Key words
thyroid nodule - hypercalcitoninemia - procalcitonin - medullary - thyroid carcinoma
Introduction
Many authors recommend the measurement of serum calcitonin (Ctn) to screen for
sporadic medullary thyroid carcinoma (MTC) in patients with thyroid nodules. Despite
the pre-analytical and analytical advantages of procalcitonin (pro-Ctn) measurement,
the clinical experience with the diagnosis and follow-up of MTC is greater for Ctn
[1]. Thus, it does not yet seem possible
to replace Ctn with pro-Ctn measurement, but the latter can complement the former
[1].
In view of the very low risk of MTC, pro-Ctn measurement without cytology or a
molecular test to suggest this cancer would be of little utility in patients with
no
“suspicious” family or personal history who have basal
Ctn<10 ng/ml. On the other hand, MTC is very likely in
patients with thyroid nodules and basal Ctn>100 pg/ml in the
absence of any other known cause (such as chronic renal failure, another secreting
tumor). In addition, at these Ctn concentrations, pro-Ctn is almost always elevated
[2]
[3]
[4]
[5]. Thus, to screen for sporadic MTC, pro-Ctn
would be more useful in patients with basal Ctn>10 pg/ml
but<100 pg/ml [1]. In
this situation, stimulation tests are traditionally recommended to distinguish
between patients with and without MTC. Pentagastrin, the most widely studied
stimulus, is not available in many countries while few studies have evaluated
calcium stimulation specifically in patients with nodular disease and mild or
moderate hypercalcitoninemia [6]. There is
also no agreement regarding the best cut-off for stimulated Ctn. Finally, adverse
reactions to these tests may occur, which are even contraindicated in some
patients.
We previously investigated patients with thyroid nodules and without a suspicion of
familial MTC or type 2 multiple endocrine neoplasia (MEN 2) who had mild or moderate
basal hypercalcitoninemia without an apparent cause [6]. The aim of the present study was to evaluate the utility of pro-Ctn
specifically in this situation.
Patients and Methods
Patients
Consecutive patients with nodular thyroid disease undergoing routine basal Ctn
measurement were selected. The following subjects were excluded: (i) children
and adolescents (age≤12 years), (ii) patients with a family history of
MTC or MEN 2 or with a clinical suspicion of the latter, (iii) patients
previously submitted to thyroid carcinoma surgery, (iv) patients who only had
hot nodules on 131I scintigraphy (performed if TSH<0.4
mIU/l) or purely cystic nodules, and (v) patients with known presence of
kidney failure, a neuroendocrine tumor, or lung cancer. The 60 patients with
basal Ctn>10 pg/ml but<100 pg/ml
(20 men and 40 women, age between 16 and 78 years) were included in this study.
The basal Ctn results of the first 41 patients had been published previously
[6].
The local research ethics committee approved the study.
Management
Patients with stimulated Ctn>100 pg/ml were submitted to
total thyroidectomy with elective dissection of the cervical lymph nodes. All
patients with non-benign cytology were operated. Only patients with stimulated
Ctn<100 pg/ml and benign cytology were not submitted to
thyroidectomy and were followed up by repetition of ultrasonography, Ctn
measurement, and fine needle aspiration (FNA) [6].
Sonography and FNA
Sonography was performed with a linear multifrequency transducer for
morphological analysis (B-mode) and for power Doppler evaluation. FNA was guided
by ultrasonography. The smears (cytology and histology) were analyzed by
pathologists experienced in thyroid pathology. In these patients with basal Ctn
between 10 and 100 pg/ml, a new FNA was performed to measure Ctn
in the needle washout (FNA-Ctn) [6].
Calcitonin and pro-calcitonin measurement
For Ctn and pro-Ctn measurement, the patients were asked to abstain from alcohol
for at least one week and to discontinue the use of proton pump inhibitors for
at least 4 weeks. None of the patients had apparent bacterial infection or
hypercalcemia or used supplements containing biotin at the time of measurement.
The serum samples were obtained in the morning (at about 8 AM) after an 8- to
10-hour fast and were analyzed immediately after collection.
Assays
Calcitonin was measured by an immunochemiluminescent assay, with a sensitivity of
2 pg/ml. Procalcitonin was measured by an automated
enzyme-linked fluorescent immunoassay with a functional sensitivity of
0.05 ng/ml [7].
Results
Histology was available for 52 patients and the eight patients not submitted to
surgery were considered not to have MTC [these patients initially had benign
cytology, stimulated Ctn<100 pg/ml, and
FNA-Ctn≤10 pg/ml; after 18 months, cytology remained
unchanged and there was no increase in basal Ctn or growth of nodules]. Nine
patients (15%) had MTC, with cytology being diagnostic in only four ([Table 1]).
Table 1 Characteristics of patients with MTC.
|
Sex
|
Age (years)
|
Ctn (pg/ml)
|
Pro-Ctn (ng/ml)
|
Cytology (Bethesda)
|
Tumor size
|
Stage
|
|
M
|
56
|
28
|
0.13
|
Benign (II)
|
12 mm
|
T1bN0M0
|
|
M
|
48
|
35
|
0.18
|
MTC (VI)
|
11 mm
|
T1bN0M0
|
|
F
|
49
|
43
|
0.23
|
Benign (II)
|
5 mm
|
T1aN0M0
|
|
F
|
53
|
56
|
0.42
|
Suspicious for PTC (V)
|
7 mm
|
T1aN0M0
|
|
F
|
56
|
72
|
0.5
|
MTC (VI)
|
16 mm
|
T1bN1aM0
|
|
M
|
63
|
81
|
0.45
|
Insufficient (I)
|
15 mm
|
T1bN1aM0
|
|
F
|
45
|
30
|
0.2
|
Indeterminate (IV)
|
7 mm
|
T1aN0M0
|
|
F
|
58
|
76
|
0.55
|
MTC (VI)
|
16 mm
|
T1bN1aM0
|
|
F
|
61
|
65
|
0.38
|
Suspicious for MTC (V)
|
13 mm
|
T1bN0M0
|
MTC: Medullary thyroid carcinoma; Ctn: Serum calcitonin; F: Female; M: Male;
Pro-Ctn: Procalcitonin; PTC: Papillary thyroid cancer.
Among the 51 patients without MTC, pro-Ctn was≤0.1 ng/ml in
46 patients and ranged from 0.12 to 0.23 ng/ml in the remaining
patients. All patients with MTC had pro-Ctn>0.1 ng/ml (range
0.13 to 0.55 ng/ml). MTC was not detected in any of the patients
with pro-Ctn≤0.1 ng/ml (n=46), while all patients
with pro-Ctn>0.25 ng/ml (n=5) and 44.4% of
patients with pro-Ctn between 0.1 and 0.25 ng/ml (n=9) had
this tumor.
Regarding basal Ctn, none of the patients with
concentrations<24.6 pg/ml (n=42) had MTC, while all
five patients with concentrations>47 pg/ml and 30.7%
of patients with concentrations between 24.6 and 47 pg/ml
(n=13) had this tumor. In the last group (intermediate basal Ctn), although
the four patients with MTC had pro-Ctn>0.1 ng/ml,
5/9 patients without this tumor also had elevated pro-Ctn. It is noteworthy
that among patients with basal Ctn>24.6 pg/ml (n=18)
pro-Ctn>0.1 ng/ml identified all cases with MTC and
64.2% of the subjects with these pro-Ctn concentrations actually had this
tumor.
In this group of patients with sporadic nodular disease and mild or moderate
hypercalcitoninemia without an apparent cause in which the frequency of MTC was
15%, the sensitivity, specificity, negative predictive value (NPV), positive
predictive value (PPV) and accuracy of pro-Ctn (cut-off 0.1 ng/ml)
were 100, 90.2, 100, 64.2 and 91.7%, respectively. Basal Ctn (cut-off
24.6 pg/ml) showed sensitivity, specificity, NPV, PPV and accuracy
of 100, 82.3, 100, 50 and 85%, respectively.
Discussion
Few studies have evaluated pro-Ctn specifically for the diagnosis of sporadic MTC
in
patients with nodular thyroid disease [5]
[8]
[9].
For the reasons mentioned above, pro-Ctn in this situation would be useful
especially for subjects with hypercalcitoninemia<100 pg/ml
[1]. A small number of patients with
nodular disease and mild or moderate hypercalcitoninemia have so far been evaluated
using pro-Ctn measurement [5]
[8]
[9].
In our series, using a cut-off of 0.1 ng/ml, pro-Ctn had 100%
sensitivity and of 90% specificity. A non-inferior performance was obtained
with basal Ctn, with a cut-off of 24.6 pg/ml resulting in
100% sensitivity and 82% specificity. In the first study of
Giovanella et al. [8], using the cut-off
proposed by the authors (0.1 ng/ml), pro-Ctn provided sensitivity
and specificity of 100%; however, the same performance would be achieved
with basal Ctn using a cut-off of 35 pg/ml. In another study using
cut-offs with better performance, Giovanella et al. [9] obtained sensitivity of 100% and specificity of 99.7%
for pro-Ctn, but basal Ctn also provided sensitivity of 100% and specificity
of 98.3%. Finally, in the recent study by Censi et al. [5], the cut-off proposed for pro-Ctn
(0.07 ng/ml) provided sensitivity of 85.7% (97.4%
for tumors>10 mm) and specificity of 98.9%. Applying the
basal Ctn cut-off of 25 pg/ml, 92% sensitivity
(97.4% for tumors>10 mm) and 96.6% specificity are
also obtained.
Since we had no cases of MTC with Ctn≤20–25 pg/ml, we
do not know whether the excellent sensitivity of pro-Ctn would persist in this
situation. Giovanella et al. [8]
[9] reported 100% sensitivity of pro-Ctn
but the patients with MTC had Ctn>20 pg/ml. In the series of
Machens et al. [3], among the seven patients
with MTC and calcitoninemia<20 pg/ml, pro-Ctn
was≤0.1 ng/ml in five patients
and≤0.15 ng/ml in all patients. In the study by Censi et al.
[5], three of the four patients with MTC
and Ctn≤20 ng/ml also had
pro-Ctn≤0.07 ng/ml and all of them had
pro-Ctn≤0.1 ng/ml. Finally, in another study, all five
patients with active MTC and basal Ctn<20 pg/ml had
pro-Ctn<0.15 ng/ml [2].
On the other hand, we did not have patients without MTC with
Ctn≥50 pg/ml and we do not know whether the specificity of
pro-Ctn would also be high in these cases. In the series of Giovanella et al. [8]
[9]
[10] in which pro-Ctn provided
specificity close to 100%, subjects without MTC had
Ctn<50 pg/ml. The same was observed in the study by Censi et
al. [5] in which all patients without MTC
exhibited Ctn<30 ng/ml and 94.3%
had<20 ng/ml.
Among patients with hypercalcitoninemia between 20 and 50 pg/ml,
pro-Ctn>0.1 ng/ml identified all four patients with MTC in
the present study, the two cases of the series of Giovanella et al. [9], and 7/10 patients in the series of
Censi et al. [5]. In our study, 5/9
patients without MTC and with Ctn between 20 and 50 ng/ml also had
pro-Ctn>0.1 ng/ml. In contrast, none of the patients without
MTC with these Ctn concentrations had elevated pro-Ctn in the series of Giovanella
et al. [8]
[9]
[10] and Censi et al. [5].
In conclusion, we did not find superiority of pro-Ctn measurement over basal Ctn for
the diagnosis of sporadic MTC in patients with nodular disease and mild or moderate
hypercalcitoninemia. More studies are necessary to specifically assess the
sensitivity of pro-Ctn in patients with MTC and mild hypercalcitoninemia
(≤20 pg/ml) and its specificity in patients without MTC with
higher Ctn concentrations (≥50 pg/ml). In the case of
patients with hypercalcitoninemia in the gray zone, pro-Ctn has an excellent NPV but
the data regarding its PPV are not uniform and further studies are needed.
