Kidney International, Vol. 65 (2004), pp.

1135–1144

PERSPECTIVES IN RENAL MEDICINE

Relationship between tonsils and IgA nephropathy as well as

indications of tonsillectomy
YUANSHENG XIE, XIANGMEI CHEN, SHINICHI NISHI, ICHIEI NARITA, and FUMITAKE GEJYO
Kidney Center of PLA, Department of Nephrology, Chinese General Hospital of PLA, Beijing, China; and Division of Clinical
Nephrology and Rheumatology, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan

Relationship between tonsils and IgA nephropathy as well as clear that a large number of cases eventually progress to
indications of tonsillectomy. Although there are many papers renal failure [7–11]. Indeed, IgAN is the main cause of
about IgA nephropathy (IgAN) and tonsils, respectively, re- end-stage renal disease (ESRD) in patients with primary
views about the relationship between tonsils, tonsillitis, tonsil-
lectomy, and IgAN are limited. In this review, we introduced glomerular disease who require renal replacement ther-
the structure, development, and function of tonsils, difference apy [12, 13]. However, the cause of primary IgAN, source
of tonsils with and without IgAN, consistency of both tonsil- of IgA deposited in glomeruli and the mechanism under-
lar IgA and glomerular IgA, the effect of tonsil stimulation, lying mesangial IgA deposition in IgAN, is unclear and
tonsil infection, and tonsillectomy on IgAN showed some evi- there is no effective treatment available for patients with
dences in which tonsils were closely related to IgAN and poly-
meric IgA1 deposited in glomerular mesangium were at least IgAN [14].
in part of tonsillar origin. Tonsillectomy can improve the uri- The IgA deposited in glomerular mesangium in pa-
nary findings, keep stable renal function, improve mesangial tients with IgAN appears to be exclusively of the IgA1
proliferation and IgA deposit, have a favorable effect on long- subclass [15] and IgA produced by tonsillar lymphcytes in
tern renal survival in some IgAN patients, and do not cause patients with IgAN is mainly polymeric IgA1, about half
significant immune deficiency and do not increase incidence of
the upper respiratory tract infections, and can be used as a po- of patients with IgAN their serum IgA levels increase
tentially effective treatment. The indications of tonsillectomy [16] and tonsillectomy decreases the levels of serum IgA,
in patients with IgAN include mainly the deterioration of uri- suggesting there is any relationship between tonsils and
nary findings after tonsillar infection, mild or moderate renal IgAN. Recently, we demonstrated that the tonsillectomy
damage. However, tonsillectomy may not be enough and may has a favorable effect on long-term renal survival in pa-
not change the prognosis in IgAN patients with marked renal
damage. tients with IgAN [17].
Although there are many papers about IgAN and ton-
sils, respectively, reviews about the relationship between
Immunoglobulin A nephropathy (IgAN), that is, tonsils, tonsillectomy, and IgAN are limited. In this re-
nephropathy with mesangial IgA-IgG deposits, was first view, we introduce the structure, development, and func-
reported by Berger and Hinglais in Frence in 1968 [1] tion of tonsils, difference of tonsils with and without
and described by Berger in English in 1969 [2]. Studies IgAN, consistency of both tonsillar IgA and glomerular
for more than 30 years demonstrated that primary IgAN IgA, the effect of tonsil stimulation, tonsil infection, and
is an immune complex–mediated glomerulonephritis de- tonsillectomy on IgAN, show some evidences in which
fined immunohistologically by the presence of glomeru- tonsils were closely related to IgAN and polymeric IgA1
lar IgA deposits [3]. It is now generally known to be deposited in glomerular mesangium were at least in part
the most common form of primary glomerulonephritis of tonsillar origin, and present the indications of tonsil-
throughout the world [4–6]. Although primary IgAN was lectomy in patients with IgAN.
considered a benign condition for many years, it is now
STRUCTURE, DEVELOPMENT, AND
FUNCTIONS OF TONSILS
Key words: tonsils, tonsillectomy, IgA nephropathy, treatment, indica-
tion.
Structure of tonsils
Human tonsils include the palatine tonsils, nasopha-
Received for publication August 3, 2003
and in revised form September 28, 2003
ryngeal tonsil (adenoid), lingual tonsil and the tubal ton-
Accepted for publication October 28, 2003 sils [18] (Fig. 1). The palatine tonsils are the largest ones in
four types of tonsils in human beings. Histologically, tonsil


C 2004 by the International Society of Nephrology tissues consist of four well-defined microcompartments,

1135
1136 Xie et al: Tonsils and IgA nephropathy

Pharyngeal tonsil decreased in acute tonsillitis [23]. The study regarding

(adenoid) the distribution and proportion of Ig subclasses produc-
Uvulae ing cells in chronic tonsillitis show that the percentage
Tubal tonsils ratios of IgG1:IgG2:IgG3:IgG4 were 53.1:35.9:4.7:6.3, re-
Palatine tonsils spectively. Proportional ratios of IgA1:IgA2 are approx-
Oral cavity
imately 80:20 [24].
Lingual tonsil
Tongue
Development of tonsils
The development of the palatine tonsils starts during
Fig. 1. Anatomy of the tonsils. the 14th gestational week when the mesenchyme underly-
ing the mucous membrane of the tonsillar cavity becomes
invaded by mononuclear wandering cells. In fetuses of
about the 16th gestational week epithelial crypts grow
Interfollicular area down into the connective tissue and are infiltrated by T
lymphocytes. At the same time, precursors of interdigi-
Mantle zone of the follicle tating cells can be identified among the epithelial cells.
Primary follicles develop in earlier fetal stages than in
Follicular germinal center all other secondary lymphoid organs. They contain pre-
cursors of dendritic reticulum cells and lymphoid cells
that belong to the B-cell line. These primary follicles may
be considered as the first assemblage of B-cell regions in
Reticular crypt epithelium human fetal lymphoid tissue [25]. The formation of the
follicular germinal centers reflecting B-cell activation by
exogenous antigens takes place shortly after birth [26].
Fig. 2. Histologic structure of tonsils. The sample originated from ton-
sil tissues after tonsillectomy because of chronic tonsillitis in 8-year-old, The immunohistochemical study show the morphomet-
male patient. The cellular nuclei of the section were stained with hema- ric features of tonsils below the age of 8 years are more
toxylin (original magnification ×50). active than those above the age of 8 years. Total number
of IgA immunocytes is the highest at the age of 5 to 7
years with a decline by age. The serum IgA and salivary
which all participate in the immune response: the retic-
secretory IgA concentrations reach to adult’s level at the
ular crypt epithelium, the interfollicular (extrafollicular)
age of 11 to 13 years. These results suggest that tonsils in
area, the mantle zone of lymphoid follicles, and the fol-
preschool children are important as a local immunologi-
licular germinal center [19] (Fig. 2). Cell biologically, im-
cal defense mechanism [27].
munocyts of tonsil tissues contain predominantly B cells
(approximately 65%), approximately 30% CD3+ T cells,
and 5% macrophages. The T cells were primarily of the Functions of tonsils
CD4+ subset (approximately 80%) [20]. Quantitative im- Tonsil tissues are located at the gateway of the respi-
munohistochemistry reveals that IgG-containing B cells ratory and alimentary tract and belong to the mucosa-
predominate in all lymphoid compartments, including associated lymphoid tissue. The generation of B cells in
follicles, extrafollicular areas, and reticular epithelium, the germinal centers of the tonsil is one of the most es-
whereas IgA cells are found predominantly in extrafol- sential tonsillar functions. The major function of tonsils
licular areas, especially subepithelial area, and IgM cells is as a first line of defense against viral, bacterial, and
are in follicles. J chain is present within IgM and some IgA food antigens that enter the upper aerodigestive system.
cells. The IgG:IgA:IgM class ratios of the overall tonsillar Secretory dimeric IgA produced by B cells has particu-
immunocyte population are 13:8:2. Cells containing IgD lar hydrophilic properties and is capable of preventing
and IgE are rare [21]. In clinically normal tonsils, the over- adsorption and penetration of bacteria and/or viruses
all percentage distribution of these cells is 65:30:3.5:1.2 into the upper respiratory tract mucosa [28]. With the
for the IgG, IgA, IgM. and IgD classes, respectively. uptake of antigen by microfold cells (membrane cell, M
In recurrent tonsillitis, these figures are 53:39:4.7:4.4; in cells) present in the cryptepithelium a process is initiated,
hyperplastic tonsillitis, 67:25:4.0:4.5; and in idiopathic which ultimately results in the generation and dissem-
tonsillar hyperplasia, 50:33:7.2:10, respectively [22]. In ination of antigen-specific memory and mainly dimeric
comparison with the clinically healthy tonsils, the num- IgA-producing effector B lymphocytes. This process re-
ber, the size of the germinal centers and the density of the quires successful cognate interactions between antigen-
immunocytes in tonsils are very large in the hyperplastic presenting cells and lymphocytes and mutually between
tonsils, large in chronic cryptic tonsillitis, but remarkably lymphocytes, which depend not only on antigen-specific
 Xie et al: Tonsils and IgA nephropathy 1137

Table 1. Difference of tonsils with (+) and without (−) IgA larged primary T nodules reminds us that extrafollicular
nephropathy (IgAN)
maturation of the stimulated B lymphocytes into plasma
Reference cells may occur more frequently in the tonsils of patients
Characteristic of tonsils IgAN(+) IgAN(−) number with IgAN than in patients with habitual tonsillitis. An-
T cell area (T nodules) Expanded Not expanded 29 other study demonstrated abnormal reticulization of ton-
Reticulization of crypt Reduced Not reduced 31 sillar crypt epithelium in patients with IgAN. Tonsils of
epithelium
IgA cells:IgG cells >1 <1 32–34 controls with recurrent tonsillitis or tonsillar hypertro-
Polymeric IgA cells Increased Not changed 32–34 phy showed well-developed reticular crypt epithelia with
Polymeric IgA:IgA Increased Not changed 35 lymphoepithelial symbiosis, and the nonreticulated area
Follicular dendritic cells IgA1+ IgA1− 36 was less than 7% of the total crypt epithelia per overall
J chain mRAN-positive cells Increased Not changed 37 section. In IgAN tonsils, however, nonreticulated crypt
Adhesion molecules CD31, Increased Not changed 38
CD54 epithelium was frequently observed and, in the advanced
CD5+ B cells Increased Not changed 39 stage of IgAN, exceeded 50% of total crypt epithelia [31].

Difference of tonsillar cells

signals, but also on the expression of various complemen-
Primary IgAN is characterized by renal deposits of
tary adhesion and costimulatory molecules [19]. In addi-
polymeric IgA (J chain-positive), the origin of which is
tional, Mitogen-triggered T cells from tonsils produced
not confirmed, yet. The study by Bene et al [32] showed
both of Th1- and Th2-type cytokines, clearly exhibit-
that in controls with recurrent tonsillitis, IgG secreting
ing their pluripotentiality for support of cell-mediated
cells were predominant (IgG secreting cells 65% and
and antibody responses. The antigen-specific T cells pro-
IgA plasma cells 29%), while in the IgAN patients, the
duced interferon-gamma (INF-c ) and lower levels of
plasma cells percentages was of an inversion (IgG 37%
interleukin-5 (IL-5). These results suggest that tonsils
and IgA 56%). This increment in the IgA population was
of the nasopharyngeal-associated lymphoreticular tissues
paralleled by an augmentation of the number of dimeric
represent a distinct component of the mucosal-associated
IgA secreting cells (75% of IgA plasma cells), stained
lymphoreticular tissues with features of both systemic
both for cytoplasmic IgA and J chain [32]. The study by
and mucosal compartments [20].
Nagy and Brandtzaeg [33] and a later multicenter study
by Bene et al [34] also demonstrated a similar result. In
DIFFERENCE OF TONSILS WITH additional, after 7 days of culture with pokeweed mito-
AND WITHOUT IgA NEPHROPATHY gen, the percentage of tonsillar cells producing polymeric
IgA is significantly higher in the IgAN patients than in the
There are significant differences in histologic structure,
controls suffering from chronic tonsillitis [35]. The IgA1
proportional ratios of cells, and cell adhesion molecules
subclass was found in follicular dendritic cells (FDC)
in tonsils with and without IgAN (Table 1).
of the tonsil of IgAN patients, but not in FDC of non-
IgAN controls. On the other hand, IgA2, IgG, IgM, and
Difference of tonsillar histologic structure C3 did not show any differences in distribution between
The enlarged primary T nodules in tonsils, which are the two groups [36]. In situ hybridization (ISH) study
defined as sum of the small areas of accumulating T lym- for the detection of J chain mRNA within IgA plasma
phocytes and apparent nodule, composed predominantly cells revealed J chain mRNA-positive cells were identi-
of small T lymphocytes, were a characteristic feature of fied in germinal centres, and within the subepithelial and
tonsils in patients with IgAN. Most T nodules in pa- interfollicular zones of tonsils. Combined immunofluo-
tients with IgAN were enlarged, especially in younger rescence and fluorescent ISH showed a greater propor-
patients, and a few T nodules contain high endothelial tion of J chain mRNA-positive interfollicular IgA cells in
venules and nonlymphoid cells. In contrast, T nodules the patient tonsils compared with the controls. In addi-
in patients with habitual tonsillitis do not expand, and tional, the finding of excess numbers of J chain-positive
nonlymphoid cells and high endothelial venules are dis- IgA-negative cells was found within germinal centers of
tributed peripherally around the nodules [29]. The basic tonsils in IgAN patients [37]. These results demonstrate
structure and functional unit of reactive lymph nodes is immune abnormalities within the tonsil as a central fea-
composed of two separate T nodules and B-lymphoid ture of abnormal polymeric IgA biology in IgAN, which
follicles. These composite nodules play a major role in is in keeping with the hypothesis favoring a tonsil origin
the triggering, helper T-cell–dependent stimulation and for the mesangial IgA present in their kidneys.
subsequent maturation of antigen-responsive B cells into Abnormalities in the partition of IgA- and IgG-
antibody-secreting plasma cells [30]. With regard to his producing cells in the tonsils of patients with IgAN have
relationship between T-cell and B-cell domains, the en- been suggested to result from a dysregulation of cell
1138 Xie et al: Tonsils and IgA nephropathy

trafficking and homing through high endothelial venules 5 minutes) and injecting hyaluronidase (2000 U/mL, each
in this lymphoid tissue. Study demonstrated a significant tonsil for 0.5 mL) into tonsils. In general, four criteria are
enhancement of cell adhesion molecules, CD31 and used to judge the results of tonsil provocation test. Any
CD54, expression on high endothelial venules of tonsils one of four criteria positive is regarded as tonsil provo-
from patients with IgAN compared with controls [38]. cation test positive: (1) an increase of white blood cell
In additional, the number of CD5+ B cells isolated from count over 1200/mm3 after 3 hours; (2) an increase in
the tonsil germinal centers of IgAN patients is increased. body temperature over 0.55◦ C after 15 minutes; (3) en-
These CD5+ B cells are likely IgA1 antibody-producing hancement of erythrocyte sedimentation rate over 12 mm
cells. Moreover, these CD5+ B cells show a reduced sus- after 1 hour; and (4) worsened skin eruption or deterio-
ceptibility to Fas-mediated apoptosis [39]. ration of urinary findings after 3 hours, which is defined
as urinary protein increased by more than 30 mg/dL or
erythrocyte count in the sediment increased by more than
RELATIONSHIP BETWEEN TONSILLAR IgA 10/hpf, as compared with that before the test [44, 45].
AND GLOMERULAR IgA
Both IgA produced by tonsil cells and IgA deposited
in glomerular mesangium with IgAN are mainly J chain- Effect of tonsil stimulation on IgAN
positive polymeric IgA [35, 37, 40, 41]. Studies demon- Although the pathogenesis of IgAN still remains un-
strated they were consistent in some cases. The antibodies certain, it is well known that IgAN patients often show
eluted from renal tissues of patients with IgAN specifi- gross hematuria or deteriorated urinary findings after
cally bound with the nuclear regions of tonsillar cells. upper respiratory tract infections such as tonsillitis, it is
The binding of eluted antibodies and tonsillar cells was supposed that tonsil inflammatory stimulation may be
completely inhibited by the addition of antihuman IgA related to IgAN. Masuda et al [46] reported that a ten-
antisera, but not inhibited by human IgA myeloma pro- dency of decreasing levels of serum complement com-
teins. The eluted antibodies bound with tonsillar cells bined with an increase of CIC was observed within 1 week
from the same patients, but only 10% of them bound after tonsil provocation test in several cases of IgAN as-
with the tonsillar cells obtained from other patients with sociated with chronic tonsillitis [46]. Shiraishi et al [44]
IgAN. This result suggests that IgA antibodies deposited performed the tonsil provocation test in 11 cases with pus-
in glomeruli specifically bind with tonsillar cells obtained tulosis palmaris et plantaris (PPP) and seven cases with
from patients with IgAN [42]. The study by Tokuda et al IgAN. Analysis of the provocation test proved positive
[43] offered another evidence of binding of IgA produced in three of 11 cases (27%) with PPP and in five of seven
by tonsillar B lymphocytes to the glomerular mesangium cases (71%) with IgAN [44]. Yamabe et al [45] studied
of IgAN. They first made heterohybridoma cells of hu- effect of ultrashort wave stimulation of tonsils on uri-
man tonsillar B lymphocytes from IgAN patient with nary findings in patients with IgAN. In 62 patients with
mouse myeloma cells and cultured them. The culture IgAN and 20 patients with other renal diseases, tonsils
medium was analyzed by Western blot analysis using an- were directly stimulated by Tonsil Provocator producing
tihuman IgA antibody, and both IgA1 and IgA2 were an ultrashort wave to 40.68 MHz each tonsil for 5 minutes.
demonstrated to be produced. The specimens of the biop- Forty (65%) of 62 patients with IgAN showed deteriora-
sied kidney tissue of IgAN were washed with 0.02 mol/L tion of urinary findings after the stimulation compared
citrate buffer (pH 3.2) to remove deposited IgA from with 6 (30%) of 20 patients with other renal diseases. The
glomerulus. The specimens were then incubated with the deterioration of urinary findings was significantly more
culture media of hybridoma cells, and immunofluores- frequent in IgAN than in other renal diseases. In addi-
cence analysis using fluorescein isothiocyanate (FITC)- tional, previous episodes of gross hematuria following
conjugated antihuman IgA antibody was performed. The upper respiratory tract infections and the level of serum
result demonstrated that IgA deposit was efficiently re- secretory IgA were higher in IgAN patients with deteri-
moved by washing with citrate buffer and was recovered oration of urinary findings after tonsil stimulation than
after incubation with the culture medium of hybridoma in those without deterioration [45]. Matsuda et al [47]
cells [43]. evaluated the effects of the mechanical tonsil stimulation
on the serum and urinary concentrations of macrophage-
colony-stimulating factor (M-CSF) in patients with IgAN
TONSIL STIMULATION AND IgAN
associated with chronic tonsillitis. The serum and urinary
Method and judging criteria of tonsil provocation test levels of M-CSF in the groups with mild and severe IgAN
The methods of tonsil provocation test include direct were significantly higher than those in the chronic tonsil-
or indirect tonsil stimulation using Tonsil Provocator pro- litis group without IgAN. Enhanced urinary excretion of
ducing an ultrashort wave (each tonsil for 5 minutes), me- M-CSF prolonged for 7 days after tonsil stimulation in
chanical tonsil stimulation (tonsil massage, each tonsil for the severe IgAN group; in contrast, the urinary M-CSF
 Xie et al: Tonsils and IgA nephropathy 1139

level was increased for only 2 days after tonsil stimulation groups compared with controls. That is, administration of
in the mild IgAN group. The urinary M-CSF level was H. parainfluenza outer membrane antigens to mice may
not changed in the chronic tonsillitis group after tonsil induce glomerular deposition of IgA and mesangial pro-
stimulation. These results suggest that tonsil stimulation liferation, resembling the changes seen in IgAN, with in-
contributes to the progression of IgAN via enhancement creases in IgA antibodies against H. parainfluenza outer
of glomerular production of M-CSF [47]. membrane antigens [54]. Furthermore, production of cy-
However, the usefulness of the tonsillar provocation tokines IL-10 and transforming growth factor-b (TGF-
test in IgAN is now doubted. Even otolaryngologists [46] b) was enhanced by stimulation with H. parainfluenza
who initially claimed that the tonsillar provocation test outer membranes in tonsillar mononuclear cells from
was of clinical value in patients with IgAN have already IgAN [55]. These results suggest that H. parainfluenza
changed their opinion. Their late results showed that antigens stimulate tonsillar T and B lymphocytes in pa-
there was no statistically significant difference between tients with IgAN to produce cytokines and IgA antibody
positive and negative patients in the rate of remission and that an immune response to H. parainfluenza anti-
of proteinuria based on any parameter of the tonsillar gens may play a role in the pathogenesis of IgAN in some
provocation test at any time after surgery [48, 49]. More- cases.
over, the Japan Society of Stomato-Pharyngology offi- In additional, Rekola et al [56] reported that 38 of 187
cially reported the lack of value of tonsillar provocation IgAN patients had possible acute glomerulonephritis at
test in determining the indications for tonsillectomy in the onset of their disease. Antistreptococcal antibodies
IgAN patients [50]. We think that an increase of white increased in forty-three percent of the patients. Thirty-
blood cell count, an increase of body temperature and three percent of the patients had different groups of beta-
enhancement of erythrocyte sedimentation rate after ton- hemolytic streptococci isolated from their throats. This
sillar provocation test may not be of any clinical value in result indicates a possible role of beta-hemolytic strepto-
patients with IgAN, but deterioration of urinary findings cocci, a most common bacterium in tonsils or throat, in
after tonsillar stimulation may be significant and suggest the pathogenesis of some IgAN cases [56].
that tonsils are related to kidneys. A questionnaire sur-
vey also showed that 51.6% of 154 medical doctors who
Tonsillar viral infection and IgAN
had reported case of IgAN answered that urine protein
was the most important factor in any estimation of the Regarding relationship between viral infection in
provocation test [51]. tonsils and IgAN, there is a adult case report in which
granular depositions of adeno- and herpes simplex vi-
ral antigens were detected in the glomerular mesangial
TONSILLAR INFECTION AND IgAN areas in IgAN patients associated with episodes of re-
Tonsillar bacterial infection and IgAN current tonsillitis and in the tonsillar epithelial cells by
Suzuki et al [52] reported that the antigen and anti- H. parainfluenza immunofluorescence [57]. The later
bodies of outer membranes of Haemophilus parainfluen- study showed that the detection ratio of Epstein-Barr
zae, a common bacterium on the tonsils, were present virus in the patients with glomerular lesions, such as IgAN
in glomerular mesangium and sera of IgAN, respectively, and membranous nephropathy, was significantly greater
suggesting that H. parainfluenza infection may have a role than those without. However, the detection of Epstein-
in the etiology of IgAN [52]. Further studies showed that Barr virus was not disease specific [58]. Kunimoto et al
tonsillar lymphocytes from patients with IgAN revealed a [59] investigated viral infections in the tonsils, pharynx,
significantly higher stimulation index to H. parainfluenza and renal tissues of patients with IgAN using cell culture,
antigens (thymidine incorporation in tonsillar lympho- polymerase chain reaction, and immunofluorescent tech-
cytes with H. parainfluenza/thymidine incorporation in niques, and measured antibody titers against numerous
unstimulated tonsillar lymphocytes) than controls. The types of viruses. As a result, no evidence was obtained that
lymphocytes from patients with IgAN also showed a the viral infections play a significant role in the pathogen-
significantly higher level of IgA antibody and IgA1 anti- esis of IgAN [59].
body against H. parainfluenza antigens in culture super-
natants than lymphocytes from controls [53]. In vivo study
EFFECT OF TONSILLECTOMY ON IgAN
showed that mouse glomerular deposition of H. parain-
fluenza outer membrane antigens and IgA, and increases Effect of tonsillectomy on immune system
in the amount of mesangial matrix were observed after As described above, human tonsil tissues are located
administration of H. parainfluenza outer membrane anti- at the gateway of the respiratory and alimentary tract,
gens orally or intraperitoneally, respectively. Levels of belong to the mucosa-associated lymphoid tissue, and
IgA antibodies against H. parainfluenza outer membrane play a role in the systemic immune and the local mu-
antigens were significantly increased in administration cosal immune. What effect does tonsillectomy have on
1140 Xie et al: Tonsils and IgA nephropathy

Table 2. Effect of tonsillectomy on urinary findings and renal function

Author Year Tonsillectomy +/− Follow-up months Urinary remission Urinary improvement Renal survival Reference number
Masuda 1988 16/0 36 56.3% 46
Tamura 1993 26/0 24 46% 71
Sugiyama 1993 28/0 61 32% 60.7% 79
Bene 1993 34/0 48 UP 3.5 g/day → 0.9 g/day Stable 69
Iino 1993 35/15 36 54.8%/53.8% 61%/46% 96.8%/76.9% 78
Tomioka 1996 104/0 12 38% 94% 80
Barta 1996 35/40 144 6m UP 1.4 g/day → 0.92 g/d 88.6%/80.0% 70
Rasche 1999 16/39 41 NS 74
Hotta 2001 250/79 75 Total 48% 81
Xie 2003 48/70 193 89.6%/63.7% 17
Abbreviations are: UP, urine protein; NS, non-statistic difference.

the systemic and local mucosal immune? Studies demon- thiocyanate, hypothiocyanite, and agglutinins, except
strated that tonsillectomy decreased the levels of serum lactoferrin, which declined significantly [66]. The effects
IgA and salivary secretory IgA, especially in children, of tonsillectomy on serum and salivary secretory IgG,
several months or years after operation [60–62]. How- IgM, and IgE remain still controversial [67, 68].
ever, these changes do not cause significant immune de-
ficiency and are clinically insignificant. Moreover, these Effect of tonsillectomy on urinary finding
alterations do not increase incidence of immunomodu- and renal function
lated diseases, such as infections of the upper respiratory
Studies demonstrated that tonsillectomy can improve
tract [63]. These may be because the concentrations of
the urinary finding and keep stable renal function in some
serum IgA and salivary secretory IgA are higher before
patients with IgAN (Table 2). Bene et al [69] followed
operation in some tonsillectomy cases than that of non-
up the evolution of urinary protein and serum creatinine
tonsillectomy controls. Tonsillectomy decreased signifi-
in 34 patients with IgAN, and Barta et al [70] followed
cantly the IgA levels compared with preoperation, but
up 35 IgAN patients after tonsillectomy. The urinary
there is no significant difference compared with normal
protein and microhematuria decreased significantly from
nonoperative controls [27]. Recent study demonstrated
6 months after tonsillectomy than that before operation,
children with chronic tonsillitis have increased levels of
and no significant variation was observed in the levels of
CD19+ B lymphocytes compared to healthy controls in
creatininemia [69]. Furthermore, tonsillectomy stopped
the pre-operative period. The percentage of B lympho-
gross hematuria in more than two thirds of patients [70].
cytes bearing CD23 was found to be significantly higher
Tamura et al [71] reported that 46% IgAN patients with
in patients, most likely representing in vivo B lymphocyte
chronic tonsillitis showed distinct improvement in urinary
activation due to chronic antigenic stimulation. After the
findings after the tonsillectomy. Akagi et al [48] followed
tonsillectomy, despite ongoing B lymphocyte activation,
up 24 patients with IgAN for more than 2 years after
CD8+ T lymphocyte levels increased and B cell levels
tonsillectomy. Remission of proteinuria was observed in
returned to normal [64].
41.7% of the patients 6 months after surgery and in 50.0%
Tonsillectomy may also lead to certain changes in the
2 years after surgery [48]. The clinical remission rate of
cellular immune systems in some boys, including slightly
urinary finding and the stable renal function rate in ton-
increased percentages of CD21+ cells, raised counts of
sillectomy patients with IgAN were significantly higher
CD4+ cells, absolute and relative increases in DR+ cells
than that in nontonsillectomy patients [72].
and a raised CD4+ DR count [63]. Peripheral blood CD8+
cells, CD45RA+ CD4+ cells, and CD8+ CD11b− cells in-
crease significantly after tonsillectomy, compared with Effect of tonsillectomy on renal histologic findings
their preoperative values in patients with IgAN accom- A repeat renal biopsy study for 35 patients demon-
panied by chronic tonsillitis. In some cases, the preop- strated that renal histologic finding improved distinctly
erative serum tumor necrosis factor-alpha (TNF-a), and after the combined therapy of methylprednisolone pulse,
INF-c levels were higher than normal before surgery, but prednisolone, antiplatelet, and tonsillectomy in IgAN pa-
decreased after surgery. These results suggest that tonsil- tients [73]. The interval between the first and second
lectomy suppresses a decrease in suppressor T cells in biopsy was 18 to 138 months (mean, 77.1 months) in that
patients with IgAN and corrects abnormal cell-mediated study. Mesangial proliferation and interstitial mononu-
immune responses in these patients [65]. clear cell infiltration were significantly reduced in sec-
In additional, tonsillectomy has no effect on com- ond biopsy specimens. Acute inflammatory glomerular
plement and saliva-derived nonimmunoglobulin host lesions, such as endocapillary proliferations, glomerular
defense factors, such as lysozyme, salivary peroxidases, tuft necrosis, and cellular crescents, were present in 32
 Xie et al: Tonsils and IgA nephropathy 1141

1979 1989 2001 in that study. Another study demonstrated that ESRD
was detected in four of 35 IgAN patients after 10 years
after tonsillectomy, in eight patients of 40 nontonsillec-
tomy controls [70]. We [17] followed up 118 patients
with idiopathic IgAN patients, including 48 tonsillectomy
patients and 70 nontonsillectomy patients, for 192.9 ±
BP (mm Hg) 132/88 120/80 120/80 74.8 months (48∼326 months). In that study, we used
UP (g/d) 0.4 0.4 0.5 three different statistical methods, including the chi-
U-RBC (/hpf) 2.3 2 2 squared test, Kaplan-Meier method with log-rank test,
Scr (mg/dL) 1.0 1.1 1.4 and Cox regression proportional hazards model in or-
Ccr (mL/min) 113 78.9 56 der to establish the efficacy of tonsillectomy in IgAN pa-
tients. Baseline characteristics at the time of renal biopsy,
Fig. 3. Effect of tonsillectomy on renal histological findings. This pa- pathologic finding, and therapy during observation were
tient was born in 1952, diagnosed with IgA nephropathy (IgAN) by the
first renal biopsy in 1979, received tonsillectomy in 1982, discovered not significant different between tonsillectomy and non-
hypertension in 1985, and received antihypertensive therapy. The pa- tonsillectomy patients. A mean 15 years after diagnostic
tient underwent the second biopsy in 1989 and the third biopsy in 2001. biopsy, only five (10.4%) of 48 tonsillectomy patients en-
Renal specimen was performed by periodic acid-Schiff (PAS) staining
(original magnification ×50). tered dialysis, whereas 18 (25.7%) of 70 nontonsillectomy
patients required dialysis, by chi-squared test, P = 0.0393.
Kaplan-Meier analysis showed the renal survival rates of
patients in first biopsy specimens, whereas these were no tonsillectomy patients were statistically higher than those
longer present in any of the second biopsy specimens. of non-tonsillectomy (log-rank test, P = 0.0329). For ex-
Although there was no significant difference in percent- ample, renal survival rates were 89.6% and 63.7% in the
age of globally sclerotic glomeruli between the first and patients with and without tonsillectomy, respectively, at
second biopsy specimens, the percentage of segmen- 240 months after renal biopsy. Cox regression analysis
tally sclerotic glomeruli was significantly lower in second showed that the relative risk for terminal renal failure in
biopsy specimens. The distribution of IgA mesangial de- patients with tonsillectomy was lower compared to non-
posits had diminished in most patients, and no IgA de- tonsillectomy patients (hazard ratio 0.22, 95% CI 0.06 to
posits were seen in second biopsy specimens from eight 0.76, P = 0.0164). The results of these three statistical
patients. Impact of isolated tonsillectomy on renal histo- analyses were consistent. All revealed that tonsillectomy
logic findings was unknown. We followed up a repeated had a favorable effect on long-term renal survival in pa-
biopsy patient with IgAN and tonsillectomy. He did not tients with IgAN [17].
receive other drug therapy except for antihypertension.
The first renal biopsy showed the marked mesangial pro- Indications and limitations of tonsillectomy
liferation, marked IgA deposit in glomerular mesangium,
and almost normal renal tubules and blood vessel. Af- Tonsillectomy and adenoidectomy procedures are
ter 10 years, the second biopsy showed the moderate among the oldest surgical procedures still performed to-
mesangial proliferation, moderate IgA deposit, mild-to- day. Otolaryngologically, the two main indications for
moderate renal tubular atrophy, and mild arteriole scle- tonsillectomy are upper airway obstruction due to ton-
rosis. The third biopsy after 22 years showed the enlarged sillar hypertrophy and recurrent acute or chronic ton-
glomeruli, mild mesangial proliferation, negative IgA de- sillitis. Adenoid hypertrophy with upper airway or eu-
posit, marked tubular atrophy, and moderate arteriole stachian tube obstruction and recurrent acute or chronic
sclerosis. The results of this patient demonstrated that adenoiditis or otitis media are main indications to per-
tonsillectomy can improve IgA deposit and mesangial form an adenoidectomy [75]. Nephrologically, indications
proliferation and cannot impact renal damage induced for tonsillectomy are to date still unclear. In fact, many
by other causes such as hypertension (Fig. 3). factors have effect on the efficacy of tonsillectomy in pa-
tients with IgAN, such as urinary finding and grades of
renal damage. In general, the efficacy of tonsillectomy
Effect of tonsillectomy on long-term renal survival in patients with hemoturia type IgAN, especially pre-
Rasche, Schwarz, and Keller [74] reported that there senting hemoturia after tonsil infection, is good [76]. We
was no significant correlation between tonsillectomy and have showed that with a mild renal damage condition,
ESRD by observing 16 IgAN patients with tonsillec- in which the amount of urine protein excretion was less
tomy and 39 patients without tonsillectomy, and intro- than 1.0 g/24 hours and global glomerular sclerosis less
duced that tonsillectomy does not reduce the risk of than 25%, none of 26 patients with tonsillectomy needed
developing renal failure [74]. The mean observation time dialysis, whereas five (13.2%) of 38 patients without ton-
after renal biopsy was relatively short (3.4 ± 4 years) sillectomy required dialysis [17]. The percentage entering
1142 Xie et al: Tonsils and IgA nephropathy

dialysis in the tonsillectomy patients with moderate renal why do IgA-producing cells be predominant in tonsils
damage, such as urinary protein was more than 1.0 g/24 with IgAN and how does IgA produced by tonsils de-
hours, but global glomerular sclerosis was less than 25% posit in mesangium. Third, tonsillectomy can improve the
of total, was less than half of that in the nontonsillec- urinary findings, keep stable renal function, and have a
tomy patients [17]. On the other hand, the patients with a favorable effect on long-tern renal survival in some IgAN
marked renal damage, in whom both the amount of urine patients. The indications of tonsillectomy in patients with
protein excretion was more than 1.0 g/24 hours and global IgAN include mainly the deterioration of urinary findings
glomerular sclerosis was more than 25% of total or cres- after tonsillar infection, mild or moderate renal damage.
cent formation was more than 25% of total might develop However, tonsillectomy may not be enough and may not
renal failure even if tonsillectomy was performed, that is, change the prognosis in IgAN patients with marked renal
tonsillectomy is mainly indicated for patients with mild damage. Unfortunately, studies regarding tonsillectomy
or moderate IgAN [17, 77, 78]. Rupture of the glomeru- were performed until now in a retrospective style and lit-
lar basement membrane occurred more frequently in the tle information has been available about the side effect
noneffective tonsillectomy than in the effective tonsillec- or complication of the operation in IgAN patients. In or-
tomy group [79, 80]. Hotta et al [81] conducted a ret- der to further clarify the clinical efficacy and security of
rospective investigation of the renal outcome in IgAN tonsillecteomy, randomized prospective controlled trials
patients with a median observation period of 75 months are necessary because of the high degree of variability of
after tonsillectomy and steroid pulse therapy. Their re- IgAN.
sults showed that there were no significant differences
between the tonsillectomy and nontonsillectomy groups
regarding the incidence of progressive renal functional ACKNOWLEDGMENTS
loss defined as a 50% increase in baseline serum crea- This study was supported by a grant from the Creative Research
tinine, but a combination of tonsillectomy and steroid Group of the National Foundation Committee of Natural Science of
P.R. China (30121005), and a grant from the Key Technologies Research
pulse therapy had a significant impact on clinical remis- and Development Programme of the Tenth Five-year Plan of P.R. China
sion by multivariate Cox regression analysis [81]. Sato (2001BA701A14a). We are grateful to Dr. Minoru Sakatsume and Dr.
et al [82] retrospectively investigated 70 patients with ad- Mitsuhiro Ueno for their help.
vanced IgAN (serum creatinine ≥1.5 mg/dL) classified Reprint requests to Professor Xiangmei Chen, Kidney Center of PLA,
into three groups according to their treatment regimens, Department of Nephrology, Chinese General Hospital of PLA, Fuxing
that is, steroid pulse with tonsillectomy (30 patients), con- Road 28, Beijing 100853, China.
E-mail: [email protected]
ventional steroid (25 patients), and supportive therapy
(15 patients). During the mean follow-up period of 70.3
(12 to 137) months, 41.4% of patients reached ESRD REFERENCES
(13.3% vs. 56.0% vs. 73.3%). The incidence of ESRD in 1. BERGER J, HINGLAIS N: Les depots intercapillaires d’IgA-IgG. J Urol
the patients treated by steroid pulse with tonsillectomy Nephrol (Paris) 74:694–695, 1968
was significantly lower than that in the patients treated by 2. BERGER J. IgA glomerular deposits in renal disease: Transplant Proc
1:939–944, 1969
conventional steroid and supportive therapy at a baseline 3. EMANCIPATOR SN: IgA nephropathy: morphologic expression and
creatinine level of 1.5 to 2 mg/dL, but no statistical dif- pathogenesis. Am J Kidney Dis 23:451–462, 1994
ference was observed at a level of >2 mg/dL [82]. These 4. D’AMICO G: The commonest glomerulonephritis in the world: IgA
nephropathy. Q J Med 245:709–727, 1987
results suggest tonsillectomy combined with steroid pulse 5. JULIAN BA, WALDO FB, RIFAI A, MESTECKY J: IgA nephropathy, the
therapy may be effective in the IgAN patients with a base- most common glomerulonephritis worldwide: A neglected disease
line creatinine level of ≤2 mg/dL, whereas when serum in the United States? Am J Med 84:129–132, 1988
6. LEVY M, BERGER J: Worldwide perspective of IgA nephropathy. Am
creatinine >2 mg/dL, tonsillectomy may not change renal J Kidney Dis 12:340–347, 1988
outcome even if that combines steroid therapy. 7. D’AMICO G, COLASANTI G, BARBIANO DI BELGIOIOSO G, et al: Long-
term follow-up of IgA mesangial nephropathy: Clinico-histological
study in 374 patients. Semin Nephrol 7:355–358, 1987
8. IBELS LS, GYORY AZ: IgA nephropathy: Analysis of the natu-
CONCLUSION ral history, important factors in the progression of renal disease,
First, human tonsils are lymphoid organs and play a and a review of the literature. Medicine (Baltimore) 73:79–102,
1994
role in production of antibodies and local mucosal im- 9. HOGG RJ, SILVA FG, WYATT RJ, et al: Prognostic indicators in chil-
mune, especially in children. Tonsillectomy decreases the dren with IgA nephropathy—Report of the Southwest Pediatric
levels of serum IgA and salivary secretory IgA, but these Nephrology Study Group. Pediatr Nephrol 8:15–20, 1994
10. RADFORD MG JR, DONADIO JV JR, BERGSTRALH EJ, GRANDE JP:
changes do not cause significant immune deficiency and Predicting renal outcome in IgA nephropathy. J Am Soc Nephrol
do not increase incidence of the upper respiratory tract 8:199–207, 1997
infections. Second, tonsils are closely related to IgAN and 11. NISHI S: The prognosis of IgA nephropathy—Favorable or poor?
Intern Med 40:679–680, 2001
polymeric IgA1 deposited in glomerular mesangium is at 12. MAISONNEUVE P, AGODOA L, GELLERT R, et al: Distribution of pri-
least in part of tonsillar origin. However, it is unclear that mary renal diseases leading to end-stage renal failure in the United
 Xie et al: Tonsils and IgA nephropathy 1143

States, Europe, and Australia/New Zealand: Results from an inter- 37. HARPER SJ, ALLEN AC, BENE MC, et al: Increased dimeric IgA-
national comparative study. Am J Kidney Dis 35:157–165, 2000 producing B cells in tonsils in IgA nephropathy determined by in
13. BRIGANTI EM, DOWLING J, FINLAY M, et al: The incidence of biopsy- situ hybridization for J chain mRNA. Clin Exp Immunol 101:442–
proven glomerulonephritis in Australia. Nephrol Dial Transplant 448, 1995
16:1364–1367, 2001 38. KENNEL-DE MARCH A, BENE MC, HURAULT DE LIGNY B, et al: En-
14. DONADIO JV, GRANDE JP: IgA nephropathy. N Engl J Med 347:738– hanced expression of CD31 and CD54 on tonsillar high endothelial
748, 2002 venules in IgA nephropathy. Clin Immunol Immunopathol 84:158–
15. LOMAX-SMITH JD, ZABROWARNY LA, HOWARTH GS, et al: The im- 165, 1997
munochemical characterization of mesangial IgA deposits. Am J 39. KODAMA S, SUZUKI M, ARITA M, MOGI G: Increase in tonsillar ger-
Pathol 113:359–364, 1983 minal centre B-1 cell numbers in IgA nephropathy (IgAN) patients
16. TOMINO Y, SAKAI H, AND SPECIAL STUDY GROUP (IGA NEPHROPATHY): and reduced susceptibility to Fas-mediated apoptosis. Clin Exp Im-
Clinical guidelines for immunoglobulin A (IgA) nephropathy in munol 123:301–308, 2001
Japan, second version. Clin Exp Nephrol 7:93–97, 2003 40. RIFAI A, SMALL PA JR, TEAGUE PO, AYOUB EM: Experimental IgA
17. XIE Y, NISHI S, UENO M, et al: The efficacy of tonsillectomy on long- nephropathy. J Exp Med 150:1161–1173, 1979
term renal survival in patients with IgA nephropathy. Kidney Int 41. HARPER SJ, FEEHALLY J: The pathogenic role of immunoglobulin A
63:1861–1867, 2003 polymers in immunoglobulin A nephropathy. Nephron 65:337–345,
18. HELLINGS P, JORISSEN M, CEUPPENS JL: The Waldeyer’s ring. Acta 1993
Otorhinolaryngol Belg 54:237–241, 2000 42. TOMINO Y, SAKAI H, ENDOH M, et al: Cross-reactivity of IgA anti-
19. VAN KEMPEN MJ, RIJKERS GT, VAN CAUWENBERGE PB: The immune bodies between renal mesangial areas and nuclei of tonsillar cells
response in adenoids and tonsils. Int Arch Allergy Immunol 122:8– in patients with IgA nephropathy. Clin Exp Immunol 51:605–610,
19, 2000 1983
20. BOYAKA PN, WRIGHT PF, MARINARO M, et al: Human 43. TOKUDA M, SHIMIZU J, SUGIYAMA N, et al: Direct evidence of the
nasopharyngeal-associated lymphoreticular tissues. Functional production of IgA by tonsillar lymphocytes and the binding of IgA
analysis of subepithelial and intraepithelial B and T cells from ade- to the glomerular mesangium of IgA nephropathy patients. Acta
noids and tonsils. Am J Pathol 157:2023–2035, 2000 Otolaryngol Suppl 523:182–184, 1996
21. MATTHEWS JB, BASU MK: Oral tonsils: An immunoperoxidase 44. SHIRAISHI S, TOMODA K, MATSUMOTO A, et al: Investigation of the
study. Int Arch Allergy Appl Immunol 69:21–25, 1982 local provocation test to PPP and IgA nephritis. Acta Otolaryngol
22. BRANDTZAEG P, SURJAN L JR, BERDAL P: Immunoglobulin- Suppl 523:178–181, 1996
producing cells in clinically normal, hyperplastic and inflamed hu- 45. YAMABE H, OSAWA H, INUMA H, et al: Deterioration of urinary find-
man palatine tonsils. Acta Otolaryngol Suppl 360:211–215, 1979 ings after tonsil stimulation in patients with IgA nephropathy. Acta
23. GOUMAS P, TROUBOUKIS D, TOSKA N, et al: Immunohistochemical Otolaryngol Suppl 523:169–171, 1996
study of the palatine tonsils. Laryngol Rhinol Otol (Stuttg) 67:34– 46. MASUDA Y, TERAZAWA K, KAWAKAMI S, et al: Clinical and immuno-
37, 1988 logical study of IgA nephropathy before and after tonsillectomy.
24. MORO I, KOMIYAMA K, IWASE T, et al: Immunoglobulin subclasses in Acta Otolaryngol Suppl 454:248–255, 1988
chronic tonsillitis. J Oral Pathol 17:475–477, 1988 47. MATSUDA M, SHIKATA K, WADA J, et al: Increased urinary excre-
25. VON GAUDECKER B, MULLER-HERMELINK HK: The development of tion of macrophage-colony-stimulating factor (M-CSF) in patients
the human tonsilla palatina. Cell Tissue Res 224:579–600, 1982 with IgA nephropathy: Tonsil stimulation enhances urinary M-CSF
26. BRANDTZAEG P: The B-cell development in tonsillar lymphoid folli- excretion. Nephron 81:264–270, 1999
cles. Acta Otolaryngol Suppl 523:55–59, 1996 48. AKAGI H, NISHIZAKI K, HATTORI K, et al: Prognosis of tonsillectomy
27. JUNG KY, LIM HH, CHOI G, CHOI JO: Age-related changes of IgA in patients with IgA nephropathy. Acta Otolaryngol Suppl 540:64–
immunocytes and serum and salivary IgA after tonsillectomy. Acta 66, 1999
Otolaryngol Suppl 523:115–119, 1996 49. AKAGI H, HATTORI K, DOI A, et al: Tonsillar provocation test for
28. BERNSTEIN JM: Mucosal immunology of the upper respiratory tract. tonsillectomy in patients with IgA nephropathy. Stomato-pharyngol
Respiration 59 (Suppl 3):3–13, 1992 11:297–305, 1999
29. KAWAGUCHI M, SAKAI T, SAKAMAKI A, et al: Expanded primary T 50. KATAURA A: Evaluation of the provocation test of tonsils: A re-
nodules in the palatine tonsils from patients with IgA nephropathy. port from the Committee for the standardization of diagnostic cri-
Acta Otolaryngol Suppl 508:36–42, 1993 teria for tonsillar focal infections. Stomato-pharyngol 9:213–221,
30. VAN DEN OORD JJ, DE WOLF-PEETERS C, DESMET VJ: The compos- 1997
ite nodule. A structural and functional unit of the reactive human 51. AKAGI H, KOSAKA M, DOI A, et al: Results of a survey of the re-
lymph node. Am J Pathol 122:83–91, 1986 lationship between palatine tonsil and IgA nephropathy. Nippon
31. SATO Y, HOTTA O, TAGUMA Y, et al: IgA nephropathy with poorly de- Jibiinkoka Gakkai Kaiho 102:305–310, 1999
veloped lymphoepithelial symbiosis of the palatine tonsils. Nephron 52. SUZUKI S, NAKATOMI Y, SATO H, et al: Haemophilus parainfluenzae
74:301–308, 1996 antigen and antibody in renal biopsy samples and serum of patients
32. BENE MC, FAURE G, HURAULT DE LIGNY B, et al: Immunoglobu- with IgA nephropathy. Lancet 343:12–16, 1994
lin A nephropathy. Quantitative immunohistomorphometry of the 53. SUZUKI S, FUJIEDA S, SUNAGA H, et al: Immune response of tonsillar
tonsillar plasma cells evidences an inversion of the immunoglobu- lymphocytes to Haemophilus parainfluenzae in patients with IgA
lin A versus immunoglobulin G secreting cell balance. J Clin Invest nephropathy. Clin Exp Immunol 119:328–332, 2000
71:1342–1347, 1983 54. YAMAMOTO C, SUZUKI S, KIMURA H, et al: Experimental nephropathy
33. NAGY J, BRANDTZAEG P: Tonsillar distribution of IgA and IgG im- induced by Haemophilus parainfluenzae antigens. Nephron 90:320–
munocytes and production of IgA subclasses and J chain in tonsil- 327, 2002
litis vary with the presence or absence of IgA nephropathy. Scand 55. FUJIEDA S, SUZUKI S, SUNAGA H, et al: Bacterial infection and pro-
J Immunol 27:393–399, 1988 duction of immunoglobulin in the tonsil. Rinsho Byori 47:724–729,
34. BENE MC, HURAULT DE LIGNY B, KESSLER M, FAURE GC: Confir- 1999
mation of tonsillar anomalies in IgA nephropathy: A multicenter 56. REKOLA S, BERGSTRAND A, BUCHT H, LINDBERG A: Are beta-
study. Nephron 58:425–428, 1991 haemolytic streptococci involved in the pathogenesis of mesangial
35. EGIDO J, BLASCO R, LOZANO L, et al: Immunological abnormalities IgA-nephropathy? Proc Eur Dial Transplant Assoc Eur Ren Assoc
in the tonsils of patients with IgA nephropathy: inversion in the 21:698–702, 1985
ratio of IgA: IgG bearing lymphocytes and increased polymeric IgA 57. TOMINO Y, YAGAME M, OMATA F, et al: A case of IgA nephropathy
synthesis. Clin Exp Immunol 57:101–106, 1984 associated with adeno- and herpes simplex viruses. Nephron 47:258–
36. KUSAKARI C, NOSE M, TAKASAKA T, et al: Immunopathological fea- 261, 1987
tures of palatine tonsil characteristic of IgA nephropathy: IgA1 lo- 58. IWAMA H, HORIKOSHI S, SHIRATO I, TOMINO Y: Epstein-Barr virus
calization in follicular dendritic cells. Clin Exp Immunol 95:42–48, detection in kidney biopsy specimens correlates with glomerular
1994 mesangial injury. Am J Kidney Dis 32:785–793, 1998
1144 Xie et al: Tonsils and IgA nephropathy

59. KUNIMOTO M, HAYASHI Y, KUKI K, et al: Analysis of viral infection in any change in long-term course of IgA nephropathy? Orv Hetil
patients with IgA nephropathy. Acta Otolaryngol Suppl 508:11–18, 137:2903–2906, 1996
1993 71. TAMURA S, MASUDA Y, INOKUCHI I, et al: Effect of and indication for
60. OSTERGAARD PA: IgA levels and carrier rate of Haemophilus in- tonsillectomy in IgA nephropathy. Acta Otolaryngol Suppl 508:23–
fluenzae and beta-haemolytic streptococci in children undergoing 28, 1993
tonsillectomy. Acta Pathol Microbiol Scand [C] 84:290–298, 1976 72. KOSAKA M: Long-term prognosis for tonsillectomy patients with
61. D’AMELIO R, PALMISANO L, LE MOLI S, et al: Serum and salivary IgA nephropathy. Nippon Jibiinkoka Gakkai Kaiho 101:916–923,
IgA levels in normal subjects: comparison between tonsillectomized 1998
and non-tonsillectomized subjects. Int Arch Allergy Appl Immunol 73. HOTTA O, FURUTA T, CHIBA S, et al: Regression of IgA nephropathy:
68:256–259, 1982 a repeat biopsy study. Am J Kidney Dis 39:493–502, 2002
62. CANTANI A, BELLIONI P, SALVINELLI F, BUSINCO L: Serum im- 74. RASCHE FM, SCHWARZ A, KELLER F: Tonsillectomy does not prevent
munoglobulins and secretory IgA deficiency in tonsillectomized a progressive course in IgA nephropathy. Clin Nephrol 51:147–152,
children. Ann Allergy 57:413–416, 1986 1999
63. BOCK A, POPP W, HERKNER KR: Tonsillectomy and the immune 75. PAULUSSEN C, CLAES J, CLAES G, JORISSEN M: Adenoids and tonsils,
system: A long-term follow up comparison between tonsillec- indications for surgery and immunological consequences of surgery.
tomized and non-tonsillectomized children. Eur Arch Otorhino- Acta Otorhinolaryngol Belg 54:403–408, 2000
laryngol 251:423–427, 1994 76. ZHOU WG, WANG TF, XUE Y, et al: Chronic tonsillitis and IgA
64. IKINCIOGULLARI A, DOGU F, IKINCIOGULLARI A, et al: Is immune nephropathy. Chin Med J (Engl) 106:770–772, 1993
system influenced by adenotonsillectomy in children? Int J Pedi- 77. HATTORI K: Therapeutic effects and prognostic factors of tonsil-
atr Otorhinolaryngol 66:251–257, 2002 lectomy for IgA nephropathy in long-term follow-up. Nippon Jibi-
65. KANAMOTO Y, SHIBATA R, OZONO Y, HARADA T: Effect of tonsillec- inkoka Gakkai Kaiho 101:1412–1422, 1998
tomy on peripheral blood T cell surface markers and cytokine pro- 78. IINO Y, AMBE K, KATO Y, et al: Chronic tonsillitis and IgA nephropa-
duction in patients with IgA nephropathy accompanied by chronic thy. Clinical study of patients with and without tonsillectomy. Acta
tonsillitis. Nippon Jinzo Gakkai Shi 36:1296–1302, 1994 Otolaryngol Suppl 508:29–35, 1993
66. KIRSTILA V, TENOVUO J, RUUSKANEN O, et al: Longitudinal anal- 79. SUGIYAMA N, SHIMIZU J, NAKAMURA M, et al: Clinicopathological
ysis of human salivary immunoglobulins, nonimmune antimicro- study of the effectiveness of tonsillectomy in IgA nephropathy ac-
bial agents, and microflora after tonsillectomy. Clin Immunol Im- companied by chronic tonsillitis. Acta Otolaryngol Suppl 508:43–48,
munopathol 80:110–115, 1996 1993
67. OSTERGAARD PA: IgA levels and carrier rate of pathogenic bacteria 80. TOMIOKA S, MIYOSHI K, TABATA K, et al: Clinical study of chronic
in 27 children previously tonsillectomized. Acta Pathol Microbiol tonsillitis with IgA nephropathy treated by tonsillectomy. Acta Oto-
Scand [C] 85:178–186, 1977 laryngol Suppl 523:175–177, 1996
68. SAINZ M, GUTIERREZ F, MORENO PM, et al: Changes in immuno- 81. HOTTA O, MIYAZAKI M, FURUTA T, et al: Tonsillectomy and steroid
logic response in tonsillectomized children. I. Immunosuppression pulse therapy significantly impact on clinical remission in patients
in recurrent tonsillitis. Clin Otolaryngol 17:376–379, 1992 with IgA nephropathy. Am J Kidney Dis 38:736–743, 2001
69. BENE MC, HURAULT DE LIGNY B, KESSLER M, et al: Tonsils in IgA 82. SATO M, HOTTA O, TOMIOKA S, et al: Cohort study of advanced
nephropathy. Contrib Nephrol 104:153–161, 1993 IgA nephropathy: Efficacy and limitations of corticosteroids with
70. BARTA J, KOVACS T, FAZEKAS A, et al: Does tonsillectomy cause tonsillectomy. Nephron Clin Pract 93:c137–c145, 2003