Incidence and Prognosis of Bone Metastasis in Newly Diagnosed Differentiated Thyroid Carcinoma: A SEER Population-Based Study

Background: Clinical characteristics and sociodemographic of differentiated thyroid carcinoma (DTC) patients with bone metastasis (BM) remain unclear. Methods: Data of patients with newly diagnosed DTC from 2010 to 2016 were retrieved from Surveillance, Epidemiology, and End Results (SEER) database. Incidence trend of BM in patients with DTC was explored by joinpoint regression. Multivariable logistic regression was utilized to gure out predictors of BM at initial presentation of DTC, and was further evaluated by receiver operator characteristics (ROC) analysis. Multivariable Cox regression was applied to reason out factors associated with overall survival and DTC-specic survival. Results: A total of 67,176 patients with DTC were identied from the database, with 0.36% (244/67176) presenting BM at the time of diagnosis. The age-adjusted incidence of BM in DTC was relatively stable during study period with an average annual percentage change (AAPC) of 2.52. Multivariable logistic regression recognized eight factors as predictors of presence of BM among the entire cohort. Multivariable Cox regression indicated multiple metastatic sites, total thyroidectomy and radioisotopes as predictors for survival of patients. Conclusions: The incidence of BM in DTC patients has been stable in recent years. Predictors and prognostic factors of BM in DTC patients were identied in this study.


Background
The overall incidence of thyroid carcinoma has been on the rise in recent decades [1], and it has been reported that the newly diagnosed thyroid carcinoma cases in 2018 was 567,233 worldwide [2].
Furthermore, new cases of thyroid carcinoma in 2020 will reach up to 52,890 in United States according to the newly published cancer statistics [3]. DTC is the most common type of thyroid cancer, as well as the most curable endocrine tumors with 10-year cancer-speci c survival of more than 70% [4]. Distant metastasis in DTC was extremely rare and has been regarded as predictor for poor prognosis [5]. It has been demonstrated that 10-year overall survival rate of DTC patient with distant metastasis was less than 50%, lower than that of no metastasis [6].
Second to lung, bone was reported to be one of the most common sites for metastasis of thyroid carcinoma, and it was still of great challenge to improve the prognosis in DTC patients with BM [7]. Spine (34.6%), pelvis (25.5%), sternum and ribs (18.3%), extremities (10.2%), shoulder girdle (5.4%), and craniomaxillofacial (5.4%) were the most common sites of BM in DTC [8]. Although BM was associated with poor prognosis, early detection and administration of appropriate therapy could signi cantly alleviate suffering of patients [9]. Treatments including radiotherapy, chemotherapy, and surgery have been applied or reported to improve outcomes of DTC patients with BM [9][10][11]. For instance, chemotherapy drugs such as Zoledronic Acid has been demonstrated to be effective in reducing skeletal-related events or delaying its appearance in patients with BM from DTC [7]. Moreover, bone stability of patients could be improved by using percutaneous osteoplasty combined with radioiodine therapy [12].
Early detection or prediction of BM in DTC was in urgent need for improving the prognosis as well as reducing unnecessary cost of patients. With early intervention, BM can be controlled or suppressed. Nevertheless, the promising data related to the incidence and prognosis of BM in newly diagnosed DTC, especially the population-based study, was extremely limited [9,13,14]. Even though characteristics of BM from DTC have been discussed in some case reports or reviews, we cannot reconcile these controversial observations to make a conclusion [15][16][17]. Thus, we extracted data from the SEER database from 2010-2016 to conduct this population-based analysis, aiming to identify the trend in incidence and investigate potential predictors and prognostic factors of BM in newly diagnosed DTC patients.

Methods
Data source SEER database was one of the most authoritative cancer databases worldwide, which collected individual data of cancer patients from various regional cancer registries throughout the United States since 1973, covering approximately 30 percent of the U.S. population [18]. Data in SEER database was deemed reliable ascribe to rigorous quality control studies and various data assessments. We obtained data of information of eligible patients in this study by using the SEER*Stat software version 8. 3.6 (Surveillance Research Program, National Cancer Institute, seer.cancer.gov/seerstat). The SEER database provided data of patients up to 2016 based on the Nov 2018 submission, and data on speci c metastatic sites including lung, bone, liver and brain was available since 2010. Therefore, the study period was set between Jan 2010 to Dec 2016.

Study design and participants
Patients with BM of newly diagnosed DTC during 2010 to 2016 were initially enrolled in this study and then screened by following the inclusion criteria below. In this study, variables collected through SEER database were composed of sociodemographic characteristics (year at diagnosis, age at diagnosis, race, gender, insurance situation and marital status), clinicopathological characteristics (laterality, histologic, American joint committee on cancer (AJCC) clinical stage, AJCC TNM classi cation, extraskeletal metastasis) and treatment-related information (surgical procedures, radiotherapy and chemotherapy). All data was con rmed by two independent authors (Lin Qi and Wenchao Zhang). Age was divided into 4 intervals (≤ 44, 45-54, 55-64 and ≥ 65) based on the major age of onset. Race was classi ed into white, black and others. Year of diagnosis was strati ed into two period (2010-2013 and 2014-2016). The primary endpoint in this study were overall survival and cancer-speci c survival, de ned as time from onset to death caused by all reasons and cancer respectively. The study design with corresponding process was shown in Fig. 1.

Statistical analysis
Description analysis was utilized to present the overall view of frequency and age-adjusted incidence of BM in newly diagnosed DTC. Frequency was de ned as the percentage of BM among the entire cohort or in the cases with distant metastasis. We principally focused on the entire cohort rather than the subset with metastatic disease because this was a population-based study. Sociodemographic, clinicopathological and treatment-related characteristics including year at diagnosis, age at diagnosis, race, gender, insurance situation, marital status, laterality, histologic, AJCC clinical stage, AJCC TMN classi cation, multifocality, extra skeletal metastases, surgery, radiation therapy and chemotherapy were employed to stratify the data. Survival data was expressed in the median with interquartile range.
Incidence of DTC patients with BM were age adjusted to the 2000 US standard population and expressed per 1 000 000 person-years. Joinpoint Regression Analysis Program (version 4.7.0.0; National Cancer Institute) was used to calculate annual percentage change (APC) and corresponding 95% CIs to present the trend in incidence.
The differences in sociodemographic, clinicopathological and treatment characteristics were compared by Pearson's chi-squared test. Univariable and Multivariable logistic regression were applied to gure out possible predictors of BM and build the predicting models. Corresponding Odds ratio (OR) along with 95% con dence intervals (CI) were also presented. The models were further assessed by using the ROC curves.
Meanwhile, the nomogram was established based on the signi cant predictors to predict the probability of BM in DTC patients. Furthermore, univariable and multivariable cox regression were conducted to identify prognostic factors for overall survival and cancer-speci c survival in DTC patients with BM. Kaplan-Meier curves along with log-rank tests were created according to the signi cant prognostic factors. All statistical analyses were done among both the entire cohort and subset with metastatic disease and were conducted utilizing SPSS version 24 for Windows (IBM, Armonk, New York, USA). Survival curves were generated by GraphPad Prism 8. The nomogram was established using R version 3.5.3. Two-sided P < 0.05 was de ned as statistical signi cance.  Moreover, ROC analysis was conducted to further assess the multivariable logistic regression models with signi cant predictors above. The multivariable logistic regression model of the entire cohort exhibited optimal performance with the value of AUC of 0.893 (Supplemental Fig. 1). Furthermore, the nomogram predicting the probability of BM in DTC patients was further established based on signi cant predictors identi ed through multivariable logistic regression (Fig. 3). Each variable had a corresponding score in the nomogram (Supplemental table 5). We could estimate the probability of undergoing amputation according to the total scores.

Survival Analysis
Findings above gave insight of BM incidence and prediction of presence in newly diagnosed DTC patients. In order to explore roles of these factors in prognosis of DTC patients with BM, we carried out the survival analysis. The univariable cox regression analysis for overall survival and cancer-speci c survival in DTC patients with BM was present in Table 3. On multivariable analysis (  (Fig. 4A), surgery (Fig. 4B) and radiation therapy (Fig. 4C), and cancer-speci c survival estimates as strati ed by metastatic site (Fig. 4D), surgery (Fig. 4E), year at diagnosis (Fig. 4F) were illustrated in the Fig. 4.   [19] and breast cancer [20,21]. Besides, BM often occurs in medullary and undifferentiated thyroid cancer [22], and limited characteristics about BM in DTC have been reported. Despite several case reports and cohort studies [23][24][25], little was known about its incidence, risk factor and prognosis in DTC. In this population-based study, we analyzed the trend in incidence, survival, risk factors and prognostic factors of BM in newly diagnosed DTC patients, aiming to promote understanding on the current situation and the association between clinical pattern and BM. To the best of our knowledge, this study included largest sample size of DTC patients with BM compared with previous ones.
In the present study utilizing SEER database, BM accounts for 0.36% (244/67176) among entire group and 34.76% (244/702) among subset with metastatic diseases in newly diagnosed DTC patients, which was lower than the proportion (2.0%) in study of Yorihisa et al. [9] and the proportion (5.51%) in study of Marie-odile et al.
[26] Sample size in these two studies was relatively small and they only included patients in a single institution; therefore the results in the above studies were not representative. In addition, the above two studies were conducted with date in early period, while our study focused on patients diagnosed during 2010 to 2016. With the advancement of screening methods, early detection of asymptomatic thyroid cancer and timely interventions such as surgery have substantially enhanced the prognosis of thyroid tumor patients. Therefore, the frequency of bone metastases in DTC patients may have reduced compared with early decades.
To explore the trend of incidence of BM in DTC patients, Jointpoint regression analysis was introduced, and result indicated that the age-adjusted incidence of BM in DTC was relatively stable during study period, which was not reported before. Besides, incidence of thyroid cancer was in constant increasing during 1974 to 2013 in United States [1].  [29]. However, as it was discussed, SEER database did not provide information about speci c anatomic sites of distant metastasis at that time, so the detailed data for BM was not getable in their analysis.
Furthermore, the optimal performance of multivariable logistic regression model in the entire cohort was authenticated by ROC analysis, in which the AUC value was 0.893. Besides, the nomogram established in this study is an e cient tool to help doctors decision-making owing to its user-friendly interface and optimal predictive ability. Nevertheless, we were not capable of guring out the most common site for BM because there was no data documented in SEER database until now.
Prognostic factors for overall survival and cancer-speci c survival of DTC patient with BM was analyzed using the multivariable Cox model. Results revealed that patients with multiple metastatic sites have higher hazard ratio for overall survival and cancer-speci c survival, which was in coherence with previous studies [10,30]. Meanwhile, signi cant decrease in the hazard ratio for overall survival and cancer-speci c survival was found in patients with total thyroidectomy and radioisotopes radiation therapy. Surgery was the mainstay of therapy for thyroid cancer patients, especially in patients without distant metastasis [31]. Surgical options include thyroid total thyroidectomy or lobectomy, and it was still in controversy on how to choose optimal surgical procedure because of the di culty in balancing surgical effect and complication [32]. Total thyroidectomy reduces recurrence and allows early detection of recurrence in the neck using ultrasonography compared with lobectomy, thereafter improving the prognosis compared with others. The most used radioisotope in DTC was radioiodine, but its using was still of wide divergence, partially due to the lack of evident prospective randomized controlled trials [33]. Patients with BM could bene t from radioiodine in both suppressing the progression of tumor and reducing recurrence of primary carcinoma as we concluded. It was testi ed in some retrospective studies that application of radioiodine could improve survival of DTC patients [9,34,35] as well as BM patients [36]. Especially, radioiodine therapy was effective in the ablation of thyroid remnant after surgery [37]. External beam radiation therapy (EBRT) was not associated with positive outcomes in our analysis. Inversely, a retrospective study enrolled 74 BM patients has demonstrated that EBRT signi cantly increased survival [38]. Nevertheless, EBRT was reported to have acute toxicities such as esophageal stricture, so it was used under relatively strict indication [39]. Besides, combining radioiodine and external beam radiation therapy seems advantageous in some cases [40]. However, the treatment effect of total thyroidectomy and radioiodine in DTC patients with BM need to be further explored in the future study. Other factors did not exhibit prognostic value in our analysis, but their role has been demonstrated in other studies. Older age, higher T stage, positive lymph node, and black race were predictors of positive outcome [24,25]. Further study is necessary to shed light on their prognostic value in BM.
Although this was a population-based study that enrolled largest cohort of DTC patients with BM, limitations in this study was unneglectable. Because it was a retrospective study based on current database record, some parameters were unavailable. Firstly, the further detailed information of BM like speci c sites was not available in SEER database, which limited further analysis. Secondly, onset of BM during follow-up period was not presented and we only had information on synchronous metastasis, so the incidence may be underestimated. Thirdly, more detailed information about radiation or chemotherapy such as chemotherapy regimens, doses, and the speci c number of cycles were not recorded in SEER database. Fourthly, except the survival, other outcomes such as complications of exposures or interventions were not reported. We could only assess the prognosis by survival data.
Finally, data on speci c metastatic sites including lung, bone, liver and brain was available in SEER database since 2010, leading to some inevitable bias. Therefore, more convincible clinical studies especially the randomized clinical trials were necessitated in the future.
In summary, results of this population-based study highlighted the incidence, predictors and prognostic factors of BM in newly diagnosed DTC patients.  Trends in annual incidence rates of DTC with different distant metastasis to speci c sites