Skip to main content
NCBI home page
The Journal of Clinical Endocrinology and Metabolism logoLink to The Journal of Clinical Endocrinology and Metabolism
. 2023 Dec 7;109(5):1231–1240. doi: 10.1210/clinem/dgad697

Characterizing Genetic Alterations Related to Radioiodine Avidity in Metastatic Thyroid Cancer

Zhuanzhuan Mu 1,2, Xin Zhang 3,4, Di Sun 5,6, Yuqing Sun 7,8, Cong Shi 9,10, Gaoda Ju 11,12,13,14, Zhentian Kai 15, Lisha Huang 16, Libo Chen 17,18, Jun Liang 19,20, Yansong Lin 21,22,✉
PMCID: PMC11031230  PMID: 38060243

Abstract

Context

Patients with differentiated thyroid cancer (DTC) with distant metastasis (DM) are usually not recognized as radioactive iodine (RAI)-refractory DTC in a timely manner. The elucidation of genetic features related to RAI uptake patterns may shed light on the early recognition of RAI-refractory DTC.

Objective

This work aimed to elucidate the underlying molecular features behind different RAI uptake patterns.

Methods

A total of 214 patients with DM-DTC were retrospectively included in the analysis. RAI uptake patterns were defined as initially RAI refractory (I-RAIR) and initially RAI avid (I-RAIA) according to the first post-treatment scan, then I-RAIA was further divided into continually RAIA (C-RAIA), partly RAIR (P-RAIR), and gradually RAIR (G-RAIR) according to subsequent scans. The molecular subtype groups—BRAFV600E mutated, RAS mutated, fusions, and others—were classified according to main driver genes status.

Results

BRAF, TERT promoter, and TP53 mutations are more frequently detected in the I-RAIR pattern while RET fusions and RAS mutations are more frequent in the I-RAIA pattern. A late-hit mutation including TERT, TP53, or PIK3CA is more common in I-RAIR than that in I-RAIA (50.0% vs 26.9%, P = .001), particularly for those with RAS mutations in the I-RAIR group, always accompanied by TERT promoter. Isolated RET fusions accounts for 10% of I-RAIR. When compared among driver gene groups, BRAFV600E-mutated tumors have a higher rate of the I-RAIR pattern (64.4%) than RAS-mutated (4.5%, P < .001) and fusion-positive (20.7%, P < .001) tumors. In I-RAIA subgroups, BRAFV600E-mutated tumors have lower prevalence of the C-RAIA pattern than those with RAS mutation or fusions.

Conclusion

Patients with the I-RAIR pattern predominantly featured mutations of the BRAF and/or TERT promoter, of which RAS mutations were usually accompanied by late-hit mutations, while fusions mostly occurred alone.

Keywords: differentiated thyroid cancer, distant metastases, genomic alteration, RAI uptake pattern

Distant metastasis (DM) is predominantly responsible for mortality in patients with differentiated thyroid cancer (DTC) (1, 2). Considering the fact that DTC is a common malignancy and the large population base in China, patients with DTC with DM (DM-DTC) become a non-negligible cluster. Radioactive iodine (RAI) therapy is generally regarded as a standard first-line treatment modality for patients with DM-DTC postoperatively. However, RAI uptake patterns among such patients may vary a lot from RAI avidity to non-RAI avidity, leading to substantial differences in prognosis. Patients with non-RAI avidity generally have a significantly worse prognosis than those with RAI avidity (3, 4). Currently, the diverse RAI responses have received wide attention, and it is recommended to stop RAI therapy for RAI-unresponsive patients (5). Nevertheless, patients have usually undergone several rounds of RAI therapies before recognition of RAI-refractory lesions, which not only exposes patients to unnecessary side effects of RAI therapy but also delays effective treatment. Therefore, there is an urgent need for an approach to identify patients who will not benefit from RAI therapy at an early stage, with no need to go through tentative treatment.

In recent years, molecular features of DTC have profoundly advanced our understanding of its clinical behavior, including RAI avidity. Among which, BRAFV600E and TERT promoter mutations are wildly explored as the principal oncogene of RAI-refractory DTC, and there is a synergistic effect on loss of RAI avidity (6, 7). Colombo et al (8) further revealed distinct molecular profiles in correlation to diverse RAI uptake patterns in a cohort of 70 patients with papillary thyroid cancer (PTC) with regional or distant metastases. Specifically, BRAFV600E mediated reduced RAI uptake, while RET/NTRK fusions were associated with RAI-avid but persistent disease. Additionally, more frequent mutations including genes encoding SWI/SNF chromatin remodeling complex and PI3K/AKT/mTOR pathway were found in non-RAI-avid tumors than that in RAI-avid disease (9). In 2014, the Cancer Genome Atlas proposed 2 types of BRAFV600E-like and RAS-like tumors based on gene expression profile. Soe et al (10) reported a higher prevalence of non-RAI-avid disease in BRAFV600E-like cancers compared with RAS-like cancers in a cohort of 78 patients with DM-DTC. In turn, findings from the above studies enlighten us to explore on a more comprehensive and detailed level. However, limited by inconsistent purposes and patients in the aforementioned studies, as well as the relatively small sample size, the relationship between mutational information, particularly BRAFV600E and TERT promoter, and RAI uptake patterns in a large cohort of patients with DM-DTC remains to be clarified. Moreover, many studies analyzed roughly on the basis of RAI-responsive or not, and the latter always displays substantial heterogeneity in RAI uptake patterns, causing a lot of information to be hidden.

Herein, we retrospectively analyzed the molecular information associated with different RAI uptake patterns among patients with DM-DTC in our institute. The aim of this study is to elucidate the underlying molecular discrepancy behind diverse RAI uptake patterns. According to the prior classification of BRAFV600E-like and RAS-like types, and given the remarkable discrepancy on RAI avidity of BRAFV600E-mutated and RET/NTRK-fusion tumors identified, we also divided patients into 4 categories according to driver oncogenes to clarify the diversity of RAI uptake patterns.

Materials and Methods

Patients Cohort

We retrospectively reviewed patients with DM-DTC attending in Peking Union Medical College Hospital between 2020 and 2022. Clinical, pathologic, and radiographic data were obtained from existing medical records. Patients had undergone thyroid surgery and RAI therapy at least once. DMs were diagnosed with either pathologic examination of surgical or fine needle aspiration specimens or evaluation according to biochemical and radiographic examinations during the follow-up period. PTC, follicular thyroid cancer (FTC), or poorly differentiated thyroid cancer (PDTC) were allowed, but other histological types were excluded. A total of 220 patients aged 18 years or above were included. After approval of the institutional review board (JS-2432) and signing written informed consent, patients in this study underwent molecular testing. A total of 281 specimens of primary thyroid tumors (n = 162) or metastatic lesions (n = 106, 13 specimens could not be identified as primary tumor or metastatic lymph node) were collected for molecular testing. Fifty patients had specimens from multiple sites; variants from different sites in the same patient were merged for the analysis.

Definition of RAI Uptake Patterns

Each patient underwent RAI therapy for the purpose of metastases treatment and a post-treatment scan was conducted. An iodine-restricted diet followed for approximately 2 weeks prior to RAI administration. The interval between the post-treatment scan and RAI administration was 5 to 7 days. According to the post-treatment scan reports, RAI uptake patterns of metastatic lesions were classified as initially RAIR (I-RAIR, lesions never uptake RAI) and initially RAIA (I-RAIA, lesions uptake RAI in the initial therapy). I-RAIA was further categorized into continually RAI-avid (C-RAIA, lesions uptake RAI continually), partly RAIR (P-RAIR, some lesions uptake RAI while others do not), and gradually RAIR (G-RAIR, lesions lose the ability to uptake RAI after previous evidence of RAI uptake). Examples of RAI uptake patterns are shown in Fig. 1. Patients with multiple metastases with mixed patterns were judged on the dominant pattern that appeared for further classification. Due to unknown results of post-treatment scans, RAI uptake patterns of 6 patients could not be determined.

Figure 1.

Figure 1.

Open in a new tab

Example images of 4 radioactive iodine (RAI) uptake patterns. Initially RAI refractory (I-RAIR) pattern (A). The patient's lung metastases did not take up RAI during the first treatment (A1); to eliminate interference from residual thyroid, a second RAI treatment was performed, but the lung metastases still did not uptake RAI (A2). Continually RAI avid (C-RAIA) pattern (B). The patient's lung metastases showed significant RAI avidity in 2 RAI treatments after successful residual thyroid ablation (B1, B2). Gradually RAI refractory (G-RAIR) pattern (C). The patient's lung metastases showed RAI avidity in the first RAI treatment after successful residual thyroid ablation (C1), but no RAI uptake was observed in subsequent treatment while structural lesions still existed (C2). Partly RAI refractory (P-RAIR) pattern (D). The patient's multiple bone metastases showed RAI uptake (D1, D2), but diffuse miliary metastases in lungs did not show RAI uptake (D1, D3).

Next-Generation Sequencing and Data Analysis

The ThyroLead panel (Topgen, China) was designed to cover the exonic region for 18 genes (HRAS, KRAS, NRAS, CDC73, CDKN1B, DICER1, IDH1, MEN1, MTOR, PIK3CA, PTEN, TP53, TSHR, CTNNB1, GNAS, PAX8, AKT1, EIF1AX), a 1000-bp region in the promoter region of TERT, and both exonic and intronic regions for 7 frequently rearranged genes (BRAF, RET, NTRK1/2/3, ALK, and PPARG). The term “mutational burden” describes the total number of genetic alterations detected in each patient that are limited to the ThyroLead panel. The details of DNA isolation, sequence library preparation, and sequencing were reported in a previous study (11). Adapter sequences and low-quality reads (ie, reads with >40% of bases failed Q25; reads shorter than 70 bp; and reads with low complexity) were removed by Fastp software. The work including sequences aligning to the reference human genome hg19, duplicates flagging, indel realignment and variant calling, and variant annotation was conducted as previously described (11). Each variant was graded according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG-AMP) guidelines and deemed to be a valid genetic alteration only when it was graded as pathogenic or likely pathogenic. Manual inspection was performed to ensure the reliability of variant calling and pathogenicity grading. Samples with uneven depth of coverage, low deduplicated depth, or aberrant insert sizes were removed and only samples that passed quality control were used for analysis.

Groups based on Driver Gene Alterations

BRAFV600E-like and RAS-like types have frequently been applied to describe the discrepancy in tumor characteristics since 2014. Genetic variations of RET/NTRK fusions have been regularly classified into the BRAFV600E-like genotype. However, the different RAI uptake patterns of RET/NTRK fusions from BRAFV600E reported in prior studies and observed in our institute suggest irrationality of the classification above. Hence, in this study, we divided patients into 4 molecular subtype groups according to driver gene mutation status. Specifically, the BRAFV600E-mutated type indicated BRAFV600E as the only early driver mutation event, the RAS-mutated type included HRAS, KRAS, and NRAS mutations, and fusions presented as RET or NTRK fusions. There were 4 cases carrying both BRAFV600E and RAS mutations whose molecular subtype group could not be determined. The remaining cases who did not meet the conditions above were grouped as “others,” including 29 with no variants detected, and 16 with detected variants other than BRAFV600E, RAS mutations, or RET/NTRK fusions.

Statistical Analysis

Data analyses were performed using SPSS Statistics (version 27.0). Demographics and clinicopathologic characteristics of patients were described using frequencies with percentages, means with SD, or medians with interquartile range as appropriate. Mann–Whitney U, Kruskal–Wallis, chi-square, and Fisher's exact test were performed to compare differences in clinicopathologic characteristics, genetic variations, and driver gene categories among RAI uptake patterns as appropriate. A 2-tailed P < .05 was considered to be statistically significant. Multiple comparisons among RAI uptake patterns were corrected using Bonferroni adjustment.

Results

Demographics and Clinicopathologic Characteristics

Of 220 adult patients with DM-DTC, 6 patients missed their RAI scan results with unknown RAI uptake pattern. We compared the clinical characteristics in terms of RAI uptake patterns initially (I-RAIR vs I-RAIA) and driver mutations (BRAFV600E mutated vs RAS mutated vs fusions vs others) and the details are shown in Table 1. Typically, patients with initially non-RAI-avid disease were older at DTC diagnosis than those with initially RAI-avid disease (P = .004). Besides PDTC, which is unsurprising, PTC accounts for more in the I-RAIR group than in the I-RAIA group. Patients with I-RAIR tend to have a higher mutational burden than I-RAIA patients (P < .001).

Table 1.

Demographics and clinicopathologic characteristics of the cohort

Characteristics Initial RAI uptake patterns Driver mutation groups
I-RAIR (n = 80) I-RAIA (n = 134) P value BRAFV600E mutated (n = 89) RAS mutated (n = 22) a Fusions (n = 60) Others (n = 45) P value
Age at diagnosis (years), median (IQR) 45.76 (34.66, 56.82) 36.79 (29.95, 49.00) .004 46.67 (35.76, 56.35) 55.32 (44.81, 62.93) 30.72 (25.92, 40.31) 36.62 (29.77, 46.94) <.001
Male, n (%) 34 (42.5%) 54 (40.3%) .752 42 (47.2%) 5 (22.7%) 23 (38.3%) 19 (42.2%) .200
Histology, n (%) .021 <.001
PTC 73 (91.3%) 114 (85.1%) 88 (98.9%) 8 (36.4%) 59 (98.3%) 34 (75.6%)
FTC 1 (1.3%) 16 (11.9%) 0 (0.0%) 12 (54.5%) 0 (0.0%) 5 (11.1%)
PTC and FTC 2 (2.5%) 1 (0.7%) 0 (0.0%) 2 (9.1%) 0 (0.0%) 1 (2.2%)
PDTC 4 (5.0%) 3 (2.2%) 1 (1.1%) 0 (0.0%) 1 (1.7%) 5 (11.1%)
T stage, n (%) .094 .247
T1-3a 35 (45.5%) 67 (57.8%) 37 (45.1%) 9 (60.0%) 33 (56.9%) 25 (62.5%)
T3b-4b 42 (54.5%) 49 (42.2%) 45 (54.9%) 6 (40.0%) 25 (43.1%) 15 (37.5%)
N stage, n (%) .018 <.001
0 2 (2.7%) 19 (15.7%) 2 (2.5%) 13 (76.5%) 1 (1.7%) 6 (14.3%)
N1a 6 (8.1%) 8 (6.6%) 8 (9.9%) 2 (11.8%) 1 (1.7%) 4 (9.5%)
N1b 66 (89.2%) 94 (77.7%) 71 (87.7%) 2 (11.8%) 57 (96.6%) 32 (76.2%)
Distant metastatic sites, n (%) .183 <.001
Lung only 58 (72.5%) 101 (75.4%) 67 (75.3%) 7 (31.8%) 55 (91.7%) 30 (66.7%)
Bone +/− Lung 8 (10.0%) 20 (14.9%) 7 (7.9%) 14 (63.6%) 2 (3.3%) 5 (11.1%)
Beyond lung and bone 14 (17.5%) 13 (9.7%) 15 (16.9%) 1 (4.5%) 3 (5.0%) 10 (22.2%)
Mutational burden, median (IQR) 2 (1, 2) 1 (1, 2) <.001 2 (1, 2) 1.5 (1, 2.25) 1 (1, 1) 0 (0, 1) <.001
Time to RAI refractoriness (months), median (IQR) 10.10 (6.37, 44.32) 38.43 (15.43, 92.33) <.001 12.23 (7.60, 72.50) 37.20 (17.73, 55.30) 10.47 (7.43, 22.10) 19.43 (8.87, 92.33) .119
Open in a new tab

Abbreviations: FTC, follicular thyroid cancer; I-RAIR, initially radioactive iodine refractory; I-RAIA, initially RAI avid; IQR, interquartile range; N, lymph node; PTC, papillary thyroid cancer; PDTC, poorly differentiated thyroid cancer; T, tumor.

a Including 54 RET fusions and 6 NTRK1 fusions.

Regarding the difference among the driver mutation groups, patients with gene fusions (54 RET fusions and 6 NTRK1 fusion) were found to be younger than those with BRAFV600E or RAS mutations in our cohort (P < .001). All FTCs were identified with RAS mutations. Tumors driven by RAS mutations were less likely to have lymph node metastases, while BRAFV600E and fusions were more likely to have metastatic lymph nodes and usually as N1b. RAS-mutated tumors commonly had bone involvement, while BRAFV600E and fusions generally involved the lung, and gene fusions rarely developed extrapulmonary metastases. Additionally, compared with patients with fusions, those with BRAFV600E and RAS mutations often co-occurred with other alterations (P < .001). Specifically, the proportions of co-occurring with other alterations were 58.4%, 50%, and 18.3% in patients with BRAFV600E, RAS mutations, and fusions, respectively. In terms of time to RAI refractoriness, patients with fusions tend to be the shortest with a median of 10.47 (7.43, 22.10) months, followed by BRAFV600E mutants (12.23 [7.60, 72.50] months), with RAS mutants the longest (37.20 [17.73, 55.30] months), though no statistically significant difference was observed.

The Correlations Between Genetic Variations and RAI Uptake Patterns

The differences with respect to the RAI uptake patterns based on independent genotypes are shown in Fig. 2. Patients with BRAF, TERT promoter, and TP53 mutations are more likely to manifest as initially non-RAI-avid pattern than those with corresponding wild genotype (P < .001, P = .003, P = .031, respectively). RET fusions and RAS mutations were more frequently identified in patients with initially RAI-avid disease (P = .005, P = .004, respectively). Patients with the I-RAIA pattern were further classified into C-RAIA, P-RAIR, and G-RAIR (Fig. 2B), showing that BRAF and TERT promoter mutations still tend to mediate negative RAI uptake as predominating in P-RAIR or G-RAIR compared C-RAIA. RET fusions show a stronger C-RAIA pattern but there were no statistical differences among the 3 subgroups when comparing pairs. The mutational rates of RAS and TP53 did not differ in these subgroups. The distributions of other genotypes were not statistically different among RAI uptake patterns. The panorama of RAI uptake patterns based on mutational data is displayed in Fig. 3.

Figure 2.

Figure 2.

Open in a new tab

Mutational rate comparison between I-RAIR and I-RAIA (A) and among 3 subgroups of I-RAIA (B). I-RAIR, initially RAI refractory; I-RAIA, initially RAI avid; C-RAIA, continually RAI avid; P-RAIR, partly RAI refractory; G-RAIR, gradually RAI refractory.

Figure 3.

Figure 3.

Open in a new tab

Chord diagram of gene alterations and radioactive iodine (RAI) uptake patterns. I-RAIR, initially RAI refractory; C-RAIA, continually RAI avid; P-RAIR, partly RAI refractory; G-RAIR, gradually RAI refractory.

Figure 4 details the genetic profile of 2 different RAI uptake patterns based on the genetic alterations with mutational rate >5%. The co-occurrence of BRAF and TERT promoter mutations have the maximum proportion in 80 patients with I-RAIR, followed by isolated BRAF mutations. RET fusions make up a non-negligible part of I-RAIR, especially when occurring exclusively (10%, 8/80). Notably, a late-hit mutation in TERT, TP53, or PIK3CA is more common in initially non-RAI-avid tumors than initially RAI-avid tumors (50.0% vs 26.9%, P = .001), particularly for those with RAS mutations that are always accompanied by TERT promoter mutations and partially overlaid with TP53 or PIK3CA mutations in the I-RAIR group.

Figure 4.

Figure 4.

Open in a new tab

The main genetic alterations of patients with initially radioactive iodine refractory (I-RAIR) pattern (A) and initially radioactive iodine avid (I-RAIA) pattern (B).

Driver Gene Groups and RAI Uptake Patterns

Considering the influence of different genetic backgrounds on tumor genesis, patients were further categorized as BRAFV600E mutated, RAS mutated, fusions (including RET, NTRK), and others. The incidence of the I-RAIR pattern was 64.4% (56/87) in BRAFV600E-mutated tumors, higher than that in RAS mutants (4.5%, 1/22; P < .001), fusion-positive (20.7%, 12/58; P < .001) tumors, and others (20.9%, 9/43; P < .001). No statistical difference was found among RAS-mutated, fusion-positive tumors, and others. When I-RAIA was further classified into 3 patterns, the proportions of C-RAIA, P-RAIR, and G-RAIR were 48.4% (15/31), 12.9% (4/31), and 38.7% (12/31) in BRAFV600E-mutated tumors, 71.4% (15/21), 19.0% (4/21), and 9.5% (2/21) in RAS-mutated tumors, 87.0% (40/46), 4.3% (2/46), and 8.7% (4/46) in fusion-positive tumors and 79.4% (27/34), 11.8% (4/34), and 8.8% (3/34) in others, respectively. The prevalence of C-RAIA pattern was evidently lower in BRAFV600E-mutated tumors than that in other 3 groups (Fig. 5).

Figure 5.

Figure 5.

Open in a new tab

The percentages of radioactive iodine (RAI) uptake patterns in different driver gene groups. The blue column regardless of filling shape represents the proportion of I-RAIA, which is the sum of C-RAIA, P-RAIR and G-RAIR. I-RAIR, initially RAI refractory; I-RAIA, initially RA avid; C-RAIA, continually RAI avid; P-RAIR, partly RAI refractory; G-RAIR, gradually RAI refractory.

Discussion

The current recognition of RAI-refractory DTC is mainly based on the RAI uptake patterns of tumors rather than histopathological evidence, which gives rise to significant heterogeneity in RAI-refractory DTCs. Different RAI uptake patterns may be indicative of various molecular backgrounds and underlying mechanisms. In this study, we set our sights on RAI uptake patterns of lesions, and investigated the molecular backgrounds corresponding to different RAI uptake patterns in an adult cohort with DM-DTC.

BRAF and TERT promoter mutations unsurprisingly show a predominate role in mediating the loss of RAI avidity in this study, which further contributes to supporting the results of our previous research (6, 12, 13). The majority of our patients with BRAF (61.1%) or TERT promoter mutations (50.7%) manifested as the I-RAIR pattern. Of note, even being observed as RAI-avid initially, patients with BRAF or TERT promoter mutations generally exhibited poorer RAI uptake ability over subsequent RAI treatments than those with RAS mutations or RET fusions. It has been revealed that the mitogen-activated protein kinase (MAPK) signaling pathway flux impacts inversely on the expression of genes in charge of iodide uptake and organification in thyrocytes (14). Activating mutations in BRAF generally yield a high MAPK signaling output because it constitutively signals as a functional monomer and is unresponsive to negative feedback by ERK, resulting in a significantly decreased ability to concentrate iodine (15, 16). A recent study, through a comprehensive genomic and transcriptomic investigation, also reported that patients unresponsive to RAI treatment were more enriched with BRAFV600E than RAI responders (17). Additionally, it further revealed that 1q-gain, mutations of genes regulating mRNA splicing, and the PI3K pathway were associated with unresponsiveness to RAI. TERT promoter mutations were proposed to provide de novo consensus motifs for the ETS family of transcription factors triggered by MAPK signaling, thereby assisting in boosting MAPK signaling output (18). Recent studies discovered that telomerase upregulation (including TERT promoter mutations and TERT overexpression) can promote the dedifferentiation and progression of thyroid cancer by regulating ribosomal RNA expression and protein synthesis, as well as triggering nontelomeric effects such as cell-autonomous and microenvironment-related consequences, which mechanistically explain the poor RAI avidity of tumors with TERT promoter mutations (19, 20).

TP53 mutations are widely reported to be enriched in advanced metastatic PTC, PDTC, and anaplastic thyroid cancer (ATC), and were even found to be the most frequent genetic alterations in several ATC cohorts (21-24). Nikitski et al (25) declared that TP53-mutated follicular adenomas may represent a precursor for the dedifferentiation of thyroid cancer, with the finding that TP53-mutated well-differentiated cancers frequently included scattered single neoplastic cells with abnormal nuclei resembling those making up ATC. All the above seem to suggest that TP53 mutations track dedifferentiation of thyroid cancer. In our study, TP53-mutated tumors are found more likely to present with the I-RAIR pattern, which provides additional clinical evidence for the crucial role of TP53 in the dedifferentiation process.

RAS mutations were noticed to be more frequent in patients with the I-RAIA pattern in our study. It has been demonstrated that activating mutations in RAS result in a modest MAPK signaling flux as mutant RAS proteins signaling through RAF dimers, which responds to negative feedback by ERK (16). Additionally, RAS mutations preferentially activate the PI3K–AKT pathway, which rarely downregulates thyroid genes (NIS, TSHR, etc.) but upregulates tumor-promoting genes (vascular endothelial growth factor A (VEGFA), MET, etc.) (26). Therefore, the ability to concentrate iodine is comparatively preserved in tumors with RAS mutations. Interestingly, we found that in I-RAIR tumors, RAS mutations always accompanied late-hit mutation events, generally the TERT promoter. Late-hit mutation events are regarded as an important contributor in the progression of thyroid cancer (27-30). In our study, the late-hit mutations (any of TERT, TP53, or PIK3CA mutations) were found to have a significant negative correlation with RAI avidity. TERT promoter was the most frequent late mutational event, followed by TP53. Unlike the mutually exclusive occurrence of driver mutations such as BRAF and RAS in thyroid cancer, the late-hit mutation events could overlap and were more common in non-RAI-avid tumors. For instance, in our DM-DTC cohort, 44.4% (data not presented in results) of patients with TP53 mutations were accompanied by TERT promoter mutations in the I-RAIR group and 20.0% in I-RAIA group. It has been widely demonstrated that a progressive accumulation of mutations of the TERT promoter, TP53, and PI3K/AKT/mTOR pathway effectors was associated with highly aggressive tumors (28, 29, 31, 32). Thus, we have further demonstrated the crucial role of late-hit mutations in the impairment of iodine uptake capability, particularly in RAS-mutated tumors.

Fusions-positive tumors were found tend to present as RAI-avid patterns, especially continually RAI-avid, different from the BRAFV600E-mutated type. Several previous studies have classified fusions (mainly RET fusions) as the BRAFV600E-like type (10, 33, 34), which is obviously inappropriate regarding RAI avidity in that BRAFV600E-mutated tumors are inclined to be non-RAI-avid whereas fusion-positive tumors tend to be RAI-avid. Herein, we have proposed fusions as an independent genotype in regard to RAI avidity. Regarding other clinicopathologic characteristics, on the one hand, fusion-driven tumors are more prone to be PTCs, commonly with lymph node metastases, which is similar to BRAFV600E mutants, implying both of them potentially share the same tumor development pathway—MAPK signaling pathway. On the other hand, patients with fusions are relatively younger than those with BRAFV600E and RAS mutations, their distant metastatic sites are frequently confined to the lung, and the lesions have persistent RAI avidity. All of the above suggest that the molecular mechanism involved in fusions is not entirely the same as BRAFV600E, and is different from RAS. Unexpectedly, though, fusion-positive patients tend to present as RAI avid, RET fusions alone are also detected in the I-RAIR pattern and accounts for 10% of that, unlike RAS mutations that are always accompanied by late-hit mutations in the I-RAIR pattern. Meanwhile, patients with RET fusions have a relatively shorter time to RAI refractoriness. All these findings suggest different mechanisms of mediating iodine uptake from RAS mutations. Furthermore, we noted that fusion-positive tumors had a lower mutational burden than BRAFV600E-mutated and RAS-mutated tumors in our cohorts. Yip et al (34) also found smaller proportions of late-hit mutations in tumors with RET and NTRK fusions compared with those with BRAFV600E and RAS mutations in a DM-DTC cohort. Thereby, we supposed that fusions alone are strong enough to induce aggressive features, such as distant metastasis and RAI refractoriness. In patients with such a molecular basis more attention should be paid to the efficacy evaluation of RAI treatment, since good RAI uptake ability in the majority of patients may conceal underlying aggressiveness and trick us into repeating RAI treatment, with increasing the risk of adverse effects but faint efficacy. Apparently, there is a heterogeneity of tumor behavior in fusions due to the part presenting with the I-RAIR pattern, which may implicate stronger underlying tumor virulence. With the exploration of redifferentiation for RAI-refractory thyroid cancer, inhibitors that target oncogenic fusion genes have also been reported as having the capacity to promote redifferentiation, especially restoring RAI uptake (35). Accordingly, combination of fusion gene inhibitors and RAI therapy could be a promising strategy and has been explored in some cases (35-38). In view of the current research, the redifferentiation effect of inhibitors on RAI uptake has been adequately confirmed, but the benefit of RAI treatment is still insufficient due to the lack of a control group and small sample size.

The retrospective nature of this study constrains a definitive causal link between RAI uptake patterns and molecular alteration. However, real-world information on RAI uptake can be provided, and the correlation between several main genetic alterations and RAI avidity obtained in this cohort are consistent with previous reports (7-10, 39, 40). A limited set of cancer genes may lead to underestimation of mutational burden and omission of interactions of specific rare gene defects, but it ensures identifying clinically common and meaningful genetic variations, like RET fusions, TP53, etc., and uncovering findings with strong universality. Moreover, we specifically discriminated fusions from the BRAFV600E-like subtype, which has been proven to be a reasonable approach. The unexpected RAI uptake patterns noted in a small subset of patients with specific genetic defects differed from that observed in the majority with the same genetic variations, such as the I-RAIR pattern found in patients with RET fusions, suggesting the possibility of other mechanisms involved that require further prospective investigation and functional validation. Regardless, this study provides valuable insights into molecular indicators of RAI avidity and lends support to individual RAI treatment recommendations.

In conclusion, our research shows the underlying distinct mutational information among different RAI uptake patterns. BRAF and TERT promoter mutations predominate in patients with the I-RAIR pattern. RAS-mutated and fusion-positive tumors generally preserve the ability to uptake RAI, but a small portion of cases exhibit a non-RAI-avid pattern, with RAS mutations involving late-hit mutations while fusions mostly work alone. The findings suggest that patients with RAI-refractory DTC should not be regarded homogeneous, which again warns us that the management of patients with RAI-refractory DTC should be based on refined assessment and precise care, and a “one size fits all” approach is not appropriate. The insight of molecular information may guide a tailored treatment strategy.

Abbreviations

ATC

anaplastic thyroid cancer

C-RAIA

continually radioactive iodine avid

DM

distant metastasis

DTC

differentiated thyroid cancer

FTC

follicular thyroid cancer

G-RAIR

gradually radioactive iodine refractory

I-RAIA

initially radioactive iodine avid

I-RAIR

initially radioactive iodine refractory

MAPK

mitogen-activated protein kinase

PDTC

poorly differentiated thyroid cancer

P-RAIR

partly radioactive iodine refractory

PTC

papillary thyroid cancer

RAI

radioactive iodine

Contributor Information

Zhuanzhuan Mu, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Xin Zhang, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Di Sun, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Yuqing Sun, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Cong Shi, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Gaoda Ju, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China; Department of Medical Oncology, Key Laboratory of Carcinogenesis & Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital and Institute, Beijing, 100142, China; Department of Oncology, Peking University International Hospital, Peking University, Beijing, 102206, China.

Zhentian Kai, Department of Bioinformatics, Zhejiang Shaoxing Topgen Biomedical Technology Co., Ltd, Shanghai, 201321, China.

Lisha Huang, Department of Medicine, Zhejiang Shaoxing Topgen Biomedical Technology Co., Ltd, Shanghai, 201321, China.

Libo Chen, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Jun Liang, Department of Medical Oncology, Key Laboratory of Carcinogenesis & Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital and Institute, Beijing, 100142, China; Department of Oncology, Peking University International Hospital, Peking University, Beijing, 102206, China.

Yansong Lin, Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College (PUMC) Hospital, Chinese Academy of Medical Sciences & PUMC, Beijing, 100730, China; Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Beijing, 100730, China.

Funding

This work was supported by Project on Inter-Governmental International Scientific and Technological Innovation Cooperation in National Key Projects of Research and Development Plan (no. 2019YFE0106400), National High Level Hospital Clinical Research Funding (no. 2022-PUMCH-B-072) and National Natural Science Foundation of China (no. 81771875).

Author Contributions

Y.S.L. conceived the study and participated in manuscript editing. Z.Z.M. conducted study design, collected and analyzed data, and wrote the manuscript. X.Z., D.S., Y.Q.S., C.S., and G.D.J. participated in data collection. Z.T.K. and L.S.H. participated in data analysis. All authors participated in data interpretation, reviewed, and approved the current manuscript.

Disclosures

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Data availability

Restrictions apply to the availability of some or all data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided. NGS data in this study can be accessed at https://ngdc.cncb.ac.cn/via accession number HRA004166 from the corresponding author on reasonable request.

References

  • 1. Sciuto  R, Romano  L, Rea  S, Marandino  F, Sperduti  I, Maini  CL. Natural history and clinical outcome of differentiated thyroid carcinoma: a retrospective analysis of 1503 patients treated at a single institution. Ann Oncol. 2009;20(10):1728‐1735. [DOI] [PubMed] [Google Scholar]
  • 2. Nixon  IJ, Whitcher  MM, Palmer  FL, et al.  The impact of distant metastases at presentation on prognosis in patients with differentiated carcinoma of the thyroid gland. Thyroid. 2012;22(9):884‐889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Sabra  MM, Ghossein  R, Tuttle  RM. Time course and predictors of structural disease progression in pulmonary metastases arising from follicular cell-derived thyroid cancer. Thyroid. 2016;26(4):518‐524. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Nunes  KS, Matos  LL, Cavalheiro  BG, et al.  Risk factors associated with disease-specific mortality in papillary thyroid cancer patients with distant metastases. Endocrine. 2022;75(3):814‐822. [DOI] [PubMed] [Google Scholar]
  • 5. Haugen  BR, Alexander  EK, Bible  KC, et al.  2015 American thyroid association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American thyroid association guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid. 2016;26(1):1‐133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Yang  X, Li  J, Li  X, et al.  TERT promoter mutation predicts radioiodine-refractory character in distant metastatic differentiated thyroid cancer. J Nucl Med. 2017;58(2):258‐265. [DOI] [PubMed] [Google Scholar]
  • 7. Liu  J, Liu  R, Shen  X, Zhu  G, Li  B, Xing  M. The genetic duet of BRAF V600E and TERT promoter mutations robustly predicts loss of radioiodine avidity in recurrent papillary thyroid cancer. J Nucl Med. 2020;61(2):177‐182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Colombo  C, Minna  E, Gargiuli  C, et al.  The molecular and gene/miRNA expression profiles of radioiodine resistant papillary thyroid cancer. J Exp Clin Cancer Res. 2020;39(1):245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Song  E, Song  DE, Ahn  J, et al.  Genetic profile of advanced thyroid cancers in relation to distant metastasis. Endocr Relat Cancer. 2020;27(5):285‐293. [DOI] [PubMed] [Google Scholar]
  • 10. Soe  MH, Chiang  JM, Flavell  RR, et al.  Non-Iodine-Avid disease is highly prevalent in distant metastatic differentiated thyroid cancer with papillary histology. J Clin Endocrinol Metab. 2022;107(8):e3206‐e3216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Tan  LC, Liu  WL, Zhu  XL, et al.  Next-Generation sequencing enhances the diagnosis efficiency in thyroid nodules. Front Oncol. 2021;11:677892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Yang  K, Wang  H, Liang  Z, Liang  J, Li  F, Lin  Y. BRAFV600E mutation associated with non-radioiodine-avid status in distant metastatic papillary thyroid carcinoma. Clin Nucl Med. 2014;39(8):675‐679. [DOI] [PubMed] [Google Scholar]
  • 13. Mu  ZZ, Zhang  YQ, Sun  D, Lu  T, Lin  YS. Effect of BRAF(V600E) and TERT promoter mutations on thyroglobulin response in patients with distant-metastatic differentiated thyroid cancer. Endocr Pract. 2022;28(3):265‐270. [DOI] [PubMed] [Google Scholar]
  • 14. Nagarajah  J, Le  M, Knauf  JA, et al.  Sustained ERK inhibition maximizes responses of BrafV600E thyroid cancers to radioiodine. J Clin Invest. 2016;126(11):4119‐4124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Durante  C, Puxeddu  E, Ferretti  E, et al.  BRAF mutations in papillary thyroid carcinomas inhibit genes involved in iodine metabolism. J Clin Endocrinol Metab. 2007;92(7):2840‐2843. [DOI] [PubMed] [Google Scholar]
  • 16. Lito  P, Pratilas  CA, Joseph  EW, et al.  Relief of profound feedback inhibition of mitogenic signaling by RAF inhibitors attenuates their activity in BRAFV600E melanomas. Cancer Cell. 2012;22(5):668‐682. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Boucai  L, Saqcena  M, Kuo  F, et al.  Genomic and transcriptomic characteristics of metastatic thyroid cancers with exceptional responses to radioactive iodine therapy. Clin Cancer Res. 2023;29(8):1620‐1630. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Bell  RJA, Rube  HT, Kreig  A, et al.  The transcription factor GABP selectively binds and activates the mutant TERT promoter in cancer. Science. 2015;348(6238):1036‐1039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Yu  P, Qu  N, Zhu  R, et al.  TERT accelerates BRAF mutant-induced thyroid cancer dedifferentiation and progression by regulating ribosome biogenesis. Sci Adv. 2023;9(35):eadg7125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Landa  I, Thornton  CE, Xu  B, et al.  Telomerase upregulation induces progression of mouse BrafV600E-driven thyroid cancers and triggers non-telomeric effects. Mol Cancer Res. 2023;21:1163‐1175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Pita  JM, Figueiredo  IF, Moura  MM, Leite  V, Cavaco  BM. Cell cycle deregulation and TP53 and RAS mutations are major events in poorly differentiated and undifferentiated thyroid carcinomas. J Clin Endocrinol Metab. 2014;99(3):E497‐E507. [DOI] [PubMed] [Google Scholar]
  • 22. Kuhn  E, Ragazzi  M, Ciarrocchi  A, et al.  Angiosarcoma and anaplastic carcinoma of the thyroid are two distinct entities: a morphologic, immunohistochemical, and genetic study. Mod Pathol. 2019;32(6):787‐798. [DOI] [PubMed] [Google Scholar]
  • 23. Yoo  SK, Song  YS, Lee  EK, et al.  Integrative analysis of genomic and transcriptomic characteristics associated with progression of aggressive thyroid cancer. Nat Commun. 2019;10(1):2764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Ma  LX, Espin-Garcia  O, Bedard  PL, et al.  Clinical application of next-generation sequencing in advanced thyroid cancers. Thyroid. 2022;32(6):657‐666. [DOI] [PubMed] [Google Scholar]
  • 25. Nikitski  AV, Nikiforova  MN, Yip  L, Karslioglu-French  E, Carty  SE, Nikiforov  YE. Can TP53-mutant follicular adenoma be a precursor of anaplastic thyroid carcinoma?  Endocr Relat Cancer. 2021;28(9):621‐630. [DOI] [PubMed] [Google Scholar]
  • 26. Xing  M. Molecular pathogenesis and mechanisms of thyroid cancer. Nat Rev Cancer. 2013;13(3):184‐199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Cancer Genome Atlas Research Network . Integrated genomic characterization of papillary thyroid carcinoma. Cell. 2014;159(3):676‐690. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Landa  I, Ibrahimpasic  T, Boucai  L, et al.  Genomic and transcriptomic hallmarks of poorly differentiated and anaplastic thyroid cancers. J Clin Invest. 2016;126(3):1052‐1066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Latteyer  S, Tiedje  V, Konig  K, et al.  Targeted next-generation sequencing for TP53, RAS, BRAF, ALK and NF1 mutations in anaplastic thyroid cancer. Endocrine. 2016;54(3):733‐741. [DOI] [PubMed] [Google Scholar]
  • 30. Bikas  A, Ahmadi  S, Pappa  T, et al.  Additional oncogenic alterations in RAS-driven differentiated thyroid cancers associate with worse clinicopathologic outcomes. Clin Cancer Res. 2023;29(14):2678‐2685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Ricarte-Filho  JC, Ryder  M, Chitale  DA, et al.  Mutational profile of advanced primary and metastatic radioactive iodine-refractory thyroid cancers reveals distinct pathogenetic roles for BRAF, PIK3CA, and AKT1. Cancer Res. 2009;69(11):4885‐4893. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Landa  I, Ganly  I, Chan  TA, et al.  Frequent somatic TERT promoter mutations in thyroid cancer: higher prevalence in advanced forms of the disease. J Clin Endocrinol Metab. 2013;98(9):E1562‐E1566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Steward  DL, Carty  SE, Sippel  RS, et al.  Performance of a multigene genomic classifier in thyroid nodules with indeterminate cytology: a prospective blinded multicenter study. JAMA Oncol. 2019;5(2):204‐212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Yip  L, Gooding  WE, Nikitski  A, et al.  Risk assessment for distant metastasis in differentiated thyroid cancer using molecular profiling: a matched case-control study. Cancer. 2021;127(11):1779‐1787. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Lee  YA, Lee  H, Im  SW, et al.  NTRK and RET fusion-directed therapy in pediatric thyroid cancer yields a tumor response and radioiodine uptake. J Clin Invest. 2021;131(18):e144847. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Groussin  L, Clerc  J, Huillard  O. Larotrectinib-Enhanced radioactive iodine uptake in advanced thyroid cancer. N Engl J Med. 2020;383(17):1686‐1687. [DOI] [PubMed] [Google Scholar]
  • 37. Groussin  L, Bessiene  L, Arrondeau  J, et al.  Letter to the editor: selpercatinib-enhanced radioiodine uptake in RET-rearranged thyroid cancer. Thyroid. 2021;31(10):1603‐1604. [DOI] [PubMed] [Google Scholar]
  • 38. Werner  RA, Sayehli  C, Hanscheid  H, et al.  Successful combination of selpercatinib and radioiodine after pretherapeutic dose estimation in RET-altered thyroid carcinoma. Eur J Nucl Med Mol Imaging. 2023;50(6):1833‐1834. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Shobab  L, Gomes-Lima  C, Zeymo  A, et al.  Clinical, pathological, and molecular profiling of radioactive iodine refractory differentiated thyroid cancer. Thyroid. 2019;29(9):1262‐1268. [DOI] [PubMed] [Google Scholar]
  • 40. Cao  J, Zhu  X, Sun  Y, Li  X, Yun  C, Zhang  W. The genetic duet of BRAF V600E and TERT promoter mutations predicts the poor curative effect of radioiodine therapy in papillary thyroid cancer. Eur J Nucl Med Mol Imaging. 2022;49(10):3470‐3481. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Restrictions apply to the availability of some or all data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided. NGS data in this study can be accessed at https://ngdc.cncb.ac.cn/via accession number HRA004166 from the corresponding author on reasonable request.

Articles from The Journal of Clinical Endocrinology and Metabolism are provided here courtesy of The Endocrine Society

RESOURCES