Efficacy and safety of multi-kinase inhibitors in patients with radioiodine-refractory differentiated thyroid cancer: a systematic review and meta-analysis of clinical trials

ABSTRACT

Objectives

Radioiodine-refractory differentiated thyroid cancer (RAI-rDTC) has frequently been associated with poor prognosis. We conducted a meta-analysis of published randomized controlled trials to evaluate multi-kinase inhibitors’ efficacy and safety profile treatment.

Methods

A comprehensive search was conducted using PubMed, Embase, Cochrane, and Medline databases. The quality of literature and trial risk of bias was assessed using the Cochrane risk of bias tool, while the results of progression-free survival (PFS), overall survival (OS), and adverse events (AEs) were evaluated using RevMan5.3 software.

Results

Treatment with MKIs significantly improved PFS and OS, but AEs were significantly higher than those in the control group (P < 0.01). The studies demonstrated the median PFS (HR 0.30, 95% CI: 0.18–0.50, P < 0.00001) and OS (HR 0.70, 95% CI: 0.57–0.88, P = 0.002) in RAI-rDTC patients treated with MKIs, and the median PFS of papillary thyroid carcinoma (HR0.28, 95% CI: 0.22–0.37, P < 0.00001) along with follicular thyroid carcinoma (HR0.14, 95%CI 0.09–0.24, P < 0.00001) were extended.

Conclusion

MKIs significantly prolonged PFS and OS in patients with RAI-rDTC (P < 0.01). Our recommendation is to use MKIs carefully in patients after evaluating their health status to maximize treatment benefits and minimize adverse effects.

KEYWORDS:

• Multi-kinase inhibitors
• Radioiodine-refractory differentiated thyroid cancer
• efficacy
• adverse events
• Meta-analysis
• systematic review

1. Introduction

Thyroid cancer (TC) is the most common endocrine cancer, accounting for approximately 1–5% of all cancers in women, accounting for less than 2% in men [1]. Differentiated thyroid carcinoma (DTC) accounts for more than 90% of all thyroid cancers, comprising papillary thyroid carcinoma (PTC), follicular thyroid carcinoma (FTC, including Hürthle cell), and poorly differentiated thyroid carcinoma (PDTC) [2]. Women are more likely to develop DTC, but have a better prognosis [3–5]. DTC responds well to current treatments, including surgery, l-thyroxine therapy, and radioactive iodine (RAI). However, after refractoriness to radioactive iodine, metastasis occurs in up to 15% of cases, followed by a poor prognosis [6,7]. The sorafenib and lenvatinib of multi-kinase inhibitors (MKIs) have been approved by FDA for treating RAI-rDTC. Vandetanib and cabozantinib were previously only approved for treating medullary thyroid carcinoma (MTC). Exelixis announced in September 2021 that FDA had approved cabozantinib for RAI-rDTC treatment. Apatinib, developed in China and released in December of that year, significantly improved PFS (P < 0.001) and OS (P = 0.04) in patients with RAI-rDTC [8–13]. RAI-rDTC remained in clinical development yet unfavorable prognosis. We conducted a meta-analysis of published articles to evaluate strengths and weaknesses of MKIs medications, providing clinical patients with a reference point for selecting them. The mechanisms of action of these MKIs depend on the following signaling pathways, including phosphatidylinositol-3 kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR), rearranged during transfection (RET)/RAS/RAF-mitogen-activated protein kinase (MAPK), janus kinase (JAK)/signal transducer and transcription activator (STAT). Vandetanib is a small molecule MKI that inhibits the vascular endothelial growth factor receptor (VEGFR)-2, the epidermal growth factor receptor (EGFR), and RET [14]. Both sorafenib and lenvatinib inhibit the signaling pathways through VEGFRs 1–3, c-KIT, and RET genes. Furthermore, sorafenib targets platelet-derived growth factor receptor‐beta (PDGFR β), RAF-1, and BRAF, while lenvatinib targets fibroblast growth factor receptor1-4 (FGFR 1–4) and platelet‐derived growth factor receptor‐alpha (PDGFR-α) [15]. Cabozantinib and apatinib inhibit VEFGR-2, but cabozantinib is also an MKI that targets MET, KIT, RET, AXL, and (FMS-like tyrosine kinase 3, FLT-3) [16,17]. Angiogenesis inhibitors inhibit endothelial cell proliferation, migration, survival, as well as vessel permeability, while EGFR blocks cell proliferation, differentiation, and survival. Inhibiting mutant genes or fusion genes can increase hypoxia, reduce tumor activity, and stimulate apoptosis in tumor cells [8–13] (Figure 1).

Figure 1. Multi-kinase inhibitors signaling pathway for vandetanib, sorafenib, lenvatinib cabozantinib, and apatinib. FL, FLT3 ligand; FLT3, FMS-like tyrosine kinase 3; GDNF, glial cell-derived neurotrophic factor; GFR, GFRα, GDNF family receptor-α; RET, rearranged during transfection; HGF, hepatocyte growth factor; MET, mesenchymal-epithelial transition factor; SCF, stem cell factor; c-KIT, Proto-oncogene proteins c-kit;c-SRC, c-SRC tyrosine kinase; STAT, signal transducer and activator of transcription; JAK, janus kinase; MKIs, multi-kinase inhibitors; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor; PDGF, platelet‐derived growth factor; PDGFR, platelet‐derived growth factor receptor; FGFR, fibroblast growth factor receptor; PI3K (phosphatidylinositol-3 kinase) /AKT/ mammalian target of rapamycin (mTOR) signaling pathway; RAS/RAF/MEK/ERK signaling pathway.

2. Methods

The literature search was conducted by using PubMed, Embase, Cochrane, and Medline databases in accordance with PRISMA guidelines. Data up to 20 December 2021, were analyzed. Medical Subject Headings (MeSH) terms and free terms were used ((Thyroid Neoplasms) OR (Neoplasm, Thyroid) OR (Thyroid Neoplasm) OR (Neoplasms, Thyroid) OR (Thyroid Carcinoma) OR (Carcinoma, Thyroid) OR (Carcinomas, Thyroid) OR (Thyroid Carcinomas) OR (Cancer of Thyroid) OR (Thyroid Cancers) OR (Thyroid Cancer) OR (Cancer, Thyroid) OR (Cancers, Thyroid) OR (Cancer of the Thyroid) OR (Thyroid Adenoma) OR (Adenoma, Thyroid) OR (Adenomas, Thyroid) OR (Thyroid Adenomas)) AND (Randomized controlled trial). Only those patients who had ‘radioiodine-refractory differentiated thyroid cancer’ and were treated with ‘multi-kinase inhibitors’ were selected in this study. Only studies designated as ‘randomized controlled clinical trials’ were included in the search strategy. Detailed search strategy is displayed in Supplementary Appendix 1.

2.1. Inclusion criteria

Clinical trials involving MKI use in treating RAI-rDTC were included if they met the following criteria: (a) Aged 18 and above; (b) DTC confirmed by histological or cytological diagnosis with no response to 131-I treatment or unable to absorb iodine or post-131-I body scan is negative [18]; (c) diseases that can be measured according to Response Evaluation Criteria in Solid Tumors version1.1 (RECIST 1.1); (d) randomized controlled clinical trials with evidence of disease progression prior to enrollment and the control group was treated with placebo; (e) prior treatment no more than two MKIs; (f) adequate function of an organ, bone marrow, and blood coagulation, with serum thyroid stimulating hormone (TSH) less than the lower limit of the reference range or below the levels of 0.50 mIU/L.

2.2. Exclusion criteria

Clinical trials were excluded if they (a) included anaplastic thyroid cancer (ATC) or medullary thyroid cancer (MTC); (b) had more than two points according to Eastern Cooperative Oncology Group performance status (ECOG PS); (c) any anticancer treatment within one month prior to randomization; (d) animal experiment and in vitro experiments, systematic reviews, meta-analyses, conference abstracts, other publications of the same experiment, and subgroup analysis; (e) literature with incomplete and inaccurate statistics; studies that did not report major outcomes; (f) sample size less than 30 patients.

2.3. Data extraction

Two reviewers (JY and YF) independently screened titles, keywords, and abstracts, then, according to the selection criteria, they performed a secondary screening of the retrieved articles by reading their complete texts, followed by a final cross-check of the included information. A third reviewer (JW) would be involved in judging any possible questions. The information obtained from the articles featured the name of the first author, publication year, study design, MKIs name, earlier treatment, study characteristics (mean age and sample size), treatment (treatment of intervention group and control group, median PFS, median follow-up period and OS), in addition to outcomes on different pathological types and major AEs.

2.4. Assessment of risk of bias

Two evaluators (JY and YF) independently assessed the risk of bias using the Cochrane Risk of Bias tool. They evaluated the quality of trials in terms of random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, data integrity, selective reporting, and other biases. Each item was evaluated as low risk, unclear risk, or high risk.

2.5. Data synthesis and analysis

Data obtained from the studies were managed using Excel 2019. RevMan5.3 was used to evaluate the heterogeneity of direct comparisons and calculate HR and 95% CI. The Chi-square test tested heterogeneity for P-value and I² statistics. P-value was bound by 0.1 and 50% for I² statistic. In the case of the Chi-square test P < 0.10 or I² > 50%, which indicates a high heterogeneity, the random-effects model would be adopted, along with sensitivity analysis and subgroup analysis. Other than that, a fixed-effect model would be used when a low heterogeneity is considered.

3. Results

3.1. Search process

The literature search was conducted by the example of the established retrieval strategy through a database search of PubMed, Embase, Cochrane, and Medline, which resulted in 3747 articles. A total of 2325 studies remained after removing duplicates. After further screening, 64 studies remained, and only six met the inclusion criteria, and none met the exclusion criteria. An illustration of PRISMA selection process and PRISMA-NMA (Network Meta-Analyses) can be found in the PRISMA flow diagram as shown in Figure 2. A checklist of Items is documented in Supplementary Appendix 1.

Figure 2. PRISMA flow diagram of search.

3.2. Quality assessment

Two reviewers independently reviewed six articles, all of which were double-blinded, randomized controlled trials. The blinding of outcome assessment was unclear. Only one article failed to mention the measure of allocation concealment and was thus considered unclear, as demonstrated in Figure 3.

Figure 3. Risk-of-bias graph.

3.3. Study characteristics

A total of 1384 patients were included in six studies, including five MKIs, vandetanib, sorafenib, lenvatinib, cabozantinib, and apatinib. All studies were randomized, double-blind Phase III studies, except the vandetanib study, a randomized double-blind Phase II study. Across all studies, the sample size was greater than 30 participants, and whether MKIs were administered or not, the median PFS was significantly prolonged (P < 0.01) in Table 1. The hazard ratio of median PFS after the application in patients with different pathological types and AEs are presented in Table 2.

Table 1. Overview of literature characteristics.

References	Study design	MKIs name	Mean age Sample size	Previous treatment
Leboulleux 2012 [8]	randomized, double-blind, phase II trial	Vandetanib	63.5 years 145 patients	6 patients one TKIs 20 patients’ chemotherapy
Brose 2014 [9]	randomized, double-blind, phase III trial	Sorafenib	63.0 years 417 patients	None
Schlumberger 2015 [10]	randomized, double-blind, phase III trial	Lenvatinib	63.0 years 392 patients	72 patients sorafenib 8 patients sunitinib 5 patients pazopanib 8 patients other TKIs
Brose 2021 [11]	randomized, double-blind, phase III trial	Cabozantinib	65.3 years 187 patients	69 patients sorafenib 74 patients lenvatinib 44 patients both TKIs
Zheng 2021 [12]	randomized, double-blind, phase III trial	Lenvatinib	60.0 years 151 patients	38 patients one TKIs
Lin 2021 [13]	randomized, double-blind, phase III trial	Apatinib	57.7 years 92 patients	4 patients sorafenib 4 patients donafenib 3 patients’ chemotherapy
References	Intervention group		Control group
Leboulleux 2012 [8]	72 patients; vandetanib 300 mg qd Median PFS 11.1 months; Median duration of follow-up was 18.9 months		73 patients; placebo qd Median PFS 5.9 months; Median duration of follow-up was 19.5 months
Brose 2014 [9]	207 patients; sorafenib 400 mg bid Median PFS 10.8 months; Median The duration of follow-up was 16.2 months		210 patients; placebo bid Median PFS 5.8 months; Median The duration of follow-up was 16.2 months
Schlumberger 2015 [10]	261 patients; lenvatinib 24 mg qd Median PFS 18.3 months; Median The duration of follow-up was 17.1 months		131 patients; placebo qd Median PFS 3.6 months; Median duration of follow-up was 17.4 months
Brose 2021 [11]	125 patients; cabozantinib 60 mg qd Median PFS NR months; Median The duration of follow-up was 7.2 months		62 patients; placebo qd Median PFS 1.9 months; Median The duration of follow-up was 7.2 months
Zheng 2021 [12]	103 patients; lenvatinib 24 mg qd Median PFS 23.9 months; Median The duration of follow-up was 14.8 months		48 patients; placebo qd Median PFS 3.7 months; Median The duration of follow-up was 15.6 months
Lin 2021 [13]	46 patients; apatinib 500 mg qd Median PFS 22.2 months; Median duration of follow-up was 18.1 months		46 patients; placebo qd Median PFS 4.5 months; Median duration of follow-up was 18.1 months

TKIs, tyrosine kinase inhibitors.

PFS, progression-free survival. NR, not reached. qd, quaqua die.

Table 2. Different types of thyroid cancer and adverse events.

References	Outcome of different types and adverse events
Leboulleux 2012 [8]	The median PFS in papillary thyroid cancer was 16.2 vs. 5.9 months (HR 0.52, 95% CI: 0.26–1.02) in follicular thyroid cancer or poorly differentiated carcinoma was 7.7 vs. 5.6 months (HR 0.83, 95% CI: 0.51–1.36) in the vandetanib and placebo arms, respectively. The most common AEs leading to discontinuation of vandetanib were QTc prolongation and diarrhea. Common adverse effects in the vandetanib group were diarrhea, hypertension, QTc prolongation, rash, acne, and decreased appetite.
Brose 2014 [9]	Median PFS in BRAF mutations was 20.5 vs. 9.4 months (HR 0.46, 95% CI: 0.24–0.90) and in the wild-type BRAF subgroup was 8.9 vs. 3.8 months (HR 0.55, 95% CI: 0.38–0.79) in Sorafenib and placebo arms, respectively. The most common adverse effects leading to discontinuation of sorafenib were the HFSR. Common adverse events in the sorafenib group were HFSR, diarrhea, alopecia, rash/desquamation, fatigue, weight loss, and hypertension.
Schlumberger 2015 [10]	The median PFS in papillary thyroid cancer was 16.4 vs. 3.5 months (HR 0.30, 95% CI: 0.20–0.44); in follicular thyroid cancer it was 18.8 vs. 2.4 months (HR 0.07, 95% CI: 0.03–0.21) in lenvatinib and placebo arms, respectively. The most common AEs leading to discontinuation of lenvatinib were asthenia and hypertension. Common AEs in the lenvatinib group were hypertension, proteinuria, arterial thromboembolic effects, venous thromboembolic effects, renal failure, liver failure, gastrointestinal fistula, corrected QT prolongation and posterior reversible encephalopathy syndrome.
Brose 2021 [11]	The median PFS in papillary thyroid cancer was NR vs. 1.8 months (HR 0.23, 95% CI: 0.12–0.44), in follicular thyroid cancer it was NR vs. 3.6 months (HR 0.22, 95% CI: 0.11–0.44) in cabozantinib and placebo arms, respectively. The most common AEs leading to discontinuation of cabozantinib were disease progression, fatigue, arthralgia, diarrhea, hypercalcemia, hypertension, large-intestine perforation, increased liver Function test, myalgia, and renal failure. The most common grade 3 or 4 AEs were palmar-plantar erythrodysaesthesia, hypertension, fatigue, diarrhea, and hypocalcemia.
Zheng 2021 [12]	Median PFS in papillary thyroid cancer was 23.9 vs. 5.5 months (HR 0.19, 95% CI: 0.11–0.33), in follicular thyroid cancer was NE vs. 1.9 months (HR 0.12, 95% CI: 0.04–0.36) in the lenvatinib and placebo arms, respectively. The most common AEs E leading to the discontinuation of lenvatinib was an emerging TEAE. The most common grade 3 or 4 TEAE were hypertension, proteinuria, and HFSR.
Lin 2021 [13]	Median PFS in papillary thyroid cancer was (HR 0.32, 95% CI 0.16–0.61). In other thyroid cancer was (HR 0.12, 95% CI: 0.02–0.58) in the apatinib and placebo arms, respectively. There were no dose reductions or discontinuations in the apatinib group. The most common grade 3 or higher-level TEAE in the apatinib group were hypertension, hand-foot syndrome, proteinuria, and diarrhea.

PFS, progression-free survival. HR, hazard ratio. CI, confidence interval. AEs, adverse events. TEAE, treatment adverse event. HFSR, hand-foot skin reaction. QT, QT interval. QTc, QTc interval. NR, not reached.NE, not estimable. vs, versus.

3.4. Clinical outcomes

3.4.1. Progression‐free survival in RAI-rDTC

The influence of the intervention groups on the median PFS of patients with RAI-rDTC was statistically significant compared to the control group (HR 0.30, 95% CI: 0.18–0.50, P < 0.00001), and the risk of death was only 30% of the control group as illustrated in Figure 4(a).

Figure 4. (a) Meta-analysis of the median PFS in patients with RAI-rDTC. (b) Meta-analysis of the OS in patients with RAI-rDTC. (c) Meta-analysis of the median PFS in patients with PTC. (d) Meta-analysis of the median PFS in patients with FTC.

3.4.2. Overall survival in RAI-rDTC

P > 0.77 and I² = 0% in the Chi-square test indicated no heterogeneity, so the random-effects model was used. Analysis results indicate that MKIs intervention of RAI-refractory cancer significantly improved OS (HR 0.70, 95% CI: 0.57–0.88, P = 0.002), whereas its risk of death was only 70% compared to the control group as displayed in Figure 4(b).

3.4.3. Overall survival in PTC or FTC

MKI treatment was used in thyroid cancer of various types, and the median PFS in PTC was (HR 0.28, 95% CI: 0.22–0.37, P < 0.00001) as depicted in Figure 4(c). The median PFS in FTC (HR 0.14, 95% CI: 0.09–0.24, P < 0.00001) were prolonged, as demonstrated in Figure 4(d).

3.4.4. Adverse events

AEs associated with tyrosine kinase inhibitors (TKIs) have been investigated in all clinical trials. Six AEs were analyzed, including hypertension, hand-foot skin reaction (HFSR), diarrhea, proteinuria, decreased appetite, and weight loss. In all studies, the intervention group experienced significantly more adverse events than the control group (hypertension: risk ratio (RR) 4.19, 95% CI: 2.89–6.08, P < 0.00001; HFSR: RR 14.95, 95% CI: 6.79–32.93, P < 0.00001; diarrhea: RR 6.09, 95% CI: 4.20–8.82, P < 0.00001; proteinuria: RR 10.30, 95% CI: 5.75–18.42, P < 0.00001; decreased appetite: RR 2.69, 95% CI: 1.55–4.69, P = 0.0005; decreased weight: RR 3.93, 95% CI: 2.98–5.20, P < 0.00001) as depicted in Figure 5.

Figure 5. Six of the most common AEs in MKIs.

4. Discussion

The current meta-analysis summarizes the efficacy of MKIs and their associated AEs in RAI-rDTC. Our study is known to be the first RCTS meta-analysis to demonstrate the efficacy of MKIs in patients with RAI-rDTC. HRs of median PFS and OS were statistically significantly higher in the MKI treatment group than in the control group (P < 0.01). On the other hand, the outcome of median PFS in RAI-rDTC with P < 0.10 and I² > 50% in the heterogeneity test exhibited severe heterogeneity. Further analysis was conducted applying the random-effects model. The sensitivity analysis found no heterogeneity. In the subgroup analysis of MKI studies, there was no heterogeneity between vandetanib and sorafenib (HR 0.60, 95% CI: 0.49–0.75, P = 0.78, I² = 0%, P < 0.00001) and lenvatinib, cabozantinib, and apatinib studies (HR 0.20, 95% CI: 0.16–0.26, P = 0.63, I² = 0%, P < 0.00001). Due to the earlier start of the vandetanib and sorafenib studies, a high number of patients discontinued the drugs for severe adverse reactions, resulting in a low median progression-free survival (PFS). Moreover, no other MKIs drugs were available at the time. Patients had not received MKIs prior to enrollment in vandetanib and sorafenib studies. However, in lenvatinib, cabozantinib, and apatinib studies, the median PFS and OS were satisfactory even though some or all patients in the previous study had received one or two TKI drugs. Tumor cells may be resistant to various drugs. We need to compare the pharmacological effects of different drugs. The absence of randomized trials comparing two TKI agents could complicate the selection of initial treatment references for patients with RAI-rDTC.

Among the most common or severe (but less common) AEs of antiangiogenic drugs are hypertension, fatigue, decreased weight, diarrhea, QTc prolongation, congestive heart failure, proteinuria, liver damage, bone marrow suppression, HFSR, and other rashes. During this phase II trial, vandetanib was associated with diarrhea (74%), hypertension (34%), decreased appetite (26%), and QTc prolongation (23%) [8]. QTc prolongation (14%) ranked first for AEs above grade 3, indicating that cardiotoxicity was the most severe AE of vandetanib. The phase III clinical trial (NCT01876784) of vandetanib showed no statistical significance in PFS (10.0 versus 5.7 months, P = 0.080) compared with that of the placebo group. The high drop-out ratio of participants could be attributed to the cardiotoxicity of vandetanib [19]. Sorafenib was the first targeted drug approved by FDA to treat RAI-rDTC in November 2013 [20]. HFSR (76.3%), the most common AE, can interrupt the dose of sorafenib [9]. Kinase inhibitors are well-known to cause cutaneous and mucosal side effects [21]. Based on a meta-analysis, HFSR was found to be an effective indicator for the treatment of hepatocellular carcinoma with sorafenib [22]. Lenvatinib has a higher average PFS and OS values, a higher objective response rate, and some manageable side effects despite these benefits [9,10,23–25]. In the treatment guidelines for thyroid carcinoma published by the National Comprehensive Cancer Network (NCCN) in September 2021, lenvatinib was listed as the first recommended regimen for RAI-rDTC. Lenvatinib benefited patients with a small baseline tumor, low tumor burden, and lung metastases above 1 cm in size [26–28]. SELECT study participants reported specific AEs with lenvatinib including hypertension (67.8%), HFSR (31.8%), and proteinuria (31.0%). According to Kiyota et al. [29], Japanese patients had a higher incidence of treatment-related hypertension in the SELECT study, but its efficacy and safety were generally comparable to the general population. Cabozantinib has been reported as a TKI regardless of the previous VEGFR targeted therapy [30]. AEs of cabozantinib are mainly related to the digestive system, including diarrhea (51%), nausea (24%), and decreased appetite (23%) [11]. Elisei et al. [31]reported gastrointestinal perforations (3.3%), hemorrhage (3.3%), and fistula development (0.5%) in the cabozantinib arm. Apatinib and lenvatinib are highly selective VEFGR-2 inhibitors [10,17], and the common AEs reported of apatinib were also hypertension (87.0%), HFSR (87.0%), and proteinuria (76.1%). A randomized controlled clinical phase II clinical study of the anlotinib in China has demonstrated significant benefit for PFS (40.54 versus 8.38 months, HR 0.21, 95% CI: 0.12–0.37, P < 0.0001) [32]. Studies on donafenib, pazopanib, sunitinib, selumetinib, and axitinib have shown promising results, but it is expected that further studies will be conducted to ascertain the clinical effectiveness of these drugs [33–37].

Similarly, we analyzed the median PFS of patients with various pathologies and found that MKIs significantly correlated with PTC and FTC (P < 0.00001). The recurrence and metastasis of thyroid cancer are associated with several specific genetic alterations. PTC is associated with BRAF mutations (29–83%), RAS mutations (0–21%), RET rearrangements (14–43%), and NTRK rearrangements (3–13%), FTC is associated with RAS mutations (40–53%) [38]. Studies have revealed that BRAF and RAS mutations were linked to poor prognosis in DTC patients [39,40]. Both studies [9,10] reported that sorafenib and lenvatinib demonstrated a higher PFS than the control group regardless of the status of BRAF and RAS. Vemurafenib and dabrafenib are type I BRAF kinase inhibitors, functioning as ATP-competitive inhibitors [41]. Two cohorts of patients in a phase II trial with metastatic BRAF V600E mutant papillary carcinoma and refractory RAI received vemurafenib in a phase II trial. A 38.5% partial response (PR) and a median progression-free survival (PFS) of 18.2 months were reported in 26 patients who had never previously taken antiangiogenic multitargeted kinase inhibitors. In 25 patients treated with antiangiogenic multitarget kinase inhibitors, 27.3% achieved a PR, and the median PFS was 8.9 months [42]. In another clinical trial conducted on 14 patients with progressive radioiodine-refractory disease, the partial response rate (PRR) of treatment with dabrafenib was 29%, and the median PFS was 11.3 months [43]. Although BRAF kinase inhibitors are used as second-line therapeutic agents after antiangiogenic kinase inhibitors, more research and data are needed to confirm their efficacy. For patients with a RET fusion gene, selpercatinib or pralsetinib is an option [25]. In Europe, larotrectinib [44] and entrectinib [45] have been approved for treating NTRK fusion gene-positive patients with DTC. Pembrolizumab can be recommended for patients with a high mutation load.

The meta-analysis showed low heterogeneity and high reliability, but our study has limitations. First, the clinical trials we selected included fewer patients with Hürthle cell and poorly differentiated thyroid carcinoma, so they may not apply to these patients. Second, three trials did not make it to the final analysis time for OS, so the results may be biased. Although apatinib has the highest efficacy of the included studies, a third consideration is that this study was only conducted in China and may not apply to other races. Lastly, no randomized trials comparing MKIs with second-line agents or any two TKI agents can complicate the selection of initial treatment references for patients with RAI-rDTC. More studies are needed, to determine a most effective MKIs drug.

5. Conclusion

MKIs are a promising treatment for patients with RAI-rDTC. Currently, MKIs are the preferred treatment for RAI-rDTC patients, the use of MKIs can significantly improve PFS and OS (P < 0.01). Both PTC and FTC patients can use MKIs, but it is necessary to pay attention to the side effects of related medications and choose medications based on their economic and health status. To achieve the optimal outcome, the goal of treatment is to improve the patient’s quality of life. To achieve a long-term sustainable outcome, efforts should also be made to manage adverse events and focus on specific efficacy, with ongoing research.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

A reviewer on this paper has received honoraria from Exelixis.

Another reviewer on this paper has served as an invited speaker or participated on the advisory board for EISAI, Bayer, Lilly and Ipsen.

The remaining reviewers have no other relevant financial relationships or otherwise to disclose.

Supplementary material

Supplemental data for this article can be accessed here

Additional information

Funding

This paper was supported by the National Natural Science Foundation of China(81904197) of Jue Wang, by Science and technology planning projects of Zhejiang Province (2018C03025) and by the Zhejiang Lin Shengyou famous Traditional Chinese Medicine expert inheritance studio project (GZS202002) of Shengyou Lin.