Thyroid 1992.2:147-152. Downloaded from online.liebertpub.com by Uc Davis Libraries University of California Davis on 02/09/15. For personal use only.
THYROID Volume 2, Number 2, 1992 Mary Ann Liebert, Inc., Publishers
Monitoring Thyroxine Treatment During Pregnancy MICHAEL M. KAPLAN
ABSTRACT Data were analyzed from 77 pregnancies in 65 hypothyroid women treated with Ievothyroxine (T4) to determine (a) how often, how severely, and when serum thyrotropin (TSH) and free T4 concentrations become abnormal during pregnancy in women taking constant T4 doses and (b) how much to increase T4 doses to normalize elevated serum TSH levels. Group I consisted of 36 women with previous thyroid ablation. Group II consisted of 29 women with Hashimoto's thyroiditis. Serum TSH levels rose above normal (> 4 piU/mL) during pregnancy more often in Group I (76%) than in Group II (47%) during treatment with constant T4 doses, and mean serum free T4 levels decreased significantly in both groups, by 47% in Group I and by 35% in Group II, compared to preconception values. Serum TSH concentrations rose above 20 piU/mL in 22% of the women, and serum free T4 concentrations fell below normal in 13%. Elevated serum TSH levels usually were detected by the first TSH test during pregnancy, as early as 4 weeks gestation. Of those women tested after an initial normal serum TSH value during pregnancy, 30% had an elevated serum TSH concentration later in the pregnancy. Of 12 women followed through two pregnancies, 10 had elevated serum TSH values in both pregnancies, 1 had normal serum TSH values in both, and 1 had discordant serum TSH values. Mean (± SD) T4 dose increments needed to normalize serum TSH values were 41 ± 24 pug T4/day for elevated serum TSH values 20 pIJ/mL. Optimal postpartum T4 doses were approximately the same as those that produced normal serum TSH levels before conception. From these findings, it seems reasonable to advise pregnant hypothyroid women treated with T4 to have serum TSH tests at 8 weeks and 6 months gestation. An elevated TSH value should be corrected by an increased T4 doses. The required dose increment can be estimated from the abnormal TSH value. After pregnancy, optimal preconception doses can be resumed.
INTRODUCTION
Women priately
in their reproductive years are
treated appro-
with
Ievothyroxine (T4) for several reasons: to correct spontaneous or iatrogenic hypothyroidism, to try to shrink a goiter or thyroid nodule, or to suppress TSH secretion after thyroid cancer surgery. Most such women have normal fertility, and their thyroid diseases do not contraindícate pregnancy. Therefore, many women taking T4 will conceive, and T4 treatment should be continued during the pregnancy.
Anecdotal information available for some years indicated that hypothyroid women need higher doses of T4 during pregnancy than in the nonpregnant state to maintain a normal serum TSH level (1,2). An.uncorrected elevation in the serum some
Farmbrook Medical Two, Southfield.
TSH concentration can be clinically important because hypothyroidism during pregnancy is associated with a higher rate of spontaneous abortion and preeclampsia, among other complications (3), and may cause the mother to have symptoms of hypothyroidism during the pregnancy. Although many reviews and current textbooks of endocrinology and thyroid diseases do not discuss a need to adjust T4 doses during pregnancy or imply that there is no such need (4-8), several studies indicate that an altered T4 dose requirement is common (9-11 ). In fact, the need to adjust the T4 dose upward during pregnancy may be more easily recognized now than in the past. The ready availability of high sensitivity serum thyrotropin (TSH) assays has made T4 treatment more precise. For this and other reasons, current average replacement doses of T4 are lower than those used
Michigan. 147
Thyroid 1992.2:147-152. Downloaded from online.liebertpub.com by Uc Davis Libraries University of California Davis on 02/09/15. For personal use only.
148
KAPLAN
before the mid-1980s. If pregnancy increases the T4 dose requirement, current prepregnancy doses would tend to be insufficient during pregnancy more frequently than the T4 doses employed previously, which were often higher than necessary. Practical guidelines are needed for monitoring and adjusting T4 therapy during pregnancy. How early in pregnancy does the dose requirement change? Is there a progressive increase in the required T4 dose as pregnancy progresses? When T4 underreplacement is identified during pregnancy, how much of an increase in T4 dose is needed to restore the patient to a euthyroid state? Is the cause of hypothyroidism—Hashimoto's thyroiditis vs thyroid ablation by radioactive iodine or surgery—a variable in patients' responses to exogenous T4 during pregnancy? To address these questions, I have analyzed serum free T4 and high sensitivity TSH results and T4 treatment doses from 77 pregnancies in 65 women treated for primary hypothyroidism.
PATIENTS AND METHODS
Group I was 36 women who had been treated with l3lI, total thyroidectomy, or both for hyperthyroidism or thyroid cancer. Group II was 29 women who had documented hypothyroidism due to Hashimoto's thyroiditis, as established by elevated serum titers of antithyroglobulin or antithyroperoxidase (microsomal) antibodies, goiter, and an abnormally high serum TSH concentration before T4 treatment. All patients took the same brand of Ievothyroxine sodium (Levothroid, Rorer, Fort Washington, Pennsylvania [manufacturer at time of this study]) at all times. Fifty-nine women stated that they took their T4 tablets regularly before and during pregnancy. The other 6 women were lost to follow-up and could not be queried about compliance. The study
groups include all women who met the criteria stated above, seen during pregnancy between 1987 and 1991 except 1, excluded because she reported skipping T4 doses frequently. Since paired sequential observations in individual patients give the most reliable information, I have used paired data for most analyses. However, the data set is incomplete for several reasons. Some Group II patients were diagnosed as having thyroid disease by their obstetricians during the pregnancy and are, therefore, lacking prepregnancy test results. Data from most of the women are obtained by retrospective review of their medical records, during a period when evaluation was advised at 6 months gestation, unless their annual examination happened to fall earlier in the pregnancy. Therefore, the timing of thyroid testing during their pregnancies was not systematic. Consequently, the numbers of observations for the various aspects of the analysis are less than the total number of study subjects. Additional unpaired comparisons were made using data from nonpregnant women with hypothyroidism due to Hashimoto's thyroiditis or thyroid ablation. Women were selected for these comparison groups according to the following criteria: they were not in the study groups, had documented hypothyroidism, were euthyroid on T4 treatment at the time of the office visit from which data were taken, were 21-39 years of age, and were not taking any drugs known to affect thyroid function tests. The comparison groups were the first 20 women with Hashimoto's thyroiditis and the first 19 women who had thyroid ablation and who met the other criteria, taken from a consecutive series of office visits in 1990.
Blood samples were obtained at least 9 h after the most recent dose. Serum free T4 concentrations were measured by a kinetic immunoextraction radioimmunoassay with a normal range of 0.6-2.2 ng/dL (Gamma-Coat two step, Clinical Assays, Cambridge, MA). TSH concentrations were measured in 5 pregnancies by an immunoradiometric assay with a detection limit of 0.1 p,U/mL and a normal range of 0.4-5.0 uXI/mL (Hybritech, San Diego, CA) and in the rest by an immunochemiluminescent assay with a detection limit of 0.05 p,U/mL and a normal range of 0.25-4.0 p-U/rnL (Magic Lite, Ciba Corning Diagnostics Corp., Medfield, MA). Interassay coefficients of variation were 10% for free T4 values within the normal range and less than 10% for both TSH assays at TSH concentrations of 0.3 p.U/mL and higher. Results presented as paired values were measured in the same assay method. Samples were assayed within 3 days of blood drawing. Statistical comparisons of means were made using two-tailed r-tests, paired or nonpaired as appropriate. TSH values below the limits of detectability were assigned the sensitivity limit value for calculations and statistical tests. Values are given as mean ± SD.
T4
RESULTS
Forty-two women had tests both before and during pregnancy while taking a constant T4 dose that produced normal serum TSH concentrations before pregnancy. The serum TSH levels were elevated above normal at some time during pregnancy in 19 of 25 such women (76%) in Group I and in 8 of 17 (47%) such women in Group II. Of the remaining women, 5 of 6 in Group I and 4 of 9 in Group II had higher serum TSH concentrations during pregnancy than before, though the levels during pregnancy remained in the normal range. No woman in either group who took a constant T4 dose developed either a serum free T4 concentration above normal or a serum TSH concentration below normal during pregnancy. Table 1 shows the mean serum free T4 and TSH concentrations in the 42 women on constant T4 doses for whom paired data before and during pregnancy were available. The mean daily T4 doses were identical in the two groups. Because of the weight gain of pregnancy, there were 6%-7% decreases during pregnancy in the mean T4 doses expressed per kilogram body weight, but mean serum free T4 concentrations decreased by a much greater extent, 47% and 35% in the two groups. Regarding the timing of the TSH elevations, in most cases, the TSH was elevated at the time of the earliest test during pregnancy: 17 in the first trimester, 14 in the second trimester, 4 in the third trimester. Six women were tested at 4-8 weeks gestation: the serum TSH level was high in 1 and normal in 1 at 4 weeks gestation, high in 1 at 6 weeks, and high in 3 at 8 weeks. However, this high frequency of abnormal TSH values early in pregnancy may not be a representative sample. Several of these women had their abnormal TSH concentrations first identified by their obstetricians. Those who had normal TSH results obtained by the obstetricians probably were not referred for further evaluation. Nineteen women had normal serum TSH concentrations in the first trimester (between 9 and 12 weeks in 18) and were tested later in gestation on the same T4 dose: 1 had an elevated TSH level in the late first trimester, 3 had elevated TSH levels in the second trimester, and 1 had an elevated TSH
MONITORING T4 IN PREGNANCY Table 1. Serum Free
T4
and
149
TSH Concentrations Before and During Pregnancy T4 Doses"
Thyroid 1992.2:147-152. Downloaded from online.liebertpub.com by Uc Davis Libraries University of California Davis on 02/09/15. For personal use only.
Group I (n
T4 dose, p-g/day T4 dose, p-g/kg/day Serum free T4, ng/dL Serum TSH, p,U/mL "Normal ranges: free T4 0.6-2.2 bp < 0.05.
cp
THYROID Volume 2, Number 2, 1992 Mary Ann Liebert, Inc., Publishers
Monitoring Thyroxine Treatment During Pregnancy MICHAEL M. KAPLAN
ABSTRACT Data were analyzed from 77 pregnancies in 65 hypothyroid women treated with Ievothyroxine (T4) to determine (a) how often, how severely, and when serum thyrotropin (TSH) and free T4 concentrations become abnormal during pregnancy in women taking constant T4 doses and (b) how much to increase T4 doses to normalize elevated serum TSH levels. Group I consisted of 36 women with previous thyroid ablation. Group II consisted of 29 women with Hashimoto's thyroiditis. Serum TSH levels rose above normal (> 4 piU/mL) during pregnancy more often in Group I (76%) than in Group II (47%) during treatment with constant T4 doses, and mean serum free T4 levels decreased significantly in both groups, by 47% in Group I and by 35% in Group II, compared to preconception values. Serum TSH concentrations rose above 20 piU/mL in 22% of the women, and serum free T4 concentrations fell below normal in 13%. Elevated serum TSH levels usually were detected by the first TSH test during pregnancy, as early as 4 weeks gestation. Of those women tested after an initial normal serum TSH value during pregnancy, 30% had an elevated serum TSH concentration later in the pregnancy. Of 12 women followed through two pregnancies, 10 had elevated serum TSH values in both pregnancies, 1 had normal serum TSH values in both, and 1 had discordant serum TSH values. Mean (± SD) T4 dose increments needed to normalize serum TSH values were 41 ± 24 pug T4/day for elevated serum TSH values 20 pIJ/mL. Optimal postpartum T4 doses were approximately the same as those that produced normal serum TSH levels before conception. From these findings, it seems reasonable to advise pregnant hypothyroid women treated with T4 to have serum TSH tests at 8 weeks and 6 months gestation. An elevated TSH value should be corrected by an increased T4 doses. The required dose increment can be estimated from the abnormal TSH value. After pregnancy, optimal preconception doses can be resumed.
INTRODUCTION
Women priately
in their reproductive years are
treated appro-
with
Ievothyroxine (T4) for several reasons: to correct spontaneous or iatrogenic hypothyroidism, to try to shrink a goiter or thyroid nodule, or to suppress TSH secretion after thyroid cancer surgery. Most such women have normal fertility, and their thyroid diseases do not contraindícate pregnancy. Therefore, many women taking T4 will conceive, and T4 treatment should be continued during the pregnancy.
Anecdotal information available for some years indicated that hypothyroid women need higher doses of T4 during pregnancy than in the nonpregnant state to maintain a normal serum TSH level (1,2). An.uncorrected elevation in the serum some
Farmbrook Medical Two, Southfield.
TSH concentration can be clinically important because hypothyroidism during pregnancy is associated with a higher rate of spontaneous abortion and preeclampsia, among other complications (3), and may cause the mother to have symptoms of hypothyroidism during the pregnancy. Although many reviews and current textbooks of endocrinology and thyroid diseases do not discuss a need to adjust T4 doses during pregnancy or imply that there is no such need (4-8), several studies indicate that an altered T4 dose requirement is common (9-11 ). In fact, the need to adjust the T4 dose upward during pregnancy may be more easily recognized now than in the past. The ready availability of high sensitivity serum thyrotropin (TSH) assays has made T4 treatment more precise. For this and other reasons, current average replacement doses of T4 are lower than those used
Michigan. 147
Thyroid 1992.2:147-152. Downloaded from online.liebertpub.com by Uc Davis Libraries University of California Davis on 02/09/15. For personal use only.
148
KAPLAN
before the mid-1980s. If pregnancy increases the T4 dose requirement, current prepregnancy doses would tend to be insufficient during pregnancy more frequently than the T4 doses employed previously, which were often higher than necessary. Practical guidelines are needed for monitoring and adjusting T4 therapy during pregnancy. How early in pregnancy does the dose requirement change? Is there a progressive increase in the required T4 dose as pregnancy progresses? When T4 underreplacement is identified during pregnancy, how much of an increase in T4 dose is needed to restore the patient to a euthyroid state? Is the cause of hypothyroidism—Hashimoto's thyroiditis vs thyroid ablation by radioactive iodine or surgery—a variable in patients' responses to exogenous T4 during pregnancy? To address these questions, I have analyzed serum free T4 and high sensitivity TSH results and T4 treatment doses from 77 pregnancies in 65 women treated for primary hypothyroidism.
PATIENTS AND METHODS
Group I was 36 women who had been treated with l3lI, total thyroidectomy, or both for hyperthyroidism or thyroid cancer. Group II was 29 women who had documented hypothyroidism due to Hashimoto's thyroiditis, as established by elevated serum titers of antithyroglobulin or antithyroperoxidase (microsomal) antibodies, goiter, and an abnormally high serum TSH concentration before T4 treatment. All patients took the same brand of Ievothyroxine sodium (Levothroid, Rorer, Fort Washington, Pennsylvania [manufacturer at time of this study]) at all times. Fifty-nine women stated that they took their T4 tablets regularly before and during pregnancy. The other 6 women were lost to follow-up and could not be queried about compliance. The study
groups include all women who met the criteria stated above, seen during pregnancy between 1987 and 1991 except 1, excluded because she reported skipping T4 doses frequently. Since paired sequential observations in individual patients give the most reliable information, I have used paired data for most analyses. However, the data set is incomplete for several reasons. Some Group II patients were diagnosed as having thyroid disease by their obstetricians during the pregnancy and are, therefore, lacking prepregnancy test results. Data from most of the women are obtained by retrospective review of their medical records, during a period when evaluation was advised at 6 months gestation, unless their annual examination happened to fall earlier in the pregnancy. Therefore, the timing of thyroid testing during their pregnancies was not systematic. Consequently, the numbers of observations for the various aspects of the analysis are less than the total number of study subjects. Additional unpaired comparisons were made using data from nonpregnant women with hypothyroidism due to Hashimoto's thyroiditis or thyroid ablation. Women were selected for these comparison groups according to the following criteria: they were not in the study groups, had documented hypothyroidism, were euthyroid on T4 treatment at the time of the office visit from which data were taken, were 21-39 years of age, and were not taking any drugs known to affect thyroid function tests. The comparison groups were the first 20 women with Hashimoto's thyroiditis and the first 19 women who had thyroid ablation and who met the other criteria, taken from a consecutive series of office visits in 1990.
Blood samples were obtained at least 9 h after the most recent dose. Serum free T4 concentrations were measured by a kinetic immunoextraction radioimmunoassay with a normal range of 0.6-2.2 ng/dL (Gamma-Coat two step, Clinical Assays, Cambridge, MA). TSH concentrations were measured in 5 pregnancies by an immunoradiometric assay with a detection limit of 0.1 p,U/mL and a normal range of 0.4-5.0 uXI/mL (Hybritech, San Diego, CA) and in the rest by an immunochemiluminescent assay with a detection limit of 0.05 p,U/mL and a normal range of 0.25-4.0 p-U/rnL (Magic Lite, Ciba Corning Diagnostics Corp., Medfield, MA). Interassay coefficients of variation were 10% for free T4 values within the normal range and less than 10% for both TSH assays at TSH concentrations of 0.3 p.U/mL and higher. Results presented as paired values were measured in the same assay method. Samples were assayed within 3 days of blood drawing. Statistical comparisons of means were made using two-tailed r-tests, paired or nonpaired as appropriate. TSH values below the limits of detectability were assigned the sensitivity limit value for calculations and statistical tests. Values are given as mean ± SD.
T4
RESULTS
Forty-two women had tests both before and during pregnancy while taking a constant T4 dose that produced normal serum TSH concentrations before pregnancy. The serum TSH levels were elevated above normal at some time during pregnancy in 19 of 25 such women (76%) in Group I and in 8 of 17 (47%) such women in Group II. Of the remaining women, 5 of 6 in Group I and 4 of 9 in Group II had higher serum TSH concentrations during pregnancy than before, though the levels during pregnancy remained in the normal range. No woman in either group who took a constant T4 dose developed either a serum free T4 concentration above normal or a serum TSH concentration below normal during pregnancy. Table 1 shows the mean serum free T4 and TSH concentrations in the 42 women on constant T4 doses for whom paired data before and during pregnancy were available. The mean daily T4 doses were identical in the two groups. Because of the weight gain of pregnancy, there were 6%-7% decreases during pregnancy in the mean T4 doses expressed per kilogram body weight, but mean serum free T4 concentrations decreased by a much greater extent, 47% and 35% in the two groups. Regarding the timing of the TSH elevations, in most cases, the TSH was elevated at the time of the earliest test during pregnancy: 17 in the first trimester, 14 in the second trimester, 4 in the third trimester. Six women were tested at 4-8 weeks gestation: the serum TSH level was high in 1 and normal in 1 at 4 weeks gestation, high in 1 at 6 weeks, and high in 3 at 8 weeks. However, this high frequency of abnormal TSH values early in pregnancy may not be a representative sample. Several of these women had their abnormal TSH concentrations first identified by their obstetricians. Those who had normal TSH results obtained by the obstetricians probably were not referred for further evaluation. Nineteen women had normal serum TSH concentrations in the first trimester (between 9 and 12 weeks in 18) and were tested later in gestation on the same T4 dose: 1 had an elevated TSH level in the late first trimester, 3 had elevated TSH levels in the second trimester, and 1 had an elevated TSH
MONITORING T4 IN PREGNANCY Table 1. Serum Free
T4
and
149
TSH Concentrations Before and During Pregnancy T4 Doses"
Thyroid 1992.2:147-152. Downloaded from online.liebertpub.com by Uc Davis Libraries University of California Davis on 02/09/15. For personal use only.
Group I (n
T4 dose, p-g/day T4 dose, p-g/kg/day Serum free T4, ng/dL Serum TSH, p,U/mL "Normal ranges: free T4 0.6-2.2 bp < 0.05.
cp
