A cohort study on 10‐year survival of sporadic medullary thyroid carcinoma with somatic RET mutation

Abstract

Somatic rearranged during transfection (RET) mutations are reported in 40–50% of sporadic medullary thyroid carcinoma (sMTC) patients with prognostic significance. As there is a lack of somatic RET mutations reported previously for the Taiwanese population, we tried to assess the presence of somatic RET mutations and evaluate the potential outcome predictors for our sMTC patients. We collected data from seven sMTC patients from the years 1997 to 2005 and analyzed their clinic‐pathological features up to 2015. All patients were still alive to follow up for 11∼18 years. Tumor DNAs were extracted to assess exons 10–11 and 13–16, and the intron‐exon boundaries of the RET gene. Six cases (86%) were screened positive of somatic RET gene mutations in hotspot regions, one at M918T, one at C620R, and three at C634S, with another two rare mutations at L629Q and V642I. Comparing the current tumor, node, metastases staging system, the 10‐year survival outcomes for our sMTC patients was not predicted by serum calcitonin and/or carcinoembryonic antigen, surgical extent, and presence of the somatic RET gene mutations. The small cohort demonstrated a relatively good outcome of sMTC patients to survive >10 years. In addition, intensive treatment with total thyroidectomy with extensive neck lymph node dissection seemed to be the critical determinant of better survival outcome for sMTC patients.

Introduction

Medullary thyroid carcinoma (MTC), arising from calcitonin‐secreting parafollicular C cells, accounts for 1–2% of all thyroid cancers. Sporadic MTC accounts for the majority (75%), while the remaining hereditary MTC is caused by germline mutations composed of multiple endocrine neoplasia (MEN) Type 2A, Type 2B, and familial MTC [[1], [2], [3]]. The rearranged during transfection (RET) proto‐oncogene mutation has been identified as the driver for the onset and progression of hereditary MTC. The RET proto‐oncogene is composed of 21 exons located on chromosome 10q11‐2 and encodes a tyrosine kinase transmembrane receptor. The receptor has three domains: (1) extracellular domain (exon 1–10) with a distal cadherin like region and a juxta‐membrane cysteine‐rich region; (2) transmembrane domain (exon 10–11); and (3) intracellular domain (exon 12–21) with tyrosine kinase activity. Previous studies have demonstrated that RET oncogene mutation results in constitutively mutant ret proteins, while activation of the ret protein induces autophosphorylation of tyrosine kinase. The continuous activation signaling could lead to unregulated cell proliferation and malignant transformations [[4], [5]].

Germline RET oncogene mutations were found in almost all hereditary MTC patients. Sporadic MTC (sMTC) is diagnosed and defined by a lack of RET gene germline mutation, no family history of MTCs, and negative clinical and laboratory evidence of other endocrine neoplasms. Somatic RET mutations are reported to occur in 40–50% of sMTC tumors with various prevalence in different ethnic populations. “Hot spot” regions of somatic RET mutation were mostly found in exon 10, exon 11, and exon 16 [[6], [7], [8], [9]]. Harboring somatic RET mutations in sMTC has been reported to signify larger tumor size, more aggressiveness, poor prognosis, and good drug response in several studies [[10], [11], [12]]. However, the clinical behavior of sMTC is still unpredictable by the current staging system [[3], [10]].

Previously, Chang et al. [13] did not report any RET mutation in hot spot regions in sMTC cases in a Taiwanese population. Our study surveyed potential common somatic RET mutations in our sMTC cases. We also compared the long‐term outcomes and explored the prognostic significance of these RET mutations among Taiwanese patients.

Methods

We collected data from seven patients (1 male, 6 females) diagnosed with sMTC at Kaohsiung Medical University Hospital within 8 years (1997–2005). All of these patients were diagnosed at the age of 49.3 ± 16.9 years (from 19 years to 76 years). Their initial manifestations were nearly all palpable neck lumps. Surgical intervention with ipsilateral or bilateral thyroidectomy was done initially, complemented by lymph node dissection (Table 1). These patients were still alive and followed in our outpatient clinic for 11–18 years up to the end of this study in 2015. They were followed in an outpatient clinic to evaluate the clinical status via scheduled thyroid ultrasound, neck computerized tomography (CT), thallium⁻²⁰¹ thyroid or whole body scan, Technetium (^99mTc) sestamibi (Tc^99m‐MIBI) combined with single photon emission computed tomography (SPECT/CT) scan in addition to the periodic evaluation of serum calcitonin or carcinoembryonic antigen (CEA) by radioimmunoassay (Cisbio Bioassays, Codolet, France). Fludeoxyglucose¹⁸ positron emission tomography scans were also used to help define free of tumor recurrence or metastasis.

Table 1.

Outcome with RET proto‐oncogene mutations in case series of sporadic medullary thyroid carcinoma.

Case	Age (y)	Sex	TNM stage a	Pathology	Initial treatment	Tumor markers (pre op)	Tumor marker (post op)	Mutation	Follow (y)	Outcome and treatment
1	76	F	T4aN0M0 Stage IVA	5 cm with extrathyroid invasion	Total thyroidectomy	CEA 1.9 ng/mL		V642I L629Q	18	Alive, no recurrence
2	46	M	T2N1bM0 Stage IVA	2.6 cm with lymph node metastasis	Total thyroidectomy + LN dissection		CEA 41.4 ng/mL b CEA 3.89 ng/mL and Ct 44 pg/mL (2 y); CEA 37.6 ng/mL and Ct 665.9 pg/mL (8 y) CEA 16.1 ng/mL (9 y) CEA 26.74 ng/mL (10 y)	M918T	13	Alive, 1st LN dissection (8 y); 2nd LN dissection (10 y)
3	53	F	T3N0M0 Stage II	2 cm with capsular invasion	Total thyroidectomy + LN dissection	Ct 111 pg/mL, c CEA 90.1 ng/mL	Ct 1.0 pg/mL and CEA 1.16 ng/mL b	C634S	11	Alive, no recurrence
4	49	F	T3N0M0 Stage II	3.2 cm with vascular, capsular and extrathyroid invasion	Left lobectomy		Ct 513.54 pg/mL and CEA 23.92 ng/mL (6 y) Ct 703.71 pg/mL (7 y) Ct 809.69 pg/mL (8 y)	C620R	11	Alive, 1st LN dissection (5 y); 2nd total thyroidectomy +LN dissection (8); no more recurrence
5	46	F	T2N0M0 Stage II	3.5 cm, encapsulated	Right lobectomy		Ct 1.0 pg/mL and CEA 1.15 ng/mL b ; Ct 18.8 pg/mL (1 y)	none	11	Alive, total thyroidectomy + LN dissection (3 y); then no recurrence
6	56	F	T2N0M0 Stage II	2 cm	Total thyroidectomy		Ct 61.9 pg/mL (3 y) Ct 31 pg/mL (4 y) CEA 0.7 ng/mL (5 y)	C634S L629Q	16	Alive, no recurrence
7	19	F	T2N0M0 Stage II	2.5 cm with vascular invasion	Total thyroidectomy		CEA 10.9 ng/mL b CEA 1.24 ng/mL (1 y)	C634S	14	Alive, no recurrence

CEA = carcinoembryonic antigen; Ct = calcitonin; LN = lymph nodes; TNM = tumor, node, metastases.

^aStaging of medullary thyroid carcinoma according to American Joint Cancer Committee 7^th edition.

^bSerum calcitonin and CEA were assessed at the initial follow up within 3 months after operation.

^cSerum Ct < 10 pg/mL and CEA < 5 ng/dL were referenced as normal.

Under the approval and supervision of the Institutional Review Board of Kaohsiung Medical University Hospital (IRB KMUH‐98‐00237), we started the study after obtaining informed consent from all patients. DNA was extracted from blood and formalin fixed paraffin embedded (FFPE) block by the QuickExtract FFPE DNA extraction kit (Epicentre, Madison, WI, USA) according to the manufacturer's protocol. Exons 10–11, 13–16, and the intron‐exon boundaries of the RET gene were amplified by semi‐nested polymerase chain reaction (PCR) to increase the sensitivity for the suboptimal DNA quality from blood and FFPE samples. Briefly, first PCR was performed with FFPE DNA plus outer forward, reverse (R) primers, and PCR master mixture. A 10‐fold dilution of the first PCR products was used as the template for a second PCR with inner forward and R primers. PCR products were confirmed by electrophoresis in 1.5% agarose gel and sequenced by an ABI 3730XL DNA analyzer (Applied Biosystems, Foster City, CA, USA). PCR amplification primers are listed in Table 2. The reference sequences of the RET gene and mRNA were NG_007489.1 and NM_020975.4.

Table 2.

Primers used for semi‐nested polymerase chain reaction amplification of human RET gene.

Primer	Sequences	Length (bp)
10_OF/IF 10_R	ACACTGCCCTGGAAATATGG/TATGGGCGCCTGGGGT CTCAGATGTGCTGTTGAGAC	263/248
11_OF/IF 11_R	CCTGGAAGGCAGCTGGG/TTGTGGGCAAACTTGTGGTA CAGAGCATACGCAGCCTGTA	388/193
13_OF/IF 13_R	TGACCTGGTATGGTCATGGA/GAAGGGGCTTCCAGGAGC CTGGGTGCAGTGGCTGC	221/203
14_OF/IF 14_R	AAGACCCAAGCTGCCTGAC/AGGGCCCCCTCTCTCCGC GCTGGGTGCAGAGCCATAT	295/266
15_OF/IF 15_R	AGTGACCGCTGCTGCCTGG/ATGGCCTGACGACTCGTGCT TTCCCAAGGGCACTGCCTGC	253/232
16_OF/IF 16_R	TGGCCAAGCTGCACAGACGG/TGTAACCTCCACCCCAAGAG TCTCCTTTACCCCTCCTTCC	231/176

bp = base pair; IF = inner forward primer; OF = outer forward primer; R = reverse primer.

This study was approved by IRB KMUH‐98‐00237and was conducted in accordance with the ethical standards of Institutional Review Board (IRB) Committee of Kaohsiung Medical University Hospital (KMUH).

Results

Within the period of 1997–2005, the incidence of newly‐diagnosed thyroid carcinoma in our hospital was around 72–90 patients/year. The incidence of MTC accounts for about 1% of all thyroid carcinoma in our hospital. All of these seven cases were still alive until 2015 with different clinical courses. As shown in Table 1, all patients survived 10 years or more, even though one was diagnosed initially with Stage IVA at 76 years old. The extrathyroid extension of MTC, which occurred in Case 1 and Case 4, was documented by initial pathologic report. The cancer work‐up of thallium⁻²⁰¹ whole body scan was performed for Case 1 and Case 5, Tc^99m‐MIBI combined with SPECT/CT scan for Case 2, Case 3, Case 4, and Case 7, neck CT scan for Case 4, and positron emission tomography scan for Case 3 and Case 4 within the follow‐up period. There was no evidence of distant metastases detected by the above routine cancer work‐up. As substantial difficulty existed for cytodiagnosis of medullary carcinoma, the preoperation calcitonin or CEA tumor marker was measured only in Case 1 and Case 2. The periodic biochemical analysis of CEA and calcitonin is shown in the supplementary figure.

Comparing index Case 2, Case 4, and Case 5 with tumor recurrence and Case 1, Case 3, Case 6, and Case 7 without tumor recurrence, it appeared that initially positive lymph node metastasis and limited surgical exploration contributed to tumor recurrence but was not related to tumor size or extrathyroid extension. Lymph node metastasis signified a risk for local recurrence in lymph nodes even after total thyroidectomy and lymph node dissection. However, total thyroidectomy might reduce recurrence even with the initial presence of capsular invasion or extrathyroid invasion. Absence of tumor recurrence in Case 1, Case 3, Case 6, and Case 7 was compatible with normal CEA during follow‐up. Continuously elevated calcitonin or CEA was significantly correlated to the presence of tumor recurrence with lymph node metastasis.

There were eight somatic RET gene mutations found in our seven patients (Figure 1). The prevalence of RET gene mutation was near 86% in our population. There were three mutations at the hot spot codon p.C634S (TGC > AGC, TGC > AGC, TGC > TCC) in exon 11. Case 4 was screened with p.C620R (TGC > CGC, homozygous) in exon 10 and Case 2 had a mutation at M918T (ATG > ACG) in exon 16. Patients with somatic mutation at p.C634S did not have recurrence during more than 10 years of follow‐up. However, cases with somatic mutations of p.M918T or p.C620R experienced much worse outcomes in terms of frequent recurrence (Table 1). There were novel mutations screened at p.L629Q (CTG > CAG) in exon 11 in Case 1 and Case 6. Case 1 also had a novel gene mutation at p.V642I (GTC > ATC, homozygous) in exon 11 (Table 1). Referring to the novel variants at p.L629Q and p.V642I in exon 11, we made “in silico” analysis by PolyPhen‐2 and Mutation Taster. The p.L629Q was predicted as “possibly damaging” with a score of 0.668 by PolyPhen‐2 and polymorphism with a score of 113 by Mutation Taster. As for p.V642I, the PolyPhen‐2 predicted it as “benign” with a score of 0.012 and Mutation Taster predicted it as polymorphism with a score of 29.

Figure 1.

Chromatogram of RET mutations in sporadic medullary carcinoma. Sanger sequence traces of the patients with RET gene mutation were performed by ABI 3730XL DNA Analyzer. These RET gene mutations included M918T (A); V642I variant (B); C634S (C); C620R (D); and L629Q and C634S mutations (E). * indicates homozygous mutation.

Discussion

The tumor, node, metastases (TNM) classification system of the 6^th edition of American Joint Committee on Cancer (AJCC) revealed 10‐year MTC survival rates for Stage I, Stage II, Stage III, and Stage IV as 100%, 93%, 71%, and 21%, respectively [14]. Overall the reported 10‐year disease‐specific survival varies from 50% to 96.6% in different series. So far, there is no reliable and concordant prognostic model to predict the risk of relapse and death [2]. The TNM/AJCC staging system could provide adequate risk stratification for disease‐specific mortality, but is unable to predict disease persistence, biochemical persistence, or recurrence at long‐term surveillance [[2], [15], [16]]. Based on the multivariate analysis achieved by the population‐based Surveillance, Epidemiology and End Results registry in America, the advanced age at diagnosis, size of primary tumor, range of surgical exploration, lymph nodes, or distant metastases appear to determine the prognosis. However, only age and stage remained significantly to be independent predictors for prognosis on multivariate analysis [14]. Age at diagnosis is an independent predictor of survival, with an increment of 5.2% risk of death for each additional year of age. Stage of distant metastasis had the highest hazard ratio at 4.47, while regional stage (extension beyond the thyroid directly into the surrounding tissues or lymph nodes) had a hazard ratio of 2.69 for decreased survival than localized tumor confined to the thyroid. Total thyroidectomy with lymph node dissection versus lesser resection (lobectomy, subtotal thyroidectomy absent, or limited lymph node dissection) also independently predicted a better outcome [14].

Reviewing our case series, there were no distant metastases and the 10 year disease‐specific survival rate seemed better than the Surveillance, Epidemiology and End Results database. The prognosis predicted by age or advanced pathologic staging by the current TNM/AJCC system could not sufficiently determine the outcome of our cases series because of neglecting the impact of serum tumor marker (calcitonin and/or CEA) and lymph node metastasis [3]. Our case series has demonstrated that continuous elevation of the calcitonin or CEA was significantly correlated to the presence of tumor recurrence with lymph node metastasis. However, the stepwise surgical interventions achieving final total thyroidectomy and radical lymph node dissection for each case may contribute to the relatively good prognosis for our patients.

Because MTC has been reported with a high rate (50–75%) of lymph node metastases and substantial incidence of multifocal tumors (0–22%), postoperative evaluation and treatment turns out to be critical for the outcome [[17], [18], [19]]. Neck ultrasound and other imaging studies (neck or chest CT scan, bone scan) are considered to evaluate the residual or metastatic tumor burden if the detectable calcitonin is still >150 pg/mL. Then, additional therapies, e.g., bilateral total thyroidectomy with central and/or localized lymph node dissection or external beam irradiation therapy should be offered to treat persistent, progressive or symptomatic locoregional tumor burden identified as more than 1 cm [[12], [20], [21]].

According to the recommendations in the American Thyroid Association guidelines, concurrent assessment of the baseline calcitonin and CEA around 2–3 months after operation is the critical step to clarify the risk status for MTC patients [[20], [21]]. Serum calcitonin and CEA were significantly correlated with tumor burden and disease progression [21]. Serum calcitonin and CEA were highly correlated in 80% of MTC patients, while disproportionately high CEA with limited calcitonin level indicates poor differentiation of the MTC. The most aggressive tumors had persistent and intense CEA staining but minimal calcitonin staining due to loss of expression during the dedifferentiation process [[3], [11], [20]]. As reported by Barbet et al. [22], doubling time of serum calcitonin (Ct DT) was regarded as the most sensitive predictor of survival, better than calcitonin, CEA, CEA DT, or current TNM staging system. Five and 10‐year survival for MTC patients was 25% and 8% if their Ct DT was <6 months. The 5‐year and 10‐year survivals improved to 92% and 37%, respectively, if the Ct DT was within 6–24 months, while all patients could survive >10 years if their Ct DT was >2 years [22]. According to 2012 European Thyroid Association guidelines, basal serum calcitonin higher than the upper limit of the normal range should be considered as “positive” and the calcitonin cutoff should also be able to discriminate C‐cell hyperplasia or microMTC from bigger tumors and neck lymph node metastases. MTC tumors bigger than 0.5 cm with lymph node metastases are never reported to have a basal calcitonin level <30–60 pg/mL [17]. However, our case series revealed different clinical courses compared with previous conclusions. For Case 1, the preoperative CEA was low and even exhibited a huge tumor burden (5 cm with extrathyroid invasion). The raised calcitonin or CEA in Case 2, Case 4, and Case 5 was compatible with lymph node metastases during postoperative follow‐up. Serum CEA or calcitonin in Case 6 and Case 7 declined gradually at 5 years and 1 year after operation without any evidence of recurrence during follow‐up beyond 10 years. Case 5 had a rapid Ct DT from 1.0 pg/mL up to 18.8 pg/mL within 1 year after lobectomy but is still alive with no recurrence beyond 11 years, even though complete total thyroidectomy and lymph node dissection was delayed for 2 years. From our limited case observations, we found that undetectable serum calcitonin or CEA after operation could not predict recurrence well. Also, initially detectable calcitonin or CEA just postoperation or within 3 months did not indicate tumor persistence. This is compatible with previous clinical investigation that postoperative 3 months is the optimal time to determine the nadir of serum calcitonin, but serum CEA may take even longer to reach a nadir [[3], [12]]. Thus we suggested periodical follow‐up of the tumor markers is necessary to compare serial alterations and verify biochemical remission or recurrence. In addition, serum calcitonin or CEA is used as a reference to judge tumor persistence or recurrence rather than as a predictor of survival.

The genotype of hereditary MTC is recognized to highly correlate with the phenotype [[9], [16]]. According to the American Thyroid Association guidelines, germline RET mutation at codon 918 (ATG to ACG, p.M918T) in exon 16 is associated with the “highest” risk status (level D) at younger onset, with the highest risk of metastases and disease‐specific mortality. Hereditary MTC with germline RET mutation at codon 634, encoding a cysteine‐rich extracellular domain, was classified as “high” risk level C, with higher transforming activity and tumor aggressiveness compared with the other codons (609, 611, 618, 620, and 630) [[3], [17]]. Existence of somatic RET mutations in sMTC was also found to correlate with higher mitotic cancer cell activity [[6], [9]]. Prevalence of the somatic RET mutation harbored in sMTC varies from 40% to 65% in different series. Somatic RET mutation was demonstrated to signify some clinicopathologic features [[6], [7], [9], [11]]. p.M918T somatic mutation was the most common mutation in sMTC and significantly contributed to a poor prognosis with larger tumors and higher lymph node and distant metastasis rates, frequent multifocal tumors, more advanced stage at diagnosis, more persistent disease after operation, more relapse, and worse outcomes [[6], [9], [17]]. In our case series only one patient (Case 2, 14%) harbored the M918T mutation, which was clinically compatible with recurrent lymph node metastases even after initial wide total thyroidectomy and lymph node dissection followed by two more lymph node dissections. Our index Case 4 harboring a somatic p.C620R mutation presented initially with extrathyroid extension, treated by left lobectomy, and experienced recurrent lymph node metastases. Total thyroidectomy and radical lymph node dissection was done 8 years later, and the outcome was still good with no tumor recurrence at the study end. All three cases (43%) having a C634S mutation had good prognosis and no recurrence beyond 10 years. Since they all were treated with bilateral total thyroidectomy at first, the clinical significance of the C634S mutation is hard to assess by our limited case experience.

Limitations of this study included small MTC cases in our thyroid cancer population. The diagnosis of sporadic MTC by clinicians was only based on the absence of family history or any associated endocrine tumor emerged among three generations within 10 years of follow up. Peripheral blood samples were not drawn to detect germline mutations at the time of data collection. Functional study was not elucidated in the pathogenesis of sMTC for the two novel variants found in this study.

Our study evidenced a relatively high frequency (86%) of the somatic RET gene mutation in sMTC. We found no obvious correlation between the somatic RET mutations and 10‐year survival outcome in cases of sMTCs. However, sMTC patients carrying the M918T mutation may have greater risk of lymph node metastases and may need more aggressive surgical intervention during follow‐up. Intensive treatment with total thyroidectomy and repeated neck lymph node dissection seemed to be the critical determinant to achieve better survival outcome for sMTC patients. However, a randomized control study should be programmed with more patients recruited for a definite conclusion. From our results, we agreed with the guideline that there is still no strong evidence or indication for routine screening of the somatic RET mutation to determine treatment strategies for sMTC patients.

Acknowledgments

We thank the patients and their families for participating in this research study. This study was approved by IRB KMUH‐98‐00237 and supported by grant KMU‐TP103D15 (to Daw‐Yang Hwang) and the Kaohsiung Medical University Hospital Nephrology Research Fund. The authors acknowledge the technical services provided by the Sequencing Core Facility of the National Yang‐Ming University Genome Research Center. The Sequencing Core Facility is supported by the National Research Program for Genomic Medicine, Ministry of Science and Technology, Taiwan, R.O.C.

Supporting information

Supplementary data

Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.kjms.2016.08.012.