2024 Volume 71 Issue 1 Pages 31-37
We analyzed the outcomes of genetic testing to study the frequency of mutations in advanced thyroid cancer in Japan. Patients (n = 96) with unresectable or metastatic thyroid carcinoma were included for retrospective chart review. Results of gene panel testing, which was performed between May 2020 and April 2023, were analyzed. The median age of the patients was 73.5 years (range, 17–88); 59 were women, and 39 were men. Overall, 17 patients had anaplastic thyroid carcinoma (ATC), 68 had papillary thyroid carcinoma (PTC), 7 had follicular thyroid carcinoma, and 6 had poorly differentiated thyroid carcinoma (PDTC). Of the 81 patients with differentiated thyroid carcinoma (DTC) and PDTC, 88.9% were radioactive iodine-refractory, and 32.7% of all cases had previously been treated with multiple kinase inhibitors. Of ATC cases, 52.9% had BRAF mutations, and 5.9% had RET fusion. Of PTC cases, 83.1% had BRAF mutations, 9.2% had RET fusion, and 1.5% had NTRK fusion. One case each of ATC and PTC had a tumor mutation burden of ≥10. ATC cases had a significantly higher prevalence of TP53 alterations than the other cases (82.3% vs. 11.8%), whereas the frequencies of TERT promoter mutations were 88.2% in ATC cases and 64.7% in the other cases, albeit without a significant difference. In conclusion, 58.8% of ATC, 93.8% of PTC, and 42.9% of PDTC had genetic alterations linked to therapeutic agents. Active gene panel testing is required to increase treatment options.
THYROID CARCINOMA is the most common endocrine cancer. According to a population-based cancer registry 2019 in Japan, the national age-adjusted incidence of thyroid cancer was 9.0/100,000 people, the mortality was 0.32/100,000 people, and the 5-year relative survival rate was 94.7% [1-3]. Differentiated thyroid carcinoma (DTC), which accounts for more than 90% of thyroid cancers have a good prognosis, and many can be cured with surgery or radioactive iodine therapy. However, a small proportion of DTC requires pharmacotherapy.
Molecular-targeted drug therapy according to tumor driver genes is used in many types of cancer and as systemic therapy for thyroid cancer. Hence, genetic testing to select the appropriate therapeutic agents has increased. Large-scale genetic alterations in thyroid cancer using The Cancer Genome Atlas Program (TCGA) data has been reported, but most of the patients have low- or intermediate-risk papillary thyroid cancer [4], and it is unknown whether it truly reflects the mutation frequency of advanced thyroid cancers requiring drug therapy. Genetic testing using next-generation sequencing (NGS) has been used to select the appropriate drug therapies. A report on the results of NGS assays on advanced thyroid cancers classified anaplastic thyroid carcinomas into clusters of BRAF and RAS mutations, and clusters in which these mutations are rare [5]. Additionally, a model of anaplastic transformation, wherein differentiated cancer acquires gene mutations such as TERT promoter or TP53 mutations, has been proposed [5].
Papillary thyroid carcinomas (PTC) account for 90% of thyroid cancers, and the most representative gene mutation is the BRAF V600E mutation, which is considered a poor prognostic factor [6, 7]. Meanwhile, the frequency of BRAF mutations is higher in Japanese populations, but the prognosis does not differ between patients with BRAF mutation-positive and wild-type [8], suggesting racial differences.
The prevalence of BRAF mutations in PTC is 74% in advanced thyroid cancers and 61.7% in the TCGA database [4, 5]. However, in the SELECT study, which is the phase 3 study of lenvatinib for DTC, only 36.9% of patients with PTC had BRAF V600E mutations, and their prognosis was better than those with wild-type mutations [9]. Due to the favorable prognosis of thyroid cancer, there are few data on unresectable or metastatic thyroid cancer, which requires drug therapy. Therefore, we analyzed the cases in which NGS was performed for the purpose of selecting the appropriate drug therapy.
This study was approved by the Kanagawa Cancer Center institutional review board (IRB no. 2022-31). Patients with unresectable or metastatic thyroid carcinoma were included for retrospective chart review. Patients with advanced thyroid cancer who are undergoing or planning to undergo molecular-targeted drug therapy were eligible for inclusion; these patients can undergo gene panel testing, which is covered by health insurance in Japan. Patients who refused to participate in this study and patients with medullary thyroid carcinoma were excluded. Gene panel testing results performed between May 2020 and April 2023 were analyzed.
Gene panel testingGene panel tests used were Oncomine DxTM Target Test (ODx), FoundationOne® CDx (F1), and FoundationOne® Liquid CDx (F1L). After preparing formalin-fixed paraffin-embedded tissues or blood at our center, nucleic acid extraction and sequencing were requested to external facilities. F1 and F1L are DNA-based NGS panel tests that include 324 genes and use tumor tissues and blood, respectively. ODx is a DNA- and RNA-based NGS panel test that uses tumor tissue and includes 46 genes.
In Japan, ODx is approved for thyroid cancer that requires drug therapy, and comprehensive genomic profiling tests, including F1 and F1L, are approved for ATC and other thyroid carcinomas after the induction of standard treatments such as tyrosine kinase inhibitors.
Regarding ODx and F1, we analyzed the storage period of tumor tissue and success of gene panel testing. The results of ODx and F1 were used for DNA analysis. Partial failure occurred when some of the target genes could not be evaluated, while failure occurred when all genes could not be analyzed. The analysis results of fusion gene detection by ODx were used for RNA analysis. Failure was defined as a qualitative or quantitative problem that prevented the test from being performed.
Statistical analysesData are reported as number and percentage for categorical variables, and as median and range for nonparametric continuous variables. Categorical variables were compared using Fisher’s exact test, and nonparametric continuous variables were compared using the Mann-Whitney U test. Statistical analysis was performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (R Foundation for Statistical Computing, Vienna, Austria). Results were considered statistically significant when p < 0.05.
Of the 281 patients who had unresectable or metastatic thyroid carcinoma, 99 underwent gene panel testing during the study period and were used in the success rate analysis. The result of three tests was not available due to qualitative or quantitative problems. Of these, two patients used ODx, and one patient used F1L. The results of the 96 patients were included in the gene alteration analysis (Fig. 1).
Patient selection and analyses
ODx was used in 64 cases, F1 in 31 cases, and F1L in 4 cases. For the F1L cases, the surgical specimens were old and repeat biopsies were difficult to perform in 3 cases. In one case, core needle biopsy was performed, which contained necrotic tissue and was inappropriate for NGS. The median age at testing was 73.5 years (range, 17–88); 59 patients were women, while 39 were men. Moreover, 17 patients had ATC, 7 had follicular thyroid carcinoma (FTC), 68 had PTC, and 6 had poorly differentiated thyroid carcinoma (PDTC). The percentage of male by pathological types was 52.9% for ATC, 33.8% for PTC, 71.4% for FTC, and 33.3% for PDTC. The median age at test was 73 (range, 51–88) for ATC, 75 (51–88) for PTC, 75 (59–80) for FTC, and 70.5 (51–83) for PDTC.
Of the 81 cases of DTC and PDTC, 88.9% were radioactive iodine (RAI)-refractory, and 32.7% of all cases had previously been treated with multiple kinase inhibitors. The median age at diagnosis was 66 years (range, 15–85), and the median time from onset to gene panel testing was 5 years (range, 0–38).
Tumor tissues submitted for examination were collected from the thyroid gland in 46 cases, cervical lymph node in 38 cases, lungs in 8 cases, and 1 each from the skin, trachea, and chest wall. General anesthesia surgery was the method of tumor tissue collection in 87.4% of cases, while biopsy was performed in 11.6% cases. The main purpose of tissue collection was NGS testing in 31 cases (32.6%), where the test was performed immediately after tissue collection. The median tissue storage time was 0.2 and 2.4 years in F1 and ODx, respectively (Table 1).
Patient characteristics according to gene panel testing
| Characteristics | Number of patients (%) | ||||
|---|---|---|---|---|---|
| FoundationOne, n = 31 |
FoundationOne liquid, n = 4 |
Oncomine Dx, n = 64 |
Total, n = 98** |
||
| Sex | Male | 15 (48.4) | 1 (25) | 23 (35.9) | 39 (39.8) |
| Female | 16 (51.6) | 3 (75) | 41 (64.0) | 59 (60.2) | |
| Age at test | Median (range) | 75 (51–88) | 64.5 (57–80) | 72.5 (17–86) | 73.5 (17–88) |
| Age at diagnosis * | Median (range) | 73 (49–78) | 57 (49–72) | 66 (15–85) | 66 (15–85) |
| Pathological type | Papillary thyroid carcinoma | 10 (32.3) | 2 (50) | 57 (89.0) | 68 (69.4) |
| Follicular thyroid carcinoma | 4 (12.9) | 1 (25) | 2 (3.1) | 7 (7.1) | |
| Anaplastic thyroid carcinoma | 16 (51.6) | 1 (25) | 0 | 17 (17.3) | |
| Poorly differentiated thyroid carcinoma | 1 (3.2) | 0 | 5 (7.8) | 6 (6.1) | |
| Previous treatment | Radioactive iodine therapy * | 14 (93.3) | 3 (100) | 56 (87.5) | 72 (88.9) |
| Multiple kinase inhibitor | 26 (83.9) | 4 (100) | 3 (4.7) | 32 (32.7) | |
| Radiation therapy for neck | 10 (32.3) | 0 | 1 (1.6) | 11 (11.2) | |
| Tissue submitted for examination | Thyroid | 24 (77.4) | 22 (34.4) | 46 (48.4) | |
| Neck lymph node | 5 (16.1) | 33 (51.6) | 38 (40.0) | ||
| Lung | 1 (3.2) | 7 (10.9) | 8 (8.4) | ||
| Skin | 0 | 1 (1.6) | 1 (1.1) | ||
| Trachea | 1 (3.2) | 0 | 1 (1.1) | ||
| Chest wall | 0 | 1 (1.6) | 1 (1.1) | ||
| Tissue collection method | General anesthesia surgery | 23 (74.2) | 60 (93.8) | 83 (87.4) | |
| Excisional or incisional biopsy | 2 (6.5) | 2 (3.1) | 4 (4.2) | ||
| Core needle biopsy | 6 (19.4) | 1 (1.6) | 7 (7.4) | ||
| Computed tomography-guided biopsy | 0 | 1 (1.6) | 1 (1.1) | ||
| Tissue storage periods (year) | Median (range) | 0.2 (0–6.3) | 2.4 (0–10.6) | 2.2 (0–10.6) | |
* Excluding anaplastic thyroid carcinoma ** In one case, Oncomine Dx was submitted after failing FoundationOne liquid.
The results of ODx and F1 were used for DNA analysis. Of these, 83.3% were successful, 14.6% were partial failures, and 2.1% were failures. The success rate was 85.7% vs. 73.7% (p = 0.299) when comparing tissue storage periods of <5 years with those of ≥5 years.
ODx was used for RNA analysis. Overall, 89.2% of cases underwent RNA testing, while 10.8% had a qualitative or quantitative problem that prevented the test from being performed. The success rate was 92% vs. 80% (p = 0.338) when comparing tissue storage periods of <5 years with those of ≥5 years.
Although the tissue storage time was up to 10.6 years and the test quality was suspected to deteriorate over time, most of the tests were successful (Fig. 2).
Storage time of specimens and genome analysis
DNA and RNA analysis status of each test is shown by classifying specimen retention periods into <1 year, 1–2 years, 3–4 years, and ≥5 years. Blue are successful cases, gray are unsuccessful cases, and orange are cases in which some of the target genes could not be evaluated.
In the following cases, actionable gene alterations, which are candidates for targeted drug therapy, were observed. Among the 17 cases of ATC, 52.9% had BRAF mutations, and 5.9% had RET fusion. Among the 65 cases of PTC, 83.1% had BRAF mutations, 9.2% had RET fusion, and 1.5% had NTRK fusion. Among the 7 cases of PDTC, 28.6% had BRAF mutations and 14.3% had RET fusion. No actionable mutations were observed among cases of FTC (Table 2). These driver genes were mutually exclusive.
Frequencies of representative gene alterations
| Genetic alteration | Number of patients (%) | ||||
|---|---|---|---|---|---|
| Total | PTC | ATC | FTC | PDTC | |
| BRAF V600E | 65 (67.7) | 54 (83.1) | 9 (52.9) | 0 | 2 (28.6) |
| RAS mutation | 11 (11.5) | 1 (1.5) | 4 (23.5) | 5 (71.4) | 1 (14.3) |
| RET fusion | 8 (8.3) | 6 (9.2) | 1 (5.9) | 0 | 1 (14.3) |
| NTRK fusion | 1 (1.0) | 1 (1.5) | 0 | 0 | 0 |
| Other | 3 (3.1) | 0 | 3 (17.6) | 0 | 0 |
| Not detected | 8 (8.3) | 3 (4.6) | 0 | 2 (28.6) | 3 (42.9) |
PTC, papillary thyroid carcinoma; ATC, anaplastic thyroid carcinoma; FTC, follicular thyroid carcinoma; PDTC, poorly differentiated thyroid carcinoma
Among the fusion genes, the most common variant was CCDC6-RET fusion, which was found in five cases, followed by two cases of ERC1-RET fusion and one each of NCOA4-RET fusion and ETV6-NTRK3 fusion. Eight cases were detected by ODx and one case was detected by F1L. The breakdown of RAS mutations was NRAS Q61R in 8 cases, NRAS Q61K in 2 cases, and HRAS Q61K in 1 case (Table 3).
Details of fusion genes and RAS mutations
| Genetic alteration | Total | Number of patients | ||||||
|---|---|---|---|---|---|---|---|---|
| Pathological type | Genome testing method | |||||||
| PTC | ATC | FTC | PDTC | ODx | F1 | F1L | ||
| CCDC6-RET | 5 | 4 | 1 | 0 | 0 | 4 | 0 | 1 |
| ERC1-RET | 2 | 1 | 0 | 0 | 1 | 2 | 0 | 0 |
| NCOA4-RET | 1 | 1 | 0 | 0 | 0 | 1 | 0 | 0 |
| ETV6-NTRK3 | 1 | 1 | 0 | 0 | 0 | 1 | 0 | 0 |
| NRAS Q61R | 8 | 0 | 3 | 4 | 1 | 2 | 5 | 1 |
| NRAS Q61K | 2 | 1 | 0 | 1 | 0 | 2 | 0 | 0 |
| HRAS Q61K | 1 | 0 | 1 | 0 | 0 | 0 | 1 | 0 |
PTC, papillary thyroid carcinoma; ATC, anaplastic thyroid carcinoma; FTC, follicular thyroid carcinoma; PDTC, poorly differentiated thyroid carcinoma; ODx, Oncomine DxTM Target Test; F1, FoundationOne® CDx; F1L, FoundationOne® Liquid CDx
We found one case each of ATC and PTC with a tumor mutation burden (TMB) of ≥10; both had BRAF mutations. Of the cases in which F1 was performed, median TMB was 2.41 mut/Mb (range, 0–11.35). There was no patient with high microsatellite instability (MSI) (Table 4).
Microsatellite instability and tumor mutation burden
| Number of patients | |||
|---|---|---|---|
| FoundationOne® CDx | FoundationOne® Liquid CDx | ||
| Microsatellite instability | High | 0 | 0 |
| Stable | 29 | 2 | |
| Not determined | 2 | 1 | |
| Tumor mutation burden (mut/Mb) | Median (range) | 2.41 (0–11.35) | 2.53 (1.26–6.32) |
Next, we compared ATC to DTC, including PDTC, for other gene alterations using the results of F1 and F1L because they target more genetic abnormalities than ODx. Genes that were altered in 15% or more cases were examined. The prevalence of TP53 alterations was 82.3% in ATC and 11.8% in DTC, indicating that cases of ATC had a significantly higher prevalence of TP53 alterations. Meanwhile, TERT promoter mutations were found in 88.2% and 64.7% of cases of ATC and DTC, respectively, albeit without a significant difference. Additionally, no difference was observed for CDKN2A, CDKN2B, and PIK3CA (Table 5).
TP53 is more frequently altered in ATC than in DTC
| Gene | Prevalence, % | p | |
|---|---|---|---|
| ATC | DTC | ||
| TP53 | 82.3 | 11.8 | 0.000086 |
| TERT promoter | 88.2 | 64.7 | 0.225 |
| CDKN2A | 23.5 | 11.8 | 0.656 |
| CDKN2B | 17.6 | 0 | 0.227 |
| PIK3CA | 17.6 | 17.6 | 1 |
We used the results of FoundationOne and FoundationOne Liquid to show the frequency of cases with each gene abnormality. P-values less than 0.05 are shown in bold.
ATC, anaplastic thyroid carcinoma; DTC, differentiated thyroid carcinoma
Regarding papillary carcinoma, we compared the frequency of genetic alterations in this study with that in the TCGA database, which mainly consists of low- to intermediate-risk cases, with an Asian frequency of 10.4%. BRAF mutations were significantly more common in cases of advanced cancer (83.1% vs. 61.7%, p = 0.0007), while RAS mutations were significantly less common (1.5% vs. 12.9%, p = 0.005). There was no significant difference in RET or NTRK fusion genes (Table 6).
Comparison between advanced and low- to intermediate-risk PTC for frequencies of representative gene alterations
| Genetic alteration | Number of patients (%) | p value | |
|---|---|---|---|
| Advanced PTC | TCGA database | ||
| BRAF mutation | 54 (83.1) | 248 (61.7) | 0.0007 |
| RAS mutation | 1 (1.5) | 52 (12.9) | 0.005 |
| RET fusion | 6 (9.2) | 33 (6.8) | 0.444 |
| NTRK fusion | 1 (1.5) | 11 (2.3) | 1 |
PTC, papillary thyroid carcinoma; TCGA, The Cancer Genome Atlas Program
P-values less than 0.05 are shown in bold.
Six cases of PTC with RET fusion gene and one case of PTC with NTRK fusion gene were detected in this cohort. The median age at diagnosis was 51 years (range, 15–70) and 67 years (range, 27–85) for patients with and without fusion genes, respectively. The age at diagnosis was significantly younger in patients with fusion gene-positive PTC (p = 0.0086, Fig. 3).
Patients with PTC with fusion genes are diagnosed at a younger age
In patients with PTC with RET or NTRK fusion genes, the median age at diagnosis was 51 (range, 15–70), whereas the median age at diagnosis in patients without fusion genes was 67 years (range, 27–85). The age at diagnosis was significantly younger in patients with fusion gene-positive PTC (p = 0.0086, Mann-Whitney U test).
Among malignant tumors, DTC has a relatively good prognosis and is characterized by a long period from initial surgery to drug therapy due to its slow progression. In this cohort, the median time from initial surgery to genomic testing was 5 years, with a maximum of 38 years. In the genome testing using formalin-fixed paraffin-embedded tissue block, longer storage period increased the decomposition and deterioration of the nucleic acid over time [10]. The nucleic acid quality may be compromised in thyroid cancer due to long storage periods of surgical specimens. However, in this study, gene panel tests by NGS were mostly successful during all storage periods, with a maximum of 10.6 years. Although the most recent tumor gene profile is desired at the start of drug therapy, using older archival specimens is an alternative when repeat biopsy is difficult.
A typical genetic alteration in thyroid carcinoma is BRAF V600E mutation, which occurs in 30%–70% of patients with PTC [4, 5, 9]. BRAF V600E mutations were also identified as a poor prognostic factor, and their presence was significantly associated with increased cancer-related mortality in patients with PTC [6, 7]. BRAF mutation-positive PTC suppresses expression and/or proper localization of the sodium–iodide symporter, impeding their ability to uptake radioiodine [11-13]. In this study, the frequency of BRAF mutations among patients with PTC was higher than previously reported. One of the reasons is that the patient population was mainly comprised of patients with a poor prognosis due to RAI-refractory DTC. Meanwhile, BRAF wild-type PTC was more common than BRAF mutation-positive PTC in the SELECT study, which is also comprised of patients with RAI-refractory DTC. It was reported that the BRAF mutation-positive rate is high in cases of advanced PTC in Japan [8]. Therefore, racial differences may be another cause.
Fusion genes are driver genes for thyroid cancer. FoundationCORE, a database of gene panel testing, found that 1.6% of thyroid cancers have NTRK fusion genes [14]. The TCGA database also found RET and NTRK fusion genes in 6.3% and 2.3% of cases of PTC, respectively [4]. In this study, the frequency of RET and NTRK fusion genes among cases of PTC was 9.2% and 1.5%, respectively, which is similar to previous reports. Fusion genes are often found in pediatric thyroid carcinoma. In a 93-case pediatric carcinoma cohort, the prevalence of RET, NTRK, and ALK fusion genes were 28%, 18%, and 6%, respectively [15]. In this study, most of the patients were adults, but the age of onset was younger in cases with fusion genes. Younger patients have a favorable prognosis, and it is expected that the period from the initial operation to the start of drug therapy will be longer. In these cases, detection of fusion genes should be performed before nucleic acid quality is compromised.
MSI and TMB are predictors of efficacy for pembrolizumab monotherapy, which is an immune checkpoint inhibitor, in patients with previously treated recurrent or metastatic advanced solid tumors [16, 17]. In this study, there were no cases of high MSI, and the prevalence of high TMB cases was 5.9%. Therefore, pembrolizumab is recommended only in a limited number of cases for thyroid cancer.
Mutations in the TERT promoter region are associated with poor prognosis in thyroid cancer [8, 18]. The prevalence of TERT mutations was high in both aggressive DTC and ATC cases. On the other hand, the prevalence of pathogenic TP53 alterations was significantly higher among cases of ATC than of DTC. In a study comparing PTC and ATC components in ATC cases with coexisting PTC, TERT promoter mutations occurred frequently in both PTC and ATC components, but TP53 abnormalities were significantly more common in ATC components [19]. TERT promoter mutations may be involved in the progression of DTC, and TP53 abnormalities may be associated with anaplastic transformation. This study also produced consistent results.
In conclusion, 58.8% cases of ATC, 93.8% cases of PTC, and 42.9% cases of PDTC had genetic alterations linked to therapeutic agents in the advanced thyroid carcinoma cohort. Active gene panel testing is required to increase the treatment alternatives.
The authors would like to thank Enago (www.enago.jp) for the English language review.
None of the authors have any potential conflicts of interest associated with this research.
