A multicenter survey of current practices of 99mTc-methoxy-isobutyl-isonitrile (MIBI) imaging for the diagnosis of thyroid nodules: more standardization is essential

Molecular imaging with 99mTc-methoxy-isobutyl-isonitrile (99mTc-MIBI, MIBI) has been used in the assessment of thyroid nodules (TNs) for more than two decades. Many studies showed that MIBI imaging is a suitable tool to rule-out malignancy when negative. However, relatively low specificity and accuracy have been described, thus, limiting its acceptance in clinical practice. Additionally, different technologies, protocols, and interpretation criteria are adopted accounting for heterogeneous data reported in the literature. Therefore, the present study was undertaken to assess the clinical use and methodology of MIBI imaging in patients with nodular thyroid disease in Europe. A questionnaire was sent to 12 European centers of Nuclear Medicine. The questionnaire encompassed ultrasound (US) and fine-needle aspiration cytology (FNAC) procedures and their evaluation as well scintigraphy imaging indications, technical procedures, and interpretation criteria of MIBI imaging. The survey showed a good agreement of different centers in approaching TNs by TSH measurement, US evaluation and 99mTc-pertechnetate thyroid scintigraphy. MIBI imaging is mainly used to assess TNs with inconclusive/indeterminate cytological findings and selection of target nodule(s) for FNAC in patients with multi-nodular goiter. Technical procedures adopted in different centers are globally comparable and the recorded differences are unlikely to impact clinical results. However, as the main result of the present study, substantial differences were found in interpretation criteria adopted in different centers. Our survey supports the urgent need of standardized interpretation criteria of thyroid MIBI imaging in order to improve its diagnostic performance and make results comparable in clinical practice.


Introduction
Because of the widespread use of ultrasound (US) thyroid nodules (TNs) are a common finding, especially in currently and previously iodine-deficient countries. Although the majority of TNs are benign, non-autonomous TNs on 99m Tc-pertechnetate or 123 I scintigraphy may need further diagnostic work-up to exclude thyroid cancer [1,2]. Depending on the detection of suspicious US features by using risk stratification systems (TIRADS) and the size of the nodule, fine-needle aspiration cytology (FNAC) represents the next step in this diagnostic algorithm [2][3][4]. However, in 25-40% the FNAC, results can be indeterminate or non-diagnostic [5]. Moreover, FNAC can be challenging in patients with multinodular goiter or nodules in an unfavorable location (i.e. below the thoracic aperture), and should not be performed on patients on anticoagulant therapy [6]. Molecular imaging with 99m Tc-methoxy-isobutyl-isonitrile ( 99m Tc-MIBI, MIBI) has been used in the assessment of TNs for more than two decades [7][8][9]. MIBI is a lipophilic cation and member of the isonitrile family. It accumulates within the mitochondria, which have a high negative transmembrane potential. Thus, MIBI uptake is related to high mitochondria 1 3 content and increased vascularity, which are commonly seen in lung and thyroid malignant tumors as well as parathyroid adenomas [10][11][12]. Many studies showed that a MIBI negative TN has a high probability of being a benign lesion [10,[13][14][15], thus, making MIBI imaging a suitable tool to ruleout malignancy. On the other hand, relatively low specificity and accuracy have been described [6][7][8]. Selecting hypofunctioning nodules for MIBI imaging increases the overall accuracy, as hyperfunctioning nodules are commonly benign but MIBI positive due to an increased vascularization and higher proliferation rate of hyperfunctioning follicular cells. Even in hypofunctioning nodules, however, different technologies (i.e., planar and SPECT images), injected activities, protocols, and image interpretation criteria account for heterogeneous results reported in the literature [7]. Therefore, the present study was undertaken to assess the clinical use and methodology of MIBI imaging in patients with nodular thyroid disease in Europe.

Materials and methods
The study was approved by the ethics committee of Magdeburg University Hospital (No. RAD 378-32/20) and the need for informed consent was waived. Firstly, 12 nuclear medicine centers offering complete clinical management of thyroid patients (i.e., including US and FNAC in addition to nuclear imaging and therapy) agreed to participate in the present study during the EANM Thyroid Committee Interesting Group meeting (Düsseldorf 2018).
A questionnaire (Table 1) was sent to these centers in December 2018. The questionnaire encompassed US and FNAC procedures and their evaluation as well as thyroid MIBI imaging indications, technical procedures, and interpretation criteria. The survey was carried out from  Items  Question/request   1  Imaging equipment (ultrasound devices and gammacameras), type, manufacturer, country  2  Thyroid ultrasound interpretation criteria  3  Indications for thyroid scintigraphy ( 99m Tc-pertechnetate)  4 Indications for FNAC and cytology reporting 5 Indications for 99m Tc-MIBI thyroid imaging 6 99m Tc-MIBI imaging protocols 1 (MIBI activities, timing of planar images, use of SPECT (-CT)  7 99m Tc-MIBI imaging protocols 2 (acquisition parameters) 8 99m Tc-MIBI imaging: interpretation criteria

Results
Technical instrumentations available in the different study centers are listed in Table 2.

Initial approach in patients with thyroid nodules
Patients with TNs first underwent TSH measurement and neck US in all centers. Then, a 99m Tc-pertechnetate scintigraphy was performed for nodules with a maximum diameter ≥ 10 mm and different TSH levels (any value in four centers, values from 0.50 to 3.0 mUI/L in others) ( Table 3). Hyperfunctioning nodules were excluded from FNAC, while the management of hypofunctioning nodules and those not evaluated by 99m Tc-pertechnetate scintigraphy (i.e., TSH above the local threshold) is centered on neck US to stratify the risk of malignancy and yardstick for additional FNAC. Among 12 participating centers, 11 (92%) use TIRADS (EU-TIRADS: eight, Kwak-TIRADS: two, ACR-TIRADS: one) for stratifying the risk of malignancy, while a descriptive report is provided by one (8%), center, respectively (Table 4). In two centers, hypofunctioning nodules larger than 10 mm underwent FNAC, one for nodules at intermediate or high risk; the other centers applied TIRADS criteria, combined with nodule's size, for stratification. In general, FNAC is not used in nodules with a low-risk TIRADS score. Intermediate-risk nodules underwent FNAC when larger than 20 mm. Finally, high-risk nodules underwent FNAC in all cases, with some differences for nodules < 10 mm where FNAC and wait-and-see strategy are applied in different centers.
In all centers, 99m Tc-MIBI imaging is only performed after exclusion of autonomously functioning thyroid nodules (AFTNs) using 99m Tc-pertechnetate scintigraphy. Administered activities of 99m Tc-MIBI ranged from 185 to 700 MBq, respectively. After tracer injection early (10-30 min) and late (60-90 min) planar images were acquired in 10 (83%) and 12 (100%) centers, respectively. An additional single-photon emission computed tomography (SPECT) is performed on a routine basis in six (50%) centers and selectively, on the decision of the attending nuclear medicine physician in two centers (16%) while it is not performed on the remaining four centers (34%) ( Table 7). Two centers also perform hybrid SPECT/CT imaging in selected patients. Acquisition protocols and parameters for planar and tomographic emission imaging as well as CT imaging are summarized in Tables 7 and 8, respectively.
The concordance rate of relevant issues is summarized in Fig. 1.

Discussion
In this manuscript, the procedures, the indications, and the imaging interpretation criteria for MIBI imaging in various areas of Europe were investigated by a survey that showed a good agreement of different centers in approaching TNs by TSH measurement, US evaluation and selective use of 99m Tc-pertechnetate thyroid scintigraphy based on nodule's size (i.e., ≥ 10 mm) and TSH levels. A notable exception is observed in the German centers, where 99m Tc-pertechnetate thyroid scintigraphy is performed independently of the TSH levels. Due to the fact, that the majority of AFTNs in Germany present with normal TSH levels [18], the German Society of Nuclear Medicine recommends thyroid scintigraphy for all TNs of 10 mm or larger independent of the TSH level [19]. In other centers, however, the adopted TSH thresholds are significantly higher than those proposed by clinical guidelines such as the 2015 ATA guideline reflecting differences in iodine intake and prevalence of AFTNs between the United States and Europe, as well as between different European regions. After excluding AFTNs, all but two centers based the decision to perform FNAC or not on ultrasound TIRADS patterns. Even if different US TIRADS and cytological reporting systems are adopted in different centers, significant differences are unlikely as all methods proved to be accurate and are comparable in terms of accuracy [20][21][22]. In this clinical context, 99m Tc-MIBI imaging is mainly used to assess TNs with inconclusive/ indeterminate cytological findings and selection target nodule(s) for FNAC in patients with multi-nodular goiter. Injected 99m Tc-MIBI activities range from 185 to 700 MBq (mean 465 MBq). Early (10-30 min after intravenous injection) and late (60-90 min after intravenous injection) anterior planar images are obtained in 10 (83%) centers, while late images only are obtained in the remaining two centers (16%), respectively. An additional SPECT is also obtained as a standard in seven centers (52%) and in selected cases in two centers (16%), respectively. Hybrid SPECT/CT is also performed in two centers (16%) in selected cases (i.e., mediastinal goiters, preoperative evaluation). When performed, SPECT and SPECT/CT are performed after the late image acquisition. Visual interpretation is based on the evaluation of intranodular 99m Tc-MIBI uptake compared to the normal  All in all, our present survey demonstrates a good agreement between the different nuclear thyroidology centers regarding the approach to nodular thyroid diseases. Adoption of different TIRADS systems and cytology reporting systems should be accounted as a potential source of heterogeneity. However, currently available systems are well comparable in terms of accuracy. Nonetheless, a standardization and harmonization are desirable to ameliorate communication between different specialists and allow a better comparison of research data. Technical procedures adopted in different centers are globally comparable and the recorded differences are unlikely to impact clinical results. However, as the main result of the present study, substantial differences were found in visual interpretation criteria adopted in different centers. Indeed, the WOI is only assessed in two centers for all TNs despite it has been supported as the preferred method for differentiating benign from malignant nodules and, especially, differentiate benign from malignant cytologically indeterminate nodules, respectively [10,13]. The adoption of WOI in daily clinical practice is likely limited by the lengthening of the image analysis times and the need for strict standardization of methods. However, while a qualitatively negative 99m Tc-MIBI scintigraphy reliably excludes malignancy, many benign follicular proliferation will frequently show isointense or hyperintense 99m Tc-MIBI uptake and will only be discriminated by a semi-quantitative assessment. In conclusion, we found satisfactory agreement of different nuclear thyroidology centers concerning indications and technical procedures. At the same time, relevant differences exist in interpretation criteria explaining differences in diagnostic performance reported in the literature.
Preliminary data on the cost-effectiveness of 99m Tc-MIBI in nodular thyroid diseases are encouraging.
Wale and colleagues evaluated the diagnostic performance of 99m Tc-MIBI calculated from a retrospective review of local data on 712 patients combined with a meta-analysis of the published literature. Decision tree analysis was used to calculate the cost-effectiveness and a combined FNAC/ MIBI investigative strategy was proved to be useful in avoiding unnecessary thyroidectomies, saving related costs and potential side effect [14].
Another study compared the cost-effectiveness of 99m Tc-MIBI imaging and the Afirma gene expression classifier for the assessment of cytologically indeterminate TNs. Costs were calculated from the perspective of the German health insurance system. A decision tree model was used and results were confirmed by the Monte Carlo simulation. Life expectancy was 34.3 years (estimated costs per patient €1459-€2224) for the MIBI scan and 34.1 years (estimated costs €3560-€4071) for the molecular test. Therefore, the authors concluded that 99m Tc-MIBI imaging is more costeffective than the gene expression classifier [23].
However, both studies referred to local costs; then for future studies, more standardized approaches will be applied, allowing the evaluation of cost-effectiveness ( 99m Tc-MIBI imaging combined with FNAC) in a larger and multicentric setting [14]. Concordance rates (%)

Conclusions
Our survey supports the urgent need for standardized interpretation criteria of thyroid 99m Tc-MIBI imaging to improve its diagnostic performance and make multicenter results comparable in clinical practice. Based on this, the EANM Thyroid Committee will promote a multicenter prospective study on the clinical results of 99m Tc-MIBI imaging by using harmonized 99m Tc-MIBI imaging interpretation criteria, aiming to develop an EANM standardized protocol for interpreting thyroid 99m Tc-MIBI imaging and improving its clinical relevance.