In order to more accurately predict the osteoporosis risk in patients with differentiated thyroid cancer treated with levothyroxine inhibition following thyroidectomy, we conducted this study. We found that patients treated with TSH suppression following thyroidectomy had a 0.27 reduction in bone mineral density T score and higher fracture risk compared with matched controls. Both PMOF with BMD and PMOF after TSH assignment increased by 0.8%. These results suggest that TSH suppression has a adverse effects on BMD and fracture risk in postoperative differentiated thyroid cancer patients. When the patients were grouped according to TSH level, the lower the TSH level, the higher the fracture risk without BMD and the fracture risk with TSH level. It can be obtained by the ROC operating curve that the FRAX score with TSH level also improved the diagnostic efficiency of osteoporosis compared with fracture risk without BMD. The FRAX by TSH weighted is convenient and quick, making it ideal for rapid clinical assessment of a patient's risk of osteophosphora fracture without waiting for BMD results, so as to facilitate early intervention treatment of osteoporotic fractures and save cost.
Our findings provide further evidence of the association between TSH suppressive therapy and osteoporosis. The study by Lin SY et al. showed a strong dose-response and duration response relationship between levothyroxine use and osteoporosis risk in patients with differentiated thyroid cancer [20]. Moreover, Heijckmann AC et al. found that patients with well-differentiated thyroid cancer did not have an increased risk of low bone mass or an increased incidence of vertebral fractures, at least when they were treated with relatively low doses of levothyroxine [24]. Our data proved that differentiated thyroid cancer patients treated with levothyroxine had an increased risk of osteoporosis, but T scores did not appear to be affected by TSH, and there was no significant difference in hazard ratio between TSH groups, and there was no clear linear trend, considering that the dosage and time in medication treatment were related. However, the fracture risk calculated by TSH-weighted FRAX score was significantly higher than that calculated by the other two methods.
In addition to the risk factors currently included in FRAX, previous studies have shown that type 2 diabetes mellitus and previous falls should be considered as independent risk factors in predicting MOF and HF [25–28]. In this study, there were concomitant diabetes in each group of patients. It may have some effects on bone mineral density, but we have tried to control the duration of diabetes in patients within 1 year to reduce the error caused by this. There is no study to show whether the TSH level of postoperative patients with differentiated thyroid cancer can be included in the risk factors of FRAX tool. However, some studies have shown that high levels of TSH were positively correlated with BMD of femoral neck and lumbar spine [29]. Similarly, some clinical studies suggest that low-normal TSH is independently associated with bone mass loss, increased bone turnover, and decreased BMD in both men and postmenopausal women [30–32]. Direct effects of TSH on bone remodeling, osteoblast bone formation, and osteoclast bone resorption are mediated by the TSH receptor (TSHR) in osteoblasts and osteoclast precursors [33, 34]. This has also been demonstrated in animal experiments where TSH from the anterior pituitary inhibits bone resorption by osteoclasts. Various evidence has shown that TSH is closely related to osteoporosis. Based on our results, clinicians assessing the risk of fracture in postoperative differentiated thyroid cancer patients should adopt TSH-weighed age in FRAX calculation according to our estimating method (when BMD is unknown) to account for the independent effect of TSH on MOF and HF. But larger studies are needed to examine whether TSH is an independent factor in improving FRAX prediction of fracture risk.
Our work has important clinical implications. We believe that FRAX score can be applied to guide physicians' treatment decisions in the treatment of osteoporotic patients with differentiated thyroid cancer after postoperative TSH suppression therapy. Due to the differences in national conditions and epidemiology, the intervention thresholds based on the cost-effectiveness of FRAX score are different in different countries [35–39]. However, there are few studies on the FRAX-based osteoporosis intervention threshold in China. A prospective cohort study of postmenopausal women at Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, showed that for Chinese postmenopausal women, the greatest clinical and economic benifits can be achieved when the FRAX intervention threshold is set at 7% [40]. From the ROC curve, it can be seen that the specificity of the FRAX after TSH weighting (without BMD) was higher than that of the FRAX without score (PMOF:0.936 VS 0.638,PHF:0.915 VS 0.596), which was close to the FRAX with only BMD (PMOF: 0.936, PHF: 0.936), but the sensitivity was slightly lower (PMOF:0.412,PHF:0.471). The FRAX cut-off value after TSH-scoring was also the highest among the three groups. Therefore, when it is used as the diagnostic criteria of osteoporosis, it will reduced the rate of misdiagnosis. But there is also the possibility of misdiagnosis. The FRAX with TSH-weighed improved the accuracy and specificity of predicting fracture risk without BMD, and provided a solid foundation for subsequent intervention and FRAX-based intervention thresholds for maximum clinical benefit in the differentiated thyroid cancer population.
However, certain limitations of our study must be acknowledged. Some patients have a brief medication duration (< 1 year) and a short duration of TSH suppression therapy, that there is no significant impact on bone density at present, and may reach a high bone density. The treatment of differentiated thyroid cancer patients with osteoporosis, active vitamin D, bisphosphonate, recombinant human parathyroid hormone, deslizumab and other osteoporosis drugs needs further study.
In summary, our findings provide additional evidence that TSH suppression therapy decreases bone mineral density in postoperative patients with differentiated thyroid cancer. The FRAX tool is suitable for postoperative differentiated thyroid cancer patients received TSH suppression therapy, and the TSH-weighed FRAX can improve the accuracy of osteoporosis diagnosis without BMD. Whether TSH can be included in FRAX as an independent risk factor for 10-year fracture risk need further research.