Management of Medullary Thyroid Carcinoma

Review of the Management of Medullary Thyroid Cancer in New Zealand: pre-operative calcitonin to guide the extent of surgery

INTRODUCTION

Medullary thyroid carcinoma (MTC) originates from parafollicular C-cells and represents 6-8% of all thyroid cancers. [1] Approximately 25% of cases are inherited as an autosomal dominant disorder, with MEN2A syndrome [2] It is an indolent disease with a 10 year survival of 69-89% [3-5] but spreads early to regional lymph nodes compared with differentiated cancers and occasionally spreads to distant sites early, prior to diagnosis. Most commonly MTC is diagnosed as a solitary thyroid nodule or palpable lymph node, although sometimes they are discovered incidentally on imaging performed for other indications. [6]

Challenges in management include distinguishing MTC from benign nodular disease and differentiated thyroid cancer; and managing post-oeperative residual MTC in asymptomatic patients. [6] Complete resection is important in curing MTC as parafollicular cells do not concentrate iodine therefore 131-I is not an option. [7] There is also a limited role for radiation in MTC. [6] Patients at high risk for local recurrence have a reduced 10 year local recurrence with external beam radiation therapy. [8]

Central and ipsilateral node nodal involvement occurs in approximately 50% of patients, whereas contralateral node involvement occurs in 25-30%. [9] >75% of patients with palpable primary tumours have cervical lymphadenopathy. [10] Nodal involvement is associated with a lower rate of biochemical cure, distant metastases and poorer long term survival. [2, 5] Standard preoperative imaging with ultrasonography (USS) can overlook lymph node metastases, as they are highly operator dependent and have low sensitivity, with 32% rate of false negative central neck examinations and 14% rate of  false negatives in the ipsilateral neck. [11] Involved lymph nodes have an abnormal appearance, including abnormal overall morphology, loss of fatty hilum, and calcifications. These may be marked on the overlying skin, and the surgeon should review the ultrasound findings with the radiologist. [4] Clinically involved lymph nodes will be missed in over a third of patients with reliance on intraoperative lymph node assessment. [10] False negative USS’s are higher for patients having reoperation (44%) for recurrent MTC. Lymph node spread occurs in a stepwise fashion with the ipsilateral central compartment (level VI) most often involved, followed by levels II through V on the ipsilateral neck, then the contralateral cervical neck and mediastinal lymph node compartment. [12] Distant compartment involvement correlates with presence of gross distant metastases. [13] Distant metastases occur in the liver, lungs, and bone are often not detectable on imaging. [4]

MTC cells produce carcinoembryoantigen (CEA), and calcitonin (CT). CT is superior to CEA in detecting C-cell pathology and extent of disease. [12] Additional imaging to investigate nodal involvement can be guided by the degree of CT elevation. [6] If postoperative serum CT is detectable then it confirms presence of persistent or recurrent tumour even if it is not clinically evident. Disease progression correlates with serum CT. [14] Normalization of post-operative CT levels is associated with increased disease free survival. [15]

Although surgery is the only effective cure, the evidence for an optimal primary operative management is lacking. Inadequate operation leads to persistent disease and disease recurrence, which is associated with re-operative surgery that rarely achieves biochemical cure and causes significant morbidity. Total thyroidectomy and central compartment is recommended when there is no evidence of local invasion or regional or distant metastases. [16]  With clinical or radiological evidence of lymph node metastases then compartment oriented surgery is recommended as survival rates and disease recurrence rates are improved compared with procedures with removal of only macroscopically involved nodes (‘berry picking’). [16] The extent of lymph node clearance performed should be a trade-off between surgical morbidity and clearance of disease.

With increasing biomarker levels, number of positive lymph nodes and lymph node compartments involved, the chance of biochemical cure reduces. [9, 13, 17] Some groups have attempted to utilize preoperative CT levels to predict the extent of disease and to tailor the extent of operation. [18, 17, 19] The Revised American Thyroid Association guidelines (2015) have specific recommendations on the extent of neck dissection to be performed in addition to thyroidectomy based on preoperative CT. 

The aims of the study were to review the management of MTC patients in several institutions in New Zealand including both major tertiary centres and provincial centres, against the current body of evidence and recommendations in the ATA guidelines. We sought to corroborate our institutional experience with existing studies using preoperative calcitonin levels in individualizing treatment. We evaluated any trends between preoperative serum biochemical marker levels (calcitonin and CEA), tumour characteristics, extent of disease, biochemical cure and recurrent or persistent disease as well We determined the utility of pre-operative calcitonin in determining lymph node positivity, distant metastases, biochemical cure and recurrent or persistent disease. Finally, we sought variables which were associated with biochemical cure and disease free survival. 

METHODOLOGY

A retrospective review of clinical notes was conducted on patients treated for MTC at seven institutions (Auckland City Hospital, North Shore Hospital, Waikato Hospital, Palmerston North Hospital, Taranaki Hospital, Christchurch Hospital and Dunedin) from 1999-2016. Patients with metastatic disease on initial presentation and patients that did not undergo operative management were included in the study. Patients diagnosed with recurrent disease but did not have initial operative treatment in the study period were excluded. Patients less than 15 years of age were excluded. Diagnosis method, and indication for neck dissection if performed was determined from clinical notes. Presenting features other than a thyroid/neck lump or a known MEN patient having prophylactic thyroidectomy were recorded. Follow up length was calculated from the date of surgery to the last date of follow up or date of death.

Hereditary MTC

Hereditary MTC was defined as MTC with an identified germline rearranged during transfection (RET) mutation. Familial MTC was defined as a phenotypic variant of MEN2A affecting at least 2 generations, without phaeochromocytomas or primary hyperparathyroidism.

Biochemical markers

Serum calcitonin was measured using either of the following immunoassays: Roche Electro chemiluminescence Immuno assay (reference range, <6 for women; <10 for men) or a DiaSorin LIASON chemiluminescence Immuno assay (reference range, <5 for women; <12 for men). CEA was measured using one of the following techniques: Roche chemiluminescence Immuno assay (reference range < 3.4ng/mL);  Seimens ADVIA centaur XP chemiluminescent Immuno assay (reference range, <3.4 ng/mL before 2007; <2.5 ng/mL after 2007); Architect Immuno assay (reference range <3ng/mL).

Surgery

The approach was decided on surgeon preference with not all performing compartment-oriented dissection. Patients that underwent two staged operation for completion lobectomy with or without lymph node dissection; or a thyroidectomy with or without further neck dissection following an incidental metastatic MTC found during neck dissection had their operations considered as one single sequence for initial management. The extent of operation and side of each compartment-oriented neck dissection was recorded from operation and pathology reports. Perithyroid nodes excised during thyroidectomy were included in the lymph node count and lymph node positivity analysis but when performed alone were not counted as a formal central compartment neck dissection. Patients that did not have lymph node dissection were considered lymph node negative if recurrence did not develop within 12 months. Undissected lymph node compartments were also considered negative during analysis.

Histopathology

Fine needle aspirate reports were reviewed for diagnostic suspicion and Bethesda score. Tumour size, number of lymph nodes harvested, number of positive lymph nodes, lymph node compartments involved, tumour size and presence of the following characteristics: multifocality, lymphovascular invasion, extrathyroidal extension, c-cell hyperplasia and immunochemical staining were ascertained from pathology reports. Distant compartment involvement was defined as involvement of the contralateral cervical neck and/or superior mediastinal lymph nodes. For patients with multiple foci of MTC, the largest tumour dimension was recorded. Features that were not documented (other than staining) were recorded as negative. Tumour staging was defined using the American Joint Committee on Cancer (AJCC) TNM staging criteria.

Disease outcomes

Recurrence was defined as local or cervical recurrence and/or distant metastatic recurrence. Persistent disease was defined as inoperable, distant metastases at initial operation and/or positive resection margins. Patient disease status was categorized by those alive with no structural evidence of disease (NED), alive with no structural disease but persistently raised calcitonin, alive with disease (AWD), dead of disease (DOD) or dead from other cause. Disease free survival was calculated from the date of surgery to date of first recurrence or date of last follow up for those remaining free of disease. For patient with persistent disease DFS was zero.

Statistics

Univariate analyses included Student’s t test, and Pearson’s correlation coefficients. Serum biomarkers were considered continuous independent variables for simple correlation analyses with tumour size and number of positive lymph nodes. Biochemical cure (with undetectable calcitonin) was considered on a per-patient basis, based on the type of assay performed. Logistic regression was used to seek factors associated with failure of biochemical cure. Age group was considered an ordinal variable, increasing with increments in age (ages 1-34, 35-54 and 55-101), while T stage (T1-T2 versus T3-T4) and AJCC stage grouping (I-II versus III-IV) were considered binary variables for logistic regression and DFS analysis. The Kaplan-Meier method was used to estimate DFS associated with relevant clinicopathological variables. Differences in the distributions were assessed by the log-rank test. Tumour size and number of lymph nodes were not included in survival analyses due to the size of the study. A multivariate analysis was performed using a Cox proportional hazards model after adjusting for variables based on clinical relevance and significance of variables on univariate analysis. 

RESULTS

Patients and pathology

There were 63 patients in the study. 30 were male, and 33 were female. Mean age was 50 (IQR = 37-62). Figure 1 shows the number of patients from each institution. There were 17 primary surgeons in total. The mean number of patients per surgeon was 3-4 (range; 1-11). 12 patients had hereditary MTC; 11 with MEN2a and 1 with familial MTC. The median tumour size was 25mm (IQR = 16-40). Table 1a-b displays clinicopathological variables that were reviewed. A fine needle aspirate (FNA) was performed in 46 patients, of which 25 were diagnostic of MTC. 34 patients gained histological diagnosis with partial or total thyroidectomy, of these, seven patients had raised biomarkers or known family history of RET mutation prior to surgery. Two patients presented with Cushing’s syndrome while two patients were initially discovered with phaeochromocytoma leading to investigation for MTC. Three patients had incidental thyroid nodules discovered on CT or FDG-PET, performed for other indications. Two patients had inoperable disease, one of which had invasion into the oesophagus determined to be unresectable following endoscopic and radiological evaluation while the other had cervical, manubrial, lung and liver metastases. Median length of follow up was 45 months (IQR=13-88 months).

Table 1a. Patient characteristics, diagnosis and stage

Patient characteristic	Measure
Age of presentation, mean +/- SD [range]	50 +/- 17.8 [15-101]
Male, n (%)	30
Female, n (%)	33
RET testing performed, n (%)	48 (76)
Presence of hereditary MTC, n (%)	12 (19)
—MEN2 (n=)	11
—Familial MTC (=)	1
Histological diagnosis	Total, n (%)
FNA	25 (39)
Core biopsy cervical node	3 (5)
Excision biopsy cervical node	1 (2)
Partial or total thyroidectomy	34 (54)
—-Biomarkers raise suspicion prior to surgery	7/34
—-Known MENIIA or family history of RET mutation prior to prophylactic surgery	5/34
FNA Bethesda score	Total (n=)
FNA’s performed	46
I	7
II	2
III	4 (Papillary neoplasm = 2, atypical epithelial cells = 2)
IV	5 (Hurthle cell =1, follicular neoplasm = 4)
V-VI	25
Other	Non-small cell carcinoma on metastatic neck node
TNM Staging	Total, n (%)
T1	22 (35)
T2	21 (33)
T3	13 (21)
T4a	7 (11)
N0	25 (40)
N1a	9 (14)
N1b	22 (35)
M0	57 (90)
M1	8 (10)
Stage grouping	Total, n (%)
I	14 (22)
II	10 (16)
III	10 (15)
IVa	20 (32)
IVb	1 (2)
IVc	8 (13)

Table 1b. Pathology variables and immunohistochemistry

Pathology variables
Median tumour size (mm), mean +/- SD [IQR]	30 +/- 20 [16-40]
Multifocality, n (%)	18 (30)
Lymphovascular invasion, n (%)	18 (30)
Extrathyroidal extension, n (%)	12 (20)
C-cell hyperplasia, n (%)	11 (18)
Margins positive, n (%)	11 (18)
Number of lymph nodes harvested, mean +/- SD [range]	34 +/- 28 [1-115]
Number of positive lymph nodes, mean +/- SD [range]	11 +/- 17 [0-86]
Immunochemistry	Total, n (%)
Staining (congo), n (%)	18/23 (78)
Staining (calcitonin), n (%)	48/52 (92)
Staining (CEA), n (%)	13/19 (68)
Staining (chromogranin a), n (%)	29/34 (85)
Staining (TTF-1), n (%)	19/26 (73)
Staining (cytokeratin), n (%)	11/20 (55)

Operative management

61 patients underwent surgery of which 1 patient underwent hemithyroidectomy and 8 patients underwent total thyroidectomy for a solitary thyroid nodule, without further lymph node management. (Table 2). 9 patients underwent diagnostic hemithyroidectomies. Of the 52 patients that underwent thyroidectomy and neck surgery, 3 underwent selective excision of nodal metastases whilst 56 underwent compartment-oriented neck dissections. Of the 56 patients that underwent compartment-oriented neck dissections, 43 patients underwent central lymph node dissection, 6 of which were ipsilateral, 29 were bilateral, and 5 patients did not have the side of central lymph node dissection recorded in the operation note or clinical notes. 4 patients had a selective excision of lateral compartment lymph nodes in addition to central lymph node dissection. 4 (in 56 patients) had one or two lateral compartments dissected without central lymph node dissection. Indications for neck dissection are detailed in Table 3.

Table 2. Procedures performed (n= 61)

Procedure	Number	Extended surgery performed	Central lymph node dissection site (n=)
Hemithyroidectomy	1
Total thyroidectomy (TT)	11*
Central lymph node dissection (CLND)	19**	Superior mediastinal lymph node dissection (n=1)	Ipsilateral (3)   Bilateral (15) Unknown: 1
Ipsilateral cervical neck dissection	14§   (10x SND, 1xMRND)	Superior mediastinal lymph node dissection (n=3)   Excision of oesophageal margin (n=1)	Ipsilateral (3)   Bilateral (5) Unknown (3)
Bilateral cervical neck dissection	16¶   (9 x bilateral SND, 2 x bilateral MRND, 5 x ipsilateral MRND + contralateral SND)	Superior mediastinal lymph node dissection (n=5)	Ipsilateral (3)   Bilateral (10) Unknown (1)

SND = selective neck dissection; MRND = modified radical neck dissection

*3 TT patients underwent lymph node sampling in the ipsilateral central or lateral compartment

**4 patients who underwent central lymph node dissection underwent lymph node sampling in the ipsilateral cervical neck

§All ipsilateral cervical neck dissection underwent CLND except 3

¶All bilateral cervical neck dissection patients underwent CLND except 1

Table 3: Indications for neck dissection (n=)

Known MENIIA or family history of RET mutation	5
Radiological evidence of metastases	12
Palpable metastases	13
Frozen section of palpable cervical disease at time of operation	4
Incidental metastatic disease during neck dissection for other H&N cancer	2
No clinical or radiological evidence of cervical metastases with calcitonin > 200	10
No clinical or radiological evidence of cervical metastases with calcitonin <200 (or no preoperative calcitonin performed)	5

Lymph node metastases

44 patients had cervical lymph node metastases; 25 with central compartment involvement and 26 with lateral neck involvement (Table 4). Of the 12 patients with hereditary MTC, 5 patients had positive cervical lymph nodes. 15 in 19 patients with positive central nodes who underwent CND with ipsilateral lateral neck dissection had ipsilateral lateral neck involvement. 5 in 14 patients with positive central nodes identified with central compartment dissection or selective node sampling did not have further neck dissection performed.  Of those who underwent central and bilateral lateral neck dissection, 11 in 13 patients with ipsilateral metastases had contralateral metastases. 3 patients who underwent CND with ipsilateral or contralateral lateral neck dissection had a ‘skip’ nodal metastasis where there was lateral compartment involvement without central compartment metastases. One patient who underwent CND with bilateral lateral neck dissection had contralateral metastases without ipsilateral lateral neck metastases. 15 patients had distant compartment involvement, 6 of whom also had distant metastases at time of operation. Average lymph node counts from each compartment cannot be described due to incomplete documentation in operative or pathology notes.

Table 4. Lymph node positivity (n=)

Lymph node dissections	52
Total with cervical lymph node metastases	44
Total with central compartment involvement	26
—Central compartment alone	9
Total with lateral neck involvement	26
—Ipsilateral alone	7
—Central and ipsilateral	10
—Central and contralateral	1
—Ipsilateral and contralateral	3
—Central, ipsilateral and contralateral	5
Total with superior mediastinal involvement	5

Calcitonin

44 patients had recorded pre-operative calcitonin levels while 57 patients had recorded post-operative calcitonin levels. On univariate analysis, preoperative calcitonin was associated with lymph node positivity (p=0.005), distant compartment involvement (p=0.020), biochemical cure (p=0.010) and recurrent or persistent disease (p=0.009) (Table 5). Presence of multifocality, lymphovascular invasion, extrathyroidal extension, c-cell hyperplasia, positive margins and distant metastases was not found to be associated with pre-operative calcitonin level. Neither tumour size (p=0.204) nor number of positive lymph nodes (p=0.107) correlated with pre-operative calcitonin levels.

Central metastases occurred in one patient with undetectable CT levels, while ipsilateral lateral neck nodes appeared at a CT level of 61. No patients with contralateral cervical neck metastases or superior mediastinal metastases had a CT less than 1100 (range; 1100-67746). A few node-negative tumours were seen with very high basal CT; one patient had a pre-operative CT level of 140000. Distant metastases appeared at a CT level >2250 in our study.

Median pre-operative calcitonin was 5002 [IQR=1243-14004] for patients with lymph node metastases and 60 [IQR=15-591] for patients without lymph node metastases.

Median pre-operative calcitonin was 9800 [IQR=2250-22500] for patients with distal compartment involvement (Table 6).

Table 7 shows pre-operative and post-operative calcitonin levels stratified into three groups: persistent disease, recurrent disease, and no persistent or recurrent disease. 9% (1 in 11) of patients with recurrence achieved biochemical cure with the initial management.

ROC curve analyses (Fig. 2a-d) determined that preoperative calcitonin predicts lymph node positivity (AUC=0.831), likelihood of biochemical cure (AUC=0.874), distant metastases (AUC=0.829) and recurrent or persistent disease (AUC=0.852).

Table 5. Preoperative serum biomarkers, pathological characteristics and lymph node involvement

Preoperative calcitonin (pg/mL)	Yes (mean calcitonin; range)	No (mean calcitonin; range)	p value
Multifocality	10233   (730-38505)	6190   (1.9-67746)	0.309
Lymphovascular invasion	7677   (20-40181)	7760   (1.9-67746)	0.984
Extrathyroidal extension	14200   (2250-38505)	5966   (1.9-67746)	0.090
C-cell hyperplasia	4222   (1.9-22500)	8687   (1.9-67746)	0.219
Positive resection margins	13200   (975-40181)	6238   (1.9-67746)	0.192
Lymph node positivity	10739   (1.9-67746)	1417   (1.9-14000)	0.005
Distant compartment involvement	17106   (1100-67746)	3138   (1.9-18600)	0.020
Failure to achieve biochemical cure	13770   (975-67746)	1921   (1.9-14300)	0.010
Distant metastases	15976   (447-40181)	7979   (1.9-81600)	0.287
Recurrent or persistent disease	17471   (447-81600)	2388   (1.9-14300)	0.009
Preoperative CEA (ng/mL)	Yes (mean CEA; range)	No (mean CEA; range)	p value
Multifocality	326   (2.9-3550)	234   (0.9-1500)	0.728
Lymphovascular invasion	158   (7.1-479)	333   (0.9-3550)	0.308
Extrathyroidal extension	164   (7.1-381)	296   (0.9-3550)	0.362
C-cell hyperplasia	61   (2.4-226)	321   (0.9-3550)	0.061
Positive resection margins	552   (7.1-3550)	189   (0.9-1500)	0.457
Lymph node positivity	353   (0.9-3550)	89.3   (2.4-605)	0.061
Distant compartment involvement	527   (7.1-3550)	164   (0.9-1500)	0.322
Distant metastases	773   (14.4-3550)	225   (0.9-2104)	0.374
Recurrent or persistent disease	497   (7.1-3550)	190   (0.9-1500)	0.266

Table 6. Biomarker levels and distant compartment involvement

Laboratory variables	Distant compartment involvement; contralateral lateral neck and/or superior mediastinal   (IQR)
Median preoperative calcitonin* (pg/mL)	9800   (2250-22500)
Median postoperative serum calcitonin** (pg/mL)	662   (98-3790)
Biochemical cure, n (%)	0%   0 in 14
Median preoperative CEA§ (ng/mL)	202   (27.9-381)
Median postoperative CEA¶ (ng/mL)	2.3   (3.5-66.4)

*14 recorded pre-operative calcitonin levels

**15 recorded post-operative calcitonin levels

§10 recorded pre-operative CEA levels

¶15 recorded post-operative CEA levels

Table 7. Biomarker levels and persistent disease or recurrence

Laboratory variables	Total (IQR)	Persistent disease (IQR)	Recurrent; cervical and/or distant metastatic recurrence (IQR)	No persistent disease and no evidence of recurrence (IQR)
Median preoperative calcitonin* (pg/mL)	1830   (83-9250)	7498   (1770-20700)	7028   (5954-13905)	281   (19-2075)
Median postoperative serum calcitonin** (pg/mL)	9.9   (undetectable-300)	300   (89-1130)	201   (40-1748)	undetectable   (undetectable – undetectable)
Biochemical cure, n (%)	47%   (27 in 57)	0 %   (0 in 13)	9%   (1 in 11)	78%   (25 in 32)
Median preoperative CEA§ (ng/mL)	10   (70-296)	105   (29-299)	204   (182-226)	18.4   (3.5-235)
Median postoperative CEA¶ (ng/mL)	3   (1.5-10)	11   (2.3-95.9)	3.1   (1.6-26)	1.9   (1.1-4.0)

*44 recorded pre-operative calcitonin levels

**57 recorded post-operative calcitonin levels

§38 recorded pre-operative CEA levels

¶46 recorded post-operative CEA levels

CEA

38 patients had recorded pre-operative CEA levels while 46 patients had recorded post-operative calcitonin levels. On univariate analysis, no clinicopathological variables were found to be associated with preoperative CEA, including multifocality, lymphovascular invasion, extrathyroidal extension, c-cell hyperplasia, positive resection margins, lymph node positivity, distant compartment involvement, distant metastases and recurrent or persistent disease (Table 5). Tumour size and number of positive lymph nodes did not correlate with pre-operative CEA levels (p=0.159, p=0.290). 9 in 21 patients with CEA greater or equal to 30 (range; 30-1500) achieved biochemical cure. 10 in 14 patients with CEA <30 achieved biochemical cure.

Median pre-operative CEA was 105 (IQR=17-340) for patients with lymph node metastases and 14 (IQR=3-70) for patients without lymph node metastases. Median pre-operative CEA was 202 (27.9-381) for patients with distal compartment involvement (Table 6).

Table 7 shows pre-operative and post-operative CEA levels stratified into three groups: persistent disease, recurrent disease, and no persistent or recurrent disease.

Biochemical cure

Following operative management, 18 of 22 patients without lymph node metastases achieved biochemical cure (median follow up 64 months; IQR 23-78) while 8 of 35 patients with lymph node metastases achieved biochemical cure (median follow up 62 months; IQR 11-94 months), but there was no statistically significant difference. 20 of 22 patients who had greater than 6 lymph nodes did not achieve biochemical cure.  No patients with distant compartment involvement achieved biochemical cure. The percentage of patients who attained biochemical cure in Stage I disease was 11/13; and in stage Iva-b 3/18. Of the 16 patients that underwent bilateral central and lateral compartment surgery, three patients were cured, with a CT <472, and 13 patients were not, with a CT >1100.Using logistic regression, no variables were found to be associated with failure to achieve biochemical cure including age group, hereditary MTC, gender, multifocality, lymphovascular invasion, c-cell hyperplasia, positive margins, T stage (T1-2 versus T3-4) and stage grouping (I-II versus III-IV).

Prognosis and survival outcomes

We only recorded 7 MTC related deaths (5 of whom had distant metastases at presentation). Table 8 shows the stage distribution of patients, data for persistent disease and recurrent disease, and the vital status of patients at last follow up date. Of the 48 patients that had R0 resections, 12 patients had recurrent disease, 11 of which had local and/or cervical recurrences. 16 patients had persistent disease defined as metastatic on initial presentation (n=8), inoperable (n=2) and R1 or R2 resections (n=6). Of the 8 patients with distant metastases at initial presentation, 7 underwent surgery for local control of disease.

On univariate analysis using the Kaplan-Meier method, male gender, c-cell hyperplasia, lymphovascular invasion, positive resection margins, failure to achieve biochemical cure, T stage (T3-4 versus T1-2), lymph node positivity and stage group (III-IV versus I-II) correlated with poorer disease free survival (Fig. 3a-h). Age, hereditary versus sporadic MTC, presence of multifocality or extrathyroidal extension did not correlate with DFS. Table 9 shows the hazard ratios of these clinicopathological variables in relation to DFS. The 5-year DFS for patients who achieved biochemical cure versus no biochemical cure was 100% and 41% respectively (p<0.0001). The 5 year DFS for negative lymph nodes and positive lymph nodes was 100% and 48% respectively (p-0.0002). After multivariate adjustment, only gender was associated with DFS with males having poorer DFS than males (hazard ratio= 0.097, p=0.023).

Table 8. Stage distribution of cohort, persistent disease, recurrence data, and vital status of patients (n=)

Stage	Total	Total persistent disease	Total recurrent disease	Local and/or cervical recurrence	Distant metastatic recurrence	CR and DM	NED	NED without biochemical cure	AWD	DOD	Dead, other
I	14	0	0	0	0	0	13	1	0	0	0
II	10	0	2	2	0	0	6	2	0	0	2
III	10	0	2	2	1	0	7	2	1	0	0
IV	29	16	8	6	0	2	1	7	12	7	2
All	63	16	12	10	1	2	27	12	13	7	4

CR = cervical recurrence; DM = distant metastatic recurrence; NED = no evidence of structural or radiological disease; AWD = alive with disease; DOD = dead of disease

Table 9. Clinicopathological variables and disease-free survival

	Hazard ratio (95% CI)	P value
Age group 1: 1-34   Group 2: 35-54 Group 3: 55-101	1:2 = 0.805 (0.249-2.60)   1:3 = 0.915 (0.259-3.23) 2:3 = 1.36 (0.368-3.50)	0.933
Gender (F:M)	0.097 (0.037-0.254)	0.0001
Hereditary MTC	0.5182 (0.160-1.68)	0.371
Multifocality	1.82 (0.629-5.25)	0.214
Lymphovascular invasion	8.32 (2.66-26.0)	<0.0001
Extrathyroidal extension	2.07 (0.580-7.39)	0.157
C-cell hyperplasia	11.94 (2.56-55.8)	<0.0001
Positive resection margins	10.5 (2.10-52.1)	<0.0001
Biochemical cure	-*	<0.0001
T stage (T3-4:T1-2)	5.39 (1.81-16.1)	<0.0001
Stage group (III-IV:I-II)	-*	<0.0002
Lymph node positivity	-*	0.0002

*All patients with biochemical cure, stage I-II, and no lymph node positivity were disease free up to last date of follow up

Management according to ATA guidelines

According to the American Thyroid Association guidelines, 28 out of 54 patients with MTC (excluding those with inoperable disease and/or distant metastases) received inadequate initial operative management (Table 10). 9 of 10 patients, with no evidence of metastases but a raised calcitonin over 60pg/mL received inadequate treatment, with 3 developing a recurrence. 11 of 21 patients with evidence of cervical metastases and a calcitonin level >200pg/mL received inadequate treatment according to the ATA guidelines, with 4 patients eveloping a recurrence while 10 of 21 patients with evidence of cervical metastases and a calcitonin level >200pg/mL received adequate treatment including central neck and bilateral cervical neck dissection, with 4 patients developing a recurrence.

Table 10. Management according to ATA guidelines.

MTC American thyroid association category	Recommended level of neck dissection	Adequate according to ATA	Inadequate according to ATA
Hemithyroidectomy	For presumed sporadic MTC completion thyroidectomy is   recommended in patients with a RET germline mutation, an elevated postoperative serum Ctn level, or imaging studies indicating residual MTC. (Recommendation 28)*	1
No evidence of metastases and normal calcitonin	Central compartment dissection   (Recommendation 24)*	7	5   5 total thyroidectomy only
No evidence of metastases with raised calcitonin (no guideline consensus on appropriate level)	Central and ipsilateral neck dissection (II-V)   (Recommendation 25)*	1	9   2 total thyroidectomy only 7 total thyroidectomy with CND
Cervical metastases	Central + ipsilateral neck dissection (II-V)   (Recommendation 26)*	6	2   1 ipsilateral MRND without CND 1 ipsilateral levels 2,3 with CND
	Plus contralateral neck dissection if calcitonin >200pg/mL   (Recommendation 26)*	10	11   1 CND only 1 CND + ipsilateral level 4 only 1 CND + superior mediastinal node dissection only 4 CND + ipsilateral neck dissection 2 CND + ipsilateral neck dissection + superior mediastinal node excision 2 ipsilateral MRND without CND
Locally advanced MTC	Less aggressive surgery in the central and lateral neck to preserve daily function. Consider external beam radiotherapy, systemic medical therapy, and other   nonsurgical therapies to achieve local tumor control. (Recommendation 27)*	1	0
Inadequate lymph node dissection at initial surgery	Needs repeat compartment-oriented dissection if pre-operative calcitonin <1000pg/mL and 5 or fewer metastatic lymph nodes were removed at initial surgery   (Recommendation 28)*	0	1   3 positive central neck nodes removed without formal neck dissection

SND= selective neck dissection; CND= central neck dissection; MRND= modified radical neck dissection; *Recommendations by revised ATA guidelines 2015

DISCUSSION

While the literature and ATA guidelines [16] are in favour of systematic neck dissection based on suspicion of nodal disease and raised CT levels, this study provides results that strongly suggest that a high percentage of MTC patients have undergone incomplete surgery. This may be explained by a high proportion of patients having initial misdiagnosis, with only a third of patients being identified by FNA, and the relatively limited exposure of single institutions and surgeons with this rare disease. 12 patients had MTC associated with a hereditary syndrome, all with germline RET mutations, amongst whom 5 had lymph node positivity. A subgroup analysis was not performed due to the low number of patients in this group. For hereditary MTC, knowledge of the specific RET mutation allows clinicians to assess risk based on age of onset and aggressiveness of MTC by the latest American Thyroid Association guidelines on MTC. Further recommendations for hereditary MTC are based on age, size of tumour, presence of palpable or radiologically identifiable lymphadenopathy and CT level. [20]

Treatment of MTC is challenging as identification of neck metastases is difficult and adjuvant treatment if of limited benefit.  USS detection practiced in the community are believed to be inferior as compared with specialized centres. [12] In this study, patients were often symptomatic, node-positive, and had advanced disease. Early detection of MTC is important as stage is associated with failure of normalization of CT [7] Elisei reported that  pre-operative CT has a greater sensitivity and specificity than FNA, achieving earlier MTC diagnosis, a greater rate of post-operative cure and improved survival. [21] Incorporating CT into the routine work up for thyroid nodules seems justified. The European Thyroid Association recommends screening with CT but this is still controversial within the ATA. [22]

Subclinical disease or occult lymph node metastases although it has a high risk for local recurrence and dissemination, is difficult to treat. [12] A compartment-oriented neck dissection is more effective at achieving local control, but an aggressive approach with extensive nodal dissection exposes the patient to increased morbidity. The need for systematic compartment dissection at least in the central neck compartment is now largely recognized, but there is  lack of data to guide the extent dissection in the lateral neck compartments [10], especially on the contralateral side of largest primary thyroid tumour [23]. This is because complete dissections of central and both lateral neck compartments have seldom been performed. In our study, this was performed in 16 patients.

Selection criteria for surgery, and surgical approach varied between different institutions in this study. Nonetheless there was evidence of a lack of consistency in approach to neck dissection by surgeons and institutions based on clinicoradiological findings and biomarkers. A number of patients underwent surgery with ‘selective nodal excision’, which is now considered to be palliative at most [24]. Patients frequently underwent ipsilateral or bilateral neck dissection without central node dissection which is inadequate based on the hypothesis of stepwise dissemination of disease starting from the central compartment. [12] Lymph status was not available in many cases as systematic lymph node dissection and meticulous histopathlogical analysis was not performed.  This was compounded by a lack of documentation when reporting sites that were dissected. The total mean yield from lymphadenectomy is 18. 3 in Surveillance, Epidemiology, and End Results program (SEER) data. [25] In our study the mean number of nodes harvested was 34 (range; 1-115) with a mean of 1 positive nodes (range; 0-86).

Failure of biochemical cure relates to lymph node positivity. It ranges from 89-95% for patients with negative nodes and 33-40% for positive nodes performed by an experienced surgeon. [9, 19, 10, 26] This data may not be applicable to individual patients. Lack of lymph node involvement even with large tumours is important as it can still predict biochemical cure and survival. In our study, 18 of 22 patients without lymph node metastases achieved biochemical cure (median follow up 64 months; IQR 23-78) while 8 of 35 patients with lymph node metastases achieved biochemical cure (median follow up 62 months; IQR 11-94 months), but there was no statistically significant difference. Scollo et al concluded that involvement of 10 or more lymph nodes or more than 2 lymph node compartments almost precludes normalization of CT [9]. The presence of more than 10 positive lymph nodes reduces recurrence free survival and disease specific survival (DSS). [12] In our study, 20 of 22 patients who had greater than 6 lymph nodes did not achieve biochemical cure.

Cervical lymph node metastases occur frequently. 44 of 52 patients who underwent lymph node dissection had positive lymph nodes. Moley et. al reported a series of 73 patients with palpable MTC who underwent thyroidectomy with concurrent or delayed central and bilateral lymph node dissection, for unilateral intrathyroid tumours. Lymph node involvement occurred in 81% in the central neck and 81% in ipsilateral II-V and in 44% in contralateral II-V. With bilateral intrathyoridal tumours, metastases were present in 78% in the central neck and 71%  in ipsilateral II-V, and in 49% contralateral II-V. [10] Once the central compartment is involved, there is a high likelihood of ipsilateral neck (81%). [7] In our study, 15 in 19 patients with positive central nodes who underwent CND with ipsilateral lateral neck dissection had lateral neck involvement on the ipsilateral side. However, a number of patients in our study with positive central nodes did not have further neck dissection performed (5 of 14).

Machens et al. showed the association between number of positive nodes in the central and lateral neck involvement in MTC. [23]  In patients who have one positive central lymph node, additional systematic dissection of the ipsilateral lateral neck compartment is mandatory because the risk of involvement is >70%. [23] With 4 or more central nodes, the rate is 98%. With one or more positive central nodes, the rate of contralateral lateral neck involvement is 38%, and this increases to 77% with ten or more positive central nodes. [23] Above a certain number of positive lymph nodes, it seems that surgical cure is not achievable [23] Quantitative analysis with average counts from each compartment cannot be described in our study due to the diversity of surgical approach and incomplete documentation in the operative notes. Meticulous recording of lymph node site and counts has been identified as area of improvement, especially as there is evidence to show that the burden of central lymph nodes predicts lateral neck involvement and ultimately survival. From a clinical perspective, when histopathological information about central node metastases is available from a previous central lymph node dissection and basal CT levels have not normalized, this will help to guide further neck dissection [23] ‘Berry picking’ will underestimate the true number of central nodes, justifying the for the need complete central dissection. Patients with central nodes have large numbers of lateral metastases and should be referred to specialist centres for reoperation to minimize surgical morbidity [23]  Machens et al. reported that  44 out of 45 patients had contralateral lateral neck nodes existing with metastases in the central and ipsilateral neck compartments. [23] In our study, of who underwent central neck dissection and bilateral lateral neck dissection, 11 of 13 patients with ipsilateral cervical metastases had contralateral cervical metastases. Contralateral neck dissection may not be warranted when central and ipsilateral nodes are negative.

In a study by Machens et al, there was a significant linear relationship between the number of involved distant lymph node compartments (contralateral cervical neck or superior mediastinal compartment), and distant metastases. [13] Distant metastases were encountered at a rate of 36% when 1 distant compartment was affected, and 51% when both were affected. [13] When both compartments were involved post-operative normalization of CT was never achieved. [13] 15 patients in our study had distant compartment involvement, none of whom achieved biochemical cure and amongst them 6 patients had distant metastases. This data suggests ipsilateral cervical lymph node compartments are the watershed between local and systemic disease. This is important clinically, as with the likely presence of systemic disease, surgical cure will not be achieved with additional extensive surgical lymphadenectomy to clear cervical disease. However, locoregional control may prevent tracheal and esophageal invasion due to tumour penetration of lymph node capsule. [13] However, when only 1 distant lymph node compartment  involved, central and bilateral lymph node compartment dissection may still achieve postoperative cure in 10% of patients. [13] There is no universal consensus on treating systemic disease whether it is sampling of gross LN or systematic LN dissection.

Extent of surgery can also be determined based on pre-operative CT level. Studies show a relationship between CT and size, lymph node, distant metastases, biochemical cure [18, 20, 17, 12]; and cancer specific survival which is worse when the CT doubling time less than 6 months. [14] Although our findings do not agree with observations that CT correlates with tumour size, European data showed that a CT level <100 was associated with median tumour of 3 mm, with 98% of tumours <1 cm. A CT level of >1000 was associated with a median = 2.5 cm. [18] We found that pre-operative CT levels correlated with extent of disease (lymph node positivity, distant compartment involvement),  likelihood of biochemical cure, and recurrent or persistent disease. ROC curves best predicted likelihood of CT in determining biochemical cure, then recurrent or persistent disease; lymph node positivity; and distant metastases.

Machens et al. showed that despite clearance of additional lymph node metastases, extensive dissection is more and more unable to secure biochemical cure in presence of higher preoperative biomarkers as they indicate occult systemic disease. [12] Bilateral compartment oriented neck surgery achieved biochemical cure in >50% of patients with >1000 CT but not in patients >10000 in a series of 300 patients. [12] In our study, of the 16 patients that underwent bilateral central and lateral compartment surgery, three patients were cured, with a CT <472, and 13 patients were not, with a CT >1100. In another study, a CT above 500pg/mL better indicated failure to achieve biochemical cure on multivariate analysis than nodal metastases. [17] In our study a CT threshold of 730 had 70% sensitivity and 100% specificity in predicting failure of biochemical cure.

The absolute level of CT in determining extent of surgery as a well as the frequency of post-operative screening for recurrence and prediction of radiological recurrence is unknown. Machens et al concluded in a study with 300 patients all of whom underwent central neck dissection, with 76% of patients undergoing ipsilateral lateral neck dissection and 72% undergoing contralateral lateral neck dissection; that patients with a CT of <20 could forgo systematic lymph node dissection. [12] Based on assessment of locoregional lymph node involvement, they recommended patients with a CT of 20-50 undergo ipsilateral central neck and lateral neck dissection; patients with CT of 50-200 undergo contralateral central neck dissection in addition. CT level >200 was the threshold for performing contralateral lateral neck dissection. [12] A basal pre-operative CT>200 was associated with a 32 % risk of contralateral neck involvement. [12] It is difficult to corrobate our data as bilateral MRND’s were seldom performed. In our study, central involvement occurred even with undetectable CT levels, whilst ipsilateral lateral neck nodes occurred at a CT level of 61. No patients with contralateral cervical neck metastases or superior mediastinal metastases had a CT less than 1100 (range; 1100-67746). A few node-negative tumours were seen with very high basal CT and CEA levels in the Machen’s series [12], as in our study, where one patient had a pre-operative  CT level of 140000 [12] The absolute level of CT for omitting CLND and MRND from the operative plan is unknown. Distant metastases appeared at basal CT levels of 150-400 in a separate study by Machen’s et al. with 385 patients. [17] Distant metastases appeared at a CT level >2250 in our study.

Preoperative CEA did not correlate with any tumour characteristics, extent of disease or chance of biochemical cure in our study. Yip found that preoperative CEA correlated with positive margins but not lymph nodes or chance of biochemical cure.[7] Machen, Holzhausen and Dralle reported that preoperative serum CEA >30 was incompatible with surgical remission. [13] The rate of central and lateral lymph node involvement in patients with a CEA >30 was 75% and this increased to to 90% when CEA was greater than 100. A CEA >100 was also associated with contralateral nodal disease and distant mets. [13] In our study, 9 patients with CEA >30 (range; 30-1500) achieved biochemical cure.

Death from MTC is uncommon unless there is evidence of distant metastases at the time of surgery. We only recorded 7 MTC specific deaths (5 of whom had distant metastases at presentation). The 5-year DFS for patients who achieved biochemical cure versus no biochemical cure was 100% and 41% respectively (p<0.0001). Although MTC spreads early to regional lymph nodes, it is an indolent disease. The 10-year DSS for no biochemical cure was 73% compared with 100% in patients who had achieved biochemical cure in another study. [27] In U.S. SEER data from 1973-2002, patients with regional disease had a 10 year overall survival of 75% while patients with metastases had a 40% 10 year survival. [28] In our study, the 5 year DFS for negative lymph nodes and positive lymph nodes was 100% and 48% respectively (p-0.0002). 23% (11 in 48 R0 resections) of patients had a cervical recurrence. This compares to the recurrence rates reported in the literature (20% to 60%) and suggests that more attention can be paid to management of neck nodes in these patients. [24] In a series of 80 patients by Yip et al. a more aggressive standardized approach was taken, with a cervical recurrence rate of 13%. Patients with sporadic MTC and no clinicoradiological evidence cervical nodal disease underwent total thyroidectomy, CND, and ipsilateral MRND; while hereditary MTC patients or sporadic MTC with clinical or radiological evidence of lymph node metastases underwent total thyroidectomy and CND and bilateral MRND [29] Reoperatiolyn by experienced surgeons for locoregional disease can achieve biochemical cure in up to one-third of patients. [4, 30] Reoperation should be considered when patients have persistent raised post-operative CT in the setting of inadequate initial surgery; recurrence or threat of compression or invasion of the trachea or major vessels.  One series reported improved rates of CT normalization and reduction in CT levels when reoperation was performed following a systematic metastatic workup before reoperation, including routine staging laparoscopy to select suitable patients. [30]

Limitations:

Retrospective design , tx selection bias.

Size limits power, constrained by low prevalence. Misclsssifcation of data

Nondissected basins assumed negative ? underestiamtation of extent of disease in some.

Prospective studies to analyse utility for screening and adjunct to determining extent of surgery as it relates to outcome.

Didn’t stratify

CONCLUSION

Current practice overall in New Zealand institutions is marked by inconsistency in management of MTC, with many patients having inadequate surgery. This is likely due to many patients being diagnosed at time of surgery and the limited exposure of individual surgeons to this rare disease. Evidence has emerged that lymph node status including site and number of lymph nodes relates to extent of disease, and the chance of biochemical cure, but often this information was not ascertained as found from the lack of detail in operation, pathology notes. As well as this, a systematic surgical approach was often not undertaken, and bilateral central neck dissection and lateral neck dissection was seldom performed in New Zealand institutions.  The extent of lateral neck dissection should be influenced by the site of known lymph node metastasis, burden of central compartment disease, preoperative calcitonin levels and pre-operative ultrasound assessment. According to the ATA recommendations based on pre-operative calcitonin levels, many patients did not have adequate surgery. Our data supports the use of preoperative calcitonin levels in guiding the extent of surgery as it reflects the extent of disease. Based on the current deficiencies in treatment of MTC, an argument can be made for centralized referral and support for MTC patients especially for those with suspected nodal metastases, positive node status or persistently raised post-operative calcitonin following operation. Surgical morbidity should be minimized without compromising cure.

It will be important to document practice prospectively in more institutions to further examine how current evidence and guidelines can be applied in utilizing pre-operative calcitonin in the management of MTC. Future prospective randomized controlled trials will be important and necessary to truly define a biomarker-based standardised approach to treating MTC, while considering the patient morbidity associated with neck dissection. This would likely need to be a multicentre study and our data would help determine the sample size (power) of such a study.

REFERENCES

1. Mazeh H, Sippel RS. Surgical management of medullary thyroid cancer. Minerva Endocrinol. 2012;37(4):329-34.

2. Raue F, Kotzerke J, Reinwein D, Schroder S, Roher HD, Deckart H. Prognostic factors in medullary thyroid carcinoma: evaluation of 741 patients from the German Medullary Thyroid Carcinoma Register Clin Invest 1993;71:7-12.

3. Kebebew E, Kikuchi S, Duh QY, Clark OH. Long-term results of reoperation and localizing studies in patients with persistent or recurrent medullary thyroid cancer. Arch Surg 2000;135:895-901.

4. Jin LX, Moley JF. Surgery for lymph node metastases of medullary thyroid carcinoma. International Journal of Endocrine Oncology. 2015;2(3):229-39.

5. Hundahl SA, Fleming ID, Fremgen AM, Menck HR. A National Cancer Data Base report on 53 856 cases of thyroid carcinoma treated in the US, 1985–1995 Cancer. 1998;83( ):2638-48.

6. Ball DW. Medullary thyroid cancer: monitoring and therapy. Endocrinol Metab Clin North Am. 2007 36(3):823-37.

7. Yip DT, Hassan M, Pazaitou-Panayiotou K, Ruan DT, Gawande AA, Gaz RD et al. Preoperative basal calcitonin and tumor stage correlate with postoperative calcitonin normalization in patients undergoing initial surgical management of medullary thyroid carcinoma. Surgery. 2011;150(6):1168-77. doi:http://dx.doi.org/10.1016/j.surg.2011.09.043.

8. Brierly J, Tsang R, Simpson WJ, Gosopodarowicz M, al e. Medullary thyroid cancer: analyses of survival and prognostic factors and the role of radiation therapy in local control. Thyroid. 1996;6(4):305-10.

9. Scollo C, Baudin E, Travagli JP, Caillou B, Bellon N, Leboulleux S et al. Rationale for central and bilateral lymph node dissection in sporadic and hereditary medullary thyroid cancer. J Clin Endocrinol Metab. 2003;88(5):2070-5.

10. Moley JF, DeBenedetti MK. Patterns of nodal metastases in palpable medullary thyroid carcinoma: recommendations for extent of node dissection. Ann Surg. 1999;229(6):880-7.

11. Kouvaraki MA, Shapiro SE, Fornage BD, al. e. Role of preoperative ultrasonography in the surgical management of patients with thyroid cancer. Surgery. 2003;134(6):946-54.

12. Machens A, Dralle H. Biomarker-based risk stratification for previously untreated medullary thyroid cancer. J Clin Endocrinol Metab. 2010;95(6):2655-63. doi:http://dx.doi.org/10.1210/jc.2009-2368.

13. Machens A, Holzhausen HJ, Dralle H. Contralateral cervical and mediastinal lymph node metastasis in medullary thyroid cancer: Systemic disease? Surgery. 2006 139(1):28-32.

14. Barbet J, Campion L, Kraeber-Bodere F, Chatal JF. Prognostic impact of serum calcitonin and carcinoembryonic antigen doubling-times in patients with medullary thyroid carcinoma. Journal of Clinical Endocrinology and Metabolism. 2005;90(11):6077-84.

15. Rowland KJ, Jin LX, Moley JF. Biochemical cure after reoperations for medullary thyroid carcinoma: a meta-analysis. Ann Surg Oncol. 2015;22(1):96-102. doi:http://dx.doi.org/10.1245/s10434-014-4102-y.

16. Kloos RT, Eng C, Evans DB, al. e. Medullary thyroid cancer: management guidelines of the American Thyroid Association. Thyroid. 2009;19:565-612.

17. Machens A, Schneyer U, Holzhausen HJ, Dralle H. Prospects of remission in medullary thyroid carcinoma according to basal calcitonin level. J Clin Endocrinol Metab. 2005;90:2029-34.

18. Cohen R, Campos JM, C CS, al e. Preoperative calcitonin levels are predictive of tumor size and postoperative calcitonin normalization in medullary thyroid carcinoma. Groupe d’Etudes des Tumeurs a Calcitonine (GETC) J Clin Endocrinol Metab. 2000;85:919-22.

19. Weber T, Schilling T, Frank-Raue K, Colombo-Benkmann M, Hinz U, Ziegler R et al. Impact of modified radical neck dissection on biochemical cure in medullary thyroid carcinomas. Surgery. 2001;130(6):1044-9.

20. Wells SA, Asa SL, Dralle H, Elisei R, Evans DB, Gagel RF et al. Revised American thyroid association guidelines for the management of medullary thyroid carcinoma. Thyroid. 2015;25(6):567-610.

21. Elisei R, Bottici V, F L, al. e. Impact of routine measurement of serum calcitonin on the diagnosis and outcome of medullary thyroid cancer: experience in 10,864 patients with nodular thyroid disorders. J Clin Endocrinol Metab 2004;89:163-8.

22. Pacini F, Schlumberger M, Dralle H, Elisei R, Smit JW, Wiersinga W. European consensus for the management of patients with differentiated thyroid carcinoma of the follicular epithelium. Eur J Endocrinol 2006;154:787-803.

23. Machens A, Hauptmann S, Dralle H. Prediction of lateral lymph node metastases in medullary thyroid cancer. Br J Surg. 2008;95(5):586-91

24. Dralle H, Damm I, Scheumann GF, Kotzerke J, Kupsch E, Geerlings H. Compartment-oriented microdissection of regional lymph nodes in medullary thyroid carcinoma. . Surg Today 1994;24:112-21.

25. Leggett MD, Chen SL, Schneider PD, Martinez SR. Prognostic value of lymph node yield and metastatic lymph node ratio in medullary thyroid carcinoma.  . Ann Surg Oncol. 2008;15:2493-9.

26. Dralle H, Scheumann GF, Proye C, Bacourt F, Frilling A, Limbert F. The value of lymph node dissection in hereditary medullary thyroid carcinoma: a retrospective, European, multicentre study. J Intern Med 1995;238:357-61.

27. Rendl G, Manzl M, Hitzl W, Sungler P, Pirich C. Long-term prognosis of medullary thyroid carcinoma. . Clin Endocrinol (Oxf). 2008;69:497-505.

28. Roman S, Lin R, Sosa JA. Prognosis of medullary thyroid carcinoma: demographic, clinical, and pathologic predictors of survival in 1252 cases. Cancer. 2006;107(9):2134-42.

29. Yen TW, Shapiro SE, Gagel RF, Sherman SI, Lee JE, Evans DB. Medullary thyroid carcinoma: results of a standardized surgical approach in a contemporary series of 80 consecutive patients. Surgery. 2003;134(6):890-9; discussion 9-901.

30. Moley JF, Dilley WG, DeBenedett MK. Improved results of cervical reoperation for medullary thyroid carcinoma. Ann Surg. 1997;225:734-40.