The proto-oncogene RET encodes a membrane TK transmembrane receptor (TK-R) that functions upstream of the MAPK and PI3K-AKT. RET is an oncogene that can be overactivated either i) by a fusion rearrangement of the TK domain of RET gene and the 5′ domain of other genes (known as RET/PTC rearrangements) or ii) by activating point mutations; RET alterations lead to activation of mitogen-activated MAPK pathway, resulting in cancer development. Nikiforov (2002) discussed the most common RET/PTC rearrangements occurring in TC: among at least 10 different types of RET/PTC, the most common types are RET/PTC1 and RET/PTC3, accounting for >90% of all rearrangements (17). He reported that the most frequent RET/PTC rearrangements occur in a follicular histological type of TC (PTC). Moreover, different types of RET/PTC correlated with distinct clinical features of PTC: RET/PTC1 leads to a benign clinical course in tumors with typical papillary growth and microcarcinomas, whereas RET/PTC3 tends to be associated with the solid variant of PTC and more aggressiveness (17). Instead, activating RET point mutations have been exclusively found in MTC: RET-activating point mutations are present at the germline level in approximately 100% of hereditary forms and 50% of sporadic cases (4). Thus, tests to identify RET germline mutations are required both in hereditary forms and in apparently sporadic cases to identify respectively gene carriers and undiagnosed familial tumors (18, 19).

Additional genomic events occurring in TC are gene amplifications resulting in the gain of genes’ copy number that encodes for TK-R, such as EGFRs, PDGFRs (PDGFR-α and PDGFR-β), VEGFRs (VEGFR1, VEGFR2, and VEGFR3), stem cell factor receptor (c-KIT), and hepatocyte growth factor (HGF) receptor (MET) (20).

Among the growth factors, vascular endothelial growth factor (VEGF) has appeared to be the most prominent due to its involvement in tumor angiogenesis and tumor growth. Indeed, the angiogenesis process becomes pathophysiological when the tumor uses it developmentally. VEGF comprises VEGF-A, VEGF-B, VEGF-C, VEGF-D, and the placental growth factor (PGF), each capable of binding different VEGFRs (21, 22). A study conducted among TC cases with follicular origin highlighted that VEGF expression was more prevalent in PTC (79%) than in FTC (50%) or PDTC (37%); moreover, more than half of these tumors co-expressed the VEGF and its receptors, going in an autocrine loop of VEGF signal (23). VEGF overexpression is a consequence of an overactivation of hypoxia-inducible factor-1 alpha (HIF-1α); this factor is expressed in TC cells (especially in ATC cells), but not in normal thyroid tissues, where it causes intratumoral hypoxia. Upregulation of HIF-1α also correlated with increased transcription of MET gene in PTC and ATC and correlated with tumor invasiveness (24). MET is a TK-R, also known as an HGF receptor due to its high affinity for HGF. The induction of MET results in the activation of signals implicated with proliferation, cell survival, cell scattering/migration, and morphogenesis. Deregulated HGF-MET signaling is implicated in oncogenesis, tumor invasiveness, and therapeutic resistance in several cancers, including TC (24).

EGFR appears to be an attractive target for molecular therapy due to its role in the pathogenesis of many types of cancer. Overexpression of EGFR in PTC has been associated with a worse prognosis (25).

The occurrence of NTRK fusion in TC is very rare. Fusions involving NTRK gene family, including NTRK1, NTRK2, and NTRK3, lead to the constitutive activation of active kinase function due to chimeric rearrangements in tropomyosin receptor kinases (TRKs) A, B, and C, respectively. NTRK fusion is an oncogenic drive mutation and occurs in approximately 0.3% of all solid tumors (26). Eight NTRK fusion types were identified in TC and seem to be associated with a 100% probability of malignancy (27). NTRK fusion is mutually exclusive with other driver mutations, but in rare cases, the co-occurrence of NTRK fusions and BRAF^V600E mutation has been reported (28, 29) A recent study investigated the impact of NTRK-rearranged tumors on the prognosis of different TC types (30). The authors detected NTRK rearrangements in 59 of 989 TC tissues, with prevalence in PTC due to high pediatric PTC sample size (n = 104), which are known to harbor NTRK fusions with a higher frequency (approximately 20%) (30).

Genetic alterations occurring in the MAPK axis play important roles in the development of cancer. Several studies showed that the main activating mutations in these transduction pathways can occur in RAS gene or BRAF proto-oncogenes (31).

RAS mutations might result in the acceleration of mitogenic activity and tumor progression. Incidences of RAS mutations correlated with the histological differentiation of TC. Evidence suggests that RAS mutations may sustain DTC dedifferentiation into PDTC and ATC (32). RAS mutations are one of the most frequent alterations found in FTC (40%–50%) (33) and in about 10% of PTC. Aside from RET point mutations in MTC, the only other alterations occurring are H- and K-mutations in RAS (reported in about 17%–80% of RET-negative sporadic MTC) (4).

Mutations in BRAF gene lead to the constitutive activation of the BRAF kinase (34). In his review Xing (2005) highlighted that among the known common oncogenic genetic alterations occurring in TC, BRAF mutations are the most frequent in PTC, with the prevalence of V600E mutation; BRAF^V600E mutation is more frequent in PTC (44% in the studied population) and in PTC-derived ATC tumors (24% of the studied population), compared to other histological subtypes, such as FTC, MTC, or benign thyroid tumors (34). Indeed, the abnormal activation of the MAPK cascade, due to mutations and/or rearrangements in RET, RAS, and BRAF genes, characterizes approximately 70% of PTC cases (31). BRAF^V600E is one of the most prevalent alterations in ATC and in PDTC; other frequent oncogene alterations in ATC are PI3CA, PTEN, IDH1, and ALK mutations (35). Xu et al. (2020) highlighted that BRAF, RAS, TP53, and TERT promoter mutations were the most common mutations that occurred in a cohort of 126 cases of ATC (45%, 24%, 63%, and 75%, respectively); in this study, they also described the very rare occurrence of NTRK fusion (3 cases) (36). While mutations of RET, BRAF, and RAS occurring in DTC and MTC are almost always mutually exclusive, ATC is characterized by a higher number of mutations in the same tissue, probably responsible for a more aggressive phenotype (12, 35, 37).

The constitutive activation of receptors or tyrosine phosphatase involved in the MAPK axis leads to activation of the MAPK pathway, which is associated with dedifferentiation, due to inhibition in the expression of thyroid hormone biosynthesis genes, including the NIS and thyroid peroxidase. In 2014, The Cancer Genome Atlas (TCGA) Research Network published a comprehensive characterization of 496 PTC (38). This study highlighted that BRAF^V600E-mutated PTC, which had the strongest activation of the MAPK pathway, showed the most dedifferentiated state, that is, low expression of some thyroid differentiation genes such as SCLC5A5 gene encoding for the NIS, thyroglobulin, or thyroid peroxidase (39). Therefore, drugs inhibiting one or most of the proteins involved in the MAPK pathway could be useful to improve iodine uptake and allow RAI therapy.

While paired box gene 8 (PAX8) is a thyroid-specific transcription factor, peroxisome proliferator-activated receptor γ (PPARγ) gene is a transcription factor ubiquitously expressed; their fusion resulting in the PAX8/PPARγ oncogene that promotes cell growth reduces rates of apoptosis and allow anchorage-independent and contact-uninhibited growth of thyroid cell lines (40). Moreover, it is associated with tumor multifocality and vascular invasion (41). PAX8/PPARγ rearrangements are detected in about one-third of FTC (35%) (40) and a follicular variant of PTC, but not in classic PTC. The major molecular alterations occurring in TC are summarized in Table 1 .

Systemic therapy should be considered in patients with progressive RAI-R DTC. The American Thyroid Association (ATA) guidelines for DTC defined RAI-R DTC as i) malignant or metastatic tissues not able to concentrate RAI, ii) tissues that have lost the ability to concentrate RAI, iii) when metastatic diseases progress despite the ability to concentrate RAI, or iv) when RAI has concentrated in some tissues and not others (9). Haugen (1999) reviewed the use of several chemotherapeutic agents for the treatment of RAI-R DTC (e.g., doxorubicin, paclitaxel, bleomycin, cisplatin, carboplatin, methotrexate, melphalan, mitoxantrone, and etoposide), which demonstrated no significant improvement in response rates (RRs) (56). Since traditional cytotoxic systemic chemotherapy has had minimal efficacy in patients with metastatic differentiated thyroid disease, ATA guidelines suggest starting treatment with a kinase inhibitor in RAI-R DTC patients with metastatic, rapidly progressive, symptomatic, and/or imminently threatening diseases not sensible to local control using other approaches (9, 57). Currently approved kinase inhibitors for the treatment of progressive DTC that is refractory to RAI treatment included sorafenib, lenvatinib, and, most recently, cabozantinib (FDA approved only). However, several other commercially available TKIs (e.g., axitinib, pazopanib, motesanib, sunitinib, vemurafenib, and selumetinib) approved for other diseases are currently undergoing clinical trials and are being tested for their use in RAI-R DTC (42).

Optimal management of patients with metastatic DTC requires careful consideration of multiple tumor-associated and patient-related factors within the context of an experienced multidisciplinary team. Optimal decision-making with regard to when to initiate TKI therapy in the metastatic disease setting requires the clinicians and the patients to thoughtfully integrate an understanding of the patients’ symptom burden, tumor progression, and potential tolerance of treatment-related side effects (58).

Moreover, it could be useful, as suggested by Tuttle et al. (2017), to use the doubling time curves to define the critical point in time when the volume and rate of progression of metastatic structural disease deserve consideration for starting a systemic therapy (58).

Patients with disease progression while on initial kinase therapy should be considered for second-line kinase inhibitor therapy. Disease progression was evaluated following Response Evaluation Criteria in Solid Tumors (RECIST) criteria (59).

In addition, as recently shown, re-challenge with lenvatinib could be an option in subjects with metastatic DTC in progression after initial response to the same TKI (60).

Since the development of MTC is often driven by a single mutation in an oncogenic kinase driver (66), its treatment is different than that of RAI-R DTC.

According to ATA guidelines, treatment with TKIs targeting both RET and VEGFR TK should be considered the treatment of choice in patients with significant tumor burden and advanced MTC. The FDA and EMA have approved vandetanib (2011) and cabozantinib (2012) for the treatment of advanced progressive MTC. Vandetanib was also approved for the treatment of aggressive and symptomatic MTC in children and adolescents aged 5 years and older (67).

In September 2021, the FDA approved cabozantinib for the treatment of RAI-R DTC that has progressed after previous treatment with a VEGFR inhibitor. This approval is based on the results of the phase III, randomized, double-blind, placebo-controlled clinical trial COSMIC-311, ClinicalTrial.gov ID: NCT03690388 (51). Enrolled patients (n = 187) were randomized 2:1 to receive cabozantinib 60 mg once daily or a placebo. The primary endpoints of the study included PFS and ORR. Cabozantinib showed significant improvement in PFS of 11 versus 1.9 months as compared to the placebo group (hazard ratio (HR), 0.22; 95% CI, 0.15–0.31). Cabozantinib could represent an important treatment option in patients affected by RAI-R DTC who have progressed following prior therapy.

Selpercatinib and pralsetinib are highly selective RET kinase inhibitors. Approval of both agents provided new treatment options for the patient population with RET-altered TC. Selpercatinib was approved by the FDA and EMA in 2020 for the treatment of RET-mutated MTC and RAI-R RET-fusion TC. The efficacy of selpercatinib in RET-mutated TC was investigated in a phase I–II study (LIBRETTO-001, ClinicalTrials.gov ID: NCT03157128), showing durable efficacy with mainly low-grade toxic effects in patients with MTC who previously received cabozantinib or vandetanib (n = 55), or were treatment naive (n = 88), and in patients harboring RET fusion-positive TC (n = 19) (53). Results of the trial documented high efficacy of selpercatinib associated with minimal side effects and excellent tolerability.

In December 2020, the FDA approved pralsetinib for advanced or metastatic RET-mutant MTC or RET-positive RAI-R TC. In March 2022, EMA approved pralsetinib for the same indications. The efficacy of pralsetinib in patients with RET-altered TC was investigated in a multi-cohort, open-label, phase I/II trial (ARROW, ClinicalTrials.gov ID: NCT03037385) (54). The clinical trial enrolled both patients with RET-mutant MTC (n = 122) who were treatment naive or had previously received cabozantinib or vandetanib, or both, as well as patients with RET-fusion positive TC (n = 20) (RR of 71%, 60%, and 89%, respectively) (54). Pralsetinib had a manageable safety profile in patients with RET-altered TC and provided meaningful clinical activity in patients irrespective of previous treatment history (54).

Although vandetanib and cabozantinib were previously approved for the treatment of metastatic MTC, some patients are not eligible for these therapies on the basis of an unacceptable risk of bleeding or concerns about wound healing (70). Indeed, treatment with both agents required dose reduction in 35% of the patients receiving vandetanib and 79% of those receiving cabozantinib and permanent discontinuation of therapy in 12% and 16% of the patients, respectively observed in the ZETA trial and EXAM trial (47, 49, 53). Until the introduction of selpercatinib and pralsetinib in clinical practice, there were no standard therapies in patients who progressed on cabozantinib or vandetanib (54). Both selpercatinib and pralsetinib showed high activity in patients who previously received cabozantinib or vandetanib, or both (ORRs of 60% and 69%, respectively), including in patients with the gatekeeper V804L/M mutation, which confers resistance to both therapies (53, 54). In the LIBRETTO-001 and ARROW clinical trials, in the arms of patients with treatment-naive RET-mutant MTC who received selpercatinib or pralsetinib, ORRs were 73% (95% CI, 62–82) and 71% (95% CI, 48–89), respectively, suggesting that these RET-targeted inhibitors might have a therapeutic advantage over available first-line MTKIs in the RET-altered population (53, 54). The utility of RET inhibitors in the population of patent with RET fusion-positive TC who previously received RAI was also demonstrated: pralsetinib showed favorable activity in these patients’ ORRs (89%), compared with rates reported in patients with radioiodine-refractory TC treated with sorafenib (12%) and lenvatinib (65%1) (54).

Since ATC is a rare and aggressive tumor, it is still challenging to predict the patient clinical therapy responsiveness. Until a few years ago, there were no efficient treatments leading to an improvement in survival in those patients. Several genetic mutations have been described in ATC, involved in different molecular pathways linked to tumor progression (shown in Table 1 ). Few clinical trials investigated the use of TKIs in the treatment of ATC (71–74), to improve the quality of life in these patients. Collectively, the results of these studies including target therapy in the treatment of ATC have been overall disappointing. Among these studies, a phase II/III trial (FACT, ClinicalTria.gov ID: NCT00507429) evaluated combretastatin A (CA4P) in combination with carboplatin and paclitaxel in patients with ATC (n = 80) and suggested a survival benefit with the combination therapy. Indeed, the median survival time was 5.2 months in the CA4P arm versus 4.0 months in the control group, showing a reduction in the risk of death of 35%. The 1-year survival was 27% in the CA4P group versus 9% in the control group.

Treatment of ATC has had major advances in the last several years (75). Ferrari et al. (2019) conducted a systematic review of studies involving the treatment of anaplastic thyroid from 1995 to 2017 (76). They highlighted that great attention has been given to the epigenetic alterations underlying thyroid carcinogenesis, including those that drive PDTC and ATC (76, 77). Among these, BRAF^V600E is a common somatic mutation occurring in ATC, which can be effectively treated with BRAF/MEK inhibitors (78). Recently, the FDA approved dabrafenib and trametinib in BRAF^V600E mutant ATC, based on a phase II trial in which the treatment using these two TKIs showed a remarkable RR of 69% in patients with ATC (n = 16), with one patient achieving complete response (52). NTRK fusion also can occur in ATC; rare ATC harboring NTRK fusion can be treated with recent approved agnostic therapies (75).

The guidelines of the American Society of Clinical Oncology and European Society for Medical Oncology recommended the use of larotrectinib and entrectinib as treatment of choice for solid tumors with NTRK gene fusions, including TC (79), but not NTRK-mutated, solid tumors (75). Regulatory approval of both agents was based on data from single-arm phase I/II studies, including tumor-agnostic basket trials that enrolled patients based on the presence of NTRK gene fusions (55).

Larotrectinib is an inhibitor of TRK 1–3 and was tested in adult, adolescent, and pediatric patients with solid tumors harboring NTRK gene fusion (a phase I study, ClinicalTria.gov ID: NCT02122913; SCOUT phase 1/2 study, ClinicalTria.gov ID: NCT02637687; NAVIGATE basket study, ClinicalTria.gov ID: NCT02576431) (79). The efficacy of larotrectinib in TC was provided in a recent pooled data analysis derived from three phase I/II clinical trials (ClinicalTria.gov ID: NCT02576431, NCT02122913, NCT02637687) that enrolled 29 patients with TRK fusion-positive TC (80). Patient stratification and the main clinical outcome of this study are shown in Table 5 .

Entrectinib inhibits TRK 1–3, as well as the ALK and ROS1 TKs; its approval was based on a pooled analysis of 3 single-arm phase 1/2 trials that enrolled a total of 54 adult patients with solid tumors harboring NTRK gene fusion (ALKA-372-001, EudraCT: 2012-000148-88, STARTRK-1, ClinicalTria.gov ID: NCT02097810, and STARTRK-2, ClinicalTria.gov ID: NCT02568267) (26). The results of this study highlighted that entrectinib is a safe and active treatment option for patients with NTRK fusion-positive solid tumors.

The efficacy and safety of entrectinib and larotrectinib highlighted the need to routinely test for NTRK fusions (26).

These tumor agnostic therapies may be relevant also in ATC with RET fusions (75).

Sorafenib and lenvatinib have been investigated for use in MTC (81, 82), and a clinical trial was conducted to investigate the use of vandetanib and cabozantinib in RAI-R DTC (51, 83, 84), which led to recent cabozantinib approval as second-line therapy of RAI-R DTC. These studies are summarized in Table 6 .

Several other TKIs approved for the treatment of other tumors by national and international regulatory agencies have been extensively investigated in clinical trials for the treatment of all types of TC (e.g., axitinib, pazopanib, motesanib, sunitinib, vemurafenib, and selumetinib). These trials, previously reviewed by other authors in the last years (41, 85, 86), are summarized in Table 7 .

TKIs can cause liver injury. Currently, the mechanism underlying hepatotoxicity is partially unknown but probably is not related to kinase inhibition in hepatocytes (108). TKI-associated hepatotoxicity may not only be due to the parent drug but also due to metabolites produced from the metabolism of cytochrome P450 3A2 (109, 110).

The onset of TKI-induced hepatotoxicity usually appears within the first 2 months of starting treatment but could be delayed and is usually reversible. Fatality from TKI-induced hepatotoxicity is uncommon but requires diligent surveillance (108).

The first clues of hepatotoxicity occurrence are vague symptoms such as fatigue, anorexia, nausea, discomfort in the right upper quadrant, and dark urine (111). Biochemical markers of liver injury include elevation of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and bilirubin, as well as an alteration in protein synthesis, which is reflected in albumin concentration and time of prothrombin (PT) (108).

Ghatalia et al. (2014) conducted a meta-analysis of randomized controlled trials to determine the RR of hepatotoxicity with VEGFR TKIs (112). High-grade ALT elevation occurred in 168 of 9,930 (1.7%) patients and AST elevation in 159 of 9,986 (1.6%) patients receiving VEGFR TKI. The incidence of high-grade liver failure/dysfunction was 0.7% (112).

In the phase I/II LIBRETTO-001 trial, approximately one-half of treated patients had elevations in transaminases during treatment with selpercatinib; they were G3 or G4 in approximately 10%. The median time to onset was approximately 4 weeks. Elevations in total bilirubin occurred in approximately one-fourth of treated patients; they were G3 or G4 in 2% (113).

In the phase I/II ARROW study (n = 438), pralsetinib-treated patients were reported to have increased AST and ALT levels in 34% and 23% of the cases, respectively (114).

Although the incidence of life-threatening liver failure reported with VEGFR TKIs is quite small, careful monitoring of hepatic function and exclusion of patients with moderate hepatic impairment may be essential in patients receiving VEGFR TKIs. Patients treated with any of these agents should have a baseline evaluation of liver function tests and periodic re-evaluation during therapy.

Gastrointestinal toxicities of TKIs include diarrhea, nausea, vomiting, and, in some cases, pancreatic atrophy.

In clinical trials, nausea of any grade has been reported in 23% to 58% of treated patients, and rates are the highest in patients treated with lenvatinib, cabozantinib, and selpercatinib (45, 49, 113). Vomiting of any grade has been reported in a range of 10% to 48% of treated patients; rates are the lowest with pazopanib and sorafenib and the highest with lenvatinib and cabozantinib (49, 106, 113).

Diarrhea is a common side effect related to TKIs. In clinical trials, diarrhea of any grade has been reported in a range of 30% to 79% of patients treated with TKI (highest rates with vandetanib). Severe forms of diarrhea (G3 or G4) occurred in 3% to 17% (115). Vandetanib acts as a moderately potent EGFR and VEGFR inhibitor; its mechanism of action explains the somewhat higher incidence of diarrhea associated with this agent (83).

In the case of G1–G2, diarrhea may improve with loperamide; if it persists, a coproculture test and antibiotic therapy are useful, followed by hydration and octreotide administration.

If antidiarrheal therapies are not sufficient, a dose reduction or discontinuation of TKI therapy could be necessary. In the case of severe toxicity, it could be necessary to restart TKI therapy at a reduced dose. It is suggested to take the medication with a large meal and plenty of water to reduce side effects (107).

Long-term therapy with sorafenib is associated with pancreatic atrophy. The possibility of pancreatic exocrine insufficiency should be considered in patients treated with sorafenib who develop refractory diarrhea.

TKIs are associated with cardiovascular toxicity, due to the involvement of TK-R in normal cellular homeostasis. Common cardiovascular events are hypertension, reduced ejection fraction, myocardial infarction, and QT prolongation. Clinical drug-by-drug cardiotoxicity incidence of TKIs has been extensively reviewed by Jin et al. (2020) (116).

Incidences of proteinuria of any grade during TKI treatment correlate with VEGFR inhibition; thrombotic microangiopathy and acute interstitial nephritis are common with sorafenib. In the SELECT trial, 10% of lenvatinib-treated patients experienced grade 3 proteinuria. While VEGF is present on glomerular podocytes, VEGFR is present on endothelial cells of the glomerulus. The alteration of VEGF and VEGFR mechanisms leads to capillary endotheliosis and proteinuria. Deprivation of endothelial fenestrations in capillaries, an increase of endothelial cells, and deprivation of podocytes inside the glomerulus are a consequence of depletion of VEGF in podocytes. The lack of interaction between podocytes and endothelium maltreats the filtration barrier, leading to proteinuria.

Monitoring of proteinuria is based on a dipstick test, which should be done before initiation of TKI and then every 14 days during treatment. In case of the absence of proteinuria, a test can be performed every 28 days. When proteinuria is G2 or higher, it is useful to evaluate the protein/creatinine ratio (PCR) using a specimen of urine collected in a universal container early in the morning. An alternative test to PCR is the albumin/creatinine ratio (ACR).

G2 proteinuria matches with PCR of 100–300 mg/mmol or ACR of 70–250 mg/mmol. G3 proteinuria matches with PCR > 300 mg/mmol or ACR > 250 mg/mmol. Treatment of high blood pressure with drugs such as ACE inhibitors or ARB can be useful in the management of proteinuria.

Hand-foot syndrome occurs commonly during therapy with sorafenib or sunitinib ( Figure 2 ). These drugs probably impact the growth of skin cells and capillaries of the hands and feet.

Symptoms include redness, swelling, tenderness, and tightness of the skin. When the syndrome is severe, peeling skin, ulcers, and intense pain leading to the limitation in the use of the hands or walking are common. Generally, hand-foot syndrome appears in the 1st to 6th week of treatment.

Prevention measures include limiting exposure to hot water, avoiding contact with chemicals used in cleaning products, and avoiding activities that cause friction on the hands or feet.

Treatment consists of i) topical anti-inflammatory medications, such as corticosteroid creams; ii) topical anesthetics, which are available as creams or patches to apply on palms or soles in the presence of pain; iii) topical moisturizing and exfoliating creams containing urea; iv) pain relievers.

In addition, when the syndrome is severe enough to affect patients’ quality of life, dose reduction or drug discontinuation is necessary until symptoms of hand-foot syndrome will improve (131).

Sorafenib is associated with hand-foot syndrome in 30% of patients. In case of a G1 toxicity, it is suggested to continue sorafenib and use topical therapy. In case of a G2 toxicity, it is suggested to use topical therapy; if no improvement within 7 days will be seen, treatment should be interrupted until toxicity resolves to grade 0–1. If G2 or G3 toxicity recurs, it can be useful to decrease the sorafenib dose by one dose level; in case of a 4th occurrence, it will be appropriate to discontinue therapy (132).

Fatigue is a common side effect related to TKIs. The etiology is complex. It can depend on cardiac dysfunction, renal dysfunction, or gastrointestinal side effects, such as diarrhea or nausea. It is very frequent in patients treated with motesanib and axitinib (133). Supportive care is necessary and includes adequate nutrition, exercise, and stress-reducing techniques (107).

Thyroid function should be monitored in case of asthenia in patients receiving TKI therapy.

During TKI therapy, thyroid function abnormalities have been observed in athyrotic patients on thyroxine substitution.

The mechanism of worsening hypothyroidism in thyroidectomized patients is still unclear. These drugs can cause an alteration in the transport and metabolism of thyroid hormones. In particular, sunitinib and sorafenib increase the activity of type 3 deiodinase (as evidenced by the decrease in T3/T4 and T3/rT3 ratios) resulting in hypothyroidism because of lower tissue availability of the active hormone T3, locally inactivated in T2 or rT3 (134).

During the treatment with TKIs, it is required to monitor TSH levels monthly and to adjust thyroid replacement medication as needed. If patients present severe chronic asthenia, primary adrenal insufficiency (PAI) should be excluded. A frequent occurrence of PAI has been reported after the first month of treatment with lenvatinib or vandetanib in patients with advanced TC. Importantly, replacement therapy is associated with substantial relief of these major AEs (135).

The mechanisms involved in the development of lenvatinib-induced PAI are unknown. It was hypothesized that the target of lenvatinib (in particular, VEGFRs and PDGFRα) may be implicated in adrenal control (136).