Full text of DOI:10.1053/ajkd.2001.27690

Tonsillectomy and Steroid Pulse Therapy Significantly Impact on
Clinical Remission in Patients With IgA Nephropathy
Osamu Hotta, MD, Mariko Miyazaki, MD, Takashi Furuta, MD, Sachiko Tomioka, MD,
Shigemi Chiba, MD, Ikuo Horigome, MD, Keishi Abe, MD, and Yoshio Taguma, MD
● We conducted a retrospective investigation of renal outcome in 329 patients with immunoglobulin A (IgA)
nephropathy with an observation period longer than 36 months (82.3 ؎ 38.2 months) in our renal unit between 1977
and 1995. Clinical remission, renal progression, and the impact of covariates were estimated by Kaplan-Meier
analysis and a Cox regression model. In 157 of 329 patients (48%), disappearance of urinary abnormalities (clinical
remission) was obtained. None of these 157 patients showed progressive deterioration, defined as a 50% increase
in serum creatinine (Scr) level from baseline, during the observation period. Conversely, in patients without clinical
remission, the Kaplan-Meier estimate of probability of progressive deterioration was 21% ؎ 5% at 10 years. In the
multivariate Cox regression model with 13 independent covariates, initial Scr level, histological score, tonsillec-
tomy, and high-dose methylprednisolone therapy had a significant impact on clinical remission, whereas protein-
uria, age, sex, levels of hematuria, blood pressure, conventional steroid therapy, angiotensin-converting enzyme
inhibitor therapy, and cyclophosphamide therapy had no significant effect. These findings indicate that interven-
tions aimed at achieving clinical remission have provided encouraging results applicable to managing patients with
IgA nephropathy.
© 2001 by the National Kidney Foundation, Inc.
INDEX WORDS: Immunoglobulin A (IgA) nephropathy; clinical remission; tonsillectomy; steroid pulse therapy.
MMUNOGLOBULIN A (IgA) nephropathy
is the most common form of glomerulone-
tients who had not achieved clinical remission
over a 20-year period developed chronic renal
failure or ESRF in the long run. In their report,
approximately 30% of patients with IgAnephrop-
athy experienced spontaneous clinical remission
during long-term follow-up (10 to 40 years). To
our knowledge, there is no evidence showing
that clinical remission terminates the progression
of nephropathy. Considering the life-long progno-
sis of patients with IgA nephropathy, if clinical
remission terminates progressive deterioration,
whether clinical remission is obtained thus ap-
pears to be very important. In this regard, factors
influencing clinical remission merit elucidation.
In the present study, we test the hypothesis
that clinical remission may terminate subsequent
progressive deterioration. We also analyzed mul-
tiple covariates, including clinical and pathologi-
cal findings at the time of renal biopsy and
therapeutic strategies, from the viewpoint of clini-
cal remission.
I
phritis worldwide,1 and it tentatively has been
established that 30% to 40% of patients reach
end-stage renal failure (ESRF) within 10 to 25
years of diagnosis as a result of the progressive
nature of this disease.2,3
The proportion of patients experiencing clini-
cal remission during the course of IgA nephropa-
thy has been uncertain.2,4 Recently, Yoshikawa et
al5 reported that clinical remission is often ob-
served in children with IgA nephropathy after
immunosuppressive therapy. Moreover, we have
experienced a similar phenomenon after inten-
sive therapy, even in adult patients.6 In contrast
to the large amount of knowledge about unfavor-
able clinical and pathological prognostic factors
associated with ESRF,1,7,8 little is known about
factors that contribute to clinical remission.
Chauveau and Droz2 reported that most pa-
From the Department of Nephrology, Sendai Shakaihoken
Hospital; and the Department of Otolaryngology, Sendai
Red Cross Hospital, Sendai, Japan.
METHODS
Patients
Received November 10, 2000; accepted in revised form
May 18, 2001.
Among 3,625 patients who underwent renal biopsy in our
renal unit at Sendai Shakaihoken Hospital (Sendai, Japan)
between 1977 and 1995, we identified 821 patients who had
histologically proven IgA nephropathy. The diagnosis was
limited to primary mesangial proliferative glomerulonephri-
tis with a predominant deposition of IgA in the mesangium.
No patient had clinical or histological evidence of such
systemic diseases as systemic lupus erythematosus, Henoch-
Scho¨nlein purpura, or chronic liver disease. Exclusion crite-
Address reprint requests to Osamu Hotta, MD, Depart-
ment of Nephrology, Sendai Shakaihoken Hospital, Tsutsumi-
machi 3-16-1, Aoba-ku, Sendai 981-8501, Japan. E-mail:
hottao@remus.dti.ne.jp
© 2001 by the National Kidney Foundation, Inc.
0272-6386/01/3804-0003$35.00/0
doi:10.1053/ajkd.2001.27690
736
American Journal of Kidney Diseases, Vol 38, No 4 (October), 2001: pp 736-743

CLINICAL REMISSION OF IgA NEPHROPATHY
737
Treatment
Supportive or immunosuppressive therapy was adminis-
tered at the treating physician’s discretion. Specific treat-
ment regimens in individual patients are listed in Table 1.
The following immunosuppressive therapies were used: (1)
high-dose methylprednisolone (0.5 g/d for 3 days for three
courses) followed by oral prednisolone at an initial dose of
0.6 mg/kg on alternate days, with a gradual decrease in
dosage over 1 year (steroid pulse therapy); (2) 1- to 1.5-year
course of oral prednisolone at an initial dose of 0.6 mg/kg/d
for 1 month (conventional steroid therapy) followed by a
gradual decrease in dosage; and (3) a 4-month course of
cyclophosphamide at a dose of 1 mg/kg/d.
Antiplatelet drugs (dipyridamole, 150 to 300 mg/d, or
dilazep dihydrochloride, 150 to 300 mg/d) were adminis-
tered to all 329 patients, at least during the initial follow-up
year. The duration of administration of antiplatelet drugs
was 31.3 Ϯ 23.2 months.
Immunosuppressive therapy basically was administered
when active glomerular lesions, such as crescents and tuft
necrosis, were present. In some patients without active
glomerular lesions, steroids, especially steroid pulse therapy,
were administered when patients sought clinical remission,
after informed consent was obtained.
In the vast majority of patients, tonsils were examined by
an otolaryngologist (S.T.). We recommended tonsillectomy
when pus was observed in tonsillar crypts. In addition, for
some patients who desired clinical remission, tonsillectomy
was performed after sufficient informed consent was ob-
tained regardless of the gross appearance of tonsils and even
in the absence of episodes of recurrent tonsillitis or gross
hematuria with tonsillitis. Two hundred fifty patients under-
went tonsillectomy. Eighteen patients had a history of tran-
sient gross hematuria synchronizing with pharyngitis. Pus
was observed in tonsillar crypts in most cases, but in some
patients, palatine tonsils had a normal gross appearance.
Antihypertensive drugs were administered when hyperten-
sion was present. Angiotensin converting-enzyme (ACE)
inhibitors were administered to 155 patients (47.1%;
Table 1).
Fig 1. Calculation of IGL. The degree of morphologi-
cal damage in each glomerulus was graded from 0 to 4
according to the percentage of injured lobules as the
result of mesangial proliferation and sclerosis. The
average for all glomeruli was calculated and regis-
tered as IGL. Abbreviations: N0-N4, numbers of glo-
meruli showing changes of grade 0 to 4, respectively:
N, N0 ؉ N1 ؉ N2 ؉ N3 ؉ N4.
ria for the present study were: (1) age younger than 15 years
or older than 60 years, (2) 24-hour urinary protein excretion
less than 0.5 g, (3) insufficient number of glomeruli (Ͻ5
glomeruli) in a biopsy specimen for light microscopic study,
(4) follow-up period less than 36 months, and (5) incomplete
data in the medical records. As a result, 329 of the 821
patients were included on the present study.
Clinical, Laboratory, and Pathological Data
Baseline data at the time of the diagnostic renal biopsy
were compiled from medical records by three trained extrac-
tors who were blinded to patients’ outcomes. Patient charac-
teristics included age, sex, blood pressure, serum creatinine
(Scr) level, degree of proteinuria and hematuria, and histolog-
ical score. Severity of the glomerular lesion (ie, histological
score) in each case was evaluated using the index of the
glomerular lesion (IGL) according to the method originally
proposed by Suwa and Takahashi,9 with a modification to
evaluate sclerotic changes as previously described10-12
(Fig 1).
Outcomes
In addition to baseline data, annual follow-up data were
assessed in the outpatient department of our institute.
The primary outcome for multiregression analysis was
either remission of urinary abnormalities or deterioration of
Table 1. Specific Treatment Regimens in Individual Patients
Therapy
Tonsillectomy (ϩ)
Tonsillectomy (Ϫ)
Total
Steroid pulse (n ؍ 197)
Cyclophosphamide (ϩ)
Cyclophosphamide (Ϫ)
54 (24)*
125 (59)
8 (4)
10 (6)
62 (28)
135 (65)
Conventional steroid (n ؍ 78)
Cyclophosphamide (ϩ)
Cyclophosphamide (Ϫ)
Steroids (؊) and cyclophosphamide (؊) (n ؍ 54)
Total
19 (10)
29 (17)
23 (8)
14 (9)
16 (5)
31 (13)
79 (37)
33 (19)
45 (22)
54 (21)
250 (118)
329 (155)
NOTE. Numbers in parentheses indicate number of patients who were given ACE inhibitors.

738
HOTTA ET AL
renal function. Remission of urinary abnormalities or clini-
cal remission was defined as negative proteinuria and hema-
turia by dipstick and urinary erythrocytes of 4/high-power
field or less. Deterioration of renal function or progression
was defined as a 50% increase in baseline Scr values. In
addition, total deaths and ESRF requiring renal replacement
therapy were included in the category of deterioration of
renal function.
RESULTS
Baseline Characteristics
The median duration of follow-up for the 329
patients with IgA nephropathy was 75 months
(range, 36 to 261 months). The man-woman ratio
was 1.2:1. Median age at renal biopsy was 36
years. Mean daily urinary protein excretion was
1.40 Ϯ 1.09 (SD) g. Massive proteinuria, defined
as 24-hour urinary protein excretion greater than
1.5 g, was observed in 33.1% of patients. Renal
dysfunction, defined as Scr level greater than 1.3
mg/dL, was observed in 20.7% of patients at the
time of renal biopsy. Elevated blood pressure,
defined as systolic blood pressure greater than
130 mm Hg or diastolic blood pressure greater
than 80 mm Hg, was present in 156 patients
(47.4%) and 111 patients (33.7%), respectively.
A high histological score, defined as IGL of 2.0
or greater, was observed in 43.5% of patients
(Table 2).
Statistical Analysis
We used the Cox proportional hazards model, choosing a
backward stepwise procedure13 to assess the impact of
multiple covariates for clinical remission. Variables used in
multivariable analysis were expressed using a binary scale,
such as absent/present, and coded as 0/1. The variables age,
sex, proteinuria, hematuria, hypertension, Scr level, histolog-
ical score, tonsillectomy, steroid pulse therapy, conventional
steroid therapy, ACE inhibitor therapy, and cyclophospha-
mide therapy were treated as categorical variables. Multi-
variable analysis results were expressed by hazard ratio as
an index of quantitative significance and 95% confidence
intervals and P as an index of statistical significance. The
influence of categorical variables on clinical remission was
tested using Cox regression analysis for each independent
variable by the enter method and for all variables by the
backward stepwise likelihood ratio method. In the Cox
analysis, exp (␤) is the hazard ratio for a case with the
characteristic compared with a case without the character-
istic.
To clarify the problem of whether clinical remission
(disappearance of prolonged urinary abnormalities) relates
to the termination of progressive deterioration of renal
function, we conducted a bivariate analysis. The clinical
remission rate was estimated by means of survival analysis
according to the Kaplan-Meier method.14 The log-rank test
was used for comparison of survival curves. For all analyses,
a two-tailed P less than 0.05 is considered statistically
significant. Correlations between dichotomous variables were
expressed by Spearman’s correlation coefficients.
Initial values for degree of proteinuria and
histological score (IGL) were significantly greater
in patients treated with steroid pulse therapy
(n ϭ 197) in contrast to the other patients (n ϭ
132) (proteinuria, 1.5 Ϯ 1.2 versus 1.2 Ϯ 0.8 g/d;
P ϭ 0.022; IGL, 2.03 Ϯ 0.63 versus 1.88 Ϯ 0.61;
P ϭ 0.040).
Clinical Remission and Outcomes
One hundred fifty-seven patients (48%) had
achieved clinical remission at the final observa-
tion, whereas another 172 patients still had pro-
longed urinary abnormalities at the end of the
observation period. Thirteen patients showed a
relapse of urinary abnormalities after transient or
prolonged remission. These 13 patients were
All statistical analyses, including chi-square test, Cox
proportional hazards model, and Kaplan-Meier analysis,
were performed using the Statistical Package for the Social
Sciences program, version 9.0 (SPSS Inc, Chicago, IL).
Table 2. Baseline Characteristics of 329 Patients With IgA Nephropathy
Cutoff for Cox Hazard
Sex (men/women)
Age (y)
Scr (mg/dL)
Proteinuria (g/24 h)
Urinary red blood cells (/HPF)
Systolic blood pressure (mm Hg)
Diastolic blood pressure (mm Hg)
IGL
178/151
36.1 Ϯ 12.8
1.14 Ϯ 0.48
Men (54.1)
Յ40 189 (57.4)
Յ1.3 261 (79.3)
Ͻ1.5 220 (66.9)
Յ29 175 (53.2)
Յ130 173 (52.5)
Յ80 218 (66.3)
Ͻ2.0 186 (56.5)
1.40 Ϯ 1.09
Median 10-29
132.3 Ϯ 17.2
78.0 Ϯ 12.3
1.97 Ϯ 0.63
Observation (mon)
82.3 Ϯ 38.2 (36-261)
NOTE. Data shown as mean Ϯ SD (range) or number (percent).
Abbreviation: HPF, high-power field.

CLINICAL REMISSION OF IgA NEPHROPATHY
739
renal dysfunction. The estimated probability of
progressive renal dysfunction at 10 years in
nonremission patients was 77.4%, but was 0%
in remission patients (log-rank test, P Ͻ 0.001;
Fig 2).
Thirty-eight of 250 patients (15.2%) showed
transient macroscopic hematuria after tonsillec-
tomy. In terms of the time to clinical remission
and percentage of patients who reached clinical
remission, there were no significant differences
between those with and without macroscopic
hematuria after tonsillectomy.
Cox Regression Analysis
Fig 2. Kaplan-Meier analysis comparing remission
with nonremission patients regarding the probability
of progressive renal dysfunction, ie, final Scr/initial
Scr > 1.50. P < 0.001, log-rank test.
Multivariate Cox regression analysis with 13
variables showed histological score (IGL) and
included with the 172 patients in the non–clinical
remission category. Among patients with clinical
remission, including 14 patients with impaired
renal function (Scr Ͼ 1.3 mg/dL) at the time of
renal biopsy, none showed progressive renal dys-
function, defined as a 50% increase in baseline
Scr values. Treatment regimens used in the 14
remission cases with initial Scr levels greater
than 1.3 mg/dL were steroid pulse therapy with
tonsillectomy in 7 patients (cyclophosphamide, 4
of 7 patients; ACE inhibitor, 2 of 7 patients),
conventional steroids with tonsillectomy in 5
patients (cyclophosphamide, 3 of 5 patients;ACE
inhibitor, 2 of 5 patients), and steroid pulse
therapy without tonsillectomy in 2 patients (cy-
clophosphamide, 1 of 2 patients; ACE inhibitor,
1 of 2 patients).
Twenty-four of 172 nonremission patients
(14%) showed progressive renal dysfunction,
and 14 of the 24 patients developed ESRF requir-
ing dialysis therapy. Seven of 118 nonremission
patients (6%) with Scr levels of 1.3 mg/dL or less
and 17 of 54 nonremission patients (31%) with
Scr levels greater than 1.3 mg/dL showed progres-
sive renal dysfunction (versus remission pa-
tients, chi-square test, P Ͻ 0.05). Five nonremis-
sion patients died of cardiomyopathy, subarachnoid
hemorrhage, cerebral infarction, abdominal aortic
aneurysm rupture, and interstitial pneumonitis be-
fore the development of ESRF.
Fig 3. Hazard ratios for the Cox regression model
in IgA nephropathy. Numbers are hazard ratios, with
95% percent confidence intervals (0.95 CI). Abbrevia-
tions: U-RBC, urinary red blood cells; hpf, high-power
field; sBP, systolic blood pressure; dBP, diastolic
blood pressure; ACEI, ACE inhibitor.*P < 0.05. **P <
0.01.
In addition, we performed a Kaplan-Meier
analysis comparing remission and nonremission
patients regarding the probability of progressive

740
HOTTA ET AL
Scr level at biopsy to be significant factors in
predicting clinical remission. Age, sex, and pro-
teinuria had no impact on clinical remission.
Systolic blood pressure less than 130 mm Hg and
diastolic pressure less than 80 mm Hg also had
no impact on clinical remission. Among charac-
teristics at the time of renal biopsy in the Cox
regression analysis with all covariates, IGL was
the predictor with the greatest influence on clini-
cal remission when the backward stepwise likeli-
hood ratio method was used. With respect to
treatment regimens, tonsillectomy and steroid
pulse therapy had significant impacts on clinical
remission. Conversely, conventional steroid, ACE
inhibitor, and cyclophosphamide therapies had
no impact on clinical remission (Fig 3).
Scr level, proteinuria, and degree of histological
score. The clinical remission rate was greater in
patients treated with the combination of steroids
(steroid pulse therapy, 153 patients; conventional
steroids, 38 patients) with tonsillectomy than
steroids (steroid pulse therapy, 12 patients; con-
ventional steroids, 22 patients) without tonsillec-
tomy (59.7% versus 35.3%; P Ͻ 0.01; Table 3).
Conversely, regarding the incidence of progres-
sive renal functional loss, there were no signifi-
cant differences between the steroid pulse therapy
versus non–steroid pulse therapy group or the ton-
sillectomy versus nontonsillectomy group (Fig 6).
Time between starting treatment and entering
clinical remission in 157 patients is shown in Fig
7. Ninety-four of 157 patients (59.9%) reached
clinical remission within 1 year after the initia-
tion of treatment, and 67 of these 94 patients
(71.3%) were treated with tonsillectomy and
steroid pulse therapy. Conversely, time to clini-
cal remission was more than 2 years in 36 of 157
patients (22.9%), of whom a combination of
tonsillectomy and steroid pulse therapy was used
in 16 patients (44.4%; incidence of tonsillectomy
with steroid pulse therapy for Ͻ1 versus Ͼ2
years, chi-square test, P Ͻ 0.01).
Correlations Between Clinical Remission Rate
and Parameters
Because Cox analysis showed Scr level and
IGL to be significant covariates in clinical remis-
sion, we analyzed relationships between these
covariates and clinical remission rates. More
than half the men and women patients with Scr
levels of 1.3 mg/dL or less and 1.0 mg/dL or less
experienced clinical remission, respectively (Fig
4). Clinical remission rates decreased in parallel
with an increase in histological score (IGL; Fig 5).
Because Cox analysis showed tonsillectomy
and steroid pulse therapy had a significant im-
pact on clinical remission, the remission rate in
steroid-treated patients with Scr levels of 1.4
mg/dL or less was compared between those with
and without tonsillectomy. Baseline data for these
two groups were statistically matched in terms of
DISCUSSION
This is the first report to address the clinical
remission of IgA nephropathy. Although clinical
characteristics of patients with IgA nephropathy
may be influenced by race, according to reports
by the Japanese Ministry of Health and Welfare15
and Japanese Society of Dialysis,16 IgA nephrop-
athy onset peaks in persons aged in the teens and
Fig 4. Correlation be-
tween serum creatinine level
at renal biopsy and percent-
age of patients experiencing
clinical remission. (■), Men
(n ؍ 178); (ᮀ), women (n ؍
151).

CLINICAL REMISSION OF IgA NEPHROPATHY
741
Fig 5. Correlation be-
tween histological score
(IGL) and percentage of pa-
tients experiencing clinical
remission. (ᮀ), Remission
(n ؍ 157); (■), nonremission
(n ؍ 172).
20s,15 and the average age at which dialysis
therapy is required because of ESRF caused by
chronic glomerulonephritis (with IgA nephropa-
thy presumably the major cause) is approxi-
mately 60 years.16 Based on these two reports,
we can assume that the duration of IgA nephrop-
athy from onset to ESRF is 30 to 40 years in
many cases. At present, predicting the course of
nephropathy beyond 30 years is essentially impos-
sible, especially for patients with IgA nephropa-
thy who do not have unfavorable prognostic
factors, such as impaired renal function, massive
proteinuria, or an advanced histological score,1,7,8
at the time of diagnosis. However, in the absence
of clinical remission, nephropathy may show an
essentially progressive course, although the speed
of this progression varies among patients.2 Con-
versely, after clinical remission is obtained, termi-
nation of the progressive decline in renal func-
tion can be expected. This concept has been
supported in other forms of glomerular disease,
such as focal segmental sclerosis17-19 and mem-
branous nephropathy.20 Our present study con-
firms this hypothesis.
In multivariate Cox regression analysis, Scr
level and histological score had a significant
impact on clinical remission. Regarding therapeu-
tic strategies for achieving clinical remission of
IgAnephropathy, tonsillectomy, and steroid pulse
therapy, but not conventional steroid, ACE inhib-
itor, or cyclophosphamide therapy, had a signifi-
cant impact on clinical remission of IgAnephrop-
athy.
The clinical benefit of tonsillectomy in pa-
tients with IgA nephropathy remains controver-
sial. With respect to clinical remission, several
studies, mostly by Japanese groups, have shown
positive results.21-24 In accordance with these
findings, the clinical remission rate was signifi-
cantly greater in patients administered steroids
with tonsillectomy than without tonsillectomy.
Conversely, with respect to progression of ne-
phropathy, a recent study using Cox regression
analysis by Rasche et al25 showed no benefit.
In Japan, the system for annual health exami-
nation is well developed. Approximately 70% of
Japanese patients with IgA nephropathy were
found to have renal disease on routine health
Table 3. Clinical Remission Rate in Steroid-Treated Patients With Scr of 1.4 mg/dL or Less
With and Without Tonsillectomy
Tonsillectomy ϩ Steroids
Steroids
No. of patients
191
34
Remission rate (%)
Age (y)
Scr (mg/dL)
Proteinuria (g/24 h)
Systolic blood pressure (mm Hg)
Diastolic blood pressure (mm Hg)
IGL
59.7
35.3*
33.5 Ϯ 12.2
0.96 Ϯ 0.22
1.38 Ϯ 1.17
127.9 Ϯ 15.0
75.5 Ϯ 11.1
1.89 Ϯ 0.55
39.6 Ϯ 13.6*
1.01 Ϯ 0.17
1.35 Ϯ 0.75
136.3 Ϯ 16.8†
78.5 Ϯ 13.5
1.85 Ϯ 0.50
*P Ͻ 0.01 at chi square test.
†P Ͻ 0.05, at chi square test.

742
HOTTA ET AL
Fig 6. (A) Kaplan-Meier analysis comparing patients with and without steroid pulse therapy regarding the
probability of progressive renal dysfunction, ie, final Scr/initial Scr > 1.50. Not significant in log-rank test, P ؍ 0.40.
(B) Kaplan-Meier analysis comparing patients with and without tonsillectomy. Not significant in log-rank test, P ؍
0.15.
examinations,15 and renal biopsies often are per-
formed even for patients with a mild degree of
proteinuria. Under such circumstances, early-
stage IgA nephropathy is found more often than
in countries with limited and strict indications
for renal biopsy, such as the United States and
Germany.
Regarding the stage of nephropathy in our
patients, only 7% reached the end point of a 50%
increase from baseline Scr level. In the study of
Pozzi et al,26 in which the same end point was
used but patients with proteinuria less than 1.0
g/d of protein were excluded, 21% of patients on
steroid pulse therapy and 33% of controls had
reached the same end point by year 5 of follow-
up. In the study by Donadio et al,27
in which the
efficacy of fish oil for IgA nephropathy was
shown, 33% of control patients (proteinuria Ն
1.0 g/d of protein) showed an increase in Scr
level of 50% or greater by 2.2 years. Therefore, it
is conceivable that the stage of nephropathy in
our patients as a whole was earlier than those in
previous studies. It is obvious that spontaneous
clinical remission is seen rather frequently in
mild IgA nephropathy. However, to avoid exces-
sive representation of mild IgA nephropathy, we
excluded patients with proteinuria less than 0.5
g/d of protein. Moreover, clinical remission was
observed even in patients with impaired renal
function, in whom spontaneous clinical remis-
sion is unlikely. In addition, degrees of protein-
uria and histological scores were significantly
greater in patients treated with steroid pulse
therapy in contrast to the others. Finally, in
patients who entered clinical remission in the
early observation period (Ͻ1 year), the combina-
tion of tonsillectomy and steroid pulse therapy
was used more often than in patients who
achieved clinical remission later (Ͼ2 years; Fig
7). We therefore speculate that the clinical remis-
sion observed in the present study was, at least in
a considerable proportion of cases, a result of
treatment intervention, especially tonsillectomy
and steroid therapy, rather than the natural course
of this disease.
Fig 7. Time between the initiation of therapy and
clinical remission in 157 remission patients. (■), Ste-
roid pulse therapy with tonsillectomy (n ؍ 97); (ᮀ),
steroid pulse therapy without tonsillectomy (n ؍ 7);
(
), tonsillectomy without steroid pulse therapy (n ؍
38), among whom conventional steroids were adminis-
tered to 27 patients; ( ), without tonsillectomy and
steroid pulse therapy (n ؍ 15), among whom conven-
tional steroids were administered to 6 patients.
Despite the significant impact of tonsillec-
tomy and steroid pulse therapy on clinical remis-

CLINICAL REMISSION OF IgA NEPHROPATHY
743
10. Sato Y, Hotta O, Taguma Y, Takasaka T, Nose M: IgA
nephropathy with poorly development lymphoepithelial sym-
biosis of the palatine tonsils. Nephron 74:301-308, 1996
11. Hotta O, Taguma Y, Kurosawa K, Sudo K, Suzuki K,
Horigome I: Predictive value of small crescents in IgA
nephropathy: Analysis of four patients showing a deterio-
rated renal function during a long follow-up period. Clin
Nephrol 40:125-130, 1993
12. Ootaka T, Suzuki M, Sudo K, Sato H, Seino J, Saito
T, Yoshinaga K: Histological localization of terminal comple-
ment complexes in renal diseases: An immunohistochemical
study. Am J Clin Pathol 91:144-151, 1989
sion, clinical outcome data failed to reach signifi-
cance. This may have occurred partly because
patients in the present study were at an early
stage of disease and the overall rate at which a
50% increase in Scr level occurred was too low,
or the population was still too small to allow the
detection of improved outcomes.
Although the present study was retrospective,
our results may provide useful information for
conducting a future prospective study. Namely,
treatment strategies leading to clinical remission
should be encouraging for patients with IgA ne-
phropathy. Terminating the progressive deteriora-
tion of renal function is the goal of effective therapy.
13. Cox DR: Regression models and life tables. J R Stat
Soc 34:187-220, 1972
14. Kaplan EL, Meier P: Nonparametric estimation for
incomplete observation. J Am Stat Assoc 53:116-122, 1958
15. Endoh M: Epidemiologic analysis of IgA nephropa-
thy in Japan [Japanese]. Kidney Dial 46:29-32, 1999
16. Japanese Society for Dialysis Therapy: An Overview
of Regular Dialysis Treatment in Japan (as of December 31,
2000) [Japanese]. Tokyo, Japan, Japanese Society for Dialy-
sis Therapy, 2001
ACKNOWLEDGMENT
The authors thank Dr S. Nakai, Nagoya University School
of Medicine, for helpful advice regarding statistical analysis,
and Dr H. Imai, Akita University School of Medicine, for
comments on this manuscript.
17. Kobet SM, Schwartz MM, Lewis EJ: Primary focal
segmental glomerulosclerosis: Clinical course and response
to therapy. Am J Kidney Dis 23:773-783, 1994
REFERENCES
1. D’Amico G, Imbasciati E, Barbiano D, Belgioioso G,
Bertoli S, Fogazzi G, Ferrario F, Fellin G, Ragni A, Colas-
anti S, Minetti L, Ponticelli C: Idiopathic IgA mesangial
nephropathy. Clinical and histological study of 374 patients.
Medicine 64:49-60, 1985
2. Chauveau D, Droz D: Follow-up evaluation of the first
patients with IgA nephropathy described at Necker Hospital.
Contrib Nephrol 104:1-5, 1993
3. Galla JH: IgA nephropathy. Kidney Int 47:377-387, 1995
4. Costa RS, Droz D, Noel LH: Long-standing spontane-
ous clinical remission and glomerular improvement in pri-
mary IgA nephropathy (Berger’s disease). Am J Nephrol
7:440-444, 1987
5. Yoshikawa N, Ito H, Sakai T, Takekoshi Y, Honda M,
Awazu M, Ito K, Iitaka K, Koitabashi Y, Yamaoka K,
Nakagawa K, Nakamura H, Matsuyama S, Seino Y, Takeda
N, Hattori S, Ninomiya M: Controlled trial of combined
therapy for newly diagnosed severe childhood IgA nephrop-
athy. J Am Soc Nephrol 10:101-109, 1999
6. Hotta O, Taguma Y, Yoshizawa N, Oda T, Nishiyama J,
Yusa N, Chiba S, Horigome I, Sudo K, Tomioka S: Long-
term effects of intensive therapy combined with tonsillec-
tomy in patients with IgA nephropathy. Acta Otolaryngol
523:S165-S168, 1996 (suppl)
7. Koyama A, Igarashi M, Kobayashi M: Natural history
and risk factors for immunoglobulin A nephropathy in Ja-
pan. Am J Kidney Dis 29:526-532, 1997
8. Yoshida H, Mitarai T, Kawamura T, Kitajima T, Miyazaki
Y, Nagasawa R, Kawaguchi Y, Kubo H, Ichikawa I, Sakai O:
Role of the deletion polymorphism of angiotensin-converting
enzyme gene in the progression and therapeutic responsiveness
of IgA nephropathy. J Clin Invest 96:2162-2169, 1995
9. Suwa N, Takahashi T: Morphological and Morpho-
metrical Analysis of Circulation in Hypertension and Is-
chemic Kidney. Mu¨nchen-Berlin-Wien, Germany, Urban &
Schwarzenberg, 1971, pp 108-116
18. Pei Y, Cattran D, Delmore T, Katz A, Lang A, Rance
P: Evidence suggesting under-treatment in adults with idio-
pathic focal segmental glomerulosclerosis. Regional Glomer-
ulonephritis Registry Study. Am J Med 82:938-944, 1987
19. Banfi G, Moriggi M, Sabadini E, Fellin G, D’Amico
G, Ponticelli C: The impact of prolonged immunosuppres-
sion on the outcome of idiopathic focal-segmental glomeru-
losclerosis with nephrotic syndrome in adults. A collabora-
tive retrospective study. Clin Nephrol 36:53-59, 1991
20. Passerini P, Pasquali S, Cesana B, Zucchelli P, Ponti-
celli C: Long-term outcome of patients with membranous
nephropathy after complete remission of proteinuria. Neph-
rol Dial Transplant 4:525-529, 1989
21. Lagrue G, Sardreoux T, Laurant J, Hirbec G: Is there
a treatment of mesangial IgA nephropathy. Clin Nephrol
14:161, 1981 (letter)
22. Sanai A, Kudoh F: Effects of tonsillectomy in chil-
dren with IgA nephropathy, purpura nephritis, or other
chronic glomerulonephritis. Acta Otolaryngol 523:S172-
S174, 1996 (suppl)
23. Sugiyama N: Clinicopathological study of the effective-
ness of tonsillectomy in IgA nephropathy accompanied by
chronic tonsillitis. Acta Otolaryngol 508:S43-S48, 1993 (suppl)
24. Iino Y, Ambe K, Kato Y, Nakai A, Toriyama M,
Saima K, Yoshimoto K: Chronic tonsillitis and IgA nephrop-
athy. Acta Otolaryngol 508:S29-S35, 1993 (suppl)
25. Rasche FM, Schwarz A, Keller F: Tonsillectomy does
not prevent a progressive course in IgA nephropathy. Clin
Nephrol 51:147-152, 1999
26. Pozzi C, Bolasco PG, Fogazzi GB, Andrulli S, Altieri P,
Ponticelli C, Locatelli F: Corticosteroids in IgAnephropathy:A
randomized controlled trial. Lancet 353:883-887, 1999
27. Donadio JV Jr, Bergstralh EJ, Offord KP, Spencer
DC, Holley KE: A controlled trial of fish oil in IgA nephrop-
athy. N Engl J Med 331:1194-1199, 1994