Most, if not all, thyroid adenomas are monoclonal, as, presumably, are most carcinomas (101). Colloid nodules may be either mono- or poly-clonal. Thus tumors represent the persistent growth of the progeny of one cell which has somehow escaped the mechanisms which maintain normal cell division at about once each 8.5 years (102).

The process of oncogenesis is conceived to be a series of events induced by genetic and environmental factors which alter growth control. At the phenomenologic level these factors may be considered as "initiators" and "promoters". Initiators include such agents as chemicals and irradiation which induce tumors, and promoters are agents such as phenobarbital, which in rats augments TSH secretion and radically increases tumor development. In man x-ray treatment is the sole known initiator, and other than elevated TSH, no promoters are known. Compounds such as phenobarbital, dilantin and PCBs, which are known thyroid tumor promoters in animals through liver microsomal hormone degrading enzyme induction leading to increased thyroid hormone metabolism, do not appear to have a detectable adverse effect in man in doses usually employed (103).

We now begin to understand oncogenesis at another level. More than 30 "oncogenes" have been recognized in the human genome. These genes, normally silent, can become activated by chromosomal translocations, deletions, or mutations, and then can "transform" normal cells into a condition of uncontrolled growth. Oncogenes appear to be closely related to normal growth factors, genes that control cell division, or to hormone receptors. In general, these genes, when turned on, promote cell growth and cell division and depress differentiation. Typically activation of one such gene may not be enough to produce malignancy, but if accompanied by expression of another oncogene, or if gene mutation or reduplication occurs, the cell may progress toward a malignant potential. Or if one growth suppresing gene is damaged, and the second copy is somehow lost, malignancy may ensue. Information on expression of oncogenes in human thyroid tissue is rapidly accumulating. Expression of c-myc is stimulated in normal thyroid cells by TSH, and the proto-oncogene is expressed in adenomas and carcinomas. Activating mutations of h-ras at codons 12, 13, and 61, and over expression of h-ras, are found in adenomas and carcinomas, but h-ras mutations are also found in nodular goiter tissue (104), suggesting that h-ras mutations could be an early event in oncogenesis (104, 105) Other studies, it should be noted, find ras mutations uncommon (106).
BRAF mutations appear to be very common in papillary thyroid cancer. For example, BRAF mutations were found in 219 of 500 cases (43.8%),primarily  the common BRAF V600E mutation . By multivariate analysis, the absence of tumor capsule was the only parameter associated (P = 0.0005) with BRAF V600E mutation. The data suggest that BRAF V600E mutation is associated with high-risk PTC and in particular in follicular variant with invasive tumor growth(106a). BRAF(V600E)'alone' does not represent a marker for poor outcome, however, when associated with alterations in other genes, such as Ras, it identifies a subset of PTCs with increased risk of recurrence and decreased survival(106b).

Vecchio and co-workers (107) cloned an oncogene which is frequently and specifically expressed in papillary thyroid cancers. This oncogene is found on chromosome 10, and involves an intrachromosomal rearrangement of the tyrosine kinase domain of the ret oncogene so that it is attached to one of three different promoters, producing retPTC-1, retPTC-2, and retPTC-3. One of these translocation products is found in 20-70% of Papillary cancers.. This rearrangement leads to constitutive expression of the oncogene. It has been shown that intra-thyroidal expression of the ret/PTC1 oncogene can induce thyroid cancer (108). BRAF mutations are also frequent in papillary carcinoma, and undifferentiated cancers that have arisen from papillary tumors(108a). Interestingly, discordant patterns of BRAF mutation were found in about 40% of the multifocal PTCs. In node metastases, BRAF mutations were, in most but not all the cases, concordant with the dominant tumor.: The heterogeneous distribution of. BRAF mutation suggests that discrete tumor foci in multifocal PTC may occur as independent tumors(108b).

A mutational change has now been associated with follicular cancers. In 5 of 8 follicular cancers, Kroll et al (109) found translocation of the DNA binding domain of PAX8 to domains A-F of the peroxisome proliferator-activater receptor (PPAR) gamma1 gene. The fusion oncogene is able to transform thyrocytes, so appears to be able to produce malignancies(109a). Although initially thought to be exclusively present in follicular cancers, it is now know to be present in follicular adenomas as well (110). Mutation or deletion of the p53 tumor suppressor gene is found in some differentiated thyroid cancers, and many undifferentiated cancers, suggesting that this genetic deletion may be one of the final steps leading to anaplastic cancer growth. (Fig.18-11)
    A proliferation of studies in this field has provided many clues to thyroid tumorigenesis. Simian virus 40-like sequences are found in many thyroid cancers, as well as other cancers, and the Tag gene sequence found is known to be oncogenic in animal models (111). Mutated and non-functional thyroid hormone receptors are recognized in up to 90% of PTC by one author, suggesting a role in oncogenesis, but other workers find these mutations to be rare. (112,112a). The tumor suppressor gene TSG101 is over-expressed in most PTCs (113). Overexpression of many other genes -galectin-3,Thymosin beta-10,hTERT, CD97, CD26, VEGF- has been detected, but of course a question always is whether these changes represent the cause or the result of oncogenesis.

Roentgen irradiation of the thyroid and administration of 131I have both induced carcinomas in the experimental animal (136). A combination of 131I injury to the thyroid cell and prolonged administration of a goitrogen is especially likely to produce carcinomas, as shown by Doniach (137). Cell metabolism is altered by 131I, even when small amounts are administered. In the rat, 5 µCi prevents subsequent response to a goitrogenic drug (140). With larger doses the colloid is sparse, the follicles are variable in size, and large eosinophilic acinar cells appear. Very large doses of 131I (producing several thousand rads) to the rat thyroid radically alter cell metabolism, liberate TG within 1 or 2 weeks, and subsequently reduce the efficiency of hormone synthesis.131Iodine irradiation in rats in doses so low as not to alter hormone biosynthesis immediately inhibits DNA synthesis and cell replication, as shown by a failure to respond to subsequent goitrogenic challenge. The cells also have a shortened life span. Similar inhibition of hyperplasia follows x-irradiation to the thyroid.

Therapeutic doses of 131I to patients also induce atypical nuclei, which may remain for many years (141). The doses of RAI needed to produce neoplastic change in the thyroid glands of animals closely parallel those given in the treatment of thyrotoxicosis in humans. The morphologic changes are intensified by a goitrogenic stimulus and reduced by thyroid hormone treatment.

The effects of radiation may be twofold. The nuclear morphologic changes presumably derive from mutations or deletions producing an abnormality in cell division or growth, which may predispose to carcinomatous change. Also, the damaged cell produces less thyroid hormone, and thereby ultimately comes under intense TSH stimulation, as in experiments with goitrogens. Thus, it seems certain that chronic TSH stimulation in animals is associated with the evolution of a neoplasm, especially if it is combined with radiation damage to the cell nuclei.

Experimental thyroid tumors induced by 131I are initially TSH dependent. At first, they can be transplanted successfully only into thyroidectomized animals that are producing much TSH. After serial passages through several generations, the tumors may become autonomous and will then grow in a normal host. Partial or complete dependence on TSH is also observed in some human papillary and follicular tumors.

Duffy and Fitzgerald (142) first made the important observation that a high proportion of children with thyroid carcinoma had received therapeutic x-irradiation to the upper mediastinum or neck during childhood for control of benign lesions such as enlarged thymus, tonsils, or adenoids. Their finding has been amply confirmed (143-150). Winship and Rosvoll (151) studied 562 children with thyroid carcinoma from all parts of the world. Among those for whom adequate historical data were available, 80% had a history of prior x-ray treatment. This relationship is not so obvious for carcinomas developing after age 35. Significant x-irradiation to the head, neck, and chest in childhood increases the frequency of thyroid cancer by 100-fold (152), and the incidence is proportional to the dose, reaching at least 1.7% at 500 rads, or 5.5 cases per million exposed persons per rad each year (Fig. 18-6). Long term evaluation of individuals who received scalp X-ray for epilation, and incidental small doses such as 7 rads to the thyroid, have confirmed a high risk (ERR/Gy of 20.2 and absolute excess risk of 9.9% per gray per 10-4 person years), with risk continuing through 40 years after exposure(152a). Our own data disclose a 7% incidence by 30 years after irradiation (153). The latent period averages 10 - 20 years, but tumors occur even after 20 - 40 years (Fig. 18-7, below). There appears to be no true threshold, since even doses as low as 9 rads increase the incidence of cancer (154). It is in fact probable that “natural” background radiation may produce many of the spontaneous tumors(154a). There is a direct dose-response relationship through 1,000 rads (155). Higher doses of irradiation also induce tumors, and the true dose-response curve in the range 1,000 - 5,000 rads in humans is not known. Benign nodules occur with nearly 10 times the frequency of cancers. Interestingly, the type of tumor induced is not different from those occurring spontaneously, and there is no relation between dose and latent period. For some reason, women are more prone to develop radiation-induced tumors than men, and both ethnic and familial factors may influence tumor development (156).

Probably any x-ray exposure of the thyroid has some carcinogenic potential, although the risk may decrease with age. Adults were extensively treated by x-irradiation for Graves' disease from 1930 to 1950. There is reported to be an increased incidence of carcinoma in these patients (157). A significant incidence of thyroid neoplasia was observed in patients who received x-ray therapy for cervical tuberculous adenitis (158). These patients were treated at ages up to 34, received 500-1,500 rads, and developed tumors 10-27 years after treatment. In a study of survivors of the atomic blasts at Nagasaki and Hiroshima, an increased incidence of thyroid cancer was found among persons who had received large amounts of radiation (159). Thus, the thyroid of the adult is sensitive to the carcinogenic action of x-rays, although not so sensitive as that of the child. Radiation-associated tumors of the thyroid continue to occur, although x-ray treatment of thymic enlargement and tonsillar or adenoid hypertrophy has been discontinued since 1959. A recent analysis of 1787 patients treated with X-ray for Hodgkin's disease found 1.7% to have thyroid cancer (160).
    The most dramatic and terrifying data are now emerging from the area around Chernobyl, where thousands of people of all ages received large doses of radiation from external fallout and ingested isotopes, especially  isotopes of iodine. In this epidemic the risk of thyroid cancer is highest among children who were under 9 years and especially under 5 years old at the time of the Chernobyl explosion, and presumably ingested iodide via milk from cows grazing on contaminated forage. The latent period in these children is amazingly short-6 to 7 years, the tumors tend to be relatively aggressive, and are frequently associated with thyroid autoimmunity (161)A strong dose-response relationship was observed between radiation dose to the thyroid received in childhood and thyroid cancer risk (P<.001). For a dose of 1 Gy, the estimated odds ratio of thyroid cancer varied from 5.5 (95% confidence interval [CI] = 3.1 to 9.5) to 8.4 (95% CI = 4.1 to 17.3). A linear dose-response relationship was observed up to 1.5-2 Gy(161a). So far there has been no evidence of increased tumor prevalence in countries in Europe that received low doses of fallout from Chernobyl.  However it has been estimated that this fall out will in fact ultimately produce up to 16000 additional cases of thyroid cancer by 2065, but that these will constitute such a small portion of the total cases that the increase will not be distinct(161.1).

 Radiation-associated tumors are generally found among younger patients. They are rarely undifferentiated, but some have been fatal. In a review of x-ray associated thyroid tumors at the University of Chicago Thyroid Clinic (162), the latent period among children treated predominantly in adolescence for tonsillar enlargement or acne averaged 20 years. It appears that the peak incidence of lesions is at 10 - 25 years after exposure (Fig. 18-7, above), and it is possible that the occurrence of new cancers decreases over time. A recent study of thyroid neoplasms (not specifically cancers) and thyroidits, showed that excess risk in exposed individuals continues through at least 30 years and actually increased in the period from 10-30 years after radiation (162a) Among 100 consecutive patients seen in 1973 and 1974, only because they knew of prior radiation exposure, 15% had lesions suggestive of tumor and 7% had cancer proven at operation (163). Favus et al (164) found a similar incidence of cancer (60/1056) in irradiated patients called back for evaluation. Although one case-controlled study suggests a lack of effect of radiation,¹⁰⁴ the evidence, reviewed by Maxon et al (165), clearly confirms the importance of this problem.

Based on these facts, it has been accepted by most physicians in the field that patients with a history of thyroid irradiation (over 20 rads, and certainly 50 rads) should be located and advised to have an assessment. This evaluation should consist at least of a physical examination and, if any thyroid abnormality is found, an ultrasound. If one or more clear-cut nodules is found, or if one or more clear-cut inactive areas is found by scan, then surgical intervention may be indicated. As noted below, needle aspiration cytology can alter this approach in some cases. Benign adenomas are also found in these glands, with an incidence much higher than that of cancers. Our own studies show that almost all clinically detectable cancers are found by careful physical examination, and that scans are of supplementary value and rarely turn up a cancer missed on palpation. Serum TG levels tend to be elevated in irradiated patients, and antithyroid antibodies are more commonly present, but these tests are not of diagnostic value. When excised these glands often show multiple benign as well as malignant nodules as well as areas of fibrosis and hyperplasia (166).

M.P., 52-Year-Old-Woman: Thyroid Radiation and Multiple Gland Abnormalities

This patient was first seen with a history of irradiation for acne during her teens. She subsequently developed telangiectasia of the skin of her face. The month before the examination, she had observed a lump on the right side of the neck. Examination disclosed a 1-cm nodule in the right lobe of the thyroid and some irregularity of the left lobe. Thyroid scintiscan showed a cold nodule of the lower pole of the right lobe. Ultrasound examination of the right lobe identified a partially cystic nodule and a small cystic structure of the left. The FTI was slightly elevated at 10.9. RAIU was above normal. Thyroid antibodies were not present. Thyroid needle aspiration showed cells indicative of malignancy.

Routine blood tests disclosed alkaline phosphatase of 107 units (normal, 25-100 units), calcium 10.9 and 11.5 mg/dl (normal, 8.5-10.2 mg/dl), and phosphate 2.7 ng/dl. Repeated assay of FTI again demonstrated an elevated value of 15 (normal = 6-10.5). The level of parathyroid hormone was 0.65 ng/ml (with a coincident calcium level of 11.5 mg/dl), values indicative of primary hyperparathyroidism.

The patient was treated with potassium iodide for 1 week and admitted for exploratory surgery. A right upper para-thyroid adenoma weighing 908 mg was found. The adenoma showed areas of cystic degeneration and fibrosis. The thyroid gland was multinodular and was suspicious on frozen section for follicular carcinoma. There was extensive fibrosis around and adherent to the thyroid gland. A near-total thyroidectomy was performed. The gland weighed 17 g. Multiple nodules in the gland measured 1-18 mm in diameter. An 18-mm nodule in the right lobe was identified as follicular carcinoma. There were, in addition, multiple follicular adenomas and multiple Hurthle cell tumors, focal hyperplasia, and colloid nodules in the right and left lobe.

Postoperatively the patient received thyroid hormone. When seen 1 month after surgery, her calcium level was 9.4 mg/dl, phosphorus 3.5 mg/dl, and parathyroid hormone 0.28 ng/ml (normal). The FTI was 9.2 while taking 0.15 mg L-T4.

This patient developed a cystic parathyroid adenoma with hyperparathyroidism, multiple adenomas of the thyroid, follicular carcinoma, and multiple functioning adenomas that produced thyrotoxicosis. All of these tumors occurred concurrently in a gland showing changes typical of prior radiation exposure.

D.C., 19-Year-Old-Girl: Development of Thyroid Carcinoma, After X-Ray Therapy, While Receiving Thyroid Hormone

At age 10 the patient had respiratory distress and was found to have a superior-anterior mediastinal mass. There was left cervical lymphadenopathy and bilateral supraclavicular lymphadenopathy. Biopsy revealed Hodgkin's disease of the nodular sclerosing variety. The result of staging laparotomy was negative. She was treated with 4,000 rads to a neck mantle field.

One year later the results of thyroid function tests were normal, but two years after x-ray treatment the FTI was 4.1 and the TSH level was 24 µU/ml. Thyroid hormone replacement therapy was begun, and the patient was carefully monitored over subsequent years with periodic FTI and TSH determinations. Six years after irradiation a 2-cm nodule was noted in the left lobe of the thyroid. This nodule was found to be cold on 123-I scan. Fine needle aspiration revealed cells suggestive of malignancy.

At surgery a papillary adenocarcinoma with capsular and vascular invasion was found, and a near-total thyroidectomy was performed. Postoperatively residual thyroid tissue was ablated by administration of 30 mCi 131I. The skeletal survey findings were negative; chest x-ray films and bone films were normal. She has remained free of evidence of thyroid carcinoma or Hodgkin's disease in the subsequent three years.

This history demonstrates the occurrence of thyroid carcinoma, in a gland heavily irradiated during therapy for Hodgkin's disease, while the patient was taking adequate replacement doses of thyroid hormone. Possibly the period of X-ray induced hypothyroidism played a role in tumor induction. The tumor developed within six years of the radiation therapy. Fortunately, the tumor was not metastatic and has presumably been eradicated by surgery and RAI ablation of residual tissue.

In our series, postradiation carcinomas averaged 1.7 cm in size (Fig. 18-8), and 14% were below 0.5 cm. The size distribution was similar to that of non-X-ray-associated tumors. They were more frequently multicentric than those in nonirradiated glands and as aggressive, or more so, in behavior than tumors arising without known irradiation (167). The tumors are mainly papillary or follicular, but an occasional anaplastic cancer is also found. In examining patients, it should be remembered that benign and malignant salivary gland neoplasms, neuromas, parathyroid adenomas (168), laryngeal cancer, skin malignancies, and breast cancer also occur with undue frequency in this group of patients.

Thyroid nodularity and cancer also occured as a sequela of nuclear fallout in the accident at Rongelap in the Marshall Islands (169),where individuals received 200-1,400 rads. The incidence of nodularity is 40%, and nearly 6% proved to have cancer. Children in Utah exposed to small amounts of fallout from atomic bomb testing  were reported to not  develop nodules or carcinomas  ¹⁰¹ . Recent re-evaluation of the population finds that there is a definite but small increase in both thyroiditis and thyroid neoplasms continuing for up to 30 years after exposure 169.1)

.

Iodine-131 treatment induces abnormalities in the thyroid gland that persist for many years (170). Giant nuclei, increased mitotic activity, hyperchromatic nuclei, and other abnormalities appear. It seems reasonable that these nuclear changes could lead to carcinomatous degeneration. Chromosomal damage in circulating lymphocytes has also been reported after 131I administration (171). Patients have developed thyroid nodules or tumors after 131I therapy for Graves' disease; it has been suggested, but not proved, that the highest incidence has been among those treated during childhood. Some of the lesions found in 131I-treated children may actually have been carcinomas, but there has been debate (172) among the various pathologists who examined the specimens. Several reports of isolated instances of cancer after 131I treatment of adults for Graves' disease have appeared, but the large United States Public Health Service cooperative study failed to show an increased risk in this group (173-175). Studies by Holm et al (175) also failed to show an increase in cancer incidence among persons given 131-I either for diagnosis or for therapy for thyrotoxicosis. These patients were adults, and usually in the 40-60-year age group. Also, very large radiation doses may be less carcinogenic than small ones, and in half or more of these patients, the thyroid has been totally destroyed. Lastly, the follow-up time in most reports averages 8-13 years, which may be too soon to see radiation-induced neoplasia. Thus the evidence is reassuring but the question cannot be considered closed.

Chronic intense stimulation of the thyroid with TSH probably can lead to carcinogenesis in humans, as it can in animals. There are several reports of intensely hyperplastic congenital goiters, untreated for long periods, in which carcinomas have finally developed (176-180). Fortunately, most patients with congenital goiter are recognized and treated with replacement thyroid hormone at some time during early childhood, so that chronic TSH stimulation does not occur. Interestingly, activating mutations of the TSH-R, which are metabolically like chronic TSH stimulation, lead to benign and not malignant change, as described above.