1.Baseline of the patients included in this study
The clinical pathological information of the patients collected for this study was retrieved from our electronic medical record system, and a total of 2844 samples information were counted. Baseline of the patients included in this study are shown in Table 1.
Table 1
Base line of the patients involved in this study (n = 2844)
Sex | | |
| Female | 1993 (70.1%) |
| Male | 851 (29.9%) |
Age | | 43.5 ± 12.6 |
Recurrence | | 224(7.8%) |
Sample resource | | |
| Surgical tissues | 2149 (75.5%) |
| FNA | 695 (24.5%) |
Sample tissues type | | |
| Thyroid tissues | 2650 (93.1%) |
| Lymph tissues | 179 (6.3%) |
| Metastasis tissues | 15 (0.5%) |
NGS panel | | |
| 30 genes | 2045 (71.9%) |
| 32 genes | 700 (24.6%) |
| 88 genes | 99 (3.5%) |
pathological features | | |
| Size of tumor | 12.9 ± 9.9 |
| Multifocality | 745 (26.2%) |
| Vascular invasion | 216 (9.3%) |
| Nerve invasion | 54 (2.3%) |
| Muscle invasion | 61 (2.6%) |
| Capsular invasion | 111 (4.8%) |
pathology | | |
| PTC | 2425 (85.3%) |
| FVPTC | 19 (0.7%) |
| FTC | 22 (0.8%) |
| MTC | 64 (2.3%) |
| ATC or PDTC | 30 (1.0%) |
| Other malignant | 14 (0.5%) |
| Benign | 170 (6.0%) |
| Uncertain malignant potential | 78 (2.7%) |
| Bethesda 1 | 22 (0.8%) |
TNM | | 2329 |
| I | 2103 (90.3%) |
| II | 148 (6.4%) |
| III | 48 (2.1%) |
| IV | 30 (1.3%) |
2. Gene Mutation Condition
A total of 2844 cases with effective gene detection reports were counted, and 2337 cases (82%) had gene changes detected. The occurrence of each type of gene mutation is shown below (Fig. 1). Genes with mutation frequency less than 0.5% are classified into OTHER. The common gene mutation types are BRAF, RAS, TERT, TP53, RET mutation, RET/PTC fusion, etc.
The most common driving genes in thyroid tumors are BRAF, RAS, and RET. We counted other rare single gene mutation and their pathological reports (Table 2). It suggests PTEN, SPOP, TSHR, ZNF148 and GNAS gene mutation may not be capable of driving the malignant progression of thyroid nodules on their own. RET/PTC fusion and NTRK3 fusion may be capable of driving the occurrence of PTC.
Table 2
The incidence of other rare gene and pathological types
Gene alteration | pathologic diagnosis |
| Benign | Uncertain | PTC | FVPTC | FTC | ATC | PDTC | OTHER |
CTNNB1 | 1 (100%) | | | | | | | |
CDKN2A | | | | | | | 1 (100%) | |
CHEK2 | | | | | | | | 1 (100%) |
EIF1AX | 2 (66.6%) | | 1 (33.3%) | | | | | |
EML4-NTRK3 | | | 3 (100%) | | | | | |
ETV6-NTRK3 | | 1 (5.3%) | 16 (84.2%) | 2 (10.5%) | | | | |
SQSTM1-NTRK3 | | 1 (100%) | | | | | | |
EZH1 | 1 (33.3%) | 1 (33.3%) | 1 (33.3%) | | | | | |
GNAS | 2 (100%) | | | | | | | |
NTRK1 | 1 (100%) | | | | | | | |
PTEN | 4 (80%) | | | | | 1 (20%) | | |
RET/PTC1 | | 3 (3.9%) | 72 (94.7%) | 1 (1.3%) | | | | |
RET/PTC3 | | 1 (6.3%) | 15 (93.7%) | | | | | |
SPOP | 5 (83.3) | | 1 (16.7%) | | | | | |
TERT | | | 3 (60%) | | 1 (20%) | 1 (20%) | | |
TP53 | | | 5 (55.6%) | | | 2 (22.2%) | 1 (11.1%) | 1 (11.1%) |
TSHR | 4 (66.6%) | 1 (16.7%) | 1 (16.7%) | | | | | |
ZNF148 | 4 (100%) | | | | | | | |
3. Application Of Ngs In Fna Samples
FNA has proved to be a rapid, economical, safe, and reliable method of investigation for the initial evaluation of patients with thyroid nodule nodules found by ultrasound. Furthermore, the method is the most appropriate diagnostic tool to distinguish the patients requiring clinical management or surgical resection[17]. The Bethesda Thyroid Cytopathology Reporting System established a standardized, category-based reporting system for FNA specimens of the thyroid gland[18]. The incidence of two types of uncertain cytology is about 10%[17] which leads to a dilemma in clinical decisions. To solve this dilemma, strategies have been proposed that combine cytopathology with molecular patterns of FNA samples to improve diagnostic accuracy[19–22]。
A total of 695 samples of this study were FNA. There were 42 Bethesda grade 1 nodules, 134 Bethesda grade 2 nodules, 114 Bethesda grade 3 or 4 nodules, and 391 Bethesda grade 5 or 6 nodules. To determine the diagnostic efficiency of NGS in FNA nodules, we compared the sensitivity and specificity of cytological diagnosis and the combination of cytology and NGS (Table 3). A total of 308 patients with NGS data for FNA nodules received thyroidectomy and obtained the final surgical pathology. The combination of cytology and NGS data significantly increased the diagnostic sensitivity from 0.76 to 0.91 with a significant decrease in specificity. However, due to sample bias (Bethesda 1–4 samples are often not subjected to NGS testing or surgical treatment), the postoperative pathological benign sample size is extremely small, resulting in low specificity, negative predictive value, and Jordan index, which cannot indicate the authenticity of this diagnostic experiment for the time being.
Table 3
Diagnostic efficiency of cytology and cytology combined with NGS
Diagnosis specificity, sensitivity, Youden Index, positive/negative predictive value |
| Cytology | |
| Bethesda V/VI | Bethesda I to IV | |
Malignant | 231 | 71 | Se 0.76 Sp 1 YI 0.76 PPV 1 NPV 0.08 |
Benign | 0 | 6 |
| NGS with cytology | |
| Bethesda V/VI or (with) pathogenic mutation | Bethesda I to IV without pathogenic mutation | |
Malignant | 277 | 25 | Se 0.91 Sp 0.33 YI 0.25 PPV 0.99 NPV 0.07 |
Benign | 4 | 2 |
4. Braf Mutation May Indicate A Better Prognosis
A total of 2026 of these 2,844 cases had BRAF mutations (71%) and a total of 1985 cases (82%) were BRAF positive in all PTC or papillary microcarcinomas. The positive rate is much higher than about 50% in previous studies[23], suggesting that the BRAF mutation may have a higher positive rate in the Chinese mainland region.
Numerous studies and meta-analyses have demonstrated the association of BRAFV600E mutations with high-risk clinical pathology features[11, 12, 24, 25]。 However, in a large multicenter retrospective study, this association was no longer statistically significant after adjustment for clinical and histopathological features of aggressive thyroid tumors. The BRAFV600E test did not increase the predictive value of PTC-related prognosis and mortality[13]。 Therefore, we analyzed the clinicopathological features of PTC patients with positive BRAF mutations (Table 4). Compared with the BRAF negative PTC patients, BRAF positive patients have lower recurrence rate, vascular invasion and tumor size, higher age and tumor multiformity. In general, there is no clear correlation between BRAF V600E and the prognosis of PTC patients.
The point mutation at base c.1799T > A of BRAFV600E accounted for this absolute advantage in all samples with BRAF gene mutations detected (n = 2024). What is worth noting was the discovery of a base c.1801A > G mutation, which ultimately led to a pathological diagnosis of a dominant follicular tumor with follicular carcinoma with minimal invasion in some areas. One case of BRAF base c.1862A > G mutation combined with the KRAS base c.34G > A mutation led to the pathological diagnosis of a rare large, poorly differentiated thyroid B-cell non-Hodgkin's lymphoma.
A total of 187 cases were BRAF combined with other mutations, and the specific co-mutations are shown in Fig. 1. The most common co-mutant gene was the TERT promoter mutation. There were 81 cases of it of which 71 cases (87.7%) showed more aggressive PTC, which is consistent with the existing research conclusions[26]. Notably, there were 10 cases with BRAF mutation and the pathological diagnosis was undifferentiated or poorly differentiated cancer. Except for one case where BRAF was combined with AKT1 mutation, the remaining 10 cases were all BRAF combined with TERT promoter mutation, and they were often combined with TP53 (n = 3) or PIK3CA (n = 3) three gene co-mutation or four gene co-mutation (n = 1). Suggesting that BRAF mutations alone may not be sufficient to cause the dedifferentiation process of thyroid cancer.
Table 4
Clinical features of BRAF + PTC patients
| BRAF+ | BRAF- | P |
number | 1985 | 440 | |
Sex(female) | 1373(69.2%) | 323(73.4%) | 0.079 |
Age | 42.84 ± 12.13 | 41.55 ± 12.72 | 0.046 |
Recurrence | 111(5.6%) | 56(12.7%) | 0 |
Size of tumor | 11.98 ± 8.17 | 13.41 ± 10.97 | 0.014 |
Multifocality | 602(34.1%) | 106(28.8%) | 0.049 |
Vascular invasion | 135(6.9%) | 57(13.5%) | 0 |
Nerve invasion | 41(2.1%) | 5(1.2%) | 0.214 |
Muscle invasion | 41(2.1%) | 10(2.4%) | 0.733 |
Capsular invasion | 76(3.9%) | 15(3.5%) | 0.737 |
Extranodal invasion | 146(7.8%) | 43(10.4%) | 0.088 |
TNM | | | |
I | 1670(91.9%) | 338(89.2%) | > 0.05 |
II | 103(5.7%) | 30(7.9%) | > 0.05 |
III | 37(2%) | 9(2.4%) | > 0.05 |
IV | 8(0.4%) | 2(0.5%) | > 0.05 |
5. Tert Promoter Mutation May Be A Late Molecular Event Of The Tumor
According to other research, in PDTC and ATC, the frequency of TERT promoter mutations significantly increased, the mutation rate in PDTC up to 40%, and the mutation frequency in ATC up to 50–70%[27–29]. Numerous studies have shown that mutations in the TERT promoter are associated with poor prognosis. The mutation of the TERT promoter often co-exists with BRAF or RAS mutation in thyroid cancer. Patients with the coexistence of the TERT promoter and BRAF mutations had significantly larger tumors, more extraglandular invasion, higher TNM staging, and higher recurrence rates compared with the patients with BRAF mutations alone. Suggesting that mutations in the TERT promoter may lead to more aggressive tumors[30, 31]. In this study, mutations in the TERT promoter were detected in 99 samples (3%), and the patients with TERT mutations positivity had extremely higher ages, recurrence rates, tumor size, tumor invasiveness, TNM staging, and were strongly associated with ATC, showing extremely poor prognostic features and were more likely to be detected in male patients (Table 5).
94 samples were positive for other mutations besides TERT, and only 5 samples had single-gene mutations of TERT. A total of 80 cases were double gene co-mutation, 11 cases were 3 gene co-mutation, and two cases were 4 gene co-mutation. The co-mutations are shown in Fig. 2. FTC, PTC, ATC, and ATC were presented in the 5 cases with only single-gene mutation of TERT, and no statistical significance clinical feature of single TERT mutation was found due to the small number of cases. It should be noted that TERT is often co-present with TP53 (n = 9) and PIK3CA (n = 5) in BRAF or RAS positive cases and that such cases of 3-gene co-mutation or 4-gene co-mutation are all ATC or spindle cell malignancies, suggesting a possible synergistic effect of TERT with TP53 or PIK3CA in tumor dedifferentiation.
Table 5
Clinical features of TERT + patients
| TERT+ | TERT- | P |
number | 99 | 2745 | |
Sex(female) | 49 (49.5%) | 1944 (70.8%) | 0 |
Age | 58.9 ± 12.7 | 42.9 ± 12.2 | 0 |
Recurrence | 42 (42.4%) | 180 (6.6%) | 0 |
Size of tumor | 25.4 ± 16.3 | 12.6 ± 9.5 | 0 |
Multifocality | 20 (46.5%) | 725 (32.6%) | 0.054 |
Vascular invasion | 13 (15.9%) | 202 (7.6%) | 0.006 |
Nerve invasion | 5 (6.1%) | 47 (1.8%) | 0.015 |
Muscle invasion | 11 (13.6%) | 50 (1.9%) | 0 |
Capsular invasion | 12 (14.8%) | 86 (3.2%) | 0 |
Extranodal invasion | 13 (22.0%) | 189 (8.0%) | 0 |
PTC | 76 (76.8%) | 2349 (85.6%) | < 0.05 |
FVPTC | 2 (2.0%) | 17 (0.6%) | |
PDTC | 3 (3.0%) | 6 (0.2%) | < 0.05 |
ATC | 14 (14.1%) | 7 (0.3%) | < 0.05 |
FTC | 3 (3.0%) | 19 (0.7%) | < 0.05 |
MTC | 0 (0%) | 64 (2.3%) | |
Other malignant | 1 (1.0%) | 13 (0.5%) | |
benign | 0 (0%) | 170 (6.2%) | < 0.05 |
Uncertain malignant | 0 (0%) | 78 (2.8%) | |
Bethesda I | 0 (0%) | 22 (0.8%) | |
TNM | | | |
I | 23 (41.8%) | 2079 (91.6%) | < 0.05 |
II | 15 (27.3%) | 133 (5.9%) | < 0.05 |
III | 8 (14.5%) | 38 (1.7%) | < 0.05 |
IV | 9 (16.4%) | 20 (0.9%) | < 0.05 |