We previously reported a partial IgG subclass deficiency (mainly involving IgG1) in proteinuric IgAN and HSP [10]. We postulated that such an IgG imbalance might favor viral and bacterial infections [11, 12], a known trigger of flare-ups in these diseases, or activate noxious isotypic compensatory mechanisms [13]. Because high-dose intravenous immunoglobulins are effective in some human immune-mediated diseases [14], we began a prospective trial of high-dose intravenous immunoglobulins in patients with severe IgAN and HSP who had indicators of poor prognosis.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

Ten men and one woman were enrolled in the study. The criteria for eligibility were idiopathic IgAN or HSP at histologic stage II/III, III, III/IV, or IV; proteinuria level greater than 2 g/d; and glomerular filtration rate greater than 35 mL/min per 1.73 m². The local ethics committee approved the trial and all patients gave informed consent. Nine had IgAN and two had HSP. Five patients with IgAN were referred by their general practitioner for heavy proteinuria, whereas the remaining four were referred for severe IgA nephropathy by the nephrology units of two general hospitals. One patient with HSP was found to have nephritis while being treated in the dermatology unit of our institution, and the second was referred to our nephrology unit by his general practitioner for HSP with renal involvement. The median age was 24 years (range, 18 to 36 years). The median time between diagnosis and therapy was 5 months (range, 1 to 18 months) (Table 1). All patients had been treated for focal infections and had tonsillectomy for chronic tonsillitis or for reasons of immunity [15]. Two patients were treated unsuccessfully with high-dose steroids at least 3 months earlier. No patient received nonsteroidal anti-inflammatory drugs in the year before the trial. Nine of the 11 patients were treated for hypertension before entry into the trial, and their blood pressure levels rapidly returned to normal during therapy with ß-blockers alone (atenolol or betaxolol) or combined with prazosin; these drugs were pursued during the trial. None of the patients had severe hypertension: Results of the fondus oculi, hemogram, and heart ultrasound examinations were normal. Treatment of hypertension did not reverse the decline in renal function, despite prolonged follow-up before immunoglobulin therapy. Nine of the 11 patients had altered renal function (median glomerular filtration rate of the group, 57.5 mL/min per 1.73 m²; range, 39 to 134 mL/min per 1.73 m²). The median monthly rate of renal function decline before therapy was –3.78 mL/min (range, –13.80 to 0 mL/min) Table 1, with a median total decline in renal function before therapy of –27.5 mL/min (range, –55 to 0 mL/min). All patients had heavy proteinuria (median protein level, 5.20 g/d; range, 2.37 to 10.70 g/d) Table 1, a high histologic grade (1 with stage II/III; 4 with stage III; 3 with stage III/IV; 3 with stage IV), a high histologic index of activity (median, 5; range, 2 to 10), and a moderate histologic index of sclerosis (median, 4; range, 0 to 9) (Table 1). All had dense IgA and C3 mesangial deposits; four patients also had IgA deposits along the glomerular capillary wall. Immunoglobulin G deposits were absent, whereas mild IgM deposits were found in 5 of the 11 patients. The two patients with HSP also had articular effusion, extensive cutaneous purpura, abdominal pain, and diarrhea.

Intervention

The patients received 1 g/kg of body weight per day (in a 12-hour infusion) of pepsin-pH 4 intravenous immunoglobulins (Biotransfusion, Roissy, France) on 2 successive days each month for 3 successive months, followed by intramuscular immunoglobulins (Polygamma, 16.5%; Association Nationale pour la Distribution des Fractions Plasmatiques Humaines, Paris, France) at a dose of 0.35 mL/kg of body weight twice each month for another 6 months. To avoid serum sickness related to formation of immune complexes between free circulating environmental antigens and infused IgG, a complication previously reported with IgG deficiency [16], the first intravenous immunoglobulin infusion was preceded by single 3-L plasmapheresis with albumin replacement. We examined patients monthly during their hospital stay for the first 3 months, as outpatients from the third to the eighth month, and in the hospital at the end of the study (ninth month). After the 9-month trial, patients received the same maintenance therapy (0.35 mL/kg of intramuscular immunoglobulins twice each month) and were examined every 3 months as outpatients. Angiotensin-converting-enzyme inhibitors, calcium channel blockers, corticosteroids, and nonsteroidal anti-inflammatory drugs were not allowed during the trial. The patients ate a normal protein diet (1.5 to 2 g/kg per day of protein).

Follow-up

We assessed clinical, renal, viral, and immunologic tolerance at each visit during the 9-month study period and then every 3 months thereafter.

Measurements

Renal Histologic Findings

Global Histologic Grading: We used the grading system of Lee and associates [7], which requires that at least nine glomeruli be scored. Biopsy specimens were blindly scored in five grades of increasing severity [7], accounting for both cellular proliferation and sclerosis. Stage I corresponded to normal glomeruli or occasional segmental mesangial thickening with or without cellularity. In stage II, fewer than one half of the glomeruli showed localized mesangial proliferation and sclerosis. Stage III was characterized by diffuse mesangial proliferation and thickening with focal and segmental variation, and focal interstitial edema and infiltrates may be present as small crescents and adhesions. In stage IV, marked diffuse mesangial proliferation and sclerosis were evident, with crescents in as many as 45% of the glomeruli. Partial or total glomerulosclerosis was frequently seen as tubular atrophy and interstitial infiltrates. Lesions classified as stage V were the same as in stage IV but more severe. Some biopsy specimens were given intermediate scores.

Histologic Activity Index: We noted proliferation of mesangial and epithelial cells blindly using a final scale of 14 points; we required that at least nine glomeruli be examined. In evaluating mesangial cell proliferation, we accounted for the intensity (absent = 0; mild = 1; moderate = 2; severe = 3) and extent of the lesions (no glomeruli affected = 0; <25% of glomeruli = 1; >25% but <50% = 2; >50% but <75% = 3; >75% = 4). We noted epithelial cell proliferation in the same way.

Histologic Sclerosis Index: We evaluated two parameters blindly (mesangial plus glomerular sclerosis, and interstitial sclerosis) and required that at least nine glomeruli be scored. Mesangial sclerosis accounted for both the intensity (absent = 0; moderate = 1; severe = 2) and extent of the lesions (no glomeruli affected = 0; <25% of the glomeruli = 1; >25% but <50% = 2; >50% but <75 % = 3; >75% = 4). With regard to interstitial sclerosis, we evaluated interstitial fibrosis on a 2-point scale (absent = 0; moderate = 1; severe = 2) and tubular atrophy in the same way (absent = 0; moderate = 1; severe = 2).

Immunofluorescence Studies: We used a direct immunofluorescence technique with fluorescein-labeled goat antisera derived from the same batches (Institut Pasteur, Marnes la coquette, France, and Behring Institut, Marburg, Germany) directed against IgA, IgG, IgM, C1q, C3, and C4. The fluorescence intensity was scored blindly from 0 to 3.

Biochemical Measurements of Renal Disease Activity

We assayed protein levels at 24 hours using a dye-based method derived from Bradford's technique. We determined the proteinuria selectivity index by studying the IgG:albumin clearance ratio, which we measured using a nephelometer and specific polyclonal antisera (Behring Institut). We measured urinary erythrocyte and leukocyte counts during a 3-hour morning rest period and assayed urinary fibrin-fibrinogen degradation products according to the method of Merskey and Kleiner [17]. Using a specific enzymatic method, we measured plasma creatinine. We assessed global renal function through measurements of creatinine clearance (according to Cockroft and Gault [18]) and an inulin analog clearance (polyfructosan). Because evaluation of renal function decline with both the 1/creatinine ratio and the slope of the glomerular clearance curve was recently criticized (because the decline in renal function is not always linear [19, 20]), we evaluated the rate of renal function decline as follows: loss in glomerular filtration rate during the period studied (mL/min per 1.73 m²) divided by the number of months.

Serologic Measurements of Activity of IgA Nephropathy

We measured IgA and IgM using a nephelometer and specific polyclonal antisera (Behring Institut). We assayed total IgE using radioimmunoassay (Pharmacia, Uppsala, Sweden). Immunoglobulin G subclasses were measured by a sandwich enzyme-linked immunosorbent assay (ELISA) in which plates were coated with affinity-purified goat IgG against mouse IgG (human protein-adsorbed, Jackson Immunoresearch, West Grove, Pennsylvania). We used a mouse monoclonal antibody specific for each IgG subclass as capture antibody (clones: HP 6186 [IgG1], GOM1 (IgG2), ZG4 (IgG3), and RJ4 [IgG4]). The revealing antibody was an alkaline phosphatase-labeled, affinity-purified, anti-human IgG goat IgG (mouse protein-adsorbed, Zymed, San Francisco, California). We used the World Health Organization reference preparation 67/97 for the calibration curves. Neopterin was assayed by radioimmunoassay (Behring), whereas ß-2 µglobulin and interleukin-2 soluble receptors were measured by specific ELISA (Pharmacia and Boehringer, Mannheim, Germany). We measured IgA immune complexes according to the method of Imai and colleagues [21] after 14% polyethylene glycol 6000 precipitation (7% final concentration) by a specific sandwich ELISA. We assayed polymeric IgA by the immunoelectrophoretic method of Meillet and coworkers [22], whereas we measured serum secretory IgA by ELISA [23]. We detected rheumatoid factors of the IgA class using an indirect ELISA according to the method of Sinico and associates [24] and Schena and colleagues [25] with the following antigens: rabbit IgG, human IgG, and human IgG Fab'2, purified by precipitation, ion exchange, and affinity chromatography, respectively. Immunoglobulin A1 antigliadin and anti--lactalbumin were assayed by an indirect ELISA [26]. We determined normal values for these variables in the sera of 30 healthy controls.

Markers of Viral, Renal, and Immunologic Tolerance of Immunoglobulin Therapy

We measured liver enzymes routinely; hepatitis B surface antigen and antibodies against hepatitis C virus, hepatitis B virus, and human immunodeficiency virus were detected using specific ELISA methods (Abbott, Abbott Park, Illinois). We assayed complement fractions C3 and C4, rheumatoid factors of IgG and IgM classes, cryoglobulin, antinuclear antibodies, and antineutrophil cytoplasmic antibodies using classic methods. We determined renal tolerance by measuring creatinine, urea, the proteinuria selectivity index, and enzymuria (alanine aminotranspeptidase; N-acetyl glucosaminidase) during the intravenous immunoglobulin infusions and by periodic analysis of creatinine and urea in the days after infusions.

Evaluation of Immunoglobulin Therapy

We compared renal histologic findings, biochemical variables of renal disease activity (except for the decline in renal function), and serum immunologic variables of IgAN activity before and after 9 months of therapy. Evaluation of the decline in renal function accounted for the latest data available for the patients' renal function.

Statistical Analyses

We analyzed the paired variables using the nonparametric Wilcoxon test (Statview 4.0; Abascus Concept) [27]. We calculated the median values of protein levels and the urinary erythrocyte and leukocyte counts in the 3 months before and after therapy. We analyzed the correlation between creatinine clearance and polyfructosan clearance in eight patients using the Spearman test (Statview 4.0) [27]. Values of the variables are expressed as medians and ranges. For IgA and C3 deposits, we did not include the values of patient 1 in the median of the group before therapy.

Results

Top Methods Results Discussion Author & Article Info References

Immunoglobulin therapy induced a decrease in proteinuria, hematuria, and leukocyturia Table 1 but no change in the proteinuria selectivity index (before therapy: median, 0.230 [range, 0.12 to 0.55]; after therapy: 0.235 [range, 0.11 to 0.50]) or in urinary fibrin-fibrinogen degradation products (before therapy: median, 0.157 µg/min [range, 0.03 to 2.0]; after therapy: 0.069 µg/min [range, 0.03 to 0.155]). We measured urinary variables a median of 8 times before therapy (range, 3 to 13 times) and 7 times after starting treatment (range, 4 to 13 times). Immunoglobulin therapy also induced a significant decrease in the histologic index of activity and an increase in the histologic sclerosis index (Table 2); (Figure 1). The median time between the renal biopsy and beginning of therapy was 4 months (range, 0.5 to 18 months) and only exceeded 6 months in two cases. Four patients had two biopsies during the course of the disease before therapy. We also observed a significant decrease in the intensity of both IgA and C3 deposits (Table 2). Immunofluorescence studies were done in three of the four cases in which IgA deposits were found along the glomerular capillary walls and were absent from the control biopsy specimens in two cases. Immunoglobulin M deposits were unmodified by immunoglobulin therapy. Immunoglobulin G deposits were absent from all the follow-up biopsy specimens. Sclerotic lesions appeared as after-effects of epithelial cell proliferation (crescent formation) and tuft adhesion to the Bowman capsule; indeed, progression toward sclerotic lesions appears to be common in proliferative glomerulonephritis, as it is with severe lupus nephritis and rapidly progressive glomerulonephritis (crescentic glomerulonephritis) [28].

View larger version (17K): [in this window] [in a new window]   Figure 1. Changes in the histologic activity index during high-dose immunoglobulin therapy. The activity index is noted on a final 14-point scale. Values for each patient before and after intervention are given on the vertical axis.

Systemic symptoms in the two patients with HSP disappeared 2 months after immunoglobulin therapy was begun.

The glomerular filtration rate did not differ before or after 9 months of therapy (before: median, 57.5 mL/min [range, 39 to 134 mL/min]; after: median, 61.5 mL/min [range, 26 to 123 mL/min]), but the rate of decline in the glomerular filtration rate was greatly reduced, from –3.78 mL/min per month before therapy to 0 mL/min per month at the final evaluation (at least 9 months after the first intravenous immunoglobulin infusion) (Table 1); Figure 2 and Figure 3. Creatinine clearance and polyfructosan clearance were highly correlated (P < 0.001, Spearman test; Spearman rank correlation coefficient: rho = 0.9).

View larger version (15K): [in this window] [in a new window]   Figure 2. Monthly rate of decline in renal function during high-dose immunoglobulin therapy. The rate of decline in renal function was evaluated as the decrease in the glomerular filtration rate during the period studied (mL/min per 1.73 m2 per number of months). Values for each patient before and after intervention are given on the vertical axis. The evaluation of the decline in renal function after intervention accounted for the latest data available for each patient. Patient 5 was omitted because we have no follow-up data on renal function more than 1 month before immunoglobulin therapy (see Table 1).

View larger version (16K): [in this window] [in a new window]   Figure 3. Change in glomerular filtration rate during high-dose immunoglobulin therapy in patients with deteriorating renal function. The glomerular filtration rate is expressed on the vertical axis in mL/min per 1.73 m2. The Y axis represents the time before and after the intervention, expressed in months. The latest data available on renal function for each patient are given. Patients 4 and 11 are omitted because their renal function was normal and stable before and after intervention. Patient 5 is omitted because we had no data earlier than 1 month before immunoglobulin therapy (see Table 1).

Viral and immunologic tolerance of immunoglobulin therapy was satisfactory. Indeed, complement fractions remained normal and no rheumatoid factors of the IgG and IgM classes, cryoglobulinemia, antinuclear antibodies, or antineutrophil cytoplasmic antibodies were detected; hepatic enzyme activities remained normal. No antibodies to human immunodeficiency virus, hepatitic C virus, or hepatitis B surface antigen were detected. Passively acquired antihepatitis B surface antigen antibodies were always present and allowed us to evaluate therapeutic compliance. We observed a slight transient increase in blood creatinine levels but not in those of urea in two patients after intravenous immunoglobulin infusion, probably a result of acute proximal tubular dysfunction shown by the increase in alanine aminotranspeptidase activity in the overall group. The good renal tolerance of intravenous immunoglobulins and the moderately acute reversible tubular dysfunction were reported previously [29]. Moderate fever and shivers occurred in most patients when the rate of infusion was too high, but they were controlled with paracetamol and reduction of the infusion rate.

Immunoglobulins A, E, G3, and G4 were unmodified, although IgM, IgG1, and IgG2 levels increased substantially (Table 3). The serum ß-2 µglobulin/creatinine ratio was the only marker of immunologic activation that was reduced during therapy (Table 3). Although analysis of the serologic variables of IgAN activity may help us understand the mechanisms of immunoglobulin therapy, they are not known to be related to clinical phenomena or to disease severity [2, 4, 5].

Discussion

Top Methods Results Discussion Author & Article Info References

In this prospective, open trial of immunomodulation with polyvalent immunoglobulins in IgAN, we observed no side effects of therapy; in particular, we found no renal degradation, flare-up of renal disease, or serum sickness.

We treated patients with a combination of intravenous immunoglobulins for 3 months followed by maintenance therapy of intramuscularly administered immunoglobulins. We postulated that intramuscular immunoglobulins might be active only after administration of intravenous immunoglobulins and that their use might decrease intravenous administration.

Intravenous immunoglobulins have been shown to be effective in several immune-mediated diseases such as autoimmune thrombocytopenia, anemia, and neutropenia; antihemophilic inhibitor diseases; neurologic diseases such as myasthenia gravis, demyelinating neuropathies, and the West syndrome; systemic lupus erythematosus; rheumatoid arthritis; and Kawasaki disease [14], but intramuscularly administered immunoglobulins have also been effective in treating myasthenia gravis and intractable epilepsy [30, 31]. Intravenous immunoglobulins have been effective in treating severe systemic lupus erythematosus and positive antineutrophil cytoplasmic antibody nephritis that is unresponsive to conventional therapy [32, 33] and in treating idiopathic membranous nephropathy, the hemolytic-uremic syndrome [34, 35], a few cases of type I membranoproliferative glomerulonephritis, and one case of IgA crescentic nephritis secondary to cirrhosis [36, 37].

The effects of our study are probably not the result of random variation because the patients included had a severe form of nephropathy, and most had declining renal function. As shown in Table 1, eight of our patients had a high rate of monthly decline in renal function before entering the trial, which was positively influenced by therapy. The course of disease in these patients when treated with immunoglobulin therapy clearly differed from the natural course of such severe forms of disease [2, 4, 7, 9]. Moreover, the hypertension in our patients was moderate and was corrected in less than 1 month using ß-blockers alone or combined with low-dose prazosin, neither of which is known to have protective effects on renal function. Despite the correction of hypertension for a few months, renal function continued to decline and the slope of the glomerular clearance curves did not change (follow-up of patients before therapy: months –10 to –4;(Figure 3). Moreover, the patients had not received nonsteroidal anti-inflammatory drugs in the year before the trial, ruling out a possible role in the decline of renal function before the trial and in its improvement during immunoglobulin therapy.

Immunoglobulin therapy in patients with IgAN only seems to arrest disease progression, as is the case in many immune-mediated diseases [14], because noncompliance in two patients (shown by the disappearance of anti-hepatitis B surface antigen antibodies) was followed by relapse 2 or 3 months later. No comparable data are available for repeated biopsies in IgAN after such a short time. Only two studies involved repeated renal biopsy in patients with IgAN, and the patients and clinical settings were different from those in our study [38, 39].

Unfortunately, immunologic therapy has not been effective in severe cases of IgAN. Indeed, only treatment of focal sepsis and management of hypertension (with angiotensin-converting-enzyme inhibitors) seem to protect renal function [3]. High-dose steroids, alone or combined with cyclophosphamide, and plasma exchange are only effective in the few cases of true crescentic IgA nephritis [40, 41]. Steroids are effective in cases of minimal-change nephropathy with IgA deposits, but their efficacy in some uncontrolled studies has not been confirmed in a controlled trial [42-45]. The results of a controlled trial of steroid therapy in severe IgAN should clarify this point [46]. Renal biopsies done after corticosteroid treatment showed identical lesions with no change in the intensity of immune deposits [39, 44] or, in one study, a decrease in the activity index with no overall change in immunofluorescence [43]. In contrast, the activity index and IgA and C3 deposits were reduced in our patients. Triple therapy with cyclophosphamide, dipyridamole, and warfarin had no meaningful protective effect on renal function, despite a slight decrease in proteinuria [47], which was attributed to dipyridamole [48]. Azathioprine and chlorambucil were ineffective in a prospective, controlled trial [49]. Finally, in a recent study, cyclosporine decreased proteinuria in patients with IgAN, but this was related to a hemodynamic mechanism (shown by the deterioration of renal function during treatment) rather than to an immunologic mechanism [50].

In IgAN, a basic dysregulation of IgA production occurs, as suggested by the increased number of IgA-bearing lymphocytes in the peripheral blood, bone marrow, and tonsils [6, 15, 51, 52]. Several autoimmune anomalies have been reported in IgAN: IgA rheumatoid factors directed against IgGFc [24] or IgG Fab'2 [25], antiendothelial cell antibodies, polyspecific IgA antibodies, anti-mouse laminin IgA, IgA antibodies against denatured DNA, and antimesangial cell antibodies of the IgG class [51-53]. The cytokine network is also abnormal, including increased interleukin-2 and interferon- production by peripheral lymphocytes [54] and increased urinary interleukin-6 excretion related to release by mesangial cells [55]. The possible mechanisms of intravenous immunoglobulins in autoimmune disorders might also be involved in IgAN and HSP therapy and they probably act synergistically. Indeed, immunoglobulin therapy inhibits B-cell differentiation and immunoglobulin production by acting on the B-cell IgG Fc receptor [14, 56] and may regulate the excessive B-cell IgA production observed in patients with IgAN and HSP [51, 52]. Thus it may reduce the levels of nephritogenic IgA antibodies. Immunoglobulins have been shown to improve T-suppressor function [14, 56, 57], which is deficient in these diseases [51, 52]. Immunoglobulin therapy modulates monocyte function, reducing monokine production [14]. Interleukin-6 and tumor necrosis factor levels are elevated in IgAN and may affect the development of glomerular lesions [55, 58]. The decrease in C3 staining we observed might be related to the ability of intravenous immunoglobulins to prevent active C3 fragments from binding to target surfaces [59], whereas the decrease in IgA staining we observed might be related to glomerular solubilization of the deposits by intravenous immunoglobulins, as shown in vitro in lupus erythematosus and membranous nephritis [60]. Intravenous immunoglobulins contain anti-idiotype antibodies that may either bind to and modulate the activity of autoreactive B and T cells or bind to circulating pathogenic antibodies [14, 56, 61]. Anomalies of the idiotypic network were recently described in IgAN [62]. Finally, the partial IgG1 deficiency in these patients may also play a role, because physiologic peripheral IgG levels have been shown to modulate cellular activities in the mouse bone marrow B-cell lineage, acting mainly on small pre-B cells [63]. The decrease in serum IgG1 might favor the expansion of the medullary IgA B-cell compartment, which is increased in IgAN [6].

Immunoglobulin therapy was effective in severe cases of IgAN and HSP in this small, uncontrolled trial, but further studies are needed to determine the optimal immunoglobulin dosages. Prospective, placebo-controlled trials are also needed to confirm the efficacy and benefit of immunoglobulin therapy in patients with IgAN. A better understanding of the effects of immunoglobulin therapy will improve our knowledge of this common form of glomerulonephritis.