Clinical Endocrinology (1987), 27, 715-720

STRUMA OVARII: A CASE REPORT

J. H . LAZARUS, A. R. RICHARDS, M. J. MACPHERSON, J. S . DINNEN, E. D. WILLIAMS, G . M . OWEN AND J. S . H . WADE

Departments of Medicine, Medical Physics, Medical Biochemistry, Pathology and Surgery, University of Wales College of Medicine, University Hospital of Wales, Heath

Park, Cardif, UK

(Received I April 1987; returned for revision 8 June 1987;Jinally revised 6 July: accepted 16 July 1987)

SUMMARY

A patient with struma ovarii and hyperthyroidism is described. She was treated for Graves’ hyperthyroidism at age 22 and received thyroxine for post-operative hypothyroidism. Twenty years later she became thyrotoxic and was treated with antithyroid drugs and radioiodine. Diagnosis of struma ovarii was made by radioiodine profile scanning and an ovarian tumour was removed. This had the pathological features of struma ovarii and autoradiographic evidence of pre-operatively administered lZ5I was seen in the lesion. The patient had positive results for long acting thyroid stimulator (LATS) and LATS-protector (LATS-P) pre- and post-operatively. Bioassays for thyroid stimulators were positive post-operatively but radioreceptor assays for TsAb were consistently negative. It is suggested that profile scanning is an appropriate investigation for diagnosis. It is not clear whether the lesion was autonomous or being stimulated by circulating thyroid stimulators.

Graves’ disease is the commonest cause of hyperthyroidism presenting to a thyroid clinic; most of the remaining hyperthyroid patients have either toxic multinodular goitre or toxic adenoma (Williams et al., 1983). A small minority of patients present with unusual types of hyperthyroidism (Hamilton & Maloof, 1973). There have been many reviews of struma ovarii (Marcus & Marcus, 1961; Nieminen et al., 1963; Bortolozzi, 1967; Fox & Langley, 1976) but the accuracy of the reported 5-15”/0 incidence of hyperthyroidism occurring in this condition is difficult to evaluate. The reason for this is that the diagnosis of hyperthyroidism has been made pre-operatively in only a few cases; also, early reported cases had insufficient or dubious thyroid function studies to either demonstrate hyperthyroidism or to confirm that the ovarian tumour was the source of the excess thyroid hormone.

The aim of this report is to document a patient with Graves’ disease who presented with hyperthyroidism due to a hyperfunctioning struma ovarii. Detailed biochemical and immunological information is available before and after treatment.

Correspondence: Dr J. H. Lazarus, Senior Lecturer in Medicine, University of Wales College of Medicine, University Hospital of Wales, Heath Park, Cardiff, UK.

715

716 J. H. Lazarus et al.

CASE REPORT

A white female aged 48 years was referred to this hospital in November 1977 for management of her hyperthyroidism.

In 1954, aged 22, she presented with right exophthalmos, tachycardia and diffuse thyroid enlargement. There was no family history of thyroid or other autoimmune disease. Four hour thyroidal I3lI uptake was 73% and partial thyroidectomy was performed.

She remained euthyroid and by 1956 the exophthalmos had regressed. In 1957, although symptomless, she was found to have hypoparathyroidism (serum calcium 7.5 mg/dl, 1.87 mmol/l) and was treated with oral calcium phosphate. Gynaecological examination because of lower abdominal pain in 1963 was normal. During normal pregnancies in 1961 and 1964 no pelvic abnormality was noted.

In September 1972 she developed mild puffiness of the eyelids and conjunctival hyperaemia. Free thyroxine index (FTI) was 40 (normal range 55--145). Treatment with L-

thyroxine 0.1 mg/day was started and reduced to 0.05 nig/day after 1 month because of tremor. Thyroxine was stopped 18 months later and she continued to have a normal FTI until she become clinically thyrotoxic in 1977, serum total T3 6.3 nmol/l (normal range 1.2-3.5 nmol/l, FTI 202. Menstruation had ceased and she had recurrent right exophthalmos, and vitiligo, with no thyroid enlargement. Thyroidal neck uptake of I3lI

was less than 5% of the administered dose at 4 h and 24 h. In December 1977 a dose of I3lI

(8 mCi) [ N 300MBql was given followed by carbimazole and propranolol for 3 months. By July 1978 she was again clinically and biochemically thyrotoxic (T3 6.2 nmol/l, FTl

17 l), TSH was 3.2 mU/l (normal range < 5 ) with no increase produced 20 min after 200 pg intravenous TRH. Haemoglobin was 11.1 g/100 ml and ESR (Westergren) 11 mm/h. Serum calcium and phosphate, albumin, bilirubin, alkaline phosphatase and aspartate transarninase were normal. Whole body profile scanning ('"I) showed the radioactivity concentrated in the pelvis with a negligible amount in the neck. Protein bound 1311

(PB 13'1) at 72 h as 0.46% dose/litre of plasma (markedly elevated). On pelvic examination a rounded mass was felt which was thought to be an ovarian

tumour prolapsed into the true pelvis. Ultrasound examination showed a large lobulated solid mass in the pelvis mainly on the left side.

At operation (November 1978) a large lobulated tumour of the left ovary was found. There was some free fluid in the pelvis but no evidence of metastasis. The right ovary and the uterus were removed in addition to the left ovarian mass. Surgical recovery was uneventful. Profile scanning 3 days post-operatively confirmed that all the radioactivity had been removed from the pelvis.

FTI fell from 290 to 192, 100 and 51, at 18 h, 72 h and 150 h, respectively, after tumour removal. Twelve weeks post-operatively the patient was hypothyroid, (TSH 25 mU/l) and required L-thyroxine. Following cessation of T4. TRH (200 pg i.v.) was given simultaneously with oral TRH 40 mg and 13'1 tracer (250 pCi) and TRH 40 mg 8 hourly continued for 3 days. TSH levels were > 40 mU/I at 20 min and 60 min after initial TRH administration and 32 mU/1 at 4 h. I3lI uptake in the neck was 1 % 4 h and 72 h. At the latter time < 1 % of administered radioactivity was in the pelvis and no concentrations of iodine were found elsewhere in the body. PB l 3 I I was subnormal at 72 h (0.02% dose/l). Currently she is well and euthyroid while taking 0.15 mg L-thyroxine daily.

Strurna ovarii: a case report

Table 1 . Antibody studies in struma ovarii

Antibody Oct 1978 Jan 1979 Oct 1979

717

Thyroglo bulin Thyroid Microsomes Gastric Parietal Cells Other Organ Specific ABs LATS*

TsAbS TsAW

LATS-Pt

TsAby

- ve - ve + ve - ve

45 5.8

- - 6.5 30

- ve

- -

10 38 - ve

6.8 1 .O pmol cAMP per cell

monolayer + ve

-Not assayed. * Normal value < 4. t Normal value i 10.

9 Cyclic AMP assay: In two separate radioreceptor assays.

Control IgG pool ( 1 mg/ml) 3.4 0.05 pmol cAMP per cell monolayer. Bovine TSH ( 1 mU/ml)-greater than 100 pmol/cAMP per cell monolayer.

7 Positive cytochemical bioassay at IgG dilutions of ljl00 and l / lOOO.

Antibody studies are shown in Table 1 . Antibodies to thyroglobulin were measured using the Burroughs Wellcome kit and to thyroid microsomes by the Fujizoki kit. Antibodies to gastric parietal cells and other organ specific antibodies were measured by immunofluorescence. Long acting thyroid stimulator (LATS) and long acting thyroid stimulator protector (LATS-P) were measured by Professor D.S. Munro (Dirmikis & Munro, 1975). Thyroid stimulating antibodies (TsAb) were also measured by radiorecep- tor assay, (Smith & Hall, 1974) the cyclic AMP response elicited from addition of IgG to a thyroid cell monolayer (Bidey & Marshall, 1983) and by cytochemical bioassay. Before the operation the concentration of the long acting thyroid stimulator (LATS) was raised but the LATS protector (LATS-P) level (Dirmikis & Munro, 1975) was markedly elevated. Approximately 1 year after operation both LATS and LATS-P were still elevated; this was associated with a positive result for thyroid stimulating immunoglobu- lins as measured by bioassay, both by a cyclic AMP response in cultured monolayers of thyroid cells (Bidey et al., 1983) and by a cytochemical bioassay. During this time the radioreceptor assay (Smith & Hall, 1974) for thyroid stimulating antibodies was consistently negative.

PATHOLOGY

Review of the thyroid histology obtained in 1954 confirmed that the thyroid showed focal areas of hyperplasia with small follicles lined by tall epithelium; some follicles showed infolding of the epithelium. There was also a slight focal lymphocytic infiltrate. The left

718 J. H. Lazarus et al.

Fig. 1. Section of ovary showing hyperplastic thyroid tissue H & E, x 200.

ovary measured 9 cm across, weighed 243 g and had a smooth nodular surface. The cut surface had a sponge-like appearance with areas resembling thyroid tissue and several large cysts. Autoradiographic examination of the operative specimen which was performed following the pre-operative administration of 1251 showed significant iodide concentration in the ‘thyroid-like’ areas. Histological examination showed that most of the ovary had been replaced by thyroid tissue (Fig. 1). The thyroid follicles varied greatly in size with some large follicles lined by flattened epithelium and some groups of small follicles lined by tall epithelium suggesting areas of hyperactivity. In several areas the epithelium was vacuolated and showed nuclear pleomorphism. This appearance is consistent with irradiation damage. Between the follicles there was oedematous fibrous tissue with several foci of bone. In the rim of ovarian tissue there were groups of eosinophilic cells with the features of luteinized stromal cells. The endometrium was in the proliferative phase with a benign polyp and showed no evidence of hyperplasia. The opposite ovary and tubes showed no significant features and there was a leiomyoma in the uterus.

DISCUSSION

Struma ovarii, defined as a teratoma of the ovary in which thyroid tissue forms the bulk of the tumour, is rare. About 400 cases have been described in the literature; some of these may not be genuine as the criteria for the definition have not been met. The incidence of hyperthyroidism due to functioning thyroid tissue within the tumour is 5 to 15% of the

Strurna ovarii: a case report 719

total (Marcus & Marcus, 1961; Bortolozzi, 1967). However, many of the cases of hyperthyroidism described in the older literature are difficult to evaluate because of the lack of adequate thyroid function tests. It is also probable that some genuine cases have not been reported.

There is good clinical and histological evidence to suggest that our patient had Graves’ hyperthyroidism when she underwent subtotal thyroidectomy in 1954. Goitre appears to occur more commonly in patients with struma ovarii than one would expect by chance (Fox, 1976) although there does not seem to be a reason for this association. The diagnosis of a hyperfunctioning struma ovarii was made in our case because of the demonstration of low radioiodine uptake in the neck associated with high I3’I uptake in the pelvis while she was clinically and biochemically hyperthyroid and having excluded other causes (Himsworth, 1985). Confirmation was obtained autoradiographically by noting a high concentration of ‘251, administered prior to operation, to be located in the strumal tissue. In addition, resolution of the hyperthyroidism occurred following resection of the tumour which itself corresponded to thyroid tissue histologically.

Scanning techniques have been used prior to surgery in one previously described case of hyperfunctioning struma ovarii (Brown et al., 1973; Yeh et al., 1973) but profile scanning has not been employed. We suggest that a suitable whole body scanning technique is a valuable aid in the diagnosis of this condition. Where studied, the iodinated components of a struma ovarii in a euthyroid patient consisted of thyroglobulin and showed a relative increase in particulate iodine, particularly in the microsomes with nearly all of the iodinated amino acid present as monoiodotyrosine (Ramagopal & Stanbury, 1963). It is probable that the iodinated compounds in our patient’s toxic struma would not be very different from those found in a toxic cervical goitre.

The pathogenesis of the hyperfunctioning state in struma ovarii is not always apparent. In most cases the pathology of the tumour is that of well differentiated thyroid tissue with cellular features suggesting activity due to adenomatous hyperplasia (Woodruff & Rauh, 1966; Kempers et al., 1970). There are no reports of struma thyroid tissue with pathological features consistent with Graves’ disease. However, in a recent case of struma ovarii occurring in a patient with Graves’ disease (Lefort et al., 198 1) it was suggested that the ‘ovarian toxic nodule’ could have been stimulated by Graves’ immunoglobulins. If the struma arose as an embryological defect in our patient it is interesting that no abnormality was found on pelvic examination in 1963. Presumably a small volume of thyroid tissue was present and this subsequently grew, perhaps in response to growth stimulating antibodies (Dexhage et al., 1980). The case described by Lefort et al. (1981) also had a normal gynaecological examination some 10 years prior to the onset of symptoms from the struma ovarii. The findings in the present case are difficult to interpret because some radioiodine therapy was given before surgery. The calculated radiation dose to the ovary was 13.3 Gy. However, allowing for radiation changes there does not seem to be any evidence of lymphoid infiltration consistent with Graves’ disease. The present case illustrates that elevated levels of TsAb characteristic of Graves’ disease may persist or regress many years after the initial therapy. It is tempting to suggest that these stimulating immunoglobulins were implicated in the hyperactivity of the struma. However, it was not possible to verify this by examining the effect of the patient’s serum on the cyclic AMP response in the excised tissue in vitro. The fact that some tests for thyroid stimulating immunoglobulins were negative suggests that TSI are a heterogeneous group of Igs and that more than one type of test may be required for their assay in a clinical situation (Hall

720 J. H. Lazarus et al.

et al., 1982). Recent refinement of some of these assays has improved the sensitivity. The recently reported case of Lefort et al. (1981) already referred to also was shown to have raised levels of TsAb as tested by the increase in CAMP production in thyroid slices.

The striking similarity between our case and that of Lefort et al. (1981) perhaps lends some support to the hypothesis that the stimulation of the struma ovarii may be due to thyroid stimulating immunoglobulins in patients with known Graves’ disease.

ACKNOWLEDGEMENTS

We thank Professor R. Ekins, Professor R. Hall and Professor D.S. Munro for assays of thyroid stimulating immunoglobulins and Dr H. J. Lloyd for referring the patient.

REFERENCES

BIDEY, S.P., MARSHALL, N.J., EKINS, R.P. (1983) Bioassay of thyroid-stimulating immunoglobulins using human thyroid cell cultures: optimization and clinical assessment. Clinical Endocrinology, 19, 193-206.

BORTOLOZZI, G. (1967) Lo struma ovarico (presentazione di 5 casi e rassegna della letteratura). Annali di Ostetricia e Ginecologia, 89, 3 1 ( r 3 I .

BROWN, W.W., SHETTY, K.R. & ROSENFELD, P.S. (1973) Hyperthyroidism due to struma ovarii: demonstration by radioiodine scan. Acta Endocrinologica, 73, 26&272.

DIRMIKIS, S.M. & MUNRO, D.S. (1975) Placental transmission of thyroid-stimulating immunoglobulins. British Medical Journal, ii, 665-666.

DREXHACE, H.A., BOTTAZZO, G.F., DONIACH, D., BITENSKY, L. & CHAYEN, J. (1980) Evidence for thyroid- growth-stimulating immunoglobulins in some goitrous thyroid disease. Lancet, ii, 287-291.

Fox, H. & LANGLEY, F.A. (1976) Tumours of the Ovary. Heinneman Medical Books, London, p.240. HALL, R., MCLACHLAN , S., MCGREGOR, A.M., WEETMAN, A.P. & REES-SMITH, B. (1982) Thyrotropin receptor

antibodies: clinico-pathological correlations. In Receptors, Anfibodies and Disease (eds D. Evered & J. Whelan), Pitman, London, (Ciba Foundation Symposium 90), pp. 133-1 52.

HAMILTON, C.R. & MALOOF, F. (1973) Unusual types of hyperthyroidism. Medicine (Balfimore), 52,195-215. HIMSWORTH, R.L. (1985) Hyperthyroidism with a low iodine uptake. In Clinics in Endocrinology and

KEMPERS, R.D., DOCKERTY, M.B., HOFFMAN, D.L. & BARTHOLOMEW, L.G. (1970) Struma ovarii-ascitic,

LEFORT, G., COMMENCES-DUCOS, M., DENECHAUD, M., RIVEL, J. & LATAPIE, J. (1981) Goitre ovarien au cows

MARCUS, C.C. & MARCUS, S.L. (1961) Struma ovarii. A report of seven cases and a review of the subject.

NIEMINEN, U., VON NUMERS, C. & WIDHOLM, 0. (1983) Struma ovarii. Acta Obstetrica et Gynecologica

RAMAGOPAL, E. & STANBURY, J.B. (1963) Studies of the distribution of iodine and protein in a struma ovarii.

SMITH, B.R. & HALL, R. (1974) Thyroid stimulating immunoglobulins in Graves’ disease. Lancet, ii, 427431. WILLIAMS, I., ANKRETT, V., LAZARUS, J.H. & VOLPE, R. (1983) Aetiology of hyperthyroidism in Canada and

WOODRUFF, J.D., RAUH, J.T. & MARKLEY, R.L. (1966) Ovarian Struma. Obstetrics and Gynaecology, 27,

YEH, E.-L., MEADE, R.C. & RUETZ, P.P. (1973) Radionuclide study of struma ovarii. Journal of Nuclear

Metabolism (ed. A.D. Toft), 14(2), 397415.

hyperthyroid, and asymptomatic syndromes. Annals of Internal Medicine, 72, 883-893.

d’une maladie de Basedow. La Nouvelle Presse Medicale, 10, 2209-221 0.

American Journal of Obstetrics and Gynecology, 81,752-761.

Scandinavia, 42, 399424.

Journal of Clinical Endocrinology, 25,526-533.

Wales. Journal of Epidemiology and Community Health, 37, 245-248.

194-201.

Medicine, 14, 118-121.