The main objective of this study was the determination of anti-PLA2R and anti-THSD7A antibodies in the serum of MN patients and control patients in a Latin population.
All included patients, regardless of the underlying disease, were negative for anti-THSD7A. Data from the literature show that no healthy individual or other proteinuric diseases other than MN tested positive for anti-THSD7A, and its specificity was close to 100% (15, 20). However, the sensitivity was low, ranging from 3% (Europe and the United States) to 9% (Japan) (15, 20–22). Considering the present sample size and the method sensitivity, the absence of positive results was expected.
On the other hand, anti-PLA2R antibody determinations were performed using two available techniques, ELISA and IFI. As described by different authors, all patients diagnosed with LMN and FSGS (control group) were negative according to the IFI technique, which provided 100% specificity in this group. When evaluated by ELISA, the specificity in this group was 97%. Only one patient with a clinical and histological diagnosis compatible with FSGS had a low-titer anti-PLA2R (23 RU/ml) result by ELISA and a negative IFI, which may be considered a false positive (23).
Among MN patients, 55% were positive according to one of the techniques employed, and similar data were found in other cohorts, where 50–80% of MN patients tested positive for anti-PLA2R antibodies (24).
Two recent meta-analyses have assessed the accuracy of anti-PLA2R in differentiating MN from other causes of proteinuria. In 2015, Dai et al. found a sensitivity of 79% and specificity of 90% (23). In 2018, Li et al. evaluated 28 studies and found a sensitivity and specificity of 65% and 97%, respectively (25).
In this sample, patients with an interval of less than 12 months between biopsy and antibody collection compared to those with an interval greater than 1 year had greater anti-PLA2R positivity (79%) and presented greater proteinuria (8.4 vs. 1.6 g/24h, p = 0.001) and lower serum albumin (1.9 vs. 3.7 g/dL, p = 0.002) at the time of inclusion in the study.
Newly diagnosed populations presented positivity between 60.2% and 80%, which was higher than that found in prevalent populations (19, 26–29). Different reasons may contribute to this variability among studies and explain the greater positivity when the interval between biopsy and antibody collection is restricted to less than 12 months. The type of test used (WB, IFI and ELISA), ethnic differences and the timing of antibody collection in relation to the course of the disease are some of the main reasons, especially the latter, which appears to be the most relevant. Moreover, patients with a shorter interval between biopsy and antibody collection are more likely to have active disease.
Studies in different populations demonstrated an association between disease activity, mainly represented by proteinuria and the presence of antibodies. Anti-PLA2R often mirrors disease activity, so when antibody titers are high, proteinuria usually follows high. It is well established that antibody levels fluctuate significantly during the course of the disease, which is why reports of sensitivity and specificity vary depending on the number of patients with nephrotic proteinuria were included in the study (30).
Dai et al. performed a meta-analysis and showed greater accuracy of anti-PLA2R discrimination according to the degree of proteinuria (23). In a subgroup analysis, the anti-PLA2R test showed greater accuracy among patients with nephrotic syndrome (AUC 0.83) than those without nephrotic proteinuria (AUC 0.47). This difference could be justified by the fluctuation of the level of the antibody according to the activity of the disease due to a spontaneous decrease of its serum level (remission) or the use of immunosuppression (27, 31, 32). In our study, patients with positive anti-PLA2R had higher proteinuria than negative patients (8.3 and 1.6 g/24h, p = 0.003).
There is a correlation between antibody positivity and proteinuria but also between antibody titers and proteinuria values. According to data obtained at the time of inclusion, a positive correlation was found between anti-PLA2R titers by ELISA and 24-hour proteinuria (Spearman rho: 0.554, p = 0.001); that is, the greater the titer of anti-PLA2R, the greater the proteinuria. This observation has been demonstrated by different authors in different populations. In 2011, Hosftra et al. found a correlation of 0.73 between the proteinuria value and anti-PLA2R titer. In 2012, Kanigicherla et al. reported a correlation of 0.51 while in 2014, Hoxha et al. indicated that a 25% reduction in proteinuria levels was associated with a 45% drop in antibody titers after 3 months of follow-up, with a greater reduction in proteinuria observed among those who received immunosuppression (43% reduction in proteinuria and 82% reduction of antibody titers). In 2015, Ruggenenti et al. found that after treatment with rituximab, a 50% reduction in antibody titers occurred 10 months before an equivalent reduction in proteinuria (27, 33–35). Accordingly, this finding may also explain the negative correlation between biopsy time and antibody titers. Patients diagnosed for a longer time are more likely to have achieved remission, either through treatment or spontaneously, thus presenting lower antibody titers.
In this study, IFI and ELISA techniques were compared regarding sensitivity and specificity, and the results varied according to the moment of determination.
At the end of the study, a sensitivity of 45% and specificity of 97% were obtained for ELISA and a sensitivity of 55% and specificity of 100% were obtained for IFI, with an agreement of 86% and a correlation of 0.72. These findings are compatible with those described in the literature, which shows a good correlation between ELISA and IFI, despite IFI having greater sensitivity (24, 30, 36–39).
In 2014, Behnert et al. compared different techniques in a cohort of 157 MN patients. There was a good correlation between the techniques employed, with an agreement between the ELISA and the IFI of 85.9% and a correlation of 0.72, which was similar to our study (39).
ELISA titers over 20 RU/ml were considered positive according to the manufacturer's instructions. However, some authors have suggested that the reference value should be reduced to 2 RU/ml or 14 RU/ml. These authors based their suggestion on studies that showed increased sensitivity without a loss of specificity (38, 39). Considering the cutoff value of 14 RU/ml, the sensitivity of the ELISA would be equivalent to that found for IFI at 55%, because all patients with a negative ELISA and positive IFI had titers greater than 14 RU/ml in our sample.
Tissue staining of the anti-PLA2R antibody was first described in 2011 and has since been used to complement the diagnosis of patients with suspected MN. Different studies have compared the sensitivity of tissue staining with serum antibody measurements, and a greater sensitivity of tissue staining was observed. This technique has also been used to differentiate primary causes from secondary causes of MN.
Fifty-two biopsies from patients with MN (N = 28), LMN (N = 17) and FSGS (N = 7) were retrospectively evaluated using the immunohistochemistry technique, which was first used for this purpose in our institution. As described in the literature, biopsy of normal kidneys shows weak positivity for anti-PLA2R, and this degree of staining is considered negative (14) (40).
Among the biopsies evaluated, 2/17 patients diagnosed with LMN and 3/7 diagnosed with FSGS were positive. Studies have shown antibody positivity for secondary causes of MN, which are mainly related to HBV, HCV and sarcoidosis and rarely described in patients with SLE or FSGS (14, 40–42).
Among the 28 patients with primary MN, 20 patients (72%) showed a positive result in the renal tissue, demonstrating that in the absence of serum antibody measurements, tissue staining may play an important role in the diagnosis of MN. In addition, tissue staining may allow for a retrospective diagnosis, which may be important for the evaluation of patients who are potential candidates for kidney transplantation.
Renal biopsy staining and serum data were obtained for 23 MN patients. Seven out of 23 patients were both negative for blood and tissue; 16 patients were positive for antibodies on renal biopsy, while 3 had negative serology and 13 presented anti-PLA2R positivity in the blood. No patients with positive serology and negative renal biopsy were found. It was not possible to establish a correlation between the sensitivity of these two methods because of the large interval between the biopsy and serum antibody determinations, which had a mean interval of 19 months. In this time interval, spontaneous or immunosuppression-induced remission may have occurred with serum antibody clearing. Only the concomitant determination of antibodies with renal biopsy would allow for an adequate comparison between the sensitivity of these two methods.
A total of 29 MN patients were included in the study, and 21 patients were followed up for 12 months, with additional antibody determinations at 6 and 12 months and concomitant evaluations of PTU, serum albumin and serum creatinine.
Throughout the follow-up, the positive relationship between the PTU and ELISA titers and the negative relationship with serum albumin were maintained. Four patients had changes in their initial serum autoantibody profile, which were accompanied by changes in proteinuria levels and preceded by the initiation or suspension of immunosuppressants.
When assessing the likelihood of remission according to the presence or absence of antibodies during follow-up, an inverse relationship was observed: patients who were positive for antibodies in both the blood and biopsy showed a lower remission rate, which was defined here as partial or complete proteinuria reduction.
Different studies have found a relationship between antibody titers and the likelihood of remission. Patients with lower ELISA titers evolved more favorably, while patients with increased titers developed a greater possibility of evolving to nephrotic syndrome and loss of kidney function. However, among our patients, there was no correlation between positive antibodies and progression to CKD (24).
Serial measurements of the antibody provide even more accurate prognostic data. Decreasing antibody levels throughout the follow-up preceded proteinuria reduction. Serial measurements have also been used for prognostic determination and treatment indication or cessation in different centers across the world.