Thyroid cancer is a common endocrine tumor, ranking seventh among all cancer types reported in China in 2022, with the fastest-growing incidence (1, 2). Differentiated thyroid cancer (DTC) is accounted for 95% of all histological types of thyroid cancer, including papillary thyroid cancer (PTC), follicular thyroid cancer (FTC), and Hürthle cell cancer (3). In general, thyroid cancer presents an excellent prognosis with standard management including surgery, radioactive iodine (RAI) treatment, and thyroid-stimulating hormone suppression therapy (4–6). However, about 20% and 10% of DTC patients are at risk of local recurrence and distant metastases respectively, and two-thirds of them lose the ability of RAI uptake initially or gradually, which was known as radioactive iodine refractory differentiated thyroid cancer (RR-DTC) (7). As limited treatment options result in poor prognosis, the 10-year survival rate is less than 10% in RR-DTC patients (8, 9).

Sorafenib, approved by the U.S. Food and Drug Administration and the State Food and Drug Administration of China for the treatment of RR-DTC, is an oral tyrosine kinase inhibitor (TKI) inhibiting vascular endothelial growth factor receptors (VEGFRs), RAF, RET, and platelet-derived growth factor receptor beta signaling (10–12). Compared with the placebo, sorafenib improved the progression-free survival (PFS) of RR-DTC patients showing efficacy and safety in DECISION trial (13). Real-world studies not only can reflect the effects of drugs and provide more important health condition information about patients in their daily life but also provide a more scientific basis guiding physicians’ clinical decision. Further demonstration of the clinical efficacy and safety of sorafenib used in different populations, centers, and age groups is essential.

Although many real-world studies on the treatment of RR-DTC with sorafenib have been reported (14–19), there are a few studies on the efficacy and adverse effects (AEs) of sorafenib at a standard dose of 400mg twice daily in Zhujiang Hospital of Southern Medical University. Our research not only analyzed the efficacy and AEs of sorafenib at a standard dose of 400mg twice daily for the treatment of RR-DTC patients at our Department of Nuclear Medicine from October 2017 to May 2020 but also summarized the real-world studies on sorafenib reported to date in the Discussion section.

Patients diagnosed with RR-DTC at our Department of Nuclear Medicine in Zhujiang Hospital of Southern Medical University from October 2017 to May 2020 were retrospectively analyzed. All patients were given sorafenib at the standard dose of 400mg, twice a day, and then the dose was adjusted according to the patient’s tolerance to side effects. Patients who received other treatments while taking sorafenib or were interrupted in follow-up were excluded. RR-DTC was defined as follows (if one of the followings is satisfied): 1) at least one target lesion without RAI uptake (never or ever); 2) progression of a target lesion despite significant RAI concentration; 3) cumulative RAI dose ≥ 22.3 GBq. Imaging examinations were performed every 3 to 6 months to assess efficacy, AEs were assessed monthly. The Response Evaluation Criteria in Solid Tumors version 1.1 was used to evaluate treatment efficacy (20): disease control rate (DCR) referred to complete response (CR) plus partial response (PR) plus stable disease (SD), objective response rate (ORR) was CR plus PR, and progression-free survival (PFS) was defined as the time from initiation of sorafenib to progression or death. Common Terminology Criteria for Adverse Events Version 5.0 was used to evaluate AEs and guide patients to adjust their medication in response to side effects: if patients had grade 1–2 AEs, the dose could be appropriately adjusted according to the patient’s ability to tolerate the AEs, if the patients had grade 3 or above AEs, the medication would be suspended until the AEs returned to grade 1–2. (Available from: https://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm#ctc_50. Accessed 15 May 2020).

Baseline characteristics of patients were described: quantitative data were reported as mean or median, and categorical variables as numbers with percentages. Waterfall plots were used to present the best tumor response. The PFS was calculated using the Kaplan–Meier method, and a log-rank test was performed to compare the difference in the PFS of variables. Cox proportional hazard model was used to estimate prognostic factors associated with PFS. All data were analyzed using SPSS version 26.0, and a p-value < 0.05 was considered statistically significant.

As shown in Table 6 , we summarize the real-world studies reported so far and the DECISION trial (13, 14, 16, 18, 19). The DCR of our study was 57.7%, similar to most of the above studies. However, the ORR in our study was slightly improved to 26.9% compared with those reported by Benekli et al., Massicotte et al., Kim et al., and Cheng et al. (14.0%,15.0%, 10.3%, and 21.7%, respectively). Furthermore, the median PFS in our study was 16.5 months, longer than those of DECISION, Kim et al., and Massicotte et al. (10.8, 9.7, and 7.2 months, respectively) (13, 14, 18), which may be due to the different tumor burdens of patients included such as those with advanced disease or distant metastases (14, 16, 18). However, the median PFS of the studies by Cheng et al. and Benekli et al. (17.6 and 21.3 months, respectively) was longer than ours, and a longer follow-up was found in the study of Cheng et al. than ours (median follow-up: 25.1 months versus 17.5 months) possibly resulting in longer PFS. Overall, both the clinical trial and real-world studies have demonstrated considerable efficacy in treating RR-DTC with sorafenib.

Studies have shown that soft tissue metastases such as lymph nodes, pleura, and pulmonary lesions presented better outcomes (16, 18, 21). Similarly, our study discovered that Sorafenib is more effective in lung metastases, all patients with PR were lung metastasis and patients with lung-only metastasis generally had a better prognosis such as typical cases shown in Figure 4 , while those with bone metastasis presented a worse prognosis ( Table 4 ). Besides, in the subgroup analysis of the relationship between PFS and clinicopathological features, as shown in Table 4 , PFS was significantly better in patients with lung-only metastases and significantly worse in patients with bone metastases, in the subgroup analysis of PFS between the presence and absence of bone metastasis groups, the bone metastasis group tended to have worse PFS. We believe that the reasons are as follows: VEGF stimulates endothelial cell proliferation and tumor angiogenesis, playing an important role in the development and progression of thyroid cancer, and its expression level correlates with advanced disease (22, 23). Sorafenib is a multitargeted TKI whose core mechanism is to block the activity of VEGFR-2 and VEGFR-3 to inhibit tumor angiogenesis and cell proliferation. VEGFR is mainly expressed in soft tissues such as heart, lung, kidney, and skeletal muscle (24). Therefore, we estimate that lower VEGFR signal transduction in bone may be responsible for the poor efficacy of bone metastases in line with the study of Cheng et al. (16). Beyond that, pathophysiological processes in the bone microenvironment promote the rapid progression of bone lesions, and complex bone composition and structure make it difficult to be targeted, which leads to the rapid development of bone metastases (25–27). Therefore, patients with lung metastases respond relatively well to sorafenib treatment. For bone metastasis, anti-angiogenic therapy alone may be less effective, and a multimodal approach including local and systemic therapies should be considered. However, in this cohort, patients with bone metastasis just tended to have worse PFS, and the p value was still greater than 0.05, which still needs more data to further confirm.

In our cohort, male had worse PFS than female ( Table 4 ; Figure 3 ). There are also various opinions on the relationship between gender and thyroid cancer prognosis. Some studies have shown that the prognosis of thyroid cancer in male patients is worse than that in female patients (28, 29), it may be related to the higher risk of radiation exposure and aggressive characteristics of male (28). However, there is no relevant study on the relationship between the efficacy of targeted drugs in RR-DTC treatment and gender, and we will further explore in our follow-up study.

Serum Tg is a well-recognized prognostic indicator for the presence of metastases or occult lesions after total thyroidectomy (6, 30, 31). However, RR-DTC lesions have various degrees of dedifferentiation, resulting in the loss of the ability to produce Tg; thus, it is controversial to use Tg to evaluate curative effects (32, 33). Our subgroup analysis revealed that patients with ≥ 60% decrease in Tg had longer PFS and better DCRs. Besides, patients in PR had a significant diminishment of the best change in Tg (more than a 90% decrease in six cases and a 58.2% decrease in one case). Kim et al. also found that a ≥ 60% decrease in Tg was associated with longer PFS and DCRs (14), and a similar conclusion was expounded by Cheng et al. (16). In contrast, Benekli et al. emphasized that Tg levels had no significance in prognosis or predicting response (19), and decreased or no change in Tg was observed in patients with advanced thyroid cancer treated with other TKIs (21, 34, 35). Whether the measurement of Tg can monitor therapeutic effects remains to be determined, but dynamic monitoring of changes in Tg such as the degree of Tg diminishment and Tg-doubling time can evaluate the efficacy of postoperative patients with thyroid cancer to a certain extent (31).

In our study, dose reduction and drug withdrawal occurred in 42.3% (n = 18) of patients due to AEs, mainly because of HFS, and a lower rate of AEs was found compared with the DECISION trial (93.1% versus 98.6%) (13), but a lower rate of AEs was found in the study of Cheng et al. (73.6%) than ours; therefore, a higher dose may result in a higher rate of AEs (16). Since common AEs were HFS, diarrhea, alopecia, abnormal liver function, generalized weakness, and hypertension, clinicians should closely monitor patients’ skin condition, blood pressure, and general situation when using sorafenib to avoid dose reduction or discontinuation.

It has been confirmed that patients treated with TKIs are at high risk of HFS (36). Virtually all patients had HFS (93.5%), most of them were grades 1–2, and only three were grades 3 and above. Moreover, the occurrence of HFS was correlated with better PFS in our analysis. Many studies have confirmed that HFS is associated with a higher cumulative sorafenib dose and with anti-VEGF and anti-VEGFR activities (17, 37). Since anti-angiogenesis is one of the main mechanisms of sorafenib in treating RR-DTC, HFS could be a significant therapeutic index during the administration of TKIs (38, 39). Thus, clinicians could estimate whether TKIs are starting to work based on the appearance of HFS.

As the sample size included in our study was not large and the median follow-up was only 17.5 months, no factors affecting PFS were identified in multivariate analysis. We will continue to follow up with patients for PFS and include more patients to analyze the factors influencing PFS.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Medical Ethics Committee of Zhujiang Hospital of Southern Medical University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

YL, WO and HF designed the study. YL and XX collected and analyzed the data and wrote the manuscript. YL, XX, and JL checked and corrected the data and manuscript. QZ, PC, LP, and ML participated in measuring and collecting the data. WO and HF reviewed and edited the manuscript and approved the final draft. All authors contributed to the article and approved the submitted version.