The immediate side effects of ¹³¹I therapy are typically minimal. As noted above, transient exacerbation of thyrotoxicosis can occur, and apparent thyroid storm has been induced within a day (or days) after 131-I therapy.  A few patients develop mild pain and tenderness over the thyroid and, rarely, dysphagia. Some patients develop temporary thinning of the hair, but this condition occurs two to three months after therapy rather than at two to three weeks, as occurs after ordinary radiation epilation. Hair loss also occurs after surgical therapy, so that it is a metabolic rather than a radiation effect. If the loss of hair is due to the change in metabolic status, it generally recovers in a few weeks or months.   However hair thinning, patchy alopecia, and total alopecia, are all associated with Graves' Disease, probably as other auto-immune processes. In this situation the prognosis for recovery is less certain, and occasionally some other therapy for the hair loss (such as steroids) is indicated. Permanent hypoparathyroidism has been reported very rarely as a complication of RAI therapy for heart disease and thyrotoxicosis.^74-76. Patients treated for hyperthyroidism with  131-I received approximately 39 microGy/MBq administered (about 0.144rad/mCi) of combined beta and gamma radiation to the testes. This is reported to cause no significant changes in FSH, but testosterone declines transiently for several  months, and there is no variation in sperm motility or % abnormal forms (76.1).
Long term studies of patients after RAI treatment by Franklyn et al (76a) show a slight increase in mortality which appears to be related to cardiovascular disease, possibly related to periods of hypothyroidism.

Fig.11-4. Transient increase in TBII followed by a decrease in patients treated with 131I. Note that in this study a lower "TSI Index" means more TBII are present.

Worsening of ophthalmopathy---In contrast to the experience with antithyroid drugs or surgery, antithyroid antibodies including TSH-RSAb levels increase after RAI.^77,78 (Fig. 11-4, above). Coincident with this condition, exophthalmos may be worsened.⁷⁹ (Fig. 11-5, below). Although we believe that this change is an immunologic reaction to discharged thyroid antigens, this is conjecture, and the relationship of radiation therapy to exacerbation of exophthalmos remains uncertain.⁷⁹ Recent data indicates that there is a significant correlation.^80,80.1 Nevertheless, we consider "bad eyes" to be a relative contraindication to RAI. Pretreatment with antithyroid drugs has been used empirically in an attempt to prevent this complication. Its benefit, if any, may be related to an immunosuppressive effect of PTU, described below. Treatment with methimazole before and for three months after I¹³¹ therapy has been shown to help prevent the treatment-induced rise in TSH-R antibodies which is otherwise seen.⁸¹

Administration of prednisone with ¹³¹I helps prevent exacerbation of exophthalmos, and  this approach is now the standard approach in patients who have significant exophthalmos at the time of treatment^82,82.1. (Fig. 11-9, below) The recommended dose is 30 mg/day for one month, tapering then over 2-3 months. Of course prednisone or other measures can be instituted at the time of any worsening of ophthalmopathy. In this instance doses of 30-60 mg/day are employed, and usually are required over several months ( See Chapter 12). IV lucocorticoids are probably more effective than oral steroids in prophylaxis or treatment, and carry some risk of serious liver damage if high doses are used. Vannucchi et al found that GO activation was observed in 7.2% of 83 Graves’ patients treated with 131-I who were thought not “at risk” and treated without steroid prophylaxis, and 33.3% of 30 patients considered at risk and treated with steroid prophylaxis, for an overall prevalence of 14.6%. GO activation occurred in 47.6% of patients treated with OGC but in none of the nine patients treated with IVGC. They note that GO may occur after RAI in approximately 15% of patients even in the absence of signs of active GO and that in this small series, IV- prophlaxis was more effective. (Vannucchi G, Campi I, Covelli D, Dazzi D, Currò N, Simonetta S, Ratiglia R, Beck-Peccoz P, Salvi MGraves' orbitopathy activation after radioactive iodine therapy with and without steroid prophylaxis. J Clin Endocrinol Metab. 2009 Sep;94(9):3381-6)

Thyroidectomy, with total removal of the gland, may be considered for patients with significant eye disease. Operative removal of the thyroid is followed by gradual diminution is TSH-R antibodies.(82.2 ), and as shown by Tallstedt is associated with a lower incidence of worsening eye problems than is initial RAI treatment.  While treatment with prednisone helps prevent eye problems, it does not appear to reduce the effectiveness of RAI in controlling the hyperthyroidism(82.3).

Figure 11-5. More patients experienced worsening of exophthalmos following 131-I therapy than after surgery or ATD treatment.

Fig. 11-9.  Redrawn from Bartelena et al, New England J. of Medicine, 338:73-78,1998. Patients with Graves' Disease were followed on methimazole, or given 131-I, or 131-I with prednisone 0.4-0.5mg/kg starting 2 or 3 days after 131-I treartment, and continuing for one month, after which the dose was tapered. Patients are grouped according to those who worsened, were unchanged, or improved during each treatment during 12 months.

Failure of ¹³¹I to cure thyrotoxicosis in 2 or 3 treatments occurs occasionally, and rarely 4 or 5 therapies are given. The reason for this failure is usually not clear. The radiation effect may occur slowly. A large store of hormone in a large gland may be one cause of a slow response. Occasional glands having an extremely rapid turnover of ¹³¹I require such high doses of the isotope that surgery is preferable to continued ¹³¹I therapy and its attendant whole body radiation. If a patient fails to respond to one or two doses of ¹³¹I, it is important to consider that rapid turnover may reduce the effective dose. Turnover can easily be estimated by measuring RAIU at 4, 12, 24, and 48 hours, or longer. The usual combined physical and biological half-time of ¹³¹I retention is about 6 days. This may be reduced to 1 or 2 days in some cases, especially in patients who have had prior ¹³¹I therapy or subtotal thyroidectomy. If this rapid release of ¹³¹I is found, and ¹³¹I therapy is desired, the total dose given must be increased to compensate for rapid release. A rough guide to this increment is as follows:

Most successfully treated glands return to a normal or cosmetically satisfactory size. Some large glands remain large, and in that sense may constitute a treatment failure. In such a situation secondary thyroidectomy could be done, but it is rarely required in practice.

Patients who have been treated with RAI should continue under the care of a physician who is interested in their thyroid problem for the remainder of their lives. The first follow-up visit should be made six to eight weeks after therapy. By this time, it will often be found that the patient has already experienced considerable improvement and has begun to gain weight. The frequency of subsequent visits will depend on the progress of the patient. Symptoms of hypothyroidism, if they develop, are usually not encountered until after two to four months, but one of the unfortunate facts of RAI therapy is that hypothyroidism may occur almost any time after the initial response.

Hazards of ¹³¹I Treatment

In the early days of RAI treatment for Graves' disease, only patients over 45 years of age were selected for treatment because of the fear of ill effects of radiation. This age limit was gradually lowered, and some clinics, after experience extending over nearly 40 years, have now abandoned most age limitation. The major fear has been concern for induction of neoplasia, as well as the possibility that ¹³¹I might induce undesirable mutations in the germ cells that would appear in later generations.

Table 11-8. Gonadal Radiation Dose (in Rads) From Diagnostic Procedures and ¹³¹I Therapy

Source: Adapted from Robertson and Gorman⁹⁵

Radiation is known to induce tumor formation in many kinds of tissues and to potentiate the carcinogenic properties of many chemical substances. Radiation therapy to the thymus or nasopharyngeal structures plays an etiologic role in thyroid carcinoma both in children and in adults.^{83-85 131}I radiation to the animal thyroid can produce tumors, especially if followed by PTU therapy.⁸⁶ Cancer of the thyroid has appeared more frequently in survivors of the atomic explosions at Hiroshima and Nagasaki than in control populations.⁸⁷ Thyroid nodules, some malignant, have appeared in the natives of Rongelap Island as the result of fallout after a nuclear test explosion in which the radiation cloud unexpectedly passed over the island.⁸⁸

The experience at 26 medical centers with thyroid carcinoma after ¹³¹I therapy was collected in a comprehensive study of the problem. A total of 34,684 patients treated in various ways were included. Beginning more than one year after ¹³¹I therapy, 19 malignant neoplasms were found; this result did not differ significantly from the frequency after subtotal thyroidectomy. Thyroid adenomas occurred with increased frequency in the ¹³¹I-treated group, and the frequency was greatest when the patients were treated in the first two decades of life.³⁹ Holm et al.⁴¹ have thoroughly examined the history of a large cohort of ¹³¹I-treated patients in Sweden and similarly found no evidence for an increased incidence of thyroid carcinoma or other tumors. For reasons that are not clear, the injury caused by ¹³¹I therapy for Graves' disease seems to induce malignant changes infrequently.. This result may occur because the treatment has largely been given to adults with  glands less sensitive to radiation, because damage from ¹³¹I therapy is so severe that the irradiated cells are unable to undergo malignant transformations or all cells are destroyed, or possibly because of the slow rate at which the dose is delivered ⁸⁹   In up to one-half of patients followed for 5-10 years, there may be no viable thyroid cells remaining. We note that two  studies reported above extend through an average follow-up period of 15 years. As described above ^44.1, a recent report by Franklyn and coworkers indicated that there is an increased (3.25 fold) risk of mortality from cancer of the thyroid (and also bowel) after RAI, detected in along term follow up of a very large patient cohort.   However it remains uncertain that this is related to hyperthyroidism per se, or radioiodine therapy.

While these data are reassuring in regard to 131-I use in adults, Chernoby made it clear that its use in children can not be considered safe. Children in the area surrounding Chernobyl have developed a hugely increased incidence of thyroid carcinoma predominately  due to ingestion of  iodine-131. ^89.1. The latency has been about 5 years, and younger children are most affected. Risk of carcinogenesis decreases with increasing age at exposure, and is much less common after age 12. However some data indicates that an increased incidence of thyroid carcinoma is seen even among adults exposed at Chernobyl.

The incidence of leukemia among patients treated with RAI for Graves' disease has not exceeded that calculated from a control group.⁹⁰ This problem was also studied by the consortium of 26 hospitals.⁹¹ The incidence of leukemia in this group was slightly lower than in a control group treated surgically, but slightly higher in the latter group than in the general population.

In the group of RAI-treated patients, there has been no evidence of genetic damage, although, as will shortly be seen, this problem cannot be disregarded. In the United States, about 100 x 10⁶ children will be born to the present population of over 200 x 10⁶ persons. Approximately 4% of these children will have some recognizable defect at birth. Of these, about one-half will be genetically determined or ultimately mutational, and represent the the effects of the baseline mutation rate in the human species. These mutations are attributed in part to naturally occurring radiation.

All penetrating radiation, from whatever sources, produces mutations. The effects may vary with rate of application, age of the subject, and no doubt many other factors, and are partially cumulative. Nearly all of these mutations behave as recessive genetic factors; perhaps 1% are dominant. Almost all are minor changes, and those produced by experimental radiation are the same as those produced by natural radiation.

Whether or not mutations are bad is in essence a philosophic question. Most of us would agree that the cumulative effect of mutations over past eras brought the human race to its present stage of development. However, most mutations, at least those that are observable, are detrimental to individual human adaptation to the present environment. In terms of the human population as a whole, detrimental mutant genes must be eliminated by the death of the carrier. We can agree that an increase in mutation rate is not desirable. It is hardly worth considering the pros and cons of the already considerable spontaneous mutation rate.

In mice, the occurrence of visible genetic mutations in any population group is probably doubled by acute exposure of each member of the group over many generations to about 30 - 40 rads, and by chronic exposure to 100 - 200 rads.⁹² This radiation dosage is referred to as the doubling dose. Ten percent of this increase might be expressed in the first-generation offspring of radiated parents, the remainder gradually appearing over succeeding generations. The change in mutation rate in Drosophila is in proportion to the dosage in the range above 5 rads. Data from studies of mice indicate that at low exposures (from 0.8 down to 0.0007 rads/min), the dose causing a doubling in the spontaneous rate of identifiable mutations is 110 rads.⁹² Linearity, although surmised, has not been demonstrated at lower doses.

At present, residents of the United States receive about 300 mrad/year, or 9 rad before age 30, the median parental age. Roughly half of this dose is from natural sources and half from medical and, to a lesser extent, industrial exposure. The National Research Council has recommended a maximum exposure rate for the general population of less than 10 rad above background before age 30. (The present level may therefore approach this limit.)

The radiation received by the thyroid and gonads during ¹³¹I therapy of thyrotoxicosis can be estimated from the following formula:

Total beta radiation dose = 73.8 x concentration of ¹³¹I in the tissue (µCi/g) x average beta ray energy (0.19 meV) x effective isotope half-life

For illustration, we can assume a gland weight of 50 g, an uptake of 50% at 24 hours, a peak level of circulating protein-bound iodide (PB¹³¹I) of 1% dose/liter, an administered dose of 5 mCi, a thyroidal iodide biologic half-life of 6 days, and a gamma dose of about 10% of that from beta rays. On this basis, the thyroid receives almost 4,100 rads, or roughly 1,600 rads/mCi retained. The gonadal dose, being about one-half the body dose, would approximate 2 rads, or roughly 0.4 rads/mCi administered.

If the radiation data derived from Drosophila and lower vertebrates are applied to human radiation exposure (a tenuous but not illogical assumption), the increased risk of visible mutational defects in the progeny can be calculated. On the basis of administration to the entire population of sufficient ¹³¹I to deliver to the gonads 2 rads or 2% of the doubling dose (assumed to be the same as in the mouse), the increase in the rate of mutational defects would ultimately be about 0.04%, although only one-tenth would be seen in the first generation. Obviously only a minute fraction of the population will ever receive therapeutic ¹³¹I. The incidence of thyrotoxicosis is perhaps 0.03% per year, or 1.4% for the normal life span. At least one-half of these persons will have their disease after the childbearing age has passed. Although most of them will be women, this fact does not affect the calculations after a lapse of a few generations. Assuming that the entire exposed population receives ¹³¹I therapy in an average amount of 5 mCi, the increase in congenital genetic damage would be on the order of 0.02 (present congenital defect rate) x 0.04 (¹³¹I radiation to the gonads as a fraction of the doubling dose) x 0.014 (the fraction of the population ever at risk) x 0.5 (the fraction of patients of childbearing age) = 0.0000056.

This crude estimate, developed from several sources, also implies that, if all patients with thyrotoxicosis were treated with ¹³¹I, the number of birth defects might ultimately increase from 4 to 4.0006%. This increase may seem startlingly small or large, depending on one's point of view, but it is a change that would be essentially impossible to confirm from clinical experience.

Unfortunately, it is more difficult to provide a reliable estimate of the increased risk of genetic damage in the offspring of any given treated patient. Calculations such as the above simply state the problem for the whole population. Since most of the mutations are recessive, they appear in the children only when paired with another recessive gene derived from the normal complement carried by all persons. Assuming that only one parent received radiation from ¹³¹I therapy amounting to 2% of the doubling dose, the risk of apparent birth defects in the patient's children might increase from the present 4.0% to 4.008%.

Similar estimates can be derived by considering the number of visible mutations derived from experimental radiation in lower species.^92,93

On this basis, the increase in the birth defect rate would be from 4.0% to 4.012%. One important observation stemming from these calculations is that large numbers of children born to irradiated parents must be surveyed if evidence of genetic damage is ever to be found. Reports of "no problems" among 30 to 100 such children are essentially irrelevant when one is seeking an increase in the defect rate of about 4.0% to about 4.008%.

These statistics are presented in an attempt to give some quantitation to the genetic risk involved in ¹³¹I therapy, and should not be interpreted as in any sense exact or final. The point we wish to stress is that radiation delivered to future parents probably will result in an increased incidence of genetic damage, but an increase so slight that it is difficult to measure. Nonetheless, the use of ¹³¹I for large numbers of women who subsequently become pregnant will inevitably introduce change in the gene pool.

In considering the significance of these risks, one must remember that the radiation exposure to the gonads from the usual therapeutic dose of ¹³¹I may be only one or two times that produced during a procedure such as a barium enema ^94,95 and similar to the 10 rads received from a CAT scan. These examinations are ordered by most physicians without fear of radiation effect (Table 11-8).

When assessing the risks of ¹³¹I therapy, one must, of course, consider the risks of any alternative choice of procedure. Surgery carries a small but finite mortality, as well as a risk of permanent hypoparathyroidism, hypothyroidism, and vocal cord paralysis. Some of these risks are especially high in children, the group in which radiation damage is most feared. Some physicians have held that ¹³¹I therapy should not be given to patients who intend subsequently to have children. In fact, there is at present little if any evidence to support this contention, as discussed above. Chapman⁴⁴ studied 110 women treated with ¹³¹I who subsequently became pregnant and were delivered of 150 children. There was no evidence of any increase in congenital defects or of accidents of pregnancy. Sarkar et al.⁹⁶ also found no evidence of excess abnormalities among children who received ¹³¹I therapy for cancer. Other studies have confirmed the apparent lack of risk.^42,43 It should be noted that no increase in congenital abnormalities has been detected among the offspring of persons who received much larger radiation doses during atomic bomb explosions.⁹⁷

Often the patient wishes to know about the possibility of carcinogenesis or genetic damage. These questions must be fully but delicately handled. It is not logical to treat a patient of childbearing age with ¹³¹I and have the patient subsequently live in great fear of bearing children. These problems and considerations must be faced each time a patient is considered for RAI therapy.

Pregnancy is an absolute contraindication to ¹³¹I therapy. The fetus is exposed to considerable radiation from transplacental migration of ¹³¹I, as well as from the isotope in the maternal circulatory and excretory systems. In addition, the fetal thyroid collects ¹³¹I after the 12th week of gestation and may be destroyed. The increased sensitivity of fetal structures to radiation damage has already been described.

Physicians treating women of childbearing age with ¹³¹I should be certain that the patients are not pregnant when given the isotope. Therapy during or immediately after a normal menstrual period or performance of a pregnancy test are appropriate precautions if pregnancy is possible. Women should be advised to avoid pregnancy for at least six months, since it usually takes this long to be certain that retreatment will not be needed.