THYROID Volume 2, Number 1, 1992 Mary Ann Liebert, Inc., Publishers
Goiter and
Pregnancy:
A New
Insight into an Old Problem
DANIEL GLINOER and MARC LEMONE
ABSTRACT Evidence is presented that pregnancy constitutes a goitrogenic stimulus, particularly in conditions with a restricted or even a marginally low iodine intake. In a series of studies carried out in a large cohort of pregnancies in the Brussels area, the authors show that an increase in thyroid volume is observed in a majority of pregnant women, leading to goiter formation at delivery in 9% of the cases. Furthermore, increments in thyroid volume were correlated with biochemical evidences of functional stimulation of the thyroid, such as an elevation in serum TG levels, preferential T, secretion, and slight increases in basal TSH at delivery. Hence, the association of biochemical features of thyroidal stimulation with volumetric changes in the gland strongly suggests that pregnancy truly induces goitrogenesis rather than vascular swelling ("intumescence") alone, at least in conditions with a low iodine intake. Finally, preliminary data from this laboratory, as well as recently published data from other investigators, suggest that goiter formation during pregnancy can easily be prevented by increasing the iodine supply during pregnancy.
HISTORICAL BACKGROUND
(5) stressed the diversity of opinions on the topic and stated that "the main question is whether goiter is an acceptable normal physiological finding in pregnancy or whether it represents
A Egyptians
quotation mentions that ancient used to tie a reed around a women's neck to assess whether she was pregnant: if the reed snapped, it was the proof that she was indeed pregnant. Even though I have never been able to confirm this test in my own practice or actually find the original reference, the quotation suggests that pharaonic Egyptians had already noted that throat enlargements were somehow associated with gestation (1,2). More recently, two apparently contradictory studies have been reported by Crooks et al. in 1967 (3) and Levy et al. in 1980 (4). Crooks et al. compared the incidence of visible and palpable goiter in preg¬ nant women between Aberdeen (with a lower iodine intake) and Reykjavik (with a higher iodine intake). They showed that the incidence was three-fold higher in Scotland compared to Iceland and that it doubled during pregnancy in the lower and remained virtually unchanged in the higher iodine area. In contrast. Levy et al. investigated 98 women from northeastern Ohio with an adequate iodine intake (> 200 µg/day) and found no difference in goiter frequency between pregnant and nonpregnant matched patients. These apparently contradictory findings have produced a decade of debate and confusion about the possible goitrogenic role of pregnancy. In a recent review article. Becks and Burrow classic historical
Université Libre de Bruxelles,
thyroid pathology."
THYROID VOLUME AND IODINE DEFICIENCY
Although North American thyroidologists suggest that goiter is infrequent (6.7), European clinicians appreciate that goiter formation is rather common during gestation, in particular in populations with a clear deficit in daily iodine intake. In the former East Germany, an area where the average iodine intake was below 30 µg/day, a goiter has been reported in 60% of pregnancies (8). Until recently, however, precise data on alterations in thyroid volume (TV) during pregnancy were scanty. Moreover, since thyroid radionuclide scans are not performed during pregnancy, the older studies had to rely solely on the clinical estimates of goiter (e.g., visible or palpable, grade), which are insensitive for detecting subtle TV changes. Today, the routine use of thyroid ultrasonography has been added to our armamentarium, allow¬ ing for precise determinations of TV, with a high sensitivity for the detection of even minor thyroid alterations (9,10). Using
Hospital Saint-Pierre, Thyroid Investigation Clinic and Department of Radiology. Brussels, Belgium. 65
66
GLINOER AND LEMONE
ultrasound determinations, it is possible to compare absolute sizes in different areas and to relate those to the environment. Figure 1 compares several areas in Europe and shows clearly that TV in "average" populations varies as a function of the iodine intake (10-12). With intake above 100 µg/day, normal TV averages 10 mL, whereas with intake around 50 µg iodine/day, TV in the normal population is almost doubled.
glandular
Table 1. Thyroid Volume: Relationship and Body Weight
Females
Males
THYROID VOLUME IN PREGNANCY In order to discuss TV alterations associated with pregnancy, three features should be considered first.
Age (years)
TV
BW
(mean mL)
(kg)
23 39 59 75 23 39 59 76
15.6 17.7 17.9 20.5 16.8 21.0 21.3 20.4
58 61 68 65 69 76 77 70
Adapted from ref.
with
Sex, Age,
Vol/wt 0.27 0.29 0.27 0.32 0.25 0.28 0.28 0.29
0.27
0.29
11.
Thyroid volume and body weight
Thyroid intumescence
subjects, the size of the thyroid is correlated with body weight. TV is larger in males than in females (11). Moreover, when the ratio of TV/body weight is considered in healthy adult subjects, the sex and age variations in TV are almost abolished (Table 1). Similarly, there is a good relation¬ ship between age or weight and TV in adolescents (13-15). Hence, the closest relationship in both adolescent and adult subjects is found with body surface or lean body mass.
Pregnancy is a clinical situation that obviously involves body modifications, and some authors have proposed the use of algorithms to allow for corrections in the calculation of TV. However, these have not provided important advantages (11). Furthermore, gestation also is associated with an overall in¬ crease in blood flow (16), and, therefore, vascular swelling, or intumescence, has been implicated as a possible cause of thyroid enlargement during pregnancy (17).
In adult
mass
Individual variations in TV In a given area with similar iodine intake levels, there is a wide range of TV in normal subjects. In 166 healthy women in the first trimester of gestation, we observed glandular volumes ranging from a low of 5 mL to 22 mL. the upper limit of normality in our area (Fig. 2). The findings were somewhat surprising, since these women had no detectable thyroid abnor¬ malities and were strictly euthyroid. The data could be inter¬ preted to indicate that small glands are capable of ensuring First trimester of pregnancy
(N=166)
Mean: 11.7 mL Median: 10.8 mL
Range: A
Sweden (Stockholm) Netherlands
I 9
4.7-21.g mL
Denmark
(Amsterdam)
(Copenhagen)
Germany »various cities)
THYROID VOLUME (ML)
IODINE EXCRETION (ug
g
créât)
FIG. 2. Distribution frequency of thyroid volume during first trimester of pregnancy in healthy subjects (n 166). The upper limit of normality in the area is 22 mL. The mean TV was 1 1.7 mL (median 10.8 mL), with a range from 4.7 mL to 21.9mL. In comparison to other European cities, the results indicate that there is no significant iodine deficiency in Brussels. =
FIG. 1. Correlation between thyroid volume and urinary iodine concentration in normal populations of Stockholm, Am¬ sterdam, Copenhagen, and various cities in Germany. The graph was reconstructed from data in refs. 10-12.
GOITER AND PREGNANCY
67
normal thyroid function, at least as long as no additional factor intervenes to disturb the homeostasis of the system. We have reason to believe, however, that small glands may be less capable of adapting thyroid hormone production to chronic stimulatory events, such as pregnancy, without marked changes in TV. In pregnancy, systematic studies of TV have been lacking. In 1988, a large prospective cohort study was undertaken in our institution to evaluate the regulation of maternal thyroid function (18). Our aim was to assess both functional and volumetric alterations of the thyroid in both healthy pregnant women (with no detectable thyroid abnormalities) and patients with mild underlying thyroid abnormalities (19). These studies repre¬ sented a significant undertaking. The original cohort encom¬ passed a total of over 730 women, investigated cross-sectionally and followed sequentially during gestation and, for some of them, also after parturition. Interestingly, the study was per¬ formed in an urban area without significant iodine deficiency but where the daily iodine intake was nevertheless marginally low, between 50 and 80 µg/day (20). Our hypothesis was that a marginally low iodine supply, representative of the situation found in many western European cities (21-23), could lead to a further decrease in its availability for the maternal thyroid during pregnancy, due to the increased iodide renal clearance (24) and losses to the fetoplacental complex during late gestation (25). Starting in a situation with a limited iodine supply, the dimin¬ ished iodine availability could lead to a state of relative iodine deficiency, thereby enhancing the physiologic changes in thy¬ roid economy. In other words, our aim was to assess whether there was a continuum from physiology to pathology, driven in part by the environmental iodine supply.
THYROID REGULATION IN PREGNANCY In pregnancy, the normal thyroid is faced with a triple challenge as a result of three separate factors acting in concert to stimulate the gland: (a) the increased thyroid hormone-binding capacity of serum due to the marked increase in serum TBG
concentrations, (b) the effects of elevated hCG levels on the thyroid (and on TSH), and (c) the reduced availability of iodine
for the maternal thyroid in conditions with a limited dietary iodine supply (18). Our studies showed that variable patterns of thyroid adjustment to these combined factors take place in normal pregnancy. Approximately one third of the subjects were characterized by having relative hypothyroxinemia, preferential T3 secretion, and a higher setting of the pituitary thyrostat, i.e., a pattern of increased glandular stimulation. In addition, an
intrinsic TSH-like activity of hCG was observed, and the data suggested strongly that hCG acts directly to stimulate the thyroid. During the first trimester of gestation, blunted serum TSH levels were found in 14% of healthy subjects (Table 2). Gestational transient thyrotoxicosis was, however, exceedingly rare, since free T4 remained in the normal range in most instances. In a few women (1.4%), high hCG levels (usually above 70,000-80,000 IU/L) were associated with gestational thyrotoxicosis (often accompanied by hyperemesis gravidarum), with free T4 levels unequivocally in the thyrotoxic range (2.5, 2.7, 3.7 ng/dL). Finally, in relation to the metabo¬ lism of iodide, the data suggested that the relative iodine deficiency in Brussels had an overall permissive role in allowing for enhancement of the stimulatory changes in the thyroidal economy.
MATERNAL REPERCUSSIONS ON THE THYROID
Repercussions on the maternal thyroid included significant increases in serum thyroglobulin (TG) levels, especially during the last trimester (Table 3). More than half of the women had elevated serum TG levels at term, with individual values up to 180 µg/L. The other major finding was that TV increased in the vast majority of women. The average increment was.20%, but it varied widely among individuals up to 130% (Table 4). True goiter, defined as TV greater than 22 mL, was found in 9% of the cohort at delivery, with volumes up to 46 mL. Furthermore, changes in size were negatively correlated to initial size, so that for TV initially under 10 mL, the average increment exceeded 30%. An increase in TV was associated with biochemical features of thyroid stimulation, such as an increased ,/T4 molar ratio or slightly increased—albeit still normal—TSH level at delivery. The most significant correlation was between TG and TV. The greater the changes in TV, the more frequently TG was elevated. TG elevation, therefore, reflects anatomic changes and constitutes a marker of an increase in TV during pregnancy (18,26). Another interesting finding was that larger volumes at term were associated with higher hCG levels during the first trimester, indicating a potential role of hCG to stimulate not only thyroid function but also glandular enlargement. By measuring TV with echography and in the absence of histologie evidence, it was impossible to distinguish glandular hyperplasia from hypertrophy or appreciate the role of changes in blood flow occurring during pregnancy. However, the asso¬ ciation of biochemical features of thyroid stimulation with volumetric changes in the gland strongly suggests that preg-
Table 2. Blunted TSH Levels During Normal Pregnancy
Blunted TSH
Increased Free
(siO.15 mU/L) First trimester Second trimester Third trimester
14%" 4% 2 ng/dL) (n)
(IU/L
3 0 0
54.5 28.2 10.0
compared to women with normal TSH
levels.
x
vs vs
vs
IO3) 33.6b 11.8 9.0
GLINOER AND LEMONE
68 Table 3. Serum TG Mean First trimester Second trimester Third trimester
Delivery "Reference limits: =s30
209 251 355 316
30 31 38 49
± ± ± ± ±
SE
in
Normal Pregnancy"
Median
2 2 2 2
% Above 30
µ^/L
35 38 49 59
23 25
30 37
Range 4-160 4-182 4-216 4-180
µg/L.
nancy truly induces goitrogenesis rather than vascular swelling alone (Fig. 3). There is, therefore, a continuum from physiology to pathology, and in areas with a marginally low iodine intake (and presumably in countries with more severe iodine deficien¬ cy), pathologic alterations of the thyroid tend to override physiologic changes related to pregnancy. Few other studies have been reported on TV changes during pregnancy, with only small groups of women investigated. Rasmussen et al. (27) reported a 30% increase in TV in 20 80 women in Denmark, with an estimated iodine intake of µg/day. In Italy, Romano et al. (28) reported a 16% increase in TV in 18 women. These authors also showed that the goitrogenic effect of pregnancy could be prevented by doubling the iodine excretion levels from 50 to 100 µg/day (matched-cases study with/without iodine supplementation). In countries with a higher iodine intake, data essentially are lacking, except for Finland, with an estimated iodine intakeofover200µg/day. Branderand Kivisaari (29) examined 21 women sequentially during gesta¬ tion and found a small, albeit significant, increase in TV between first and third trimesters. ~
patients with underlying (and most frequently unknown) thyroid abnormalities compared to healthy subjects. in
REVERSIBILITY DURING POSTPARTUM Hence, pregnancy does constitute a goitrogenic stimulus, at least where iodine supply is limited. We finally examined the question of the reversibility of" the alterations found in preg¬ nancy, during the postpartum period. Thyroid volume was reevaluated 12 months after parturition in a small number of selected women, who had a normal TV in the first trimester but in whom TVs had increased above average (by 55%) during gestation (30). In the majority of the cases, TV did not revert to the values found during early gestation, and the goiters that had formed only partially regressed after parturition. These data strongly suggest that the alterations in TV associated with pregnancy are not limited to the period of gestation. In the study of Rasmussen et al., slightly increased TV during gestation apparently normalized 12 months after parturition (27). How¬ ever, the iodine excretion in Copenhagen was 50% higher than that in Brussels.
ALTERATIONS IN TV IN PATIENTS WITH UNDERLYING THYROID ABNORMALITIES We also examined the effects of pregnancy in 44 euthyroid patients with a preexisting diffuse goiter and 20 with thyroid nodules without glandular enlargement (19). It was observed that goiter size increased significantly in one third of the goitrous women, associated with biochemical evidence of functional stimulation of the thyroid. Moreover, in patients with nodules (mainly microcysts), a striking finding was an increase in both their number and size, respectively in 20% and 60% of the patients. The data, therefore, suggest that pregnancy is associ¬ ated with an even greater thyroid risk from glandular stimulation
DIAGNOSTIC PROCEDURES AND THERAPEUTIC ATTITUDE
Thyroid enlargement occurring in pregnancy should be inves¬ tigated along the usual lines, including thyroid function tests, echography, and screening for positive thyroid autoantibodies (31). In our experience, the most sensitive marker of an excessive glandular stimulation is provided by an increase in the molar T,/T4 ratio above 0.025 (18). Thyroid function tests should be interpreted with care, owing to the changes associated with elevated TBG levels (32). Similarly, TSH levels should be
Table 4. Thyroid Volume"
Left lobe (mL) First trimester Second trimester Third trimester
Delivery
168 172 33 179
5.7 5.9 6.8 7.0
2.3 2.4 2.8 3.1
Right
lobe
(mL) 6.3 ± 2.6 6.9 ± 2.4 7.2 ± 2.5 8.0 ± 4.1
Total vol
(mL) 12.1 12.8 13.9 15.0
± ± ± ±
4.5 4.5 4.8 6.8
NS
Goiter and
Pregnancy:
A New
Insight into an Old Problem
DANIEL GLINOER and MARC LEMONE
ABSTRACT Evidence is presented that pregnancy constitutes a goitrogenic stimulus, particularly in conditions with a restricted or even a marginally low iodine intake. In a series of studies carried out in a large cohort of pregnancies in the Brussels area, the authors show that an increase in thyroid volume is observed in a majority of pregnant women, leading to goiter formation at delivery in 9% of the cases. Furthermore, increments in thyroid volume were correlated with biochemical evidences of functional stimulation of the thyroid, such as an elevation in serum TG levels, preferential T, secretion, and slight increases in basal TSH at delivery. Hence, the association of biochemical features of thyroidal stimulation with volumetric changes in the gland strongly suggests that pregnancy truly induces goitrogenesis rather than vascular swelling ("intumescence") alone, at least in conditions with a low iodine intake. Finally, preliminary data from this laboratory, as well as recently published data from other investigators, suggest that goiter formation during pregnancy can easily be prevented by increasing the iodine supply during pregnancy.
HISTORICAL BACKGROUND
(5) stressed the diversity of opinions on the topic and stated that "the main question is whether goiter is an acceptable normal physiological finding in pregnancy or whether it represents
A Egyptians
quotation mentions that ancient used to tie a reed around a women's neck to assess whether she was pregnant: if the reed snapped, it was the proof that she was indeed pregnant. Even though I have never been able to confirm this test in my own practice or actually find the original reference, the quotation suggests that pharaonic Egyptians had already noted that throat enlargements were somehow associated with gestation (1,2). More recently, two apparently contradictory studies have been reported by Crooks et al. in 1967 (3) and Levy et al. in 1980 (4). Crooks et al. compared the incidence of visible and palpable goiter in preg¬ nant women between Aberdeen (with a lower iodine intake) and Reykjavik (with a higher iodine intake). They showed that the incidence was three-fold higher in Scotland compared to Iceland and that it doubled during pregnancy in the lower and remained virtually unchanged in the higher iodine area. In contrast. Levy et al. investigated 98 women from northeastern Ohio with an adequate iodine intake (> 200 µg/day) and found no difference in goiter frequency between pregnant and nonpregnant matched patients. These apparently contradictory findings have produced a decade of debate and confusion about the possible goitrogenic role of pregnancy. In a recent review article. Becks and Burrow classic historical
Université Libre de Bruxelles,
thyroid pathology."
THYROID VOLUME AND IODINE DEFICIENCY
Although North American thyroidologists suggest that goiter is infrequent (6.7), European clinicians appreciate that goiter formation is rather common during gestation, in particular in populations with a clear deficit in daily iodine intake. In the former East Germany, an area where the average iodine intake was below 30 µg/day, a goiter has been reported in 60% of pregnancies (8). Until recently, however, precise data on alterations in thyroid volume (TV) during pregnancy were scanty. Moreover, since thyroid radionuclide scans are not performed during pregnancy, the older studies had to rely solely on the clinical estimates of goiter (e.g., visible or palpable, grade), which are insensitive for detecting subtle TV changes. Today, the routine use of thyroid ultrasonography has been added to our armamentarium, allow¬ ing for precise determinations of TV, with a high sensitivity for the detection of even minor thyroid alterations (9,10). Using
Hospital Saint-Pierre, Thyroid Investigation Clinic and Department of Radiology. Brussels, Belgium. 65
66
GLINOER AND LEMONE
ultrasound determinations, it is possible to compare absolute sizes in different areas and to relate those to the environment. Figure 1 compares several areas in Europe and shows clearly that TV in "average" populations varies as a function of the iodine intake (10-12). With intake above 100 µg/day, normal TV averages 10 mL, whereas with intake around 50 µg iodine/day, TV in the normal population is almost doubled.
glandular
Table 1. Thyroid Volume: Relationship and Body Weight
Females
Males
THYROID VOLUME IN PREGNANCY In order to discuss TV alterations associated with pregnancy, three features should be considered first.
Age (years)
TV
BW
(mean mL)
(kg)
23 39 59 75 23 39 59 76
15.6 17.7 17.9 20.5 16.8 21.0 21.3 20.4
58 61 68 65 69 76 77 70
Adapted from ref.
with
Sex, Age,
Vol/wt 0.27 0.29 0.27 0.32 0.25 0.28 0.28 0.29
0.27
0.29
11.
Thyroid volume and body weight
Thyroid intumescence
subjects, the size of the thyroid is correlated with body weight. TV is larger in males than in females (11). Moreover, when the ratio of TV/body weight is considered in healthy adult subjects, the sex and age variations in TV are almost abolished (Table 1). Similarly, there is a good relation¬ ship between age or weight and TV in adolescents (13-15). Hence, the closest relationship in both adolescent and adult subjects is found with body surface or lean body mass.
Pregnancy is a clinical situation that obviously involves body modifications, and some authors have proposed the use of algorithms to allow for corrections in the calculation of TV. However, these have not provided important advantages (11). Furthermore, gestation also is associated with an overall in¬ crease in blood flow (16), and, therefore, vascular swelling, or intumescence, has been implicated as a possible cause of thyroid enlargement during pregnancy (17).
In adult
mass
Individual variations in TV In a given area with similar iodine intake levels, there is a wide range of TV in normal subjects. In 166 healthy women in the first trimester of gestation, we observed glandular volumes ranging from a low of 5 mL to 22 mL. the upper limit of normality in our area (Fig. 2). The findings were somewhat surprising, since these women had no detectable thyroid abnor¬ malities and were strictly euthyroid. The data could be inter¬ preted to indicate that small glands are capable of ensuring First trimester of pregnancy
(N=166)
Mean: 11.7 mL Median: 10.8 mL
Range: A
Sweden (Stockholm) Netherlands
I 9
4.7-21.g mL
Denmark
(Amsterdam)
(Copenhagen)
Germany »various cities)
THYROID VOLUME (ML)
IODINE EXCRETION (ug
g
créât)
FIG. 2. Distribution frequency of thyroid volume during first trimester of pregnancy in healthy subjects (n 166). The upper limit of normality in the area is 22 mL. The mean TV was 1 1.7 mL (median 10.8 mL), with a range from 4.7 mL to 21.9mL. In comparison to other European cities, the results indicate that there is no significant iodine deficiency in Brussels. =
FIG. 1. Correlation between thyroid volume and urinary iodine concentration in normal populations of Stockholm, Am¬ sterdam, Copenhagen, and various cities in Germany. The graph was reconstructed from data in refs. 10-12.
GOITER AND PREGNANCY
67
normal thyroid function, at least as long as no additional factor intervenes to disturb the homeostasis of the system. We have reason to believe, however, that small glands may be less capable of adapting thyroid hormone production to chronic stimulatory events, such as pregnancy, without marked changes in TV. In pregnancy, systematic studies of TV have been lacking. In 1988, a large prospective cohort study was undertaken in our institution to evaluate the regulation of maternal thyroid function (18). Our aim was to assess both functional and volumetric alterations of the thyroid in both healthy pregnant women (with no detectable thyroid abnormalities) and patients with mild underlying thyroid abnormalities (19). These studies repre¬ sented a significant undertaking. The original cohort encom¬ passed a total of over 730 women, investigated cross-sectionally and followed sequentially during gestation and, for some of them, also after parturition. Interestingly, the study was per¬ formed in an urban area without significant iodine deficiency but where the daily iodine intake was nevertheless marginally low, between 50 and 80 µg/day (20). Our hypothesis was that a marginally low iodine supply, representative of the situation found in many western European cities (21-23), could lead to a further decrease in its availability for the maternal thyroid during pregnancy, due to the increased iodide renal clearance (24) and losses to the fetoplacental complex during late gestation (25). Starting in a situation with a limited iodine supply, the dimin¬ ished iodine availability could lead to a state of relative iodine deficiency, thereby enhancing the physiologic changes in thy¬ roid economy. In other words, our aim was to assess whether there was a continuum from physiology to pathology, driven in part by the environmental iodine supply.
THYROID REGULATION IN PREGNANCY In pregnancy, the normal thyroid is faced with a triple challenge as a result of three separate factors acting in concert to stimulate the gland: (a) the increased thyroid hormone-binding capacity of serum due to the marked increase in serum TBG
concentrations, (b) the effects of elevated hCG levels on the thyroid (and on TSH), and (c) the reduced availability of iodine
for the maternal thyroid in conditions with a limited dietary iodine supply (18). Our studies showed that variable patterns of thyroid adjustment to these combined factors take place in normal pregnancy. Approximately one third of the subjects were characterized by having relative hypothyroxinemia, preferential T3 secretion, and a higher setting of the pituitary thyrostat, i.e., a pattern of increased glandular stimulation. In addition, an
intrinsic TSH-like activity of hCG was observed, and the data suggested strongly that hCG acts directly to stimulate the thyroid. During the first trimester of gestation, blunted serum TSH levels were found in 14% of healthy subjects (Table 2). Gestational transient thyrotoxicosis was, however, exceedingly rare, since free T4 remained in the normal range in most instances. In a few women (1.4%), high hCG levels (usually above 70,000-80,000 IU/L) were associated with gestational thyrotoxicosis (often accompanied by hyperemesis gravidarum), with free T4 levels unequivocally in the thyrotoxic range (2.5, 2.7, 3.7 ng/dL). Finally, in relation to the metabo¬ lism of iodide, the data suggested that the relative iodine deficiency in Brussels had an overall permissive role in allowing for enhancement of the stimulatory changes in the thyroidal economy.
MATERNAL REPERCUSSIONS ON THE THYROID
Repercussions on the maternal thyroid included significant increases in serum thyroglobulin (TG) levels, especially during the last trimester (Table 3). More than half of the women had elevated serum TG levels at term, with individual values up to 180 µg/L. The other major finding was that TV increased in the vast majority of women. The average increment was.20%, but it varied widely among individuals up to 130% (Table 4). True goiter, defined as TV greater than 22 mL, was found in 9% of the cohort at delivery, with volumes up to 46 mL. Furthermore, changes in size were negatively correlated to initial size, so that for TV initially under 10 mL, the average increment exceeded 30%. An increase in TV was associated with biochemical features of thyroid stimulation, such as an increased ,/T4 molar ratio or slightly increased—albeit still normal—TSH level at delivery. The most significant correlation was between TG and TV. The greater the changes in TV, the more frequently TG was elevated. TG elevation, therefore, reflects anatomic changes and constitutes a marker of an increase in TV during pregnancy (18,26). Another interesting finding was that larger volumes at term were associated with higher hCG levels during the first trimester, indicating a potential role of hCG to stimulate not only thyroid function but also glandular enlargement. By measuring TV with echography and in the absence of histologie evidence, it was impossible to distinguish glandular hyperplasia from hypertrophy or appreciate the role of changes in blood flow occurring during pregnancy. However, the asso¬ ciation of biochemical features of thyroid stimulation with volumetric changes in the gland strongly suggests that preg-
Table 2. Blunted TSH Levels During Normal Pregnancy
Blunted TSH
Increased Free
(siO.15 mU/L) First trimester Second trimester Third trimester
14%" 4% 2 ng/dL) (n)
(IU/L
3 0 0
54.5 28.2 10.0
compared to women with normal TSH
levels.
x
vs vs
vs
IO3) 33.6b 11.8 9.0
GLINOER AND LEMONE
68 Table 3. Serum TG Mean First trimester Second trimester Third trimester
Delivery "Reference limits: =s30
209 251 355 316
30 31 38 49
± ± ± ± ±
SE
in
Normal Pregnancy"
Median
2 2 2 2
% Above 30
µ^/L
35 38 49 59
23 25
30 37
Range 4-160 4-182 4-216 4-180
µg/L.
nancy truly induces goitrogenesis rather than vascular swelling alone (Fig. 3). There is, therefore, a continuum from physiology to pathology, and in areas with a marginally low iodine intake (and presumably in countries with more severe iodine deficien¬ cy), pathologic alterations of the thyroid tend to override physiologic changes related to pregnancy. Few other studies have been reported on TV changes during pregnancy, with only small groups of women investigated. Rasmussen et al. (27) reported a 30% increase in TV in 20 80 women in Denmark, with an estimated iodine intake of µg/day. In Italy, Romano et al. (28) reported a 16% increase in TV in 18 women. These authors also showed that the goitrogenic effect of pregnancy could be prevented by doubling the iodine excretion levels from 50 to 100 µg/day (matched-cases study with/without iodine supplementation). In countries with a higher iodine intake, data essentially are lacking, except for Finland, with an estimated iodine intakeofover200µg/day. Branderand Kivisaari (29) examined 21 women sequentially during gesta¬ tion and found a small, albeit significant, increase in TV between first and third trimesters. ~
patients with underlying (and most frequently unknown) thyroid abnormalities compared to healthy subjects. in
REVERSIBILITY DURING POSTPARTUM Hence, pregnancy does constitute a goitrogenic stimulus, at least where iodine supply is limited. We finally examined the question of the reversibility of" the alterations found in preg¬ nancy, during the postpartum period. Thyroid volume was reevaluated 12 months after parturition in a small number of selected women, who had a normal TV in the first trimester but in whom TVs had increased above average (by 55%) during gestation (30). In the majority of the cases, TV did not revert to the values found during early gestation, and the goiters that had formed only partially regressed after parturition. These data strongly suggest that the alterations in TV associated with pregnancy are not limited to the period of gestation. In the study of Rasmussen et al., slightly increased TV during gestation apparently normalized 12 months after parturition (27). How¬ ever, the iodine excretion in Copenhagen was 50% higher than that in Brussels.
ALTERATIONS IN TV IN PATIENTS WITH UNDERLYING THYROID ABNORMALITIES We also examined the effects of pregnancy in 44 euthyroid patients with a preexisting diffuse goiter and 20 with thyroid nodules without glandular enlargement (19). It was observed that goiter size increased significantly in one third of the goitrous women, associated with biochemical evidence of functional stimulation of the thyroid. Moreover, in patients with nodules (mainly microcysts), a striking finding was an increase in both their number and size, respectively in 20% and 60% of the patients. The data, therefore, suggest that pregnancy is associ¬ ated with an even greater thyroid risk from glandular stimulation
DIAGNOSTIC PROCEDURES AND THERAPEUTIC ATTITUDE
Thyroid enlargement occurring in pregnancy should be inves¬ tigated along the usual lines, including thyroid function tests, echography, and screening for positive thyroid autoantibodies (31). In our experience, the most sensitive marker of an excessive glandular stimulation is provided by an increase in the molar T,/T4 ratio above 0.025 (18). Thyroid function tests should be interpreted with care, owing to the changes associated with elevated TBG levels (32). Similarly, TSH levels should be
Table 4. Thyroid Volume"
Left lobe (mL) First trimester Second trimester Third trimester
Delivery
168 172 33 179
5.7 5.9 6.8 7.0
2.3 2.4 2.8 3.1
Right
lobe
(mL) 6.3 ± 2.6 6.9 ± 2.4 7.2 ± 2.5 8.0 ± 4.1
Total vol
(mL) 12.1 12.8 13.9 15.0
± ± ± ±
4.5 4.5 4.8 6.8
NS
