Clinical Thyroidology March 2001 Volume 13 Issue 1thyroid.org/wp-content/uploads/publications/clinthy/clinthy_v131.pdf ·

THYROID DISEASES

Approximately Half of Patients with Hyperthyroidism orHypothyroidism Have Neuromuscular Dysfunction . . .1

Radiation Therapy in Children and Adolescents is Assoc-iated with Risk for All Types of Thyroid Disease . . . . .2

HYPERTHYROIDISM

Subclinical Hyperthyroidism Causes Symptoms and HasDeleterious Cardiovascular Effects . . . . . . . . . . . . . . . .3

High Serum Triiodothyronine Concentrations Are a RiskFactor for Coronary Events . . . . . . . . . . . . . . . . . . . . . .4

Liver Dysfunction in Patients with Hyperthyroidism . . .5

Propylthiouracil, but Not Methimazole, Is Associated withAntineutrophil Cytoplasmic Antibodies . . . . . . . . . . . . .6

Mothers Taking Methimazole Can Nurse Their InfantsSafely . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

GRAVES' OPHTHALMOPATHY

Surgical and Radiation Therapy Lowers Intraocular Pressure in Patients with Graves' Ophthalmopathy . . . .8

HYPOTHYROIDISM

Thyroxine Therapy Lowers Serum Cholesterol Concentrations in Patients with SubclinicalHypothyroidism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9

Hypothyroidism Caused by Rapid Thyroid HormoneDestruction in an Infant with Hepatic Hemangiomas .10

Fetal Loss is Increased in Pregnant Women with Hypothyroidism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

THYROID CANCER

Iodine-123 Is Superior to Iodine-131 in Detecting Thy-roid Remnants in Patients with Differentiated ThyroidCarcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12

Presence of PAX8-PPARγ1 Gene TranslocationDistinguishes Follicular Carcinoma from Other FollicularLesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

Clinical Findings and Treatment of Patients with ThyroidCancer Treated at Over 1500 Hospitals in the UnitedStates in 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14

Paclitaxel Has Some Efficacy in Patients with AnaplasticThyroid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . .15

THYROID HORMONE ACTION

Thyroid Hormone Receptors Are Present in the BrainVery Early in Fetal Life . . . . . . . . . . . . . . . . . . . . . . . .16

A Thyroid Hormone Analogue with Differential Cardiacand Hepatic Actions . . . . . . . . . . . . . . . . . . . . . . . . . . .17

THYROID AUTOIMMUNITY

Low Frequency of Anti-Sodium-Iodide TransporterAntibodies in Autoimmune Thyroid Disease . . . . . . . .18

Hypothyroidism Can Be Transferred by Bone MarrowTransplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19

Successful Treatment of Cushing’s Disease EvokesThyroid Autoimmunity . . . . . . . . . . . . . . . . . . . . . . . .20

CLINICALTHYROIDOLOGY

VOLUME XIII ● ISSUE 1 MARCH 2001

A publication of the American Thyroid Association

Editor-in-Chief Robert D. Utiger, M.D. Thyroid DivisionDepartment of MedicineBrigham & Women's Hospital 77 Avenue Louis Pasteur Boston, MA 02115(617) 525-5171 Telephone (617) 731-4718 Fax [email protected]

Founding EditorRidha Arem, M.D.

PresidentWilliam W. Chin, M.D.

Secretary Paul W. Ladenson, M.D.

Treasurer David S. Cooper, M.D.

President Elect Carole A. Spencer, Ph.D.

Publications ChairLynn A. Burmeister, M.D.

AdministratorDiane P. MillerAmerican Thyroid Association55 Old Nyack Turnpike, Suite 611Nanuet, NY 10954-2454(845) 623-1800 Telephone (845) 623-3736 Fax [email protected]

Published BySaratoga Graphics7 Kaatskill WayBallston Spa, NY 12020(518) 583-0243Art Direction: Kandra L. [email protected]

Clinical ThyroidologyCopyright © 2001American Thyroid Association, Inc Printed in the USA. All rights reserved

A New Clinical Thyroidology

This issue marks a new beginning for Clinical Thyroidology. It is now a publication ofthe American Thyroid Association, and it has a new appearance and a new editor. Nowbeginning its 13th year of publication, this journal was launched by Dr. Ridha Arem forthe purpose of informing endocrinologists about the best clinical studies in thyroidolo-gy - wherever published - and providing commentary about those studies. As foundingeditor, Dr. Arem nourished the journal wisely, so that it has become a valuable source ofinformation about new studies in the field.

The purpose of this format is not only to inform readers of interesting and usefulnew studies of thyroid pathophysiology and disease (the summaries), but also to put thestudies in perspective and offer editorial opinion about them (the commentaries). Ideally,readers will be stimulated to look further. Journals such as this do not take the place ofthose that contain reports of original research, indeed they are totally dependent uponreports of original research for their existence. What you will get here is only a summ-ary and commentary, and no matter how carefully an article is summarized or how good(or bad) the commentary may be, you will not see all the data you ought to see beforedrawing your own conclusions about the value of a particular study and whether it isapplicable to your work or practice.

The focus of Clinical Thyroidology will continue to be on clinical studies, and the searchfor such studies will be broad, extending beyond journals of endocrinology as it has inthe past. In addition, I think that there is a place for summaries and commentary aboutpreclinical studies. As examples, there is in this issue a summary of an article describingthe identification of thyroid hormone receptors in the brain of fetuses and another ofan article describing the effects of a new thyroid hormone analogue in mice and rats.

This journal was supported initially by the Daniels Pharmaceutical Company, and thesupport continued when Daniels was bought by Jones Pharma, Inc. That company hasin effect given the journal to the association, and has generously agreed to provide sup-port for it for five years. I have no financial or other relationships with Jones Pharma.

Clinical Thyroidology is the second scientific publication, after Thyroid, of the AmericanThyroid Association. It will be distributed three times yearly to all members of the asso-ciation, other endocrinologists, and anyone else who requests a copy, and it will be avail-able soon on the association’s Web site. It is my hope that it will complement Thyroidwell.

I am pleased to be the editor of this journal. I think it will be fun, informative andstimulating. You may not find the contents fun, but I hope you will find them informa-tive and stimulating, and will let me know if they are not.

Robert D. Utiger, M.D.

VOLUME XIII ● ISSUE 1 MARCH 2001

ATA News & Upcoming Events

Clinical Thyroidology

The journal is available without charge. Thosewho wish to be added to the mailing list shoulde-mail the administrative office ([email protected]), including your full name, institution,address, telephone and fax numbers, and e-mailaddress.

73rd Annual Meeting of the Association

The annual meeting will be held at the OmniShoreham Hotel in Washington, DC fromSeptember 12 to 16, 2001. The deadline for sub-mission of abstracts is Wednesday, May 9, 2001.All abstracts must be submitted electronicallyusing the form on the ATA web site. Pre-registration material for the meeting will bemailed in May 2001.The preliminary scientific program for 2001 isavailable online (www.thyroid.org).

CLINICALTHYROIDOLOGY

HYPERTHYROIDISM AND HYPOTHYROIDISM

Approximately half of patients with hyperthyroidism or hypothyroidismhave neuromuscular dysfunction

Duyff RF, Van den Bosch J, Laman DM, Potter van Loon B-J, Linssen WHJP. Neuromuscular findings in thyroid dys-function: a prospective clinical and electrodiagnostic study. J Neurol Neurosurg Psychiatry 2000;68:750-5.

SUMMARY

Background Patients with either hyperthyroidism orhypothyroidism may have symptoms and signs of neuro-muscular dysfunction, but there have been few systematicstudies of neuromuscular function in these patients.

Methods Comprehensive clinical and electrodiagnosticstudies were done in 24 consecutive patients with hyperthy-roidism (mean age 48 years) and 21 patients with hypothy-roidism (mean age 43 years) before and after treatment.The studies included detailed assessment of musclestrength and sensation and electrodiagnostic studies (elec-tromyography and measurements of nerve conductiontimes).

Results Among the 24 patients with hyperthyroidism, 67percent had complaints of muscle weakness, 62 percent hadobjective muscle weakness, mainly in the proximal musclesof the legs, 19 percent had symmetric distal sensory abnor-malities and depressed distal tendon reflexes, 38 percenthad generalized hyperreflexia and 76 percent had tremor.During or after treatment (an antithyroid drug or radioio-dine) the muscle weakness resolved in an average of 4months and the sensory abnormalities resolved in an aver-age of 7 months.

Among the 21 patients with hypothyroidism, 79 percenthad symptoms of muscle dysfunction (weakness 54 per-cent, fatigability, muscle cramps, pain or stiffness 42 per-cent). Muscle weakness was detected in 38 percent, largelyin proximal muscles, symmetric distal sensory abnormalitiesand depressed reflexes in 42 percent and the carpal tunnelsyndrome in 29 percent. Seven patients had high serumcreatine kinase concentrations, which did not correlate withmuscle weakness. During treatment the muscle symptomsresolved in most patients in an average of 7 months.

Conclusion The majority of patients with hyperthy-roidism or hypothyroidism have symptoms and signs ofneuromuscular dysfunction.

COMMENTARY

These 45 patients were seen in amedical outpatient department in theNetherlands, and their age (mean 46years), sex (80 percent women) and biochemical abnormalities were charac-teristic of the two thyroid disorders.The study has several limitations. Thepatients were recruited from a largergroup of 141 patients (most of theothers declined to participate becauseof the electrodiagnostic studies). Thosewho participated may have had more, ormore severe, symptoms, and thereforethe study patients may not be represen-tative of all patients with these disorders.

Also, similar studies were not donesimultaneously in a group of normalsubjects of the same age and sex.

Whatever the limitations of thestudy, there is considerably moreinformation in the article than summa-rized above. It includes detailed sum-maries of different symptoms and signs,the results of systematic grading ofmuscle strength and of measurements ofmuscle strength by dynamometry, theresults of the electrodiagnostic studiesand the relationships between neuro-muscular symptoms and the overallduration of thyroid disease. The resultsthus provide a useful picture of thebreadth of neuromuscular symptoms

and signs that occur in patients withhyperthyroidism and hypothyroidismand the extent to which they arereversible.


CLINICAL THYROIDOLOGY ● VOLUME XIII ● ISSUE 1 ● 1

SUMMARY

Background Most patients with Hodgkin's disease receiveradiation therapy to the neck, which may subsequentlyaffect thyroid function and cause thyroid tumors. The typesof thyroid disease and the risk factors for them have notbeen evaluated in a large cohort followed for many years.

Methods The Childhood Cancer Survivor Study is acohort study of children, adolescents and young adults (age<21 years) treated for cancer between 1970 and 1986 at 25centers in the United States who survived for at least 5 yearsafter diagnosis. This study focused on the 1791 patients(959 males and 82 females) with Hodgkin's disease in thecohort for whom follow-up data were obtained. The medi-an age at diagnosis was 14 years (range 2 to 20), and themedian age at follow-up was 30 years (range 12 to 47).

Follow-up data were obtained by mail questionnaire con-taining specific questions about thyroid dysfunction andtumors; diagnoses of thyroid cancer were verified by reviewof pathology reports. The same questionnaire was sent to348 siblings (150 males and 198 females, median age 25years) of the patients with Hodgkin’s disease.

Results The incidence of hyperthyroidism, hypothy-roidism and thyroid nodules was significantly higher in thepatients with Hodgkin's disease than their siblings (P<0.001for all comparisons) (table).

Hyperthyroidism. The mean interval between diagnosis ofHodgkin's disease and hyperthyroidism was 8 years (range 0to 22). Factors associated with increased risk were timeafter diagnosis of Hodgkin's disease and higher radiationdoses.

Hypothyroidism. Hypothyroidism was diagnosed a mean of7 years (range 0 to 27 years) after diagnosis of Hodgkin'sdisease. Hypothyroidism was more common in females,and was associated with older age at and increasing timeafter diagnosis of Hodgkin’s disease. Among patients whoreceived 3500 to 4499 cGy and those who received ≥ 4500cGy or more, the actuarial risks of hypothyroidism at 20years were 30 percent and 50 percent, respectively.

Thyroid Nodules. Thyroid nodules were detected on aver-age 14 years (range 0 to 27) after diagnosis of Hodgkin'sdisease. Factors associated with increased risk were femalesex, time after diagnosis of Hodgkin's disease and radiationdoses ≥ 2500 cGy. The actuarial risk of a thyroid nodule ina female 20 years after radiation therapy was 20 percent.

Thyroid Cancer. Among the 146 patients who had thyroidnodules, 11 (7.5 percent) had thyroid cancer; 9 otherpatients reported having thyroid cancers but not thyroidnodules.

Conclusion Thyroid diseases of all types are common inpatients with Hodgkin's disease who receive radiationtherapy.

HYPERTHYROIDISM AND HYPOTHYROIDISM

Radiation therapy in children and adolescents is associated with risk for alltypes of thyroid disease

Sklar C, Whitton J, Mertens A, Stovall M, Green D, Marina N, Greffe B, Wolden S, Robison L. Abnormalities of thethyroid in survivors of Hodgkin’s disease: data from the Childhood Cancer Survivor study. J Clin Endocrinol Metab2000;85:3227-32.

COMMENTARY

Thyroid nodular disease, includingthyroid cancer, is a complication of bothlow- and high-dose radiation therapy tothe head and neck region in young peo-ple. Hypothyroidism is a complicationof high-dose radiation therapy to thisregion in patients of any age, not only inpatients with Hodgkin's disease but alsoin those who receive radiation therapyfor cancer of the larynx or other head

and neck structures.That patients with Hodgkin's dis-

ease who receive radiation therapy havean increased risk of hyperthyroidism isless well known. There is no informa-tion about the causes of hyperthy-roidism in this study, but it was probablyGraves' disease in most of the patients. Presumably, in some patients who aregenetically susceptible to Graves' diseaseradiation therapy injures the T cells thatmaintain the patient's tolerance to their

own thyroid antigen(s). Perhapshypothyroidism in radiation-treatedpatients is due not to a direct destructiveeffect of radiation on thyroid cells, butrather radiation-induced injury of T cellsthat maintain tolerance to the thyroidantigen(s) involved in thyroid-cell cyto-toxicity or antibody-mediated thyroid-cell injury.


2 ● CLINICAL THYROIDOLOGY ● VOLUME XIII ● ISSUE 1

Table. Incidence of Thyroid Disease in Surviving Patients with Hodgkin’s Disease.

Disease Survivors Siblings Relative Risk*Cases Rate/1000 py** Cases Rate/1000 py

Hyperthyroidism 82 1.6 13 0.2 8.0 (4.6-15.1)Hypothyroidism 456 9.6 39 0.6 17.1 (12.5-24.3)Nodule(s) 146 2.9 7 0.1 27.0 (13.6-63.9)

*95 percent confidence intervals in parenthesis.**py denotes person years.

HYPERTHYROIDISM

Subclinical hyperthyroidism causes symptoms and has deleterious cardio-vascular effects

Biondi B, Palmieri EA, Fazio S, Cosco C, Nocera M, Sacca L, Filetti S, Lombardi G, Perticone F. Endogenoussubclinical hyperthyroidism affects quality of life and cardiac morphology and function in young and middle-agedpatients. J Clin Endcrinol Metab 2000;85:4701-5.

COMMENTARY

Some notable features of this studywere that the patients had endogenoussubclinical hyperthyroidism, they had nochanges in serum free T4 and free T3concentrations in the six months beforethey were studied and the study includeduse of the Short Form 36 to assess wellbeing. The overall findings are similar tothose found in patients with subclinicalhyperthyroidism caused by exogenousthyroxine (T4) therapy by the sameinvestigators (1); the Short Form 36scores provide further evidence thatsubclinical hyperthyroidism may be asso-ciated with some disability. In anotherstudy of T4-treated patients, however,the patients had substantially lowerscores on the Hyperthyroid SymptomScale (2). Unfortunately, none of thesestudies included a group of patientswith overt hyperthyroidism.

When subclinical hyperthyroidism is

caused by too much T4, the cardiovascu-lar changes, and presumably the risk ofovert cardiovascular problems, can bereduced by reducing the dose of T4 (3)or giving a β-adrenergic antagonist drug.Symptoms should improve too. There isno reason to doubt that similar improve-ment would occur in patients with sub-clinical hyperthyroidism caused by thy-roid nodular disease, but the study hasn’tbeen done. It should be a controlled study with methimazole orplacebo, or possibly radioiodine andplacebo, and last at least six months.


References

1. Biondi B, Fazio S, Palmieri EA, et al.Effects of chronic subclinical hyperthy-roidism from levothyroxine on cardiacmorphology and function. Cardiologia1999;44:443-9.

2. Shapiro LE, Sievert R, Ong L, et al.Minimal cardiac effects in asymptomaticathyreotic patients chronically treatedwith thyrotropin-suppressive doses of L-thyroxine. J Clin Endocrinol Metab1997;82:2592-5.

3. Mercuro M, Panzuto MG, Bina A,et al. Cardiac function, physical exercisecapacity, and quality of life during long-term thyrotropin-suppressive therapywith levothyroxine: effect of individualdose tailoring. J Clin Endocrinol Metab2000;85:159-64.

SUMMARY

Background Subclinical hyperthyroidism is defined as alow serum thyrotropin (TSH) concentration in an otherwisebiochemically euthyroid patient. The extent to which itcauses symptoms or affects organ function is debated.

Methods The presence of symptoms and cardiovascularfunction were determined in 23 patients (20 women and 3men, mean [±SD] age 40 ± 10 years) with subclinical hyper-thyroidism caused by a solitary thyroid adenoma or amultinodular goiter and 23 age- and sex-matched normalsubjects. All the patients had subclinical hyperthyroidismfor at least 6 months. Symptoms were assessed using aHyperthyroid Symptom Scale and the Short Form 36Health Survey. Cardiac function was assessed by electro-cardiography and echocardiography.

Results The mean symptom score was 10 in the patientsand 4 in the normal subjects (P<0.001) (compared with≥20 in patients with overt hyperthyroidism). The scores onboth the mental and physical scales of the Short Form 36were lower in the patients, indicating more disability. The

mean 24-hour heart rates were 70 and 82 beats/minute inthe normal subjects and the patients (P<0.01), but therewere no differences in the frequency of atrial or ventricularpremature beats. Echocardiography revealed a greater ven-tricular mass (162 g versus 132 g) and end-systolicdiameter, increased systolic contractility and decreased dias-tolic relaxation in the patients, as compared with the normalsubjects.

Conclusion Patients with subclinical hyperthyroidismcaused by thyroid nodular disease have some symptoms ofhyperthyroidism, are less healthy, and have abnormalcardiac function.


SUMMARY

Background Hyperthyroidism alters cardiovascular func-tion in ways that may be detrimental, but the relationshipsbetween hyperthyroidism and angina pectoris and myocar-dial infarction are unclear.

Methods Serum free thyroxine (T4), free triiodothyronine(T3) and thyrotropin (TSH) were measured in all 1049patients aged 40 years or older (median age 70 years, 51 per-cent men) seeking care in a university hospital emergencydepartment during a 4-month period in 1995. Among the185 patients who had angina or a myocardial infarction atthis time, 181 (98 percent) were reevaluated three years later.

Results There were 76 patients in the myocardial infarctiongroup, 109 patients in the angina group, and 864 otherpatients. Serum free T3 concentrations were high in 10.5percent, 8.2 percent and 5.0 percent, respectively. Serumfree T4 concentrations were high in 10.5 percent, 8.2 per-cent and 9.5. percent, respectively. Serum TSH concentra-tions were low in 7.9 percent, 7.3 percent and 10.9 percent,respectively. Among the patients with low serum TSH val-ues, most had high serum free T3 or T4 values. Consideringserum free T3 concentrations as a continuous variable, theodds ratio for coronary artery disease at the time of hospi-talization in those patients who had a high serum free T3

concentration was 2.6 (95 percent confidence interval, 1.3to 5.2, P=0.007).

Among the 181 patients with coronary artery disease whowere evaluated 3 years later, 22 had some subsequent coro-nary event. As compared with the patients who had nocoronary event,, more patients in the coronary-event grouphad high baseline serum free T3 concentrations (22.7 per-cent vs. 6.9 percent, odds ratio 4.0) and low baseline serumTSH concentrations (22.7 percent vs. 12.6 percent, oddsratio 3.3); the proportion with high baseline serum free T3concentrations was similar (9.1 percent and 9.4 percent).

Conclusion More patients hospitalized for angina ormyocardial infarction have biochemical evidence of hyper-thyroidism, especially high serum free T3 concentrations,than do other patients seeking emergency care. Among theangina-myocardial infarction group, high serum free T3 andlow serum TSH concentrations are risk factors for coronaryevents during follow-up.

COMMENTARY

It is difficult to know how to inter-pret these results. The measurementswere done only once, no actual hormon-al values are given, and unexplained (andunexplainable) abnormalities are com-mon in acutely ill patients. The findingthat high serum free T3 concentrationswere associated with angina or myocar-dial infarction at the time of admission(10.5 percent of patients) may bebecause the proportion of patients with-out angina or myocardial infarction whohad high serum free T3 concentrationswas low (5.0 percent), since the propor-tions of patients in the two groups withhigh serum free T4 and low TSH con-centrations were similar. If morepatients in the latter group had been sicklonger, were sicker, or were taking drugsthat inhibit extrathyroidal conversion ofT4 to T3, then more of them would

have normal or even low serum free T3concentrations (no data regarding lowvalues are given).

Looked at broadly, these results donot suggest that abnormalities compati-ble with hyperthyroidism are more com-mon in patients presenting to emergencydepartments with angina or myocardialinfarction. Possibly relevant to this con-clusion is the fact that in patients withcoronary artery disease a high dose ofT3 given immediately after bypass sur-gery does not precipitate coronaryevents (1,2). The follow-up data in theangina-myocardial infarction group aremore suggestive of a relationshipbetween hyperthyroidism and futurecoronary events, but this conclusion isweakened substantially by the problemsmentioned above.

Robert D. Utiger, M.D

References

1. Klemperer JD, Klein I, Gomez M, etal. Thyroid hormone treatment aftercoronary bypass surgery. N Eng J Med1995;333:1522-7.

2. Bennett-Guerrero E, Jimenez JL,White WD, et al. Cardiovascular effectsof intravenous triiodothyronine inpatients undergoing coronary arterybypass graft surgery. A randomized,double-blind, placebo-controlled trial.JAMA 1996;275:687-92.

HYPERTHYROIDISM

High serum triiodothyronine concentrations are a risk factor for coronaryevents

Peters A, Ehlers M, Blank B, Exler D, Falk C, Kohlmann T, Fruehwald-Schultes B, Welhoener P, Kerner W, Fehm HL.Excess triiodothyronine as a risk factor for coronary events. Arch Intern Med 2000;160:1993-9.


SUMMARY

Background Hyperthyroidism has been associated withabnormalities in tests of liver function, but only rarely withliver disease. Whether the frequency of these abnormalitieshas changed as more sensitive tests for identifying patientswith hyperthyroidism is not known.

Methods Serum total alkaline phosphatase, aspartateaminotransferase, alanine aminotransferase, gamma-glu-tamyl transferase and bilirubin were measured at the time ofdiagnosis in 30 consecutive ambulatory patients with hyper-thyroidism caused by Graves' disease. There were 27women and 3 men, ranging in age from 25 to 60 years. Notall patients had all tests, and the isoforms of serum alkalinephosphatase were not measured separately.

Results Overall, 15 patients (50 percent) had high serumconcentrations of one or more of the enzymes or bilirubin.Four patients had only high serum alkaline phosphataseconcentrations.

The results of the individual tests are shown in the table.

Conclusion Minor abnormalities in hepatic function testsare present in approximately one third of patients withhyperthyroidism.

COMMENTARY

The frequency of abnormalities inliver function in these 30 patients, asassessed by these tests and excludingserum alkaline phosphatase because ithas several sources, was very similar tothat found in two similar studies in the1970s and 1980s (1,2). No details wereprovided about the severity of thehyperthyroidism or the presence ofother factors that might contribute toliver dysfunction in these 30 patients,but it seems reasonable to assume thatthey had what might be called ‘typical'hyperthyroidism as encountered amongoutpatients in the 1990s. If oneassumes that it was easier to identifypatients with hyperthyroidism in the1990s and that hyperthyroidism alonecaused the abnormalities, then a lowerfrequency of abnormalities might havebeen expected.

Liver dysfunction in patients with

hyperthyroidism has been explained onthe basis of hypoxia (increased need foroxygen but no increase in splanchnicblood flow), hepatic congestion second-ary to cardiac dysfunction (the frequencyof abnormal tests is indeed higher inhyperthyroid patients who also havecongestive heart failure [2]), and - inpatients with Graves' hyperthyroidism -autoimmune hepatitis. The abnormali-ties are reversible (1).

There is no evidence that the pres-ence of these minor abnormalities pre-dicts the occurrence of hepatotoxicity inpatients treated with an antithyroid drug,so there is no reason to assess liver func-tion in patients with hyperthyroidism.When the tests are done, for whateverreason, and minor abnormalities arefound, no further action need be takenuntil the patient's hyperthyroidism hasbeen treated.


References

1. Thompson P Jr, Strum D, Boehm T,et al. Abnormalities of liver functiontests in thyrotoxicosis. Mil Med 1978;143:548-51.

2. Fong T-L, McHutchinson JG,Reynolds TB. Hyperthyroidism andhepatic dysfunction. A case series analy-sis. J Clin Gastroenterol 1992;14:240-4.

HYPERTHYROIDISM

Liver dysfunction in patients with hyperthyroidism

Biscoveanu M, Hasinski S. Abnormal results of liver function tests in patients with Graves' disease. Endocr Pract2000;6:367-9.


Table. Liver Function Test Results in Patients with Hyperthyroidism.

Serum Test Values Normal No. Abnormalin Patients Range /Tested (%)

Alkaline phosphatase (IU/L) 124-283 0-117 10/30 (33)Aspartate aminotransferase (U/L) 36-71 0-35 5/30 (17)Alanine aminotransferase (U/L) 45-157 0-40 6/23 (26)γ-Glutamyl transferase (U/L) 69-331 0-68 6/25 (24)Total bilirubin (mg/dL) 2.5 0.1-0.5 2/24 (8)

To convert serum bilirubin values to µmol/L multiply by 17.1.

SUMMARY

Background Occasional patients with hyperthyroidismtreated with an antithyroid drug develop vasculitis withantineutrophil cytoplasmic antibodies, but the frequencywith which these antibodies can be detected in patients withhyperthyroidism before and during treatment withpropylthiouracil or methimazole is not known.

Methods Serum antineutrophil cytoplasmic antibodies(ANCA) were measured in three groups of patients withhyperthyroidism caused by Graves' disease. One groupconsisted of 42 patients who had not received any treat-ment, a second group of 56 patients who had receivedpropylthiouracil for 3 to 138 months (mean 57), and a thirdgroup of 21 patients who had received methimazole for 6to 146 months (mean 44). The patients' serum was testedfor ANCA reacting with both myeloperoxidase (MPO) andproteinase-3 using immunoassay kits. These antibodies stainthe perinuclear (P-ANCA) and cytoplasmic (C-ANCA)region of neutrophils, respectively, when tested by immuno-fluorescence.

Results Using two different immunoassay kits, high serumMPO-ANCA concentrations were found in 18 (32 percent)and 13 (23 percent) of the 56 patients treated with

propylthiouracil, but none of the untreated patients andnone of the patients treated with methimazole. Among thepatients treated with propylthiouracil, the results of the twoassays were correlated (P<0.001). The frequency of highconcentrations increased with the duration of propylth-iouracil treatment; the mean (±SD) duration of treatmentwas 78 ± 48 months in the patients with high serum MPO-ANCA concentrations and 46 ± 34 months in those withnormal concentrations.

Two of the patients with high serum MPO-ANCA concen-trations had been tested before treatment, and were normal.Nine of the patients in this group had myalgia, arthralgia,purpura and recurrent influenza-like symptoms, but nonehad proteinuria.

Approximately 65 percent of the patients in each group hadhigh serum antithyroid peroxidase antibody concentrations,and no patient in any group had a high serum concentrationof the proteinase-3 ANCA.

Conclusion A substantial proportion of patients withGraves’ hyperthyroidism who are treated with propylth-iouracil have high serum MPO-ANCA concentrations, andsome of them have symptoms consistent with vasculitis.

COMMENTARY

Vasculitis - a better term is micro-scopic polyangiitis - with the symptomslisted above, fever, glomerulonephritisand high serum concentrations of MPO-ANCA is a reasonably well-defined syn-drome. It is a known although infre-quent complication of antithyroid drugtherapy. A 1999 summary identified 27patients with hyperthyroidism treatedwith an antithyroid drug who developedvasculitis with high serum ANCA con-centrations (nearly all the antibodies wereMPO-ANCA) (1). Among them all butthree were taking propylthiouracil. Thecause of hyperthyroidism was Graves'disease in 9 patients, a multinodular goi-ter in 1 and unstated in 17; 48 percentwere Japanese.

Japanese patients with hyperthy-roidism seem to be more likely to have not only vasculitis but also high serumANCA concentrations. Evidence for the

latter comes from both this study by Seraet al and another study of 51 Japanesechildren and adolescents with Graves'hyperthyroidism (age range 3 to 15 yearsat the time of diagnosis) (2). Amongthese 51 patients, high serum MPO-ANCA concentrations were found in 1of 16 untreated patients (6 percent), 16of 25 patients treated with propylth-iouracil (64 percent) (mean duration oftreatment 4.0 ± 3.6 years) and 0 of 10patients treated with methimazole (meanduration 2.1 ± 2.8 years). However, nosurveys of comparable size from else-where have been reported.

Given the apparent rarity of vasculi-tis, the results of these two studies mightbe interpreted to mean that MPO-ANCA are necessary but not sufficientto cause vasculitis. Whether true or not,the association of both the antibodiesand the disease with propylthiouracil isstriking, and another reason - in additionto ease of administration, rapidity ofaction and probably lower frequency of

other side effects - to prefer methimazolefor patients with hyperthyroidism whochoose to be treated with an antithyroiddrug for a prolonged period.


References

1. Gunton JE, Stiel J, Caterson RJ, et al.Anti-thyroid drugs and antineutrophilcytoplasmic antibody positive vasculitis.A case report and review of the litera-ture. J Clin Endocrinol Metab 1999;84:13-6.

2. Sato H, Hattori M, Fujieda M, et al.High prevalence of antineutrophil cyto-plasmic antibody positivity in childhoodonset Graves' disease treated withpropylthiouracil. J Clin EndocrinolMetab 2000;85:4270-3.

HYPERTHYROIDISM

Propylthiouracil, but not methimazole, is associated with antineutrophilcytoplasmic antibodies

Sera N, Ashizawa K, Ando T, Abe Y, Ide A, Usa T, Tominaga T, Ejima E, Yokoyama N, Eguchi K. Treatment withpropylthiouracil is associated with appearance of antineutrophil cytoplasmic antibodies in some patients with Graves'disease. Thyroid 2000;10:595-9.


SUMMARY

Background Pregnant women taking an antithyroid drugare usually advised not to nurse their infants after deliverybecause their milk might contain sufficient amounts of thedrug to cause hypothyroidism in the infant. However, thedata to support this advice are sparse.

Methods Serum thyrotropin (TSH), free thyroxine (T4)index and free triiodothyronine (T3) index were measuredmonthly for 6 months in 51 mothers taking 5 mg methima-zole daily after delivery and their breast-fed infants. Allthese women had received methimazole during pregnancy.The same measurements were done before and serially for12 months during methimazole treatment in 88 motherswho developed hyperthyroidism 2 to 8 months after deliv-ery. 46 of the mothers were treated with 10 mg daily for 2months and 5 or 10 mg daily thereafter, and 42 were treat-ed with 20 mg daily for 1 month, 10 mg for 1 month and 5to 10 mg daily thereafter. Fourteen of the infants whosemothers had received methimazole and 17 other infants ofsimilar sex, age and socioeconomic status had tests of intel-lectual development at age 48 to 74 months.

Results The 51 infants who were nursed from the time ofdelivery had normal serum TSH concentrations and free T4

index and free T3 index values at all times. Among the 88mothers in whom treatment was begun after pregnancy,most had normal serum free T4 index and free T3 index val-ues after treatment for 2 months and all had normal valuesafter 3 months. The infants of these 88 mothers had nor-mal values for all tests before and at all times during treat-ment of their mothers. This includes the tests done in 6infants at a time when their mothers had transient hypothy-roidism.

There were no differences in the full-scale, verbal and per-formance intelligence quotient (IQ) scores (mean IQ score103 in both groups) or the subscales of the verbal and per-formance IQ tests, for example the information, vocabu-lary, sentences, picture completion and mazes subscales.

Conclusion Thyroid function is normal in breast-fedinfants whose mothers are taking low or moderate doses ofmethimazole, as is their intellectual development at 48 to 74months of age.

COMMENTARY

The still-widespread recommenda-tion (1) that women who take an antithy-roid drug not nurse their infants seems tobe based primarily on the presumptionthat any amount of drug would be harm-ful for infants. With respect to drug con-centrations in milk, the little evidencethat is available suggests that breast milkcontains relatively more of a dose ofmethimazole than of propylthiouracil(milk to plasma ratio 1 versus <0.1), butthe amount of either drug a breast-fedinfant receives is very small. Further,small studies have revealed no deleteriouseffects of either drug in breast-fedinfants, even infants whose mothers weretaking 750 mg propylthiouracil daily (2,3).

This study, the largest yet reported,provides further evidence that breast-fedinfants of mothers who are taking smallor moderate doses of methimazole haveno changes in thyroid function or inintellectual development. Parallel studies

were not done in infants of mothers withhyperthyroidism taking methimazole whowere not breast fed, but it is very hard tobelieve their thyroid function would havebeen more normal. The results of thestudies of intellectual development, albeitof relatively few infants, are consistentwith the findings of normal thyroid func-tion during breast-feeding in theseinfants.

Women with hyperthyroidism whoare taking moderate doses of eithermethimazole or propythiouracil can beencouraged to breast feed their infants,like other women.


References

1. Lee A, Moretti E, Collantes A, et al.Choice of breastfeeding and physician'sadvice: a cohort study of women receiv-ing propylthiouracil. Pediatrics 2000;106:27-30.

2. Lamberg B-A, Ikonen E, OsterlundK, et al. Antithyroid treatment of mater-nal hyperthyroidism during lactation.Clin Endocrinol 1984;21:81-7.

3. Momotani N, Yamashita R, Makino F,et al. Thyroid function in wholly breast-feeding infants whose mothers take highdoses of propylthiouracil. Clin Endo-crinol 2000;53:177-81.

HYPERTHYROIDISM

Mothers taking methimazole can nurse their infants safely

Azizi F, Khoshniat M, Bahrainian M, Hedayati M. Thyroid function and intellectual development of infants nursed bymothers taking methimazole. J Clin Endocrinol Metab 2000;85:3233-8.


SUMMARY

Background Some patients with Graves' ophthalmopathyhave high intraocular pressures, which may be caused byorbital venous congestion, compression of the eye byenlarged or fibrotic extraocular muscles, or orbital inflam-mation, or it may be coincidental. There is little informa-tion about the effect of orbital decompression, extraocular-muscle surgery and orbital radiation therapy on intraocularpressure.

Methods The study subjects were patients with Graves'ophthalmopathy who underwent decompression of themedial and inferior orbital walls (80 patients, 116 eyes) orrecession of the inferior and medial rectus muscles (24patients, 32 eyes), or who received orbital radiation therapy(2000 cGy in 10 daily fractions) (28 patients, 56 eyes).Intraocular pressure was measured just before treatmentand 2 to 6 weeks after surgery or 6 to 12 weeks after the lastradiation treatment. Patients taking a glucocorticoid ormedication for glaucoma were not excluded unless theirtherapy was changed during the observation period.

Results The preoperative and postoperative intraocularpressure results in the three groups are shown in the table.

The decreases in intraocular pressure on forward gaze andalso on upward gaze in the decompression and extraocularmuscle surgery groups were highly statistically significant(P<0.01). Among the 116 patients in the decompressiongroup, the fall in intraocular pressure was greater in the 65patients who had values ≥21 mm Hg before surgery and inthe 35 patients who were receiving topical therapy for glau-coma.

Conclusion In patients with Graves' ophthalmopathyorbital decompression surgery and extraocular muscle sur-gery reduces intraocular pressure.

COMMENTARY

Most if not all patients with Graves'ophthalmopathy who undergo orbitaldecompression or extraocular musclesurgery have severe ophthalmopathy,however that is defined, usually withconsiderable inflammation and edema ofthe extraocular muscles and adipose tis-sue and therefore venous congestionwithin the orbit. These changes raiseepiscleral pressure, which in turn slowsremoval of fluid by the trabecular net-work of the eye and raises intraocularpressure. It is reassuring that the twooperations resulted in a fall in intraocularpressure, but unfortunate that thepatients were not followed longer. It isperhaps surprising that orbital radiationtherapy did not lower intraocular pres-sure, since it can reduce retrobulbarinflammation. There was no control

group in this study to determine theeffect of time alone on intraocular pres-sure, but given the characteristics of thepatients assembling a control groupmight be difficult.

The frequency of intraocular hyper-tension, usually defined as intraocularpressure greater than 20 or 22 mm Hg, isincreased in patients with Graves' oph-thalmopathy, at least in those patientswho are referred to ophthalmologists.For example, in a survey of 500 consec-utive patients with Graves' ophthalmopa-thy, 120 (24 percent) had intraocularpressure values greater than 22 mm Hgbut less than 30 mm Hg (1); the frequen-cy in the general population was 5 per-cent in one study (2). It is important tonote that intraocular hypertension is notsynonymous with glaucoma; that diagno-sis is based on the presence of a highcup-to-disk ratio and progressive visual

field deficits and not on the intraocularpressure alone. Glaucoma is not oftenmentioned in studies of Graves' ophthal-mopathy, but the possibility of its occur-rence is one reason why patients withmore than minimal Graves' ophthal-mopathy should be evaluated by anophthalmologist.


References

1. Cockerham KP, Pal C, Jani B, et al.The prevalence and implications of ocu-lar hypertension and glaucoma in thy-roid-associated orbitopathy. Ophthal-mology 1997;104:914-7.

2. Sommer A. Intraocular pressure andglaucoma. Am J Ophthalmol 1989;107:186-8.

GRAVES' OPHTHALMOPATHY

Surgical and radiation therapy lowers intraocular pressure in patients withGraves' ophthalmopathy

Danesh-Meyer HV, Savino PJ, Deramo V, Sergott RC, Smith AF. Intraocular pressure changes after treatment forGraves' orbitopathy. Ophthalmology 2001;108:145-50.


Table. Mean Intraocular Pressure Values before and after Treatment(Forward Gaze).

Decompression Extraocular Muscle RadiationSurgery Surgery Therapymm Hg mm Hg mm Hg

Pretreatment 21.6 18.5 19.4Post-treatment 17.5 16.1 18.4Difference (%) -4.1 (-18.9) -2.4 (-13.3) -1.0 (-5.1)

SUMMARY

Background Patients with mild thyroid failure, defined thesame way as is subclinical hypothyroidism (high serum thy-rotropin (TSH) and normal serum free thyroxine (T4) con-centrations), may have high serum total and low-density-lipoprotein (LDL) cholesterol concentrations. However,the extent to which the values fall when the patients aretreated with T4 has varied considerably in different studies.

Methods In an extensive review of the literature theauthors identified 13 studies in which serum total choles-terol was measured prospectively in a total of 247 patientswith subclinical hypothyroidism before and during treat-ment with T4. Serum LDL cholesterol and high-density-lipoprotein (HDL) cholesterol and other lipids also weremeasured in some of the studies. The studies were evaluat-ed using 13 criteria of validity, including randomized con-trolled design, treatment for at least 3 months, mean serumTSH concentration <2 mU/L during treatment, individualdosing of T4, and recruitment from a population-basedsample.

Results No study fulfilled all 13 of the criteria of validity.For example, only 3 were randomized trials, in only 6 wasthe mean serum TSH concentration <2 mU/L at the timeof follow-up, and in only 1 were the study subjects drawnfrom a population-based sample.

The mean baseline serum total cholesterol concentration inthe 13 studies was 241 mg/dL (6.2 mmol/L). During T4treatment it decreased in 11 studies and increased in 2 stud-ies. The overall mean decrease was 7.9 mg/dL (0.2mmol/L), and it was greater in patients with higher initialvalues. The baseline serum LDL cholesterol concentration,measured in 9 studies, was 158 mg/dL (4.1 mmol/L). Dur-ing T4 treatment it decreased in 7 studies and increased in 2;the overall mean decrease was 10 mg/dL (0.3 mmol/L).The mean baseline serum HDL cholesterol concentration,measured in 10 studies, was 51 mg/dL (1.3 mmol/L). Dur-ing T4 treatment; it increased in 5 of the studies; the overallmean increase was 3.1 mg/dL (0.1 mmol/L). The 95 per-cent confidence intervals for the overall change excluded 0(no change) for the 3 measurements. There was no changein serum triglyceride concentrations, measured in 12 studies.

Conclusion T4 therapy lowers serum total and LDL cho-lesterol concentrations slightly in patients with subclinicalhypothyroidism.

COMMENTARY

Whether patients with subclinicalhypothyroidism should be treated withT4 is debated. That treatment mightlower serum total and LDL cholesterolconcentrations is one argument in favorof treatment, but its efficacy in thisregard has been debated. This articledescribes a systematic review of the lit-erature on the effect of T4 treatment onserum lipid concentrations in thesepatients. The criteria used to evaluate thevalidity of the studies might be consid-ered utopian, so it is perhaps not sur-prising that no study met all of them.

This study focused on changes dur-ing treatment, not the baseline values.The baseline values in the different stud-

ies ranged from 196 to 310 mg/dL (5.1to 8.0 mmol/L) for serum total choles-terol and from 123 to 204 mg/dL (3.2 to5.3 mmol/L) for serum LDL choles-terol. The values on the whole werehigher than would be expected in nor-mal subjects. However, many patientswith subclinical hypothyroidism havenormal values. As an example, in arecent study in Austria, the mean serumcholesterol concentration in 4866 nor-mal subjects was 217 mg/dL (5.6mmol/L) and it was 219 mg/dL (5.7mmol/L) in 1055 patients with subclini-cal hypothyroidism, the largest groupever reported (1). So one cannot pre-scribe T4 for patients with subclinicalhypothyroidism on the assumption thatthey have high serum cholesterol con-

centrations, but if the concentrations arehigh T4 treatment will lower them slight-ly.


References

1. Vierhapper H, Nardi A, Grosser A,et al. Low-density lipoprotein choles-terol in subclinical hypothyroidism.Thyroid 2000;10:981-4.

HYPOTHYROIDISM

Thyroxine therapy lowers serum cholesterol concentrations in patients withsubclinical hypothyroidism

Danese MD, Ladenson PW, Meinert C, Powe NR. Effect of thyroxine therapy on serum lipoproteins in patients withmild thyroid failure: a quantitative review of the literature. J Clin Endocrinol Metab 2000;85:2993-3001.


HYPOTHYROIDISM

Hypothyroidism caused by rapid thyroid hormone destruction in an infantwith hepatic hemangiomas

Huang SA, Tu HM, Harney JW, Venihaki M, Butte A, Kozakewich HPW, Fishman SJ, Larsen PR.. Severe hypothy-roidism caused by type 3 iodothyronine deiodinase in infantile hemagiomas. N Engl J Med 2000;343:185-9.

SUMMARY

Background Thyroid hormone degradation is not thoughtto be an important determinant of thyroid hormone pro-duction. When changes in degradation do occur compen-sation by the pituitary-thyroid system usually maintains thy-roid secretion within the normal range. The infantdescribed in this report had multiple hepatic hemagiomasthat degraded thyroid hormone so rapidly that the infanthad hypothyroidism.

Care Report A 6-week-old boy developed abdominal dis-tension, and was found to have hepatomegaly. Liver biop-sy revealed hemangioma. The infant's serum thyrotropin(TSH) concentration was 156 mU/L and his serum free thy-roxine (T4) concentration was low. He was treated withprednisolone and T4.

At age 3 months, he was hospitalized for increasing abdom-inal distension and respiratory distress. He had bradycardia,hypothermia and massive hepatomegaly; the thyroid glandwas not enlarged. The liver contained multiple large hema-giomas. Serum T4 was 2.5 µg/dL (32 nmol/L), triiodothy-ronine (T3) <15 ng/dL (0.23 nmol/L), reverse T3 413ng/dL (6.4 nmol/L), TSH 177 mU/L and thyroglobulin1014 ng/mL (normal 6 to 87).

The infant was treated with mechanical ventilation, 6-methylprednisolone, interferon-2 alfa and intravenous T3.His clinical condition improved, and his serum TSH con-centration 27 hours later was 79 mU/L. In the next 18 days,he received intravenous T3 continuously in doses up to 96µg/day, and intravenous and then jejunal T4 in doses up to50 µg/day. This treatment raised the infant's serum T3 con-centration to normal in 5 days and lowered his serum TSH

concentration to normal in 17 days; both remained normalthereafter. Serum T4 concentrations remained low. Serumreverse T3 concentrations initially declined when T3 wasgiven, but then increased to near the baseline value when T4was added.

The hepatic hemangiomas were treated by embolization.Staphlococcal bacteremia was diagnosed on hospital day 19.The infant's clinical condition continued to deteriorate, andhe died 40 days after admission

Special Studies Studies of hemagioma tissue from theinfant revealed a very high level of type 3 deiodinase activ-ity. The maximal velocity of deiodination of T3 was 0.78nmol/mg protein, 7.5 times higher than is found in placen-ta, the normal tissue with the highest activity. The tissuealso contained a large quantity of mRNA for this deiodi-nase; in situ hydridization studies revealed the mRNA to bein the hemangioma cells.

A review of hospital records revealed that, among 1555children with hemangiomas, serum TSH had been meas-ured in 92. Two of them had values two or more times theupper limit of normal for age and low serum T4 concentra-tions; both had massive hepatic hemagiomas. No tissuefrom these children was available for study.

Assays of hemangioma tissue from 5 other childrenrevealed type 3 deiodinase activity in amounts comparableto that in placental tissue in 3.

Conclusion Some hemagiomas contain large amounts oftype 3 iodothyronine deiodinase activity, which in one infantwith a large tumor deiodinated both T4 and T3 sufficientlyrapidly to cause hypothyroidism.


COMMENTARY

Infants with congenital hypothy-roidism are usually treated with T4 in adose of 25 µg/kg/day, about 35 percentof which is converted to T3. Thus, thisinfant's requirement for thyroid hor-mone was vastly greater than expected.Furthermore, this treatment did notraise the infant's serum T4 concentra-tion, it raised his serum T3 concentra-tion much less than expected, and itraised his serum reverse T3 concentra-tion to well above normal. Another fac-tor undoubtedly contributing to the lowserum T3 concentrations was the infant'ssevere nonthyroidal illness.

The results pointed to rapiddestruction of both T4 and T3, and thusto an increase in the activity of type 3deiodinase, which catalyzes inner ringdeiodination (5-deiodination) of eachhormone to form, respectively, reverseT3 and 3,3 -diiodothyronine. This wasconfirmed by the invitro studies. Thepresence of a high level of type 3 deiod-inase activity in the hemangioma tissueof this and several other children is aunique finding.

The infant's thyroidal secretion wascertainly highly stimulated, witness thehigh serum TSH and thyroglobulin con-centrations, but was unable to producesufficient quantities of T4 and T3 to pre-

vent hypothyroidism. No other situationin which T4 and T3 are so rapidlydegraded that a person who has normalpituitary-thyroid function does notremain euthyroid has been described.This study will undoubtedly stimulatemore studies of pituitary-thyroid func-tion in patients with not only heman-giomas but also other tumors. It may bethat hypothyroidism due to rapid tumor-mediated degradation of T4 and T3 isnot so rare.


SUMMARY

Background Hypothyroidism in pregnant women hasbeen associated with several maternal and fetal complica-tions of pregnancy in observational studies. However, therehave been no population-based studies in which the out-come of pregnancy was related to the presence of hypothy-roidism during the pregnancy.

Methods Serum thyrotropin (TSH) was measured between15 and 18 weeks gestation and the outcome of pregnancywas determined from state records in 9403 women with sin-gleton pregnancies undergoing screening for neural tubedefects and Down's syndrome in Maine between 1990 and1992. Serum free thyroxine (T4) was measured in thewomen with serum TSH concentrations ≥6 mU/L andtwice that number of women with serum TSH concentra-tions <6 mU/L.

Results Among the 9403 pregnant women, 209 (2.2 per-cent) had serum TSH concentrations ≥6 mU/L. The valueswere 6 to 9.9 mU/L in 172 women (1.8 percent) and ≥10mU/L in 37 women (0.4 percent). The two groups ofwomen with high serum TSH concentrations had lowermean serum free T4 concentrations than the women who

had normal serum TSH concentrations, although theirmean values were within the normal range. In addition,more of the women with high serum TSH concentrationshad high serum antithyroid peroxidase antibody concentra-tions (60 percent versus 9 percent).

Among the 9194 women with normal serum TSH values,there were 83 fetal deaths (0.9 percent), as compared with 8fetal deaths among the 209 women with serum TSH values≥6 mU/L (3.8 percent) (odds ratio 4.4, 95 percent confi-dence interval, 1.9 to 9.5). The frequency of fetal death washigher in the women with serum TSH values ≥10 mU/L(8.1 percent) than in those with values of 6 to 9.9 mU/L(2.9 percent). There were no differences in the frequency ofgestational hypertension, abruptio placenta or cesareandelivery in the women in the three groups, or in gestationalage at delivery, mean birth weight, Apgar scores at 5 min-utes or neonatal deaths.

Conclusion The rate of fetal death but not other maternalor fetal outcomes is increased in women who have highserum TSH concentrations during the second trimester ofpregnancy.

COMMENTARY

Most of the women with highserum TSH concentrations hadsubclinical hypothyroidism, and theywere studied only once, so it might havebeen transient. Nonetheless, the findingof an increasing rate of fetal loss as afunction of the degree of elevation inserum TSH concentration (and decreasein serum free T4 concentration) suggeststhat mild hypothyroidism has a deleteri-ous effect on fetuses.

The lack of an increased frequencyof gestational hypertension, abruptioplacenta, premature birth and low birth-weight, as reported in observationalstudies of women with hypothyroidism(1), may be due to differences in severityof hypothyroidism or to differences inascertainment of the complications ofpregnancy. Overt hypothyroidism isassociated with increased peripheral vas-cular resistance, and might be expectedto reduce placental blood flow, butwhether the same change occurs in sub-

clinical hypothyroidism is not known. The evidence from this study that

hypothyroidism in pregnant women isassociated with decreased fetal survivalcomplements the results of a study ofan earlier cohort of women in Maine(2). That study revealed that the infantsof women who were hypothyroid at 15to 18 weeks gestation had slightly lowerscores on intelligence tests at age 7 yearsthan the infants of normal women.These new results may strengthen theargument for screening all pregnantwomen for hypothyroidism, but it wouldseem prudent to determine if interven-tion is beneficial before widespreadscreening is undertaken.


References

1. Leung AS, Millar LK, Koonings PP,et al. Perinatal outcome in hypothyroidpregnancies. Obstet Gynecol 1993;81:349-53.

2. Haddow JE, Palomaki GE, AllanWC, et al. Maternal thyroid deficiencyduring pregnancy and subsequent psychological development in the child.N Engl J Med 1999;341:549-55.

HYPOTHYROIDISM

Fetal loss is increased in pregnant women with hypothyroidism

Allan WC, Haddow JE, Palomaki GE, Williams JR, Mitchell ML, Hermos RJ, Faix JD, Klein RZ. Maternal thyroid defi-ciency and pregnancy complications: implications for population screening. J Med Screening 2000;7:127-30.


SUMMARY

Background Iodine-123 is considered the radionuclide ofchoice for studying patients with thyroid nodular disease,but has not been widely used in patients with thyroid carci-noma.

Methods Total body scans were done in 14 consecutivepatients with differentiated thyroid carcinoma within 7weeks after near-total thyroidectomy. All the patients werehypothyroid (serum thyrotropin concentrations greater than40 mU/L) and had eaten a low-iodine diet for 1 week. Thefirst scan was done 5 hours after oral administration of 48to 56 MBq (1.3 to 1.5 mCi) iodine-123 (I-123). The patientsthen were given 111 MBq (3 mCi) iodine-131(I-131) and another scan was done 42 to 44 hours later. Thepatients who had uptake in the neck were given 3700 to5500 MBq (100 to 150 mCi) I-131 and a third scan was done7 days later.

Results Thirteen patients had at least one focus of uptakein the thyroid bed or neck on the diagnostic scans, and over-all 35 foci were seen. All 35 foci were seen on the I-123scans, as compared with 32 foci on the I-131 scans. The 3foci not seen on the latter scans were in different patients.The scans done after I-131 therapy were similar in 11patients. In 1 patient a new focus of uptake was seen in the

lateral neck; this focus was not seen on either diagnosticscan and proved to be carcinoma. In another patient uptakein the left thyroid bed on the diagnostic scans was no longerseen. The I-123 scans were judged to be of better quality.

Conclusion Iodine-123 is superior to iodine-131 for imag-ing the neck in patients with differentiated thyroid carcino-ma.

COMMENTARY

The strength of this small study isthat the diagnostic I-123 and I-131 scanswere done in the same patients. Theresults therefore add in a unique way tothe existing evidence that diagnosticscanning in patients with thyroid carcinoma should be done with I-123.As compared with I-131 scans, I-123scans reveal a few more foci of uptake.Futhermore, I-123 does not damage ( 'stun' ) thyroid tissue so that itbecomes more difficult to destroy with ahigh dose of I-131 (1), it is safer, and ityields better-quality scans. It is alsoabout twice as expensive and less readilyavailable than I-131.

The 'stunning' phenomenon seemsto apply more to thyroid remnants thancarcinoma. Therefore, the benefit of I-123 for scanning may be greatest for thefirst scan. Of course, the risk of stun-ning can be avoided entirely by notdoing a scan before I-131 is given todestroy thyroid remnants, but this meanssome patients would be treated unneces-sarily.


References

1. Park HM, Park YH, Zhou XH.Detection of thyroid remnant/metasta-sis without stunning: an ongoing dilem-ma. Thyroid 1997;7:277-80.

THYROID CANCER

Iodine-123 is superior to iodine-131 in detecting thyroid remnants in patientswith differentiated thyroid carcinoma

Mandel SJ, Shankar LK, Benard F, Yamamoto A, Alavi A. Superiority of iodine-123 compared with iodine-131 scanningfor thyroid remnants in patients with differentiated thyroid carcinoma. Clin Nucl Med 2001;26:6-9.


THYROID CANCER

Presence of PAX8-PPARγγ1 gene translocation distinguishes follicular carci-noma from other follicular lesions

Kroll TG, Sarraf P, Pecciarini L, Chen C-J, Mueller E, Spiegelman BM, Fletcher JA. PAX8-PPARγ1 fusion in oncogenehuman thyroid carcinoma. Science 2000;289:1357-60.

COMMENTARY

Approximately 10 to 20 percent offine-needle aspiration biopsies of thy-roid nodules are assigned terms such asfollicular tumor, follicular lesion, andfollicular neoplasm (1,2). The majorcytological characteristics of these nod-ules are a large number of follicularcells, the presence of microfollicles, andlittle or no colloid. The term "suspi-cious" is applied to biopsy results havingthese characteristics in some studies, butthat term is also applied to biopsies thathave cells with some but not all of thecharacteristics of papillary carcinomas.In short, the terminology used to cate-gorize the cytological findings in fine-needle aspiration biopsies is not stan-dardized.

From 10 to 25 percent of folliculartumors prove to be follicular carcinomasand follicular variants of papillary carci-noma when removed surgically andexamined histologically (1,2). Theremainder are follicular adenomas and

hyperplastic nodules. In some patientswho undergo surgery the nodule isthought to be benign by the surgeon andpathologist on the basis of examinationof frozen sections, and therefore only athyroid lobectomy is done, but subse-quent pathological examination revealsfollicular carcinoma (3). Then, a secondoperation to remove the contralateralthyroid lobe is necessary.

Given this background, a reliableprocedure that allowed cells of follicularcarcinomas to be identified in thyroidcells obtained by fine-needle aspirationbiopsy would be very useful. Althoughthis study was done using tissue sections,it should be feasible to detect the fusiononcogene in isolated cells, by expansionof the DNA with the polymerase chainreaction, in situ hydridization orimmunocytochemistry. The practicalquestion is whether the findings will beconfirmed when many more folliculartumors are studied.


References

1. Gharib H, Goellner JR. Fine-needleaspiration biopsy of the thyroid: anappraisal. Ann Intern Med 1993;118:282-89.

2. Ravetto C, Colombo L, DottoriniME. Usefulness of fine-needle aspira-tion in the diagnosis of thyroid carcinoma. A retrospective study in37,895 patients. Cancer (CancerCytopathol) 2000;90:357-63.

3. Chen H, Nicol TL, Udelsman R.Follicular lesions of the thyroid. Doesfrozen section evaluation alter operativemanagement? Ann Surg 1995;222:101-6.


SUMMARY

Background It is not possible to distinguish betweenbenign and malignant follicular tumors of the thyroid glandon the basis of cytology and it may be difficult to distin-guish between them by histology. Molecular analysis for achromosomal translocation that has been identified in fol-licular carcinomas may be useful in distinguishing betweenbenign and malignant follicular tumors. This translocation,(t(2;3)(q13;p25), results in fusion of part of the DNA bind-ing region of the PAX8 gene to the peroxisome prolifera-tor-activator γ1 (PPARγ1) gene. PAX8 is a transcriptionfactor that plays an important role in the development andfunction of thyroid follicular cells, and PPARγ1 is a tran-scription factor that stimulates cell differentiation andinhibits cell growth.

Methods The presence of PAX8-PPARγ1 DNA and pro-tein was studied in sections of 8 follicular carcinomas, 20follicular adenomas, 10 papillary carcinomas and 10 hyper-plastic nodules by molecular and immunochemical meth-ods. In addition, the biological activity of the fusion pro-

tein was evaluated by determining its ability to activateresponse elements in DNA that are activated by wild-typePPARγ1 in the presence of troglitazone.

Results PAX8-PPARγ1 DNA was detected by reversetranscriptase-polymerase chain reaction analysis in 5 of the8 follicular carcinomas (62 percent), but in none of the fol-licular adenomas, papillary carcinomas or hyperplastic nod-ules. Two of the other follicular carcinomas contained adifferent rearrangement the two genes. Immunochemicalstudies using a monoclonal antibody against PPARγ1revealed strong diffuse staining of the nuclei of 7 of the 8follicular carcinomas, but only faint focal staining of thenuclei of the follicular adenomas, papillary carcinomas andhyperplastic nodules. In transfection studies in culturedcells, PAX8-PPARγ1 was biologically inactive and blockedthe activity of PPARγ1, an effect that would be expected toblock cell differentiation and promote cell growth.

Conclusion Many follicular carcinomas of the thyroidcontain a PAX8-PPARγ1 fusion oncogene.

SUMMARY

Background Thyroid cancer is sufficiently rare that fewinstitutions accumulate large numbers of patients, and thefindings in patients evaluated and treated at those placesmay be biased by the fact that many of the patients werereferred there. Therefore, information on the characteris-tics of the patients evaluated and treated at a large numberof hospitals may provide a broader insight into how thepatients are being treated and what needs to be done to raisethe overall standard of care nationally.

Methods A prospective cohort study was conducted atover 1500 hospitals in the United States in 1996 under theauspices of the American College of Surgeons Commissionon Cancer. The hospitals reported demographic, diagnos-tic-test, pathological and treatment data on 5583 patients.

Results Of the 5583 patients, 79 percent were diagnosedand treated at the same institution. 4522 patients (81 per-cent) had papillary carcinomas, 583 (10 percent) had follic-ular carcinomas, 205 (4 percent) had Hurthle-cell carcino-mas, 177 (3 percent) had medullary carcinomas (15 percentof which were familial), and 96 (2 percent) had anaplasticcarcinomas.

Some characteristics of the patients in the papillary-carci-noma and follicular-carcinoma groups are shown in thetable.

The papillary carcinomas tended to be smaller, 16 percentbeing >3 cm vs. 39 percent of the follicular carcinomas, butwere more often multifocal, 29 percent vs. 16 percent.

With respect to treatment, approximately 75 percent of thepatients in both groups had near-total or totalthyroidectomy, and nearly the same proportions were con-sidered to have no residual tumor after surgery. No lymphnodes were examined at the time of surgery in 54 percentand 77 percent, respectively. Approximately 50 percent ofthe patients received radioiodine at least once and approxi-mately 50 percent received thyroid hormone. The 30-daypostoperative mortality was 0.2 percent.

Conclusion The diagnostic tests done in patients with thy-roid carcinoma and the treatments varied considerably inthe United States in 1996. Areas identified where care needsto be improved include more frequent use of fine-needleaspiration biopsy and preoperative laryngoscopy and morefrequent resection of lymph nodes.

COMMENTARY

The 5583 patients in this studyfrom over 1500 hospitals represent onlyabout one third of the estimated 15,600new cases of thyroid carcinoma in theUnited States in 1996 (1) (the estimatefor 1999 was 18,100 [2]). Where andhow the other 10,000 or so patientswere evaluated and treated is not com-mented upon in the article, and is notobvious.

While the report does not thereforeprovide a complete picture of thyroid

carcinoma in this country in 1996, itnonetheless contains a wealth of infor-mation about all types of thyroid carci-noma. The demographic and clinicalfindings in the patients with the differenttypes of carcinoma were similar to thosein single-center studies. However, someaspects of diagnosis and treatment seemin disagreement with practices believedby many to simplify diagnosis andimprove survival. These include theinfrequent use of fine-needle aspirationbiopsy and the rather frequent use ofincisional biopsy and of CT or

MR imaging, and the rather low frequen-cy of postoperative radioiodine and thy-roid hormone therapy.


References

1. Parker SL, Long T, Bolden S, WingoPA. Cancer statistics, 1996. CA CancerJ Clin 1996;65:5-27.

2. Landis AH, Murray T, Bolden S,Wingo PA. Cancer statistics, 1999. CACancer J Clin 1999;49:8-31.


THYROID CANCER

Clinical findings and treatment of patients with thyroid cancer treated atover 1500 hospitals in the United States in 1996

Hundahl SA, Cady B, Cunningham MP, Mazzaferri E, McKee RF, Rosai J, Shah JP, Fremgen AM, Stewart AK, Holzer S,for the U.S. and German Thyroid Cancer Study Group. Initial results from a prospective cohort study of 5583 cases ofthyroid carcinoma treated in the United States during 1996. Cancer (Cytopathol) 2000;89:202-17.

Table. Characteristics of Patients with Differentiated Thyroid Carcinoma.Characteristic Papillary Carcinoma Follicular Carcinoma Age 30 to 49 years 49% 42%Women 77% 71%History of radiation 5% 4%Family history of thyroid cancer 5% 4%Symptoms and signsThyroid mass 75% 78%Hoarseness or dysphagia 18% 24%Lymphadenopathy 27% 22%Diagnostic testsNeedle aspiration of thyroid 53% 53%Lymph node aspiration 7% 5%Incisional biopsy 13% 11%Thyroid ultrasonography 38% 41%Radionuclide scan 39% 42%CT or MRI of neck 18% 18%Laryngoscopy 6% 6%

COMMENTARY

Paclitaxel is obviously not highlyeffective in patients with anaplasticcarcinoma, but it is a step forward. Theproportion of patients having a partialor complete response was as high orhigher than in patients treated with otherchemotherapy regimens, many of themmore complex and toxic (reviewed in 1).In some patients tumor size decreasedand then increased within the 3-weektreatment interval, a dramaticdemonstration of the rapid growth ofanaplastic carcinomas.

This study serves as a model ofhow chemotherapy regimens ought tobe devised, from tests in cultured tumor

cells to animals to patients. Perhaps theanimal studies can be omitted at somepoint. The first steps along this pathhave been taken with regard to genetherapy, in that the gene for the tumorsuppressor molecule p53 (which is oftenmutated in anaplastic carcinomas) hasbeen inserted into cultured anaplasticcarcinoma cells, which then died (1).


References

1. Ain KB. Anaplastic thyroid carcino-ma: behavior, biology, and therapeuticapproaches. Thyroid 1998;8:715-26.

2. Nagayama Y, Yokoi H, Takeda K,et al. Adenovirus-mediated tumorsuppressor p53 gene therapy foranaplastic thyroid carcinoma in Vitro andin Vivo. J Clin Endocrinol Metab 2000;85:4081-6.

THYROID CANCER

Paclitaxel has some efficacy in patients with anaplastic thyroid carcinoma

Ain KB, Egorin MJ, DeSimone PA, for the Collaborative Anaplastic Thyroid Cancer Health Intervention Trials(CATCHIT) Group. Treatment of anaplastic thyroid cancer with paclitaxel: phase 2 trial using ninety-six-hour infusion.Thyroid 2000;10:587-94.


SUMMARY

Background There is no effective therapy for patientswith anaplastic carcinoma of the thyroid. Pacllitaxel hasactivity against anaplastic carcinoma cells cultured in vitroand grown in nude mice.

Methods Nineteen patients (6 women, 13 men, mean age62 years, range 47 to 86) with anaplastic carcinoma in whomthe tumor persisted or recurred after surgery and radiationtherapy were given continuous intravenous infusions ofpaclitaxel in doses 120 mg/m2 (7 patients) or 140 mg/m2

(12 patients) for 96 hours. Up to six infusions were given at3-week intervals. Eight patients had been treated by totalthyroidectomy and 3 by partial thyroidectomy; theremaining 8 patients had only a diagnostic biopsy. Responsewas assessed by measurements of tumor in the thyroid bedin 14 patients, elsewhere in the neck in 7 patients, in thelungs in 11 patients, and in the mediastinum and abdomenin 5 patients by physical examination and imaging after thesecond, fourth and sixth treatments.

Results Ten of the 19 patients (53 percent) respondedtransiently to paclitaxel. One patient had completedisappearance of tumor and 9 patients had partialresponses (50 percent or more reduction in the sum of thediameters of all measurable tumor deposits); 1 other patienthad stabilization of tumor. The median survival afterinitiation of paclitaxel therapy was 24 weeks (32 weeks in

the patients who responded vs. 10 weeks in those who didnot, P=0.40). Women responded more often than men, andyounger patients more often than older patients. There wasno correlation between plasma paxlitaxel clearance(measured in 10 patients) and responses to therapy. Theadverse effects were nausea, vomiting, diarrhea, fever,alopecia and stomatitis, none of which was severe.

Conclusion Paclitaxel has some efficacy in patients withanaplastic carcinoma of the thyroid.

SUMMARY

Background Thyroid hormones stimulate neural develop-ment early in gestation, but little is known about theontogeny of thyroid hormone receptors in fetal neuraltissue.

Methods RNA was extracted from brain tissue of 13fetuses aborted at from 10 to 16 weeks gestation. The RNAwas reverse transcribed and amplified using primers for theα1, α2, β1 and β2 isoforms of the receptor. In addition,brain tissue obtained from 11 fetuses aborted in the firsttrimester (gestational age 11 to 13 weeks), 12 fetuses abort-ed in the second trimester (gestational age 15 to 25 weeks),21 stillborn fetuses (gestational age 26 to 40 weeks) and 18stillborn fetuses with intrauterine growth retardation (gestational age 23 to 29 weeks, birthweight <10th per-centile) were studied for thyroid receptor isoforms byimmunochemistry using antibodies against each isoform.

Results mRNA for all four isoforms of the thyroid recep-tor was detected in brain tissue from all the 10- to 16-weekfetuses studied. The immunochemical studies revealed

some α1 and α2 receptors but no β1 or β2 receptors in thebrain tissue of first-trimester fetuses. There was staining forall the receptors in the brains of approximately 70 percentof the second-trimester fetuses and all the third-trimesterfetuses. The staining was largely confined to the pyramidalcells of the cerebral cortex and the Purkinje cells of thecerebellum, and there was little staining of glial cells or cellsof the choroid plexus. The proportion of neurons thatstained for each isoform increased progressively withincreasing gestational age, so that by the third trimester allthe neurons in each section of cerebral cortex and cerebel-lum were stained. The intensity of staining for all isoformswas less in the brains of the fetuses with intrauterine growthretardation, as compared with the fetuses of the same ges-tational age.

Conclusion mRNA and protein for all four isoforms ofthe thyroid receptor are present in neurons of the cerebralcortex and cerebellum early in gestation, and the proportionof neurons containing receptor protein increases through-out gestation.

COMMENTARY

The presence of both thyroidreceptor mRNA and protein in the brainvery early in gestation provides strongevidence that the receptors are producedin brain tissue and that the tissue isresponsive to thyroid hormone, andtherefore that the hormone plays a rolein very early development of the brain.In another recent study in which themRNAs for the α1, α2 and β1 isoformsof the thyroid receptor were measuredin brain tissue of fetuses (1), the contentof the α isoforms increased progressive-ly from 8 to 14 weeks, whereas that ofthe β1 isoform was curvilinear, beinglower at 10 to 12 weeks than earlier orlater.

If thyroid receptors are there, theymust have some thyroid hormone,specifically triiodothyronine (T3), to acti-vate them (excluding the α1 isoformthat does not bind T3). Binding of T3may either activate the receptors, so thatthey have transcriptional activity, orreverse the transcriptional activity of

unliganded receptors. Initially, this T3has to come from the mother, eitherdirectly or via deiodination of maternalthyroxine (T4) in the brain of the fetus.

The finding of lower levels of allreceptor isoforms in the brains of thefetuses with intrauterine growth retarda-tion suggests that decreased thyroid hor-mone action could contribute to themental retardation that occurs in someof these infants. Quantitatively, thebrain contains more of the α than the βisoforms, and their distribution is notuniform. This is perhaps most evidentfor the β2 isoform, which is foundmostly in the hypothalamus (and pitu-itary), but it is likely that there areregion-specific differences in the distri-bution of all the isoforms and also thatthe different isoforms are linked to dif-ferent neurochemical reactions. Thelesser abundance of the β isoformsseems to fit with the clinical observa-tions that patients who have inactivatingmutations in the gene for the β isoformsof the receptor have a low incidence ofmental retardation (2). These patients,who have high serum thyroid hormone

concentrations, have a high incidence ofhyperactivity, which could reflect activa-tion of the more abundant α1 isoformof the thyroid receptor.


References

1. Iskaros J, Pickard M, Evans I, et al.Thyroid hormone receptor gene expres-sion in first trimester human fetal brain.J Clin Endocrinol Metab 2000;85:2620-23.

2. Refetoff S. Resistance to thyroidhormone. In Braverman LE, Utiger RD,eds. The thyroid. A fundamental andclinical text. 8th ed. Philadelphia: Lip-pincott Williams and Wilkins, 2000,1028-43.


Thyroid hormone receptors are present in the brain very early in fetal life

Kilby MD, Gittoes N, McCabe C, Verhaeg J, Franklyn JA. Expression of thyroid receptor isoforms in the human fetalnervous system and the effects of intrauterine growth restriction. Clin Endocrinol 2000;53:469-77.


COMMENTARY

The recognition that there are mul-tiple TRs and that their tissue distribu-tion differs has provided impetus to thelong-sought goal of finding thyroid hor-mone analogues with different potencyin different tissues. In older studies, oneanalogue, D-thyroxine (T4), proved to beas active in stimulating cardiac functionas in lowering serum cholesterol concen-trations. Another, triiodothyroaceticacid, did seem to have more potenthepatic and skeletal actions than cardiacactions (1).

GC-1 has yet to be tested inhumans, but it seems to have somedesirable properties. As more is learnedabout the three-dimensional structure of

the receptors, it should be possible todesign other analogues with even moredifferential activity. Furthermore, thereare other ways that analogues with dif-ferent potency in different tissues mightbe devised. For example, analoguesmight be devised such that TR co-activa-tor and co-repressor molecules bind dif-ferently to analogue-TR complexes thanT3-TR complexes. Another way to alterT3 action differentially might be to alterextrathyroidal T3 production by devisingcompounds that selectively inhibit oneof the two deiodinases that catalyze con-version of T4 to T3.


References

1. Sherman SI, Ringel MD, Smith MJ,et al. Augmented hepatic and skeletalthyromimetic effects of tiratricol incomparison with levothyroxine. J ClinEndocrinol Metab 1997;82:2153-8.


A thyroid hormone analogue with differential cardiac and hepatic actions

Trost SU, Swanson E, Gloss B, Wang-Inverson DB, Zhang H, Volodarsky T, Grover GJ, Baxter JD, Chiellini G, ScanlanTS, Dillman WH. The thyroid hormone receptor-(-selective agonist GC-1 differentially affects plasma lipids and cardiacactivity. Endocrinology 2000;141:3057-64.


SUMMARY

Background Thyroid hormones have multiple actions.These actions are mediated by three different thyroid hor-mone nuclear receptors (TR-α1, TR-β1 and TR-β2) that aredistributed differently among different tissues (there is aTR-α2 isoform that does not bind thyroid hormone). Forexample, cardiac tissue contains relatively more TR-αwhereas the liver contains more TR-β. The structure of thetriiodothyronine (T3)-binding region of TR-β1 and -β2 isthe same, but that of TR-α1 is slightly different, making itpossible to design ligands that preferentially activate TR-αor the two isoforms of TR-β.

Methods GC-1 is T3 analogue in which the iodine atomsof the inner ring are replaced by methyl groups and theiodine atom on the outer ring is replaced by an isopropylgroup. The ether link between the rings is replaced by amethylene link and the amino side chain by an oxyaceticacid group. The affinity of GC-1 for the β isoforms of thereceptor is 10 times greater than for the α1 isoform. Thecardiac and hepatic actions of GC-1 were compared withthose of T3 in hypothyroid mice and in normal rats withdiet-induced hypercholesterolemia, and tissue and plasmaconcentrations of T3 and GC-1 were measured in the rats.

Results In hypothyroid mice given T3 or GC-1 for 4 weeks,T3 increased heart rate and cardiac contractility more thandid equimolar amounts of GC-1. It was also more potent

in raising the myocardial content of the mRNAs for myosinheavy chain-α and -β, sarcoplasmic reticulum adenosinetriphosphatase, and HCN2, a cardiac pacemaker channel.In these latter actions, T3 was approximately 9 times morepotent than an equimolar amount of GC-1. In contrast, inthese mice, T3 and GC-1 were equipotent in lowering serumcholesterol concentrations, and GC-1 was more potent inlowering serum triglyceride concentrations.

In hypercholesterolemic rats given T3 or GC-1 for 7 days,the dose of GC-1 needed to lower serum cholesterol con-centrations was approximately 10 times higher than that ofT3 and the dose needed to lower serum TSH concentrationsby 30 percent was approximately 20 times higher. In con-trast, the dose of GC-1 needed to increase the heart rate by15 percent was greater than 120 times higher. As comparedwith T3, the tissue to plasma ratio of GC-1 was slightlylower in the liver and much lower in the heart.

Conclusion The T3 analogue GC-1 lowers serum lipidconcentrations more effectively than it stimulates cardiacfunction, indicating that its ability to activate TR isoformsdiffers from that of T3.

SUMMARY

Background Some patients with autoimmune thyroid dis-ease have serum antibodies that react with the sodium-iodine transporter in thyroid cells, but the frequency ofdetection of the antibodies in patients with Graves' diseaseor Hashimoto's disease is uncertain.

Methods Serum samples from 177 patients with hyperthy-roidism caused by Graves' disease, 72 patients withHashimoto's disease and 165 normal subjects were testedfor anti-sodium/iodine transporter antibodies. The ratio ofwomen to men in the three groups was 4 to 1, 3 to 1 and 1to 4, respectively). The serum samples were analyzed in aradioligand assay using recombinant radiolabeledsodium/iodine transporter as antigen. Because antibodieswere readily detected in a few normal subjects, the test wasconsidered positive if the serum sample bound more radio-labeled antigen than the samples from 95 percent of thenormal subjects (threshold value defined as 20 arbitraryunits). Serum antithyroid peroxidase and thyrotropin (TSH)

receptor-binding inhibitory antibodies were also measured.

Results Nineteen of the 177 patients with Graves' disease(11 percent), 15 of the 72 patients with Hashimoto's disease(21 percent) and 8 of the 165 normal subjects (5 percent)had a positive test for sodium/iodine transporter antibod-ies. The proportion of patients with positive testsdecreased substantially if the threshold value for a positivetest was raised. Antithyroid peroxidase antibodies weredetected in 100 percent of the patients with Hashimoto'sdisease, and 90 percent of the patients with Graves' hyper-thyroidism had TSH receptor-binding inhibitory antibodies.

Conclusion Most patients with Graves' hyperthyroidismor Hashimoto's thyroiditis do not have high serum concen-trations of anti-sodium/iodine transporter antibodies.

COMMENTARY

Since identification of thesodium/iodine transporter, severalgroups of investigators have looked forantibodies to the transporter or peptidecomponents of it in the serum ofpatients with autoimmune thyroid dis-ease using immunoassays and also bioas-says in which the serum is tested for itsability to inhibit iodide transport in cellstransfected with the transporter gene(1-4). Given that many patients withautoimmune thyroid disease have serumantibodies to multiple components ofthyroid tissue, it is not surprising thatthey have antibodies to the transporteror peptide components of it. The fre-quency of positive tests varied consider-ably in different studies, perhaps notsurprisingly because a different assay wasused in each.

More important is whether sodi-um/iodine transporter antibodies haveany biological importance. Antibodiesthat inhibit iodide transport might con-tribute to thyroid hypofunction in

patients with Hashimoto's thyroiditis.There is reason, however, to be skepticalthat any transporter antibodies inhibitiodide transport, because in a recentstudy serum samples that inhibitediodide transport lost activity when dia-lyzed and retained it after theimmunoglobulins were extracted (4).These results do not deny the existenceof transporter antibodies as detected byimmunoassay, just their biologicalactivity.


References

1. Endo T, Kogai T, Nakazato M, et al.Autoantibody against Na+/I- symporterin the sera of patients with autoimmunethyroid disease. Biochem Biophys ResComm 1996;224:92-5.

2. Morris JC, Bergert ER, Bryant WP.Binding of immunoglobulin G frompatients with autoimmune thyroid dis-ease to rat sodium-iodide symporter

peptides: evidence for the iodide trans-porter as an autoantigen. Thyroid1997;7:527-34.

3. Allan RA, Findlay C, Metcalfe RA, etal. The modulation of the human sodi-um iodide symporter activity by Graves'disease sera. J Clin Endocrinol Metab1992;83:1217-21.

4. Chin HS, Chin DKH, MorgenthalerNG, et al. Rarity of anti-Na+/I- (NIS)antibody with iodide uptake inhibitingactivity in autoimmune thyroid disease(AITD). J Clin Endocrinol Metab2000;85:3937-40.

AUTOIMMUNE THYROID DISEASE

Low frequency of anti-sodium-iodide transporter antibodies in autoimmunethyroid disease

Seissler J, Wagner S, Schott M, Lettman M, Feldkamp J, Scherbaum WA, Morgenthaler NG. Low frequency of autoanti-bodies to the human Na+/I- symporter in patients with autoimmune thyroid disease. J Clin Endocrinol Metab2000;85:4630-4.


COMMENTARY

The authors suggest that the sameantibodies caused the different distur-bances in thyroid function in the sisterand her brother, but that seems unlikely.The TSH receptor antibodies that haveTSH agonist actions and cause hyperthy-roidism and those that have TSH antago-nist actions and cause hypothyroidism areprobably structurally similar, but notidentical.

The donor had two episodes ofGraves' hyperthyroidism, between whichshe had an episode of subclinicalautoimmune hypothyroidism. Therefore,she had T and B lymphocytes sensitizedto epitopes of the TSH receptor relatedto both receptor activation and blockade,

but predominantly activation. What thesister gave to her brother, however, wereT lymphocytes primarily from clonescapable of inducing the production ofTSH receptor-blocking or cytotoxic anti-bodies or T lymphocytes capable ofdirect cytotoxic actions. When theseclones expanded, he developed hypothy-roidism. The fact that he had subclinicalrather than overt hyperthyroidism 2months after cessation of T4 therapysuggests he did not have severe thyroidinjury, and that his hypothyroidism wascaused more by TSH receptor-blockingantibodies than thyroid cytotoxicity.

This was a rather unique series ofevents in this man and his sister, butGraves' hyperthyroidism has been report-ed in both donors and recipients of bone

marrow transplants. They include adonor-recipient (sister-brother) pair inwhich the donor and recipient first devel-oped hyperthyroidism 96 months and112 months, respectively, after transplan-tation (1).


References

1. Holland FJ, McConnon JK, Volpe R,Saunders EF. Concordant Graves' dis-ease after bone marrow transplantation:implications for pathogenesis. J ClinEndocrinol Metab 1991;72:837-40.

THYROID AUTOIMMUNE DISEASE

Hypothyroidism can be transferred by bone marrow transplantation

Marazuela M, Steegman JL. Transfer of autoimmune hypothyroidism following bone marrow transplantation from adonor with Graves' disease. Bone Marrow Transplantation 2000;26:1217-20.


SUMMARY

Background Both Graves' disease and chronic autoim-mune thyroiditis may be caused by autoantibodies that bindto the thyrotropin (TSH) receptor, and both disorders mayoccur sequentially in the same patient. Here, a man whoreceived a bone marrow transplant from his sister, who hadGraves' hyperthyroidism and was in remission, developedautoimmune hypothyroidism 10 months after transplanta-tion. After donating the marrow the sister developed sub-clinical hypothyroidism and then hyperthyroidism.

Case Reports

The Recipient. A 23-year-old man with acute lymphoblas-tic leukemia in relapse was treated with cyclophosphamideand 12 Gy total body radiation and then received a bonemarrow transplant (3.8 x 108 mononuclear cells/kg bodyweight) from his ABO- and HLA-identical sister. Hisserum TSH and thyroxine (T4) concentrations were normaland antithyroid microsomal antibodies were not detected inserum. Despite prophylaxis, he had an episode of graft-ver-sus-host disease and then a cytomegalovirus infection, bothof which responded to appropriate therapy. At 5 monthsafter transplantation, his serum TSH concentration wasnormal, and no antithyroid or other autoantibodies weredetected. At 8 months, immunologic reconstitution wasnormal; serum antithyroid microsomal antibodies weredetected in a titer of 1:400. At 10 months, he complainedof weakness and somnolence. His thyroid gland was notenlarged. His serum TSH concentration was 60 mU/L, hisserum free T4 concentration was low-normal (0.9 ng/dL[11.6 pmol/L]), and tests for antithyroid peroxidase,antithyroglobulin and TSH receptor-binding inhibitory anti-

bodies were positive. There was complete chimerism ofperipheral blood mononuclear and polymorphonuclearcells. He was treated with T4. At 16 months, T4 therapy wasstopped; 2 months later he had subclinical hypothyroidism(serum TSH concentration 9.6 mU/L, serum free T4 con-centration 1.0 ng/dL [12.9 pmol/L]), and anther test forTSH receptor-binding inhibitory antibodies was positive.T4 therapy was resumed. At 24 months he was asympto-matic, with no evidence of leukemia.

The Donor. The donor was the recipient's 19-year-old sis-ter. She had developed Graves' hyperthyroidism, with pos-itive tests for antithyroid microsomal and TSH receptor-binding inhibitory antibodies, two years before donating thebone marrow. She was treated with an antithyroid drug for18 months. At the time of marrow donation her serumTSH concentration was 3.5 mU/L and her antithyroidmicrosomal antibody titer was 1:6400. At 12 months aftermarrow donation her serum TSH concentration was nor-mal, but at 15 months she had subclinical hypothyroidism(serum TSH 11.7 mU/L, serum free T4 0.8 ng/dL [10.3pmol/L]), a very high titer of serum antithyroid microsomalantibodies (1:409,600), and a positive test for TSH receptor-binding inhibitory antibodies. At 18 months her serumTSH was 4.7 mU/L. At 21 months she was clinically andbiochemically hyperthyroid (serum TSH 0.01 mU/L, serumfree T4 3.5 ng/dL [35.1 pmol/L]).

Conclusion The occurrence of hypothyroidism caused bychronic autoimmune thyroiditis in a recipient of marrowfrom a sibling with Graves' disease suggests that TSHreceptor antibodies can act as an agonist in one person (themarrow donor) and an antagonist in another (the marrowrecipient).

SUMMARY

Background Glucocorticoid excess ameliorates thyroidautoimmune disease. Conversely, in occasional patientsrestoration of normal adrenal function has evoked thyroidautoimmune disease. There has not, however, been a sur-vey of the effect of treatment of patients with glucocorti-coid excess on thyroid autoimmunity.

Methods Serum free thyroxine (T4), free triiodothyronine(T3), thyrotropin (TSH) and antithyroglobulin and antithy-roid peroxidase antibodies were measured and thyroid ultra-sonography was done in 20 patients (16 women, 4 men, agerange 18 to 38 years) with Cushing's disease and 40 normalsubjects (32 women, 8 men, age range 18 to 38 years). Thepatients were studied before treatment and after they hadnormal adrenal function for 6 months. All the patients hada corticotroph adenoma of the pituitary and underwenttranssphenoidal adenomectomy; 6 were not cured, ofwhom 5 were successfully treated by radiation and 1 by totaladrenalectomy.

Results The mean serum free T4 and free T3 concentra-tions were similar in the patients both before and after treat-ment and in the normal subjects. Their serum TSH andthyroid antibody concentrations are shown in the table.

Before treatment, 12 patients (60 percent) had low serumTSH concentrations. After treatment 7 (35 percent) hadhigh concentrations. Ultrasonography revealed a nodulargoiter in 6 of the normal subjects (15 percent) and 9 of thepatients (45 percent) before treatment.

After treatment, these 9 patients still had a nodular goiter,and 6 of them and 1 other patient also had a diffuse hypoe-choic pattern suggestive of thyroiditis. These 7 patientswith thyroiditis were the ones with high serum TSH con-centrations. They also had the highest serum concentra-tions of both antibodies both before and after treatment,and all had biopsies that revealed chronic lymphocytic thy-roiditis.

Conclusion Restoration of normal adrenal function inpatients with Cushing's disease is associated with markedincreases in serum antithyroid antibody concentrations anda high incidence of hypothyroidism.

COMMENTARY

Glucocorticoids have both directand indirect effects on pituitary-thyroidfunction. The direct effects includeinhibition of thyrotropin-releasing hor-mone and TSH secretion, particularlynocturnal TSH secretion, inhibition ofthe production of thyroid hormonetransport proteins, and inhibition ofexrathyroidal conversion of T4 to T3.Thus, many patients with Cushing's syn-drome have not only slightly low serumTSH concentrations, but also low-nor-mal serum free T4 and free T3 concen-trations (1).

The indirect effects include inhibi-tion of thyroid autoimmune disease, as

amply documented in this study inwhich many patients had rises in serumantithyroid antibody concentrations and7 patients developed subclinical or overthypothyroidism after successful treat-ment of Cushing's disease. This is anunusually high frequency, and suggeststhat glucocorticoid excess somehowevokes but at the same time inhibits thy-roid autoimmunity, which then becomesclinically evident after the excess is treat-ed. This pattern is in some ways similarto that of postpartum thyroiditis.


References

1. Benker G, Raida M, Olbricht T, et al.TSH secretion in Cushing's syndrome:relation to glucocorticoid excess, dia-betes, goitre, and the ‘sick' euthyroidsyndrome. Clin Endocrinol 1990;33:777-86.

THYROID AUTOIMMUNITY

Successful treatment of Cushing's disease evokes thyroid autoimmunity

Colao A, Pivonello R, Faggiano A, Filippella M, Ferone D, Di Somma C, Cerbone G, Marzullo P, Fenzi G, Lombardi G.Increased prevalence of thyroid autoimmunity in patients successfully treated for Cushing's disease. Clin Endocrinol2000;53:13-9.


Table. Mean (±SD) Serum TSH and Thyroid Antibody Concentrations beforeand after Treatment in Patients with Cushing's Disease and Normal Subjects.

Serum TSH Serum Antithyroglobulin Serum Antithyroid(mU/L) Antibodies (U/mL) Peroxidase

Antibodies (U/mL)

Patients

Pretreatment 0.4±0.05 66±16 82±19Post-treatment 4.7±1.1 387±100 421±131

Normal subjects 2.3±0.4 82±15 99±12

Thyroid Review Articles

Glinoer D, Delange F. The potential repercussions of maternal, fetal, and neonatal hypothyroxine-mia on the progeny. Thyroid 2000;10:871-87.

Klein I, Ojamaa K. Thyroid hormone and the cardiovascular system. N Engl J Med 2001;344:501-9.

Morreale de Escobar G, Obregon MJ, Escobar del Rey F. Is neuropsychological development relatedto maternal hypothyroidism or to maternal hypothyroxinemia? J Clin Endocrinol Metab 2000;85:3975-87.

Palazzo FF, Hammond LJ, Goode AW, Mirakian R. Death of the auto-immune thyrocyte: is itpushed or does it jump? Thyroid 2000;10:561-72.

Documents

Clinical Thyroidology March 2001 Volume 13 Issue 1thyroid.org/wp-content/uploads/publications/clinthy/clinthy_v131.pdf ·