One can speculate that tonsillectomy represents an easy means to eliminate a pathogen source. However, several studies suggest that abnormalities of the mucosal-associated lymphoid tissue (MALT) are critical for the development of IgAN, with infections representing the simple role of triggering an event [1]. Experimental IgAN can be produced in animals after abrogation of the natural process of mucosal tolerance, which favors the host defense against pathogens [8]. In this context, IgAN is likely to develop because of a failure of mucosal antigen elimination and altered IgA synthesis, leading to the production of nephritogenic IgA. On the other hand, studies with experimental IgAN also demonstrated that chronic mucosal infection is not required for nephritogenic IgA production [9]. Moreover, transient mucosal activation of a pattern recognition receptor (PRR), such as Toll-like receptor (TLR), by pathogen-associated molecular patterns in IgAN-prone mice is sufficient to exacerbate this disease with rapid serum elevation of IgA [9, 10]; this suggests that preexisting mucosal B-cell clones produce nephritogenic IgA. Palatine tonsils have a unique cellular composition in the reticulated subepithelium, which is ideal for productive antigen sampling. One of the most important characteristics of the palatine tonsils is that very rich B-cell lymphoid follicles at the subepithelial space foster the development of memory B cells and plasma cells. This is very different from other tonsils in Waldeyer’s ring. Japanese nephrologists and otolaryngologists are aware that the beneficial effect of tonsillectomy in IgAN patients is independent of the size of the palatine tonsils and the presence of abscesses. Therefore, if the responsible B cells producing nephritogenic IgA are localized in the palatine tonsils, tonsillectomy may abrogate mucosal antigen encounters to such B cells, even if not chronic, leading to acute elevation of nephritogenic serum IgA [9, 10] and their clonal expansion [11].

It is now accepted that galactose-deficient IgA1 (GdIgA1) and GdIA1 and immune complexes (IC) with endogenous antiglycan antibodies are essential nephritogenic molecules initiating IgAN [12, 13]. Although there is no study clearly demonstrating the type of mucosal B cells involved, the recent studies revealed that the palatine tonsils are, at least in part, delivery sources of GdIgA1 under abnormal cytokine conditions [14–18]. Total IgA is decreased by 10 % on average after tonsillectomy alone in IgAN patients; because patients who showed tonsillar activation of innate immunity and a large decrease of serum IgA after the tonsillectomy had a better clinical outcome, it was thought that the palatine tonsils may be the major delivery source of nephritogenic IgA [19]. One recent study directly demonstrated that tonsillectomy alone rapidly decreased serum levels of GdIgA1 in patients who showed rapid improvement of hematuria after this therapy [20].

However, when considering tonsillectomy for reducing MALT surface, we should consider that tonsils represent only 0.5 m² of the entire 400 m² of the total mucosal surface in humans. In a recent study [21], IgAN patients who underwent tonsillectomy showed a long-term reduction, but not normalization, of GdIgA1, signs of persistently activated MALT, ongoing oxidative stress, and increased expression of TLR4. The ligand of TLR4 is lipopolysaccharide (LPS), produced by gram-negative intestinal bacteria. Notably, a correlation was found between progressive cases of IgAN and genetic polymorphism of the membrane receptor for LPS (CD14–159). A renewed interest toward intestinal immunity has been recently raised by a genetic-wide association study (GWAS), showing a strong association between IgAN and genes of intestinal MALT response [22]. These recent data, together with past reports on the role of dietary antigens in IgAN [23], tend to limit the extent of the rationale for tonsillectomy in patients with IgAN.

In contrast, other studies further supported tonsillectomy in nephropathy because serum levels of IgA and GdIgA1 were found to be correlated with tonsillar TLR9 overexpression [19, 20], and the TLR9 genotype was strongly associated with histological severity of IgAN [9]. Patients who did not show a decrease in GdIgA1 after tonsillectomy did so after the first additional steroid pulse therapy with improvement of hematuria, indicating GdIgA1 production outside the tonsils [20]. This finding suggests that some parts of the responsible cells producing GdIgA1 in such patients may be disseminated to MALT other than the tonsils and other lymphoid organs such as the bone marrow or spleen [24, 25]. It is known that tonsillectomy with steroid pulse therapy (TSP) leads to a better clinical outcome than tonsillectomy alone [26]. These clinical and experimental findings suggest that additional steroid pulse therapy on tonsillectomy may target these disseminated extratonsillar responsible cells [27].

In the absence of large series studies or randomized clinical trials (RCTs), nephrologists in Western countries tend to consider tonsillectomy in IgAN as a procedure to be performed only in patients with macroscopic hematuria and clinically evident recurrent tonsillitis. However, in eastern countries tonsillectomy continues to be regularly performed. In 2003, a breakthrough report by Xie et al. from Japan resurrected the interest for tonsillectomy in IgAN because for the first time, the procedure was associated with a better outcome at long-term follow-up [28]. A Japanese retrospective analysis of 118 patients reported a significant effect of tonsillectomy on survival from dialysis.

Here, the scientific conflict regarding the benefits of tonsillectomy between Western and Asian countries began. In Europe, retrospective analyses of single medical centers showed negative results in tonsillectomy [29, 30]. However, reports from Japan showed a benefit of tonsillectomy, particularly with steroid pulse therapy versus steroid therapy alone [31–33]. A benefit of tonsillectomy was also shown in IgAN recurrent in grafted kidneys while receiving a standard immunosuppressive regimen [34].

A recent RCT including 72 cases of IgAN in Japan was conducted. The patients received methylprednisolone pulse alone or in combination with tonsillectomy; some positive effects in reducing proteinuria at 1-year follow-up were observed [35]. However, this RCT did not provide data on the benefits on renal function decline and a longer follow-up is needed.

Due to these conflicting reports, the Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guidelines did not recommend the tonsillectomy as a therapeutic approach for IgAN [36].

Conflicting research findings highlight the differences between the European and American perceptions of tonsillectomy as a treatment for IgAN. The European-American practice, which limits tonsillectomy to cases with an obvious focal infection, is in clear disagreement with the common indication in Asia to perform tonsillectomy upon diagnosis of IgAN.

Why is there an underlying regional scientific conflict regarding the benefits of tonsillectomy? Although this conflict is based on the discrepancy between the abovementioned two European retrospective studies [29, 30] and some Asian studies [28, 31, 32], the clinical stages of the IgAN patients in these studies appear to be different. Rasche et al. included patients with relatively advanced stages: 55 % of the participants had hypertension, 35 % had elevated serum creatinine (>150 mmol/L), 62 % had heavy proteinuria (>1.5 g/day), and 25 % of the participants reached ESRD within 2.3 years after tonsillectomy [29]. Although Piccoli et al. selected only CKD stage 1 and 2 patients with tonsillectomy, they analyzed the efficacy of the tonsillectomy in both IgAN and non-IgAN patients [30].

Meanwhile, Xie et al. examined 118 Japanese patients in a moderate stage: 38.1 % had <0.5 g/day proteinuria and the mean serum creatinine was 1.07 mg/dl; in addition, a better outcome for tonsillectomy during the long-term follow-up was observed [28]. A recent study by Maeda et al. included a relatively early stage of IgAN patients; they demonstrated the efficacy of tonsillectomy adjusted for the known risk factors including blood pressure, proteinuria, and histological findings [31]. Therefore, perceived discrepancies in efficacy between Europe and Asia may partly be because of the clinical stage of IgAN at the time of intervention [5].

In Europe and the USA, the guidelines for tonsillectomy provided by Ears, Nose, and Throat (ENT) International Societies do not recommend tonsillectomy in patients who do not present with recurrent tonsillitis, including fever, pharyngeal or tonsillar erythema, enlarged tonsils, tonsillar exudate, cervical adenopathy, and supportive microbiological test results. Moreover, watchful waiting is recommended for patients with recurrent throat infections if they have had fewer than seven episodes in the previous year. At least 12 months of observation is recommended before considering tonsillectomy. No specific benefit is supported by nephrology guidelines; therefore, tonsillectomy in a patient with IgAN without clinical signs could be questionable in case of complications, even malpractice.

Following Hotta’s 2001 report [32], tonsillectomy in Japan is frequently performed in combination with steroid pulse therapy. According to a nationwide survey in more than 350 Japanese teaching hospitals, TSP is now performed in approximately 65 % of the center hospitals [37], with remission of hematuria or proteinuria in more than 50 % patients. These uncontrolled clinical data represent the basis for TSP as the standard therapy for IgAN in Japan.

The majority of Japanese IgAN patients are less than 40 years old. 70 % of the patients were diagnosed at early stage of this disease [38] through chance hematuria in screening urinalysis. Although the KDIGO guidelines recommend renin-angiotensin system blockade (RASB) agents as the first choice of treatment for IgAN patients [39], Japanese medical guidelines do not permit the use of RASB in IgAN patients without hypertension. These guidelines in addition to teratogenic side effects further nourish a hesitation for RASB usage in normotensive young females with IgAN in common Japanese medical practice.

The difficultly in prescribing RASB along with well-demonstrated benefits of TSP are factors that favor tonsillectomy in early phases of IgAN in Japan, making it difficult to justify multicenter RCTs. However, a large, nationwide prospective cohort study in Japan began in 2008 and will provide some much needed clarity to this matter.

Tonsillectomy as monotherapy in IgAN has only a small possibility to be proven beneficial by RCT. It would not be feasible for opposite reasons in the two parts of the world.

In Europe and Asia, no ethical committee would permit a potentially invasive surgical intervention, which is not recommended by ENT or nephrology guidelines. No ENT specialist could be convinced to participate without a strong, evidence-based recommendation.

On the opposite side, in Japan, social medicine does not support RASB in normotensive young IgAN patients. Medical practitioners and patients generally believe that tonsillectomy and TSP are safe and beneficial. Tonsillectomy/TSP is the standard of practice; therefore, it would be not easy to conduct an RCT.

The abovementioned differences in clinical practice of tonsillectomy in IgAN provoke a still-passionate discussion at international meetings about this procedure in patients with IgAN. The interest remains strong and we do not abandon the hope to find a scientific solution.