Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
0013-7227/03/$15.00/0 Printed in U.S.A. The Journal of Clinical Endocrinology & Metabolism 88(5):2100 –2105 Copyright © 2003 by The Endocrine Society doi: 10.1210/jc.2002-021799 Clinical Features and Outcome of Subacute Thyroiditis in an Incidence Cohort: Olmsted County, Minnesota, Study VAHAB FATOURECHI, JAROSLAW P. ANISZEWSKI, GUITI Z. EGHBALI FATOURECHI, ELIZABETH J. ATKINSON, AND STEVEN J. JACOBSEN Division of Endocrinology, Diabetes, Metabolism, Nutrition and Internal Medicine (V.F., J.P.A., G.Z.E.F.), Division of Biostatistics (E.J.A.) and Division of Epidemiology (S.J.J.), Mayo Clinic, Rochester, Minnesota 55905 Subacute thyroiditis (SAT), or granulomatous thyroiditis, is an inflammatory thyroid condition associated with pain and systemic symptoms. Few community studies are available. We studied the 160 patients with SAT in Olmsted County, Minnesota, seen between January 1, 1960, and December 30, 1997. Subjects were identified through the medical diagnostic index of the Rochester Epidemiology Project. The overall ageand sex-adjusted incidence from 1960 through 1997 was 4.9 cases per 100,000/yr. In the most recent 28-yr period (1970 – 1997), 94 patients were identified. In this group, pain was the presenting symptom in 96%. SAT recurred in 4% of the patients 6 –21 yr after the initial episode. Corticosteroid therapy was given to 36%. Early-onset hypothyroidism occurred both in S UBACUTE THYROIDITIS (SAT) (also called de Quervain thyroiditis or granulomatous thyroiditis) is a selflimited, possibly viral, inflammatory thyroid disorder usually associated with thyroid pain and systemic symptoms (1–3). Clinical features of SAT are well known and include thyroid pain with symptoms of hyperthyroidism, suppressed levels of TSH, low thyroid uptake of radioactive iodine, and elevated erythrocyte sedimentation rate (4, 5). Diagnosis is based on clinical and laboratory data. Tissue diagnosis is rarely needed (4, 6, 7). Although the clinical features and outcome have been described in many nonpopulation-based studies (8 –11), community and incidence cohort studies are scarce (12, 13). The epidemiology of SAT from 1960 –1967 in Olmsted County, Minnesota, was previously reported (13). We used the database of the Rochester Epidemiology Project to study SAT and its outcome in the Olmsted County community from 1970 through 1997. Subjects and Methods Study setting This study was carried out within the boundaries of Olmsted County, Minnesota. Epidemiological research is possible in Olmsted County because the county is relatively isolated from other urban centers and because nearly all medical care is delivered to local residents by a small number of providers. With the exception of a higher proportion of the working population employed in the health care industry, the characteristics of the population of Olmsted County are similar to those of United States whites (14). Mayo Clinic provides approximately half of the primary care and nearly all of the specialty care for the community. Abbreviations: NSAID, Nonsteroidal anti-inflammatory drug; SAT, subacute thyroiditis. patients receiving corticosteroid therapy (29%) and in those not receiving corticosteroid therapy (37%). At latest follow-up, significantly more patients who had received corticosteroid therapy had a diagnosis of hypothyroidism than the group without corticosteroid therapy (25% vs. 10%, P < 0.05; overall rate of hypothyroidism, 15%). Early transient hypothyroidism is common in SAT. Permanent hypothyroidism is less common, and only 15% of the patients are receiving T4 therapy after 28 yr of follow-up. Symptomatic relief is achieved with corticosteroid therapy, but such therapy does not prevent early- and late-onset thyroid dysfunction. (J Clin Endocrinol Metab 88: 2100 –2105, 2003) Olmsted Medical Center and its affiliated hospital, along with Mayo Clinic, provide comprehensive care for the region in every discipline. The epidemiological potential in the community is enhanced by the fact that each provider uses a comprehensive medical record system, whereby all data about an individual are assembled in one place. Thus, the details of every inpatient and outpatient encounter, including visits to the emergency department, laboratory results, pathology reports, and correspondence, can be accessed. Moreover, Mayo Clinic has indexed all diagnoses and procedures documented in the medical records since the early 1900s. Since 1966, similar indexes have been developed for nonMayo Clinic providers under the aegis of the Rochester Epidemiology Project. The result is the linkage of medical records from all sources of medical care used by the Olmsted County population. This linkage system, therefore, constitutes a unique opportunity to ensure complete ascertainment of disease occurrence and clinical outcomes in a geographically defined population. Study cohort Data from the medical diagnostic index of Mayo Clinic and the Rochester Epidemiology Project were used to identify a cohort of patients with SAT, among residents of Olmsted County, during a 38-yr interval (January 1, 1960, through December 30, 1997). During the study period, the population of Olmsted County grew from approximately 75,000 to 100,000. The inclusion criteria for diagnosis of SAT were any one of the following: 1) pathological diagnosis (surgical specimen or biopsy); 2) painful thyroid and either suppressed uptake of iodine 131 or elevated erythrocyte sedimentation rate, or both; or 3) unilateral thyroid pain with abnormally low uptake of iodine 131 and elevated erythrocyte sedimentation rate. Data were obtained by review of each patient’s medical records and follow-up information. Inpatient and outpatient records from all providers in the county were included. For the present report, demographic clinical and laboratory data and information related to treatment patterns and follow-up thyroid function were used. Follow-up was completed in 1997 by review of records and telephone interview when needed. A total of 160 patients were identified in the community from 1960 through 1997 and 94 patients from 1970 through 1997. The insti- 2100 Fatourechi et al. • Subacute Thyroiditis J Clin Endocrinol Metab, May 2003, 88(5):2100 –2105 2101 tutional review boards of Mayo Foundation and Olmsted Medical Center approved the protocol. Laboratory methods At the beginning of the study period, serum total T4 was measured with a competitive protein-binding assay; beginning in 1976, an RIA was used; and beginning in 1990, a commercial chemiluminescence system was used. Serum TSH was measured initially with a first-generation RIA and later, starting in 1986, with a third-generation sensitive RIA. In 1989, a double-antibody chemiluminescent assay was introduced, which used a modified automated chemiluminescence system (ACS: 180; Bayer Corp. Diagnostics, Tarrytown, NY) and modified ACS: 180 TSH reagents. Serum total triiodothyronine was measured initially with an RIA and, in recent years, with an automated chemiluminescent immunoassay (ACS: 180; Bayer Corp. Diagnostics). Serum free T4 was measured with a dialysis method, using an RIA system, starting in 1974 and, more recently, with a commercial system (AxSYM analyzer; Abbott Laboratories, Abbott Park, IL). Antimicrosomal antibody was measured with a microtitration assay based on agglutination of gelatin particles (SeraTek; Fujirebio Inc., Tokyo, Japan). Serum thyroglobulin was measured with a double-antibody chemiluminescence assay. Other than a change in the lower limit of detection of serum TSH, despite some changes in assay methodology, the reference ranges did not change during the study period (Table 1). Incidence calculation In calculating the incidence rates (1960 –1997), the entire population of Olmsted County was considered to be at risk. Age- and sex-specific person-year denominators were estimated from decennial census data for Olmsted County, with linear interpolation between census years (15). It was assumed that, with a fixed number of person-years, the number of incidence cases followed a Poisson distribution. This assumption allowed for the estimation of se and the calculation of confidence intervals for incidence rates. The rates were age- and sex-adjusted to the distribution of the 1990 United States white population by expressing adjusted rates as weighted averages of crude rates. The se and confidence intervals of adjusted rates were based on the same assumptions. The relationships of crude incidence rates to age, sex, seasonality, and decade of diagnosis were assessed by use of generalized linear models, assuming a Poisson error structure (16). Clinical features and outcome The data from 94 cases seen from January 1, 1970, through December 30, 1997, were evaluated in detail for clinical and laboratory features and outcomes. Descriptive statistics, the 2 test, and the Student t test were used for comparison of differences between groups and for other analyses of these data. Results Incidence The overall age- and sex-adjusted incidence rate from 1960 through 1997 was 4.9 cases per 100,000/yr (95% confidence interval, 4.1–5.7). The total age- and sex-adjusted incidence rate decreased from 8.7 in the 1960s, to 5.6 in the 1970s, to 3.2 in the 1980s (P ⬍ 0.01). In the 1990s, the incidence rate remained stable at 3.6. The main difference in rates was between the 1960s and the following years. Differences in rates among the 1970s, 1980s, and 1990s were only marginally significant (P ⫽ 0.05). Incidence rates for women in the 1960s, 1970s, 1980s, and 1990s were 14.1, 9.0, 4.7, and 4.6, respectively; for men, 2.8, 2.1, 1.5, and 2.8, respectively. Rates between the sexes were significantly different (P ⬍ 0.01). The decrease in the incidence was statistically significant in the female population between the 1960s and the 1990s (P ⬍ 0.01) but not in the male population (P ⫽ 0.46). The peak incidence was in the fifth decade of life for both sexes. Seasonal variation Of the 160 patients in 38 yr, 46 were seen in the spring, 45 in the fall, 36 in the summer, and 33 in the winter. In spring and summer, more patients were seen in May and August than in any other 2 months (Fig. 1). Although there was a trend toward more cases in fall and spring, Poisson regression did not show statistical significance (P ⫽ 0.26). There was no obvious geographic clustering within the county, no apparent familial aggregation of cases, and no involvement of more than 1 member in a household. Clinical features (1970 –1997) Demographic data. There were 94 patients in the cohort within the 28-yr period: the female-male ratio was 3.5; mean age was 46 yr (median, 45 yr; range, 14 – 87 yr); 76 patients were residents of the city of Rochester, and 18 lived in other areas of the county. Four patients were never seen at Mayo Clinic; the diagnoses for those 4 were made elsewhere. Symptoms at onset. A history of upper respiratory tract infection in the 30 d before the onset of SAT was recorded for 20 patients; a history of upper respiratory tract infection in TABLE 1. Laboratory data at diagnosis for 94 patients with SAT Laboratory test n Reference range Total T4 (g/dl) Free T4 (ng/dl)b Total triiodothyronine (ng/dl)c TSH (IU/ml)d Thyroglobulin (ng/ml)f Erythrocyte sedimentation rate (mm/h) 78 20 10 44 5 53 Uptake of iodine 131 (%) 79 5.0 –12.5 0.8 –1.8 80 –180 0.30 –5.0 ⱕ59.4 0 –22 (male) 0 –29 (female) 8 –29 a a Mean ⫾ SD Median 12.1 ⫾ 43 2.8 ⫾ 2.0 180 ⫾ 75 0.9 ⫾ 1.6e 383 ⫾ 295 51 ⫾ 26 11.1 2 134 0.05 271 44 3.2 ⫾ 3.5 1.4 To convert to nM, multiply by 12.87. To convert to pM, multiply by 12.87. To convert to nM, multiply by 0.0154. d To convert to mIU/liter, multiply by 1.0. e Mean serum concentration of TSH is higher than expected because, between 1974 and 1986 when the sensitive TSH assay was not available, the lower limit of detection of serum TSH was higher. f To convert to g/liter, multiply by 1.0. b c 2102 J Clin Endocrinol Metab, May 2003, 88(5):2100 –2105 Fatourechi et al. • Subacute Thyroiditis typical of patients with SAT. We excluded from the study 3 patients who had a clinical diagnosis of SAT not confirmed by fine-needle aspiration biopsy. FIG. 1. Number of patients with SAT seen in Olmsted County, Minnesota, in different months of the year, from 1960 to 1997. the 30 –90 d before the onset of SAT was recorded for 4 patients. Mild to severe spontaneous thyroid pain was present in 90 patients and was absent in 4. Pain was documented in 1 of these 4 patients at a follow-up visit. Radiation of pain to the jaw was documented in 12 patients and to the ears in 18 patients. At presentation, dysphagia was present in 30 patients, arthralgia in 3, myalgia in 12, tremor in 19, sweating in 22, and weight loss (4.5–18.2 kg) in 15. Signs. Of 41 patients who had body temperature recorded, 27 had temperatures between 37 C and 38 C, and 7 had temperatures between 38 C and 39.5 C. Only 5 patients did not have pain on palpation of the thyroid. For cases in which clinical estimation of goiter weight was recorded, the estimate was between 20 and 70 g in 57 patients and less than 20 g in 19 patients. Tenderness involved the entire gland in 44 patients, the left lobe in 14, and the right lobe in 16. Localized tenderness of 1 of the lobes was noted in 10 patients; the area of tenderness was not documented in the remainder. No cases of tracheal compression or deviation were documented. Palpable nodules were noted in 21 patients, with nodules varying in size between 5 and 30 mm. Laboratory and imaging data. Initial laboratory data at diagnosis are shown in Table 1. Because the study encompassed a period when serum TSH or sensitive TSH measurement was not available, some patients did not have a serum TSH measurement. Antimicrosomal antibodies were measured for 14 patients: for 11 patients, they were weakly positive at a 1:100 dilution; for 3 patients, antimicrosomal antibodies were highly positive at dilutions of 1:25,600, 1:1,600, and 1:400. Only 2 patients had thyroid-stimulating immunoglobulin measurements, and both were negative. Technetium 99 thyroid scanning was done in 14 patients; in each of the 14, the thyroid was not visualized or was visualized poorly. Thyroid ultrasonography was done in 6 patients; in each, it showed mild thyroid enlargement with inhomogeneous echo texture and hypoechoic areas consistent with inflammation. Fine-needle aspiration biopsy of the thyroid was done in 13 patients: results for 11 patients were typical for SAT; for 1, results were nondiagnostic; and for 1, results were consistent with Hashimoto thyroiditis. We included these last 2 patients because they were otherwise Therapy. Table 2 shows the types of therapy for the patients. Among 94 patients, 34 (36%) received prednisone therapy (daily starting dose: mean, 40 mg; median, 42 mg). Therapy with nonsteroidal anti-inflammatory drugs (NSAIDs) was the only therapy for 39 patients; therapy with NSAIDs and corticosteroids was given to 18 patients. Acetylsalicylic acid was given as therapy to 43 patients; 14 received other NSAIDs. For the group receiving NSAIDs and acetylsalicylic acid, the mean period from the start of therapy to complete alleviation of pain was 35 d (median, 21 d; range, 7–150 d). The duration of therapy with high doses of prednisone (30 – 40 mg/d) was 3–30 d (median, 7 d; range, 4 – 43 d). After initial high-dose therapy, prednisone therapy was tapered and continued for a mean of 34 d (median, 30 d). Five patients required restarting prednisone therapy because of early recurrence within 2–10 months. Excluding patients who had early recurrence (within the first 6 –12 months), the median time to complete resolution of pain, after beginning therapy with steroids, was 4 d (mean, 8 d). -Blockers were given to 11 patients. Recurrence. Early initial recurrence resulted in restarting corticosteroid therapy in 10% of the patients who had received corticosteroid therapy initially. SAT recurred after 6 –21 yr in 4 patients (4% of the patients) (Table 3). A fifth patient had a clinical diagnosis of recurrence 11 yr after the initial episode, but the data were not adequate for inclusion in this group. The recurrences were milder than the initial episodes. One patient had a documented episode of SAT before the study period and was seen after 6 yr with recurrence. She had two undocumented episodes before the two documented episodes. Despite corticosteroid therapy, hypothyroidism developed in two of the patients who had recurrence. TABLE 2. Therapy for 94 patients with SAT Patients Therapy NSAIDs alone NSAIDs alone or NSAIDs ⫹ corticosteroids Corticosteroids alone Corticosteroids alone or with other therapy Acetaminophen Thyroidectomy n % 39 57 15 34 21 1 41 61 16 36 22 1 TABLE 3. Outcome among 94 patients with SATa Outcome b Early recurrence Early hypothyroidismb Late recurrencec Late hypothyroidismc Nodular goiter Associated autoimmune disease Malignancy Patients (%) 10 34 4 15 5 5 11 a None had thyroid cancer, Graves’ disease, or Graves’ ophthalmopathy. b Early, within the first 6 –12 months. c Late, after 1 yr. Fatourechi et al. • Subacute Thyroiditis Follow-up of thyroid function. At some time during follow-up, serum TSH concentrations were more than 5 IU/ml (5 mIU/liter) in 32 patients (34%); in 16 of those patients (17%), serum TSH was more than 10 IU/ml (10 mIU/liter). Among the 34 patients who had received corticosteroid therapy, 10 had TSH levels greater than 5 IU/ml (5 mIU/liter) (29%). Of the 60 patients not receiving corticosteroids, 22 (37%) had TSH elevation at follow-up; 11 of those (18%) had serum TSH greater than 10 IU/ml (10 mIU/liter). TSH elevation was documented in the first year in 27 patients. In 4 patients, TSH elevation persisted after the first year, and they received T4 therapy. Five patients were treated within the first year, and treatment was continued long term. Of the 27 patients with elevated TSH levels in the first year, only 9 were receiving T4 therapy at latest follow-up; among the 18 who did not receive therapy, TSH levels normalized during follow-up. For the 27 patients, mean TSH (⫾sd) was 16 ⫾ 14 IU/ml (16 ⫾ 14 mIU/liter), and median TSH was 9.6 IU/ml (9.6 mIU/liter). For the patients who had TSH elevation in the first year, the elevation occurred after a mean of 84 d (median, 68 d). Late-onset hypothyroidism occurred 2–24 yr after SAT in 5 patients. At latest follow-up, for 14 patients (15%), there was definite previous documentation of hypothyroidism; 3 other patients were receiving therapy without adequate documentation. With the exclusion of patients with early (in the first year) TSH elevation, 9 patients (9.5%) were receiving T4 therapy for late (after the first year) hypothyroidism. With the exclusion of the 3 cases without adequate documentation, T4 therapy was given to 25% of the group (8 patients) receiving corticosteroid therapy and 10% of the group (6 patients) not receiving corticosteroid therapy. The corticosteroid therapy group had a higher rate of hypothyroidism on long-term follow-up (P ⬍ 0.05). No patient had Graves’ disease, Graves’ ophthalmopathy, or thyroid cancer on follow-up. One patient continued to have persistent pain beyond 1 yr and had a thyroidectomy. Five had nodular goiter, and 1 had a subsequent diagnosis of Hashimoto thyroiditis. Associated conditions. There was one case of seronegative rheumatoid arthritis, one case of Sjögren syndrome, one case of chronic ulcerative colitis, one case of urticaria, one case of contact dermatitis, and one case of psoriasis. However, nonthyroid malignancies were common (11.4% of patients). Either before or after SAT, three patients had breast cancer and eight patients had other malignancies (prostate, lung, endometrial, cervical, gastric, and pancreatic cancer; non-Hodgkin lymphoma; and systemic mastocytosis). It is not clear whether this rate of malignancy is higher than in the general population or reflects the long follow-up period of 28 yr. Discussion SAT (de Quervain thyroiditis) is a transient, inflammatory thyroid disease associated with pain and tenderness of the thyroid as well as with generalized somatic symptoms, causing great discomfort for weeks or months if left untreated. SAT is most likely the result of a viral infection (17–19). There is no definitive cure for painful SAT, but there is effective treatment that will ameliorate the symptoms and allow the J Clin Endocrinol Metab, May 2003, 88(5):2100 –2105 2103 disease to run its course in an asymptomatic fashion. Diagnosis of SAT excludes forms of infective bacterial and fungal thyroiditis. The term “SAT” is not usually applied to silent, painless thyroiditis with lymphocytic pathological features or to postpartum thyroiditis (6, 20). Salicylates and NSAIDs can be given to patients with mild and moderate forms of the disorder (20). In severe forms, therapy with corticosteroids generally provides rapid relief of symptoms within 24 – 48 h (20). Prednisone therapy may be initiated in dosages of 40 mg daily, with a gradual reduction in dosage thereafter over several weeks (20). Corticosteroid therapy was given to 36% of our patients; symptoms resolved in a few days. Thyroidectomy should be considered for only a small minority of patients who have repeated relapses despite appropriate treatment (21). Only one patient in our series required thyroidectomy. Few epidemiological studies of SAT have been reported (12, 22–26). Only 1 study has attempted to collect and report all cases seen in a specific community (13). In that report, from our institution, cases seen between 1960 and 1967 in our county were studied. The incidence was higher in females (19.1 per 100,000/yr) than in males (4.4 per 100,000/yr). The higher incidence in females in the present study is similar to the findings of previous reports (5). Although the cause is most likely viral, SAT, like all other thyroid conditions, occurs most commonly (in 80% of cases) in women who are 40 –50 yr old (5). The reason for the reduced incidence after 1970 is not clear. However, the use of more refined laboratory tests in recent years, resulting in the exclusion of some cases, may be an explanation. Similar to the results of the previous Mayo Clinic study, the results of our present study did not show statistically significant seasonal variation over the extended 38-yr period. However, there was a slight trend toward aggregation of cases in May and August and also in the fall and spring seasons. Familial or geographic aggregation of cases was not noted either. In some reports, seasonal clustering has been reported; but because these were not community studies, the importance is not clear (12, 25). Other studies show no seasonal clustering of cases (13, 22). In our present study, associated autoimmune conditions did not seem more common than those observed in the general population. Transient early hypothyroidism and late-onset (or permanent) hypothyroidism can occur (23, 27–30). Reports indicate that the hyperthyroid phase is often followed by a transient phase of hypothyroidism, in which an exaggerated TSH response to TSH-releasing hormone was noted in 56% of the patients (29). Permanent hypothyroidism is also a known complication and has been reported in 5% of cases (29). During the shorter follow-up of the previous Olmsted County study, only 3 of the 70 patients (4%) required T4 therapy. In the present report with longer follow-up and more patients, the incidence of hypothyroidism was higher (9.5% had documented elevated TSH after the first year). Among our patients, 34% had subsequent abnormal TSH elevation, the majority occurring within the first year after onset of SAT symptoms. There was no significant difference in the overall rate of early-onset hypothyroidism between the group receiving corticosteroid therapy and the group not receiving corticosteroid therapy. However, at latest follow- 2104 J Clin Endocrinol Metab, May 2003, 88(5):2100 –2105 up, more patients who had received corticosteroid therapy were receiving T4 therapy, which indicates that corticosteroid therapy does not prevent development of permanent hypothyroidism. The reason for the higher degree of permanent hypothyroidism in the corticosteroid-treated group is not completely clear, but patients with more severe disease may be more likely to be treated with corticosteroids and thus be more likely to have hypothyroidism develop. One study of 29 patients from France reported an incidence of hypothyroidism of 31% (23). Development of autoimmune hypothyroidism, initiated by a thyroid-destructive insult, can explain this finding. One could argue that the hypothyroid cases were transient thyroid failures inappropriately treated with long-term T4. However, only 5 of the 27 patients with early TSH elevation received T4 therapy in the first year; and in the 22 patients who were observed, thyroid function later normalized in 18 and hypothyroidism persisted in 4, who required therapy. To determine the actual rate of irreversible hypothyroidism, discontinuation of T4 therapy and reassessment would be necessary for the early therapy group, but this was not done for our cohort. In the pathogenesis of late-onset hypothyroidism, autoimmunity and development of antithyroid and blocking antibodies have been speculated (28, 30 –32). SAT may trigger autoreactive B cells to produce TSH-receptor antibodies, resulting in TSHreceptor antibody-associated thyroid dysfunction in some patients (31). In our community study, we did not see the development of Graves’ disease and Graves’ ophthalmopathy that have been reported in isolated cases (33–35); thus, in community practice, development of subsequent Graves’ disease is unlikely. In a recent report of SAT, 22% of 36 patients had early recurrence, necessitating restoration of a higher dose of corticosteroids (8). In our study, early recurrence within the subacute phase occurred in 10% of the group receiving corticosteroid therapy. SAT, which is considered to be a viral disease (17–19), rarely recurs after a complete recovery, possibly because of immunity to the offending virus. However, late recurrence after several years has been reported (36 –39). In 1 study, 4 recurrent episodes of SAT occurred in 3 of 222 patients (1.4%). The recurrent episodes were similar to the first episodes of SAT. The titers of various viral antibodies were not increased during the clinical course of recurrence (39). In a larger study that evaluated data for 3,344 patients with SAT between 1970 and 1993, SAT recurred in 48 of 3,344 patients (1.4%) (mean, 14.5 ⫾ 4.5 yr after the first episode). Five patients experienced a third episode (mean, 7.6 ⫾ 2.4 yr after the second episode) (39). Our cohort study showed a 4% recurrence rate after many years. One patient had 3 recurrences, with several years between episodes. Theoretically, late recurrence possibly occurs after the disappearance of immunity to the previous viral infection. Hypothyroidism developed in 2 of 4 patients with late recurrence; but from these 2 cases, we cannot conclude that recurrence predisposes to late-onset hypothyroidism. Our study has some limitations. Data for frequency of symptoms, in the medical records, may have been incomplete. Diagnostic practices have changed; some of the earlier cases did not have a sensitive TSH measurement. Not all the Fatourechi et al. • Subacute Thyroiditis cases had results for erythrocyte sedimentation rate. Underestimation may have been a problem, and patients with milder disease may not have sought medical help. Some of the patients receiving over-the-counter medications may not have supplied an adequate drug history. There are also some limitations to the conclusions derived from this retrospective study. It is probable that patients with severe disease were more likely to be treated with corticosteroids, and the severity of the thyroid inflammation may have been the cause of more frequent hypothyroidism in this group. Also, the greater incidence of hypothyroidism in this cohort could be an overestimation, because it is not clear whether some patients had temporary disease in the first year and were treated without a need for lifelong therapy. Our patients were from a mostly white population, and generalization to other ethnic groups may not be appropriate. However, a strong point of our study is that the high quality of primary care in the community reduces the possibility of missed cases. In addition, care was given by a limited number of providers, and the majority of patients were referred to subspecialists for management of thyroid disorder. In conclusion, our cohort study shows that, in the first year of onset of SAT, transient hypothyroidism is very common. There is no significant difference in the incidence of mild thyroid failure, in this early phase, between patients receiving corticosteroid therapy and those not receiving corticosteroid therapy. In contrast, long-term hypothyroidism requiring T4 therapy is significantly more common in the group receiving corticosteroid therapy. Therefore, corticosteroid therapy does not prevent subsequent hypothyroidism. Thus, we believe that corticosteroid therapy should be given to improve the symptoms and quality of life without an expectation of reducing long-term thyroid dysfunction. Also, because of the high incidence of transient hypothyroidism in the first year, T4 therapy should not be considered lifelong if started in the first year. Acknowledgments Received November 14, 2002. Accepted February 6, 2003. Address all correspondence and requests for reprints to: Vahab Fatourechi, M.D., Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905. This work was supported in part by the Richard Emslander Clinical Investigator Award and Scholarly Clinician Award of the Mayo Foundation and the U.S. Public Health Service, National Institutes of Health (AR-30582). These data were presented in part at the 74th Annual Meeting of the American Thyroid Association, November 9, 2001, Washington, D.C., and at the 84th Annual Meeting of The Endocrine Society, June 19, 2002, San Francisco, California. References 1. Volpe R 1979 Subacute (de Quervain’s) thyroiditis. Clin Endocrinol Metab 8:81–95 2. Walfish PG 1997 Thyroiditis. Curr Ther Endocrinol Metab 6:117–122 3. Ross DS 1998 Syndromes of thyrotoxicosis with low radioactive iodine uptake. Endocrinol Metab Clin North Am 27:169 –185 4. Intenzo CM, Park CH, Kim SM, Capuzzi DM, Cohen SN, Green P 1993 Clinical, laboratory, and scintigraphic manifestations of subacute and chronic thyroiditis. Clin Nucl Med 18:302–306 5. Singer PA 1991 Thyroiditis: acute, subacute, and chronic. Med Clin North Am 75:61–77 6. Volpe R 1981 Subacute thyroiditis. Prog Clin Biol Res 74:115–134 7. Slatosky J, Shipton B, Wahba H 2000 Thyroiditis: differential diagnosis and management. Am Fam Physician 61:1047–1052, 1054 Fatourechi et al. • Subacute Thyroiditis 8. Mizukoshi T, Noguchi S, Murakami T, Futata T, Yamashita H 2001 Evaluation of recurrence in 36 subacute thyroiditis patients managed with prednisolone. Intern Med 40:292–295 9. Kitchener MI, Chapman IM 1989 Subacute thyroiditis: a review of 105 cases. Clin Nucl Med 14:439 – 442 10. Hwang SC, Jap TS, Ho LT, Ching KN 1989 Subacute thyroiditis— 61 cases: review. Zhonghua Yi Xue Za Zhi (Taipei) 43:113–118 11. Jayaram G, Marwaha RK, Gupta RK, Sharma SK 1987 Cytomorphologic aspects of thyroiditis. A study of 51 cases with functional, immunologic and ultrasonographic data. Acta Cytol 31:687– 693 12. Dulipsingh L, Ikram Z, Malchoff CD, Shoukri KC 1998 A cluster of cases of subacute and silent thyroiditis in the northern Connecticut, Greater Hartford area. Conn Med 62:395–397 13. Furszyfer J, McConahey WM, Wahner HW, Kurland LT 1970 Subacute (granulomatous) thyroiditis in Olmsted County, Minnesota. Mayo Clin Proc 45: 396 – 404 14. Melton III LJ 1996 History of the Rochester Epidemiology Project. Mayo Clin Proc 71:266 –274 15. Bergstralh EJ, Offord K, Chu CP, Beard CM, O’Fallon WM, Melton LJ 1992 Calculating incidence, prevalence and mortality rates for Olmsted County, Minnesota: an update. Technical Report Series No. 49. Rochester: Mayo Clinic 16. McCullagh P, Nelder JA 1983 Generalized linear models. London: Chapman and Hall 17. Brouqui P, Raoult D, Conte-Devolx B 1991 Coxsackie thyroiditis. Ann Intern Med 114:1063–1064 18. Volpe R, Row VV, Ezrin C 1967 Circulating viral and thyroid antibodies in subacute thyroiditis. J Clin Endocrinol Metab 27:1275–1284 19. Sato M 1975 Virus-like particles in the follicular epithelium of the thyroid from a patient with subacute thyroiditis (de Quervain). Acta Pathol Jpn 25:499 –501 20. Volpe R 1993 The management of subacute (de Quervain’s) thyroiditis. Thyroid 3:253–255 21. Duininck TM, van Heerden JA, Fatourechi V, Curlee KJ, Farley DR, Thompson GB, Grant CS, Lloyd RV 2002 de Quervain’s thyroiditis: surgical experience. Endocr Pract 8:255–258 22. Oksa H, Jarvenpaa P, Metsahonkala L, Pasternack A, Leinikki P 1989 No seasonal distribution in subacute de Quervain’s thyroiditis in Finland. J Endocrinol Invest 12:495 23. Cordray JP, Nys P, Merceron RE, Augusti A 2001 Frequency of hypothyroidism after de Quervain’s thyroiditis and contribution of ultrasonographic thyroid volume measurement [French]. Ann Med Interne (Paris) 152:84 – 88 24. de Bruin TW, Riekhoff FP, de Boer JJ 1990 An outbreak of thyrotoxicosis due to atypical subacute thyroiditis. J Clin Endocrinol Metab 70:396 – 402 J Clin Endocrinol Metab, May 2003, 88(5):2100 –2105 2105 25. Martino E, Buratti L, Bartalena L, Mariotti S, Cupini C, Aghini-Lombardi F, Pinchera A 1987 High prevalence of subacute thyroiditis during summer season in Italy. J Endocrinol Invest 10:321–323 26. Nordyke RA, Gilbert Jr FI, Lew C 1991 Painful subacute thyroiditis in Hawaii. West J Med 155:61– 63 27. Gozariu L, Stroe M, Vladutiu T, Yepez-Escobar NG, Simionescu L, Szantay I, Dumitru E, Florescu O 1986 Late hypothyroidism following subacute thyroiditis. Exp Clin Endocrinol 87:48 –52 28. Iitaka M, Kakinuma S, Yamanaka K, Fujimaki S, Oosuga I, Wada S, Katayama S 2001 Induction of autoimmune hypothyroidism and subsequent hyperthyroidism by TSH receptor antibodies following subacute thyroiditis: a case report. Endocr J 48:139 –142 29. Lio S, Pontecorvi A, Caruso M, Monaco F, D’Armiento M 1984 Transitory subclinical and permanent hypothyroidism in the course of subacute thyroiditis (de Quervain). Acta Endocrinol (Copenh) 106:67–70 30. Tikkanen MJ, Lamberg BA 1982 Hypothyroidism following subacute thyroiditis. Acta Endocrinol (Copenh) 101:348 –353 31. Iitaka M, Momotani N, Hisaoka T, Noh JY, Ishikawa N, Ishii J, Katayama S, Ito K 1998 TSH receptor antibody-associated thyroid dysfunction following subacute thyroiditis. Clin Endocrinol (Oxf) 48:445– 453 32. Nakamura S, Saio Y, Suzuki E 1996 Subacute thyroiditis with thyroid-stimulation blocking antibodies: a case report. Endocr J 43:185–189 33. Wartofsky L, Schaaf M 1987 Graves’ disease with thyrotoxicosis following subacute thyroiditis. Am J Med 83:761–764 34. Fukata S, Matsuzuka F, Kobayashi A, Hirai K, Kuma K, Sugawara M 1992 Development of Graves’ disease after subacute thyroiditis: two unusual cases. Acta Endocrinol (Copenh) 126:495– 496 35. Werner SC 1979 Graves’ disease following acute (subacute) thyroiditis. Arch Intern Med 139:1313–1315 36. Tauveron I, Thieblot P, Marcheix JC 1991 Recurrence after 12 years of de Quervain-Crile subacute thyroiditis [French]. Rev Med Interne 12:396 37. Bauman A, Friedman A 1983 Recurrent subacute thyroiditis: a report of three cases. NY State J Med 83:987–988 38. Iitaka M, Momotani N, Ishii J, Ito K 1996 Incidence of subacute thyroiditis recurrences after a prolonged latency: 24-year survey. J Clin Endocrinol Metab 81:466 – 469 39. Yamamoto M, Saito S, Sakurada T, Tamura M, Kudo Y, Yoshida K, Kaise K, Kaise N, Fukazawa H, Itagaki Y, Yoshinaga K 1988 Recurrence of subacute thyroiditis over 10 years after the first attack in three cases. Endocrinol Jpn 35:833– 839