DETERMINATION OF MANDIBULAR BONE CHANGES IN PATIENTS WITH PRIMARY HYPOTHYROIDISM TREATED WITH LEVOTHYROXINE SODIUM

Abstract

Background

This study aimed to assess fractal dimension (FD) and the radiomorphometric indexes on the digital panoramic radiography (DPR) of patient with primary hypothyroidism receiving levothyroxine sodium replacement therapy.

Methods

A total of 115 subjects were included in this cross sectional retrospective study. According to the results of the thyroid function tests, the subjects were divided into two groups as primary hypothyroidism (levothyroxine sodium replacement therapy given), (n = 57) and the healthy control group (n = 58). The fractal dimension (FD), panoramic mandibular index (PMI), mandibular cortical width (MCW), gonial index (GI) and mandibular cortical index (MCI) values of all patients were calculated on DPRs. The statistical analysis of all data was performed with SPSS version 22.

Results

The distributions of age and gender in the primary hypothroidism group were similar to control group (p = 0.19 and p = 0.62, respectively). The two groups did not differ statistically significantly in terms of FD, PMI, MCW, GI, and MCI.

Conclusion

We determined that mandibular cortical and trabecular bone structure did not significantly differ between healthy individuals and patients receiving drug replacement theraphy due to hypothyroidism, but our results should be further supported with the investigation of clinical parameters.

INTRODUCTION

Thyroid hormones play a key role in skeletal development and are important regulators of bone health and homeostasis maintenance in adults (1). Primary hypothyroidism (PH) is an endocrinopathy that is common in the general population and can be seen in up to 10% of adults, mostly in women and different patient groups (2-4). Untreated overt hypothyroidism causes deterioration of endochondral ossification in children, epiphyseal dysgenesis, delayed skeletal development, short stature, enamel hypoplasia, delayed tooth eruption, anterior open bite, and micrognathia, while reducing both osteoblastic and osteoclastic activity in adults (5-7).

Although there are studies showing skeletal damage caused by untreated hypothyroidism (8, 9), only few clinical studies have been undertaken to investigate the effects of hypothyroidism on the skeletal system in humans (2,10).

Osteoporosis is a metabolic bone disease in which bone microarchitecture is impaired, its mass is reduced, and consequently the risk of fracture associated with them increases. Fractures are an important cause of morbidity and mortality, especially in the adult population (11). Despite adequate hormone replacement therapy, osteoporosis and osteopenia have been reported in 20-50% of patients with PH (2,10). The recommended and frequently used method for bone mineral density (BMD) measurement is dual-energy X-ray absorptiometry (DEXA), in which osteoporosis is diagnosed according to the scales determined using the lumbar vertebra for trabecular bone and the femoral neck for cortical bone (12). However, the most important disadvantages of the DEXA method are the inability to distinguish between cortical and trabecular bones, expensive nature, and increased bone mineral density value in osteoarthritis and aortic calcification and false positive results (13). Dental panoramic radiographs (DPRs) are generally used in dental clinical practices. DPR can be used for determining and evaluating high-risk osteoporosis cases for bone densitometry. DPR allows for comprehensive anatomical evaluation of the jaws in the general population, making it a valuable diagnostic tool in identifying osteoporosis from the alveolar bones (14). On DPR, fractal dimension (FD) calculated using image processing and analysis methods and various mandibular indices can be used to evaluate the microarchitecture of the mandibular bone mass and trabecular bone, and offer an idea about osteopenia and osteoporosis diseases can be obtained (15-17).

It is known that bone metabolism is affected in patients with PH, but there are only few studies showing the effect of levothyroxine sodium (L-thyroxine sodium) treatment on bone metabolism according to literature reviews (18,19). In addition to our knowledge no study has been conducted to evaluate both FD and mandibular indexes together using DPR. Therefore the current study aimed to determine FD, panoramic mandibular index (PMI), mandibular cortical width (MCW), gonial index (GI), and mandibular cortical index (MCI) using DPR in patients with PH, who received 25-100 microg LT4 daily for at least one year and compare these values with the healthy control group.

MATERIALS AND METHODS

Study Design

The Clinical Research Ethics Committee of X University Medical Faculty confirmed this study (70904504/108, 2021). The study was conducted in accordance with the ethical rules the Declaration of Helsinki. In the study, subjects who presented to the Oral and Maxillofacial Radiology Department of X University X Faculty between January 2017 and December 2018 for various reasons and underwent DPR for any reason were retrospectively evaluated. The study consisted of two groups: the PH group undergoing LT4 replacement therapy and the healthy control group. Subjects with systemic diseases, such as chronic kidney disease, diabetes mellitus, and hyperthyroidism, those with a history of bone diseases such as osteoporosis, women in the postmenopausal period, Paget’s disease, patients with a history of radiotherapy-chemotherapy, those with long-term corticosteroid, bisphosphonate or denosumab use, those using antiresorptive drugs, those with cysts, tumors, and lesions and those with DPRs without sufficient diagnostic quality were excepted from the study. The study was carried out with a total of 115 subjects, consisting of 57 patients in the PH group and 58 in the healthy control group.

The cause of hypothyroidism in all patients was chronic autoimmune thyroid disease. Patients with a diagnosis of hypothyroidism for at least 3 years, thyroid-stimulating hormone, T3 and T4 values within the normal range with LT4 replacement treatment for the past one year, no change in the dose of additional medication, and patients in euthyroid state were included. All patients were using the daily dose of 25-100 microg in the treatment of LT4 replacement.

Obtaining Radiographs

All DPRs were obtained by the same operator using the same panoramic radiography device, (Planmeca Promax Oy, 00880 Helsinki, Finland), with the parameters of 66 kVp, 7 mA, and 16 second exposure. The patient's Frankfurt plane was positioned parallel to the ground, and the sagittal plane parallel to the vertical plane of the device. All DPRs were recorded in the JPEG format.

All the images were analyzed using ImageJ version 1.3 (National Institutes of Health, Bethesa, MD, USA). ImageJ was used to measured linear distance and calculate to fractal dimension by using box counting method recommended by White and Rudolph (20). Regions of interest (ROI) were selected from four different 60x60 pixel regions in the right mandible on DPRs. The subcortical area of the condyle (ROI 1), the supracortical area in the angulus (ROI 2), the mesial area of the first molar (ROI 3), and the apical area of the incisors (ROI 4) (Fig. 1).

Figure 1.

Fractal dimension measurement from four different regions of interest on DPR.

After ROI were selected, they were duplicated, and the Gaussian blur filter was applied. To eliminate differences in brightness caused by soft tissue and bone density. The resulting images were extracted from the original images, and 128 shades of gray were added to each pixel. Bone trabeculae were distinguished to distinctively visualize the spaces between trabeculae. After the images were converted to black and white with the threshold option, the noise was reduced with the ‘erode’ and ‘dilate’ options. After applying the ‘invert’, ‘skeletonize’ and ‘analyze’ functions, FD was calculated (Fig. 2).

Figure 2.

The processing parts of fractal analysis.

MCW: It is the height of the cortical bone on the perpendicular line drawn from the lower border of the mental foramen to the tangent line drawn from the lower border of the mandibular basis (21) (Fig. 3).

Figure 3.

Sample measurements of PMI, MCW, GI, and MCI on DPR.

PMI: The method used by Benson et al. (22) was used to calculate this index. PMI was calculated separately as PMI-inferior (PMI-i) and PMI-superior (PMI-s). PMI-i is the ratio of mandibular cortical width (MCW) to the length of the perpendicular line drawn from the inferior border of the mental foramen to the imaginary line tangent to the mandibular base and PMI-s, is the ratio of mandibular cortical width to the length of the perpendicular line drawn from the superior border of the mental foramen to the imaginary line tangent to the mandibular base (22) (Fig. 3).

GI: It is the cortical thickness corresponding to the line defining the bisector of the gonial angle (23) (Fig. 3).

MCI: It is a method based on visual morphological evaluation of the mandibular cortex distal to the mental foramen on DPR. The types of MCI are as: C1, the endosteal margin of the mandibular cortex is continuous and sharp; C2, the endosteal margin of the mandibular cortex appears as semi-lunar defect or endosteal cortical residues; C3, the cortical layer is highly endosteal cortical residues and clearly porous (24) (Fig. 3).

All measurements were made by a single dentomaxillary radiologist. The FD, PMI, MCW, MCI, and GI measurements were undertaken at the same time. In order to test the intraobserver reliability in ROI selection, index measurements and fractal analysis, 20% of the images were re-evaluated by the same radiologist one month later.

Statistical analyses

Statistical analysis was performed using SPSS v. 22 for Windows (IBM Corp., Armonk, NY, USA). The normality assumption was evaluated using the Shapiro- Wilk method. For the analysis of inter-group differences, the independent samples t-test was applied for data with a normal distribution. The differences between the PH and control groups in terms of age, gender, FD, MCW, and GI were evaluated with the Mann Whitney U test, and the differences in the MCI and PMI superior measurements were evaluated with Student’s t- test. Basic descriptive statistical analysis and normality tests of all variables were performed. The statistical significance level was taken as p < 0.05.

The Intraclass Classification Correlations (ICC) indicated good reliability for FD (ICC value was 0.98) and MCW, GI, PMI-i, PMI-s (ICC values were 0.985, 0.984, 0.981, 0.975, respectively. The kappa coefficient for MCI was 0.971.

RESULTS

Of the 57 individuals in the PH group, 36 were female, and 21 were male, and of the 58 subjects in the control group, 34 were female and 24 were male. There was no statistical difference between the two groups in terms of gender (p = 0.61) (Table 1).

Table 1.

Distribution of the patient and healthy group in terms of age and gender

		Control Group		PH Group
		N	%	N	%	p
Gender	Female	36	63.1 %	34	58.6 %	0.61*
	Male	21	36.9 %	24	41.4 %
Total		57	100	58	100
Age	mean ±SD	36.81 ± 10.29		34.38 ± 10.53		0.35*

^*: Mann Whitney U Test, SD: Standart Deviation, N: Noun.

The mean age was 36.81 ± 10.29 years for the PH group and 34.38 ± 10.53 years for the control group. There was no significant difference in age between the two groups (p = 0.35) (Table 1). Although the mean MCW of the PH group (3.91 ± 0.84 mm) was lower than the control group (4.09 ± 0.72 mm), there was no statistically significant difference (p = 0.24). The mean PMI-i and PMI-s values of the PH group (0.37 ± 0.07 and 0.29 ± 0.05, respectively) were similar to those of the control group (0.37 ± 0.08 and 0.29 ± 0.06, respectively) (p = 0.22 for both). Although the mean GI value of the PH group (1.19 ± 0.28) was lower than that of the control group (1.35 ± 0.31), no significant difference was found (p = 0.28) (Table 2).

Table 2.

Descriptive statistics between groups and comparison of results

	Control Group				PH Group				p
	Mean	SD	Min	Max	Mean	SD	Min	Max
MCW	3.91	0.84	2.10	6.01	4.09	0.72	2.90	6.25	0.24*
PMI Inf	0.37	0.07	0.19	0.52	0.37	0.08	0.25	0.53	0.22*
PMI Sup	0.29	0.05	0.16	0.41	0.29	0.06	0.10	0.41	0.16**
GI	1.19	0.28	0.8	1.70	1.35	0.31	0.93	2.01	0.28*
ROI 1 FD	1.498	0.104	1.355	1.692	1.478	0.101	1.208	1.614	0.88*
ROI2 FD	1.475	0.145	1.058	1.700	1.467	0.158	0.976	1.676	0.29*
ROI3 FD	1.538	0.073	1.396	1.669	1.525	0.107	1.072	1.668	0.28*
ROI4 FD	1.475	0.094	1.240	1.648	1.493	0.103	1.150	1.640	0.60*
Mean FD	1.496	0.062	1.356	1.630	1.491	0.068	1.286	1.609	0.63*

^*: Mann whitney U test, **Student t test, PMI: Panoramic mandibular index, MCW: mandibular cortical width, GI: Gonial index, ROI 1FD: Fractal dimension of the subcortical area of the condyle, ROI2 FD: Fractal dimension of the supracortical part in the angulus, ROI3 FD: Fractal dimension of the mesial area of the first molar, ROI4 FD: Fractal dimension of the apical area of the incisors.

In the PH group, MCI was classified as C1 in 24 patients, C2 in 30, and C3 in three, while in the control group, the C1 type was present in 22 subjects, C2 in 32 and C3 in four. No statistically significant difference was observed in the MCW, PMI -i, PMI-s, GI, and MCI values of the two groups (p = 0.86) (Table 3).

Table 3.

Distribution of MCI by groups

		PH Group		Control Group		p
		N	%	N	%
MCI	C1	24	42.1	22	37.9
	C2	30	52.6	32	55.1	0.86**
	C3	3	5.26	4	6.89
Total		57		58

^**Student t test, MCI: Mandibular cortical index, N: Noun.

When the PH and control groups were compared in relation to FD, there was no statistically significant difference in the measurements performed from between four different ROI (ROI 1 to 4) and mean value (p = 0.88, p = 0.29, p = 0.28, p = 0.60, and p = 0.62, respectively). The mean FD values calculated for ROI 1, ROI 2 and ROI 3 were found to be higher in the control group than the PH group (Table 2).

DISCUSSION

To the best of our knowledge, this is the first study in the literature to calculate both FD and radiomorphometric indices together on DPR in patients with PH receiving LT4 replacement therapy. Since thyroid hormone receptors are existing in human bone, it has been suggested that thyroid hormones may be effective straight on bone cells through specific nuclear receptors or indirectly by rising the secretion of growth hormone (25,26).

PH is defined as an increased thyroid stimulating hormone (TSH) level with T4 and T3 values below laboratory reference values, and histological evaluations in hypothyroidism indicate decreased osteoblastic and decreased osteoclastic activity with low bone turnover (27). Although desired TSH levels can not always be achieved with LT4 replacement therapy, it has also been reported that osteoporosis may occur with high doses of overtreatment (2). In a study on rats, Gecgelen Cesur et al. reported that prenatal and postnatal untreated hypothyroidism caused osteoporotic changes in the mandibular and maxillary bones (8). Ribeiro et al. also found that hormone replacement did not potentiate the formation of osteopenia in the bone for up to 120 days in female rats (9). Similarly, Talejipour et al. reported that thyroxine decreased bone density in the mandible more than other anatomical points such as skull, hard palate, and alveolar bone in rats, and therefore, mandible radiography could be a predictor for the reduction of bone density and osteoporosis (28). Gonzalez Rodriguez et al. showed that LT4 had no negative effect on bone health. Similarly we observed no difference between the PH and the control groups in accordance with the markers evaluating the bone quality of LT4 replacement therapy (29).

In the literature, there are studies in which FD was used in the evaluation of trabecular bone structure in patients with different systemic diseases, such as, hyperparathyroidism [30], osteoporosis (17), Type-1 and Type-2 diabetes mellitus (31), chronic renal failure (32), osteogenesis imperfecta (15), celiac disease (33), Medication-related osteonecrosis of the jaws (MRONJ) (34), thalassemia (35), sickle cell anemia (36), and hemoglobinopathy (37), as well as in those using drugs that affect bone metabolism, such as aromatase inhibitors (38), and bisphosphonate (39).

Gümüşsoy et al. calculated FD on the panoramic radiographs of patients with chronic kidney disease and healthy individuals and found that FD was lower the former which they attributed to hyperparathyroidism and vitamin D deficiency in patient group with chronic kidney disease (32). In another study, Ergün et al. demonstrated an increase in postoperative bone density in a patient with hyperparathyroidism with adrenal adenoma, comparing the bone mineral level before and after parathyroidectomy with FD on DPR. The authors associated the increase in FD after parathyroidectomy with the reduction of bone resorption as a result of the treatment of hyperparathyroidism (32).

Kurşun-Çakmak EŞ et al., in their study, calculated FD in four different areas on panoramic radiography in patients with Type-1 and Type-2 diabetes mellitus and presented no difference between the diabetes and control groups in terms of FD values (31). Similarly, Neves et al. did not find any statistical difference in FD between 40 women with and without celiac disease (33).In our study, there was no important difference between the PH and control groups in terms of the FD values we measured from four different areas.

In order to evaluate bone mineral density, mandibular radiomorphometric indices, including MCW, PMI, GI, and MCI, have been used in different conditions and diseases, such as osteogenesis imperfecta (15), osteoporosis-osteopenia (40), cemento-osseous dysplasia (41), edentulism (42), and postmenopause (21). The result of these studies indicate that these indexes are reliable in evaluating bone mineral density, and morphometric indices evaluated by DPR in the adult population are correlated with osteoporotic changes (14, 22, 42).

Khaitan et al., evaluating the reliability of PMI in determining BMD and sexual dimorphism, determined that PMI was higher in men than in women, and although it could not be used for gender discrimination, it was reliable in evaluating BMD (43). Kurşun-Çakmak et al. did not observe any statistical difference in MCI, MCW and PMI in a total of 104 individuals, of whom 26 had were Type 1 diabetes, 26 had Type 2 diabetes, and 52 were in the control group (31).

In other study, Rahangdale et al. measured MCW and PMI indices in a primary hypothyroid group of 52 patients receiving thyroxine replacement therapy and a control group consisting of 50 healthy individuals, and reported that both indices were lower in the primary hypothyroid group but there was no significant difference compared to the control group (19). Similarly, we detected no statistical difference between the two groups in relation to PMI and MCW, but the MCW value was lower in the PH group.

This study has several limitations. Firstly, it had a single-center, retrospective and cross-sectional design. Cross-sectional investigations lack the power to show true negative results and biases can appear if the case and control groups are not properly matched. Secondly, the study included a small number of patients, therefore decreasing the chance of finding significant differences and increasing the risk of missing real differences.

In conclusion, in this study, there was no difference in radiological markers evaluating bone quality between the patients with PH who underwent LT4 replacement theraphy and healthy individuals. We consider that in the future studies the evaluation of pre- and post-treatment radiographs in patients with PH who have received the same dose and duration of LT4 replacement therapy in comparison with a control group of similar age and body mass index with no additional systemic disease can contribute to the literature by supporting our results with clinical-laboratory parameters.

Conflict of interest

The authors declare that they have no conflict of interest.