Treatment of patients with Graves’ disease in Sweden compared to international surveys of an ‘index patient’

Abstract Introduction The treatment strategies for a 42‐year‐old female index patient with moderate Graves’ disease (GD) vary according to several international surveys. The important question whether surveys of treatment preferences in theoretical patient cases also match how real patients are treated has not yet been addressed. Materials and Methods From a Swedish cohort of 1186 GD patients (TT‐12 cohort), 27 women were identified using the same criteria as from the index patient surveys from the European and American Thyroid Associations. This ‘index patient cohort’ was age 40–45, otherwise healthy female, with two children and uncomplicated GD. The applied first‐line treatment of the patients in the index cohort, together with its variations, was compared with the treatment preferences according to international surveys. A comparison with the TT‐12 cohort was also performed. Results In the ‘Index cohort’, 77.8% were treated with antithyroid drugs (ATD), and 22.2% were treated with radioiodine (131I). This preference for ATD is in line with most countries/regions, with the exception of USA and the Middle East/North Africa, where 131I was preferred. The distribution of treatment in the TT‐12 cohort did not significantly differ from the index cohort. ATD was the preferred treatment in male and young (age 19–22) patients, as was RAI in old (age 69–73) patients. The age‐related, but not the gender‐related, cases differed significantly from the entire TT‐12 cohort. Conclusion The treatment choice in an index patient in Sweden seems in line with European practice, where ATD is the preferred first choice. This differs compared to US and North African survey intentions, where 131I is more often used. Age more than gender influences the treatment choice of GD patients. This is, to our best knowledge, the first time an index patient from ‘real life’ has been presented and compared to treatment preferences of international thyroid association surveys.


| INTRODUC TI ON
Graves' disease (GD) is a common autoimmune disease with an incidence of 6-93 per 100,000 population. [1][2][3] It has three main treatment options: antithyroid drugs (ATD), radioactive iodine ( 131 I) or surgery. GD mostly affects women at 30-60 years of age. 2 The treatment preferences of a hypothetical 'index patient' vary internationally, as illustrated by several authors. [4][5][6][7][8][9] Several factors influence the choice of treatment, such as age, sex, the patient's preferences, planned pregnancy, smoking, eye symptoms and costeffectiveness 10,11 ; additionally, local traditions and access to treatments can influence the choice of treatment. These issues may explain why different treatment strategies are used for an 'index case' of GD. [4][5][6][7][8][9] Several clinical studies showed that ATD is a common first choice of treatment of GD worldwide, and treatment strategies imply a higher use of ATD in South East Asia and in Europe than in the USA, where 131 I is predominantly used. [12][13][14][15][16] Surgery, on the other side, is less frequently used as a first-line treatment option, but it is more common in France (6.1%) and Sweden (4.6%) than in the USA, Taiwan and South Korea (2.0-2.9%). [12][13][14][15][16] These clinical studies do not take into account the previously  (Table 1)
The important question, whether a survey of treatment preferences based on theoretical patient cases also matches how real patients, similar to an 'index patient', is treated today, has not been elucidated. In this study, we therefore present the actual treatment choices in Sweden for patients who meet the criteria for an index patient used in international surveys.
We aimed to study the following: 1. The choice of treatment of consecutively registered patients, who fulfilled the criteria for the index patient and its variations, in a Swedish national cohort.

| Subjects
The patients were collected from an original Swedish cohort (the TT-12 cohort) consisting of 1186 adult GD patients followed up 6-10 years after diagnosis from 2003-2005. 16 The initial cohort was collected over three years from hospitals covering approximately 40% of the Swedish population. With a loss of approximately 46% young (age [19][20][21][22] patients, as was RAI in old (age 69-73) patients. The age-related, but not the gender-related, cases differed significantly from the entire TT-12 cohort.

Conclusion: The treatment choice in an index patient in Sweden seems in line with
European practice, where ATD is the preferred first choice. This differs compared to US and North African survey intentions, where 131 I is more often used. Age more than gender influences the treatment choice of GD patients. This is, to our best knowledge, the first time an index patient from 'real life' has been presented and compared to treatment preferences of international thyroid association surveys.

K E Y W O R D S
antithyroid drugs, Graves' disease, hyperthyroidism, index patient, international surveys, longterm follow-up, radioactive iodine, thyroidectomy, treatment options to follow-up, this cohort corresponds to approximately 65% (40% of population x 3 years x 54%) of the total yearly incidence of GD in Sweden.

| Definition of the index patient
Patients as similar as possible to the index case of a 42-to 43-year-old otherwise healthy woman, with two children, and uncomplicated GD, were extracted from the database with the same criteria used as in the previous ETA and ATA surveys 4-7 ( Table 1). The definitions of the index patient have, in small ways, changed over time and between the different surveys. For example, the age of the index patient changed from the first study to more recent studies, from 43-year-old to 42-year-old patient. Where such differences occur, the original definition from the first ETA survey 4 has been used.
Taken together, there are 10 variations of the index case, and we collected data for four sub-cohorts of actual patients fulfilling these criteria.
Since most patients in Sweden have GD without a large goitre, we presume all selected patients had a goitre size less than 40-50 g. The mean thyroid weight in GD of cases undergoing surgery in Sweden is 30 g. 26 There were no records of symptoms or the age of the woman's two children, and therefore, these criteria had to be excluded.
Additionally, when selecting the patients, only five 42-year-old and three 43-year-old female patients were found. The age criterion was extended to include all female patients of close ages, that is, 40-45 years old, which resulted in 27 patients. This adjustment is considered appropriate, because the age range in the surveys was arbitrarily selected, and with a small variation in age definition, the survey responses would probably not change significantly.
Selected patient groups: male, young, old and patients with a relapse.
We also selected young, old and male patients and patients with relapse for separate analyses. There was also, in the later surveys, a case with a woman who wished to get pregnant. Since there were no questions about pregnancy wishes in our questionnaire, these patients were assumed to be equivalent to the young patients, supposing all young women may have a wish for pregnancy.
TT-12 old female cohort: The old index patient in previous investigations was defined as a 71-year-old woman, but since there were only five patients fulfilling this criterion in our study, we expanded the group to an age 69-73, resulting in 33 patients.
TT-12 relapse female cohort: Concerning the cases with relapse after ATD and surgery, we can only report on the relapse after ATD since there were nine female patients between 40-45 years old who later had a relapse after primary ATD treatment in this cohort, but none after surgery.

| Statistical analysis
To calculate significant differences between the different groups, which consist of different types of data, 95% confidence intervals (CIs) of proportions were calculated for our index groups. If, then, the entire original cohort or any of the results from the IPQ surveys falls outside that range, they are considered to have a significant difference.
The IPQ surveys all have a low proportion of surgically and conservatively (symptomatic treatment with beta-blockers) treated patients, and this is also true for the whole TT-12 cohort, so the only real difference is in the ratio between ATD and 131 I. 4

| Ethics
This study was approved, as a part of the TT12 project, by the Regional Ethics Committee in Uppsala (Dnr 2012/035, 2012 April 4).
The study was performed according to the Declaration of Helsinki.
There was no significant difference compared to the entire TT-12 cohort, of which 65.3% received ATD and 27.3% 131 I treatment (shown in Figure 1). A comparison between the TT-12 index patient cohort and the IPQ surveys can be seen in Figure 2. ATD was used at a similar frequency compared to the TT-12 index patient cohort in most countries/regions, with the exception of the USA and Middle East/ North Africa, where 131 I was preferred.

| Special patient groups in the Swedish cohort
ATD treatment was the most common (67.7% CI 51.2-84.2%) treatment in the 31 male patients aged 40-45 years, with 131 I as the second most common treatment (29.0%). No male patient received surgical treatment, but 3.2% of the patients were treated conservatively with beta-blockers. This male group did not significantly differ from the entire TT-12 cohort (shown in Figure 1). This is not significantly different from most IPQ surveys, with the exception of the USA and Spain.
All 13 patients aged 19-22 years were treated with ATD (100% CI 100-100%). This differs significantly from the entire TT-12 cohort (shown in Figure 1) and all IPQ surveys, except Spain (shown in Figure 3).  Figure 1). The TT-12 old female patient adhered to the preferences of Europe and South America, but not to surveys from Japan or the USA (shown in Figure 3).
In the nine patients who had a relapse after ATD treatment, the secondary treatments were equally divided (33.3% CI 2.5-64.1%) between ATD, 131 I and surgery. This distribution was significantly different from the entire TT-12 cohort (shown in Figure 1). This is not significantly different compared to most IPQ surveys, with the exception of the USA.  veys. 4,6,8,9,[12][13][14]16,27 In the end, there is a difference between studying constructed cases and treated populations; 'fiction' does differ from reality. IPQ surveys may reflect the treatment intention of the medical community, but to study what treatment patients receive is the only way is to conduct clinical studies. The differences between countries, as illustrated in the IPQ surveys, are possibly due to different circumstances in both the community and the health-financial system. It may, to some extent, also illustrate a difference in the perception of cure between different countries/regions.  Zealand (1991, 2014), a trend of using more ATD can be seen. 4,6,8,9,17,29 Studies showing that RAI can increase the risk of Graves' orbitopathy may be one explanation for this trend away from RAI. [30][31][32][33] Another pattern may be a growing popularity in the use of long-term ATD therapy. 11,28 The same pattern seems to hold when looking at a 42-year-old male, where most countries/ regions prefer ATD before 131 I. 4,5,18,19,21 This is also in line with the results of the TT-12 male cohort. The US centres predominantly uses 131 I for the same patients. 6 Perhaps it is a step towards a more internationally aligned treatment strategy.

| Other clinical studies
Antithyroid drugs is the main choice for the TT-12 young (19-22 years old) female patient, as it is in most countries/ regions, [4][5][6]8,9,[18][19][20][21][22]24,25 but an interesting trend in these patients is that surgery has gained popularity in some of these cases, for example, Europe and the USA show a 20-30% operation preference of patients in this age group. 8,9 The change over time could be attributed to the difference in the variations of the index cases introduced over time. In the later studies, both the young age and the desire for pregnancy are taken into account, as opposed to only the young age in the previous studies. Since, clinically, one cannot ignore a potential wish for pregnancy in young women, this point is probably moot. This also sheds light on another flaw in the index patient, as women today may have a reasonable wish for pregnancy up to 35-45 years of age. 34 When comparing the treatment of older patients (69-73 years old), most countries, including the TT-12 old female cohort, prefer 131 I rather than ATD, and only Spain (in 1997, but not in 1987) and Japan use ATD as the primary treatment. [4][5][6][18][19][20][21][22] Second-line treatment after relapse after ATD treatment is predominately 131 I in Europe, USA and South America, which is not significantly different from the TT-12 relapse female cohort. However, in contrast, a second ATD treatment is more common in Japan. [4][5][6]18,19,21,22 This is a striking difference. Japan has a history of aversion to radioactivity, given the two atomic bombs 20 and, most recently, a nuclear accident.
Since both the old and relapse case is not reiterated in later surveys, these results are old (before year 2000) and much has happened to treatment strategies since then. Therefore, the results from these IPQs are not up to date, but the comparison with patients from This can also be seen when looking at other clinical studies. One of the most interesting findings is South Korea, where both an IPQ survey and a clinical study have been conducted. 13,20 The South Korean IPQ survey shows a 20% significantly lower preference for ATD compared to what the patients received in the correlating clinical study. This could depend on the differences in the composition of the cohort and the index criteria, like the fact that 25% of the South Korean cohort has TAO. 13 The same discrepancy between clinical studies and IPQ surveys can also be seen in America where only 16% of patients received ATD in an American clinical study 14 compared to 30% and 54% in the early and late ATA surveys. 6,8 These studies, like many population studies, are, however, single centre studies with a relatively more homogenous treatment practice than in the IPQ surveys.
Nevertheless, the same pattern can still be seen when comparing IPQ surveys with national multicentre studies in Europe. There are differences in the preference for ATD between the ETA surveys 4,9 (77% to 83.8% preference for ATD) in comparison with national multicentre studies both in French (91% preference for ATD) 12 and Sweden (65% preference for ATD). 16 Index patient questionnaire surveys span, not only over countries, but sometimes continents, making the comparisons even with nationwide studies difficult. A student report, in form of a IPQ survey, from Sweden in 2005 (not published) shows regional differences in treatment strategies, with a range of ATD preference in an index patient from about 15%-100%. This is also corroborated in the precursor of the TT-12 cohort where regional differences in treatment can be seen, 1 making even the comparison of national IPQ surveys and populations studies hard. At the end, since the validity of the IPQ surveys is in doubt, a proper validation of the IPQ surveys should be done.
The small size of the selected groups provides a wide confidence interval, which makes the results more uncertain.
Nevertheless, differences between the groups, the IPQ surveys and the entire TT-12 cohort, can still be seen. The database also did not contain all of the variables for the selection, so the groups may not be quite the same as the index cases. Another limitation is that treatment strategies have changed since 2003-2005, when patients received their first treatment in this study. They are, however, compared with physicians' treatment preferences from a 25-year perspective, where these years are in the centre. This is both a strength and a limitation. No survey is conducted exactly at the same time as this study, which complicates comparisons. However, most surveys were conducted less than 10 years before or after this study, which means that comparisons could be viewed as a type of mean.
The main strength of this investigation is the large database that makes it possible to select patients according to these criteria and obtain analysable groups. Even though the index patient has been described many times in the literature, this is, to our best knowledge, the first time an index patient from 'real life' has been presented and compared to the treatment preferences of IPQ surveys.

| S TATEMENT OF E THI C S
The study was approved, as a part of the TT12 project, by the Regional Ethics Committee in Uppsala (Dnr 2012/035, 2012 April 4). The study was performed according to the Declaration of Helsinki.

AUTH O R CO NTR I B UTI O N S
Gabriel Sjölin has made a substantial contribution to the design, acquisition, analysis and interpretation of data, as well as drafting the work. Kristina Byström has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Mats Holmberg has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Ove Törring has made a substantial contribution to the conception and design of the project, acquisition and interpretation of data, as well as revising it critically for important intellectual content. Selwan Khamisi has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Jan Calissendorff has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Mikael Lantz has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Bengt Hallengren has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Helena Filipsson Nyström has made a substantial contribution to the acquisition and interpretation of data, as well as revising it critically for important intellectual content. Tereza Planck has made a substantial contribution to the interpretation of data, as well as revising it critically for important intellectual content. Göran Wallin has made a substantial contribution to the design, acquisition and interpretation of data, as well as revising it critically for important intellectual content. All approve the final version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request due to privacy/ethical restrictions.