Endocrine DOI 10.1007/s12020-014-0510-9
META-ANALYSIS
Reliability of real-time elastography to diagnose thyroid nodules previously read at FNAC as indeterminate: a meta-analysis Pierpaolo Trimboli • Giorgio Treglia • Ramin Sadeghi • Francesco Romanelli Luca Giovanella
•
Received: 30 October 2014 / Accepted: 10 December 2014 Ó Springer Science+Business Media New York 2014
Abstract The main limit of thyroid fine-needle aspiration cytology (FNAC) is represented by indeterminate report. Recently, real-time elastography (RTE) has been described in the management of these cases. Here, we performed a metaanalysis of published studies specifically focused on the use of RTE in indeterminate thyroid nodules. A comprehensive literature search of PubMed/MEDLINE and Google Scholar databases was conducted by using the combination of the terms ‘‘thyroid’’ and ‘‘indeterminate’’ and ‘‘elastography.’’ Pooled sensitivity, specificity, accuracy, PPV and NPV of RTE as predictor of malignancy in thyroid nodules with indeterminate FNAC were calculated, including 95 % confidence intervals (95 % CI). The area under the summary ROC curve (AUC) was also assessed. Databases found 572 papers, and eight were included in the meta-analysis. Of these, six studies had prospective design and two were retrospective. Pooled malignancy rate was 31 %. As common denominator, all studies set the prevalence of hardness within the nodule as risk factor for malignancy of the lesion. Sensitivity of RTE ranged from 11 to 89 % (pooled estimate of 69 %; 95 % CI
P. Trimboli (&) Section of Endocrinology and Diabetology, Ospedale Israelitico di Roma, Via Fulda, 14, 00148 Rome, Italy e-mail: [email protected] P. Trimboli G. Treglia L. Giovanella Department of Nuclear Medicine and Thyroid Centre, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland R. Sadeghi Nuclear Medicine Research Center, Mashhad University of Medical Sciences, Mashhad, Iran F. Romanelli Department of Experimental Medicine, Sapienza University, Rome, Italy
55–82 %), specificity varied from 6 to 100 % (pooled estimate of 75 %; 95 % CI 42–96 %), and accuracy was comprised between 35 and 94 % (pooled estimate of 73 %; 95 % CI 54–89 %). The AUC was 0.77. RTE has suboptimal diagnostic accuracy to diagnose thyroid nodules previously classified as indeterminate. Then, RTE alone should not be used for selecting these patients for surgery or not. We advice for further studies using other elastographic approaches and combined RTE and B-mode ultrasonography. Keywords Thyroid Ultrasonography Elastography Indeterminate Fine-needle aspiration cytology (FNAC)
Introduction Traditionally, ultrasonography (US) is the main tool in selecting thyroid nodules for fine-needle aspiration cytology (FNAC) [1, 2]. Several papers have reported the specific features by US and color flow Doppler (CFD) that are suggestive of malignancy (i.e., hypoechogenicity, blurred margins, microcalcifications, taller shape, and vascular signals) [3, 4]. Recently, in addition to conventional US, the real-time elastography (RTE) has improved the management of thyroid lesions, increasing the sensitivity of US plus CFD examination [5–7]. This evaluation consists of a real-time analysis of an area during the application of an external force by US probe. The probe is placed on the neck of the patient, a light pressure is applied, and the software tracks the tissues displacement displaying that by colors according to the tissue hardness. Several color score systems were proposed for thyroid lesions (i.e., by 3-, 4-, 5-, or 6-scale) and blue color is generally associated to hard tissue, while green and/or red reveal softness. Classically, thyroid cancer manifests as a hard and firm lesion. Thus, ‘‘blue’’ thyroid nodule at RTE should be
123
Endocrine
considered at risk for cancer. Consequently, regardless of the above different color scales adopted in the studies, thyroid lesions with a prevalent (i.e., more than 50 %) blue component were assessed as suspicious for malignancy [5–7]. The main limit of FNAC is represented by indeterminate report assessed as Thy 3, Class 3, or Category III–IV according to British Thyroid Association [8], American Association Clinical Endocrinologists/Associazione Medici Endocrinologi/European Thyroid Association [1], or Bethesda System for Reporting Thyroid Cytopathology [9], respectively. These lesions occur for about 10–15 % of thyroid cytologies and require diagnostic thyroidectomy. However, only one in four of these nodules is a cancer and to identify predictors of malignancy or benignancy is one of the most critical challenges in thyroidology [10–14]. In particular, clinical, instrumental, or molecular features associated with benignancy with high negative predictive value (NPV) should allow to avoid surgery for diagnosis [10–14]. The above-mentioned RTE has been investigated as potential predictor of malignancy in patients with thyroid nodule and/or thyroid indeterminate cytologic report, with controversial findings. Thus, the use of RTE in presurgical selection of indeterminate thyroid lesions (Thy 3, Class 3, Category III– IV) is still a matter to be clarified. Here, in order to provide more robust estimates on the use of RTE in diagnosing thyroid nodules with prior indeterminate FNAC report, we performed a meta-analysis of published studies on this topic.
thyroid malignancy in patients with prior indeterminate FNAC report. Also, studies investigating elastography performed by manual US probe compression and displayed by color scale (real-time elastography, RTE) were selected. Only nodules with final postoperative histologic follow-up were included in the meta-analysis. The exclusion criteria were as follows: (a) articles not within the field of interest of this review; (b) review articles, editorials or letters, comments, and conference proceedings; (c) case reports or case series; and (d) articles not in English language. Two authors independently reviewed titles and abstracts of the retrieved articles, applying the above-mentioned criteria. Articles were rejected if they were clearly ineligible. Then, the same two researchers independently reviewed the full-text version of the remaining articles to determine their eligibility for inclusion. Disagreements were resolved in a consensus meeting.
Methods
Quality assessment
Search strategy
The 2011 Oxford Center for Evidence-Based Medicine checklist for diagnostic studies was used for quality assessment of the included studies [Oxford Center for Evidence-Based Medicine checklist for diagnostic studies appraisal. Available at: http://www.cebm.net/index.aspx?o_1025 Accessed October 20th 2014.]. This checklist has five major parts as follows: representative spectrum of the patients, consecutive patient recruitment, ascertainment of the gold standard regardless of the index test results, independent blind comparison between the gold standard and index test results, and enough explanation of the test to permit replication.
This review was performed according to the ‘‘Preferred Reporting Items for Systematic Reviews and Meta-Analyses’’ (PRISMA) statement which describes an evidence-based minimum set of items that have to be reported in systematic reviews [15]. A comprehensive literature search of PubMed/ MEDLINE and Google Scholar databases was conducted to find relevant published articles on the use of RTE as predictor of malignancy in thyroid nodules with previous indeterminate FNAC report. The search was made on October 20, 2014. Initially, we used a search algorithm that was based on the combination of the terms ‘‘thyroid’’ and ‘‘indeterminate’’ and ‘‘elastography.’’ To expand our search, the references lists of the retrieved articles, and citation database of Google Scholar were also screened for additional studies. Study selection Studies or subsets in studies were eligible for inclusion when reported data about the use of elastography as predictor of
123
Data extraction For each included study, information was collected concerning basic study (authors, journals and year of publication, country of origin, and study design), patients characteristics (number, mean age, and gender), RTE scoring classification adopted, RTE classes recorded in cancers and benign lesions, and diagnostic accuracy data on RTE in predicting malignancy and benignancy. For each study, the number of true positive, false positive, true negative, and false negative findings for RTE was recorded on a per lesion-based analysis.
Statistical analysis Sensitivity, specificity, accuracy, and positive and negative predictive values (PPV and NPV) of RTE in predicting the malignancy of thyroid nodules with indeterminate FNAC were obtained from individual studies on a per lesion-based analysis. A random-effects model was used for statistical pooling of the data. Pooled data were presented with 95 % confidence intervals (95 % CI). An I-square (I2) index was
Endocrine Table 2 True and false results of RTE recorded in the eight studies First author
Cancers
Benign nodules
TP
TN
FP
FN
Rago [16]
31
111
24
110
1
7
Lippolis [17]
36
66
32
4
62
4
Unluturk [18]
9
9
6
7
2
3
Trimboli [19]
16
26
12
18
8
4
Ragazzoni [20]
7
17
5
11
6
2
Russ [21]
9
25
1
19
6
8
Mohey [22] Garino [23]
4 34
12 74
3 25
12 66
0 9
1 8
TP true positives, TN true negatives, FP false positives, FN false negatives
Fig. 1 Flow chart of the search for eligible studies on the diagnostic accuracy of RTE as predictor of malignancy in thyroid nodules with indeterminate FNAC
used to test for heterogeneity between studies (a significant heterogeneity was present if I2 [ 50 %). The area under the summary ROC curve (AUC) was calculated to measure the diagnostic accuracy of RTE. Statistical analyses were performed using StatsDirect statistical software (StatsDirect Ltd, Altrincham, UK) and Meta-DiSc statistical software version 1.4 (Unit of Clinical Biostatistics, Ramo´n y Cajal Hospital, Madrid, Spain).
Results Eligible articles Database of PubMed found 21 papers, while 551 articles were retrieved by Google Scholar. Of these 572, several Table 1 Main characteristics of the included studies
First author
Year
papers investigated RTE only for the initial assessment of thyroid nodules and were excluded. A number of nine articles were initially eligible for the review [16–24]; one paper was excluded because reported quantitative elastographic data [24]. Finally, the review comprised eight papers investigating RTE as predictor of malignancy in thyroid nodules with prior indeterminate (Thy 3, Class 3 or Category III–IV) FNAC report (Fig. 1). Qualitative analysis Regarding the geographical distribution of the eight articles, five were from Italy, one from Egypt, one from France, and another from Turkey. Six studies had a prospective design, while a retrospective setting was applied in the other two. All but one was single-center. In almost all cases, nodules had histologic verification after surgery. The pooled malignancy rate was 31 % ranging from 22 to 50 %. Table 1 details the main characteristics of the seven studies included in the meta-analysis. Of the included studies, three [16, 17, 23] investigated RTE in a series of thyroid lesions with previous indeterminate cytology, while in the other four papers the cases of indeterminate FNAC represented a subanalysis of the entire series of Country
Study setting
Elastographic score system
Nodules
Rago [16]
2010
Italy
Retrospective
3 classes
142
Lippolis [17]
2011
Italy
Prospective
4 classes
102
Unluturk [18]
2012
Turkey
Prospective
3 classes
18
Trimboli [19]
2012
Italy
Prospective
4 classes
42
Ragazzoni [20]
2012
Italy
Prospective
4 classes
24
Russ [21]
2013
France
Prospective
2 classes
34
Mohey [22]
2013
Egypt
Prospective
5 classes
16
Garino [23]
2014
Italy
Retrospective
4 classes
Tot
108 486
123
Endocrine Fig. 2 Plots of individual studies and pooled sensitivity, specificity, and accuracy of RTE in the differential diagnosis between malignant and benign thyroid nodules with indeterminate FNAC
123
Endocrine Fig. 3 Plots of individual studies and pooled positive and negative predictive value of RTE in the differential diagnosis between malignant and benign thyroid nodules with indeterminate FNAC
thyroid nodules. As accessory information, one study [16] reported RTE reliability also in a subgroup of nodules with prior inadequate (Thy 1) FNAC. Different RTE classification systems were adopted in the studies. Four studies classified nodules by four categories and the classes 3–4 as a whole were defined as at risk of malignancy [17, 19, 20, 23]; the study by Russ et al. [21] used a 2-classes score which is quite similar to the latter one in four classes; a 3-classes scoring was used by two studies [16, 18]; and a 5-classes system was report by the remaining one paper [22]. Common denominator
between all these studies was that the prevalence of hardness within the nodule was set as a risk factor for malignancy of the lesion, and the diagnostic accuracy of RTE was calculated as a consequence. Some studies investigated both RTE and US in the same series of indeterminate nodules [16–18, 21, 23]. The results showed that microcalcifications [17, 18], small nodule’s size [17], loss of halo sign [18], irregular margins [18, 23], and intranodular vascularization [22] were significantly associated with cancers; furthermore, TI-RADS gray-scale score was reported as highly sensitive and poorly specific in detecting malignancy [21].
123
Endocrine
Quantitative analysis (meta-analysis) Table 2 and Figs. 2, 3, and 4 show the diagnostic accuracy of RTE in the selected studies. Sensitivity of RTE ranged from 11 to 89 % with a pooled estimate of 69 % (95 % CI 55–82 %) (Fig. 2). The included studies were statistically heterogeneous in their estimate of sensitivity (I2: 68 %). Specificity varied from 6 to 100 % with a pooled estimate of 75 % (95 % CI 42–96 %) (Fig. 2). The included studies were statistically heterogeneous in their estimate of specificity (I2: 95 %). Accuracy was comprised from 35 and 94 % with a pooled estimate of 73 % (95 % CI 54–89 %) (Fig. 2). The included studies were statistically quite heterogeneous in their estimate of accuracy (I2: 94 %). Pooled PPV and NPV were 63 % (95 % CI 41–82 %) and 82 % (95 % CI 72–90 %), respectively (Fig. 3). The AUC was 0.77, which is considered a not optimal value for a diagnostic examination.
Discussion To preoperatively diagnose thyroid nodules cytologically classified as indeterminate (Thy 3, Class 3, or Category III– IV) represents one major challenge in thyroidology. In fact, a 75 % rate of benignancy is expected at surgery; thus, to exclude a cancer in these patients is a relevant scope which should allow to avoid a diagnostic surgery [25]. Many potential predictors of malignancy and benignancy have
Fig. 4 Summary ROC curve of RTE in the differential diagnosis between malignant and benign thyroid nodules with indeterminate FNAC. The curves represent the summary ROC curve (middle) and 95 % confidence intervals
123
been analyzed, such as nodule’s ultrasonographic features, clinical characteristics of the subject, molecular evaluation on FNAC, microhistologic examination by core needle biopsy, and molecular imaging (i.e., 99mTc-sestaMIBI scan, 18FDG-PET/CT) [10–14, 26–28]. However, surgery is still often necessary [1, 2, 8, 9]. Since its advent, thyroid RTE has promised to give accurate risk stratification in initial assessment of thyroid lesions. In particular, those nodules indentified as hard were frequently reported as cancers [5–7]. Also, as an extension of the first studies, RTE seems to increase the sensitivity of US and this combination should be used to carefully select nodules for FNAC [19]. In this context, a few papers investigated the reliability of RTE as predictor tool of malignancy/benignancy in patients with thyroid indeterminate cytology. Here we reviewed studies on this topic, and eight papers were meta-analyzed [16–23]. The main findings of the present study were that RTE has very low sensitivity, specificity, and accuracy when is used to diagnose thyroid lesions with prior indeterminate (Thy 3, Class 3, or Category III–IV) FNAC report. Of relevance, also PPV and NPV were low. Thus, RTE was demonstrated as poor predictor of malignancy/benignancy of these thyroid lesions. Some comments on this finding have to be addressed; interestingly, one paper [23] reported that both sensitivity and NPV increased by combining RTE and US. The latter should be in agreement with a previous report focused on thyroid nodule assessment [19]. Also, here we selected only papers using elastography by manual compression. However, it has to be cited the paper by Cantisani et al. [24] which investigated Q-elastography in 140 indeterminate nodules. There, 87 % sensitivity, 92 % specificity, 81 % PPV, and 94 % NPV might be obtained by using 2.05 cut-off of strain ratio. The combination of US and RTE features and other elastographic approaches have to be investigated in future studies. Thyroid elastographic examination has been diffused due to its easy use and brief time for the examination per patient (i.e., 5 min). Also, elastography is an application of US systems, which is worldwide the main thyroid imaging tool. However, a not negligible number of limitations of thyroid gland elastography (and of the published studies on this topic) has to be discussed. First, several nodule’s conditions represent an exclusion criteria in most studies; in particular, nodules with micro- and macro-calcifications and cystic nodules were often not included due to potential bias, but they have not been systematically examined by elastography [29, 30]. In addition, different elastographic measurement methods have been described, such as real time by manual compression, tissue-to-tissue strain ratio, calculation by quantitative assessment, shear wave, carotid pulsation, and others [30]. The first relevant meta-analysis of seven studies reported that elastography can be used
Endocrine Table 3 Reference standard and quality of the studies First author
Spectrum of patients
Consecutive or random selection of patients
Reference standard
Application of reference standard regardless of indexed test
Enough explanation of the index test to ensure reproducibility
Independent blind comparison between index test and reference standard
Level of evidence
Rago [16]
Patients with thyroid nodules and indeterminate or nondiagnostic cytology
N/A
Histology
Yes
Yes
Yes
3
Lippolis [17]
Patients being operated on for thyroid nodule with indeterminate cytology
Consecutive
Histology
Yes
Yes
Yes
2
¨ nlu¨tu¨rk U [18]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
Consecutive
Histology
Yes
Yes
Yes
2
Trimboli [19]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
Consecutive
Histology
Yes
Yes
Yes
2
Ragazzoni [20]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis) Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
Consecutive
Histology
Yes
Yes
Yes
2
Consecutive
Histology
Yes
Yes
Yes
2
Monhey [22]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
N/A
Histology
Yes
Yes
Yes
3
Garino [23]
Patients being operated on for thyroid nodule with indeterminate cytology
Consecutive
Histology
Yes
Yes
Yes
2
Russ [21]
with high sensitivity (92 %) and specificity (90 %) in the initial work-up of thyroid nodules [31]. More recently, Veer and Puttagunta [32] meta-analyzed thirty-eight studies and recorded lower sensitivity (87 %) and poorer specificity (80.6 %); in the latter study, the results obtained in the papers using elastography for the assessment of nodules showed no difference from that focusing on indeterminate lesions. Unfortunately, data on indeterminate nodules were not specified. Remarkably, both meta-analyses included studies with different elastographic systems and color scales. Overall, low evidence on thyroid elastography has been achieved to date. Several papers investigated US features in diagnosing indeterminate thyroid nodules [11]. Even if some B-mode US nodule’s risk factors (i.e., irregular margins, microcalcifications, and small size) were identified, no
significant evidence has been obtained and US is not recommended in the decision making of indeterminate thyroid lesions [1, 2, 8, 9]. To the best of our knowledge, no metaanalyses or systematic reviews on RTE on this topic have been published to date. For the latter and the above reasons, present data achieve interest for clinical practice and advice for further studies. Some potential limitations of the present study should be briefly discussed. Heterogeneity between studies may represent a source of bias in a meta-analysis. In herein pooled results, the studies were statistically heterogeneous regarding sensitivity, specificity, and accuracy. The latter is likely to arise through diversity in methodological aspects between different studies. The baseline differences among patients in the included studies, the reference standard used, and the study quality (Table 3) may have contributed
123
Endocrine
to the observed results; however, heterogeneity between studies was accounted for in a random-effects model. Publication bias is a major concern in all meta-analyses as studies reporting significant findings are more likely to be accepted for publication than those reporting non-significant (i.e., negative and non-confirmatory) results. Indeed, it is not unusual for small-sized early studies to report a positive relationship that subsequent larger studies fail to replicate. A possible publication bias could not be excluded in our meta-analysis. Discrepancy was present between the studies regarding some data. The malignancy rate varied from 21 to 50 %, being pooled cancer rate (31 %) higher than that expected in indeterminate thyroid nodules [1, 2, 8–10], and the percentage of positive RTE exams was heterogeneous (from 17 to 92 %). All these findings should be due to the surgical series described in the eight studies. Specifically, the reported series could not include all patients with indeterminate FNAC recorded at each one Institute. Also, the selection of patients was not specified in two of these papers [16, 22] (view Table 3). To identify those patients with indeterminate thyroid nodules (Thy 3, Class 3, or Category III–IV) who could avoid diagnostic surgery because of certain preoperative assessment of benignancy is still a major challenge. Present data from eight papers showed that RTE, performed by manual compression and displayed by color scale, has suboptimal diagnostic accuracy in discriminating cancers from benign thyroid nodules which were previously cytologically classified as indeterminate. Thus, RTE alone should not be used for selecting patients with these nodules for surgery or not. More advanced elastographic approaches (i.e., quantitative, shear wave, etc.), as well as the combined use of RTE and B-mode US, have shown promise in improving thyroid nodule’s risk assessment and deserve further studies in indeterminate lesions.
3.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13. Conflict of interest declare.
The authors have not conflicts of interest to 14.
References 1. H. Gharib, E. Papini, R. Paschke, D.S. Duick, R. Valcavi, L. Hegedu¨s, P. Vitti, AACE/AME/ETA Task Force on Thyroid Nodules, American Association of Clinical Endocrinologists, Associazione Medici Endocrinologi, and European Thyroid Association Medical guidelines for clinical practice for the diagnosis and management of thyroid nodules: executive summary of recommendations. Endocr. Pract. 16, 468–475 (2010) 2. D.S. Cooper, G.M. Doherty, B.R. Haugen, R.T. Kloos, S.L. Lee, S.J. Mandel, E.L. Mazzaferri, B. McIver, F. Pacini, M. Schlumberger, S.I. Sherman, D.L. Steward, R.M. Tuttle, Revised American Thyroid Association management guidelines for
123
15.
16.
patients with thyroid nodules and differentiated thyroid cancer. Thyroid 19, 1167–1214 (2009) L. Solbiati, V. Osti, L. Cova, M. Tonolini, Ultrasound of thyroid, parathyroid glands and neck lymph nodes. Eur. Radiol. 11, 2411–2424 (2001) E.K. Kim, C.S. Park, W.Y. Chung, K.K. Oh, D.I. Kim, J.T. Lee, H.S. Yoo, New sonographic criteria for recommending fineneedle aspiration biopsy of nonpalpable solid nodules of the thyroid. Am. J. Roentgenol. 178, 687–691 (2002) A. Lyshchik, T. Higashi, R. Asato, S. Tanaka, J. Ito, J.J. Mai, C. Pellot-Barakat, M.F. Insana, A.B. Brill, T. Saga, M. Hiraoka, K. Togashi, Thyroid gland tumor diagnosis at US elastography. Radiology 237, 202–211 (2005) T. Rago, F. Santini, M. Scutari, A. Pinchera, P. Vitti, Elastography: new developments in ultrasound for predicting malignancy in thyroid nodules. J. Clin. Endocrinol. Metab. 92, 2917–2922 (2007) M. Friedrich-Rust, A. Sperber, K. Holzer, J. Diener, F. Gru¨nwald, K. Badenhoop, S. Weber, S. Kriener, E. Herrmann, W.O. Bechstein, S. Zeuzem, J. Bojunga, Real-time elastography and contrast-enhanced ultrasound for the assessment of thyroid nodules. Exp. Clin. Endocrinol. Diabetes 118, 602–609 (2010) Royal College of Physicians of London, Guidelines for the Management of Thyroid Cancer, 2nd edn. (The Lavenham Press, Suffolk 2007) E.S. Cibas, S.Z. Ali, The bethesda system for reporting thyroid cytopathology. Thyroid 19, 1159–1165 (2009) P. Trimboli, G. Treglia, L. Guidobaldi, E. Saggiorato, G. Nigri, A. Crescenzi, F. Romanelli, F. Orlandi, S. Valabrega, R. Sadeghi, L. Giovanella, Clinical characteristics as predictors of malignancy in patients with indeterminate thyroid cytology: a metaanalysis. Endocrine 46, 52–59 (2014) P. Trimboli, E. Condorelli, A. Catania, S. Sorrenti, Clinical and ultrasound parameters in the approach to thyroid nodules cytologically classified as indeterminate neoplasm. Diagn. Cytopathol. 37, 783–785 (2009) A. Bartolazzi, F. Orlandi, E. Saggiorato, M. Volante, F. Arecco, R. Rossetto, N. Palestini, E. Ghigo, M. Papotti, G. Bussolati, M.P. Martegani, F. Pantellini, A. Carpi, M.R. Giovagnoli, S. Monti, V. Toscano, S. Sciacchitano, G.M. Pennelli, C. Mian, M.R. Pelizzo, M. Rugge, G. Troncone, L. Palombini, G. Chiappetta, G. Botti, A. Vecchione, R. Bellocco, Italian Thyroid Cancer Study Group (ITCSG), Galectin-3-expression analysis in the surgical selection of follicular thyroid nodules with indeterminate fine-needle aspiration cytology: a prospective multicentre study. Lancet Oncol. 9, 543–549 (2008) Z.W. Baloch, S. Fleisher, V.A. LiVolsi, P.K. Gupta, Diagnosis of ‘‘follicular neoplasm’’: a gray zone in thyroid fine-needle aspiration cytology. Diagn. Cytopathol. 26, 41–44 (2002) E. Saggiorato, T. Angusti, R. Rosas, M. Martinese, M. Finessi, F. Arecco, E. Trevisiol, N. Bergero, B. Puligheddu, M. Volante, V. Podio, M. Papotti, F. Orlandi, 99 mTc-MIBI Imaging in the presurgical characterization of thyroid follicular neoplasms: relationship to multidrug resistance protein expression. J. Nucl. Med. 50, 1785–1793 (2009) A. Liberati, D.G. Altman, J. Tetzlaff, C. Mulrow, P.C. Gøtzsche, J.P. Ioannidis, M. Clarke, P.J. Devereaux, J. Kleijnen, D. Moher, The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J. Clin. Epidemiol. 62, e1–e34 (2009) T. Rago, M. Scutari, F. Santini, V. Loiacono, P. Piaggi, G. Di Coscio, F. Basolo, P. Berti, A. Pinchera, P. Vitti, Real-time elastosonography: useful tool for refining the presurgical diagnosis in thyroid nodules with indeterminate or nondiagnostic cytology. J. Clin. Endocrinol. Metab. 95, 5274–5280 (2010)
Endocrine 17. P.V. Lippolis, S. Tognini, G. Materazzi, A. Polini, R. Mancini, C.E. Ambrosini, A. Dardano, F. Basolo, M. Seccia, P. Miccoli, F. Monzani, Is elastography actually useful in the presurgical selection of thyroid nodules with indeterminate cytology? J. Clin. Endocrinol. Metab. 96, e1826–e1830 (2011) 18. U. Unlu¨tu¨rk, M.F. Erdog˘an, O. Demir, S. Gu¨llu¨, N. Bas¸ kal, Ultrasound elastography is not superior to grayscale ultrasound in predicting malignancy in thyroid nodules. Thyroid 22, 1031–1038 (2012) 19. P. Trimboli, R. Guglielmi, S. Monti, I. Misischi, F. Graziano, N. Nasrollah, S. Amendola, S.N. Morgante, M.G. Deiana, S. Valabrega, V. Toscano, E. Papini, Ultrasound sensitivity for thyroid malignancy is increased by real-time elastography: a prospective multicenter study. J. Clin. Endocrinol. Metab. 97, 4524–4530 (2012) 20. F. Ragazzoni, M. Deandrea, A. Mormile, M.J. Ramunni, F. Garino, G. Magliona, M. Motta, B. Torchio, R. Garberoglio, P.P. Limone, High diagnostic accuracy and interobserver reliability of real-time elastography in the evaluation of thyroid nodules. Ultrasound Med. Biol. 38, 1154–1162 (2012) 21. G. Russ, B. Royer, C. Bigorgne, A. Rouxel, M. Bienvenu-Perrard, L. Leenhardt, Prospective evaluation of thyroid imaging reporting and data system on 4,550 nodules with and without elastography. Eur. J. Endocrinol. 168, 649–655 (2013) 22. N. Monhey, T.A. Hassana, S. Abdel-Bakib, Role of combined grey scale US and US tissue elastography in differentiating solid thyroid nodules. Egypt. J. Radiol. Nuclear. Med. 44, 505–512 (2013) 23. F. Garino, M. Deandrea, M. Motta, A. Mormile, F. Ragazzoni, N. Palestini, M. Freddi, G. Gasparri, E. Sgotto, D. Pacchioni, P.P. Limone, Diagnostic performance of elastography in cytologically indeterminate thyroid nodules. Endocrine (2014). doi:10.1007/ s12020-014-0438-0
24. V. Cantisani, S. Ulisse, E. Guaitoli, C. De Vito, R. Caruso, R. Mocini, V. D’Andrea, V. Ascoli, A. Antonaci, C. Catalano, F. Nardi, A. Redler, P. Ricci, E. De Antoni, S. Sorrenti, Q-elastography in the presurgical diagnosis of thyroid nodules with indeterminate cytology. PLoS One 7, e50725 (2012) 25. T.J. Sebo, What are the keys to successful thyroid FNA interpretation? Clin. Endocrinol. 77, 13–17 (2012) 26. P. Trimboli, A. Crescenzi, Thyroid core needle biopsy: taking stoke of the situation. Endocrine (2014). doi:10.1007/s12020014-0382-z 27. L. Giovanella, L. Ceriani, G. Treglia, Role of isotope scan, including positron emission tomography/computed tomography, in nodular goitre. Best. Pract. Res. Clin. Endocrinol. Metab. 28, 507–518 (2014) 28. G. Treglia, C. Caldarella, E. Saggiorato, L. Ceriani, F. Orlandi, M. Salvatori, L. Giovanella, Diagnostic performance of (99m)TcMIBI scan in predicting the malignancy of thyroid nodules: a meta-analysis. Endocrine 44, 70–78 (2013) 29. E. Szczepanek-Parulska, K. Wolin´ski, A. Stangierski, E. Gurgul, M. Ruchała, Biochemical and ultrasonographic parameters influencing thyroid nodules elasticity. Endocrine 47, 519–527 (2014) 30. M. Andrioli, L. Persani, Elastographic techniques of thyroid gland: current status. Endocrine 46, 455–461 (2014) 31. J. Bojunga, E. Herrmann, G. Meyer, S. Weber, S. Zeuzem, M. Friedrich-Rust, Real-time elastography for the differentiation of benign and malignant thyroid nodules: a meta-analysis. Thyroid 20, 1145–1150 (2010) 32. V. Veer, S. Puttagunta, The role of elastography in evaluating thyroid nodules: a literature review and meta-analysis. Eur. Arch. Otorhinolaryngol. (2014). doi:10.1007/s00405-014-3155-7
123
META-ANALYSIS
Reliability of real-time elastography to diagnose thyroid nodules previously read at FNAC as indeterminate: a meta-analysis Pierpaolo Trimboli • Giorgio Treglia • Ramin Sadeghi • Francesco Romanelli Luca Giovanella
•
Received: 30 October 2014 / Accepted: 10 December 2014 Ó Springer Science+Business Media New York 2014
Abstract The main limit of thyroid fine-needle aspiration cytology (FNAC) is represented by indeterminate report. Recently, real-time elastography (RTE) has been described in the management of these cases. Here, we performed a metaanalysis of published studies specifically focused on the use of RTE in indeterminate thyroid nodules. A comprehensive literature search of PubMed/MEDLINE and Google Scholar databases was conducted by using the combination of the terms ‘‘thyroid’’ and ‘‘indeterminate’’ and ‘‘elastography.’’ Pooled sensitivity, specificity, accuracy, PPV and NPV of RTE as predictor of malignancy in thyroid nodules with indeterminate FNAC were calculated, including 95 % confidence intervals (95 % CI). The area under the summary ROC curve (AUC) was also assessed. Databases found 572 papers, and eight were included in the meta-analysis. Of these, six studies had prospective design and two were retrospective. Pooled malignancy rate was 31 %. As common denominator, all studies set the prevalence of hardness within the nodule as risk factor for malignancy of the lesion. Sensitivity of RTE ranged from 11 to 89 % (pooled estimate of 69 %; 95 % CI
P. Trimboli (&) Section of Endocrinology and Diabetology, Ospedale Israelitico di Roma, Via Fulda, 14, 00148 Rome, Italy e-mail: [email protected] P. Trimboli G. Treglia L. Giovanella Department of Nuclear Medicine and Thyroid Centre, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland R. Sadeghi Nuclear Medicine Research Center, Mashhad University of Medical Sciences, Mashhad, Iran F. Romanelli Department of Experimental Medicine, Sapienza University, Rome, Italy
55–82 %), specificity varied from 6 to 100 % (pooled estimate of 75 %; 95 % CI 42–96 %), and accuracy was comprised between 35 and 94 % (pooled estimate of 73 %; 95 % CI 54–89 %). The AUC was 0.77. RTE has suboptimal diagnostic accuracy to diagnose thyroid nodules previously classified as indeterminate. Then, RTE alone should not be used for selecting these patients for surgery or not. We advice for further studies using other elastographic approaches and combined RTE and B-mode ultrasonography. Keywords Thyroid Ultrasonography Elastography Indeterminate Fine-needle aspiration cytology (FNAC)
Introduction Traditionally, ultrasonography (US) is the main tool in selecting thyroid nodules for fine-needle aspiration cytology (FNAC) [1, 2]. Several papers have reported the specific features by US and color flow Doppler (CFD) that are suggestive of malignancy (i.e., hypoechogenicity, blurred margins, microcalcifications, taller shape, and vascular signals) [3, 4]. Recently, in addition to conventional US, the real-time elastography (RTE) has improved the management of thyroid lesions, increasing the sensitivity of US plus CFD examination [5–7]. This evaluation consists of a real-time analysis of an area during the application of an external force by US probe. The probe is placed on the neck of the patient, a light pressure is applied, and the software tracks the tissues displacement displaying that by colors according to the tissue hardness. Several color score systems were proposed for thyroid lesions (i.e., by 3-, 4-, 5-, or 6-scale) and blue color is generally associated to hard tissue, while green and/or red reveal softness. Classically, thyroid cancer manifests as a hard and firm lesion. Thus, ‘‘blue’’ thyroid nodule at RTE should be
123
Endocrine
considered at risk for cancer. Consequently, regardless of the above different color scales adopted in the studies, thyroid lesions with a prevalent (i.e., more than 50 %) blue component were assessed as suspicious for malignancy [5–7]. The main limit of FNAC is represented by indeterminate report assessed as Thy 3, Class 3, or Category III–IV according to British Thyroid Association [8], American Association Clinical Endocrinologists/Associazione Medici Endocrinologi/European Thyroid Association [1], or Bethesda System for Reporting Thyroid Cytopathology [9], respectively. These lesions occur for about 10–15 % of thyroid cytologies and require diagnostic thyroidectomy. However, only one in four of these nodules is a cancer and to identify predictors of malignancy or benignancy is one of the most critical challenges in thyroidology [10–14]. In particular, clinical, instrumental, or molecular features associated with benignancy with high negative predictive value (NPV) should allow to avoid surgery for diagnosis [10–14]. The above-mentioned RTE has been investigated as potential predictor of malignancy in patients with thyroid nodule and/or thyroid indeterminate cytologic report, with controversial findings. Thus, the use of RTE in presurgical selection of indeterminate thyroid lesions (Thy 3, Class 3, Category III– IV) is still a matter to be clarified. Here, in order to provide more robust estimates on the use of RTE in diagnosing thyroid nodules with prior indeterminate FNAC report, we performed a meta-analysis of published studies on this topic.
thyroid malignancy in patients with prior indeterminate FNAC report. Also, studies investigating elastography performed by manual US probe compression and displayed by color scale (real-time elastography, RTE) were selected. Only nodules with final postoperative histologic follow-up were included in the meta-analysis. The exclusion criteria were as follows: (a) articles not within the field of interest of this review; (b) review articles, editorials or letters, comments, and conference proceedings; (c) case reports or case series; and (d) articles not in English language. Two authors independently reviewed titles and abstracts of the retrieved articles, applying the above-mentioned criteria. Articles were rejected if they were clearly ineligible. Then, the same two researchers independently reviewed the full-text version of the remaining articles to determine their eligibility for inclusion. Disagreements were resolved in a consensus meeting.
Methods
Quality assessment
Search strategy
The 2011 Oxford Center for Evidence-Based Medicine checklist for diagnostic studies was used for quality assessment of the included studies [Oxford Center for Evidence-Based Medicine checklist for diagnostic studies appraisal. Available at: http://www.cebm.net/index.aspx?o_1025 Accessed October 20th 2014.]. This checklist has five major parts as follows: representative spectrum of the patients, consecutive patient recruitment, ascertainment of the gold standard regardless of the index test results, independent blind comparison between the gold standard and index test results, and enough explanation of the test to permit replication.
This review was performed according to the ‘‘Preferred Reporting Items for Systematic Reviews and Meta-Analyses’’ (PRISMA) statement which describes an evidence-based minimum set of items that have to be reported in systematic reviews [15]. A comprehensive literature search of PubMed/ MEDLINE and Google Scholar databases was conducted to find relevant published articles on the use of RTE as predictor of malignancy in thyroid nodules with previous indeterminate FNAC report. The search was made on October 20, 2014. Initially, we used a search algorithm that was based on the combination of the terms ‘‘thyroid’’ and ‘‘indeterminate’’ and ‘‘elastography.’’ To expand our search, the references lists of the retrieved articles, and citation database of Google Scholar were also screened for additional studies. Study selection Studies or subsets in studies were eligible for inclusion when reported data about the use of elastography as predictor of
123
Data extraction For each included study, information was collected concerning basic study (authors, journals and year of publication, country of origin, and study design), patients characteristics (number, mean age, and gender), RTE scoring classification adopted, RTE classes recorded in cancers and benign lesions, and diagnostic accuracy data on RTE in predicting malignancy and benignancy. For each study, the number of true positive, false positive, true negative, and false negative findings for RTE was recorded on a per lesion-based analysis.
Statistical analysis Sensitivity, specificity, accuracy, and positive and negative predictive values (PPV and NPV) of RTE in predicting the malignancy of thyroid nodules with indeterminate FNAC were obtained from individual studies on a per lesion-based analysis. A random-effects model was used for statistical pooling of the data. Pooled data were presented with 95 % confidence intervals (95 % CI). An I-square (I2) index was
Endocrine Table 2 True and false results of RTE recorded in the eight studies First author
Cancers
Benign nodules
TP
TN
FP
FN
Rago [16]
31
111
24
110
1
7
Lippolis [17]
36
66
32
4
62
4
Unluturk [18]
9
9
6
7
2
3
Trimboli [19]
16
26
12
18
8
4
Ragazzoni [20]
7
17
5
11
6
2
Russ [21]
9
25
1
19
6
8
Mohey [22] Garino [23]
4 34
12 74
3 25
12 66
0 9
1 8
TP true positives, TN true negatives, FP false positives, FN false negatives
Fig. 1 Flow chart of the search for eligible studies on the diagnostic accuracy of RTE as predictor of malignancy in thyroid nodules with indeterminate FNAC
used to test for heterogeneity between studies (a significant heterogeneity was present if I2 [ 50 %). The area under the summary ROC curve (AUC) was calculated to measure the diagnostic accuracy of RTE. Statistical analyses were performed using StatsDirect statistical software (StatsDirect Ltd, Altrincham, UK) and Meta-DiSc statistical software version 1.4 (Unit of Clinical Biostatistics, Ramo´n y Cajal Hospital, Madrid, Spain).
Results Eligible articles Database of PubMed found 21 papers, while 551 articles were retrieved by Google Scholar. Of these 572, several Table 1 Main characteristics of the included studies
First author
Year
papers investigated RTE only for the initial assessment of thyroid nodules and were excluded. A number of nine articles were initially eligible for the review [16–24]; one paper was excluded because reported quantitative elastographic data [24]. Finally, the review comprised eight papers investigating RTE as predictor of malignancy in thyroid nodules with prior indeterminate (Thy 3, Class 3 or Category III–IV) FNAC report (Fig. 1). Qualitative analysis Regarding the geographical distribution of the eight articles, five were from Italy, one from Egypt, one from France, and another from Turkey. Six studies had a prospective design, while a retrospective setting was applied in the other two. All but one was single-center. In almost all cases, nodules had histologic verification after surgery. The pooled malignancy rate was 31 % ranging from 22 to 50 %. Table 1 details the main characteristics of the seven studies included in the meta-analysis. Of the included studies, three [16, 17, 23] investigated RTE in a series of thyroid lesions with previous indeterminate cytology, while in the other four papers the cases of indeterminate FNAC represented a subanalysis of the entire series of Country
Study setting
Elastographic score system
Nodules
Rago [16]
2010
Italy
Retrospective
3 classes
142
Lippolis [17]
2011
Italy
Prospective
4 classes
102
Unluturk [18]
2012
Turkey
Prospective
3 classes
18
Trimboli [19]
2012
Italy
Prospective
4 classes
42
Ragazzoni [20]
2012
Italy
Prospective
4 classes
24
Russ [21]
2013
France
Prospective
2 classes
34
Mohey [22]
2013
Egypt
Prospective
5 classes
16
Garino [23]
2014
Italy
Retrospective
4 classes
Tot
108 486
123
Endocrine Fig. 2 Plots of individual studies and pooled sensitivity, specificity, and accuracy of RTE in the differential diagnosis between malignant and benign thyroid nodules with indeterminate FNAC
123
Endocrine Fig. 3 Plots of individual studies and pooled positive and negative predictive value of RTE in the differential diagnosis between malignant and benign thyroid nodules with indeterminate FNAC
thyroid nodules. As accessory information, one study [16] reported RTE reliability also in a subgroup of nodules with prior inadequate (Thy 1) FNAC. Different RTE classification systems were adopted in the studies. Four studies classified nodules by four categories and the classes 3–4 as a whole were defined as at risk of malignancy [17, 19, 20, 23]; the study by Russ et al. [21] used a 2-classes score which is quite similar to the latter one in four classes; a 3-classes scoring was used by two studies [16, 18]; and a 5-classes system was report by the remaining one paper [22]. Common denominator
between all these studies was that the prevalence of hardness within the nodule was set as a risk factor for malignancy of the lesion, and the diagnostic accuracy of RTE was calculated as a consequence. Some studies investigated both RTE and US in the same series of indeterminate nodules [16–18, 21, 23]. The results showed that microcalcifications [17, 18], small nodule’s size [17], loss of halo sign [18], irregular margins [18, 23], and intranodular vascularization [22] were significantly associated with cancers; furthermore, TI-RADS gray-scale score was reported as highly sensitive and poorly specific in detecting malignancy [21].
123
Endocrine
Quantitative analysis (meta-analysis) Table 2 and Figs. 2, 3, and 4 show the diagnostic accuracy of RTE in the selected studies. Sensitivity of RTE ranged from 11 to 89 % with a pooled estimate of 69 % (95 % CI 55–82 %) (Fig. 2). The included studies were statistically heterogeneous in their estimate of sensitivity (I2: 68 %). Specificity varied from 6 to 100 % with a pooled estimate of 75 % (95 % CI 42–96 %) (Fig. 2). The included studies were statistically heterogeneous in their estimate of specificity (I2: 95 %). Accuracy was comprised from 35 and 94 % with a pooled estimate of 73 % (95 % CI 54–89 %) (Fig. 2). The included studies were statistically quite heterogeneous in their estimate of accuracy (I2: 94 %). Pooled PPV and NPV were 63 % (95 % CI 41–82 %) and 82 % (95 % CI 72–90 %), respectively (Fig. 3). The AUC was 0.77, which is considered a not optimal value for a diagnostic examination.
Discussion To preoperatively diagnose thyroid nodules cytologically classified as indeterminate (Thy 3, Class 3, or Category III– IV) represents one major challenge in thyroidology. In fact, a 75 % rate of benignancy is expected at surgery; thus, to exclude a cancer in these patients is a relevant scope which should allow to avoid a diagnostic surgery [25]. Many potential predictors of malignancy and benignancy have
Fig. 4 Summary ROC curve of RTE in the differential diagnosis between malignant and benign thyroid nodules with indeterminate FNAC. The curves represent the summary ROC curve (middle) and 95 % confidence intervals
123
been analyzed, such as nodule’s ultrasonographic features, clinical characteristics of the subject, molecular evaluation on FNAC, microhistologic examination by core needle biopsy, and molecular imaging (i.e., 99mTc-sestaMIBI scan, 18FDG-PET/CT) [10–14, 26–28]. However, surgery is still often necessary [1, 2, 8, 9]. Since its advent, thyroid RTE has promised to give accurate risk stratification in initial assessment of thyroid lesions. In particular, those nodules indentified as hard were frequently reported as cancers [5–7]. Also, as an extension of the first studies, RTE seems to increase the sensitivity of US and this combination should be used to carefully select nodules for FNAC [19]. In this context, a few papers investigated the reliability of RTE as predictor tool of malignancy/benignancy in patients with thyroid indeterminate cytology. Here we reviewed studies on this topic, and eight papers were meta-analyzed [16–23]. The main findings of the present study were that RTE has very low sensitivity, specificity, and accuracy when is used to diagnose thyroid lesions with prior indeterminate (Thy 3, Class 3, or Category III–IV) FNAC report. Of relevance, also PPV and NPV were low. Thus, RTE was demonstrated as poor predictor of malignancy/benignancy of these thyroid lesions. Some comments on this finding have to be addressed; interestingly, one paper [23] reported that both sensitivity and NPV increased by combining RTE and US. The latter should be in agreement with a previous report focused on thyroid nodule assessment [19]. Also, here we selected only papers using elastography by manual compression. However, it has to be cited the paper by Cantisani et al. [24] which investigated Q-elastography in 140 indeterminate nodules. There, 87 % sensitivity, 92 % specificity, 81 % PPV, and 94 % NPV might be obtained by using 2.05 cut-off of strain ratio. The combination of US and RTE features and other elastographic approaches have to be investigated in future studies. Thyroid elastographic examination has been diffused due to its easy use and brief time for the examination per patient (i.e., 5 min). Also, elastography is an application of US systems, which is worldwide the main thyroid imaging tool. However, a not negligible number of limitations of thyroid gland elastography (and of the published studies on this topic) has to be discussed. First, several nodule’s conditions represent an exclusion criteria in most studies; in particular, nodules with micro- and macro-calcifications and cystic nodules were often not included due to potential bias, but they have not been systematically examined by elastography [29, 30]. In addition, different elastographic measurement methods have been described, such as real time by manual compression, tissue-to-tissue strain ratio, calculation by quantitative assessment, shear wave, carotid pulsation, and others [30]. The first relevant meta-analysis of seven studies reported that elastography can be used
Endocrine Table 3 Reference standard and quality of the studies First author
Spectrum of patients
Consecutive or random selection of patients
Reference standard
Application of reference standard regardless of indexed test
Enough explanation of the index test to ensure reproducibility
Independent blind comparison between index test and reference standard
Level of evidence
Rago [16]
Patients with thyroid nodules and indeterminate or nondiagnostic cytology
N/A
Histology
Yes
Yes
Yes
3
Lippolis [17]
Patients being operated on for thyroid nodule with indeterminate cytology
Consecutive
Histology
Yes
Yes
Yes
2
¨ nlu¨tu¨rk U [18]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
Consecutive
Histology
Yes
Yes
Yes
2
Trimboli [19]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
Consecutive
Histology
Yes
Yes
Yes
2
Ragazzoni [20]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis) Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
Consecutive
Histology
Yes
Yes
Yes
2
Consecutive
Histology
Yes
Yes
Yes
2
Monhey [22]
Patients with thyroid nodules (only patient with indeterminate cytology were included in the metaanalysis)
N/A
Histology
Yes
Yes
Yes
3
Garino [23]
Patients being operated on for thyroid nodule with indeterminate cytology
Consecutive
Histology
Yes
Yes
Yes
2
Russ [21]
with high sensitivity (92 %) and specificity (90 %) in the initial work-up of thyroid nodules [31]. More recently, Veer and Puttagunta [32] meta-analyzed thirty-eight studies and recorded lower sensitivity (87 %) and poorer specificity (80.6 %); in the latter study, the results obtained in the papers using elastography for the assessment of nodules showed no difference from that focusing on indeterminate lesions. Unfortunately, data on indeterminate nodules were not specified. Remarkably, both meta-analyses included studies with different elastographic systems and color scales. Overall, low evidence on thyroid elastography has been achieved to date. Several papers investigated US features in diagnosing indeterminate thyroid nodules [11]. Even if some B-mode US nodule’s risk factors (i.e., irregular margins, microcalcifications, and small size) were identified, no
significant evidence has been obtained and US is not recommended in the decision making of indeterminate thyroid lesions [1, 2, 8, 9]. To the best of our knowledge, no metaanalyses or systematic reviews on RTE on this topic have been published to date. For the latter and the above reasons, present data achieve interest for clinical practice and advice for further studies. Some potential limitations of the present study should be briefly discussed. Heterogeneity between studies may represent a source of bias in a meta-analysis. In herein pooled results, the studies were statistically heterogeneous regarding sensitivity, specificity, and accuracy. The latter is likely to arise through diversity in methodological aspects between different studies. The baseline differences among patients in the included studies, the reference standard used, and the study quality (Table 3) may have contributed
123
Endocrine
to the observed results; however, heterogeneity between studies was accounted for in a random-effects model. Publication bias is a major concern in all meta-analyses as studies reporting significant findings are more likely to be accepted for publication than those reporting non-significant (i.e., negative and non-confirmatory) results. Indeed, it is not unusual for small-sized early studies to report a positive relationship that subsequent larger studies fail to replicate. A possible publication bias could not be excluded in our meta-analysis. Discrepancy was present between the studies regarding some data. The malignancy rate varied from 21 to 50 %, being pooled cancer rate (31 %) higher than that expected in indeterminate thyroid nodules [1, 2, 8–10], and the percentage of positive RTE exams was heterogeneous (from 17 to 92 %). All these findings should be due to the surgical series described in the eight studies. Specifically, the reported series could not include all patients with indeterminate FNAC recorded at each one Institute. Also, the selection of patients was not specified in two of these papers [16, 22] (view Table 3). To identify those patients with indeterminate thyroid nodules (Thy 3, Class 3, or Category III–IV) who could avoid diagnostic surgery because of certain preoperative assessment of benignancy is still a major challenge. Present data from eight papers showed that RTE, performed by manual compression and displayed by color scale, has suboptimal diagnostic accuracy in discriminating cancers from benign thyroid nodules which were previously cytologically classified as indeterminate. Thus, RTE alone should not be used for selecting patients with these nodules for surgery or not. More advanced elastographic approaches (i.e., quantitative, shear wave, etc.), as well as the combined use of RTE and B-mode US, have shown promise in improving thyroid nodule’s risk assessment and deserve further studies in indeterminate lesions.
3.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13. Conflict of interest declare.
The authors have not conflicts of interest to 14.
References 1. H. Gharib, E. Papini, R. Paschke, D.S. Duick, R. Valcavi, L. Hegedu¨s, P. Vitti, AACE/AME/ETA Task Force on Thyroid Nodules, American Association of Clinical Endocrinologists, Associazione Medici Endocrinologi, and European Thyroid Association Medical guidelines for clinical practice for the diagnosis and management of thyroid nodules: executive summary of recommendations. Endocr. Pract. 16, 468–475 (2010) 2. D.S. Cooper, G.M. Doherty, B.R. Haugen, R.T. Kloos, S.L. Lee, S.J. Mandel, E.L. Mazzaferri, B. McIver, F. Pacini, M. Schlumberger, S.I. Sherman, D.L. Steward, R.M. Tuttle, Revised American Thyroid Association management guidelines for
123
15.
16.
patients with thyroid nodules and differentiated thyroid cancer. Thyroid 19, 1167–1214 (2009) L. Solbiati, V. Osti, L. Cova, M. Tonolini, Ultrasound of thyroid, parathyroid glands and neck lymph nodes. Eur. Radiol. 11, 2411–2424 (2001) E.K. Kim, C.S. Park, W.Y. Chung, K.K. Oh, D.I. Kim, J.T. Lee, H.S. Yoo, New sonographic criteria for recommending fineneedle aspiration biopsy of nonpalpable solid nodules of the thyroid. Am. J. Roentgenol. 178, 687–691 (2002) A. Lyshchik, T. Higashi, R. Asato, S. Tanaka, J. Ito, J.J. Mai, C. Pellot-Barakat, M.F. Insana, A.B. Brill, T. Saga, M. Hiraoka, K. Togashi, Thyroid gland tumor diagnosis at US elastography. Radiology 237, 202–211 (2005) T. Rago, F. Santini, M. Scutari, A. Pinchera, P. Vitti, Elastography: new developments in ultrasound for predicting malignancy in thyroid nodules. J. Clin. Endocrinol. Metab. 92, 2917–2922 (2007) M. Friedrich-Rust, A. Sperber, K. Holzer, J. Diener, F. Gru¨nwald, K. Badenhoop, S. Weber, S. Kriener, E. Herrmann, W.O. Bechstein, S. Zeuzem, J. Bojunga, Real-time elastography and contrast-enhanced ultrasound for the assessment of thyroid nodules. Exp. Clin. Endocrinol. Diabetes 118, 602–609 (2010) Royal College of Physicians of London, Guidelines for the Management of Thyroid Cancer, 2nd edn. (The Lavenham Press, Suffolk 2007) E.S. Cibas, S.Z. Ali, The bethesda system for reporting thyroid cytopathology. Thyroid 19, 1159–1165 (2009) P. Trimboli, G. Treglia, L. Guidobaldi, E. Saggiorato, G. Nigri, A. Crescenzi, F. Romanelli, F. Orlandi, S. Valabrega, R. Sadeghi, L. Giovanella, Clinical characteristics as predictors of malignancy in patients with indeterminate thyroid cytology: a metaanalysis. Endocrine 46, 52–59 (2014) P. Trimboli, E. Condorelli, A. Catania, S. Sorrenti, Clinical and ultrasound parameters in the approach to thyroid nodules cytologically classified as indeterminate neoplasm. Diagn. Cytopathol. 37, 783–785 (2009) A. Bartolazzi, F. Orlandi, E. Saggiorato, M. Volante, F. Arecco, R. Rossetto, N. Palestini, E. Ghigo, M. Papotti, G. Bussolati, M.P. Martegani, F. Pantellini, A. Carpi, M.R. Giovagnoli, S. Monti, V. Toscano, S. Sciacchitano, G.M. Pennelli, C. Mian, M.R. Pelizzo, M. Rugge, G. Troncone, L. Palombini, G. Chiappetta, G. Botti, A. Vecchione, R. Bellocco, Italian Thyroid Cancer Study Group (ITCSG), Galectin-3-expression analysis in the surgical selection of follicular thyroid nodules with indeterminate fine-needle aspiration cytology: a prospective multicentre study. Lancet Oncol. 9, 543–549 (2008) Z.W. Baloch, S. Fleisher, V.A. LiVolsi, P.K. Gupta, Diagnosis of ‘‘follicular neoplasm’’: a gray zone in thyroid fine-needle aspiration cytology. Diagn. Cytopathol. 26, 41–44 (2002) E. Saggiorato, T. Angusti, R. Rosas, M. Martinese, M. Finessi, F. Arecco, E. Trevisiol, N. Bergero, B. Puligheddu, M. Volante, V. Podio, M. Papotti, F. Orlandi, 99 mTc-MIBI Imaging in the presurgical characterization of thyroid follicular neoplasms: relationship to multidrug resistance protein expression. J. Nucl. Med. 50, 1785–1793 (2009) A. Liberati, D.G. Altman, J. Tetzlaff, C. Mulrow, P.C. Gøtzsche, J.P. Ioannidis, M. Clarke, P.J. Devereaux, J. Kleijnen, D. Moher, The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J. Clin. Epidemiol. 62, e1–e34 (2009) T. Rago, M. Scutari, F. Santini, V. Loiacono, P. Piaggi, G. Di Coscio, F. Basolo, P. Berti, A. Pinchera, P. Vitti, Real-time elastosonography: useful tool for refining the presurgical diagnosis in thyroid nodules with indeterminate or nondiagnostic cytology. J. Clin. Endocrinol. Metab. 95, 5274–5280 (2010)
Endocrine 17. P.V. Lippolis, S. Tognini, G. Materazzi, A. Polini, R. Mancini, C.E. Ambrosini, A. Dardano, F. Basolo, M. Seccia, P. Miccoli, F. Monzani, Is elastography actually useful in the presurgical selection of thyroid nodules with indeterminate cytology? J. Clin. Endocrinol. Metab. 96, e1826–e1830 (2011) 18. U. Unlu¨tu¨rk, M.F. Erdog˘an, O. Demir, S. Gu¨llu¨, N. Bas¸ kal, Ultrasound elastography is not superior to grayscale ultrasound in predicting malignancy in thyroid nodules. Thyroid 22, 1031–1038 (2012) 19. P. Trimboli, R. Guglielmi, S. Monti, I. Misischi, F. Graziano, N. Nasrollah, S. Amendola, S.N. Morgante, M.G. Deiana, S. Valabrega, V. Toscano, E. Papini, Ultrasound sensitivity for thyroid malignancy is increased by real-time elastography: a prospective multicenter study. J. Clin. Endocrinol. Metab. 97, 4524–4530 (2012) 20. F. Ragazzoni, M. Deandrea, A. Mormile, M.J. Ramunni, F. Garino, G. Magliona, M. Motta, B. Torchio, R. Garberoglio, P.P. Limone, High diagnostic accuracy and interobserver reliability of real-time elastography in the evaluation of thyroid nodules. Ultrasound Med. Biol. 38, 1154–1162 (2012) 21. G. Russ, B. Royer, C. Bigorgne, A. Rouxel, M. Bienvenu-Perrard, L. Leenhardt, Prospective evaluation of thyroid imaging reporting and data system on 4,550 nodules with and without elastography. Eur. J. Endocrinol. 168, 649–655 (2013) 22. N. Monhey, T.A. Hassana, S. Abdel-Bakib, Role of combined grey scale US and US tissue elastography in differentiating solid thyroid nodules. Egypt. J. Radiol. Nuclear. Med. 44, 505–512 (2013) 23. F. Garino, M. Deandrea, M. Motta, A. Mormile, F. Ragazzoni, N. Palestini, M. Freddi, G. Gasparri, E. Sgotto, D. Pacchioni, P.P. Limone, Diagnostic performance of elastography in cytologically indeterminate thyroid nodules. Endocrine (2014). doi:10.1007/ s12020-014-0438-0
24. V. Cantisani, S. Ulisse, E. Guaitoli, C. De Vito, R. Caruso, R. Mocini, V. D’Andrea, V. Ascoli, A. Antonaci, C. Catalano, F. Nardi, A. Redler, P. Ricci, E. De Antoni, S. Sorrenti, Q-elastography in the presurgical diagnosis of thyroid nodules with indeterminate cytology. PLoS One 7, e50725 (2012) 25. T.J. Sebo, What are the keys to successful thyroid FNA interpretation? Clin. Endocrinol. 77, 13–17 (2012) 26. P. Trimboli, A. Crescenzi, Thyroid core needle biopsy: taking stoke of the situation. Endocrine (2014). doi:10.1007/s12020014-0382-z 27. L. Giovanella, L. Ceriani, G. Treglia, Role of isotope scan, including positron emission tomography/computed tomography, in nodular goitre. Best. Pract. Res. Clin. Endocrinol. Metab. 28, 507–518 (2014) 28. G. Treglia, C. Caldarella, E. Saggiorato, L. Ceriani, F. Orlandi, M. Salvatori, L. Giovanella, Diagnostic performance of (99m)TcMIBI scan in predicting the malignancy of thyroid nodules: a meta-analysis. Endocrine 44, 70–78 (2013) 29. E. Szczepanek-Parulska, K. Wolin´ski, A. Stangierski, E. Gurgul, M. Ruchała, Biochemical and ultrasonographic parameters influencing thyroid nodules elasticity. Endocrine 47, 519–527 (2014) 30. M. Andrioli, L. Persani, Elastographic techniques of thyroid gland: current status. Endocrine 46, 455–461 (2014) 31. J. Bojunga, E. Herrmann, G. Meyer, S. Weber, S. Zeuzem, M. Friedrich-Rust, Real-time elastography for the differentiation of benign and malignant thyroid nodules: a meta-analysis. Thyroid 20, 1145–1150 (2010) 32. V. Veer, S. Puttagunta, The role of elastography in evaluating thyroid nodules: a literature review and meta-analysis. Eur. Arch. Otorhinolaryngol. (2014). doi:10.1007/s00405-014-3155-7
123
