Fibrillary glomerulonephritis (FGN) is a rare disease that is identified in 0.5–1% of renal biopsies.1,2 It is characterized by the presence of nonamyloid fibrils, typically of 10–30 nm in diameter, lacking a hollow center, located in the glomerulus, particularly the mesangium and the glomerular basement membrane. They stain intensely for IgG, kappa, lambda and C3 and are usually negative for Congo red.1–4

FGN has been associated with various autoimmune diseases, monoclonal gammopathy, neoplasia and infections, notably hepatitis C. However, in more than half of cases, the cause remains unidentified, suggesting that the pathogenic mechanisms underlying FGN are not yet understood.5,6 Clinically, it usually manifests at approximately 60 years of age nonspecifically; up to 70% of patients have renal failure, hematuria and proteinuria, the last of which is frequently in the nephrotic range. In 15–36% of cases, the disease may begin as a pure nephrotic syndrome.1,2,6,7

Recently, it was shown that the DNAJB9 protein, which under physiological conditions can be found at low levels in tubular cells, podocytes, mesangial cells and renal endothelial cells, is located exclusively as extracellular deposits in the glomeruli in FGN. This finding has transformed the diagnosis of the disease, allowing it to be identified with a sensitivity of 98% and a specificity greater than 99%, eliminating the need for ultrastructural analysis, the classic diagnostic method.8–11

To date, patients experience unfavorable outcomes, as more than half will progress to stage 5 chronic kidney disease (CKD) within 5 years after diagnosis.6 Evidence on optimal treatment remains limited, and immunosuppressive management, including mainly steroids, cyclophosphamide, mycophenolate and rituximab (RTX), has shown variable results in different cohorts.1,6 Recent studies have suggested that RTX can slow disease progression; however, evidence on its efficacy is scarce.12,13

In the current study, we described the main clinicopathological characteristics of patients diagnosed with FGN in our hospital, analyzed the risk factors associated with its progression and explored the impact of treatment on its course.

We conducted a retrospective cohort study that included 13 patients diagnosed with FGN by renal biopsy, followed by DNAJB9 testing for confirmation, in our center between 2019 and 2024.

Demographic data, body mass index (BMI), comorbidities associated with the diagnosis of FGN, laboratory results (both at baseline and at follow-up), treatments used (both supportive and immunosuppressive) and characteristics of the renal biopsy were collected from a review of the medical records. Due to its prognostic relevance, proteinuria was included as an additional baseline clinical variable and classified into 2 categories: less than or greater than 3.500 mg/g. With respect to the renal biopsy, the number of glomeruli per biopsy, the percentage of sclerosed glomeruli, the lesion pattern on optical microscopy and the location of the fibrils on electron microscopy were recorded, if available. Inflammation and tubular atrophy were categorized as absent (0%), mild (1–25%), moderate (25–50%) or severe (>50%).

Information, including proteinuria levels and their changes, albumin levels and changes in renal function was collected at follow-up visits at 3, 6 and 12 months and subsequently annually until censoring or the end of the study period. Patients were classified into groups according to the treatments received (supportive, immunosuppressive or their combination) and analyzed according to their outcomes. The outcomes of the study included complete remission, defined as a reduction in proteinuria to less than 500 mg/g and normal renal function; partial remission, defined as a reduction in proteinuria by more than 50% with respect to baseline to lower than 2.000 mg/g and stable renal function (no increase greater than 20% with respect to baseline); and no remission, which included patients who did not achieve complete or partial remission, as well as those who progressed to stage 5 CKD or died during follow-up.

Statistical analysis was performed using Python (version 3.13). Categorical variables are described as frequencies and percentages, while the continuous variables are expressed as means ± standard deviations (SDs) or medians with interquartile ranges (IQRs), according to the nature of the distribution as evaluated by the Shapiro‒Wilk normality test; given the small number of patients, cases with missing values were excluded from the corresponding analysis. For survival analyses, univariable Cox proportional hazards regression models were applied to identify risk factors for progression to stage 5 CKD or death, considering the time from diagnosis until the event (stage 5 CKD, death) or censoring. The proportionality of the risks was evaluated using the Schoenfeld test, and sensitivity analyses were performed to verify the robustness of the findings after adjusting for different hypoalbuminemia thresholds (<3 vs. <2.5 g/dL) and proteinuria (median 2.318 mg/g). No multivariable analysis was performed because of the limited sample size. Statistical significance was defined if p < 0.05.

In this study, we described the clinical characteristics, management and outcomes of patients diagnosed with FGN at our center. Our cohort, with a mean age of 61 years, was predominantly male, which interestingly, contrasts the results of other published series, which usually reported a higher proportion of women.14,15 The majority of our patients had arterial hypertension and diabetes mellitus, denoting a relevant comorbidity profile; however, we found no association of FGN with malignancies or autoimmune diseases, in contrast to the findings in the literature, where these conditions are present in 5–10% and 5–15% of patients, respectively.1,6,9 This interesting finding could suggest that our population may be uniquely characterized and further explored in larger studies.

Similar to what has been reported in other studies, our patients had impaired renal function at the time of diagnosis, along with significant proteinuria, which in 38% of the patients reached the nephrotic range, highlighting the severity of FGN at the time of diagnosis. However, unlike what has been described in the literature, only 38% of our patients had hematuria at the time of diagnosis of the disease, which is lower than what is usually reported.1,6,9

In our cohort, the most common histopathological pattern was membranoproliferative, followed by mesangial. In agreement with other series, patients with a membranoproliferative pattern tended to progress to kidney failure; however, in our study, this trend did not reach statistical significance.2,15,16 In addition, one patient in our cohort exhibited a membranous pattern and a slow disease course, marked by significant renal disease progression and death due to multiple comorbidities, an outcome that aligns with the findings of several case reports in the literature that have described an unfavorable disease course in patients with this pattern.17,18 Our patients presented an immunofluorescence pattern characterized primarily by IgG, kappa, lambda and C3 positivity that coincided with that described in other cohorts.1,9,15 All our patients also had positive for DNAJB9, highlighting the importance of this marker in the diagnosis of the disease.3,8,10,11

The patient outcomes reflect the severity of FGN in our cohort, with progression to kidney failure or death reported for 38.4% of the patients, a finding that aligns with the most numerous international cohorts, highlighting the poor outcomes of this disease.1,6 The presence of hypoalbuminemia was associated with a worse outcome in this cohort. In addition, proteinuria greater than 3.500 mg/g and decline in kidney function at the time of diagnosis tended to be associated with an increased risk of a poor outcome, although the associations were not significant. Notably, proteinuria has presented variable results in other cohorts9,15,19 and has been identified as a risk factor in some studies but not in a multicenter cohort by Andeen et al.,6 which adds complexity to the interpretation of these findings. However, the results related to the degree of deterioration of renal function at the time of diagnosis are consistent with those reported in other cohort studies1,6,9 as well as with those observed in other glomerulopathies.20,21 Kidney biopsy data, such as the degree of tubular atrophy, the percentage of glomerular sclerosis and the histopathological patterns, were not significantly related to patient outcomes.

In our center, the most commonly used immunosuppressive strategy is RTX treatment, which is generally used as a single therapy, but in some cases is combined with another immunosuppressant. Partial remission was observed in 4 of the 13 patients (30.8%), exclusively in those treated with RTX alone. These results are consistent with findings from other studies,6,8,12 which have associated RTX with a greater probability of remission and a slowing of disease progression, although larger studies are needed to confirm the efficacy of this treatment.

Our study has several limitations. The small size of our cohort results in limited statistical power in detecting significant associations, as observed for proteinuria and renal function. Therefore, all the findings should be considered exploratory and hypothesis-generating, and extrapolations or causal interpretations should be avoided. Furthermore, residual confounding cannot be ruled out; thus, caution is recommended in the interpretation of the observed associations. In addition, as this was a retrospective study of the data from a single center, selection bias and variability in clinical management could have influenced the outcomes. The heterogeneity in the treatments, the absence of a control group and the lack of a multivariable analysis for adjusting for confounding factors also limit our conclusions.

In conclusion, our findings reflect the notable severity of FGN, with rates of kidney disease progression and mortality consistent with those reported in other cohorts, highlighting the relevance of the identified risk factors and the limited therapeutic potential of RTX, despite our limited sample size. These results emphasize the need for multicenter prospective studies, new therapeutic strategies and larger populations to validate our findings and improve the management of this disease.

María José Soler Romeo: 000-0003-3621-0766

This study was funded by a research grant from Boehringer Ingelheim. JLR has a Río Hortega scholarship from the Carlos III Health Institute (CM23/00213). The authors are also beneficiaries of other research grants from the Carlos III Health Institute (PI21/01292, AC22/00029, INT23/00078, PI24/01510). The Nephrology and Transplantation group is part of the RICORS2040 network (RD21/0005/31 and RD24/0004/0031) and is recognized as a consolidated group by the Agency for University and Research Aid Management of Catalonia (2021 SGR 00883).