TO THE EDITOR:
While reading the European Association of Nuclear Medicine (EANM) and Society of Nuclear Medicine (SNM) procedure guidelines for therapy with 131I (1,2), I was concerned to find that both organizations recommend the use of pregnancy testing (type unspecified) a few days before 131I treatment. The implication is that the pregnancy test will preclude pregnancy, which is contraindicated because of the high radiation dose to the embryo or fetus.
Standard pregnancy tests are intended to assess pregnancy status after a missed period rather than to exclude pregnancy at other points in the menstrual cycle. The standard pregnancy tests detect human chorionic gonadotropin (hCG) in the urine with a sensitivity of ∼20–50 mIU/mL. Using such tests, urinary hCG is detectable only from ∼3 d after the fertilized ovum is implanted in the uterine wall. Typically, implantation occurs ∼21 d into the cycle, so that it is ∼24 d before measurable urinary hCG is produced. However, even a urinary hCG assay 100 times more sensitive than the standard assays failed to detect 10% of pregnancies on the first day after a missed period (i.e., ∼29 d after the last menstrual period [LMP]) and 3% of pregnancies 7 d after a missed period, presumably because of variability in times of fertilization and implantation (3).
Therefore, if we look at the expected performance of a standard pregnancy kit for urinary hCG throughout the cycle, from the early phase of the cycle until day 10 after LMP, it is considered reasonable to assume that the woman has not ovulated and is not pregnant and that, therefore, a standard pregnancy test would yield a true-negative result. If we assume that the ovulation and fertilization occur ∼11–14 d after LMP, then from that time until ∼day 24, a standard pregnancy test is likely to yield a negative result although the woman is pregnant, that is, a false-negative result. Thereafter, the findings would become increasingly true positive, although as indicated above, false negatives can occur even up to 7 d after a missed period. Therefore, if our aim is to exclude pregnancy before therapy, a standard pregnancy test is likely to mislead us for approximately half the menstrual cycle. With a more sensitive serum assay for hCG, the period of uncertainty might be reduced but not eliminated.
The other way in which we try to exclude the possibility of pregnancy is to ask the patient if she can categorically state that she is not pregnant. There is at least one reference in the literature where, despite a pretherapy negative pregnancy test and a written signed declaration stating that she was not pregnant, the woman in question was subsequently found to be pregnant after radioiodine therapy (4). The typical patient’s assessment of her own pregnancy status may be inaccurate, perhaps because she overestimates the efficiency of the method of contraception used. Because many methods of “reversible” contraception, such as contraceptive pills or condoms, are not 100% efficient, it is unlikely that a woman who has had intercourse “mid cycle” and is using “reversible” contraception can categorically state that she is not pregnant. This is supported by an interesting statement in a recent paper on contraceptive failure to the effect that “half of all pregnancies in the United States are unintended. Of these, half occur to women who were practicing contraception in the month they conceived” (5).
Our practice to date has been to ask the woman to sign a declaration that she is definitely not pregnant. If she has done this, we have proceeded with the therapy; however, if she is in any doubt, we have applied the “10-d rule”; that is, performed the therapy during the 10 d after the onset of a period (6). Maybe we should apply the 10-d rule across the board for all sexually active women of childbearing age undergoing radioiodine therapy, irrespective of whether they indicate they are pregnant. Even this protocol will not be foolproof, given that thyrotoxicosis can be associated with menstrual cycle irregularities and that cycle variability generally affects the earlier follicular phase of the cycle, or the phase associated with our “10 d,” rather than the later luteal phase (7). Thus, the 10-d rule might be too long for someone with a short cycle.
The considerations above indicate the difficulties in being proscriptive when administering radioiodine to female patients of childbearing age. Therefore, perhaps it would be best to administer therapies on the basis on the 10-d rule but modified as required after the prescribing physician asks detailed questions about the patient’s menstrual history. This protocol would require modifications to both EANM and SNM guidelines.
REPLY:
Dr. Watson correctly makes 3 points regarding the inherent difficulties of absolutely excluding pregnancy before the administration of radioactive iodine in all women of childbearing age.
First, pregnancy testing alone will not exclude all early pregnancies. However, if pregnancy testing were not performed, it is reasonable to expect that more pregnant hyperthyroid patients would receive 131I therapy than is the case when pregnancy testing is done. The Society of Nuclear Medicine (SNM) guidelines are deliberately not too specific regarding method and timing of the pregnancy test. Of interest, the American College of Radiology (ACR) Practice Guideline for the Performance of Therapy with Unsealed Radiopharmaceutical Sources (effective Jan.1, 2001) also states that “pregnancy may be ruled out by a negative hCG [human chorionic gonadotropin] test obtained within 48 hrs prior to administration of the radiopharmaceutical, or by documented history of hysterectomy or tubal ligation, by a postmenopausal state with absence of menstrual bleeding for 2 years, or by premenarche in a child” (1). Institutions should have their own policies for excluding pregnancy and can make them more stringent than what is found in the guidelines. However, the SNM guidelines reflect the consensus opinion of numerous practitioners from various regions of the United States who were involved in the writing, review, and approval of this document.
Second, historical data also are not perfect. The SNM guidelines state, “Occasionally, when historical information clearly indicates pregnancy is impossible, a pregnancy test may be omitted at the discretion of the treating physician.” All physicians should know that condoms and birth control medication are fallible. Again, all institutions should have their own specific criteria.
Third, likewise, limiting radioiodine therapy to the first 10 d of the menstrual cycle will also not exclude all pregnancies for the reasons given by Dr. Watson. This and the limitations of even the highly sensitive serum β-hCG testing in the earliest stages of pregnancy (approximately 7 d after conception) underscore the value of the treating physicians’ clinical interview and judgment. Some practitioners ask the patient to abstain from intercourse for a week before therapy and still obtain a pregnancy test just before treatment.
We believe that Dr. Watson’s practice of relying on the patient’s “declaration that she is definitely not pregnant” is insufficient in the current regulatory and medicolegal environment in the United States. Aside from the obvious goal of never administering 131I therapy to a potentially pregnant patient, in the United States, fetal exposure greater than 50 mSv (which is likely to occur with 131I therapy) must be reported to regulatory agencies; thus, it seems reasonable to take every measure to minimize the likelihood of this occurrence.
Strictly adhering to the policies outlined in our first and second points, plus having the patient sign a consent stating she is definitely not pregnant, will make the occurrence of treating a pregnant woman with radioiodine an exceedingly rare event. Obviously, the given criteria must be rigidly followed.