A Novel N Staging System for Predicting Survival in Patients with Medullary Thyroid Cancer

Abstract

Introduction
Despite the crucially prognostic value of lymph node metastasis (LNM) in patients with medullary thyroid cancer (MTC), only the LNM compartment alone was reflected in the 8th edition of the American Joint Committee on Cancer (AJCC) system.

Objective
This study aimed to incorporate the metastatic lymph node number and metastatic lymph node ratio to generate a more accurate and appropriate N staging system for patients with MTC based on recursive partitioning analysis.

Design, Setting, and Patients
Two cohorts were included in the analysis, including 1374 MTC patients from the Surveillance, Epidemiology, and End Results database as the derivation cohort, and 164 patients from Fudan University Shanghai Cancer Center as the validation cohort. The predictive performance of the alternative proposed N staging system was compared with that of the 8th AJCC system by using the Harrell concordance index (C-index) and the area under the receiver operating characteristic curve (AUC).

Results
In the derivation cohort, the C-index and the AUC at 10 years were 0.778 and 0.789, respectively, for the novel N staging system, and 0.749 and 0.741, respectively, for the 8th AJCC N staging system. Similar trends were also observed in the validation cohort. The proposed N staging system had a better prognostic performance.

Conclusion
With some improvements, the novel N staging system for MTC suggested from this research may be assessed for potential adoption in the next edition of the AJCC N staging system.

Full text

ORIGINAL ARTICLE – ENDOCRINE TUMORS
A Novel N Staging System for Predicting Survival in Patients
with Medullary Thyroid Cancer
Lili Chen, MM
1,2
, Kai Qian, MD
3
, Kai Guo, MM
3
, Xiaoke Zheng, MM
1,2
, Wenyu Sun, MM
1,2
,
Tuanqi Sun, MD
1,2
, Yunjun Wang, MD
1,2
, Duanshu Li, MM
1,2
, Yi Wu, MD
1,2
, Qinghai Ji, MD
1,2
, and
Zhuoying Wang, MD, PhD
1,2,3
1
Department of Head and Neck Surgery, Fudan University Shanghai Cancer Center, Shanghai, China;
2
Department of
Oncology, Shanghai Medical College, Fudan University, Shanghai, China;
3
Department of Head and Neck Surgery, Renji
Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
ABSTRACT
Introduction. Despite the crucially prognostic value of
lymph node metastasis (LNM) in patients with medullary
thyroid cancer (MTC), only the LNM compartment alone
was reflected in the 8th edition of the American Joint
Committee on Cancer (AJCC) system.
Objective. This study aimed to incorporate the metastatic
lymph node number and metastatic lymph node ratio to
generate a more accurate and appropriate N staging system
for patients with MTC based on recursive partitioning
analysis.
Design, Setting, and Patients. Two cohorts were included
in the analysis, including 1374 MTC patients from the
Surveillance, Epidemiology, and End Results database as
the derivation cohort, and 164 patients from Fudan
University Shanghai Cancer Center as the validation
cohort. The predictive performance of the alternative pro-
posed N staging system was compared with that of the 8th
AJCC system by using the Harrell concordance index (C-
index) and the area under the receiver operating charac-
teristic curve (AUC).
Results. In the derivation cohort, the C-index and the
AUC at 10 years were 0.778 and 0.789, respectively, for
the novel N staging system, and 0.749 and 0.741,
respectively, for the 8th AJCC N staging system. Similar
trends were also observed in the validation cohort. The
proposed N staging system had a better prognostic
performance.
Conclusion. With some improvements, the novel N stag-
ing system for MTC suggested from this research may be
assessed for potential adoption in the next edition of the
AJCC N staging system.
Thyroid cancer is one of the most common malignant
tumors in the endocrine system, with an increasing global
incidence.
1
As the third most common histology of thyroid
cancers, medullary thyroid cancer (MTC) is a rare neu-
roendocrine tumor that originates from parafollicular
(C) cells.
2,3
The clinicopathological and biologic charac-
teristics of MTC are intensively different from those of
differentiated thyroid cancer (DTC). Currently, an
increasing number of technical methods, such as advanced
ultrasonography, fine needle aspiration biopsy, serum
levels of calcitonin (Ctn) and carcinoembryonic antigen
(CEA) measurements, and RET germline mutation analy-
sis, have allowed MTC to be diagnosed much more
accurately in the earlier stages of disease.
4–8
MTC only
accounts for 1–2% of thyroid carcinomas of all types, but
leads to 13.4% of all thyroid cancer-related deaths. Despite
different biological features between MTC and DTC, the
8th edition of the American Joint Committee on Cancer
(AJCC) N staging system of MTC is extrapolated from that
of DTC and remains controversial in predicting patient
survival.
9
Lili Chen, Kai Qian, and Kai Guo contributed equally to this work.
ÓSociety of Surgical Oncology 2019
First Received: 27 April 2019
Z. Wang, MD, PhD
e-mail: zhuoyingwang@hotmail.com
Ann Surg Oncol
https://doi.org/10.1245/s10434-019-07871-1

Lymph node metastasis (LNM) status gradually appears
to be an important predictor of outcome for many cancers,
such as lung cancer, other head and neck cancers, and
MTC.
10–14
For patients with hypopharyngeal cancer, the
number of positive lymph nodes is an effective predictor of
survival and has been demonstrated to further stratify
prognoses.
13
In this study, we assumed that metastatic
lymph node number (MLNN) and metastatic lymph node
ratio (MLNR) could be prognostic metrics.
15,16
MLNR was
defined as the ratio of MLNN to the examined LN number
(ELNN). However, the impact of quantitative LNM (i.e.
MLNN and MLNR) for predicting the survival of MTC
patients has rarely been researched, and a modified N
staging system of MTC is seldom proposed. Therefore, we
believe that this study is necessary to some degree. We
aimed to define a special N staging system for MTC
through the population-based Surveillance, Epidemiology,
and End Results (SEER) database and the Fudan University
Shanghai Cancer Center (FUSCC) database. In this study,
we developed and validated the N staging system for pre-
dicting the survival of MTC patients.
PATIENTS AND METHODS
Patients and Outcomes
Two large cohorts of patients with MTC, including a
derivation cohort from the SEER database (1998–2014)
and a validation cohort from FUSCC (2002–2015), were
collected. In the derivation cohort, pathologically con-
firmed MTC patients whose primary site was limited to
‘C73.9—thyroid gland’ and codes 8345/3, 8510/3, and
8512/3 from the International Classification of Diseases for
Oncology, 3rd Edition (ICD-O-3) histology were included.
The extent of operation was recognized according to the
RX Summ–Surg Prim Site (1998 ?) and regional nodes
positive (1988 ?). Only patients who received total thy-
roidectomy as their primary treatment were included. The
exclusion criteria were patients with an unknown cause of
death, distant metastasis, no LN resection or ELNN = 0,
undetermined MLNN or ELNN, incomplete follow-up
details, or those who received oncological therapy before
surgery. In the validation cohort, all patients were also
identified via pathological diagnoses. The primary treat-
ment was limited to total thyroidectomy. All available data
of these patients were collected. We used the same
exclusion criteria from the derivation cohort for the vali-
dation cohort. Patients from both cohorts were followed for
at least 1 year after the initial treatment, and were restaged
according to the 8th AJCC staging system definitions.
In the SEER cohort, cancer-specific survival (CSS),
which was calculated from the diagnosis date to the cancer-
specific date of death or the last follow-up, was used as the
primary endpoint for MTC patients. In the FUSCC cohort,
the main endpoint was disease-free survival (DFS), the
period from the first surgery to the recurrence, or the most
recent follow-up visit. The MTC recurrence was confirmed
by both histology and radiography.
This study was approved by the Ethics Committee of
FUSCC.
Statistical Analyses
In the baseline characteristics, continuous variables are
described as the means, and categorical variables are
described as frequencies and percentages. Cut-off values of
MLNN and MLNR with the highest sensitivity and speci-
ficity were calculated using the X-tile program (Yale
University School of Medicine, New Haven, CT, USA).
CSS curves and DFS curves were drawn using the Kaplan–
Meier method and tested using the log-rank tests.
Recursive partitioning analysis (RPA), a method of
building decision trees to model predictors,
17
was applied
to develop a novel N staging system. A conditional infer-
ence tree was created using MLNN and MLNR, estimated
by binary recursive partitioning.
Risk factors associated with CSS from the SEER data-
base were enrolled in the Cox proportional hazards model.
To judge the discriminatory ability of the novel N staging
system and the 8th AJCC N staging system, we calculated
the concordance index (C-index) and time-dependent
receiver operating characteristic (td-ROC) curves for two
cohorts. Analyses were performed using SPSS version 22.0
(IBM Corporation, Armonk, NY, USA) and R version 3.4.0
(Bell Laboratories, Murray Hill, NJ, USA; https://www.r-
project.org/). Survival curves were generated using the R
package survival and survminer, and C-indices were cal-
culated using the R package Hmisc and survival. Td-ROCs
were derived from R package survivalROC and Tidyverse
(https://cran.r-project.org/src/contrib/Archive/). A two-
sided pvalue \0.05 was considered statistically
significant.
RESULTS
Patient Clinical Characteristics
Overall, 1179 patients with pathologically confirmed
MTC were screened from the SEER database, approxi-
mately half of whom had LNM. The clinical features of
MTC patients are shown in Table 1. The mean ELNN
value was 21.53 (median 14). For MLNN, the mean value
was 4.53 (median 0), and the MLNR was 0.19 (median
0.00). The mean age at diagnosis was 49.0 years (median
L. Chen et al.

50 years). There were more female patients than male
patients (female:male ratio of 1.3:1.0). The mean CSS
value was 82.4 months (median 73.0 months).
Identification of Cut-Off Values for Metastatic Lymph
Node Number and Metastatic Lymph Node Ratio
To discover the optimal cut-off values for MLNN and
MLNR, an X-tile program was used to analyze the data of
patients with MTC from the SEER database. The results
showed that 0 and 12 were the best cut-off values for
MLNN (p\0.001) (Fig. 1a). When MLNN was B12, the
optimal cut-off value for MLNR was 0.6 (p\0.001)
(Fig. 1b), and when MLNN was [12, the optimal cut-off
value for MLNR was 0.75 (p\0.001) (Fig. 1c).
Proposed N Staging System
Using the RPA, we generated a novel N staging system
(Fig. 1d) for patients from the SEER cohort. First, based on
the MLNN, patients were divided into the following three
groups: MLNN = 0, 1 BMLNN B12, and MLNN [12.
Patients in the 1 BMLNN B12 group were then divided
into two subgroups as follows: MLNR \0.6 and MLNR
C0.6. In addition, patients in the MLNN [12 group were
divided into the following two subgroups: MLNR \0.75
and MLNR C0.75. Kaplan–Meier estimates for the five
subgroups (MLNN = 0; 1 BMLNN B12 and MLNR \
0.6; 1 BMLNN B12 and MLNR C0.6; MLNN [12
and MLNR \0.75; and MLNN [12 and MLNR C0.75)
indicated that there were no significances between patients
from the two subgroups of 1 BMLNN B12 and
TABLE 1 Clinicopathologic
characteristics of the SEER and
FUSCC cohorts
Clinical characteristics Total SEER cohort FUSCC cohort pvalue
[1343 (100%)] [1179 (87.8%)] [164 (12.2%)]
Age, years
a
0.008
B63 1076 (80.1) 932 (79.1) 144 (87.8)
[63 267 (19.9) 247 (20.9) 20 (12.2)
Sex 0.002
Female 771 (57.4) 695 (58.9) 76 (46.3)
Male 572 (42.6) 484 (41.1) 88 (53.7)
Race
White 999 (84.7)
Black 92 (7.8)
Others 88 (7.5)
T stage 0.009
T1 ?T2 963 (71.7) 843 (71.5) 120 (73.2)
T3 ?T4 304 (22.6) 261 (22.1) 43 (26.2)
Unknown 76 (5.7) 75 (6.4) 1 (0.6)
N stage \0.001
N0 651 (48.5) 599 (50.8) 52 (31.7)
N1a 192 (14.3) 166 (14.1) 26 (15.9)
N1b 306 (22.8) 222 (18.8) 84 (51.2)
NX 194 (14.4) 192 (16.3) 2 (1.2)
Tumor size, cm 0.366
0–2 650 (48.4) 562 (47.7) 88 (53.7)
2–4 454 (33.8) 400 (33.9) 54 (32.9)
[4 205 (15.3) 185 (15.7) 20 (12.2)
Unknown 34 (2.5) 32 (2.7) 2 (1.2)
Extracapsular extension 0.445
Yes 173 (12.9) 151 (12.8) 22 (13.4)
No 1155 (86.0) 1013 (85.9) 142 (86.6)
Unknown 15 (1.1) 15 (1.3) 0 (0)
Data are expressed as n(%)
SEER Surveillance, Epidemiology, and End Results, FUSCC Fudan University Shanghai Cancer Center
a
The age cut-off value was 63 years and was derived from the X-tile program
N-Staging System for Medullary Thyroid Cancer

FIG. 1 X-tile analyses
identifying optimal MLNN
(a) and MLNR [bMLNN B12;
cMLNN [12] cut-off values
based on cancer-specific
survival. Defining a novel LN
staging system for MTC by
recursive partitioning analysis
based on the MLNN and MLNR
(d). The differences between the
two systems are shown in (e).
LNs lymph nodes, MLNN
metastatic lymph node number,
MLNR metastatic lymph node
ratio, MTC medullary thyroid
cancer, CSS cancer-specific
survival, AJCC American Joint
Committee on Cancer, AUC
area under the receiver
operating characteristic curve,
SEER surveillance,
epidemiology, and end results,
FUSCC Fudan University
Shanghai Cancer Center
L. Chen et al.

MLNR C0.6, and MLNN [12 and MLNR \0.75.
Therefore, we combined the two subgroups and then gen-
erated the following novel N staging system: N0:
MLNN = 0; N1: 1 BMLNN B12 and MLNR \0.6; N2:
1BMLNN B12 and MLNR C0.6 or MLNN [12 and
MLNR \0.75; and N3: MLNN [12 and MLNR C0.75.
Impact of the Novel N Staging System on CSS
Prediction
In the Cox model, univariate analysis showed that this
novel N staging system was found to be a strong predictor
of CSS (p\0.001). Multivariate analysis implied that
after adjusting for other clinical and demographic factors,
the novel N staging system (N1: hazard ratio [HR] 2.889,
95% confidence interval [CI] 1.248–6.689, p= 0.013; N2:
HR 7.005, 95% CI 3.033–16.182, p\0.001; and N3: HR
29.162, 95% CI 8.316–102.263, p\0.001) could still
independently predict CSS (Table 2). Age [63 years
(p\0.001) and larger tumor size (p\0.001) were also
independently associated with poorer prognosis.
Comparison of the Novel N Staging System Versus
the 8th American Joint Committee on Cancer N Staging
System
To identify whether the novel N staging system
improved predictive ability compared with the 8th AJCC N
staging system for MTC, Kaplan–Meier estimates of the
novel system (Fig. 2a, c) and the AJCC system (Fig. 2b, d)
TABLE 2 Univariate and
multivariate Cox analysis for
cancer-specific survival in the
SEER cohort
Variables Univariate analysis Multivariate analysis
HR (95% CI) pvalue HR (95% CI) pvalue
Age, years
a
B63 1 1
[63 2.322 (1.479–3.645) \0.001 2.406 (1.405–4.119) 0.001
Sex
Female 1 1
Male 1.890 (1.202–2.970) 0.006 1.142 (0.663–1.968) 0.663
Year of diagnosis
1998–2003 1 1
2004–2009 0.601 (0.356–1.015) 0.057 0.742 (0.387–1.422) 0.368
2010–2014 0.547 (0.254–1.182) 0.125 0.553 (0.211–1.445) 0.227
T stage
T1 ?T2 1 1
T3 ?T4 5.037 (2.973–8.535) \0.001 1.316 (0.492–3.519) 0.585
Tumor size, cm
0–2 1 1
2–4 2.699 (1.447–5.036) 0.002 3.851 (1.1.787–8.300) 0.001
[4 6.129 (3.259–11.524) \0.001 3.872 (1.388–10.798) 0.010
Extracapsular extension
No 1 1
Yes 4.767 (2.996–7.853) \0.001 1.323 (0.592–2.958) 0.496
ELNN
b
B14 1 1
[14 2.173 (1.352–3.492) 0.001 1.047 (0.581–1.887) 0.879
Novel N stage
N0 1 1
N1 4.562 (2.129–9.777) \0.001 2.889 (1.248–6.689) 0.013
N2 12.268 (5.581–25.725) \0.001 7.005 (3.033–16.182) \0.001
N3 76.100 (30.068–192.603) \0.001 29.162 (8.316–102.263) \0.001
HR hazard ratio, CI confidence interval, ELNN examined lymph node number, SEER surveillance, epi-
demiology, and end results
a
The age cut-off value was 63 years and was derived from the X-tile program
b
The ELNN cut-off value was 14 and was derived from the X-tile program
N-Staging System for Medullary Thyroid Cancer

L. Chen et al.

were illustrated. The novel system showed preferable dis-
crimination of the survival curves for both the SEER and
FUSCC cohorts. The same trend was found in the 10-year
CSS of patients in each substage from the SEER cohort
(Fig. 1e). According to the novel system, the 10-year CSS
rates for N0, N1, N2, and N3 were 97.6%, 89.8%, 74.0%,
and 23.9%, respectively. In contrast, the results for N0,
N1a, N1b, and NX from the 8th AJCC system were 97.6%,
93.8%, 71.2%, and 77.8%, respectively. The results of the
univariate Cox regression analysis also demonstrated that
the novel system showed better discrimination in risk
stratification than the AJCC system (Table 3).
The predictive metric C-indices of two different N
staging systems are presented in Fig. 1e. The respective
C-indices for the SEER cohort using the novel and AJCC
systems were 0.778 (95% CI 0.775–0.781) and 0.749 (95%
CI 0.746–0.753). For the FUSCC cohort, when predicting
DFS, a similar result was obtained: the C-index of the
novel system was 0.716 (95% CI 0.712–0.720) and the
C-index of the AJCC was 0.670 (95% CI 0.666–0.674).
When analyzing the evaluation ability of the systems,
the novel system performed better than the AJCC system in
both the SEER cohort (5-year area under the curve (AUC)
0.803 vs. 0.771; 10-year AUC 0.789 vs. 0.741) (Fig. 2e)
and the FUSCC cohort (5-year AUC 0.801 vs. 0.699;
10-year AUC 0.788 vs. 0.735) (Fig. 2f) for CSS (DFS).
DISCUSSION
This study was performed in the context that the 8th
AJCC N staging system fails to fully use the state of LNM
to reflect the prognosis of MTC. MTC differs from DTC in
many aspects; however, the latest N staging system for
MTC is still restricted to the same concepts that were ini-
tially developed for DTC.
3,4,18
It is very significant to
propose a more accurate and appropriate N staging system
for MTC. Based on the relatively large sample of patients
from the SEER and FUSCC databases, this study proposed
a novel N staging system. In this system, MLNN and
MLNR were applied to predict patient prognosis. From our
perspective, the joint use of MLNN and MLNR could
potentially show two significant factors, i.e. regional
metastasis
19,20
and surgical approach,
4,21
which were
essentially regarded as prognostic factors. In addition,
these two variables were easy to calculate, and it was
convenient to classify patients based on these vari-
ables.
22,23
Therefore, we thought it was reasonable to use
MLNN and MLNR for the new N staging system.
Notably, the number and ratio of LNM are important
prognostic factors in many cancers.
16,24–26
However, owing
to the low incidence and limited available data, there were
few studies related to the connection between MTC prog-
nosis and LNM. Either C16 or C10 positive LNs were
identified as the optimal MLNN values in some studies that
analyzed patients from the SEER database.
25,27
One study
suggested a ratio of 0.5 as the optimal MLNR to predict the
prognosis of MTC patients with stage IV disease.
28
Machens and Dralle demonstrated that MLNN was related
to distant metastasis (p= 0.003) and proposed an N staging
system that did not show the relationship between MLNN
and CSS.
15
However, in all of the abovementioned studies,
the validation methods were limited to the Kaplan–Meier
analyses, patients were from a single center, and MLNN
and MLNR were not jointly used. In this study, we col-
lected patients with MTC from two databases (the SEER
and FUSCC databases). The former database includes
approximately 28% of the American population, and the
latter is a high-volume cancer center in China. The X-tile
program was used to choose the best cut-off values for
MLNN and MLNR. Using RPA, we proposed a new MTC-
specific N staging system based on MLNN and MLNR.
To assess whether the new system was better in pre-
dicting prognosis, several contrasts to the current system
were implemented. Compared with the Kaplan–Meier
survival curves obtained using the 8th AJCC N staging
system, those obtained using the RPA-derived N staging
system were sufficiently separated between the N substages
of the SEER and FUSCC cohorts. The proposed system
showed preferable discriminatory and predictive capacities
(the C-index: SEER cohort 0.778 [novel] vs. 0.749 [8th];
FUSCC cohort 0.716 [novel] vs. 0.670 [8th]). Nevertheless,
the LNM novel classification system seems to be better
than the AJCC system in some ways. First, this novel
classification system is a distinct system used for patients
with MTC, rather than one extrapolated from a DTC sys-
tem. In addition, a novel system abandoned the NX stage,
which was defined as regional LNs that could not be
assessed. Patients in the NX stage accounted for 16.3% of
patients in the SEER database, and, although the location
of LNM was unknown, the MLNN and MLNR values for
these patients were easy to record and calculate. This result
indicated that MLNN and MLNR were more useful in
clinical practice. We believe that the novel N staging
bFIG. 2 Kaplan–Meier survival curves for MTC patients using the
modified N staging system and the 8th AJCC N staging system for the
a,bSEER and c,dFUSCC cohorts. Time-dependent ROC curves of
the novel and 8th AJCC N staging systems in the prognostic
prediction of MTC patients at 1-, 3-, 5-, and 10-year points in the
eSEER and fFUSCC cohorts. MTC medullary thyroid cancer, AJCC
American Joint Committee on Cancer, SEER surveillance,
epidemiology, and end results, FUSCC Fudan University Shanghai
Cancer Center, ROC receiver operating characteristic curve, AUC
area under the receiver operating characteristic curve
N-Staging System for Medullary Thyroid Cancer

system will not only improve survival prognosis but will
also more accurately recognize patients who need further
treatment and close follow-up.
Limitations
This study has three limitations. First, the two cohorts
are different in some aspects, such as race and sex ratios,
and these clinicopathologic characteristics may affect the
accuracy of the staging system. Second, the study is limited
by its retrospective nature. It would be prudent to confirm
our results in a prospective cohort study or by using
another large, multi-institutional database, such as the
National Cancer Database. Finally, although the SEER
database maintains highly accurate records, incorrect cod-
ing or erroneous data are still possible.
CONCLUSIONS
The results of the Cox analysis demonstrated that the
novel system predicts survival independently and signifi-
cantly (p\0.001). A standardized surgical strategy may
account for this phenomenon. In addition, the collection
method for specimens, and the method used by pathologists
for verifying positive LNs could also affect the results;
thus, there must be standards for surgeons and pathologists.
Therefore, we must comply with standardized conditions
for precisely using MLNN and MLNR values to predict
survival, or the prognostic significance will be limited. Age
and tumor size could also predict CSS. The T staging
system has been considered in relation to the size of the
tumor, and perhaps, in the future, we could explore the
efficiency of age in stratifying different risk levels of MTC
patients.
TABLE 3 Comparison of the prognostic performances of different N staging systems in the derivation and validation cohorts
Cohort Staging
system
No. of
patients
Death (recurrence)
[n(%)]
CSS (DFS)
10-year (%) AUC
Cox analysis [HR (95% CI)] C-index (95% CI)
SEER Novel 0.784 0.778
(0.775–0.781)
N0 600 9 (1.5) 97.6 1
N1 359 25 (7.0) 89.8 4.562 (2.129–9.777)
N2 193 32 (16.6) 74.0 12.268 (5.581–25.725)
N3 27 10 (37.0) 23.9 76.100 (30.068–192.603)
8th 0.743 0.749
(0.746–0.753)
N0 600 9 (1.5) 97.6 1
N1a 166 8 (4.8) 93.8 4.209 (1.619–10.943)
N1b 221 14 (6.3) 71.2 6.111 (2.629–14.205)
NX 192 45 (23.4) 77.8 10.974 (5.326–22.500)
FUSCC Novel 0.788 0.716
(0.712–0.720)
N0 52 0 (0.0) 100 1
N1 50 8 (16.0) 63.2 48085.126 (0.000–7.325E?71)
N2 58 20 (34.5) 41.5 142137.370
(0.000–2.225E?72)
N3 4 4 (100.0) 0.00 585403.168
(0.000–9.190E?72)
8th 0.735 0.670
(0.666–0.674)
N0 52 0 (0.0) 100 1
N1 26 4 (15.4) 72.2 94663.868 (0.000–4.247E?79)
N1b 84 28 (33.3) 44.3 138092.231
(0.000–6.179E ?79)
NX 2 0 (0.0) 100 0.928 (0.000–4.259E ?271)
CSS cancer-specific survival, DFS disease-free survival, CI confidence interval, HR hazard ratio, AUC area under the receiver operating
characteristic curve, SEER surveillance, epidemiology, and end results, FUSCC Fudan University Shanghai Cancer Center
L. Chen et al.

ACKNOWLEDGEMENT The authors harbor sincere gratitude for
the efforts of the SEER program in the establishment of the SEER
database.
FUNDING This work was supported by the National Natural Sci-
ence Foundation of China (81772852), and the Science and
Technology Commission of Shanghai Municipality (16ZR1406600).
DISCLOSURES Lili Chen, Kai Qian, Kai Guo, Xiaoke Zheng,
Wenyu Sun, Tuanqi Sun, Yunjun Wang, Duanshu Li, Yi Wu, Qinghai
Ji, and Zhuoying Wang have no disclosures to declare.
REFERENCES
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA: A
Cancer Journal for Clinicians. 2019;69(1):7-34.
2. Brose MS, Bible KC, Chow LQM, et al. Management of treat-
ment-related toxicities in advanced medullary thyroid cancer.
Cancer Treat Rev. 2018;66:64–3.
3. Ernani V, Kumar M, Chen AY, Owonikoko TK. Systemic
treatment and management approaches for medullary thyroid
cancer. Cancer Treat Rev. 2016;50:89–8.
4. Wells SA Jr, Asa SL, Dralle H, et al. Revised American thyroid
association guidelines for the management of medullary thyroid
carcinoma. Thyroid. 2015;25(6):567–10.
5. Castinetti F, Waguespack SG, Machens A, et al. Natural history,
treatment, and long-term follow up of patients with multiple
endocrine neoplasia type 2B: an international, multicentre, ret-
rospective study. Lancet Diabetes Endocrinol. 2019;7(3):213–20.
6. Accardo G, Conzo G, Esposito D, et al. Genetics of medullary
thyroid cancer: an overview. Int J Surg. 2017;41:S2–S6.
7. Bae YJ, Schaab M, Kratzsch J. Calcitonin as biomarker for the
medullary thyroid carcinoma. In: Raue F (ed). Medullary thyroid
carcinoma: biology—management—treatment. Springer, Cham;
2015, pp. 117–137.
8. Cabanillas ME, Houston MR, Hu MI, et al. Prognostic signifi-
cance of circulating RET M918T mutated tumor DNA in patients
with advanced medullary thyroid carcinoma. J Clin Endocrinol
Metab. 2017;102(9):3591–9.
9. Berg AN, Seethala RR. The American Joint Committee on cancer
eighth edition: changes in thyroid carcinoma staging and an
update on reporting. AJSP Rev Rep. 2018;23(3):145–8.
10. Ding N, Pang Z, Zhang X, et al. Prognostic and predictive effects
of positive lymph node number or ratio in NSCLC. Sci Rep.
2017;7(1):584.
11. Kang SI, Kim D-W, Shin E, et al. Clinical significance of lymph
node metastasis in the mesentery of the terminal ileum in patients
with right-sided colon tumors at different locations. Dis Colon
Rectum. 2018;61(6):692–7.
12. Ho AS, Kim S, Tighiouart M, et al. Metastatic lymph node
burden and survival in oral cavity cancer. J Clin Oncol.
2017;35(31):3601–9.
13. Ho AS, Kim S, Tighiouart M, et al. Association of quantitative
metastatic lymph node burden with survival in hypopharyngeal
and laryngeal cancer. JAMA Oncol. 2018;4(7):985–9.
14. Machens A, Lorenz K, Dralle HJE. Histology-proven recurrence
in the lateral or central neck after systematic neck dissection for
medullary thyroid cancer. Endocrine. 2018;61(3):428–9.
15. Machens A, Dralle H. Prognostic impact of N staging in 715
medullary thyroid cancer patients: proposal for a revised staging
system. Ann Surg. 2013;257(2):323–9.
16. Adam MA, Pura J, Goffredo P, et al. Presence and number of
lymph node metastases are associated with compromised survival
for patients younger than age 45 years with papillary thyroid
cancer. J Clin Oncol. 2015;33(21):2370–5.
17. Hothorn T, Hornik K, Zeileis A. Unbiased recursive partitioning:
a conditional inference framework. J. Comput. Graph. Stat.
2006;15(3):651–74.
18. Tuttle RM, Haugen B, Perrier ND. Updated American Joint
Committee on cancer/tumor-node-metastasis staging system for
differentiated and anaplastic thyroid cancer (eighth edition): what
changed and why? Thyroid. 2017;27(6):751–6.
19. Nixon IJ, Shaha AR. Management of regional nodes in thyroid
cancer. Oral Oncology. 2013;49(7):671–5.
20. Moley JF. Medullary thyroid carcinoma: management of lymph
node metastases. J Natl Compr Canc Netw. 2010;8(5):549–56.
21. Pacini F, Castagna MG, Cipri C, Schlumberger M. Medullary
Thyroid Carcinoma. Clin Oncol. 2010;22(6):475–85.
22. Qian K, Sun W, Guo K, et al. The number and ratio of positive
lymph nodes are independent prognostic factors for patients with
major salivary gland cancer: results from the surveillance, epi-
demiology, and end results dataset. Eur J Surg Oncol.
2019;45(6):1025–32.
23. Tang X, Chen Y, Guo L, Zhang J, Wang C. Prognostic signifi-
cance of metastatic lymph node number, ratio and station in
gastric neuroendocrine carcinoma. J Gastrointest Surg.
2015;19(2):234–41.
24. Zhou Y-Y, Du X-J, Zhang C-H, et al. Comparison of three lymph
node staging schemes for predicting the outcome in patients with
small bowel adenocarcinoma: a population-based cohort and
international multicentre cohort study. EBioMedicine.
2019;41:276–85.
25. Randle RW, Balentine CJ, Leverson GE, et al. Trends in the
presentation, treatment, and survival of patients with medullary
thyroid cancer over the past 30 years. Surgery.
2017;161(1):137–46.
26. Kuo EJ, Sho S, Li N, Zanocco KA, Yeh MW, Livhits MJ. Risk
factors associated with reoperation and disease-specific mortality
in patients with medullary thyroid carcinoma. JAMA Surg.
2018;153(1):52–9.
27. Meng K, Luo H, Chen H, Guo H, Xia W. Prognostic value of
numbers of metastatic lymph node in medullary thyroid carci-
noma: a population-based study using the SEER 18 database.
Medicine (Baltimore). 2019;98(1):e13884.
28. Tingyin Jiang CH, Yuan X. Ratio of positive lymph nodes: the
prognostic value in stage IV thyroid cancer. Oncotarget.
2017;8(45):79462–8.
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
N-Staging System for Medullary Thyroid Cancer