Acute renal failure that is associated with macroscopic

Factors That Determine an Incomplete Recovery of Renal Function in Macrohematuria-Induced Acute Renal Failure of IgA Nephropathy Eduardo Gutiérrez,* Esther González,* Eduardo Hernández,* Enrique Morales,* M. Ángel Martínez, Gabriel Usera, and Manuel Praga* Departments of *Nephrology and Pathology, Hospital Universitario 12 de Octubre, Madrid, Spain Acute renal failure that is associated with macroscopic hematuria (ARF-MH) is a widely known complication of IgA nephropathy (IgAN). Although spontaneous recovery of renal function after cessation of MH has been described, no long-term outcome studies have been performed. The outcome of patients who had biopsy-proven IgAN and presented an ARF-MH episode in the period 1975 through 2005 was studied. Thirty-six episodes of ARF-MH that occurred in 32 patients were identified. A complete recovery of baseline renal function after cessation of MH was observed in 27 (group 1); in the remaining nine episodes (25%; group 2), estimated GFR (egfr) did not reach the baseline value. Final egfr was 89 28 ml/min per 1.73 m 2 in group 1 patients and 38 12 ml/min per 1.73 m 2 in group 2 patients (P 0.0005). The duration of MH was significantly longer in group 2 patients: 33.7 25.3 versus 15.4 18.4 d (P 0008). A high proportion of tubules that were filled by red blood cell casts and had signs of acute tubular necrosis were the most striking histologic abnormalities. In conclusion, a significant proportion (25%) of ARF-MH in IgAN did not recover the baseline renal function after the disappearance of MH. Duration of MH longer than 10 d, age >50 yr, decreased baseline egfr, absence of previous episodes of MH, and the severity of tubular necrosis were significant risk factors for an incomplete recovery of renal function. Clin J Am Soc Nephrol 2: 51 57, 2007. doi: 10.2215/CJN.02670706 Acute renal failure that is associated with macroscopic hematuria (ARF-MH) is a widely known complication of IgA nephropathy (IgAN) (1 8). Because the glomerular abnormalities usually are mild, most of the authors agree that hematuria by itself is responsible for the ARF s worsening, through a tubular damage that is induced by intratubular erythrocytic casts and a possible nephrotoxic effect of the hemoglobin that is released by these casts (1 6). Most but not all of the previously reported cases spontaneously showed a complete recovery of baseline renal function. As a consequence, ARF-MH is considered a benign complication of IgAN, although some patients can need short-term hemodialysis as a result of the severity of renal failure. However, no studies have analyzed the long-term evolution of renal function or the factors that influence its recovery after this particular type of ARF. In this study, we analyzed the long-term outcome of 36 episodes of ARF-MH that occurred in 32 patients with IgAN, the largest series gathered so far. Importantly, a considerable proportion (25%) of patients did not recover their previous renal function, arguing against the benign character of this complication that so far has been considered. We analyzed factors that were associated with a lack of renal function recovery. Received July 31, 2006. Accepted October 30, 2006. Published online ahead of print. Publication date available at www.cjasn.org. Address correspondence to: Dr. Manuel Praga, Servicio de Nefrología, Hospital Universitario 12 de Octubre, Avda. de Córdoba s/n, 28041 Madrid, Spain. Phone: 34-91-390-8208; Fax: 34-91-390-8383; E-mail: mpragat@senefro.org Materials and Methods We performed a retrospective analysis of all patients who had biopsy-proven IgAN and presented an ARF-MH in the period 1975 through 2005. ARF-MH was defined as an increase of serum creatinine (Scr) of at least 0.5 mg/dl above baseline values that coincided with an episode of MH of glomerular origin. Renal biopsy was obtained in every patient who presented an ARF-MH 3 to 65 d after the onset of macroscopic hematuria (18.5 17.8). No renal biopsies had been performed in any patient before the appearance of ARF-MH. The diagnosis of IgAN was based on the finding of mesangial glomerulonephritis with predominant mesangial deposits of IgA on immunofluorescence. Patients who received a diagnosis of Schönlein-Henoch purpura, acute or chronic liver diseases, alcoholism, systemic lupus erythematosus, or any other systemic disease were excluded. In all of the patients, the presence of urinary tract infection, including renal tuberculosis, was ruled out by urine cultures. Repeated measurements of 24-h urinary calcium and uric acid excretions were performed to exclude hematuria that was induced by hypercalciuria or hyperuricosuria. Intravenous urographies and ultrasound examination of kidneys and urinary tract were normal in every case. Urine phase-contrast microscopy was performed in 25 (69%) ARF-MH episodes; in every case, 80% of erythrocytes were dysmorphic, indicating a glomerular hematuria. The medical records of the patients were reviewed. All of them had been admitted to the hospital during ARF-MH. All of the patients were discharged after complete resolution of MH, which was defined by the complete and persistent disappearance of dark urine. During hospitalization, macroscopic appearance of urine was examined daily. After resolution of MH, the patients were followed at regular intervals (usually every 2 wk in the first visits and thereafter every 6 mo once Scr remained stable) at our outpatient clinic. The following data at admission were obtained from medical records and analyzed for this study: Copyright 2007 by the American Society of Nephrology ISSN: 1555-9041/201 0051

52 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 2: 51 57, 2007 Age, gender, BP, abnormal findings on physical examination, history of previous MH bouts, existence of associated infections that could induce MH, baseline treatment, and day of MH appearance. Analytical study included a complete blood count, routine serum biochemistry profile (including Scr, glucose, uric acid, calcium, phosphorus, sodium, potassium, cholesterol, triglycerides, total proteins, serum albumin, bilirubin, alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, and lactate dehydrogenase), urine sediment examination, urine cultures, 24-h proteinuria, and 24-h urinary excretion of calcium and uric acid. Evolution of these analytical parameters during admission and thereafter during follow-up also were recorded. Estimated GFR (egfr) was calculated by the Modification of Diet in Renal Disease (MDRD-4) formula. All medications that were administered during the ARF-MH, particularly steroids, were recorded carefully, registering their dosage and duration. Duration of MH was determined in every ARF-MH, recording the day of appearance of MH and the day of its disappearance. All patients had at least one Scr determination before the episode of ARF-MH. Last determination of Scr before the onset of MH was defined as baseline Scr. Initial Scr was defined as the value that was obtained at admission. The highest Scr value that was registered during ARF-MH was recorded, as well as the need for acute dialysis. Final Scr was defined as the value that was obtained 6 mo after the episode of ARF-MH. An incomplete recovery of baseline renal function was defined by a Scr value that was higher than at least 25% of the baseline value. Patients were divided in two groups according to whether the recovery of baseline renal function had taken place (group 1, ad integrum recovery of renal function; group 2, incomplete recovery of renal function). Proteinuria levels at the onset of ARF-MH (initial proteinuria), the highest proteinuria value obtained in every case (highest proteinuria), and proteinuria level 6 mo after ARF-MH (final proteinuria) also were recorded. Intervals between onset of MH and highest Scr, cessation of MH and onset of Scr decrease, and cessation of MH and Scr stabilization were calculated in every ARF-MH episode. Renal biopsy specimens were revised for our study by two different pathologists. The severity of mesangial proliferation, interstitial fibrosis, and tubular necrosis was graduated between 0 and (absent, mild, moderate, and severe). The presence of glomerular crescents and red blood cell (RBC) casts that occluded renal tubules was expressed as the percentage of glomeruli or tubules that presented these lesions. Statistical Analyses Data are means SD. For statistical analysis, we used paired and unpaired t test and nonparametric Mann-Whitney test when appropriate. Qualitative variables were analyzed by Fisher test and 2 test. The analysis of data evolution throughout the study period was performed by ANOVA for repeated measurements. Stepwise multiple logistic regression analysis was performed to determine the influence of various parameters on the absence of a complete recovery of baseline renal function. Correlations between quantitative variables were performed with Pearson correlation coefficient. P 0.05 was considered statistically significant. Statistics were calculated using SPSS for Windows, version 11 (SPSS, Chicago, IL). Results Thirty-six episodes of ARF-MH that occurred in 32 patients with biopsy-proven IgAN were identified. Mean age at the onset of ARF-MH was 46.5 20.8 yr (range 18 to 83 yr; Table 1). More than half of the patients (n 19) were older than 50 yr at the onset of episode. Most of the episodes (30 [83%] of 36) occurred in men. A measurement of Scr before the onset of MH (baseline Scr) was available in every episode. In five patients, baseline Scr was 1.3 mg/dl (1.4, 1.4, 1.5, 1.5, and 1.7 mg/dl), expressing a baseline status of chronic renal insufficiency. In the remaining cases, Scr was 1.3 mg/dl. A previous episode of MH was referred by the patient in 47% of ARF-MH. Four patients had two episodes each of ARF-MH. The most common precipitating events of ARF-MH were upper respiratory tract infection in 34%, tonsillitis in 28%, and urinary tract infection in 12%. As expressed in Table 1, mean initial Scr (at admission) was 3.3 2.4 mg/dl (range 0.9 to 12 mg/dl) and highest Scr was 4.8 3.1 mg/dl (range 1.4 to 13.3 mg/dl). Four patients needed several sessions of hemodialysis because of the severity of worsening of renal function. Mean duration of MH was 20 21.5 d (range 3 to 75 d). The mean interval between the onset of MH and highest Scr was 13.1 14.9 d (range from 2 to 50 d). Table 1. Clinical characteristics of the 36 episodes of ARF-MH a Characteristic Value Age (yr) 46.5 20.8 (range 18.0 to 83.0) Gender (M/F) 30/6 Baseline egfr (ml/min per 1.73 m 2 ) 82 26 (range 37 to 156) Baseline Scr (mg/dl) 1.1 0.2 (range 0.7 to 1.7) Initial Scr (mg/dl) 3.3 2.4 (range 0.9 to 12.0) Highest Scr (mg/dl) 4.8 3.1 (range 1.4 to 13.3) Final Scr (mg/dl) 1.3 0.5 (range 0.7 to 2.6) Final egfr (ml/min per 1.73 m 2 ) 76 33 (range 23 to 154) Initial proteinuria (g/24 h) 2.1 2.1 (range 0.3 to 10.5) Highest proteinuria (g/24 h) 2.7 2.4 (range 0.3 to 11.0) Final proteinuria (g/24 h) 0.4 0.6 (range 0.1 to 3.0) Duration of MH (d) 20.0 21.5 (range 3.0 to 75.0) Interval onset MH, highest Scr (d) 13.1 14.9 (range 2.0 to 50.0) Interval end MH, onset of Scr decrease (d) 4.5 10.4 (range 2.0 to 45.0) a ARF-MH, acute renal failure associated with macroscopic hematuria; egfr, estimated GFR; Scr, serum creatinine.

Clin J Am Soc Nephrol 2: 51 57, 2007 Recovery of Renal Function in ARF of IgAN 53 Renal function started to improve after the disappearance of MH in every ARF-MH episode. The mean interval between the end of MH and the onset of Scr decrease was 4. 5 10.4 d (range 2 to 45 d). Stabilization of renal function was observed 41.7 62.9 d after the disappearance of MH (see Table 1) A complete recovery of baseline renal function after urine normalization was observed in 27 ARF-MH (group 1). In the remaining nine (25%) episodes (group 2), Scr did not reach the baseline values, remaining at least 25% higher than the baseline values. Final Scr (value of Scr 6 mo after ARF-MH episode) was 1.0 0.2 mg/dl (range 0.7 to 1.4 mg/dl) in group 1 patients and 1.9 0.4 mg/dl (range 1.2 to 2.6 mg/dl) in group 2 patients (P 0.0001). Clinical characteristics of group 1 and group 2 are shown in Table 2. Group 2 patients were significantly older than those of group 1 (67.2 10.6 versus 39.6 18.6 yr; P 0,01). Baseline Scr was higher in group 2 (1.2 0.3 versus 1 0.2 mg/dl; P 0.04). The duration of MH was significantly longer in group 2 patients (33.7 25.3 versus 15.4 18.4 d; P 0008). A duration of MH longer than 10 d was observed in 88.9% of group 2 patients versus 40.7% in group 1 (P 0.01). The proportion of ARF-MF episodes that lasted 15 d was 77.8% in group 2 and 25.9% in group 1 (P 0.009)). As is shown in Table 2, initial (at admission) Scr and initial proteinuria were significantly higher among group 2 patients. Of the four patients who required acute hemodialysis, an incomplete recovery of baseline Scr (group 2) was observed in two. By univariate analysis, age older than 50 yr, baseline Scr, absence of previous MH episodes, and duration of MH longer than 10 d were the only clinical factors that were statistically associated with the risk for incomplete recovery of previous renal function (Table 3). By multivariate analysis, only a duration of MH longer than 15 d (odds ratio 12.3; 95% confidence interval 1.06 to 143.5; P 0.04) remained as statistically significant risk factor. As expressed in Figure 1, a significant correlation between the duration of MH and the final Scr (Scr 6 mo after the episode of MH) was found (r 0.45; P 0.01). Final Scr was 1.4 0.5 mg/dl after MH longer than 15 d versus 1.0 0.2 mg/dl after episodes shorter than 15 d (P 0.029). Steroid treatment was prescribed in 14 (38.9%) of ARF-MH episodes. The initial dosage usually was 1 mg/kg per d administered orally and then tapered for 2 mo. As is shown in Table 2, there were no significant differences between group 1 and group 2 regarding the percentage of patients who were treated with steroids. By multiple logistic regression analysis, no significant influences of steroid treatment on the recovery of renal function were found. However, steroids in general prescribed were in ARF-MH episodes of long duration: Mean duration of MH before the onset of steroid therapy was 25.5 24.6 (range 5.0 to 75.0) in the 14 episodes of ARF-MH in which they were administered. MH disappeared 13.7 17.4 d (range 1.0 to 45.0 d) after the onset of steroids. A renal biopsy was performed in every ARF-MH episode 18.5 17.8 d (range 3.0 to 65.0) after the onset of MH. Table 4 summarizes the main histologic findings. Glomerular crescents were observed in 15 biopsies (41.8%), but the percentage of glomeruli that were affected by crescents was 25% or less (16.0 6.3 glomeruli; range 0.0 to 25.0). The crescents were cellular, small, and segmental in every case, without compression of the glomerular tuft. The glomeruli showed mild or moderate diffuse mesangial proliferation in all of the cases. Neither proliferative necrotizing lesions nor glomeruli with FSGS were observed. The percentage of glomeruli that showed global glomerulosclerosis was higher in group 2 than in group 1 (16.3 versus 4.1%; Table 4) but without statistical significance. Immunofluorescence showed predominant mesangial deposits of IgA in every case. The most striking histologic abnormalities were observed in renal tubules (Figure 2): Almost a half (45.1 24.5%) were filled by Table 2. Clinical characteristics of group 1 (complete recovery of baseline renal function) and group 2 (incomplete recovery) patients Characteristic Group 1 (n 27) Group 2 (n 9) P Age (yr) 39.6 18.6 (18 to 73) 67.2 10.6 (55 to 83) 0.01 Patients 50 yr (%) 37 100 0.001 Gender (M/F) 23/4 7/2 NS Baseline Scr (mg/dl) 1.0 0.2 (0.7 to 1.5) 1.2 0.3 (0.7 to 1.7) 0.04 Baseline egfr (ml/min per 1.73 m 2 ) 88 24 64 23 0.01 Highest Scr (mg/dl) 4 2.8 (1.4 to 13.3) 7.2 3 (4.4 to 12.1) 0.003 Final Scr (mg/dl) 1 0.2 (0.7 to 1.4) 1.9 0.4 (1.2 to 2.6) 0.0001 Final egfr (ml/min per 1.73 m 2 ) 84 28 38 12 0.0005 Duration of MH (d) 15.4 18.4 (3.0 to 75.0) 33.7 25.3 (7.0 to 75.0) 0.008 Initial proteinuria (g/d) 1.9 2.1 (0.3 to 10.5) 2.5 2.2 (0.2 to 7.0) NS Final proteinuria (g/d) 0.3 0.3 (0.1 to 1.5) 0.9 1 (0.2 to 3.0) 0.03 Steroid treatment (%) 33.3 55.6 NS Previous MH episode (%) 48.1 11.1 0.05 Duration of MH 7 d (%) 63.0 100.0 NS Duration of MH 10 d (%) 40.7 89.0 0.01 Duration of MH 15 d (%) 25.9 77.8 0.009

54 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 2: 51 57, 2007 Table 3. Variables that influence the risk for an incomplete recovery of renal function a Variable Univariate Analysis Multivariate Analysis P OR 95% CI P OR 95% CI Age 0.46 Age 50 yr 0.001 9.2 1.3 to 62.0 NS Gender 0.618 Baseline Scr 0.002 1.22 1.0 to 1.7 NS Baseline egfr 0.02 1.04 1.0 to 1.1 NS Previous MH episode 0.05 0.98 0.96 to 0.99 NS Severity of tubular necrosis 0.03 6.4 1.1 to 37.1 NS Steroid treatment 0.24 Duration of MH 10 d 0.01 11.6 1.2 to 106.0 NS Duration of MH 15 d 0.005 13.8 1.7 to 110.2 0.04 12.3 1.06 to 143.5 a CI, confidence interval; OR, odds ratio. Figure 1. Correlation between duration of macroscopic hematuria (MH; days) and final serum creatinine (Scr; Scr 6 mo after onset of acute renal failure associated with MH). RBC casts. Signs of acute tubular necrosis (dilation of the lumen, flattening and desquamation of the lining epithelial cells, cytoplasmic degeneration, nuclear pyknosis, and multiple mitosis) were observed in renal tubules, mainly in those that were filled by RBC casts. There was a significant difference in the number of renal biopsies that showed moderate tubular damage in group 1 (28.5%) and group 2 (77.8%; P 0.02), and by univariate analysis, the severity of tubular necrosis was statistically associated with the risk for incomplete recovery of previous renal function (Table 3). No significant differences in the degree of mesangial proliferation, global glomerulosclerosis, crescents, tubules with RBC casts, or interstitial fibrosis were found (Table 4). Table 5 summarizes the main pathologic findings of renal biopsies that were obtained in group 2 patients (incomplete recovery of renal function after MH cessation). Discussion The most salient finding of our study is that a considerable proportion (25%) of patients who had IgAN and experienced an ARF-MH episode do not completely recover the baseline renal function, showing chronic renal insufficiency of variable severity after the resolution of the episode. These findings are important because previous descriptions have reported a spontaneous recovery of normal renal function after cessation of gross hematuria, with no need for any therapeutic intervention (1 8). Bouts of MH are a characteristic complication of IgAN. Most of them are triggered by infections, mainly located in the upper respiratory tract, and may have a recurrent presentation (9,10). Although alarming and disturbing for the patient, the longterm repercussion of these episodes seems to be relatively benign. Studies that have been performed in large cohorts of patients with IgAN have failed to find any long-term detrimental influence of these episodes (11). However, coinciding with an episode of gross hematuria, some patients develop an acute worsening of renal function of variable severity. Although the number of MH bouts that precipitate a severe, oligoanuric, ARF in IgAN is considered to be low, a previous study of our group suggested that milder forms can be more common than suspected (2). Previous descriptions of ARF-MH, including ours, stressed the spontaneous recovery of baseline renal function after cessation of gross hematuria (1 8). Only two cases, previously reported by our group (12), showed a persistent chronic renal insufficiency after an initial improvement of renal function that coincided with the end of MH. In this study, we found that one fourth of the patients in our series did not recover their baseline renal function. Duration of MH was associated with the risk for an incomplete recovery of renal function, by both univariate and multivariate analysis. As is shown in Table 2, all group 2 patients (with an incomplete recovery of renal function) had duration of MH longer than 7 d, and it was longer than 15 d in almost 80% of them. In addition, a significant correlation between the duration of MH and the final Scr was found (Figure 1). However, it should be stressed that age 50 yr, baseline Scr and egfr, and the absence of previous episodes of MH also were significant clinical risk factors by univariate analysis. These findings indicate that previously damaged kidneys (by

Clin J Am Soc Nephrol 2: 51 57, 2007 Recovery of Renal Function in ARF of IgAN 55 Table 4. Histologic findings a Finding Group 1 (n 27) Group 2 (n 9) P Mesangial proliferation (%) NS absent 4.8 0 mild 37 77.8 moderate 37 22.2 severe 0 0 Glomeruli with global glomerulosclerosis (%) 4.1 16.3 NS Glomeruli with crescents (%) 3.6 7.4 NS Tubules with RBC casts (%) 36.6 53.7 NS Tubular necrosis (%) 0.02 absent 9.6 0 mild 61.9 11.1 moderate 28.5 77.8 severe 0 11.1 Interstitial fibrosis (%) NS absent 48.2 22.2 mild 48.1 55.6 moderate 3.7 22.2 severe 0 0 a RBC, red blood cell. Figure 2. Tubules filled by red blood cells, showing diminished number of lining epithelial cells and nuclear pyknosis. Magnification, 250 (hematoxylin-eosin). age or hypertension-related nephrosclerosis) seem to be more sensitive to the detrimental effect of a prolonged MH. The finding of duration of MH as a significant risk factor to determine the recovery of previous renal function agrees well with the pathogenesis proposed for these episodes of ARF: Renal biopsies that are performed during these ARF-MH of IgAN characteristically show important tubular lesions, whereas the glomerular abnormalities cannot explain the acute worsening of renal function (1 6). Tubular abnormalities consist in a high proportion of tubules that are filled by RBC casts and signs of tubular necrosis that are more evident in tubules that are occupied by RBC casts. On the basis of these findings, we and others proposed that glomerular hematuria might damage tubular cells through several mechanisms: Direct tubular obstruction by RBC casts, tubulotoxic effect of hemoglobin or other substances that are released by broken erythrocytes in the tubular lumen, and erythrophagocytosis by tubular epithelial cells (1,2,4,13). In this way, if glomerular bleeding by itself is responsible for ARF-MH episodes through its tubulotoxic effects, then a prolonged duration of hematuria could induce a more severe and partially irreversible tubular damage. Renal tubules of elderly patients with preexisting chronic renal insufficiency might be more unable to recover their functional integrity after such prolonged insult. It is interesting to remark that similar cases of ARF as a result of acute tubular necrosis by RBC casts have been reported in patients with glomerular hematuria caused by glomerular diseases other than IgAN, such as Henoch-Schönlein purpura, acute postinfectious glomerulonephritis, focal necrotizing glomerulonephritis, and Goodpasture syndrome (4,14 16). The histologic findings of our study are in keeping with previous studies of our group and other investigators: A considerable proportion of renal tubules (42.6 24.5%) were filled by RBC casts, and characteristic lesions of acute tubular necrosis were observed mainly in tubules that were occupied by RBC casts. The histologic severity of tubular necrosis was significantly worse in group 2 patients (Table 4), and it was a significant risk factor for an incomplete recovery of renal function (Table 3). Glomerular lesions were mild in general, showing the typical mild or moderate mesangial proliferation of IgAN. It has been reported that renal biopsies that are performed closely to an episode of MH in patients with IgAN show glomerular

56 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 2: 51 57, 2007 Table 5. Histologic findings in group 2 patients (incomplete recovery of renal function) a Patient Duration of MH (d) Mesangial Proliferation %GS % Glomeruli with Crescents % Tubules with RBC Casts Tubular Necrosis Fibrosis Final Scr b (mg/dl) 1 37 0 0 100 1.8 2 21 50 0 50 2.6 3 11 61 23 50 1.9 4 15 25 19 50 2.1 5 75 0 25 50 1.9 6 60 0 0 50 2.2 7 17 0 0 30 1.2 8 7 11 0 50 1.7 9 60 0 0 50 2.0 a GS, glomerulosclerosis. b Scr 6 mo after the onset of ARF-MH. crescents in a higher proportion than those that are performed out of gross hematuria bouts (17); however, with the exception of some patients with very aggressive clinical courses, the number of glomeruli that are affected by crescents usually is low. Our study confirms this point: Crescents, when present, affected only a minority of glomeruli. The direct relationship between the duration of MH and the persistence of renal damage that we have found in this study has important clinical implications. According to our data, ARF-MH episodes that last longer than 10 to 15 d, particularly in elderly patients and in patients with previously damaged kidneys, are accompanied by a high risk for an incomplete recovery of baseline renal function. Therefore, therapeutic interventions that aim to reduce the duration of MH clearly would be indicated. The problem is that clinical data about therapeutic interventions that could shorten the duration of gross hematuria episodes in IgAN are scarce. Intuitively, highdosage steroids could induce this effect; however, although steroid administration has proved to be beneficial for the longterm outcome of patients with IgAN (18), its effect on the duration of gross hematuria episodes has not been tested in an appropriate prospective design. In our study, 14 (38.9%) ARF-MH episodes were treated with steroids. No beneficial effect of steroid treatment on the recovery of renal function was observed, but it should be considered that steroids were administered in cases of ARF-MH of longer duration, introducing a bias that invalidates statistical analysis. Our data might suggest, however, that steroid administration shortened the duration of MH: Duration of gross hematuria before the onset of steroid treatment was 25.5 24.6 d in the 14 episodes of ARF-MH that were treated; it disappeared 13.7 17.4 d after the onset of steroid treatment. Therefore, it might be reasonable to recommend the administration of a course of steroids in ARF-MH episodes that are longer than 10 to 15 d, particularly in patients who are older than 50 yr or in those with baseline renal insufficiency, considering the high risk for persistent chronic renal insufficiency after such prolonged episodes and the ominous significance for long-term prognosis of such chronic renal insufficiency in IgAN (19). However, no conclusive recommendations can be extracted from our retrospective data; prospective and randomized trials are necessary to ascertain the influence of steroids in ARF-MH of IgAN. Our study also provides new clinical information about the characteristics of ARF-MH of IgAN. Although the duration of MH episodes is considered to be short usually, resolving in a few days, our study shows that they can be strikingly prolonged, lasting for 2 mo in some cases. In addition, recurring MH episodes are classically considered as a characteristic presentation of young patients in the earlier course of IgAN (9 11). By contrast, more than half of ARF-MH episodes in our study developed in patients who were older than 50 yr, the oldest one being 83 yr of age. Finally, because recurring MH is considered a complication of the earlier stages of IgAN, ARF-MH episodes are thought to appear almost exclusively in patients with baseline normal renal function. Our study shows that this statement is not always correct: Five ARF-MH episodes developed in patients with a Scr 1.3 mg/dl. However, considering the significant number of elderly patients in our cohort, this baseline renal insufficiency might be secondary to age-related nephrosclerosis rather than to advanced IgAN by itself. Conclusion A significant proportion (25%) of ARF-MH in IgAN did not recover the baseline renal function after the disappearance of MH. A longer duration of MH, together with age 50 yr, decreased baseline egfr, the absence of previous MH episodes, and the finding of a severe tubular necrosis in renal biopsy were the most conclusively implicated parameters in this untoward outcome. Acknowledgments Portions of this study were presented at the American Society of Nephrology Renal Week, November 8 through 13, 2005, Philadelphia, PA; and at the Congress of the European Renal Association-European Dialysis and Transplant Association, July 15 to 18, 2006, Glasgow, Scotland. Disclosures None.

Clin J Am Soc Nephrol 2: 51 57, 2007 Recovery of Renal Function in ARF of IgAN 57 References 1. Kincaid-Smith P, Bennett WM, Dowling JP, Ryan GB: Acute renal failure and tubular necrosis associated with hematuria due to glomerulonephritis. Clin Nephrol 19: 206 210, 1983 2. Praga M, Gutierrez-Millet V, Navas JJ, Ruilope LM, Morales JM, Alcazar JM, Bello I, Rodicio JL: Acute worsening of renal function during episodes of macroscopic hematuria in IgA nephropathy. Kidney Int 28: 69 74, 1985 3. Delclaux C, Jacquot C, Callard P, Kleinknecht D: Acute reversible renal failure with macroscopic haematuria in IgA nephropathy. Nephrol Dial Transplant 8: 195 199, 1993 4. Fogazzi GB, Imbasciati E, Moroni G, Scalia A, Mihatsch MJ, Ponticelli C: Reversible acute renal failure from gross haematuria due to glomerulonephritis: Not only in IgA nephropathy and not associated with intratubular obstruction. Nephrol Dial Transplant 10: 624 629, 1995 5. Praga M, Costa JR, Shandas GJ, Martinez MA, Miranda B, Rodicio JL: Acute renal failure in cirrhosis associated with macroscopic hematuria of glomerular origin. Arch Intern Med 147: 173 174, 1987 6. Packham DK, Hewitson TD, Yan HD, Elliott CE, Nichols K, Becker GJ: Acute renal failure in IgA nephropathy. Clin Nephrol 42: 349 353, 1994 7. Talwalkar YB, Price WH, Musgrave JE: Recurrent resolving renal failure in IgA nephropathy. J Pediatr 92: 596 597, 1978 8. Francois B, Cohen R: Acute renal failure and IgA nephropathy. Clin Nephrol 19: 269 270, 1983 9. Galla JH: IgA nephropathy. Kidney Int 47: 377 387, 1995 10. Donadio JV, Grande JP: IgA nephropathy. N Engl J Med 347: 738 748, 2002 11. Droz D: Natural history of primary glomerulonephritis with mesangial deposits of IgA. Contrib Nephrol 2: 150 157, 1976 12. Mazuecos A, Ramos PC, Campo C, Sanchez R, Martinez MA, Praga M: Chronic renal insufficiency after and episode of macroscopic hematuria in IgA nephropathy. Nephron 64: 452 455, 1993 13. Hill PA, Davies DJ, Kincaid-Smith P, Ryan GB: Ultrastructural changes in renal tubules associated with glomerular bleeding. Kidney Int 36: 992 997, 1989 14. Kobayashi Y, Omori S, Kamimaki I, Ikeda M, Akaoka K, Honda M, Ogata K, Morikawa Y: Acute reversible renal failure with macroscopic hematuria in Henoch-Schonlein purpura. Pediatr Nephrol 16: 742 744, 2001 15. Michail S, Stathakis C, Nakopoulou L, Aroni K, Revenas K, Avdikou K, Vaiopoulos G: Acute renal failure in an adult patient with Henoch-Schonlein episode of macroscopic hematuria. Ren Fail 21: 107 111, 1999 16. Walker RG, Becker GJ, d Apice AJ, Kincaid-Smith P: Treatment of oligoanuric anti-gbm nephritis. Lancet 1:508.1986 17. Bennett WH, Kincaid-Smith P: Macroscopic hematuria in mesangial nephropathy: Correlation with glomerular crescents and renal dysfunction. Kidney Int 23: 393 400, 1983 18. Pozzi C, Bolasco PG, Fogazzi GB, Andrulli S, Altieri P, Ponticelli C, Locatelli F: Corticosteroids in IgA nephropathy: A randomised controlled trial. Lancet 13: 883 887, 1999 19. Bartosik LP, Lajoie G, Sugar L, Cattran DC: Predicting progression in IgA nephropathy. Am J Kidney Dis 38: 728 735, 2001