Monitoring suPAR levels in post-kidney transplant focal segmental glomerulosclerosis treated with therapeutic plasma exchange and rituximab

We conducted both retrospective and prospective studies of patients with FSGS as the cause of end stage renal disease (ESRD). Only patients who received TPE alone or together with rituximab for recurrent FSGS treatment were enrolled. All subjects received immunosuppressive therapy consisted of induction with thymoglobulin (1.5 mg/kg/day) and steroid (500 mg/day), followed by maintenance treatment with tacrolimus, mycophenolate mofetil and prednisone. Both studies were approved by the institutional review board (IRB) of Johns Hopkins Hospital. In the retrospective study cohort, we included 19 adult renal transplant recipients who underwent renal transplantation between September 1, 2008 to December 31, 2011 in our center and developed rFSGS after kidney transplantation. For the purpose of our prospective study, we enrolled 15 patients who were transplanted in our center between August 1, 2011 and May 31, 2015 and developed rFSGS post transplant. We followed these participants as per our research protocol and collected serum, and plasma waste samples before and after each TPE. We also evaluated the subjects’ response to TPE treatment. Response to therapy is defined as reduction of proteinuria to less than 1 g/g, or the reduction of proteinuria by more than 50%. The demographic and clinical characteristics of both the retrospective and the prospective cohorts are shown in Table 1.

The diagnosis of recurrent and de novo FSGS was made by the new onset of proteinuria as measured by urine protein-creatinine ratio (UPCR) of more than 1 g/g and confirmed by kidney biopsy. Kidney biopsy in early stages showed only significant podocyte effacement on electron microscopy. Classical findings of light microscopic changes are also seen.

The measurement of serum suPAR was performed using a Human uPAR Quantikine ELISA kit (R&D Systems Inc) following the manufacturer’s instruction [9, 10]. Standards were run three times to calculate the intra-assay coefficient of variation (CV). The mean and SD for standard 1, standard 2, and so forth were used to derive the CV before averaging the CV of each standard. The inter-assay CV was derived by calculating the mean and SD for standard 1 (e.g., measurement day 1 and day 2), standard 2 (day 1 and day 2), and so forth to derive the CV and then average the CV. Both the intra-assay and inter-assay CVs were < 5% for suPAR.

To semi-quantitatively examine the effect of FSGS patient sera on podocyte β3 integrin activity, a human podocyte cell line was cultured at 37 °C for 14 days for complete differentiation [17]. The cells were then incubated in 5% of FSGS patient serum for 24 h with lipopolysaccharide (LPS) as a positive control. Next, the cells were fixed with 4% paraformaldehyde (PFA) and processed for immunofluorescence staining for AP5 (Blood Center of Wisconsin) and paxillin (Millipore). AP5 is an antibody detecting the active state of β3 integrin by recognizing the unfolding N-terminal epitope GPNICT upon the activation of the integrin [18]. After immunostaining, confocal (Leica) images were taken to quantify the AP5 and paxillin intensity for each sample treatment. Paxillin signal was used to correct AP5 signal for each treatment. The relative AP5 signal (AP5/paxillin ratio) from each patient serum was then normalized against that of normal blood donor included in each assay for final report [15]. To control for suPAR specificity, the cells were co-incubated with both FSGS sera and suPAR blocking antibody. The normalized AP5 value from normal serum treated podocytes was 1. The relative AP5 value of 1.05 or more obtained from patient serum treated podocytes was considered abnormal.

For continuous variables, data are expressed as mean ± SEM or median with interquartile range as appropriate. Categorical variables were expressed as percentages. The demographic and clinical characteristics of patient and control participants were compared using the t test, or the Fisher’s exact test for categorical variables. Multiple linear or logistic regression analyses were performed to evaluate the association between serum suPAR and the variables of interest while controlling for age, sex, and other potential confounders with SPSS software. The relative change in suPAR after TPE treatment was calculated as per 10% reduction from before TPE treatment. The relative change of proteinuria in terms of UPCR was calculated as 100 x (UPCR before treatment-UPCR after treatment)/UPCR before treatment. All statistical tests were two tailed. P values < 0.05 were considered significant.

To look at the immediate effect of TPE on serum suPAR levels, we compared serum suPAR right before and after a single course of TPE. We found that single course of TPE could remove on average 37% of serum suPAR (Fig. 1a). Simultaneously, suPAR was detected in the pheresis waste bags, ranging from 1149 pg/ml to 2417 pg/nl with an average suPAR value of 1848 pg/ml. This is in consistent with previous reports [9, 14‐16], and indicates that TPE could effectively decrease serum suPAR levels by removing suPAR from the blood circulation.

We performed a retrospective study, in which we analyzed 19 post-transplant (post-Tx) FSGS patients, 17 of them had rFSGS, and 2 had de novo FSGS (Table 1), 42% of them were male. The mean ± SD age at transplant was 40 ± 12 years old. Median time to post-Tx FSGS (IQR) was 31 (5, 238) days. All patients were treated with TPE, and 12 received rituximab as well. Serum was collected before transplantation, after transplantation at the time of post-Tx FSGS diagnosis, and before and after TPE sessions for suPAR measurement. Each patient received a median (IQR) of TPE 14 (7, 82) sessions (Table 1). As shown in Fig. 1b, mean serum suPAR was higher before transplantation comparing to that at the time of post-Tx FSGS diagnosis (8131 ± 1300 pg/ml before Tx versus 5551 ± 429.2 pg/ml post-Tx, p = 0.056). One TPE session resulted in significant reduction in serum suPAR level (5551 ± 429 pg/ml before TPE vs 4532 ± 351 pg/ml post TPE, p < 0.05). By the end of treatment with TPE and rituximab, the mean proteinuria significantly decreased from 4.84 ± 0.76 g/g to 2.06 ± 0.47 (p < 0.01, Fig. 2a). The median serum creatinine levels were significantly decreased as well (2.5 mg/dL before therapies versus after 1.8 mg/dL after therapies, p < 0.05, Fig. 2b).

To validate our findings of the retrospective study and assess the role of suPAR in monitoring the response to TPE in post-Tx FSGS, we performed a prospective study of 15 patients with rFSGS. As indicated in Table 1, 67% of them were male and mean ± SD age at transplantation was 43 ± 15 years old. The median time to rFSGS was 33 (5, 299) days. All these patients received TPE and rituximab. The median (IQR) TPE was 10 (10, 52) sessions. Serum was collected to measure serum suPAR and creatinine levels, and urine was collected for proteinuria assessment before and after TPE treatment. We found that one TPE resulted in significant reduction in serum suPAR levels to from 3240 ± 409.4 pg/ml, to 2486 ± 229 pg/ml (p < 0.05, Fig. 3a). Similarly, proteinuria decreased significantly as well (4.19 ± 1.14 g/g before vs 2.07 ± 0.75 g/g after treatment, p < 0.01, Fig. 3b). The median serum creatinine reduced but did not reach any statistical significance (Fig. 3c). Building on our previous findings that elevated circulating suPAR could induce podocyte β3 integrin activity and thus kidney injuries [9], we performed cultured human podocyte based β3 integrin activity assay by incubating podocytes with post-Tx FSGS patient sera harvested before and after TPE treatments (Fig. 4). The generated AP5 activity was corrected against that of normal sera from healthy subjects. Co-incubation of the patient sera with anti-human suPAR antibody was applied in order to control for the suPAR effect (Fig. 4). In consistent with the changes of serum suPAR levels and proteinuria, the mean β3 integrin activity in terms of AP5 intensity significantly decreased after treatment (1.10 ± 0.08) as compared to before treatment (1.34 ± 0.08, p < 0.05, Fig. 3d). Before TPE and rituximab combined therapies, 11 out of 12 patients had high podocyte AP5 value, while 7 of 12 patients had high AP5 after treatment. Out of 5 patients with AP5 normalized after treatment, only 1 (20%) did not respond to TPE and rituximab. Among the 7 patients with persistent high podocyte AP5 values after treatment, 5 (71.4%) did not respond to the combined therapies.

Next, we analyzed the distribution of both retrospective and prospective cohorts and found that the demographic and clinical characteristics are similar (Table 1), thus we combined the two cohorts together for further analysis. First, we examined the bivariate correlations between the continuous variables, including serum suPAR, serum creatinine, eGFR, and UPCR before TPE treatment; age at diagnosis, age at transplantation, transplant number, TPE course numbers, as well as, the reduction in serum suPAR (dsuPAR), the reduction in UPCR (dUPCR) after treatment (Table 2). Pre-TPE levels of serum suPAR were associated with serum creatinine (r = 0.37, p < 0.05). Pre-TPE UPCR was correlated with serum creatinine (r = 0.47, p < 0.01), and transplant number (r = 0.38, p < 0.05). The reduction in UPCR after treatment was correlated with Pre-TPE UPCR (r = 0.42, p < 0.05), and the reduction in suPAR (r = 0.48, p < 0.01).

Then, we performed multiple linear regression analysis to evaluate the reduction in UPCR after treatment, controlling for Pre-TPE suPAR, UPCR, serum creatinine, eGFR, and patient gender, TPE number, age at transplant, transplant number, as well as the reduction in suPAR post-TPE treatment. We observed that the reduction in serum suPAR levels alone accounted for 23% proteinuria reduction. The reduction in serum suPAR levels and baseline UPCR together accounted for 35% proteinuria reduction after treatment, with the reduction in serum suPAR still predicting stronger than baseline UPCR (Table 3). Lastly, we performed a logistic multiple regression to look at the response to TPE and rituximab combined therapies, which was defined as UPCR < 1 g/g and/or the reduction in UPCR more than 50%, controlling for variables including the reduction in serum suPAR levels, Pre-TPE UPCR, eGFR, TPE number and age at transplant. We found that only the reduction in serum suPAR significantly predicts the response to the therapies with an odds ratio of 1.43, 95% CI (1.02, 2.00), p < 0.05.