Abstract
As a Visiting Endocrinology Fellow at the Massachusetts General Hospital (MGH) during the 1963–1964 academic year, I had the pleasure of interacting with Dr. Stanbury and many other distinguished physician–scientists at the Thyroid Clinic and Laboratory who had also greatly contributed to the understanding of thyroid diseases. In particular, I had the opportunity of meeting the founder of the Thyroid Clinic at the MGH, namely James Howard Means, who was still attending many of the weekly conferences held at the unit. These conferences were also attended by such distinguished physician–scientists including Jacob Lerman, Farahe Maloof, Earle Chapman, and Leslie DeGroot. Their contributions, and those of many others who were part of the history of the MGH, have been summarized in Dr. Stanbury's publication, “A Constant Ferment” (1). Meeting such outstanding physician–scientists was a wonderful and exciting experience for me and familiarized me with the role of the physicians as scientists in applying innovative technology to improving patient care. Among their many achievements was John Stanbury's contributions to the understanding of the physiology of iodine metabolism and its application to the pathophysiology of endemic goiter (2). Throughout his career in the study of thyroid diseases, he followed the concept that “insight into the nature of thyroid disease, as regards both cause and pathologic physiology, has developed through an understanding of the normal function of the thyroid gland” (2). His pioneering studies, conducted in the tradition of James Howard Means, led to his clinical research studies in the province of Mendoza, Argentina, to understand the defects in iodine metabolism in a region of endemic goiter. He focused on the importance of iodine deficiency in endemic goiter regions and recognized the association of impaired growth and development to include myxedematous, neurologic, or mixed subtypes of cretinism. This knowledge also led him to further investigate and discover the enzymatic defects of thyroid hormonogenesis and possible genetic causes. Subsequently, his recognition of the role of inheritance in metabolic defects led to the seminal publication of the classic textbook which he co-edited within James B. Wyngaarden and Donald S. Fredrickson, “The Metabolic Basis of Inherited Disease” (3). Dr. Stanbury also helped to organize the International Council for the Control of Iodine Deficiency Disorders—an organization that has played an important catalytic function in the control of iodine deficiency in many developing countries where iodine deficiency is a major cause of disability, pregnancy loss, and premature death, as well as a great impediment to social and economic development.
While at the MGH, I had the opportunity to observe the diagnostic techniques that were being used at the Thyroid Clinic to assess thyroid function and assist in clinical care. At that time, the basal metabolic rate test based upon oxygen consumption was still being performed as an index of thyroid function. The circulating total level of thyroid hormone was estimated using crude indirect methods by measuring the protein-bound iodine level, or when iodine contamination was suspected, a butanol extractable iodine. These levels in combination with a triiodothyronine (T3) binding test using red blood cells or resin assisted in estimating “free hormone” levels. A 24-hour thyroid uptake of radioactive 131I was also utilized to measure thyroid gland activity and perform a radioactive scintiscan to assess the function of thyroid gland nodules. Wang and Vickery, at the MGH, were performing a Vim Silverman cutting needle core biopsy on thyroid nodules. However, they as well as Crile and Hawke at the Cleveland Clinic reported concerns about the possible implantation of thyroid cancer in the needle tract. These reports as well as the overall morbidity of such a procedure discouraged the general application of needle biopsy in the assessment of thyroid nodules in North America.
Before arriving at the MGH, I had previously developed an interest in thyroid diseases when I was an Endocrine Research Fellow at the Toronto General Hospital in 1959. At that time I investigated the newly reported
Facilitated by my MGH experiences and supported by an R. Samuel McLaughlin Travelling Fellowship, I returned to Toronto in 1964 as a full-time physician–scientist in the Department of Medicine at the University of Toronto and Mount Sinai Hospital. In 1965, I established a Nuclear Medicine Department at Mount Sinai and began using 131I scintiscanning for the study of thyroid nodules. In collaboration with Dr. Harry Strawbridge in our Department of Pathology, we assessed the feasibility of applying fine-needle aspiration biopsy to evaluate hypofunctioning thyroid nodules (7). Dr. Strawbridge changed the cytologic staining technique from May-Grunwald Giemsa to a modified Papanicolau stain which greatly improved nuclear detail. To our surprise, papillary thyroid cancer could be readily recognized and/or strongly suspected when there was increased cellularity and/or atypical nuclear features without any significant risk of needle tract cancer implantation or morbidity and we recommended routine application of the technique in North America (8). At the same time, the dogma regarding the <2% risk of cancer in thyroid cysts was challenged by our observations that large 3-cm recurrent hemorrhagic cysts had an approximately 30%–40% risk of cancer and 70% risk of an underlying neoplasm (9). Moreover, the diagnostic features of cysts could be further clarified by the discovery of clear colorless fluid that was indicative of a parathyroid cyst (10), whereas purulent fluid could lead to a diagnosis of acute suppurative thyroiditis and be sent for aerobic and anaerobic bacterial cultures (11). When thick gelatinous butterscotch fluid was obtained and the associated characteristic cholesterol crystals identified by polarizing light, a diagnosis of a branchial cleft cyst could be made.
In collaboration with Dr. Miskin in the Department of Radiology at Mount Sinai Hospital, we began to look in 1972 at the application of neck ultrasound in assessing thyroid nodules. Beginning with the very primitive A-Mode method followed by the Gray Scale and subsequent B-Mode, we were able to distinguish between predominantly solid, cystic, or mixed lesions (12). This permitted a classification of hypofunctioning thyroid nodules and a new paradigm of thyroid nodule management (13). The development of high resolution technology further refined assessment for vascularity and calcification as well as atypical appearance compatible with malignancy and widely expanded the utility of ultrasound as a useful modality in the detection of thyroid malignancy before or after thyroidectomy.
After a visit to Toronto by Rosalyn Yalow and Solomon Berson to receive their Gairdner Foundation International Prize in the early 1970s, I began to apply radioimmunoassay (RIA) techniques to establish an Endocrine Hormone Laboratory at Mount Sinai Hospital. Assays for insulin, cortisol, growth hormone, and thyroid hormone levels were then established for clinical application. Studies by Inder Chopra showed that antibodies could be made to thyroid hormones by injecting thyroglobulin in Freund's adjuvant and that alanine naphthalene sulfonate could saturate thyroid hormone binding sites to facilitate the direct measurement of T4 and T3 levels in unextracted serum. These technological modifications encouraged me to develop serum assays for total T4 and T3. Since there was already a commercial Murphy Pattee competitive protein binding assay method for the clinical measurement of serum T4, I was mainly interested in achieving a T3 RIA by developing antibodies to T3. However, the antibodies produced by this technique were not sufficiently specific to achieve a T3 RIA method. By obtaining a T3 conjugate as a gift from Gharib and Mayberry from the Mayo Clinic, a T3 antibody was produced that permitted studies on the pathophysiology of total serum T3 and T4 in health and disease (14).
Since we had already achieved a T4 RIA method, my laboratory began, in 1972, researching the feasibility of measuring T4 using both heel prick dried blood discs and cord blood of newborn infants delivered at our Mount Sinai Hospital Obstetrical Unit. From these preliminary studies on the successful measurement of total T4, we extended its application to all newborn infants delivered in Metropolitan Toronto. From these preliminary findings, we convinced the Ontario Ministry of Health to introduce this newborn screening test for all newborns delivered in the province of Ontario. A prevalence of primary congenital hypothyroidism was documented to occur at 1 per 3,000–4,000 newborn infants (15). In collaboration with the Toronto Hospital for Sick Children, the various subtypes of congenital primary hypothyroidism were defined to include athyreosis, hypoplasia, goitrous, and ectopic (lingual) (16,17). From further advances in the availability of a sensitive antiserum for the measurement of thyrotropin (TSH) from dried blood discs, we were able to establish in 1983 the first primary TSH screening program in North America (18). This replaced a previous test program of an initial T4 with a supplementary TSH on the lowest 10%–12% of T4 values in a run (19,20). These strategies reduced the recall burden to detect congenital primary hypothyroidism to 0.2% of all births, and permitted treatment of affected infants prior to 3–4 weeks of age. This strategy has now been adopted worldwide and has virtually eliminated the major neurologic sequelae of congenital hypothyroidism.
In our early research on congenital hypothyroidism, the mothers of recalled infants suspected to have a low T4 value were also tested for possible concurrent abnormalities in their thyroid function. At the time, we inadvertently discovered that several mothers had elevated thyroid function values that spontaneously resolved to hypothyroidism before recovery to euthyroidism 6–9 months later. Stimulated by an observation of transient postpartum hypothyroidism by Amino and coworkers in Japan, in combination with the growing recognition of a painless thyroiditis syndrome with spontaneously resolved hyperthyroidism, we identified the postpartum thyroiditis syndrome and reported on this observation in 1977 (21). In a subsequent study, the prevalence of this syndrome detected in mothers delivering at our Mount Sinai Hospital Obstetrical Unit was approximately 6%–8% of all deliveries (22). Follow-up studies on mothers with postpartum thyroiditis strongly supported an autoimmune basis through measurements of thyroid anti-microsomal antibodies and HLA typing (23).
In 1988, I took a sabbatical to work in the laboratory of Vincent Giguere at the Hospital for Sick Children, Toronto, to learn more about the recently discovered families of nuclear hormone receptors. After a fortuitous interaction with Bert O'Malley and Tauseef Butt, I began to determine whether thyroid action could be reconstituted in a yeast model system. Subsequently, a collaboration with Dr. Michael Stallcup at the University of Southern California permitted the demonstration that the coexpression of a p160 coactivator such as GRIP1 or SRC1 in the presence of a T3 receptor could dramatically restore a T3 reporter gene response to thyroid hormone (T3 or Triac) in a yeast model system (24). This observation opened the possibility of applying yeast genetics to further elucidate the role of post receptor multimeric transcriptional complexes and chromatin remodeling in facilitating hormone action by thyroid and retinoic acid receptors (25). Additional studies to determine the effects of adenovirus 5-E1A on reporter gene activation in yeast led to an unsuspected hormone independent activation of a T3 reporter gene. Moreover, this constitutive activation could be suppressed by added T3, suggesting a model of negative regulation (26). Further mutagenic analyses established that the N-terminus of adenovirus 5-E1A had essential sequences similar to a CoRNR box motif. Interestingly, we also found that N-CoR and SMRT corepressors had a similar series of CoRNR box motifs in their nuclear receptor interacting domains. Consequently, studies on N-CoR and SMRT corepressors devoid of their N-terminal repressors domains could be shown to paradoxically function in yeast as a thyroid hormone receptor coactivator (27). Since others had previously recognized the existence of natural splice variants of N-CoR and SMRT that were devoid of their N-terminus repressor domains, a putative transcriptional mechanism was discovered in yeast wherein we could postulate that such spliced variants might play a role as hormone independent gene activators and be down-regulated by T3 and represent a model of negative regulation (28).
Based upon the established reports of others that thyroglobulin (Tg) could be measured in the serum and respond to TSH stimulation, I began to routinely measure Tg levels in patients with well-differentiated thyroid cancer (WDTC) several months after a near or total thyroidectomy just prior to the administration of radioactive iodine (RAI) (29). Later, we analyzed the significance of this serum Tg measurement as a predictor of clinical course after a single dose of RAI in both high- and low-risk WDTC patients (30). Further retrospective analyses demonstrated that patient age and tumor size did not reliably predict long-term clinical outcomes as measured by stimulated Tg after one dose of RAI (31). At the same time we noted that many low-risk WDTC patients had a stimulated Tg that was either negative or weakly positive just prior to the first dose of RAI. The latter observation led to a prospective study (32) to apply follow-up serial stimulated Tg and ultrasound testing to select those patients who might avoid routine RAI therapy. This RAI selection strategy offers a personalized approach for each low-risk WDTC patient to possibly avoid RAI treatment when they have already been essentially rendered free of disease by an undetectable stimulated Tg result. Moreover, the side effects and costs associated with routine RAI therapy can be averted and the global shortage of isotopes now restricting the supply of 131I could be ameliorated. This strategy challenges the current guidelines applied for RAI therapy in low-risk WDTC. Advances in early detection by ultrasound and fine-needle aspiration biopsy, in conjunction with improved surgical technology in most academic centers that perform a total thyroidectomy and therapeutic central compartment neck dissection when indicated, has led to approximately to 60%–70% of all detected WDTC being of low-risk, i.e., confined to the thyroid gland and Level VI (central compartment) lymph nodes. Accordingly, a major reduction in RAI remnant ablation will have a considerable impact on the management of low-risk WDTC. Hence a new paradigm for RAI selection in the postsurgical management of low-risk WDTC appears to be justified and a revision of current RAI selection guidelines is urgently indicated.
Stimulated by the exemplary careers of Dr. John Stanbury and several other physician–scientists at the MGH Thyroid Unit, and as outlined in this article, I have been motivated to advance our understanding of the pathophysiology of several thyroid diseases. It has been a special pleasure to see how technological advances have moved from the bench to the bedside and how, as discoveries are made, they challenge existing dogma and replace outdated concepts of patient care. I and many of my colleagues have been fortunate to benefit from the rapid advances in scientific knowledge and technology occurring in the latter part of the 20th century and to work within and in communication with excellent academic centers and scientists interested in pursuing similar goals. Our work would not have been possible without the generous collaborative support of many medical colleagues, scientists, postdoctoral fellows, and technicians, as well as governmental and private funding agencies. A productive career in pathophysiology requires a combination of scientific curiosity unrestricted by preconceived ideas, intuition based upon a familiarity with current methods of clinical care and available technological advances, and finally, well-defined project goals and the perseverance to successfully achieve them.