Key words papillary microcarcinoma - aggressive subtype - clinical outcome
Introduction
Papillary thyroid microcarcinoma (PTMC) is defined as papillary thyroid carcinomas
(PTC) measuring ≤ 1 cm in size [1 ]. The clinical importance of PTMCs is
controversial with many reports suggesting that it may represent a subclinical
disease that is non-progressive and has no significant effect on morbidity or
survival [2 ]. However, even though the
clinical course of PTMCs is generally indolent, a small number of PTMCs,
particularly those with histologically aggressive subtypes may be associated with
an
aggressive clinical course [3 ].
The histologically aggressive subtypes of PTC are the diffuse sclerosing, tall cell,
columnar cell, solid, and hobnail subtypes [3 ]
[4 ]. Studies suggest higher rates
of bilateral disease, multifocality, extrathyroidal extension [ETE, either
microscopic (mETE) or macroscopic], locoregional recurrence, neck lymph-node
(LN)/distant metastasis, decreased survival, and in some cases, absence of
avidity to radioiodine (RAI) in patients with aggressive subtypes of PTC [3 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ].
Therefore, the aggressive subtypes of PTC have been classified as
“intermediate risk” for recurrence in the latest American Thyroid
Association (ATA) guidelines, irrespective of their size [5 ]. Unfortunately, despite total thyroidectomy
and central LN dissection have been recommended for the treatment of aggressive
subtypes of PTMC, subtyping of PTC by cytopathological evaluation of the fine needle
aspiration (FNA) specimens is challenging, even for the most experienced
cytopathologists [6 ]
[12 ]. Therefore, most of the cases are diagnosed
after the histopathological evaluation of surgical specimens [6 ]
[12 ].
To date, only a few studies have evaluated the clinical outcomes of histologically
aggressive PTMC subtypes. In a population-level analysis, aggressive PTMC subtypes
were suggested to exhibit more aggressive pathologic characteristics than classic
PTMCs, but the overall and disease-specific survival rates were similar [6 ]. Considering the recent rising trend towards
active surveillance as well as thermal ablation rather than surgical treatment, and
challenges in the cytological diagnosis of aggressive subtypes in FNA specimens,
determining the differences in clinical outcomes such as
persistent/recurrent disease, disease-free survival (DFS) rate, and dynamic
risk stratification between histologically aggressive and non-aggressive PTMC
subtypes is of paramount importance [13 ]
[14 ]. However, although several studies were
carried out to compare the clinical outcomes between histologically aggressive and
classic subtypes of larger PTCs, the difference in clinical outcomes between
patients with histologically aggressive and non-aggressive PTMC subtypes have not
been exclusively studied [3 ]
[15 ]
[16 ].
Therefore, in this multicenter cohort study, we intended to find answers to certain
questions such as: do histologically aggressive subtypes of PTMC have worse clinical
outcomes than non-aggressive subtypes? If so, should patients receive a more
aggressive treatment based on the presence of aggressive histology? We also aimed
to
evaluate whether aggressive histology is an independent predictor of
persistent/recurrent disease when controlling for other contributing risk
factors.
Materials and Methods
Patient selection and study design
The computer-recorded data collected from 9 referral hospitals were analyzed. The
patients included in this study consisted of consecutive individuals who
underwent hemithyroidectomy or total thyroidectomy for toxic nodular thyroid
disease, nodular thyroid disease with suspicious FNAC results, or
Graves’ disease, between January 2000-January 2021, and were diagnosed
as PTMC by histopathological analysis. The primary objectives of this study were
to determine the differences in structural persistent/recurrent disease
and DFS rates, and the secondary objective was to assess the dynamic risk
stratification results, between patients with histologically aggressive and
non-aggressive PTMC subtypes. The following data were recorded for each patient:
Patient and tumor characteristics, the extent of surgery, ETE, central or
lateral neck LN involvement or distant metastasis at diagnosis, radioactive
iodine (RAI) ablation, persistent disease, and development of locoregional
recurrence or distant metastasis during the follow-up period. To assess whether
the diagnosis was compatible with the recent WHO classification of tumors [17 ], the tissue sections of the patients
with a histopathological diagnosis of aggressive subtypes were re-examined by
experienced cytopathologists in each center. All patients were classified
according to the 8th edition of the American Joint Committee on Cancer
(AJCC)/Union for International Cancer Control TNM staging system and the
2009 revised ATA management guidelines for thyroid nodules [18 ]
[19 ]. However, due to the lack of information regarding the size of
metastasis in resected neck LNs, the patients could not be stratified according
to the 2015 ATA guidelines for thyroid nodules [5 ]. For those patients with more than one PTMC, the largest tumor was
recorded and considered for analysis, unless the histopathological analysis
results were consistent with one of the aggressive subtypes. Patients
with<1 year of follow-up data, patients with any missing data, patients
with concomitant thyroid carcinomas larger than 1 cm,
patients<18 years old, and patients who were diagnosed as non-invasive
follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) after
2016, were not included in the study. This study was performed in line with the
principles of the Declaration of Helsinki. Approval was granted by the Ethics
Committee of Tekirdag Namık Kemal University, School of Medicine.
Treatment and follow-up for clinical outcomes
The treatment and follow-up protocols of the patients were carried out as
previously reported [20 ]. All patients had
undergone total thyroidectomy or hemithyroidectomy. Decisions for the extent of
surgery, completion of thyroidectomy, central or lateral cervical LN dissection,
and RAI treatment were made based on the institutional guidelines at the time of
the initial surgery. All patients were followed with serum TSH, free T4,
thyroglobulin (Tg), and anti-Tg antibody levels at 3- to 6-month intervals as
well as ultrasonographic examinations of the neck at least once a year.
Stimulated Tg level assessment and other imaging studies including diagnostic
iodine 131 (131 I) whole-body scans (WBS), 18-FDG PET/CT,
computed tomography, magnetic resonance imaging, and bone scan were also
performed as needed. During the follow-up period, neck LNs with any suspicious
or indeterminate appearance on ultrasonography underwent fine-needle aspiration
cytology and/or measurement of Tg in needle washout fluid. Patients were
accepted as “disease free” if they had no cytopathologic and
imaging evidence of disease, undetectable suppressed Tg, and anti-Tg antibody
levels in those who had undergone total thyroidectomy with or without RAI
treatment. A cut-off value of≤30 ng/dl for Tg was used
to define “disease-free status” in patients who underwent
hemithyroidectomy. For the assessment of dynamic risk stratification, patients
were classified into four response groups, according to the dynamic risk
stratification system (excellent, indeterminate, biochemical incomplete, and
structural incomplete responses) [5 ].
Statistical analysis
Predictive Analytics Software Statistics 18 (International Business Machines
Corporation) for Windows was used for data input and statistical analysis. An
independent sample t -test or Mann–Whitney U-test was used to
compare 2 groups, whereas an analysis of variance was used to compare≥3
groups, followed by a Tukey (or Tamhane T2) test for subgroup comparisons. A
chi-square test was used for categorical data comparison. DFS was estimated
using the Kaplan–Meier method. Cox’s proportional hazards model
was used to evaluate the effect of age, sex, tumor size, neck LN involvement,
multifocality, tumor stage, mETE, and aggressive histology on
persistent/recurrent disease. All tests used were two-sided and A
p- value of<0.05 was considered statistically significant.
Results
After excluding 131 patients with insufficient follow-up data, a total of 1585
patients with PTMC [female 1340 (84.5%), male 245 (15.5%)], mean age
47.9±11.63 years, min. 18, max. 80 years) with a mean follow-up time of
66.55±37.16 months (min. 13, max. 240 months) were included in the study.
The baseline clinical and histopathological characteristics and follow-up outcomes
of the cohort according to the extent of surgery are presented in [Table 1 ]. In this study, 98 (6.2%)
cases with PTMCs were histopathologically diagnosed as aggressive (diffuse
sclerosing 2.8%, tall cell 2.6%, columnar cell 0.4%, and
solid subtype 0.3%), and 1487 cases (93.8%) as non-aggressive
subtypes ([Table 1 ]). Patients with
aggressive PTMC subtypes tended to be younger compared to patients with
non-aggressive subtypes (42.6±12.4 vs. 48.24±11.5 years,
p<0.001), while no significant difference was found in terms of sex between
both groups (p=0.50) ([Table 2 ]). On
the other hand, patients with aggressive PTMC subtypes tended to have larger tumor
sizes, bilateral and multifocal disease, mETE, as well as higher rates of central
and lateral neck LN involvement than patients with non-aggressive subtypes
(p<0.001 for all comparisons) ([Table
2 ]). Furthermore, more patients with aggressive PTMC subtypes had
undergone completion thyroidectomy, central and lateral neck LN dissection as well
as RAI ablation than patients with non-aggressive PTMC subtypes (p=0.005,
p<0.001, p=<0.001, and p<0.001, respectively).
Likewise, more patients with aggressive PTMC subtypes had stage II disease than
patients with non-aggressive subtypes (p=0.005) ([Table 2 ]). Persistent/recurrent
disease was observed in 33 (33.6%) and 41 (2.8%) patients with
aggressive and non-aggressive PTMC subtypes, respectively (p<0.001) ([Table 2 ]). Patients with aggressive PTMC
subtypes also tended to have lower rates of excellent response and higher rates of
incomplete biochemical and structural response rates than patients with
non-aggressive subtypes (p<0.001) ([Table
2 ]). Distant metastasis at the initial diagnosis was present in one
patient with a non-aggressive subtype and one patient with a tall cell subtype,
while one patient with a non-aggressive and one with a tall cell subtype developed
distant metastasis during follow-up. On the other hand, 5 (0.3%) patients
with non-aggressive subtypes had a macroscopic ETE, while no patient with an
aggressive subtype had macroscopic ETE. Considering patients with aggressive PTMC
subtypes, no marked difference was observed between patients with tall cell and
diffuse sclerosing subtypes in terms of age, sex, type of surgery, bilateral and
multifocal disease, neck LN involvement, mETE, tumor stage, and initial ATA risk
stratification (p>0.05 for all comparisons) ([Table 3 ]). However, more patients with the
tall cell subtype had a tumor size of≥5 mm than patients with the
diffuse sclerosing subtype (p=0.039), and although statistically only
marginally significant, persistent/recurrent disease and incomplete
structural response rates were more common in patients with tall cell than in
patients with diffuse sclerosing subtype (p=0.05, and p=0.07,
respectively) ([Table 3 ]). Although the
number of patients with columnar and solid PTMC subtypes was not adequate for
within-group analysis, persistent/recurrent diseases were observed in 2
patients with a solid subtype, while no patient with a columnar subtype had
persistent/recurrent disease.
Table 1 Baseline clinical and histopathological
characteristics of 1585 patients with PTMC according to the extent of
surgery.
Variables
Hemithyroidectomy n=160 (10%)
Total Thyroidectomy n=1425 (90%)
Total Cohort n=1585 (100%)
Age (years)
46 .03±11.44
48.1±11.64
47.9±11.63
<55
127 (79.4)
1024(71.9)
1151 (72.6)
>55
33 (20.6)
401(28.1)
434 (27.4)
Sex
Female
133 (83.1)
1207(84.7)
(84.5)
Male
27 (16.9)
218(15.3)
245 (15.5)
Completion thyroidectomy
65 (4.1)
–
–
Neck LN Dissection
Central
–
145(10.1)
145 (9.1)
Lateral
–
11(0.77)
11 (0.7)
Central and Lateral
–
96(6.7)
96 (6.1)
Histopathologic subtype
Classic
109 (68.1)
916(64.3)
1025(64.7)
Follicular
32 (20)
341(23.9)
373(23.5)
Oncocytic
5 (3.1)
73(5,1)
78 (4.9)
Diffuse sclerosing
3 (1.9)
42(2.9)
45 (2.8)
Tall cell
7 (4.4)
35(2.5)
42 (2.6)
Columnar cell
1 (0.6)
5(0.4)
6 (0.4)
Solid subtype
1 (0.6)
4 (0.3)
5 (0.3)
Clear cell
1 (0.6)
3 (0.2)
4(0.2)
Warthin-like
1 (0.6)
6(0.4)
7(0.4)
Tumor diameter (mm)
5.75±2.91
5.81±2.79
5.8±2.8
<5 mm
82 (51.2)
653(45.8)
735(46.4)
≥5 mm
78 (48.8)
772(54.2)
850(53.6)
Tumor foci and location
Unilateral
160 (100
1044(73.3)
1204 (76)
Bilateral
–
381(26.7)
381 (24)
Multifocal
41(25.6)
630(44.2)
671 (42.3)
Unifocal
119(74.4)
795(55.8)
914 (57.7)
Microscopic ETE
15(9.4)
115(8.1)
130 (8.2)
Neck LN involvement at diagnosis
–
107 (7.5)
107 (6.8)
Central (N1a)
–
71(5)
71 (4.5)
Lateral (N1b)
–
36(2.5)
36 (2.3)
RAI treatment TNM Staging
#
–
562(39.4)
562 (35.5)
Stage I
160(100)
1406(98.7)
1566 (98.8)
Stage II
–
19(1.3)
19 (1.2)
Initial ATA risk stratification
Low
134(83.7)
1225(86)
1359 (85.7)
Intermediate
26(16.3)
197(13.8 )
223 (14.1)
High
–
3 (0.2)
3 (0.2)
Follow-up time
65.03(42.61)
66.72(36.51)
66.5±37.16
Persistent/recurrent disease
13(8.1)
61(4.3)
74 (4.7%)
ATA Dynamic risk stratification
Excellent response
145(90.6)
1146(80.4)
1291(81.5)
Indeterminate response
9(5.6)
227(15.9)
236(14.9)
Incomplete biochemical response
6(3.8)
41(2.9)
47(3)
Incomplet structural response
–
11(0.8)
11(0.7)
PTMC: Papillary thyroid microcarcinoma; LN: Lymph node.;
#
No patient in the cohort was classified as stage III or IV; ATA: American
Thyroid Association.
Table 2 Clinical and histopathological characteristics of 1585
patients with histologically aggressive and non-aggressive PTMC
subtypes.
Variables
Aggressive subtype n=98 (6.1%)
Non-aggressive subtype n=1487 (93.9%)
p-Value
Age (years)
42.6±12.4
48.2±11.5
<0.001
<55
83(84.7)
1068(71.8)
>55
15(15.3)
419(28.2)
0.005
Sex
Female
83 (84.7)
1257 (84.5)
Male
15 (15.3)
230 (14.5)
0.5
Type of surgery
Total thyroidectomy
86 (87.8)
1139 (90)
Hemithyroidectomy
12 (12.2)
148 (10)
0.48
Completion thyroidectomy
10 (10.2)
55 (3.7)
0.005
Neck LN Dissection
49 (50)
203 (13.6)
<0.001
Central
35(35.7)
110 ( 7.4)
<0.001
Lateral
2 (2)
9 (0.6)
<0.001
Central+Lateral
12 (12.2)
84 (5.6)
<0.001
Tumor size (mm)
7.16±2.26
5.71±2.81
<0.001
<5
25(25.5)
710(47.7)
≥5
73(74.5)
777(52.3)
<0.001
Tumor foci and location
Unilateral
52( 53.1)
1151 (77.4)
Bilateral
46(46.9)
335 (22.6)
<0.001
Multifocal
58(59.2)
613(41.2)
Unifocal
40(40.8)
874(58.8)
<0.001
Microscopic ETE
27 (27.6)
103 (6.9)
<0.001
Neck LN involvement at diagnosis
Central (N1a)
18 (18.4)
53 (3.6)
<0.001
Lateral (N1b)
10 (10.2)
26 (1.7)
<0.001
RAI treatment
85 (86.7)
477 (32.1)
<0.001
TNM Staging
#
Stage I
93(94.9)
1473 (99.1)
Stage II
5 (5.1)
14 (0.9)
0.005
Initial ATA risk stratification
Low
–
1359 (91.4)
Intermediate
96 (98)
127 (8.5)
High
2 (2)
1 (0.1)
<0.001
Persistent/recurrent disease
33 (33.7)
41(2.8)
<0.001
ATA dynamic risk stratification
Excellent response
65 (66.3)
1226 (82.4)
Indeterminate response
13 (13.4)
223 (15)
Incomplete biochemical response
11 (11.2)
36 (2.4)
Incomplete structural response
9 (9.2)
2 (0.1)
<0.001
PTMC: Papillary thyroid microcarcinoma; LN: Lymph node; ETE: Extrathyroidal
extension.; # No patient in the cohort classified as stage III or
IV; ATA: American Thyroid Association.
Table 3 The differences in histopathological characteristics
and clinical outcomes of the 98 patients with aggressive PTMC
subtypes.
Variables
Diffuse Sclerosing n=45 (45.9%)
Tall Cell n=42 (42.8%)
p-Value
Solid subtype n=5 (5.1%)
Columnar cell n=6 (6.1%)
Age (years)
43.20±11.84
41.81±12.93
0.6
42.80±18.8
43.33±9.24
<55
37(82.2)
36(85.7)
4(80)
6(100)
>55
8(17.8)
6(14.3)
0.44
1(20)
–
Sex
Female
41 (91.1)
33(78.6)
4(80)
5(83.3)
Male
4(8.9)
9(21.4)
0.09
1(20)
1(16.7)
Type of surgery
Total Thyroidctomy
42 (93.3)
35(83.3)
4(80)
5(83.3)
Hemithyroidectomy
3 (6.7)
7 (16.7)
0.13
1(20)
1(16.7)
Completion thyroidectomy
3(6.7)
6 (14.2)
0.2
1(20)
–
Neck LN Dissection
Central
14(31.1)
17(40.5)
1(20)
3(50)
Lateral
2(4.4)
–
–
–
Central and Lateral
6(13.3)
5(11.9)
0.47
1(20)
–
Tumor diameter (mm)
6.8±2.5
7.5±1.9
0.136
7.4±3.4
7.1±1.4
<5 mm
16(35.6)
7(16.7)
1(20)
1(16.7)
≥5 mm
29(64.4)
35(83.3)
0.039
4(80)
5(83.3)
Tumor foci and location
Unilateral
22(48.9)
22(52.4)
4(80)
4(66.7)
Bilateral
23(51.1)
20(47.6)
0.45
1(20)
2(33.3)
Multifocal
26 (57.8)
28(66.7)
1(20)
3(50)
Unifocal
19 (42.2)
14(33.3)
0.26
4(80)
3(50)
Microscopic ETE
15 (33.3)
12 (28.6)
0.4
–
–
TNM Staging
#
Stage I
41(91.1)
41(97.6)
5(100)
6(100)
Stage II
4 (8.9)
1(2.4)
0.2
–
–
Neck LN involvement at diagnosis
Central (N1a)
7 (15.6)
8 (19)
2(40)
–
Lateral (N1b)
7 (15.6)
3 (7.1)
0.45
–
–
RAI treatment
38(84.4)
37(88.1)
0.42
5(100)
6(100)
Initial ATA risk stratification
Low
–
–
–
–
Intermediate
44(97.8)
41(97.6)
5(100)
6(100)
High
1 (2.2)
1(2.4)
0.73
–
–
Persistent/recurrent disease
12 (26.7)
19(45.2)
0.05
2(40)
–
ATA Dynamic risk stratification
Excellent response
33(73.3)
25(59.5)
4(80)
3(50)
Indeterminate response
7(15.6)
5(11.9)
1(20
–
Incomplete biochemical
Response
4(8.9)
4(9.5)
–
3(50)
Incomplete structural
response
1(2.2)
8(19)
0.07
–
–
The number of patients with columnar and solid subtypes was not eligible for
statistical analysis.; PTMC: Papillary thyroid microcarcinoma; ETE:
Extrathyroidal extension; LN: Lymph node.; # No patient in the
cohort classified as stage III or IV.
Kaplan–Meier and Cox regression analysis
In the Kaplan–Meier analysis, the overall DFS rate was 95.4%.
However, DFS was markedly lower in patients with aggressive than in those with
non-aggressive PTMC subtypes [66.3%, estimated median DFS 171.5 months
(95% CI: 151–191.9) vs 94.8%, estimated median DFS 193
months (95% CI: 188.7–197.27), respectively (Log-Rank
p<0.001) ([Fig. 1 ]). Considering
Cox’s model, in univariate analysis, male sex, tumor
size≥5 mm, mETE, neck LN involvement, tumor stage, and
multifocality as well as aggressive histology were all independent predictors of
persistent/recurrent disease, while age>55 was not an
independent predictor of persistent/recurrent disease. However, in
multivariate analysis, male sex, neck LN involvement, tumor stage, as well as
aggressive histology were independent predictors of persistent/recurrent
disease while tumor size≥5 mm, mETE, and multifocality were not
independent predictors of persistent/recurrent disease ([Table 4 ]).
Fig. 1 Kaplan–Meier curves displaying the estimated
persistent/recurrent disease-free survival probability in patients with
histologically aggressive and non-aggressive PTMC subtypes.
Table 4 The effect of aggressive histology as well as age,
sex, tumor size, neck lymph node involvement, microscopic
extrathyroidal extension, multifocality, and stage on
persistent/recurrent disease in patients with PTMC according
to Cox’s proportional hazard model.
Variables
HR (95% CI)
p-Value
Univariate analysis
Age (years)
Ref.
>55
1.07 (0.62–1.85)
0.78
Sex
Ref.
Male
2 (1.14–3.57)
0.015
Tumor size
Ref.
≥5 mm
3.2 (1.84–5.55)
<0.001
Neck LN involvement at diagnosis
Ref.
12.78 (8.0–20.3)
<0.001
Microscopic ETE
Ref
5.18 (3.2–8.3)
<0.001
Multifocal tumor
Ref
3.1 (1.94–5.14)
<0.001
Stage
Ref
II
10.4 (4.7–22.7)
<0.001
Aggressive histology
Ref.
12.3 (7.77–19.6)
<0.001
Multivariate analysis
Sex
Ref.
Male
1.86 (1.04–3.33)
0.037
Tumor size
Ref.
≥5 mm
1.72(1.06–3.22)
0.079
Neck LN involvement at diagnosis
Ref.
4.4 (2.31–8.47)
<0.001
Microscopic ETE
Ref
1.27 (0.72–2.26)
0.4
Multifocal tumor
Ref
1.3 (0.75–2.24)
0.34
Stage
Ref
II
3.12 (1.13–8.3)
0.028
Aggressive histology
Ref.
5.78 (3.32–10)
<0.001
PTMC: Papillary thyroid microcarcinoma; ETE: Extrathyroidal
extension.
Discussion
In the present study, persistent/recurrent disease as well as incomplete
biochemical and structural response rates were more common in patients with
aggressive than in patients with non-aggressive PTMC subtypes ([Table 2 ]). The DFS rate was markedly lower in
patients with aggressive PTMC subtypes compared to patients with non-aggressive PTMC
subtypes as well. Moreover, in multivariate analysis, aggressive histology was an
independent predictor of persistent/recurrent disease, after controlling for
other contributing factors ([Table 2 ]
[4 ]).
In recent years, an increase in the frequency of PTMC has been observed [21 ]. Generally, studies report excellent
clinical outcomes and DFS rates in patients with PTMC. In a recent meta-analysis
comparing DFS among patients with PTMC who underwent thyroidectomy, the 10-year DFS
rates were found as 95% and 92% after total thyroidectomy and
hemithyroidectomy, respectively [22 ]. However,
a recurrence rate as high as 19% has been reported as well [23 ]. To the best of our knowledge, no study has
evaluated the impact of aggressive histology on the clinical outcomes of PTMC so
far. As in prior studies [22 ], the overall
persistent/recurrent disease and DFS rates in our study were 4.7 and
95.3%, respectively. However, when persistent/recurrent disease and
DFS rates were analyzed according to the histopathological subtypes,
persistent/recurrent diseases were significantly more common and DFS rates
were markedly lower in patients with histologically aggressive than in those with
non-aggressive PTMC subtypes (2.8 vs. 33.7%, p<0.001, and 97.3 vs
66.3%, p<0.001, respectively).
Since most of the PTMCs have an indolent clinical course, it has raised the question
of which treatment method is a more appropriate option for these patients [24 ]. Nowadays, active surveillance is being
recommended over immediate surgery in patients with low-risk PTMC [21 ]. In a recent systematic review, Chou et al.
suggested that in patients with small low-risk DTC, active surveillance and
immediate surgery may have a similar mortality rate and risk of recurrence [25 ]. In a study by Kudo et al., no significant
difference was observed in disease-specific and overall survival among patients with
classic, tall cell, and diffuse sclerosing PTMC subtypes, but patients with tall
cell and diffuse sclerosing subtypes tended to have more frequent mETE, nodal
metastasis, and multifocality than patients with classic PTMC [6 ]. However, they haven’t analyzed DFS
rates and dynamic risk stratification results between classic and aggressive PTMC
subtypes [6 ]. Our study results were in line
with that reported by Kudo et al. Neck LN involvement, mETE, multifocal and
bilateral disease were all more common in patients with aggressive PTMC subtypes
([Table 2 ]). Furthermore, patients with
aggressive PTMC subtypes tended to have larger tumor sizes, more common
stage–II disease, as well as higher rates of completion thyroidectomy and
RAI treatment than patients with non-aggressive subtypes ([Table 2 ]). In addition to the findings of Kudo
et al., our study results suggest a markedly lower DFS and higher
persistent/recurrent diseases in patients with aggressive PTMC subtypes
[6 ]. Therefore, aggressive PTMC subtypes
share many clinical and histopathological characteristics with their identical
tumors>1 cm in size and may not be appropriate for active
surveillance and should be treated as histologically identical
tumors>1 cm in size.
Tall cell and diffuse sclerosing subtypes are the most frequently observed aggressive
subtypes of PTMC [6 ]. Studies indicated that
angiolymphatic and parenchymal invasion, ETE, neck LN involvement, locoregional
recurrence, and distant metastasis are more frequent in patients with tall cell than
in non-aggressive PTC subtypes [26 ].
Furthermore, tall cell histology was suggested to be an independent predictor of
neck LN involvement in patients with PTMC [27 ]. In a recent meta-analysis, comparing the tall cell subtype and classic
PTC, multifocality, ETE, neck LN involvement, distant metastasis, and cancer-related
mortality were all significantly more common in patients with tall cell subtype than
in patients with classic PTC [28 ]. Moreover,
tumor recurrence was more common in patients with the tall cell than in those with
classic PTC as well (OR 5.12, 95% CI 1.7–15.44, p=0.004)
[24 ]. The diffuse sclerosing subtype is
consisting about %6 of PTCs. In a recent meta-analysis including 732
patients with diffuse sclerosing subtype, vascular invasion, ETE, neck LN
involvement, and distant metastasis were all more common in patients with diffuse
sclerosing than patients with classic PTC [29 ]. Moreover, persistent/recurrent disease was observed in
22% of the patients with diffuse sclerosing and 10.7% of patients
with classic PTC (OR 2.83; 95% CI 1.59–5.05). In our study,
persistent/recurrent disease and incomplete structural response rates were
observed in 45.2%, 19%, 26.7, and 2.2% of the patients with
tall cell and diffuse sclerosing subtypes, respectively (p=0.05 and
p=0.07). Therefore, the presence of tall cell histology, even in patients
with PTMC is associated with higher rates of persistent/recurrent diseases
and structural incomplete response and should be taken into consideration when
deciding on surgical treatment or active surveillance.
Solid and columnar subtypes of PTC are the other rare subtypes of PTC. The solid
subtype is associated with a markedly higher risk for vascular invasion, tumor
recurrence, and cancer mortality than classic PTC [30 ]. On the other hand, the columnar subtype is variable in biological
behavior, some are clinically aggressive, whereas others are more clinically
indolent [31 ]. In our study, only 5 patients
had solid and 6 had columnar PTMC subtypes. Persistent/recurrent disease was
observed in 2 patients with a solid PTMC subtype and non of the patients with a
columnar subtype. However, the number of patients with solid and columnar PTMC
subtypes was not adequate to make a within-group analysis.
Neck LN involvement is an independent predictor of recurrence in patients with PTMC
[32 ]. PTMC with lateral LN involvement is
more likely to have biochemical or structural persistence or recurrence compared
with PTMC without LN involvement [33 ]. On the
other hand, male sex, age<55 years, multifocality, tumor
size>5 mm, and ETE are all found to be independent risk factors of
central and lateral neck LN metastasis [34 ]
[35 ]. On the contrary, bilateral
multifocality was also suggested to be only an indicator of aggressiveness but not
an independent predictor of worse clinical outcomes [36 ]. Although in the recent ATA guidelines, aggressive histology and mETE
have been classified as ''intermediate risk'' for
persistent/recurrent disease [5 ], a
recent study suggested that mETE in small PTCs is not an independent risk factor for
persistent/recurrent disease, and therefore, could be classified and treated
as ''low risk'' tumors [37 ]. In our study, male sex, neck LN
involvement, tumor stage as well as aggressive histology were independent predictors
of persistent/recurrent disease, while age at diagnosis, multifocality,
mETE, and size≥5 mm were not independent predictors of
persistent/recurrent disease. Therefore, according to our study results too,
PTMCs with mETE could be classified as “low-risk” tumors.
Unfortunately, despite aggressive PTC subtypes having unique cytopathologic
characteristics, preoperative diagnosis of aggressive PTMC subtypes is difficult,
and it is more common for patients to be diagnosed postoperatively after the
histopathologic examination [6 ]
[8 ]
[38 ].
Nevertheless, multiple mutations could be present in aggressive subtypes of PTC
[21 ]
[39 ], and the molecular and genetic analysis of the FNAC may provide some
clues regarding the histopathological subtype and aggressive clinical behavior [40 ]. For instance, BRAFV600E , TERT
promoter, and TP53 mutations are highly prevalent in the tall cell, columnar cell,
and solid subtypes. On the other hand, RET/PTC rearrangements appear to
predominate, while BRAFV600 mutation may be present in only 20%
of the cases with diffuse sclerosing subtype [17 ]. Therefore, further molecular analysis may help to identify them
before surgery.
Our study has several limitations. First of all, the retrospective nature of our
study makes it subject to bias. In addition, because the FNAs of the patients with
non-incidentally identified PTMCs were performed in different centers, we could not
analyze the FNAC results. Furthermore, only patients with sufficient follow-up data
were included in the study. Therefore, a significant number of patients, even those
with a proven persistent/recurrent disease had to be excluded. And finally,
molecular and genetic analysis had not been performed in any of the cases. However,
the strength of our study derives from the inclusion of a relatively large number
of
patients with aggressive PTMC subtypes defined based on the newly published WHO
criteria. Therefore, we believe that our study is an obvious example of daily
practice.
Conclusions
Our study results suggest that histologically aggressive PTMC subtypes are associated
with a markedly lower DFS rate as well as a higher frequency of
persistent/recurrent disease than non-aggressive subtypes. Aggressive
histology is an independent predictor of persistent/recurrent diseases as
well. Moreover, incomplete biochemical and structural response rates are more common
in patients with aggressive PTMC subtypes than in those with non-aggressive
subtypes. Aggressive PTMC subtypes share many characteristics with histologically
identical tumors>1 cm in size. Therefore, the histopathological
subtype of PTMC must be taken into serious consideration to tailor a personalized
management plan. Further studies are needed to improve our understanding of the
clinical outcomes of aggressive PTMC subtypes.
