Paschke_editQ8_Layout 1 02/09/2009 15:15 Page 61
Guideline-orientated Diagnosis of Thyroid Nodules
studies (level of evidence: 3 according to AACE scale), the ETA one Table 1: Ultrasound Characteristics Associated with
retrospective observational study (level of evidence: 3) and one
Thyroid Cancer
consensus (level of evidence: 4) and the AACE/AME five prospective
observational studies (level of evidence: 3) and six reviews (level of
Ultrasound Sensitivity Specificity Positive Negative
evidence: 4). For FNAB, the ATA reports three observational studies
Characteristics (%) (%) Predictive Predictive
Value (%) Value (%)
(level of evidence: 3), the ETA three observational studies (level of
Microcalcifications
1–5
26–59 86–95 24–71 42–94
evidence: 3) and one review (level of evidence: 4) and the AACE/AME
Hypoechogenicity
2–5
27–87 43–94 11–68 74–94
six observational studies (level of evidence: 3) and seven reviews
Irregular margins or 17–78 39–85 9–60 39–98
with a pooled analysis (level of evidence: 3–4).
no halo
2–5
Solid
4–6
69–75 53–56 16–27 88–92
It is noteworthy to observe that the evidence reported by the ATA, Intranodule 54–74 79–81 24–42 86–97
the ETA and the AACE/AME guidelines on thyroid ultrasound and
vascularity
3,6
FNAB lacks consistency. The three guidelines share only: The ability of single thyroid ultrasound criteria to predict malignancy has been summarised
in the recent consensus statement by radiologists. It documented low sensitivity and low
34
specificity for all single ultrasound criteria.
one reference in ATA and ETA (Marqusee et al.
31
);
two references in ATA and AACE/AME (Tan et al.
32
and Hagag et
Table 2: Summary of Characteristics for Thyroid
al.
33
); and
Fine-needle Aspiration
no references in the ETA Consensus and the AACE/AME GL.
Feature Mean (%) Range (%) Definition
With regard to the ability of thyroid ultrasound to predict malignancy,
Sensitivity 83 65–98 Likelihood that patient with disease
the recent consensus statement by radiologists documented low
has positive test results
sensitivity and low specificity for all single ultrasound criteria,
34
as Specificity 92 72–100 Likelihood that patient without
outlined in disease has negative test resultsTable 1. However, as suggested by several studies, most
likely the combination of several ultrasound criteria together with
Positive 75 50–96 Fraction of patients with positive test
clinical criteria suggestive of malignancy will lead to a better
predictive value results who have disease
selection of thyroid nodules for FNAB.
35,36 False-negative 5 1–11 Fine-needle aspiration negative;
rate histology positive for cancer
False-positive 5 0–7 Fine-needle aspiration positive;
According to the ATA guidelines, in the presence of a low or low to
rate histology negative for cancer
normal serum TSH concentration, a radioiodine scan should be
Fine-needle aspiration biopsy is currently the most sensitive and specific test to distinguish
performed and directly compared with the ultrasound images to 26
benign and malignant thyroid nodules.
determine the functionality of each nodule larger than 1–1.5cm. The
AACE/AME and the ETA extend the indication to radioisotope scan, the test increase significantly. Accordingly, as the use of thyroid
suggesting that thyroid scintigraphy should be performed for a ultrasound by endocrinologists is becoming more widespread, the
multinodular goitre in iodine-deficient areas even if the TSH level is AACE/AME guidelines suggest ultrasound FNA in the following
still in the normal range in order to identify the presence of an clinical settings: any size nodule with a history of radiation, family
autonomous nodule. Concordance between the guidelines is very history of mycobacterium tuberculosis complex (MTC) or family
high for the suggested actions even if the clinical evidence available history of multiple endocrine neoplasia type 2 (MEN2); any size
for using radioisotope scan is just fair (ATA: grade B; AACE/AME: nodule with suspicious ultrasound features; nodules with extra-
grade B and C). capsular growth or cervical nodes; and impalpable or small (<1cm)
nodule. The other two guidelines do not make specific
Literature linked to the issue of thyroid scintigraphy is scarce. The recommendations for ultrasound-FNA.
ATA guidelines quote no references, the ETA mentions one
reference (Pacini
37
) and the AACE/AME reports seven articles (one If the FNA result is benign, 59 and 27% would ask this patient to
review and six observational studies). In fact, the quality of EBM return in six to 12 months for thyroid palpation + ultrasound or
evidence about the use of radioisotope scans is quite low due to thyroid palpation + ultrasound + FNA, respectively. The AACE/AME
the absence of level 1 and 2 evidence, and at present guidelines suggest simple follow-up for cytologically benign thyroid
recommendations are based mostly on expert opinion and largely nodules; repeat ultrasound was not recommended. The ATA
accepted thyroid practice. guidelines suggest clinical follow-up at six to 18 months, without
ultrasound monitoring, for easily palpable benign nodules. Opinion
FNAB is currently the most sensitive and specific test to distinguish on re-aspiration of benign nodules remains divided. The AACE/AME
benign and malignant thyroid nodules (see Table 2). For the further suggests re-aspiration only for enlarging nodules, recurrent cysts or
work-up of the example patient mentioned above, 74% selected FNA nodules not shrinking after T4 therapy; the ATA guidelines suggest
with ultrasound guidance. This seems surprising, considering that the either re-aspiration or surgery for growing nodules. Wiersinga has
nodule was easily palpable. However, as the vast majority of recommended repeat palpation and FNA one year after a benign
European experts use ultrasound for the initial investigation of such FNA result.
39
Lucas et al. re-biopsied 116 patients with benign FNA
patients, it appears logical that they would also perform an and found no missed malignancy, concluding that re-aspiration is
ultrasound-guided FNA rather than a palpation-directed FNA. Several not necessary.
40
On the other hand, Chehade et al. followed 235
recent reports suggest that ultrasound–FNA is more reliable than patients with benign FNA for an average of 2.9 years, and on repeat
palpation–FNA.
26,38
With the use of ultrasound guidance, the FNA found malignancy in one (0.4%), concluding that re-biopsy
sensitivity, positive predictive value and negative predictive value of reduces false-negative rates.
41
EUROPEAN ENDOCRINOLOGY 61