Time-averaged albumin predicts the long-term prognosis of IgA nephropathy patients who achieved remission

The study comprised 878 IgAN patients recruited between January 2005 and December 2010 from the Department of Nephrology, Ren Ji Hospital, Shanghai, China. All the patients had definitive pathological data with predominant (at least 1+) mesangial staining for IgA according to immunofluorescence staining and electron-dense deposits within the mesangium according to electron microscopy analysis. Patients who presented at younger than 18 years of age (n = 30), were pregnant (n = 3), had acute kidney failure (n = 12), had less than 36 months of follow-up (n = 160), had systemic inflammation such as Henoch–Schönlein purpura (n = 6), or had chronic advanced liver disease (n = 35) were excluded. A total of 632 patients were included. The study was approved by the Ethics Committee of Ren Ji Hospital, and all the participants gave written informed consent. All the kidney biopsy slides were reviewed by an experienced renal pathologist.

Only those subjects in whom the treatment response evaluated at 6 months after the initiation of therapy met the criteria for a complete remission (CR) or partial remission (PR) were included in the present study. A CR was defined as the absence of proteinuria, the normalization of all biochemical findings, and a lack of worsening of renal function. A PR was defined as a > 50% reduction in proteinuria from baseline. No response (NR) was defined as a < 50% reduction in proteinuria or an increase in proteinuria, with or without renal deterioration. In the end, 369 patients who met the remission criteria were enrolled.

At the end of the follow-up, the estimated glomerular filtration rate (eGFR) values were evaluated. The primary outcome was a doubling of the baseline serum creatinine (SCr); secondary outcomes included ESRD and death. Patients were classified with progression when their eGFR values decreased (below the normal range) ≥ 5 ml/min/1.73 m²/year or reached the endpoint during the follow-up period. The patients who stabilized or had a slower decrease in the eGFR were considered as non-progression patients. Among the 369 remission patients, 61 were progression patients, and 308 subjects were non-progression patients (Figure 1).

The patients had regular follow-up visits at intervals of at least 3–6 months. All the data were collected prospectively. The baseline clinical and demographic data were recorded at the time of the renal biopsy. The demographic data included age and gender. The clinical parameters included systolic blood pressure (SBP), diastolic blood pressure (DBP), blood erythrocytes (RBC), blood leukocytes (WBC), thrombocytes (PLT), hemoglobin (Hb), SCr, serum uric acid (UA), blood urea nitrogen (BUN), ALB, glutamic-pyruvic transaminase (ALT), glutamic-oxaloacetic transaminase (AST), triglyceride levels (TG), serum total cholesterol levels (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL), the erythrocyte sedimentation rate (ESR), complement 3 (C3) levels, serum immunoglobulin (IgA, IgG, IgM) levels, 24-h urinary protein excretion (UPE), hematuria (uRBC/HP), and the eGFR at the diagnosis of IgAN and during the follow-up.

The eGFR was calculated using the Modification of Diet in Renal Disease (MDRD) study equation: eGFR (ml/min/1.73 m²) = 1.8 × [SCr (mg/dl)]^-1.154 × (age)^-0.203 × (0.742 if female) [19]. Chronic kidney disease (CKD) was classified based on the Kidney Disease Outcomes Quality Initiative (K/DOQI) practice guidelines.

For each patient, proteinuria was measured via the UPE. An average UPE was determined for each 3-month block during the follow-up; the average of the UPE values for every 3-month period was represented by the time-averaged UPE (TA-UPE) [20]. The time-averaged serum albumin (TA-ALB), time-averaged serum creatinine (TA-SCr), time-averaged blood urea nitrogen (TA-BUN), and time-averaged eGFR (TA-eGFR) values were calculated using the same method as that used for the TA-UPE.

Most of the patients were treated according to the accepted standards at our center as follows: (1) full doses of ACEis (angiotensin converting enzyme inhibitors) and/or ARBs (angiotensin receptor blockers) were recommended for all the patients with proteinuria or hypertension; (2) steroids were used in cases of massive proteinuria (>1 g/d); and (3) other immunosuppressive agents were considered in patients with impaired kidney function or rapidly progressing kidney function decline (an increase in SCr > 15% in the year before entry into the trial) [21]. Renal lesions were graded according to the Lee’s classification at the time the database was established [22].

Analyses of the data were carried out using SPSS software (version 13: SPSS, Chicago, IL). For comparison of the continuous variables between the groups, Student’s t-test was used if the variables had a normal distribution, and a Mann–Whitney U-test was used if there was not a normal distribution. The chi squared test was employed for the categorical variables. Cox’s proportional hazards models for estimating the hazard ratios and the 95% confidence intervals (CI) were used to identify the predictive factors for the development of IgAN progression. The multivariate models used a stepwise forward selection procedure based on a likelihood-ratio test with P > 0.10 for the removal and P < 0.05 for the entry of the variables. Receiver operating characteristic curves (ROC) were established to determine the optimal cutoff value of a risk factor for predicting renal progression. In addition, ROC curves were also analyzed when two risk factors were combined. Renal survival was estimated with the Kaplan–Meier method, and comparisons were performed using the log-rank test. The results were reported as the means ± standard deviation (SD). Values of P less than 0.05 were considered statistically significant.

From 2005 to 2010, 878 patients with primary IgAN were recorded in the Registry Database, among whom 246 cases were excluded and 632 cases were included in this study. The data from a total of 369 patients with IgAN who achieved remission 6 months following therapy were utilized in the analyses. The clinical characteristics at baseline and during the follow-up were summarized in Tables 1 and 2. All the subjects were Chinese. There were 184 males and 185 females in the study, and the mean age at biopsy was 37.2 ± 12.3 years. The mean SBP and DBP at baseline were 123.8 ± 15.9 mm Hg and 79.3 ± 9.5 mm Hg, respectively. The mean eGFR was 83.6 ± 28.4 ml/min/1.73 m². The median and interquartile range (IQR) values for the initial proteinuria were 2.1 and 1.2-3.2 g/d, respectively. Moreover, 17% of the patients had ALB levels less than 35 g/L, and the mean ALB level was 39.4 ± 5.5 g/L. The patients were followed for 49.0 (38.0-65.8 months) months, and 61 of these patients reached the endpoint of the follow-up.

We compared the baseline characteristics of the progression and non-progression groups (Table 1). Our data revealed that the patients with renal progression had a worse baseline renal function, as represented by the eGFR (66.7 ± 7.8 ml/min/1.73 m² versus 87.7 ± 3.9 ml/min/1.73 m²; P = 0.03), the SCr level (112.8 ± 11.9 μmol/L versus 86.2 ± 4.7 μmol/L; P = 0.02), and the BUN (7.7 ± 0.9 mmol/L versus 5.8 ± 0.3 mmol/L; P = 0.03). There were no significant differences in the other baseline characteristics, including gender, age, blood pressure, ALB levels, and so on. The classifications of the renal histopathology were summarized in Table 1, and no significant differences were seen between the progression and non-progression groups.

The clinical profiles present during the follow-up after diagnosis were shown in Table 2. The renal function, as represented by the eGFR, SCr and BUN values, was significantly different between the two groups, similar to what was observed at baseline. The level of ALB at 6 months was lower (40.5 ± 1.3 versus 43.5 ± 0.6 g/L; P = 0.02) in patients in the renal progression group than those in the non-progression group. The median value of the UPE at 3 months after the initial biopsy was higher [1.3 (0.7-2.0) g/d versus 0.6 (0.2-1.0) g/d; P = 0.04] in the progression group than in the non-progression group. No significant difference was found in the TA-UPE between the two groups.

The time-averaged characteristics of the two groups were also summarized in Table 2, which indicated that the progressing patients had a lower level of TA-ALB (39.4 ± 1.2 g/L versus 42.7 ± 0.5 g/L; P = 0.01), a lower level of TA-eGFR (61.9 ± 8.1 ml/min/1.73 m² versus 95.1 ± 4.5 ml/min/1.73 m²; P = 0.002), a higher level of TA-SCr (136.7 ± 21.0 μmol/L versus 80.5 ± 4.2 μmol/L; P = 0.002), and a higher level of TA-BUN (8.6 ± 0.9 mmol/L versus 6.1 ± 0.2 mmol/L; P = 0.004) than the non-progressing patients.

Both univariate and multivariate Cox analyses were performed to evaluate the impact of the potential predictors of renal progression. As shown in Table 3, in the univariate analyses, the SCr, BUN, and eGFR values at the baseline and follow-up time periods, the ALB level at the sixth month, and the TA-ALB were statistically significant. Those factors that were significantly correlated with progression on the basis of the univariate analysis were further evaluated with a multivariate analysis. The results revealed that only the TA-ALB and TA-SCr levels were independently associated with renal progression. The adjusted multivariate Cox analysis model indicated that each 1 g/L drop in the TA-ALB level was associated with a 14% increase in the risk of renal progression.

As mentioned previously, the TA-ALB value was used to quantify the level of ALB during the follow-up of patients with IgAN, and the ROC curves for the TA-ALB were established to determine the optimal cutoff values for predicting a worse outcome (Table 4). The area under the ROC curve (AUC) was 0.82 when the TA-ALB was combined with the ALB value at baseline, at the third month and at the sixth month, indicating that the TA-ALB had a high diagnostic accuracy for an unfavorable renal outcome (sensitivity 90%, specificity 83%).

Based on the Kaplan-Meier analyses, the actual renal survival according to the estimated 36-month risk was plotted in Figures 2 and 3. The optimal cutoff for TA-ALB was 38 g/L, indicating that the relationship between the TA-ALB values and renal outcomes was dramatically altered at levels as low as 38 g/L. Those with SCr levels > 120 μmol/L at baseline, the third month, and the sixth month and ALB levels < 39 g/L at the sixth month had an actual cumulative incidence of reaching the primary endpoint.