on which alloimmune process it accompanies, and in this cohort, that process was predominantly TCMR rather than ABMR.

Microvascular inflammation occurred in 15% of patients in this single-center cohort of immunologically low-risk recipients. The authors classified all g+ptc≥2 patient cases by concomitant diagnosis (Table 2 of Wong et al.), but their “microvascular inflammation” label is mostly ptc only with g=0 in 66%, hence microvascular inflammation with TCMR really means ptc ≥2 with no glomerulitis. Moreover, outcome analyses only ever test microvascular inflammation in the context of TCMR (Figure 3 of Wong et al. for the time-dependent models). ABMR-associated microvascular inflammation is essentially not analyzed as an outcome group with only 5 ABMR-alone patient cases, thus preventing an actual comparison of TCMR- versus ABMR-associated microvascular inflammation.

The study cohort does not include presensitized patients, who are known to be at higher risk of early ABMR, and 42% of the analyzed biopsies are protocol biopsies, which more often reveal subclinical TCMR than ABMR. Accordingly, 53 of the 106 microvascular inflammation biopsies had g=0 and ptc≥2 in the setting of borderline/TCMR, and these TCMR-associated g+ptc≥2 events occurred early (median 5.8 months after transplant). By contrast, the 5% (5/106) of microvascular inflammation patient cases that were seen in the context of ABMR alone occurred much later (median 32.8 months). In other words, microvascular inflammation accompanying TCMR/borderline seems to be a different phenotype than microvascular inflammation accompanying ABMR. This overlap of microvascular inflammation between TCMR and ABMR is well recognized and reflected in the current Banff rules for diagnosing TCMR and ABMR: For the ABMR/microvascular inflammation categories, when borderline or acute TCMR (or infection) is present, g+ptc≥2 is insufficient to diagnose microvascular inflammation and g ≥1 is required because ptc cannot be confidently attributed to antibody-mediated injury in the presence of tubulointerstitial inflammation.

Regardless of how microvascular inflammation is defined in this study, the null finding for g+ptc≥2 rests on limited power. Its independent hazard is 1.77 (95% CI, 0.84 to 3.73), an interval wide enough to encompass both no effect and clinically important harm; by the authors' own Freedman calculation, roughly 2500 recipients would have been needed to detect a hazard of 1.7 with 80% power, against the 689 studied. With only 60 death-censored losses, the study seems underpowered for its primary research question, and the DSA-negative, C4d-negative subanalysis, resting on just 40 patient cases, is insufficient to support a robust conclusion, a limitation the authors acknowledge. These same 60 events support roughly 20 estimated parameters, approximately 3 events per parameter, well below the conventional threshold of 10 events per variable.

Moreover, isolating an independent effect of microvascular inflammation requires patients in whom microvascular inflammation and TCMR can be distinguished, thus cases where one occurred without the other. Such separation is intrinsically difficult because g+ptc≥2 is recognized to co-occur with TCMR, and in this cohort, that overlap is further increased by a broad g+ptc≥2 definition that also counts ptc-driven patient cases with g=0 arising alongside TCMR. The independent microvascular inflammation effect is therefore mostly estimated from the small subset with isolated microvascular inflammation.

Beyond the statistics, there is a biological reason to think microvascular inflammation is not inconsequential. Recent mechanistic and spatial studies show that the infiltrate in microvascular inflammation can be composed mainly of monocytes, macrophages, and natural killer cells, an observation consistent with an alloimmune basis for HLA mismatch–associated microvascular inflammation. The authors' own finding that HLA-DR/DQ molecular mismatch independently predicts g+ptc≥2-free survival echoes prior reports linking HLA mismatch to DSA-negative microvascular inflammation.9 Together, these support microvascular inflammation as primarily an alloimmune process rather than an epiphenomenon. Mechanistic insights into natural killer cell–mediated missing self-recognition,10 T-cell transcriptional signatures,11 and myeloid allorecognition continue to clarify why these lesions matter and how they might be treated in a targeted fashion, independent of the underlying cause.

The felzartamab trial indicates that ABMR with microvascular inflammation is amenable to treatment.6 Determining which other microvascular inflammation phenotypes might respond will require microvascular inflammation to be defined clearly and consistently across studies. The present study illustrates the complexity: Much of what is captured here as g+ptc≥2 rests on ptc alone, with glomerulitis absent, in the setting of TCMR. Consistent application of a standardized definition is therefore a prerequisite for the trials that will determine who benefits from emerging therapies. Adequately capturing this microvascular inflammation heterogeneity will require large, multicenter cohorts spanning a broad immunologic risk spectrum, to better define the different microvascular inflammation phenotypes and evaluate their response to emerging therapies.4 The study by Wong et al. highlights why such effort is needed.

Supplementary Material

jasn-37-2370-s001.pdf (1.4MB, pdf)

Acknowledgments

The content of this article reflects the personal experience and views of the authors and should not be considered medical advice or recommendation. The content does not reflect the views or opinions of the American Society of Nephrology (ASN) or JASN. Responsibility for the information and views expressed herein lies entirely with the authors.

Footnotes

See related article, “Microvascular Inflammation in Kidney Transplantation,” on pages 2529–2537.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at https://links.lww.com/JSN/G36.

Author Contributions

Writing – original draft: Alexandre Loupy, Michael Mengel.

Writing – review & editing: Alexandre Loupy, Michael Mengel.

Funding

None.

References

  • 1.Loupy A Aubert O Orandi BJ, et al. Prediction system for risk of allograft loss in patients receiving kidney transplants: international derivation and validation study. BMJ. 2019;366:l4923. doi: 10.1136/bmj.l4923 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Klein A Loupy A Stegall M, et al. Qualifying a novel clinical trial endpoint (iBOX) predictive of long-term kidney transplant outcomes. Transpl Int. 2023;36:11951. doi: 10.3389/ti.2023.11951 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Hogan J Divard G Aubert O, et al. Validation of a prediction system for risk of kidney allograft failure in pediatric kidney transplant recipients: an international observational study. Am J Transplant. 2023;23(10):1561–1569. doi: 10.1016/j.ajt.2023.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Sablik M Sannier A Raynaud M, et al. Microvascular inflammation of kidney allografts and clinical outcomes. N Engl J Med. 2025;392(8):763–776. doi: 10.1056/NEJMoa2408835 [DOI] [PubMed] [Google Scholar]
  • 5.Raynaud M Truchot A Naser S, et al. Qualified prediction system for allograft failure in real world settings: extended validation study. BMJ Med. 2026;5(1):e002088. doi: 10.1136/bmjmed-2025-002088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Mayer KA Schrezenmeier E Diebold M, et al. A randomized phase 2 trial of felzartamab in antibody-mediated rejection. N Engl J Med. 2024;391(2):122–132. doi: 10.1056/NEJMoa2400763 [DOI] [PubMed] [Google Scholar]
  • 7.Naesens M Roufosse C Haas M, et al. The Banff 2022 kidney meeting report: reappraisal of microvascular inflammation and the role of biopsy-based transcript diagnostics. Am J Transplant. 2024;24(3):338–349. doi: 10.1016/j.ajt.2023.10.016 [DOI] [PubMed] [Google Scholar]
  • 8.Wong ETY Balshaw R Gibson IW, et al. Microvascular inflammation in kidney transplantation. J Am Soc Nephrol. 2026;37(11):2529–2537. doi: 10.1681/ASN.0000001125 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Senev A Lerut E Coemans M, et al. Association of HLA mismatches and histology suggestive of antibody-mediated injury in the absence of donor-specific anti-HLA antibodies. Clin J Am Soc Nephrol. 2022;17(8):1204–1215. doi: 10.2215/CJN.00570122 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Koenig A Chen CC Marçais A, et al. Missing self triggers NK cell-mediated chronic vascular rejection of solid organ transplants. Nat Commun. 2019;10(1):5350. doi: 10.1038/s41467-019-13113-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Cristoferi I Varol H van Baardwijk M, et al. Multiomic profiling of transplant glomerulopathy reveals a novel T-cell dominant subclass. Kidney Int. 2024;105(4):812–823. doi: 10.1016/j.kint.2023.11.026 [DOI] [PubMed] [Google Scholar]

Articles from Journal of the American Society of Nephrology : JASN are provided here courtesy of American Society of Nephrology

RESOURCES