A Systematic Review Examining the Effects of Therapeutic Radioactive Iodine on Ovarian Function and Future Pregnancy in Female Thyroid Cancer Survivors

Anna M. Sawka; Deepak C. Lakra; Jane Lea; Bandar Alshehri; Richard W. Tsang; James D. Brierley; Sharon Straus; Lehana Thabane; Amiram Gafni; Shereen Ezzat; Susan R. George; David P. Goldstein

Disclosures

Clin Endocrinol. 2008;69(03):479-490. 

In This Article

Results

After reviewing 342 abstracts retrieved from an electronic search and seven citations retrieved from a manual search, the full-text review was restricted to 61 potentially relevant full-text papers, and ultimately 16 papers were included in the review[6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21] (Fig. 1). All studies included in the review were from observational studies, with no long-term randomized controlled trial data available. Data were included from a total of 3023 women surviving DTC (including at least 275 women who did not receive RAI treatment). The kappa statistic measuring initial agreement between two independent reviewers was 0·813 (0·721-0·906) at the abstract review stage and 0·514 (0·239-0·790) at the full-text review stage. A final consensus was achieved for the inclusion of papers in the final review. Included studies were from various parts of the world including: North and South America, Europe, Asia, and the Middle East ( Table 1 ). All of the studies were composed of clinical populations, either followed longitudinally or sampled in a cross-section. A description of the study populations and the control groups, and losses to follow-up, is shown in Table 1 . In the majority of studies, data for RAI-treated patients were obtained from a clinical cohort followed longitudinally.[6,7,8,9,10,12,13,14,15,17,18,19,20,21] However, a cross-sectional sample from a clinical population was examined in the studies by Einhorn et al.[11] and Sarkar et al.[16] Sioka et al.[18] compared outcomes between a cross-section of the clinical RAI-treated cohort relative to an external group of contemporaneous controls. Statistical comparisons of outcomes in treatment and control groups were performed in only some of the included studies.[8,9,10,12,15,17,18,19,20] Eight of the studies included comparisons with contemporaneous controls.[8,9,10,14,15,17,18,20] In six studies, the goal of levothyroxine therapy was reported to be suppression of TSH[6,7,8,14,17,20] ( Table 1 ), and of these, TSH levels were reported to be undetectable only in the study by Ceccarelli et al.[8] The age at first RAI treatment varied from 8 to 50 years and the cumulative activities of RAI administered for treatment varied from 30 to 1099 mCi (for studies reporting these data) ( Table 2 ). Data on the use of oral contraceptives at the time of RAI treatment were reported in only two studies.[15,19] Reporting on a history of external beam radiation therapy was also variable ( Table 2 ).

Figure 1.

Process of study selection.

We examined the relationship between RAI treatment for DTC and subsequent gonadal function, as reflected by measurements of serum gonadotrophins or sex steroids, menstrual history, and menopausal age. Gonadal function after RAI treatment was examined in seven studies[8,12,15,18,19,20,21] ( Table 3 ). Serum FSH measurements were found to be significantly elevated in some women within 1 year after RAI treatment, relative to pretreatment values in two[15,19] of three studies[12,15,19] examining this outcome. Furthermore, serum LH measurements were significantly elevated relative to pretreatment in one[15] of two studies in which this comparison was made.[12,15] In a study by Raymond et al.,[15] significant rises in serum FSH and LH were found in women who experienced transient absence of menstrual periods after RAI treatment. Furthermore, Souza Rosario et al.[19] reported that in a study of 50 women with DTC who all had normal FSH measurements prior to RAI, 28% (14) had FSH values that were elevated above 15 IU/l at 6 months (P < 0·00001 compared to pre-RAI) but all normalized at 12 months (NS compared to pre-RAI). Esfahani et al.[12] did not observe significant changes in serum oestrogen or progesterone within 1 year after RAI, measured 9 days after a menstrual period. They also found no significant correlation between serum levels of hormones (FSH, LH, oestrogen, progesterone) and cumulative dose of RAI received (P > 0·15).

Changes in menstrual function after RAI treatment for DTC were examined in four studies.[15,18,19,21] Between 12% and 31% of women experienced new menstrual changes, including transient absence of menses, or changes in menstrual timing or flow at some point after RAI treatment,[15,18,19,21] when compared to menstrual histories in external control groups or women with DTC prior to RAI treatment. The absolute rate of transient absence of menstrual periods varied from 8% to 27% within 1 year of RAI treatment for DTC.[15,19,21] Transient oligomenorrhoea resolved within 1 year after RAI treatment in all affected women.[15,19,21] If menstrual periods were missed after RAI treatment, this occurred between 1 and 6 months or cycles after RAI treatment and lasted between 1 and 10 months or cycles.[19,21] There were three studies in which a relationship between the age at RAI treatment and subsequent transient menstrual changes were explored.[15,18,21] Raymond et al.[15] reported that the average age of the 18 women who had transient oligomenorrhoea after RAI was significantly higher (mean 38·9 years, SD 7·1) compared to the 48 women with DTC who did not (mean 32·2 years, SD 6·1) (P < 0·001). Furthermore, Vini et al.[21] reported that the women who experienced transient absence of menstrual periods after RAI treatment were older (mean age of 36 years) than women who experienced milder menstrual irregularities (mean age 28 years) or no menstrual changes (mean age 31 years). Sioka et al.[18] reported a trend for a positive correlation of menstrual irregularities after RAI treatment with increasing age, although the results did not reach statistical significance. Thus, either a significant association or a trend was reported for the relationship between increasing age at the time of RAI treatment and subsequent transient menstrual changes in all three studies examining this relationship.[15,18,21] Conflicting data were found in two studies examining the relationship between the activity of RAI administered and the incidence of transient menstrual changes.[15,21] Vini et al.[21] concluded that patients experiencing transient menstrual disturbances after RAI had received a greater cumulative radiation dose compared to those with normal menses. In the study by Vini et al.,[21] the median cumulative dose activity of RAI received according to menstrual history after RAI ranged as follows: normal menses 81 mCi (range 30-972 mCi), irregular menses 378 mCi (range 30-810 mCi), and transient absence of menses 378 mCi (range 230-1593 mCi) (no statistical comparison described). By contrast, Raymond et al.[15] reported that there was no significant difference in the mean dose activity of RAI received between 18 women who experienced transient ovarian failure (278 mCi, SD 5 mCi) and the 48 women with DTC who did not (265 mCi, SD 62 mCi) (P = NS). The potential relationship between oral contraceptive use at the time of RAI treatment and protection of subsequent menstrual function was examined only in the study by Raymond et al.,[15] who reported that the use of an oestrogen/progestin contraceptive at the time of RAI therapy was not protective for continuation of menses after adjustment for age.

We examined whether women with DTC treated with RAI experienced menopause at a younger age if they were treated with RAI (compared to women not treated with RAI). In both studies examining this issue,[8,20] women with DTC who had been treated with RAI for DTC experienced menopause at a statistically significantly younger age (mean or median age late forties) compared to women with goitre treated with levothyroxine[8] or those with DTC who were not treated with RAI[20] (median or mean age of menopause in the early fifties for the respective control groups). Furthermore, Souza Rosario et al.[20] reported that the menopausal age of 36 women treated with RAI activities of 100 mCi was 48 years (SD 3·6) compared to 45 years (SD 3·3) in women treated with activities > 100 mCi (P < 0·003), suggesting a dose effect. However, Ceccarelli et al.[8] reported that there was no significant difference in menopausal age in the RAI-treated women when subdivided according to age at initial treatment (< 40 or ≥ 40 years), age at time of administration of last RAI treatment (< 45 or ≥ 45 years), the number of administrations (one or more), or the cumulative dose activity received (< 100 mCi or > 100 mCi). Thus, although RAI treatment was associated with a slightly earlier age of menopause, there were conflicting data with respect to any potential dose effect.

Reproductive outcomes after RAI treatment for DTC were reported in nine studies[6,7,9,10,11,13,14,16,17] ( Table 4 ). Reproductive outcomes after RAI treatment were collected from a total of at least 591 pregnancies and 496 live births (not all studies reported the total number of pregnancies) ( Table 4 ). The time from RAI treatment until conception varied from 0 to 222 months ( Table 4 ). There were two studies specifically reporting fertility or infertility rates after RAI treatment.[10,16] In the study by Dottorini et al.,[10] the fertility rate (defined by the authors as ratio of live births per 1000 fertile female years per year) was 23 in RAI-treated women and 19 in women with DTC not treated with RAI (P = 0·39 for the difference). Sarkar et al.[16] assumed a diagnosis of infertility if conception had not occurred after more than a year of marriage in the absence of contraception. In this study, the infertility rate in RAI-treated women was 10% (2/20); relative to a reported 'normal population' infertility rate of 12% (normal population rate not sex specific). Although not specifically defined as 'fertility outcomes', the number of live births or proportion of live births from all pregnancies was compared to a control group in another five studies.[6,9,11,14,17] In a study by Chow et al.,[9] the percentage of live births in women treated with RAI was 68% (78/116), whereas the rate in women with DTC who did not receive RAI was 74% (81/110) (no statistical comparison performed, data include induced abortions in the pregnancy rate). However, upon excluding induced abortions from the data by Chow et al.,[9] the live birth rate per pregnancy is estimated to be 92% (81/88) in RAI-treated women and 84% (78/93) in women with DTC not treated with RAI. In a study by Schlumberger et al.,[17] the percentage of live births per total number of pregnancies was 80% (164/206) in RAI-treated women, compared to 77% (51/66) in women with DTC who did not receive RAI (no statistical comparison, data exclude induced abortions). Pomorski et al.[14] reported that seven RAI-treated women who became pregnant all gave birth to live children, and that 12 women with DTC not treated with RAI who became pregnant also all gave birth to healthy children, suggesting that 100% of pregnancies resulted in live births in each group. In studies by Balenovic and Kusic[6] and Einhorn et al.,[11] the number of live births for women treated with RAI was compared to the number prior to RAI treatment, but the total number of pregnancies was not specified. In the Einhorn study,[11] assuming the number of pregnancies is equal to the number of spontaneous abortions plus live births, the estimated live birth rates are 80% (12/15) pre-RAI and 90% (9/10) after RAI. To summarize, after excluding induced abortions, the long-term fertility rates and the percentage of live births per pregnancy are generally not dissimilar in women with DTC treated with RAI compared to various control groups.

The rates of miscarriages or induced abortions were reported in seven studies[6,9,10,11,13,16,17] ( Table 4 ). The rates of miscarriage in women who had been treated with RAI varied from 2·2% to 20% of subsequent pregnancies.[6,9,10,11,13,16,17] The miscarriage rates were generally not clinically or statistically significantly higher in women who had a history of RAI treatment compared to reference populations or other women with DTC who were not treated with RAI,[6,9,10,11,13,16,17] with the possible exception of the Schlumberger study,[17] in which the rates of miscarriage were reported to be: 20% in women with DTC treated with RAI, 22% in women with DTC not treated with RAI, and 10% in women with DTC prior to any treatment. In this study, the rates of miscarriage in women with DTC treated with or without RAI were similar, but the rates were approximately double that of women prior to any DTC treatment; these data were not subject to a statistical comparison. The induced abortion rates in pregnancies after RAI were examined in studies by Chow et al.[9] and Schlumberger et al.[17] The rates of induced abortion in women with DTC treated with RAI were approximately one in five pregnancies, and these rates were similar in women with DTC not treated with RAI in both studies.[9,17] However, in women who had DTC but had not yet been treated by any modality, the induced abortion rate was reported to be about 10% by Schlumberger et al.[17] The relationship between the time interval from RAI treatment until conception and miscarriage or induced abortion outcomes was examined in the studies by Chow et al.,[9] Lin et al.[13] and Schlumberger et al.[17] Schlumberger et al. reported that RAI exposure within 1 year preceding conception was significantly associated with an increased risk of miscarriage [40% (4/10) of pregnancies in exposed women compared to 18% (36/202) of pregnancies in unexposed women], with the 95% confidence interval for the difference in risk being 10-70%.[17] This association remained significant after adjustment for the mother's age at conception and the cumulative dose of RAI.[17] Schlumberger et al.[17] also reported a higher frequency of induced abortions in 20 pregnancies in patients who had received > 10 mCi of RAI during the year preceding conception [50% (10/20) of pregnancies compared to 18% (45/247) of pregnancies in women who did not receive RAI (P < 0·001)]. Chow et al.[9] also reported that about half of pregnancies conceived within 1 year of RAI treatment for DTC were terminated by induced abortion (8/15) but no miscarriages were reported in this time interval [compared to an induced abortion rate of 16% (16/101) and a miscarriage rate of 13% (13/101) for pregnancies conceived more than 1 year after RAI treatment]. By contrast, Lin et al.[13] reported the miscarriage rate to be the same (14%) for pregnancies conceived within 6 months after RAI (1/7) or > 6 months after RAI (7/51).

There were nine studies in which pregnancy and infant outcomes, such as preterm births, stillbirths or infant mortality, were compared for RAI-treated women with DTC relative to control groups[6,7,9,10,11,13,14,16,17] (excluding miscarriage and induced abortion outcomes) ( Table 4 ). The risk of preterm births (≤ 37 weeks gestation, or undefined by primary investigators) per total live births was found to be significantly clinically or statistically increased in only one[9] of five studies examining this outcome.[9,10,14,16,17] Chow et al.[9] reported that the rate of preterm births (≤ 37 weeks) was significantly higher in pregnancies of RAI-treated women at 9% (7/78 births) compared to pregnancies of women who did not receive RAI at 1·2% (1/110) (P = 0·04). The rate of stillbirth in pregnancy was not significantly clinically or statistically increased in RAI-treated women in all four studies examining this outcome[9,10,14,17] ( Table 4 ). The risk of neonatal mortality for offspring of RAI-treated women was not significantly clinically or statistically increased in four studies comparing this outcome with a control group[9,11,14,17] ( Table 4 ). There were five studies in which the rate of congenital birth defects was compared in offspring of RAI-treated women to offspring of controls[9,10,14,16,17] ( Table 4 ). In a study by Dottorini et al.,[10] one of the 65 children born to RAI-treated women (1·5%) had a ventricular septal defect and patent ductus arteriosis; the historical European rate for cardiac malformations was quoted to be 44·3 per 100 000 live births (0·4%). In addition, in a study by Sarkar et al.,[16] one of the 43 live children (2·3%) born to RAI-treated women had a congenital cardiac anomaly and this result was compared to a historical 'normal population' congenital anomaly prevalence rate of 1-2%. However, in the respective studies by Chow et al.[9] and Pomorski et al.,[14] no congenital anomalies were observed in any of the groups. Furthermore, Schlumberger et al.[14] reported the following rates of congenital anomalies: 4% (57/1384) before any treatment DTC, 3% (6/215) after surgery for DTC, 4% (2/51) for no RAI, and 2% (4/164) for any RAI (with no cardiac malformations in the RAI group). Thus, an increased risk of congenital anomalies in offspring of women with DTC was not confirmed, particularly in studies using contemporaneous (non-historical) controls.[9,14,17]

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....