Kidney Transplant Rejection and Peritubular Capillary Basement Membrane Multilayering: The Ongoing Search for Diagnostic Guidelines

Microvascular injury to the glomerular and peritubular capillaries is a major cause of late kidney transplant loss. Endothelial activation and sustained injury caused by toxins, complement dysregulation, inflammation, antibodies, and others can lead to excess capillary basement membrane formation with multilayering. In kidney transplants, there are 2 histological lesions interrelated with capillary multilayering: transplant glomerulopathy (TG) and peritubular capillary basement membrane multilayering (PTCML). Both lesions are the principal morphological features of chronic antibody-mediated rejection (cAMR). PTCML was first recognized in 1990 in an electron microscopy study of kidney allografts with TG.1 Since then, the use of electron microscopy as an ancillary tool for diagnostic work-up of kidney transplant rejection has gained increased attention. Different groups reported PTCML in cAMR with varying degrees of severity but found that some kidney transplants without cAMR and native kidneys with different diseases also had PTCML. PTCML was common in native kidney diseases with generally up to moderate severity: obstructive nephropathy, tubulointerstitial nephritis, and immune-mediated glomerulonephritis.2-4 However, chronic thrombotic microangiopathy was an exception, which commonly showed severe PTCML.4 Based on the cumulative evidence, the Banff 2013 classification recommended only severe PTCML, as evidence for cAMR, defined as ≥7 layers in 1 cortical peritubular capillary and ≥5 in 2 additional capillaries.5 Subsequently, Dobi et al6 showed that Banff 2013 criteria identified only 7% of early cAMR (TG score cg1) and 52% of advanced cAMR (cg2–3), questioning its validity as it seems that Banff criteria are missing early cAMR cases when they are likely treatable. Sis et al7 showed that 91% of TG cases (cg1–3) had PTCML, which was commonly mild (2–4 layers) or moderate (5–6 layers) PTCML in 32% and 36% of TG, respectively, and was below the Banff diagnostic threshold. Data from other centers suggested that 1 capillary with 5 layers,6 a minimum of 3 capillaries with 5–6 layers,3 or 2 capillaries with ≥4 layers8 may be a better PTCML cutoff for the diagnosis of cAMR. Furthermore, Roufosse et al9 showed that mild PTCML predicts subsequent TG with an increased risk for every additional capillary with ≥3 layers. However, the Banff 2013 criteria were not designed to diagnose early cAMR; rather, they were intended to avoid potential overdiagnoses of cAMR. In addition, the comparison of PTCML with TG as a reference test for cAMR should be interpreted with caution, as they represent different compartments, and we know that not all TG cases are cAMR. In their recent article, Nankivell et al10 revisit PTCML cutoffs. They address a long-standing question: what is the minimal degree of PTCML that can serve as a sensitive and specific test for cAMR? As previously summarized, many groups explored this issue for decades. With the new data by Nankivell et al, are we closer to a final verdict? The answer is: perhaps. Their study reaffirms the established observation that PTCML reflects various forms of endothelial injury in native and transplant kidneys, with chronic thrombotic microangiopathy and donor-specific antibodies (DSAs) being the strongest drivers. Yet, detailed analyses of PTCML in biopsies with advanced chronic injury (eg, arterionephrosclerosis or diabetic nephropathy) are lacking. This omission leaves an unresolved important question: how does severe chronic injury influence PTCML? The issue is clinically relevant in marginal donor organs and older grafts, where nonrejection-related complex chronic changes render diagnoses of cAMR more challenging. Prior studies by Liapis et al,4 including large cohorts of native biopsies, showed a “background noise” of PTCML with up to 4–5 layers in cases of chronic parenchymal injury without DSAs. These findings contributed to the Banff recommendations on PTCML in 2013.5 Reassessing PTCML would be valuable for establishing updated cutoffs in the modern diagnostic era. Do higher Banff chronicity scores in transplants necessitate higher PTCML thresholds, or can universal guidelines suffice as suggested by Nankivell et al? Correlation studies on PTCML are difficult because the causal insult (eg, transient DSA-mediated endothelial injury) is temporally separated from the effect (chronic accumulation of basement membrane layers). Questions remain about whether class I versus class II DSAs differ in their ability to induce PTCML, how antibody strength and duration of exposure matter, and how repetitive or smoldering injury shapes basement membrane remodeling. To deal with this challenging problem, Nankivell et al used TG (≥cg1a) as a surrogate marker for cAMR and correlated it with PTCML. They found, unsurprisingly, that basement membrane multilayering was often—but not always—present in both compartments, with differences noted in the severity of TG compared with PTCML. Statistical analysis suggested a diagnostic cutoff of 3–4 circumferential PTCML layers in “multiple” capillaries, yielding 83% sensitivity and 73% specificity against cg1a used as a surrogate marker for presumed cAMR. They showed that the Banff 2013 PTCML criteria had 30% sensitivity and 91% specificity, which improved to 55% sensitivity and 90% specificity with the refinement of the PTCML criteria as 1 capillary with ≥7 layers or 2 capillaries with ≥5 layers. How the proposed PTCML cutoff— ≥ 3 circumferential layers in “multiple” peritubular capillaries—will perform has to be seen. Questions remain: how reliable are the proposed PTCML cutoffs in reflecting antibody-induced basement membrane remodeling? Nankivell et al largely based their analysis on comparative studies of 2 anatomic compartments—TG and PTCML—while exhaustive studies into the underlying cause, that is DSA-driven PTCML, were lacking. Furthermore, what exactly qualifies as “multiple” capillaries in the proposed cutoffs? And will these diagnostic thresholds prove robust across diverse settings, including patients with more pronounced chronic graft injury? Time will tell whether the study by Nankivell et al finally resolves the enduring debate over the optimal diagnostic cutoff for PTCML. We propose that milder degrees of PTCML may be used as a surrogate lesion to prompt testing for DSA and risk stratification. Evidence suggests that advanced cAMR displays more severe PTCML at a stage when diagnosis is not difficult, but often treatments are not effective. Thus, more sensitive approaches (mild PTCML) may enable early diagnosis and treatment of cAMR in the appropriate clinicopathologic context if other diseases are reasonably excluded.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.