2024 Volume 71 Issue 6 Pages 617-621
To establish an easy way to perform volumetry of the thyroid gland using ultrasonography, we evaluated the accuracy of the products of the depth and width of the right thyroid lobe as indices of thyroid volume. The depth and width of both thyroid lobes were measured using ultrasonography before surgery in 193 patients with Graves’ disease. The products were compared with the weight of the thyroid obtained from operative records. We also evaluated the depth and width of the right thyroid lobe in 312 subjects who presented without any thyroid disease. The products of depth and width of the right and left lobes of patients with Graves’ disease correlated similarly well with the weight of the thyroid obtained from operative records (ρ = 0.896 for right, ρ = 0.886 for left, p < 0.0001). Because the right lobes were larger than the left lobes, the products of the depth and width of the right lobe were adopted as novel parameters for an easy volumetric approach. The relationship between the weight and the measurements of the right lobe was described using the following regression equation: weight (g) = [11.8 × depth (cm) × width (cm)] – 16.0. The products of the subjects without any thyroid diseases were distributed between 0.6 cm2 and 4.4 cm2, with a median of 2.0 cm2. The upper limit of these values in these subjects was estimated to be 3.8 cm2. This easy ultrasonographic volumetric technique makes it possible to perform a semi-quantitative assessment of thyroid volume and to differentiate diffuse goiter from normal-sized thyroids.
DIFFUSE ENLARGEMENT of the thyroid gland is a common sign observed in patients with autoimmune thyroid diseases and supports their diagnosis. Several methods are known for expressing the size of enlarged thyroid glands. Classification of the degree of enlargement based on inspection and palpation is a conventional and simple method. However, this method is relatively inaccurate.
When ultrasonography is used, the horizontal and vertical dimensions of the thyroid lobes can represent the degree of enlargement. Estimation of thyroid volume by ellipsoidal approximation [1-4] is regarded as a reliable method, but it is cumbersome. In clinical settings, common methods to describe the size of swollen thyroid glands are not widely shared among doctors.
This retrospective study established an easy method for evaluating and describing the size of the thyroid gland in patients with diffuse goiter and in those without thyroid diseases. Our volumetric approach for the thyroid gland is easy and convenient to perform. Furthermore, it permits a semi-quantitative evaluation of thyroid volume.
A total of 193 patients (157 women and 36 men) with Graves’ disease who were 13–75 (median, 34 years) were included in this study. Diagnosis of Graves’ disease was based on the guidelines of the Japan Thyroid Association [5]. All patients with Graves’ disease included in this study underwent thyroidectomy following treatment with antithyroid drugs.
Further 312 subjects (238 women and 74 men) who were 18–86 (median, 38 years) without thyroid disease were included as controls. These subjects visited our hospital with several complaints; their serum levels of free T3, free T4, and TSH were within the reference ranges, and anti-thyroglobulin antibodies and anti-thyroid peroxidase antibodies were negative. No abnormal findings were confirmed by ultrasonography.
This study was approved by the Ethical Committee of our hospital (approval number 041) and conducted in accordance with the Declaration of Helsinki. The patients were given the right to opt out of the study.
Measurement of the size of the thyroid gland by ultrasonographyBefore surgery, patients were examined in the sitting position using ultrasonography. The width and depth of each thyroid lobe and isthmus were measured in the transverse plane with the maximum cross-sections. The lengths were also measured in the longitudinal planes of the two lobes and isthmus (Fig. 1). These parameters were rounded to one decimal place (cm). The products of the width and depth of the right lobe, those of the left lobe, and the sum of both lobes were obtained. The volumes of each thyroid lobe and isthmus were estimated using ellipsoidal approximation [1-4]. The total thyroid volume was defined as the sum of the volumes of each thyroid lobe and isthmus (conventional volumetry). We also determined the distribution of the products of the width and depth of the right lobe in 312 control subjects.
Transverse scan of the thyroid (a) and longitudinal scan of the right lobe of the thyroid (b). The depth (D), width (W), and length (L) of each lobe and isthmus are required to estimate the thyroid volume using ellipsoidal approximation. For the method employed in this study, only the depth (D) and width (W) of the right lobe were necessary.
To evaluate inter-observer differences in the measurements, eight examiners measured the thyroid gland of the same subject, and the coefficient of variation was determined.
The ultrasound equipment used in this study was Aplio 500 (Canon Medical Systems Corporation, Ohtawara, Japan) and Philips iU 22 (Philips Japan, Ltd., Tokyo, Japan).
Statistical analysesThe correlation between parameters obtained by ultrasonography and thyroid weight obtained from operative records was evaluated in 193 patients with Graves’ disease. Correlations were analyzed using the Spearman’s rank test. The difference in the mean products of the depth and width of both lobes was analyzed using the Wilcoxon signed-rank test. The mean products of the depth and width for men and women were analyzed using the Wilcoxon rank-sum test. The value at 2.5% from the maximum measurements was regarded as the upper limit of the products of the depth and width of the control subjects, since these data were not normally distributed. Finally, the cutoff value to discriminate Graves’ disease from the control was determined using an ROC analysis.
Statistical analyses were performed using SAS JMP software program, ver. 11.0.0 (SAS Institute Inc., Cary, NC, USA).
Assays for serum hormones and anti-thyroid autoantibodiesFree T3, free T4, TSH, and anti-TSH receptor antibodies were measured using Roche Elecsys FT3, FT4, TSH, and TRAb (Roche Diagnostics, Basel, Switzerland), respectively. Anti-thyroglobulin antibodies and anti-thyroid peroxidase antibodies were measured using CL AIA-PACK TgAb and TPOAb (Tosoh Corporation, Tokyo, Japan), respectively.
The estimated thyroid volumes by ellipsoidal approximation (conventional volumetry) correlated well with the weights obtained from the operative records (ρ = 0.970, p < 0.0001) (Fig. 2). The products of the depth and width of the right lobe, those of the left lobe, and the sum of those of both lobes were also well correlated with the weights obtained from the operative records, as shown in Table 1 and Fig. 3. The products of depth and width of the right and left lobes ranged from 2.3 to 18.9 (median 6.2) cm2 and 1.1 to 13.7 (median 5.2) cm2, respectively, and the former was significantly (p < 0.0001) larger than the latter. The relationship between the weights obtained from operative records and the products of depth and width of the right lobe was described by the following regression equation: weight (g) = [11.8 × depth (cm) × width (cm)] – 16.0.
Correlation of the thyroid volume estimated by ellipsoidal approximation with the weight obtained from operative records in 193 patients with Graves’ disease.
Correlation coefficients between the thyroid weight obtained from operative records and the products of the depth and width of the right and left lobes, and the sum of those of both lobes in 193 patients with Graves’ disease.
| Depth × Width (cm2) | Correlation coefficients | p value |
|---|---|---|
| Right lobes | 0.896 | p < 0.0001 |
| Left lobes | 0.886 | p < 0.0001 |
| Sum of both lobes | 0.918 | p < 0.0001 |
Correlation of the product of the depth and width of the right lobe with the weight obtained from operative records in 193 patients with Graves’ disease.
Patients were divided according to the product of the depth and width of the right lobe, as shown in Table 2. A product value of 3.8 cm2 was the upper limit for subjects without any thyroid diseases, as described below. A product value of over 10.0 cm2 roughly indicated that the thyroid gland weighed over 100 g.
The median and IQR of the thyroid weight obtained from operative records of five groups divided by the product of depth and width in 193 patients with Graves’ disease.
| Depth × Width (cm2) | n | Median and IQR in parentheses of weight (g) |
|---|---|---|
| ≤3.8 | 30 | 24.2 (21.9–27.9) |
| >3.8–6.0 | 64 | 40.5 (33.6–51.1) |
| >6.0–8.0 | 48 | 61.5 (51.1–74.8) |
| >8.0–10.0 | 34 | 79.6 (68.8–92.0) |
| >10.0 | 17 | 131.0 (108.2–169.7) |
IQR, interquartile range
In the 312 subjects without any thyroid diseases, the products of depth and width were distributed between 0.6 and 4.4 cm2, with a median value of 2.0 cm2. The upper limit was estimated to be 3.8 cm2 for all the control subjects (Fig. 4). The results for women ranged from 0.6–4.2 (median 1.8) cm2, which were significantly lower (p < 0.0001) than those for men which ranged from 1.4–4.4 (median 2.5) cm2. The upper limit for women was estimated to be 3.2 cm2 (Fig. 5).
The product of the depth and width of 312 control subjects. The median and range were 2.0 cm2 and 0.6–4.4 cm2, respectively.
The products of the depth and width of women were significantly lower than those of men.
According to our analysis, the cutoff value that discriminated Graves’ disease from control subjects was determined to be 3.1 cm2, where the AUC was 0.986. The sensitivity and specificity were 94.8% and 93.9%, respectively.
Inter-observer variationThe mean and standard deviation of the volume obtained by 8 independent examiners using conventional volumetry were 17.80 and 1.65 g, respectively. The coefficient of variation was 9.3%. In contrast, the mean and standard deviation of the products of the depth and width were 2.85 and 0.20, respectively, yielding a coefficient of variation of 7.0%.
Thyroid gland volume in patients with autoimmune thyroid disease provides important clinical information. In patients with Graves’ disease, gland size is a predictive marker of remission after anti-thyroid drug treatment [6]. The amount of radioactive iodine to be administered is dependent on gland size in radioactive iodine therapy for Graves’ disease [7, 8]. In Japan, the Shichijo classification [9] has been widely used to describe the size of the diffusely enlarged thyroid glands. However, substantial errors are inevitable owing to the subjective nature of the evaluation, which involves inspection and palpation.
Ultrasound examinations are widely used to treat thyroid diseases. Ultrasonography is an informative modality for identifying and characterizing thyroid nodules. In addition, in autoimmune thyroid diseases, ultrasonographic findings provide key information such as speculated histological changes, vascularity, and gland size. An ellipsoidal approximation is regarded as an accurate method for estimating the thyroid volume using ultrasonography. However, this method is cumbersome and time consuming. Therefore, it is not widely accepted by doctors. Therefore, an easier method for semiquantitative volumetry is required.
In the present study, we established an easy method for performing volumetry using ultrasound. The products of the depth and width of the right and left lobes were similarly correlated with the weights obtained from the operative records. The correlations between the products of the single lobes and the weights of the thyroid were found to be inferior to those of conventional volumetry or the sum of the products of both lobes. However, the accuracy of this method is sufficient to meet the clinical requirements. The availability of this method is advantageous, as it overwhelms its slight inaccuracy.
The right thyroid lobes were larger than the left lobes in this study, consistent with the results of previous reports [4, 10]. To avoid underestimation, the product of the depth and width of the right lobe was adopted in this study as a novel parameter for an easy volumetric approach. Measurement of only two dimensions on the ultrasonogram of the right lobe is required to determine the volume of the thyroid; therefore, general practitioners and endocrinologists can evaluate the thyroid volume in their consulting rooms using handheld ultrasound equipment, rendering high-end ultrasonic equipment no longer necessary in this setting.
This index can be used to discriminate between normal thyroid and diffuse goiter during medical checkups. Thus, subjects who may have autoimmune thyroid disease can be efficiently identified using this method. The cutoff value for separating Graves’ disease from the control was determined as 3.1 cm2, which was lower than the upper limit of the products of the depth and width of the control subjects. This is because the thyroid glands of some patients with Graves’ disease do not enlarge. It should be noted that this method may produce false negative results when screening patients with autoimmune thyroid disease.
Several limitations of the present study warrant mention. First, this study did not include patients with Hashimoto thyroiditis. Second, the number of control subjects without thyroid disease was rather small and comprised patients who had visited the hospital with several complaints, meaning that they could not be considered healthy controls.
In conclusion, the product of the depth and width of the right thyroid lobe on ultrasonography is an easy and efficient index for determining thyroid volume. This indicator correlated well with the weight obtained from operative records. This method is suitable for clinical practice in the era of point-of-care ultrasound [11]. The common notation of thyroid volume should be widely shared among general practitioners and endocrinologists.
None of the authors have any potential conflicts of interest associated with this research. Tsukasa Murakami is a member of the Editorial Board of Endocrine Journal.
