High-dose intravenous immunoglobulin pulse therapy in patients with progressive immunoglobulin A nephropathy: a long-term follow-up

Abstract

In progressive immunoglobulin A nephropathy (IgAN), intravenous immunoglobulin (IVIg) treatment has been used to delay disease progression, but the long-term efficacy is largely unknown. We report the clinical outcomes after IVIg therapy in six male patients with progressive IgAN [median glomerular filtration rate (GFR) 31 ml/min per 1·73 m²] followed for a median observation period of 8 years. In this single-arm, non-randomized study, IVIg was given monthly at a dose of 2 g/kg body weight for 6 months. The course of renal function was assessed by linear regression analysis of GFR and proteinuria, and was compared to eight patients with IgAN (median GFR 29 ml/min per 1·73 m²) without IVIg as a contemporaneous control group. IgAN disease progression was delayed after IVIg therapy on average for 3 years. The mean loss of renal function decreased from − 1·05 ml/min per month to − 0·15 ml/min per month (P = 0·024) and proteinuria decreased from 2·4 g/l to 1·0 g/l (P = 0·015). The primary end-point (GFR < 10 ml/min or relapse) occurred 5·2 years (median; range 0·4–8·8) after the first IVIg pulse, and after 1·3 years (median; range 0·8–2·4) in the control group (P = 0·043). In Kaplan–Meier analysis, the median renal survival time with IVIg was prolonged by 3·5 years (IVIg 4·7 years versus control 1·2 years; P = 0·006). IVIg pulse therapy may be considered as a treatment option to reduce the loss of renal function and improve proteinuria in patients with progressive IgAN.

Introduction

Immunoglobulin A nephropathy (IgAN) is the most common form of primary glomerulonephritis [1,2]. At least 20% of affected patients have a progressive course with complete loss of renal function [3,4]. The pathogenesis of IgAN remains unclear, and specific antigens or autoantibodies have not been detected. Polyclonal B cell stimulation and T cell proliferation, aberrant glycosylation of IgA molecules [5] and IgA receptors on myeloid, epithelial or mesangial cell surfaces, e.g. Fc alpha/mµ receptors [6], may play a part. This results in glomerular deposition or mesangial cell interactions with immune complexes containing predominantly IgA1/IgG [7].

Immunosuppressive therapy with steroids and/or cyclophosphamide prolongs renal survival time in patients with progressive IgAN [8–11], but treatment regimes with lower toxicity appear desirable in young patients, in patients with diabetes mellitus or females with childbearing potential. Polyclonal intravenous high-dose immunoglobulin (IVIg) preparations consist of normal IgG from plasma pools from healthy donors, and have been investigated for immunomodulation in various autoimmune diseases [12]. One study reported nine patients with IgAN and moderate renal impairment who had a decrease in proteinuria and an improvement in renal function and histological activity index after 6 months of IVIg and subsequently intramuscular injection of immunoglobulins [13]. The long-term effects of IVIg in patients with progressive IgAN have not been described so far.

We report here a long-term follow-up of six consecutive IgAN patients with a progressive course treated with IVIg pulse therapy.

Subjects and methods

Patients

Between February 1994 and February 1995, we recruited six patients with a progressive course of IgAN for this study (Table 1). Inclusion criteria were primary IgAN diagnosed by renal biopsy, serum creatinine higher than 180 µmol/l [glomerular filtration rate (GFR) > 25 ml/min per 1·73 m²] and/or a loss of GFR of more than 25% in the previous 3 months and normal kidney morphology and size (> 9 cm by ultrasound examination). Renal biopsies were performed on average 1·7 years (range, 0·1–6·5) before IVIg started. Renal biopsies were examined by one reference histopathologist, and were classified according to Haas' classification system [14] and grading system according to Lee [15]. The six patients were characterized by Lee's grading III–IV or Haas' classification III–IV.

Table 1.

Baseline characteristics of six patients treated with intravenous high dose immunoglobulins (IVIg) and eight untreated patients with progressive IgAN.

Patient	Age (years)/ gender*	GFR (ml/min per 1·73 m2)*	Loss of renal function (ml/min per month)*	Proteinuria (g/l)*	Serum protein (g/l)*	Blood pressure (mmHg)*	ACE inhibitor (enalapril)¶
IVIg
1	46 m	20	−1·48	3·8	72	140/80	None†
2	50 m	33	−1·19	2·4	72	160/100	10 mg
3	46 m	33	−0·72	2·4	72	160/100	10 mg
4	42 m	17	−0·90	3·2	65	120/80	10 mg
5§	35 m	36	−2·03	1·9	65	140/90	5 mg
6‡	55 m	29	−0·67	1·1	67	140/90	None†
Median	46	31	−1·05	2·4	70	140/90
Control
7	46 m	36	−0·55	2·0	68	150/80	10 mg
8	75 m	19	−1·01	2·4	68	160/80	None†
9	21 m	42	−2·26	2·5	57	130/90	None
10	44 m	40	−2·21	2·4	57	140/100	None
11	54 m	30	−1·18	1·25	66	160/100	10 mg
12	48 m	27	−1·92	2·2	72	180/115	None
13	34 f	25	−0·83	2·4	68	120/100	5 mg
14	53 f	29	−0·77	2·2	63	140/80	10 mg
Median	47	29	−1·09	2·3	67	145/95

ACE: angiotensin converting enzyme. Glomerular filtration rate (GFR) was estimated by the modification of diet in renal disease (MDRD)-2 formula.

^§Patient 5 received cyclophosphamide orally (month 58) after IVIg.

^‡Patient 6 received intravenous cyclophosphamide pulses (CyP) (month 63) after IVIg.

Follow-up period including subsequent cyclophosphamide therapy after IVIg-1 pulse.

^†Chronic cough or hyperkalaemia.

^*Age, GFR, loss of renal function, proteinuria, serum protein and blood pressure were not significantly different between IVIg and the control group in the Mann–Whitney U-test (P > 0·05).

^¶Distribution of the patients with ACE-I between the IVIg and the control group was not significant using the χ² test (P > 0·05); m: male; f: female.

As a contemporaneous control group, eight patients with progressive IgAN and identical inclusion criteria were studied (Table 1). Following their individual preferences, these eight patients did not receive treatment with IVIg or other immunomodulatory regimes.

Exclusion criteria were age < 18 years, serum creatinine exceeding 500 µmol/l (GFR < 15 ml/min per 1·73 m²), kidney size < 9 cm on ultrasound, extracapillary proliferative and rapidly progressive forms of IgAN, secondary mesangioproliferative IgA nephropathy due to systemic diseases, Schoenlein–Henoch purpura, acute or chronic infections (including human immunodeficiency virus, hepatitis B and C virus) and carcinoma.

Treatment protocol

Six intravenous high-dose human polyvalent immunoglobulin pulses (IVIg) were given monthly in a dose of 2 g/kg body weight (Venimmun™; Aventis Behring, Frankfurt, Germany) containing 50 mg/ml polyvalent immunoglobulins, 25 mg/ml glycine and maximum fraction of IgA 6·1 mg/ml. Only one patient received sucrose-containing IVIg once (Sandoglobin™; Novartis, Nürnberg, Germany, containing 60 mg/ml polyvalent immunoglobulins and 100 mg/ml sucrose, maximum fraction of IgA 2·4 mg/ml). No maintenance therapy with oral prednisolone or intramuscular immunoglobulins was prescribed. Treatment and side effects were recorded every second week. Primarily, angiotensin converting enzyme (ACE) inhibitors (ACE-I) were given to treat hypertension at the first visit before IVIg therapy started (Table 1), and further anti-hypertensive drugs were added as required. Dietary protein restriction (< 0·8 g per body weight/day) or salt restriction (< 6 g/day) was not prescribed. The IVIg treatment protocol was approved by the institutional review board and the ethics committee of the University of Ulm, and informed consent was obtained from all patients.

Outcome parameters and statistical analysis

GFR were estimated using the modification of diet in renal disease formula (MDRD 2) [16]. The course of renal function was analysed with linear regression analysis of GFR levels before (months − 3 to − 0), during IVIg (months 0–6) and after therapy (months 6–12, 12–24, 24 or 36). The beginning of the observation period in the control group was set at a GFR of 29 ml/min per 1·73 m² (MDRD 2), which was comparable with the median GFR of the IVIg group. The primary end-point of the observation period was a serum creatinine > 620 µmol/l (GFR MDRD 2 < 10 ml/min per 1·73 m²) or further relapse (decrease of GFR > 15%) associated with the necessity for other immunomodulatory therapy, e.g. cyclophosphamide. Proteinuria (g/l) was assessed in the spontaneous morning urine in our study, because 24-h urine collections or protein/creatinine urine ratio were not available in all patients. Differences between the slopes of linear regressions analysis, proteinuria (month 0 versus 6, 12, 24 or 36), serum protein (month 0 versus 6, 12, 24 or 36) and systolic and diastolic blood pressure (month 0 versus 6, 12, 24 or 36) were tested with nonparametric tests (Wilcoxon test, Friedman test). Differences between the IVIg group and the control group were tested with Mann–Whitney U-test. The cumulative probability of renal survival for IgAN patients with or without IVIg was calculated as per Kaplan–Meier analysis. Statistical significance for all tests was set at a level of P < 0·05. Statistical analysis was performed using the spss version 8·0 software package (SPSS Inc., Chicago, IL, USA). If not indicated otherwise, data are given as median with minimum and maximum, or mean ± standard deviation (s.d.).

Results

The median observation time from the first visit of the six patients with IVIg was 8·0 years (range 3·0–10·0 years) and 2·4 years (range 0·8–6·0 years) for the eight control patients. The baseline characteristics of all patients are summarized in Table 1, showing no differences in pretreatment GFR, loss of renal function (ml/min per month), proteinuria and blood pressure between the IVIg group and the selected contemporaneous control. The individual course of renal function (GFR MDRD 2) in each patient is shown in Fig. 1a (IVIg patients) and Fig. 1b (control patients).

Fig. 1.

Individual courses of renal function [glomerular filtration rate calculated by modification of diet in renal disease formula (GFR MDRD 2)] of six patients with progressive immunoglobulin A nephropathy (IgAN) (a) treated with high-dose human immunoglobulin pulses (IVIg) and eight control patients with progressive IgAN (b). Cumulative renal survival in Kaplan–Meier analysis (c) of the IVIg patients and the control patients. IVIg patients had a median of 3·5 years significantly longer renal survival time compared to the control group (log rank test, P = 0·0062). Loss of renal function in linear regression analysis (d) of the IVIg and the control group. Baseline loss of renal function was not significantly different between the groups (P = 0·8), but IVIg therapy significantly decreased the median loss of renal function compared to the control group (P = 0·02, Mann–Whitney U-test). R: regression coefficient before/after IVIg in linear regression analysis.

The primary end-point (serum creatinine > 620 µmol/l or GFR MDRD 2 < 10 ml/min per 1·73 m², or beginning of cyclophosphamide therapy) occurred in IVIg patients median after 5·2 years (range 0·4–8·8) after the first IVIg pulse. The control group showed a significantly shorter period to the primary end-point with 1·3 years (range, 0·8–2·4; Mann–Whitney U-test, P = 0·043). In Kaplan–Meier analysis, the median renal survival time from the beginning to the primary end-point with IVIg was significantly prolonged by 3·5 years (IVIg 4·7 years versus control 1·2 years; log rank test, P = 0·0062, Fig. 1c).

Course of renal function in linear regression analysis

The loss of renal function in linear regression analysis in each patient and in the median of the IVIg group before and after therapy and the control group (median) is shown in Fig. 1d. Baseline loss of renal function was not significant different between the groups (P = 0·8), but IVIg therapy decreased significantly the loss of renal function compared to the control group (P = 0·02; Mann–Whitney U-test, Fig. 1d). The median loss of renal function in the IVIg patients decreased significantly from − 1·05 ml/min per month (range − 2·03 to − 0·67; regression coefficient R = 0·86) before IVIg to − 0·15 ml/min per month 36 months after the first IVIg pulse (range – 0·85–0·11; regression coefficient R = 0·83; Friedman test, P = 0·026; Fig. 2a), and an improvement of renal function was noted in four patients, median 2 months after the first IVIg pulse (range 1–2·4 months). After 36 months the loss of renal function increased to − 0·48 ml/min per month; however, it was significantly lower before therapy (range − 0·66 to − 0·28; regression coefficient R = 0·89; P = 0·043 Wilcoxon test, P = 0·028 Friedman test; Fig. 2a). Differences in the slopes of linear regression analysis before IVIg and in the period 0–6 months (P = 0·4), 6–12 months (P = 0·043), 12–24 months (P = 0·043), 24–36 months (P = 0·043) and 36 months (P = 0·043) to the end of the study were tested for significance (Wilcoxon test). In the control group the loss of renal function was unchanged during the observation period (median − 1·09, range − 2·26 to − 0·55 ml/min per month, regression coefficient R = 0·97, Fig. 1d).

Fig. 2.

Loss of renal function per month of the IVIg patients in linear regression analysis (a) and proteinuria (b), before and with IVIg pulse therapy. Box plots show the median, interquartile range and outliers. R: regression coefficient in linear regression analysis. Significances were calculated by the Friedman test # and Wilcoxon test* (number of patients: n = 5).

Proteinuria, blood pressure and ACE inhibitors

Median proteinuria decreased significantly from median 2·4 g/l (range, 1·1–3·8) before IVIg to 1·0 g/l (range, 0·4–1·9) at 36 months (P = 0·015, Friedman test), and to 1·3 g/l (range, 0·4–1·5) at the end of the study (P = 0·023, Friedman test; Fig. 2b). The first decrease of proteinuria was noted 2 months (median, range, 1–3·7) after the first IVIg pulse. There was a trend in the decrease of proteinuria 6 months after IVIg pulse 1 (median 1·6 g/l, range 1·0–2·8; P = 0·11; Wilcoxon test) and 12 months after IVIg pulse 1 (median 1·9 g/l, range 0·6–2·7; P = 0·068; Wilcoxon test); however, proteinuria decreased significantly 18 months after IVIg (median 1·0 g/l, range 0·5–1·44; P = 0·043; Wilcoxon test) after IVIg, which lasted to the end of the study (median 1·3 g/l, range 0·4–1·5 g/l; P = 0·043; Wilcoxon test). Serum protein was stable during the observation period at 70 g/l (range, 65–72) to 72 g/l (range 65–72; P > 0·05 Friedman and Wilcoxon tests). No significant differences were observed on proteinuria between the IVIg and the control group at beginning (Table 1) and after 1 year (control group: median 2·0 g/l range 1·3–2·2 g/l; P = 0·6, Mann–Whitney U-test).

Blood pressure was not different between the IVIg and the control group (P > 0·05; Mann–Whitney U-test, Table 1) at baseline, whereas at the end of the study a significantly lower systolic blood pressure was observed in the IVIg group (median, IVIG 120/80 mmHg versus control 148/85 mmHg, systolic pressure P = 0·043, diastolic pressure P > 0·05; Mann–Whitney U-test). In the IVIg patients the blood pressure decreased significantly during the study period (median before IVIg 140/90 mmHg versus 120/80 mmHg at the end of the study, systolic pressure P = 0·027, diastolic pressure P = 0·041, Wilcoxon test), whereas in the control group the systolic blood pressure increased at the end of the study (median, 145/85 mmHg versus 148/85 mmHg at the end of the study, systolic pressure P = 0·012, diastolic pressure P > 0·05, Wilcoxon test).

ACE-I were used in all patients as first-line therapy for arterial hypertension, and treatment was initiated with a median of 2 months before IVIg. Two patients of the IVIg and five of the control group did not tolerate ACE-I due to chronic cough or severe hyperkalaemia (Table 1) and were switched to other anti-hypertensive medications. Angiotensin receptor blockers (ARBs), now considered a therapeutic alternative for ACE-I, were not available at the time our study was initiated. There was no difference of the frequencies of ACE-I between the control and the IVIg group (P = 0·6, χ² test, Table 1).

Side effects of IVIg therapy

One patient (patient 6) with severely impaired renal function at the start of therapy (serum creatinine of 375 µmol/l) developed acute renal failure immediately after the first sucrose-containing IVIg-pulse, due presumably to osmotic nephrosis [17,18], and the GFR decreased from 16·5 ml/min per 1·73 m²−6·2 ml/min per 1·73 m² 5 days after administration. After three sessions of haemodialysis renal function recovered (GFR 11·7 ml/min per 1·73 m²), and he was switched to a sucrose-free formulation of IVIg. In all other patients we used the same preparation of IVIg with a sucrose-free formulation (Venimmun™). Another patient (patient 1) suffered from deep vein thrombosis of the right leg after the second IVIg pulse, but no embolic complication occurred. No anaphylactic reactions, myalgia or other side effects were observed.

Extended clinical follow-up

Due to a relapse, defined as a decrease of GFR > 15% after IVIg therapy, two of the IVIg patients received further immunosuppressive therapy with cyclophosphamide. Patient 5 received oral cyclophosphamide 57 months after IVIg-1, starting with cyclophosphamide 100 mg orally (Endoxan™, Baxter Oncology, Halle, Westfahlen, Germany) adjusted to the leucocyte count. Before oral cyclophosphamide the loss of renal function was −0·28 ml/min per 1.73 m² per month (regression coefficient R = 0·39, Fig. 1a), and increased further with oral cyclophosphamide to −0·56 ml/min per 1.73 m² per month (R = 0·57). Patient 6 received a cyclophosphamide pulse therapy (750 mg/m² body surface area intravenously monthly for 6 months, CyP 1–6) 63 months after IVIg-1. The loss of renal function improved with CyP from −0·66 ml/min per 1.73 m² per month (R = 0·99, Fig. 1a) to −0·08 ml/min per 1.73 m² per month (R = 0·38). Patient 5 developed end-stage renal disease 67 months after IVIg-1, and patient 6 developed end-stage renal disease 98 months after IVIg-1.

Discussion

Immunosuppressive intervention in patients with IgAN is still controversial [9,19,20]. ACE-I and/or ARBs are able to retard the loss of renal function in IgAN patients and are now recommended and established as the first-line therapy in IgAN [21]. In the last 5 years, immunosuppression with cyclophosphamide [22] or high-dose steroid pulses [8,23] has shown some evidence to reduce the loss of renal function in randomised, controlled studies in IgAN patients with mild and moderate renal impairment. In addition, cyclophosphamide pulse therapy has been shown to have beneficial effects on renal function in IgAN patients with a more advanced stage of renal disease and progressive loss of renal function [10].

In 1995 we started IVIg therapy in our patients with advanced progressive IgAN, long before the results of controlled studies with steroids [23] or cyclophosphamide [22] were available. As a long-term follow-up we demonstrate here that IVIg pulse therapy slowed the progressive loss of renal function for 3 years in five of six patients with progressive IgAN. We observed a reduction in the loss of renal function and proteinuria in our patients beginning 2 months after IVIg pulse 1 and with a maximum effect 2 years after IVIg pulse 1. It is probable that the decrease of proteinuria was not related to interference with a previous ACE-I therapy. The effects of IVIg therapy were not permanent but lasted for an average of 3 years. It is currently unknown whether subsequent immunosuppression using azathioprine or mycophenolate is suitable to further arrest disease progression after IVIg pulse therapy.

In previous reports, beneficial effects after IVIg therapy on proteinuria, haematuria and the loss of renal function have been demonstrated in nine patients with IgAN and renal impairment (GFR > 35 ml/min per 1.73 m²), who were treated with IVIg pulse therapy (1 g/kg of body weight per day on 2 successive days each month for 3 successive months, followed by intramuscular immunoglobulins in a dose of 0·35 ml/kg of body weight twice each month over 6 months) [13]. In that study, the median follow-up was 14 months, and the longest follow-up was 19 months. In our study the median follow-up time was 5·2 years after the first IVIg pulse (range, 0·4–8·8 years). Beneficial effects of IVIg were demonstrated without prescribed ACE I/ARBs in that study [13].

Several mechanisms of action of IVIg therapy have been proposed, including the reduction in nephritogenic IgA antibodies by inhibiting B cell differentiation and the activation and effector functions of B and T lymphocytes, e.g. antigen presentation, blockade of the Fc receptors and expression of cytokines, inhibition of the complement system and neutralization of pathogenic autoantibodies, bacterial toxins and superantigens, inhibition of apoptosis and direct administration of soluble factors with the capacity to modulate immune response [12]. In autoimmune diseases such as IgAN, infusion of exogenous immunoglobulins may enhance intracellular degradation of endogenous and potentially pathogenic immunoglobulins through competitive interaction with the intracellular Fc receptor responsible for recycling [24]. Additionally, in patients with progressive IgAN there was a significant reduction in elevated proinflammatory cytokines such as total tumour necrosis factor (TNF) and interleukin (IL)-6, and levels of soluble TNF receptors were increased on immunoglobulin therapy [25].

The therapeutic use of ACE-I and blood pressure control in our patients were performed according to the accepted standards of clinical care at the time our study was designed, including the restricted use of ACE-I in individuals with advanced renal failure (GFR 30 ml/min per 1.73 m² or less) as in our patients. Since then, several clinical trials [21,26] have demonstrated the efficacy of the ACE-I on the progression of IgAN, so that ACE-I is now accepted as standard therapy in mild to severe IgAN, even in the absence of arterial hypertension [11]. In addition, with the advent of ARBs in the last decade, another potent therapeutic option is now available in particular for patients who do not tolerate ACE-I. In our study, there was no difference in the use of ACE-I between the IVIg pateints and the control patients, but the interval at the start between ACE-I and IVIg therapy was short in our patients. There was a small but significant reduction of systolic blood pressure at the end of the study in the patients with IVIg. It is possible that this observation may be attributed to differential patient adherence to ACE-I medication, or the less progressive course of renal deterioration in IVIg patients compared to the control group.

The safety of high-dose intravenous immunoglobulin therapy seems to be related to the formulation of the IVIg preparation and individual characteristics of the patient, e.g. degree of renal impairment and hydration. Acute renal failure after sucrose-containing IVIg was observed in one patient with low baseline GFR (17 ml/min per 1·73 m²) [17,18]. This severe side effect is most probably attributable to osmotic nephrosis and may be prevented by the use of sucrose-free formulations [17,18]. Accordingly, sucrose-free IVIg preparations were well tolerated in all subsequent patients.

Our observations are limited by the small numbers of treated patients and by a single-arm, non-randomized study design. A larger randomized controlled trial of high-dose IVIg in progressive IgAN would be ideal to confirm our observations, but seems impractical at present given the relative paucity of affected patients, the considerable cost of IVIg treatment and the advent of alternative therapies in recent years [8,9,11,22,27,28].

In conclusion, our data indicate that IVIg therapy may be a therapeutic option in patients with progressive IgAN to delay the loss of renal function, particularly in individuals with contraindications to cyclophosphamide or high-dose steroid pulse therapy, who did not respond to standard therapy and aggressive blood pressure control.