Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
451
Original Article
SMJ
Phatharaporn Kiatpanabhikul, M.D., M.Sc.
Department of Internal Medicine, Charoenkrung Pracharak Hospital, Medical Service Department, Bangkok Metropolitan Administration, Bangkok
10120, ailand.
Relapse Rate and Clinical Risk Factors Affecting
the Treatment of Graves’ Disease
ABSTRACT
Objective: To determine the relapse rate of Graves’ disease (GD) and identify important clinical risk factors for relapse.
Materials and Methods: is was a 10-year retrospective cohort study. Information was collected with ICD10 E050
codes for Graves’ hyperthyroidism among ai patients of both sexes and all ages with no history of pregnancy,
thyroid storm or antithyroid drug (ATD) allergy.
Results: e 286 included GD patients had a relapse rate of 35% aer ATD withdrawal for one year. e clinical
risk factors associated with relapse were male sex (p = 0.014), smoking (p = 0.001), serum free T4 (FT4) levels >
2 times the upper normal range at diagnosis (p = 0.005), duration for maintenance treatment < 6 and 9 months
(p < 0.005) compared with remission. A TSH level < 1 mIU/L (p = 0.060) and MMI > 2.5 mg per day before ATD
withdrawal (p = 0.094) trended toward associations with relapse. e clinical factors that predicted GD relapse
were serum FT4 levels at diagnosis (p = 0.006) and serum free T3 (FT3) levels before ATD withdrawal (p = 0.019).
Conclusion: Male sex, smoking and serum FT4 levels at diagnosis > 2 times the normal range were signicant
clinical factors for GD relapse in ai patients. To reduce the relapse rate in the rst year, MMI should be used in
maintenance periods for 9 to 12 months with serum FT3 levels within low-normal ranges before ATD withdrawal.
is would promote future guidelines for GD management in ailand.
Keywords: Relapse; Graves’ disease; antithyroid drugs; risk factors (Siriraj Med J 2021; 73: 451-461)
Corresponding author: Phatharaporn Kiatpanabhikul
E-mail: kiatpanap@gmail.com
Received 15 February 2021 Revised 23 March 2021 Accepted 25 March 2021
ORCID ID: http://orcid.org/0000-0001-6914-7092
http://dx.doi.org/10.33192/Smj.2021.59
INTRODUCTION
Graves’ disease (GD) is the most common cause of
thyrotoxicosis. ere are 3 modalities for management:
antithyroid drugs (ATDs), radioactive iodine (RAI) and
surgery. According to survey results for the management
of GD conducted by members of the Endocrine Society
of the USA, the proportions of respondents with ATDs,
RAI therapy and thyroid surgery as the preferred modes
of therapy for uncomplicated GD are 53.9%, 45% and
0.7%, respectively. However, among members of the
Endocrine Society of ailand, ATDs are the preferred
choice of therapy (90.8%), with RAI therapy (9.2%) being
the second most preferred. In a corresponding survey
of the members of the European yroid Association,
ATDs were preferred by 83.8% of respondents and RAI
therapy by 14.1% of respondents. ATDs have been used
to treat hyperthyroidism for more than 75 years. In
ailand, 2 ATD types are used: methimazole (MMI)
and propylthiouracil (PTU). Generally, the guidelines
suggest MMI for treatment of GD or other causes of
hyperthyroidism for all patients except patients with
a rst-trimester pregnancy, thyroid storm or minor
reaction to MMI.
1,
e treatment goal is restoration to
normal thyroid function as soon as reasonably possible
even though this mode is not curative among GD patients
but suppresses their autoimmunity for only a short time.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
452
e GD relapse and remission rates vary among
geographical areas due to varying in the duration of
treatment and dierent iodine status.
,,,,,,,
In the USA, the
remission rate aer ATD use for at least 12-18 months is
20-30%
10
, while that in Europe is 50-60% aer medication
use for 5-6 years.
11
In Sweden, the relapse rate is 30.2% aer
stopping ATDs within 1 year.
12
In Japan, the remission rate
is 68% for at least 2 years aer treatment with maintenance
doses of MMI for a minimum of 19 months.
13
ere are
many clinical factors for predicting GD relapse, such as
male sex, smoking, a large thyroid goiter, the FT3/FT4
ratio, and continuous TSH suppression.
,,,,,
In principle, ATDs are used in most patients with
varying characteristics and are easy to access. However,
the responses to ATD treatment dier according to race
and continental regions worldwide.
10-14,18
e previous
study in ailand showed the low remission rate of
children with GD only 18.8% aer ATD treatment about
3.5 years. A recent study in adult showed the early relapse
of hyperthyroidism within a year aer MMI withdrawal,
27.8% of median duration of treatment being nearly 2 years.
However, there is no clear guideline for the treatment of
such patients. Current management practices are based
on those of outside countries; thus, there are variations
in remedies, especially regarding treatment dosages and
durations. us, this study aimed to evaluate the GD
relapse rate in ailand and clinically relevant factors
aecting the disease that can be used as basic information
to properly set treatment guidelines and reduce the cost
of screening for TRAb levels before discontinuing ATDs
in some groups of patients. e primary objective was to
determine the GD relapse rate. e secondary objective
was to identify important clinical factors for GD relapse.
MATERIALS AND METHODS
Study population
is retrospective cohort study included consecutive
searches of an electronic database with ICD10 E050 for
new diagnoses of GD, was 526 patients, that had been
referred to and/or treated in all clinics at Charoenkrung
Pracharak Hospital, Bangkok, ailand, between 1
st
January 2007 and 31
st
December 2017. e inclusion
criteria were ai patients of any age and gender. e
exclusion criteria were patients with thyroid storms, major
side eects of ATDs, a history of prior RAI treatment
and thyroidectomy, loss to follow-up before remission,
pregnancy and death before remission. Of those, 286
patients were eligible by reviewing all documentations
carefully. GD was diagnosed based on clinical diagnosis
based on indicators including high serum FT4 and/or
serum FT3 with suppressed TSH accompanied by one of
the following clinical characteristics for at least 3 months:
symptoms and signs of hyperthyroidism (pulse >100
bpm and/or ne tremor and/or warm and moist skin
and/or onycholysis and/or proximal muscle weakness),
diuse goiter with or without bruit, thyroid-associated
ophthalmopathy (TAO), and thyroid dermopathy
(pretibial myxedema and/or thyroid acropachy). Goiter
was categorized by physical examination by a physician:
normal gland (10-20 gm), mild-to-moderate goiter (25-
60 gm) and large goiter (>60 gm). TAO was classied
according to nonspecic signs, lid lag and retraction,
and specic signs, including exophthalmos.
14
An ATD
titration regimen was commonly used. e duration of
ATDs was divided into two periods: initial treatment and
maintenance treatment. e initial period consisted of
MMI > 5 mg/day or PTU > 100 mg/day followed by a
maintenance period that consisted of MMI ≤ 5 mg/day
or PTU ≤ 100 mg/day until withdrawal therapy. is
study was approved by the Human Research and Ethics
Committee of Bangkok Metropolitan Administration,
Bangkok, ailand.
Laboratory measurement
Serum FT4, FT3 and TSH levels were measured
by electrochemiluminescent immunoassay (ECLIA)
methods using Roche Diagnostics with normal ranges
of 0.93 - 1.70 ng/dL, 2.20 - 4.40 pg/mL and 0.270 - 4.200
mIU/L, respectively. If any result value is too large or
too small to measure, I will display only the highest or
lowest reading for analysis.
Denitions of clinical outcomes
Remission was dened as euthyroidism with normal
FT4, FT3 and TSH maintained for at least one year aer
ATD withdrawal. Relapse was dened as recurrence of
hyperthyroidism, including low TSH with or without
high FT4 and/or FT3 during follow-up aer withdrawal
therapy for less than one year.
Statistical analyses
e relapse rate and remission rate are presented as
percentages. Continuous variables are described as means
with standard deviations or medians with interquartile
ranges (IQRs) and were analyzed using Student’s t test
or the Mann-Whitney U test. Categorical variables are
presented as percentages and were examined with the
chi-square test or Fisher’s exact test. e multivariate
analysis included age, sex, smoking, the presence of TAO,
goiter size, serum FT4 and FT3 levels at diagnosis and
aer ATD withdrawal, serum TSH levels aer stopping
ATDs and ATD dosage before withdrawal, which were
Kiatpanabhikul.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
453
Original Article
SMJ
analyzed by using logistic regression; the data are presented
as the odds ratio (OR), 95% condence interval (CI) and
p-value. All two-sided p-values <0.05 were considered
statistically signicant. Statistical analyses were carried
out using SPSS version 26.0.
RESULTS
Baseline and clinical characteristics of patients with GD
e mean age at diagnosis of the 286 GD patients
was 43.5 ± 13.6 years (minimum, 8 years; maximum, 80
years), and 77.3% were female. Most were non-smokers
(94.1%), and 48.3% had no underlying disease. Among
patients with underlying disease, the most common
disease was dyslipidemia, 36% (Table 1).
Regarding the clinical characteristics, laboratory
results and treatments, there was no TAO in 82.5% of
cases and mild-to-large goiter at diagnosis in 68.5% of
cases. e mean levels of serum FT4, FT3 and TSH at
diagnosis were 5.38 ± 2.14 ng/dL, 18.30 ± 8.87 pg/mL and
0.0096 ± 0.023 mIU/L, respectively. e mean serum FT4,
FT3 and TSH levels aer ATD withdrawal were 1.22 ±
0.29 ng/dL, 2.88 ± 0.49 pg/mL and 3.235 ± 0.289 mIU/L,
respectively. In total, 92.7% were treated with MMIs. e
median initial treatment duration was 7 months (IQR
4 - 14.25 months). e median maintenance treatment
duration was 19 months (IQR 12 - 27.25 months). e
median follow-up duration was 32 months (IQR 16 - 55
months) (Table 1).
GD patient baseline and clinical characteristics in the
relapse and remission groups
e relapse rate of GD was 35% aer ATD withdrawal
for one year. We divided patients based on clinical features
into the relapse group (n = 100) and remission group
(n = 186). e baseline and clinical features, including
laboratory results and treatments, are shown in Table 2.
e relapse group had a mean age at diagnosis of 45 ±
15 years, most patients had no TAO (84.4%), however
there were more TAO in relapse group (21%) than in
remission group (15.6%) statistical insignicantly and a
majority had a goiter of approximately 25-60 gm (68.5%),
which was not signicantly dierent from that of the
remission group.
Clinical factors, including male sex (OR 2.01; 95%CI
1.13 – 3.53; p = 0.014) and smoking (OR 6.80; 95%CI 2.15
– 21.46; p = 0.001), were signicantly dierent between
the two groups. ere was more underlying diabetes
mellitus, hypertension, dyslipidemia in relapse group
than in remission group but not signicantly. e mean
serum FT3 and FT4 level at diagnosis in the relapse group
was insignicantly higher than that of remission group.
e mean serum FT4 level before stopping ATDs in the
relapse group (1.17 ± 0.25 ng/dL) was signicantly lower
than that of the remission group (1.25 ± 0.31 ng/dL) (p
= 0.047). e mean dose of MMI before stopping ATDs
in the relapse group (3.4 ± 2.4 mg/day) was signicantly
higher than that in the remission group (2.8 ± 1.7 mg/
day) (p = 0.014). e median duration of remission in
the relapse group was 4 months, in contrast to 1 year
and 11 months in the remission group (Table 2).
Further analyses of the factors expected to inuence
future treatment were conducted. is was partly based
on a study of a predictive model for GD recurrence based
on clinical features or the Graves’ Recurrence Events
Aer erapy (GREAT) score (Table 3). e relapse
group had serum FT4 levels > 2 times the normal ranges
at diagnosis (OR 2.49; 95%CI, 1.30 - 4.77; p = 0.005)
and intervals of the maintenance periods of ATDs < 6
and 9 months, which were signicantly dierent from
those of the GD remission group (OR 0.4; 95%CI, 0.21
- 0.78; p = 0.006 and OR 0.56; 95%CI, 0.31 - 0.99; p =
0.043, respectively). e relapse group tended to have
TSH levels < 1 mIU/L before discontinuation and MMI
doses > 2.5 mg/day before discontinuation in contrast
to those of the remission group, but the dierence was
not statistically signicant.
Other factors considered included age < 40 versus
> 40 years, the thyroid gland abnormality at diagnosis,
serum FT3 three or more times the normal value, and
an ATD maintenance period duration less than or equal
twelve months, but the comparisons between relapse
and remission groups revealed no statistically signicant
dierences.
As shown in Table 4, there were three clinical factors
that predicted the likelihood of GD recurrence, namely,
serum FT3 levels at diagnosis, serum FT4 levels at diagnosis
and serum FT3 levels prior to discontinuation, and
dierences were statistically signicant (p <0.05). It was
found that when serum FT3 at diagnosis was reduced
by 1 pg/mL, there was a chance of reducing GD relapse
by 0.84 (OR 0.84; 95%CI, 0.71 - 0.99; p = 0.036). When
serum FT4 at diagnosis was increased by 1 ng/dL, there
was an increased risk of GD relapse by 2.35 (OR 2.35;
95%CI, 1.28 -4.32; p = 0.006). When serum FT3 levels
before discontinuation increased by 1 pg/mL, there was
an increased risk of relapsed GD by 3.85 (OR 3.85; 95%CI,
1.24 – 11.93; p = 0.019), so serum FT3 was considered
the most important variable. If the serum FT3 value
changed, it might also change the course of the disease.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
454
TABLE 1. Baseline and clinical characteristics, including laboratory results and treatments, of 286 patients with GD.
Characteristics Total, N = 286 (%)
Baseline characteristics
Age at diagnosis (years), min–max 43.5 ± 13.6, 8–80
Sex
Female 221 (77.3)
Male 65 (22.7)
Smoking
Never smoker 269 (94.1)
Smoker 15 (5.2)
Former smoker 2 (0.7)
Coexisting disease
None 138 (48.3)
Hypertension 90 (31.5)
Diabetes mellitus 42 (14.7)
Dyslipidemia 103 (36.0)
Cardiac diseases 11 (3.8)
Cerebrovascular diseases 8 (2.8)
Renal diseases 2 (0.7)
Liver diseases 4 (1.4)
Others 57 (19.9)
Clinical characteristics
Thyroid-associated ophthalmopathy at diagnosis
Absent 236 (82.5)
Lid lag and/or lid retraction 25 (8.7)
Exophthalmos 25 (8.7)
Goiter at diagnosis
Normal (10–20 gm) 83 (29.0)
Mild to moderate (25–60 gm) 196 (68.5)
Large (> 60 gm) 7 (2.5)
Laboratory results and treatments
Mean levels of thyroid function at diagnosis, min–max
FT4 (ng/dL)* 5.38 ± 2.14, 1.16–7.77
FT3 (pg/mL)** 18.30 ± 8.87, 3.40–32.55
TSH (mIU/L) 0.0096 ± 0.023, 0.002–0.300
Mean levels of thyroid function at ATDs withdrawal, min–max
FT4 (ng/dL)
#
1.22 ± 0.29, 0.25–2.94
FT3 (pg/mL)
##
2.88 ± 0.49, 1.21–4.39
TSH (mIU/L)
###
3.235 ± 0.289, 0.032–84.960
Type of ATDs
MMI 265 (92.7)
PTU 21 (7.3)
Median initial duration of treatment (months), IQRs 7, 4–14.25
Median maintenance duration of treatment (months), IQRs 19, 12–27.25
Median total duration of treatment (months), IQRs 27.5, 24–38
Median follow-up duration of treatment (months), IQRs 32, 16–55
Median duration of remission (months), IQRs 14, 5–31
*N = 277, **N = 261,
#
N = 239,
##
N = 202,
###
N = 284 due to missing data or data not obtained
GD, Graves’ disease; ATDs, antithyroid drugs; MMI, methimazole; PTU, propylthiouracil; IQR, interquartile range
Kiatpanabhikul.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
455
Original Article
SMJ
TABLE 2. Baseline and clinical characteristics, including laboratory results and treatments, of the relapse group versus the remission group of patients with GD.
Characteristics Relapse group, Remission group, p-value, odds ratio
n = 100 (%) n = 186 (%) (95% CI)
Baseline characteristics
Age at diagnosis (years), min–max 45.0 ± 15.0, 12–80 42.7 ± 12.7, 8–76
15 or under 2 (2.0) 6 (3.2)
16–25 10 (10.0) 15 (8.1)
26–35 11 (11.0) 27 (14.5)
36–45 30 (30.0) 60 (32.3) 0.187
§
46–55 24 (24.0) 49 (26.3)
56–65 15 (15.0) 23 (12.4)
Over 65 8 (8.0) 6 (3.2)
Sex
Female 69 (69.0) 152 (81.7)
0.014
†
, 2.01 (1.13–3.53)
Male 31 (31.0) 34 (18.3)
Smoking
Never smoker 87 (87.0) 182 (97.8)
Smoker 11 (11.0) 4 (2.2) 0.001
†
, 6.80 (2.15–21.46)
Former smoker 2 (2.0) 0 (0.0)
Coexisting disease
None 42 (42.0) 96 (51.6)
Hypertension 40 (40.0) 50 (26.9)
Diabetes mellitus 22 (22.0) 20 (10.8)
Dyslipidemia 40 (40.0) 63 (33.9)
Cardiac diseases 3 (3.0) 8 (4.3) 0.121
†
, 1.47 (0.90–2.41)
Cerebrovascular diseases 4 (4.0) 4 (2.2)
Renal diseases 2 (2.0) 0 (0.0)
Liver diseases 1 (1.0) 3 (1.6)
Others 24 (24.0) 33 (17.7)
Clinical characteristics
Thyroid-associated ophthalmopathy at diagnosis
Absent 79 (79.0) 157 (84.4)
Lid lag and/or lid retraction 12 (12.0) 13 (7.0) 0.348
†
, 1.44 (0.77–2.68)
Exophthalmos 9 (9.0) 16 (8.6)
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
456
TABLE 2. Baseline and clinical characteristics, including laboratory results and treatments, of the relapse group versus the remission group of patients with GD. (Continue)
Characteristics Relapse group, Remission group, p-value, odds ratio
n = 100 (%) n = 186 (%) (95% CI)
Goiter at diagnosis
Normal (10–20 gm) 28 (28.0) 55 (29.6)
Mild to moderate (25–60 gm) 69 (69.0) 127 (68.3) 0.882
†
, 1.08 (0.63–1.85)
Large (> 60 gm) 3 (3.0) 4 (2.2)
Laboratory results and treatments
Mean levels of thyroid function at diagnosis
FT4 (ng/dL)* 5.66 ± 1.94 5.24 ± 2.24 0.155
‡
FT3 (pg/mL)** 18.79 ± 8.48 18.06 ± 9.08 0.548
‡
TSH (mIU/L) 0.01 ± 0.03 0.01 ± 0.02 0.372
‡
Mean FT3/FT4 ratio at diagnosis (pmol/L) 0.38 ± 0.10 0.41 ± 0.10 0.317
‡
Mean levels of thyroid function at ATD withdrawal
FT4 (ng/dL)
#
1.17 ± 0.25 1.25 ± 0.31 0.047
‡
FT3 (pg/mL)
##
2.94 ± 0.54 2.84 ± 0.47 0.423
‡
TSH (mIU/L)
###
4.45 ± 10.64 2.58 ± 1.83 0.317
‡
Type of ATD
MMI 93 (93.0) 172 (92.5)
0.871
†
, 0.93 (0.36–2.37)
PTU 7 (7.0) 14 (7.5)
Mean dose of ATDs at diagnosis (mg/day)
MMI 37.1 ± 9.7 49.5 ± 10.9 0.962
‡
PTU 158.2 ± 107.9 204.3 ± 89.2 0.219
‡
Mean dose of ATDs before withdrawal (mg/day)
MMI 3.4 ± 2.4 2.8 ± 1.7 0.014
‡
PTU 67.9 ± 42.6 41.1 ± 12.4 0.091
‡
Median initial duration of treatment (months), IQR 8, 5–16 7, 4–13.25 0.249
‡
Median maintenance duration of treatment (months), IQR 19, 8–31 19, 14–25.25 0.804
‡
Median total duration of treatment (months), IQR 28, 24–43.5 27, 24–37 0.445
‡
Median duration of remission (months), IQR 4, 3–7.75 23, 14.75–40.25 <0.001
‡
§
Independent t-test,
†
chi-square test,
‡
Mann-Whitney U-test
*N = 277, **N = 261,
#
N = 239,
##
N = 202,
###
N = 284 due to missing data or data not obtained
GD, Graves’ disease; 95% CI, 95% condence interval; IQR, interquartile range
Kiatpanabhikul.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
457
Original Article
SMJ
TABLE 3. Specic clinical risk factors that are expected to inuence future treatment in the relapse group versus remission group of patients with GD.
Characteristics Relapse group, Remission group, p-value, odds ratio
n = 100 (%) n = 186 (%) (95% CI)
Baseline and clinical characteristics
Age at diagnosis
< 40 vs ≥ 40 years 39 (39.0) vs 61 (61.0) 74 (39.8) vs 112 (60.2) 0.897
†
, 1.03 (0.63–1.70)
Goiter at diagnosis
Normal vs abnormal glands 28 (28.0) vs 72 (72.0) 55 (29.6) vs 131 (70.4) 0.780
†
, 1.08 (0.63–1.85)
Laboratory results and treatments
Level of thyroid function at diagnosis
FT4 < 2 vs ≥ 2 times the normal range 14 (14.6) vs 82 (85.4)* 54 (29.8) vs 127 (70.2)* 0.005
†
, 2.49 (1.30–4.77)
FT3 < 3 vs ≥ 3 times the normal range 23 (26.4) vs 64 (73.6)* 64 (36.8) vs 110 (63.2)* 0.095
†
, 1.62 (0.92–2.86)
Level of thyroid function at ATDs withdrawal
TSH ≤ 1 vs > 1 mIU/L 22 (22.2) vs 77 (77.8) 25 (13.5) vs 160 (86.5) 0.060
†
, 0.55 (0.29–1.03)
Dose of MMI before withdrawal
≤ 2.5 vs > 2.5 mg/day 63 (63.9) vs 37 (37.0) 135 (72.6) vs 51 (27.4) 0.094
†
, 1.56 (0.93–2.61)
Duration of maintenance treatment
≤ 6 vs > 6 months 23 (23.0) vs 77 (77.0) 20 (10.8) vs 166 (89.2) 0.006
†
, 0.40 (0.21–0.78)
≤ 9 vs > 9 months 28 (28.0) vs 72 (72.0) 33 (17.7) vs 153 (82.3) 0.043
†
, 0.56 (0.31–0.99)
≤ 12 vs > 12 months 31 (31.0) vs 69 (69.0) 41 (22.0) vs 145 (78.0) 0.096
†
, 0.63 (0.36–1.86)
*Missing data or data not obtained,
†
chi-square test
GD, Graves’ disease; 95% CI, 95% condence interval; ATDs, antithyroid drugs
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
458
TABLE 3. Demonstrated clinical factors for prediction of relapsed GD by Wald (forward stepwise) logistic regression.
Inuencing clinical risk factors (variables) β S.E. Wald p-value Odds ratio 95% CI for odds ratio
Lower Upper
Constant -4.252 2.510 2.869 0.090 0.014
Age 0.021 0.016 1.689 0.194 1.021 0.989 1.055
Sex -0.136 0.549 0.061 0.805 0.873 0.297 2.562
Smoking 1.116 1.077 1.074 0.300 3.052 0.370 25.187
TAO 0.411 0.721 0.325 0.569 1.508 0.367 6.199
Goiter size -0.543 1.221 0.198 0.656 0.581 0.053 6.360
Serum FT3 level at diagnosis -0.181 0.086 4.399 0.036 0.835 0.705 0.988
Serum FT4 level at diagnosis 0.855 0.310 7.602 0.006 2.352 1.281 4.321
Serum FT3 level before ATDs withdrawal 1.349 0.577 5.465 0.019 3.852 1.243 11.933
Serum FT4 level before ATDs withdrawal -2.141 1.326 2.607 0.106 0.117 0.009 1.581
Serum TSH level before ATDs withdrawal 0.000 0.050 0.000 1.000 1.000 0.907 1.103
Doses of ATDs before withdrawal 0.072 0.064 1.254 0.263 1.075 0.947 1.219
GD, Graves’ disease; 95% CI, 95% condence interval; TAO, yroid-associated ophthalmopathy; ATDs, antithyroid drug
Kiatpanabhikul.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
459
Original Article
SMJ
DISCUSSION
is 10-year retrospective study of 286 rst-diagnosed
ai patients with GD focused on ATD therapy only. e
relapse rate was 35% among patients with GD aer rst-
year ATD withdrawal therapy. Overall, this study had a
median follow-up of 2 years and 8 months, the relapse
group was found to comprise males and smokers, and the
mean MMI doses before discontinuation were greater than
those in the remission group (Table 2). From which the
mean serum FT4 level before stopping ATDs in the relapse
group was signicantly lower than that of the remission
group while the mean serum FT3 and FT4 level at diagnosis
in the relapse group was higher than that of remission
group but insignicantly (Table 2), probably due to T3
toxicosis was predominated in the relapse group. When
subgroup analysis for each factor that was expected to
aect future treatment was performed, the relapse group
had a serum FT4 level at diagnosis ≥ 2 times the normal
range, and the maintenance treatment duration, less than
6 and 9 months, was signicantly dierent from that of
the remission group. However, there was no signicant
dierence in use of this treatment for up to 12 months
compared with more than 12 months. Additionally, the
relapse group tended to have serum TSH levels ≤ 1 mIU/L
and doses of MMI before discontinuation of more than
2.5 mg/day compared with the remission group, but
the dierence was not statistically signicant (Table 3).
When important clinical factors were used in relation to
GD relapse during the rst year aer discontinuation,
elevations in serum FT4 levels at diagnosis of 1 ng/dL
and elevations in serum FT3 levels before discontinuation
of 1 pg/mL increased the risk of GD recurrence by 2.35
and 3.85, respectively (Table 4).
Compared to previous studies in Asia, the relapse
rate was 32 - 52% after one to two years of ATD
discontinuation.
13,14,23,26,27
e clinical factors related to
relapse were younger age
27
, large thyroid gland
14,27
, TAO
aer drug treatment
14
, higher serum FT3 and FT4 levels
27
,
FT3/FT4 (pmol/l) ratio
14
, continued TSH suppression
levels
14
, TRAb levels
14
and a duration of ATD use in the
maintenance period of less than 6 months.
26
e clinical
factors related to remission were older age (45.6 ± 10.3
years)
14
and a minimum dose of MMI of 5 mg every
other day for at least 19 months.
13
Compared with rates
from previous studies in Europe and Africa, the relapse
rate was 30 - 49% aer one year of discontinuation. ,,
e clinical factors related to relapse were < 40 years of
age; smoking; TAO; thyroid size by ultrasound; serum
FT4, FT3, and TRAb levels; and total T4 levels ≥ 2 times
the upper normal range.
30
e clinical factors related to
remission were female sex, nonsmoking status, no TAO
and duration of treatment of more than 2 years.
28
is study
showed a relapse rate similar to that of previous studies
and conrmed that some clinical risk factors inuence
relapse in both Asian and Western countries. However,
we found that ATDs should be taken as maintenance
treatment for more than 9 months and a trend toward
having TSH > 1 mIU/L before discontinuation medications
to reduce relapse.
In ailand, a recent study showed the relapse rate
37% within rst year of ATD withdrawal as well as this
study (35%). However, the factors associated with early
relapse in that study showed patients 40 years of age or
less and the highest quartile of serum T3 level at the
time of diagnosis
23
, in contrast with this study showed
no statistically signicant of those factors.
When statistically signicant clinical factors associated
with relapsed GD aer one year of ATD withdrawal were
assessed in the context of previous studies, we made
several observations. First, male sex: many previous
studies have shown that male sex was a factor associated
with the GD relapse. One study showed that men usually
have a larger thyroid gland and more of a family history
of autoimmune thyroid disease than women.
31
Another
study found that men have a larger thyroid gland and still
smoked more than women.
32
is was consistent with
the results of our study showing that males smoked more
cigarettes (24.6%) than females (0.5%), but there was no
dierence in thyroid size in our study. Second, smoking:
in addition to being a major risk factor for developing
GD
33
, studies comparing smokers to nonsmokers found
that elimination of serum FT4, serum FT3 and TRAb
among smokers was slow, causing antibodies to be present
longer, leading to disease progression.
34
is support the
results of our study showing that smoking is associated
with relapse. ird, serum FT4 levels at diagnosis ≥ 2
times the normal range: a previous study showed a positive
correlation between FT4 and TRAb levels, indicating
that higher FT4 levels were also associated with a high
TRAb level
35
, thus promoting disease recurrence. is
was consistent with our study results. Fourth, higher
mean serum FT3 levels before discontinuation: a previous
study found that the remission phase is related to normal
thyroid hormone levels regardless of the doses received.
36
is supports the results of our study, which show that
more of such factors are associated with a higher risk of
relapsed GD. Fih, administration of more than 2.5 mg/
day MMI prior to discontinuation and a maintenance
treatment duration less than 6-9 months: these ndings
are consistent with a recent study showing that a 2.5-mg
daily dose of MMI for periods of less than 6 months is
associated with relapse
28
and that treatment with MMI
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
460
5 mg every other day for no longer than 6 months has
a higher rate of relapse.
13
According to the American yroid Association
1
and
the European yroid Association
37
, it is recommended
that the total dose of ATDs is given for approximately
12 to 18 months and discontinued aer the TSH and
TRAb levels are normal; the consequent remission
rate is 20-30%. Based on studies in 1994
38
and 2004
39
,
treatment with ATDs for 18 months versus 6 months has
lower rates of recurrence. However, in a meta-analysis
published in 2005
40
, the remission rate did not increase
aer taking ATDs for more than 18 months. Consistent
with the results of this study, the median total duration
of ATDs was 28 months, and no dierence was found
between the relapse and remission groups. However, this
study provides additional data showing that low-dose
ATDs should be taken for 9-12 months before stopping
medications to reduce the relapse rate.
Benets
1. Knowledge of the relapse rate and the important
clinical factors for GD relapse in ai patients.
2. Ability to provide ATD treatment with appropriate
dosages and durations of administration to minimize
the likelihood of GD relapse in ai patients.
3. Usefulness as the basis for planning suitable guidelines
in the management of GD in ailand that lead to a
reduction in costs associated with measuring TRAb levels
before discontinuation of ATDs for patients among whom
these measurements are unnecessary, such as females,
nonsmokers, those with serum FT4 levels at diagnosis
< 2 times the normal range, and those with low serum
FT3 levels before discontinuation.
Limitations
is is a 10-year retrospective study using data
from past medical records. Patients were examined by
physical examination, which might lead to inaccuracies
regarding certain clinical features. Some information
was considered missing data, namely, in cases in which
data was lost or cases that were not fully investigated,
with data including serum FT4, FT3 and TSH levels at
diagnosis and before ATD withdrawal. However, these
cases were rare and comprised missing data for only
select hormones in a few patients but not all hormones
at once. erefore, only the information contained in
the dataset was analyzed.
Suggestions
Further studies should include a prospective trial with
physical examination by any internist or endocrinologist
and set the protocol for the laboratory test. For further
reliability, a Hertel ophthalmometer and ultrasound may
be used for the measurement of thyroid gland size and
additional TRAb levels to help provide more accurate
data.
ACKNOWLEDGMENTS
Thank you, Charoenkrung Pracharak Hospital
Foundation, for providing partial funding.
REFERENCES
1. Ross DS, Burch HB, Cooper DS, Greenlee MC, Laurberg P,
Maia AL, et al. 2016 American yroid Association Guidelines
for Diagnosis and Management of Hyperthyroidism and Other
Causes of yrotoxicosis. yroid 2016;26:1343-421
2. Burch HB, Burman KD, Cooper DS. A 2011 survey of clinical
practice patterns in the management of Graves’ disease. J Clin
Endocrinol Metab 2012;97:4549-58.
3. Sriphrapradang C. Diagnosis and management of Graves’
disease in ailand: A Survey of current practice. Journal of
yroid Research, vol.2020, Article ID 8175712, 8 pages.
4. Bartalena L, Burch HB, Burman KD, Kahaly GJ. A 2013 European
survey of clinical practice patterns in the management of
Graves’ disease. Clin Endocrinol (Oxf) 2016;84:115-20.
5. Burch HB, Cooper DS. Antithyroid drugs therapy: 70 years
later. Eur J Endocrinol 2018;179:R261-74.
6. Kahaly GJ, Bartalena L, Hegedus L, Leenhardt L, Poppe K,
Pearce SH. 2018 European yroid Association guideline
for the management of Graves' hyperthyroidism. Eur yroid
J 2018;7:167-86.
7. Laurberg P. Remission of Graves' disease during anti-thyroid
drug therapy. Time to reconsider the mechanism? Eur J
Endocrinol 2006;155:783-6.
8. Klein I, Becker DV, Levey GS. Treatment of hyperthyroid
disease. Ann Intern Med 1994;121:281-8.
9. Mazza E, Carlini M, Flecchia D, Blatto A, Zuccarini O, Gamba S,
et al. Long-term follow-up of patients with hyperthyroidism
due to Graves' disease treated with methimazole. Comparison
of usual treatment schedule with drug discontinuation vs
continuous treatment with low methimazole doses: a retrospective
study. J Endocrinol Invest 2008;31:866-72.
10. Mohlin E, Filipsson Nystrom H, Eliasson M. Long-term prognosis
aer medical treatment of Graves' disease in a northern Swedish
population 2000-2010. Eur J Endocrinol 2014;170:419-27.
11. Konishi T, Okamoto Y, Ueda M, Fukuda Y, Harusato I,
Tsukamoto Y, et al. Drug discontinuation aer treatment with
minimum maintenance dose of an antithyroid drug in Graves'
disease: a retrospective study on eects of treatment duration
with minimum maintenance dose on lasting remission. Endocr
J 2011;58:95-100.
12. Liu L, Lu H, Liu Y, Liu C, Xun C. Predicting relapse of Graves'
disease following treatment with antithyroid drugs. Exp er
Med 2016;11:1453-8.
13. Park S, Song E, Oh H-S, Kim M, Jeon MJ, Kim WG, et al.
When should antithyroid drug therapy to reduce the relapse
rate of hyperthyroidism in Graves’ disease be discontinued?
Endocrine 2019;65:348-56.
14. Anagnostics P, Adamidou F, Polyzos SA, Katergari S, Karathanasi
Kiatpanabhikul.
Volume 73, No.7: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
461
Original Article
SMJ
E, Zouli C, et al. Predictors of long-term remission in patients
with Graves' disease: a single center experience. Endocrine
2013;44:448-53.
15. Lawgaly SA, Kallousa HA, Gerryo S. Predictors of relapse in
Graves' hyperthyroidism aer treatment with antithyroid
drugs. Ibnosina J Med Biomed Sci 2018;10:205-8.
16. Abraham P, Avenell A, Park CM, Watson WA, Bevan JS. A
systematic review of drug therapy for Graves' hyperthyroidism.
Eur J Endocrinol 2005;153:489-98.
17. Kimball LE, Kulinskaya E, Brown B, Johnston C, Farid NR.
Does smoking increase relapse rates in Graves' disease? J
Endocrinol Invest 2002;25:152-7.
18. Allahabadia A, Daykin J, Holder RL, Sheppard MC, Gough SC,
Franklyn JA. Age and gender predict the outcome of treatment
for Graves' hyperthyroidism. J Clin Endocrinol Metab 2000;85:
1038-42.
19. Nedrebo BG, Holm PI, Uhlving S, Sorheim JI, Skeie S, Eide
GE, et al. Predictors of outcome and comparison of dierent
drug regimens for the prevention of relapse in patients with
Graves' disease. Eur J Endocrinol 2002;147:583-9.
20. Orunesu E, Bagnasco M, Salmaso C, Altrinetti V, Bernasconi
D, Monte PD, et al. Use of an articial neural network to
predict Graves' disease outcome within 2 years of drug withdrawal.
Eur J Clin Invest 2004;34:210-7.
21. Bolaños F, González-Ortiz M, Durón H, Sánchez C. Remission
of Graves' hyperthyroidism treated with methimazole. Rev
Invest Clin 2002;54:307-10.
22. Harinsoot Somnuke P, Pusuwan P, Likitmaskul S, Santiprabhob
S, Sawathiparnich P. Treatment outcome of Graves’ disease
in ai children. J Med Assoc ai 2007;90:1815-20.
23. ewjitcharoen Y, Karndumri K, Chatchomchuan W, Porramatikul
S, Krittiyawong S, Wanothayaroj E. et al. Serum T3 level and
duration of minimum maintenance dose therapy predict relapse
in methimazole-treated Graves’ disease. J Endocr Soc 2021;5:
1-11.
24. Davies TF LP. yrotoxicosis. In: Kronenberg HM, Melmed S,
Polonsky KS, Larsen PR. 11th Williams textbook of endocrinology:
Philadephia: Saunders, 2008.p.342-55.
25. Vos XG., Endert E., Zwinderman AH., Tijssen JGP., Wiersinga
WM. Predicting the risk of recurrence before the start of
antithyroid drug therapy in patients with Graves' hyperthyroidism.
J Clin Endocrinol Metab 2016;101:1381-9.
26. Park S, Song E, Oh H-S, Kim M, Jeon MJ, Kim WG, et al.
When should antithyroid drug therapy to reduce the relapse
rate of hyperthyroidism in Graves’ disease be discontinued?
Endocrine 2019;65:348-56.
27. Shi M, Sheng R, Hu Y, Jiang L, Wang Z, Cui D. Risk factors
for the relapse of Graves’ disease treated with antithyroid drugs:
A systematic review and meta-analysis. Clin er 2020;42:662-
75.e4.
28. Anagnostics P, Adamidou F, Polyzos SA, Katergari S, Karathanasi
E, Zouli C, et al. Predictors of long-term remission in patients
with Graves' disease: a single center experience. Endocrine
2013;44:448-53.
29. Struja T, Fehlberg H, Kutz A, Guebelin L, Degen C, Mueller B,
et al. Can we predict relapse in Graves’disease? Resullts from
a systematic review and meta-analysis. Eur J Endocrinol
2017;176:87-97.
30. Lawgaly SA, Kallousa HA, Gerryo S. Predictors of relapse in
Graves' hyperthyroidism aer treatment with antithyroid
drugs. Ibnosina J Med Biomed Sci 2018;10:205-8.
31. Magri F, Zerbini F, Gaiti M, Capelli V, Ragni A, Rotondi M,
et al. Gender inuences the clinical presentation and long-term
outcome of Graves' disease. Endocr Pract 2016;22:1336-42.
32. Zuhur SS, Yildiz I, Altuntas Y, Bayraktaroglu T, Erol S, Sahin
S, et al. e eect of gender on response to antithyroid drugs
and risk of relapse aer discontinuation of the antithyroid
drugs in patients with Graves' hyperthyroidism: a multicentre
study. Endokrynol Pol 2020;71:207-212.
33. Brix TH, Hansen PS, Kyvik KO, Hegedus L. Cigarette smoking
and risk of clinically overt thyroid disease: A population-based
twin case-control study. Arch Intern Med 2000;160:661-6.
34. Nyirenda MJ, Taylor PN, Stoddart M, Beckett GJ, To AD.
yroid-stimulating hormone-receptor antibody and thyroid
hormone concentrations in smokers vs nonsmokers with
Graves' disease treated with carbimazole. JAMA 2009;301:
162-4.
35. Bano A, Gan E, Addison C, Narayanan K, Weaver JU, Tsatlidis
V, et al. Age may inuence the impact of TRAbs on thyroid
function and relapse-risk in patients with Graves' disease. J
Clin Endocrinol Metab 2019;104:1378-85.
36. Laurberg P. Remission of Graves' disease during anti-thyroid
drug therapy. Time to reconsider the mechanism? Eur J
Endocrinol 2006;155:783-6.
37. Kahaly GJ, Bartalena L, Hegedus L, Leenhardt L, Poppe K,
Pearce SH. 2018 European yroid Association guideline
for the management of Graves' hyperthyroidism. Eur yroid
J 2018;7:167-86.
38. Klein I, Becker DV, Levey GS. Treatment of hyperthyroid
disease. Ann Intern Med 1994;121:281-8.
39. Abraham P, Avenell A, Watson WA, Park CM, Bevan JS.
Antithyroid drug regimen for treating Graves' hyperthyroidism.
Cochrane Database Syst Rev 2004 (2);CD003420.
40. Abraham P, Avenell A, Park CM, Watson WA, Bevan JS. A
systematic review of drug therapy for Graves' hyperthyroidism.
Eur J Endocrinol 2005;153:489-98.
