Estimating Prognosis in Anti–Glomerular Basement Membrane Disease

One of the most vexing issues in clinical nephrology occurs when a patient with anti-glomerular basement membrane (GBM) disease presents late in the course with advanced kidney failure, often requiring the immediate initiation of dialysis. Patients with anti-GBM disease discovered earlier in the course of disease evolution most often are promptly treated with standard-of-care using steroid, immunosuppression, and plasma exchange combination regimens. Such early discovered disease usually responds sufficiently to avoid dialysis, and patients can often regain adequate renal function lasting for many years. Relapses are uncommon in these patients, but they can occur.¹ Thus, a validated tool to stratify patients with anti-GBM disease for prediction of renal survival would be very valuable to better guide treatment interventions. Such a tool is already available for ANCA-associated vasculitis, but not for anti-GBM disease.²

A superbly conducted multi-institutional retrospective study by a group of investigators from the United Kingdom, Europe, and the United States have now filled this gap, at least in a partial but helpful way.³ Floyd and colleagues collected and analyzed reasonably complete clinical and histopathological data on 174 patients with anti-GBM disease (61.5% with anti-GBM disease only and 38.5% with combined anti-GBM and ANCA disease) presenting for care over a 22-year period (1998-2020) that embraces the modern era of management of anti-GBM disease. Not unexpectedly, most of the patients (74.1%) presented with advanced kidney failure requiring immediate initiation of RRT. With such progressive disease, it is also no surprise that most (74.4%) of those requiring RRT failed to recover kidney function, despite adequate therapy. Immunosuppressive therapy was provided in 96.1% of the patients, including plasma exchange in 82.2%. Overall, 30.5% died at some point during follow-up. By any measure, this cohort represented patients with severe, life-threatening disease managed in a conventional manner.

One of original goals of the study was to see whether the Renal Risk Score (RRS) tool, originally developed for prognostic stratification in ANCA-associated vasculitis² could be extended to anti-GBM disease. The study succeeded in this venture, but the additional goals of identifying patients most likely to benefit from immunosuppressive therapy (including plasma exchange) and to develop useful prediction tools for prognostic stratification succeeded in an even more important way, at least in my opinion. Using the RRS, kidney survival was 100%, 62.4%, and 20.7% in the low-, moderate-, and high-risk RRS categories, with no differences in patient mortality between these categories of RRS. Interestingly, these outcomes were independent of clinical presentation. Pulmonary involvement (alveolar hemorrhage) was seen in a somewhat higher frequency in those with an ESKD outcome (27%) compared with those with renal recovery (18%).

A key finding was the prognostic significance of the percentage of “normal” glomeruli (without any scarring, crescents or fibrinoid necrosis within the tufts; N0=>25%; N1=10–25%; N2=<10%) in the initial biopsy specimen. End-stage kidney disease occurred in 15.2%, 53.3%, and 76.6% of N0, N1, and N2 groups, respectively. A low percentage of normal glomeruli and the need for dialysis contributed in a major way to defining a poor prognosis, even better than the RRS. No difference in outcome could be found between those patients with kidney biopsies exhibiting zero normal glomeruli and those with <10% but non-zero normal glomeruli. The presence or absence of ANCA did not seem to influence outcome. Models with the dialysis need and percentage of normal glomeruli achieved an excellent level of discrimination (Harrell C-statistic=0.840, P<0.001).

These findings are of great clinical value and emphasize several points. First, kidney biopsy provides additional information, beyond diagnosis, that can be useful in overall management of patients with anti-GBM disease, providing that the sample size is adequate (>10 glomeruli). Second, the requirement for dialysis at presentation, not just serum creatinine level or eGFR, is a major determinant of outcome. This serves to stress, again, the overriding importance of early detection with prompt and aggressive treatment (steroids, cyclophosphamide, and daily plasma exchange) of anti-GBM disease (with or without alveolar hemorrhage or concomitant ANCA positivity). This study did not analyze the additional influence of high anti-GBM levels or oligoanuria, both of which have been associated with a poorer prognosis for eventual renal recovery, perhaps independent of kidney biopsy findings.^4,5 The utility of repeat kidney biopsy was also not examined.

The authors have quite appropriately signaled that this risk stratification system has very limited utility, if any, in declaring futility in individual patients. A 36-month renal survival rate of 14.1% in the worst prognostic group should be enough to dissuade the application of this tool for declaration of futility, which is a highly individualized decision. Other factors, such as extensive tubulointerstitial fibrosis, glomerulosclerosis and tubule atrophy, persistent oligoanuria, risks of immunosuppression, likelihood of availability of a kidney transplant, and the wishes and desires of the patient need to be taken into account. The KDIGO clinical practice guidelines (2021 version) recommended that aggressive immunosuppressive therapy be withheld in patients with a dialysis requirement at presentation, zero normal glomeruli or >50% global glomerulosclerosis, and no pulmonary hemorrhage.⁶ This analysis calls into question the validity of this advice.

Owing to the rarity of anti-GBM disease, it was not possible to conduct a validation study of this system. So for the time being at least, this tool should be regarded as nonvalidated, but it is clearly the best prognostic tool available and it should be widely used, in my opinion.

Whether new advances in treatment regimens, such as combined cyclophosphamide and anti-CD20 monoclonal antibodies or streptococcal endopeptidase (imlifidase) or double filtration plasmapheresis, immunoadsorption devices, will make this tool obsolete remains unknown.^7-9 All such prognostic tools have a half-life of utility as knowledge accrues. This study has provided a new benchmark for prognostication in anti-GBM disease that is both practical and useful.

Disclosures

R.J. Glassock reports consultancy: Agenx, Alexion. Chinook, Am J Nephrol, Anteris Bio, Arrowhead Bio, Aurinia, BioCryst, Cal!iditas, ChemoCentryx, Equillium, Forsee Pharma, Horizon, Sentien, Ionis, Karger Publications, Midornid, NIH, Novartis, Omeros, Otsuka Pharma, RenaSight (Natera), River3Renal, Therini Bio, Travere (Retrophin), Up To Date (Wolters-Kluwer), Vera, Vertex, Vivace, Walden Bioscience; ownership interest: Reata, Inc; honoraria: Agenx, Alexion, Alexion, Anteris, Arrowhead, Aurinia, Biocryst, Chemocentryx, EcoR1, Horizon, Karger Publications, Midornid, Natera, Novartis, Omeris, Wolters-Kluwer (UpToDate); advisory or leadership role: American Journal of Nephrology, BioCryst, Calliditas, JASN, Novartis, Otsuka, RenaSight, Travere, University Kidney Research Organization, UpToDate; speakers bureau: Aurinia; and other interests or relationships: ASN-Open Forum Communities.

Funding

None.

Author Contributions

R.J. Glassock conceptualized the study, was responsible for formal analysis, wrote the original draft, and reviewed and edited the manuscript.