Tonsillectomy and steroid pulse (TSP) therapy for patients with IgA nephropathy: a nationwide survey of TSP therapy in Japan and an analysis of the predictive factors for resistance to TSP therapy

Tonsillectomy and steroid pulse (TSP) therapy was proposed as a curative treatment for IgA nephropathy by Hotta et al. (Am J Kidney Dis 38:736–742, 2001) based on data that about 50% of patients achieved clinical remission (CR) of urinary abnormalities.

As a primary survey, we sent a questionnaire and letter to 848 hospitals in Japan, each of which employed a Fellow of the Japanese Society of Nephrology between October and December of 2006, in order to gather information about the prevalence and efficacy of TSP therapy for patients with IgA nephropathy. As a secondary survey, we collected data from both low- and high-CR-rate groups to determine which factors predicted resistance to TSP therapy.

A total of 2,746 patients received TSP therapy between 2000 and 2006. The CR rates, calculated by measuring urinary criteria 6 and 12 months after TSP therapy, were 32.0% (347/1,085) and 45.6% (452/991), respectively. Analysis of the 30 hospitals in which TSP therapy had been performed on at least ten patients revealed that the CR rates varied from below 10% to 100%. A secondary survey of ten hospitals revealed that, after correction of the CR rate from each hospital, patients could be categorized into three groups: those with a low CR rate (122 patients in four hospitals), a middle CR rate (78 patients in four hospitals), and a high CR rate (103 patients in two hospitals). The CR rate of all patients (N = 303) was 54.1%. A comparison of patient data between the low- and high-CR-rate groups showed a significant difference in age at onset (years; P = 0.05), amount of proteinuria (g/day; P = 0.02), total protein (g/dl; P = 0.02), pathological grade (P = 0.009), and prognostic score as described by Wakai et al. [Nephrol Dial Transplant 21:2800–2808, 2006, (P = 0.04)]. Univariate analysis revealed that there was a significant difference between non-CR and CR subgroups in duration from diagnosis until TSP therapy (6.9 ± 6.8 versus 5.3 ± 5.2 years; P = 0.02), amount of proteinuria (1.5 ± 1.6 versus 0.8 ± 0.8 g/day; P < 0.0001), serum creatinine (0.99 ± 0.40 versus 0.87 ± 0.34 mg/dl; P = 0.006), pathological grade (P = 0.0006), and Wakai et al.’s prognostic score (37.4 ± 17.8 versus 28.1 ± 15.1; P < 0.0001). A multivariate logistic analysis demonstrated that resistance to TSP therapy depends on age at onset, amount of proteinuria, hematuria grade, and pathological grade, and a score predicting resistance to TSP therapy could be derived by the formula: [(−0.0330) × (age) + (0.4772) × log (amount of proteinuria) − (0.0273) × (hematuria grade: 0, 1, 2, and 3) + (0.7604) × (pathological grade: 1, 2, 3, and 4) − 0.1894]. A receiver operating characteristic (ROC) curve showed that patients with a resistance score of greater than −0.02 easily resist TSP therapy (sensitivity 69%, specificity 75%, positive likelihood ratio 2.76).

TSP therapy shows promise as a treatment that can bring about CR of urinary abnormalities, but unfortunately the average CR rate is about 50% at 1 year after treatment. Predictive factors for resistance to TSP therapy are age at onset, amount of proteinuria, hematuria grade, and pathological grade. The present study suggests that patients with either early-stage or mild to moderate IgA nephropathy easily achieve CR following TSP therapy, whereas patients with late-stage or severe disease are prone to TSP therapy resistance.