Key points

Question: Is there a relationship between clinicopathological features of recurrent or metastatic RAI-negative DTC and FDG-avidity on PET/CT?

Introduction

Thyroid carcinoma is one of the most popular endocrine cancer worldwide. Differentiated thyroid cancer (DTC) accounts for 90% of all thyroid cancer types [1]. Approximately 5% of patients with DTC follow a more aggressive course with radioiodine (RAI)-refractory or RAI-negative disease, often becoming the cause of mortality associated with tumor recurrences and distant metastases [2]. These DTC patients were predicted to have poorer prognosis and limited effective treatments including surgery, radiation therapy, chemotherapy, immunotherapy and tyrosine kinase inhibitors [3]. As a result, risk stratification and prognostic evaluation are required to identify high-risk patients and guide the appropriate treatment modality.

Histopathologic and biomolecular markers play an important role in the improvement of risk stratification in DTC. Histopathologic subtype of thyroid cancer is grouped as classical subtype (well-differentiated thyroid carcinoma) and aggressive subtype (poorly differentiated, tall cell, columnar cell, hobnail variants of DTC) [4] . Classical subtype, characterized by a papillary and follicular variant, has an excellent prognosis, while patients with aggressive histopatholologic features in the primary tumor are considered at risk of developing RAI-refractory DTC [5]. BRAFV600E is the most common mutation observed in DTC and triggers tumorigenesis through the mitogen-activated protein kinase (MAPK) pathway [6]. Multiple studies have shown BRAFV600E mutation is associated with poor clinicopathologic outcomes, larger tumor size, local recurrence and distant metastases [7, 8].

[18F] F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) is a noninvasive diagnostic modality and beneficial for localizing residual or recurrent disease, particularly when iodine avidity of disease has been lost (RAI-negative) [9, 10]. The level of metabolic activity of the disease on FDG PET/CT is independently associated with the patient survival, hence may guide to individualize the intensity of follow-up and treatment of these patients [11, 12]. A limited number of studies have investigated the association between FDG-avidity of the tumors with BRAFV600E mutation status and clinicopathological features [13,14,15]. While some studies suggest primary or recurrent tumors bearing BRAFV600E mutation may have higher FDG-avidity [16, 17], data on the correlation between other clinicopathological features and FDG-avidity are rather sparse. Therefore, we aimed to determine the relationship between clinicopathologic factors, BRAFV600E mutation status and FDG avidity in a rather homogenous group of patients with recurrent or metastatic RAI-negative DTC.

Methods

Patients

This is a retrospective review of all patients from 2015 to 2018 who had 1) prior RAI ([131I] treatment, 2) had negative diagnostic or post-treatment iodine scan following suspected recurrent disease based on rising thyroglobulin (Tg) or ultrasound findings, 3) underwent FDG PET/CT for detection of the site of recurrence. The suspected site of recurrence on either FDG PET or contrast-enhanced component of PET/CT were biopsied with BRAFV600E testing on the sample.

The present study protocol was reviewed and approved by the Institutional Review Board of Viet Nam Ministry of Science and Technology (approval No. 02/HĐTK-ĐTCT-KC.10.03/16–20).

Procedures

PET/CT examination was performed, using GE Discovery 710, according to the European Association of Nuclear Medicine (EANM) guidelines, version 1.0 [18]. For patient preparation, the serum glucose level was checked to exclude hyperglycemia. Afterward, the patients rested in the waiting room before intravenous injection of 2.5 MBq/kg body weight (±10%) of FDG. Contrast-enhanced CT with 100 ml of the contrast material with a scan delay of 30 s and an injection rate of 3 ml/s from the skull base to the mid-thigh was performed 60 min after FDG injection. The parameters of the CT scan were as follows: 120 kVp, 100 mA, the helical thickness of 3.75 mm and 0.5 s/rotation. PET images were reconstructed using an iterative algorithm with attenuation correction with CT.

Histopathologic type of thyroid cancer was divided into classical type (well-differentiated thyroid carcinoma) and aggressive histologic types (poorly differentiated, tall cell, columnar cell, hobnail variants of DTC). BRAFV600E mutation was analyzed by using immunohistochemistry and real-time PCR. The immunohistochemical method was performed using anti-BRAFV600E (VE1) antibody (Ventana Medical System) on automated BenchMarch Ultra (Ventana Medical System, USA). DNA for real-time PCR was extracted from formalin-fixed, paraffin-embedded tissue obtained from core needle biopsy (n = 44) and percutaneous needle aspiration (n = 19) specimens. Each DNA sequence was read on an ABI-PRISM 3100 automatic sequencer (Applied Biosystems) in order to determine the presence of the BRAF mutation. The positive BRAFV600E mutation was decided by the concordance between the histopathologist and molecular biologist in 108 hospital.

Qualitative and semiquantitative assessment

FDG PET/CT images were evaluated by two nuclear medicine physicians (NMPs) blinded to the clinical data. FDG-avid lesions are defined as uptake above that of mediastinal blood pool activity by the consensus of two NPMs. Semiquantitative analysis of highest hypermetabolic lesion, determined by maximum standardized uptake value (SUVmax) was assessed by automated polygonal regions of interest (ROIs) drawing on attenuation-corrected PET images using the GE workstation (version 4.7, GE Healthcare). For suspected malignant lesions on contrast-enhanced CT with no visual FDG uptake on PET, the manual ROIs were drawn on CT and cloned to co-registered PET images to record the SUVmax. In cases of multiple malignant lesions, an ROI was drawn on a lesion with the highest SUVmax on PET.

Statistical analysis

Categorical values were compared utilizing the Chi-squared test or Fisher’s exact test. Continuous variables following normal distribution were compared with paired t-test or repeated measure ANOVA and for variables not following a normal distribution with Wilcoxon signed-rank test or Friedman test. To analyze the relationship between clinicopathologic variables and SUVmax, the univariate analysis was performed by logistic regression. The significance threshold was set at P ≤ 0.05. The statistical software Statistical Package for the Social Sciences (SPSS) 20.0 (SPSS Inc., Chicago IL, USA) and GraphPad Prism (version 8.0 Graphpad Software, Inc., USA) have been used to analyze the data.

Results

Patients characteristics

Sixty-three consecutive patients, 55(84.8%) female, and 8 (15.2%) male, with the median age of 48 (range,17–81), were included in the study. The patients had received median 3 doses (range 1–9) of I-131 and the median thyroglobulin (Tg) at the time of suspected recurrence was 262.7 ng/ml (Table 1). The majority of the patients, 55/63 (89%), had BRAFV600E mutation. In regard to histopathologic variants, the proportion of classical DTC was 45/63 (71.4%) and higher than those of aggressive subtypes 18/63 (28.6%). There was no significant difference in the prevalence of BRAFV600E mutation by the histopathologic subtype of DTC, 40/45 (88.9%) in classical subtype vs. 15/18 (83.3%) in aggressive subtype (P = 0.6), respectively (Table 2).

Table 1 Clinicopathologic features of radioiodine-negative DTC patients
Full size table
Table 2 The relationship between BRAFV600E mutation and histopathologic types of DTC
Full size table

FDG PET detected sites of hypermetabolic recurrence or metastatic disease in most patients 58/63 (92.1%), while the remainder were diagnosed based on contrast-enhanced CT component of the study. Fifty-four/63 (85.7%) patients had regional recurrent disease (thyroid bed or cervical lymph nodes) while 9/63 (14.2%) patients had distant metastatic. Of the latter group, almost all except one had also evidence of local recurrence too. The location of metastatic lesions was mostly in cervical lymph nodes (cervical lymph node alone was seen in 44/63 [69.8%]) while the thyroid bed and distant metastases alone were noted in 2/63 (3.2%) and 1/63 (1.6%), respectively (Table 1).

Relationship between SUVmax, histopathologic types and BRAFV600E mutation

The median SUVmax of the BRAFV600E mutation tumors was not significantly higher than those of wild type tumors, 9.5 vs. 6.8, P = 0.1 (Fig. 1). In contrast, the median SUVmax of the aggressive histopathologic subtype was significantly higher than the classical subtype, 13.9 vs. 7.2, P = 0.0004 (Fig. 1). No significant difference in FDG uptake was found between subgroups of classical DTC with or without BRAFV600E mutation. However, we observed a significant difference of SUVmax in aggressive subtypes with BRAFV600E mutation compared to those with classic subtypes with BRAFV600E mutation, 15.1 vs. 6.2 vs. P = 0.0028 (Fig. 1).

Fig. 1
figure 1

Distribution of SUVmax stratified by histopathologic types of DTC and BRAFV600E mutation status. * significant difference. ns: not significant

Full size image

To assess the partial volume effect and SUVmax, we compared the SUVmax between tumors > 10 mm, < 10 mm. SUVmax of lesions > 10 mm was significantly higher than those of lesions < 10 mm, 13.4 vs. 6.3, P = 0.0003 (Fig. 2). Consistent results were seen in the subgroup of BRAFV600E positive, between tumors larger than 10 mm and smaller than 10 mm, 14.3 vs. 6.5, P = 0.0002 (Fig. 2).

Fig. 2
figure 2

Distribution of SUVmax of DTC tumors stratified by tumor size and BRAFV600E mutation status. * significant difference. ns: not significant

Full size image

The relationship between clinicopathologic factors and SUVmax is detailed in Table 3. In univariate analysis, FDG uptake was predicted by lesion size, histopathologic type, and distant metastases, p = 0.04, p = 0.001 and p = 0.004 respectively. There was no relationship between SUVmax and BRAFV600E mutation or Tg level on univariate analysis (Figs. 3 and 4).

Table 3 Logistic regression analysis for the assessment of association between various factors and SUVmax
Full size table
Fig. 3
figure 3

Forty-year-old female with papillary thyroid cancer, aggressive histopathologic type, underwent total thyroidectomy and then received 131I therapy with a total dose of 550 mCi. Diagnostic 131I whole-body scan after the 3rd treatment showed no abnormal uptake and the patient’s stimulated Tg was > 1000 ng/ml. PET (panel A and D), CT (panel B) and fused PET/CT (panel D) images showed markedly FDG-avid metastatic disease (SUVmax up to 35.55, cross-hair). Surgical resection of that lesion confirmed metastases of papillary thyroid carcinoma, immunohistochemistry and real-time PCR showed BRAF V600E mutation

Full size image
Fig. 4
figure 4

FDG PET/CT with contrast enhancement of a sixty-eight-year-old female with a diagnosis of papillary thyroid carcinoma, previously treated with total thyroidectomy. The patient presented with serum thyroglobulin level of 500 ng/ml and negative post-therapy 131I whole-body scan after 3rd treatment. The PET image (panel A and D) illustrated multifocal FDG uptake with the hottest lesion seen in the right axillary region (SUVmax: 13.59) correlated with 11 × 20 mm lymph node on contrast-enhanced CT (panel B, crosshair) and fused PET/CT image (panel C, crosshair). The histopathologic assessment of the lymph node confirmed thyroid carcinoma and immunohistochemistry and real-time PCR was negative for BRAF V600E mutation

Full size image

Discussions

In this study, we have shown the independent association between FDG avidity and aggressive histopathologic subtypes in patients with recurrent or metastatic RAI-negative disease. The strength of this study was that all patients had histopathologic confirmation of recurrence or metastases and underwent BRAF mutation testing by immunohistochemically and real-time PCR method. In the majority (92%) of the patients, FDG PET detected sites of recurrent disease in RAI-negative patients. This was consistent with the study performed by Rivera et al. which showed that 77% of FDG-avid recurrent or metastases in patients with RAI-refractory disease are of aggressive subtype [19]. However, that study did not describe a comprehensive definition of RAI-refractory state and also did not assess the relationship between SUVmax and histopathologic factors. The underlying mechanism of FDG avidity in aggressive DTC cells is likely the upregulation of glucose transporter-1 (GLUT1) and reduced expression of sodium-iodide symporter (NIS) and this phenomenon proposed that FDG PET/CT is an effective diagnostic tool in aggressive subtypes of DTC [20, 21].

To our surprise, this study did not show the association between BRAFV600E mutation and FDG-avidity. Previous studies have suggested that the SUVmax in BRAFV600E mutation bearing tumors was significantly higher than wild type variants in papillary thyroid cancer (PTC) [22, 23]. It has also been shown that BRAF V600E mutation is significantly associated with the expression of GLUT-1 and glycolysis in thyroid cancer [24, 25]. The prevalence of BRAFV600E mutation was 87.5% in this study which is consistent with other studies in poorly differentiated thyroid carcinoma and tall cell PTC [26,27,28]. However, the lower number of patients in the non-BRAF mutation may have limited the statistical power of our study. Furthermore, the result of The Cancer Genome Atlas project indicated that BRAFV600E PTC represents a diverse group of tumors, consisting of multiple molecular subtypes, with variable degrees of thyroid differentiation [29]. Their results also suggested more refined reclassification of thyroid cancers into molecular subtypes that better reflect their underlying signaling and differentiation properties, and potentially better informing the management of these patients. Whether FDG-avidity can be used as an additional clinical tool to further refine the subtypes of tumors with BRAFV600E mutation may warrant further studies.

Expectedly, our study showed that lesion size is an independent factor of FDG positivity. FDG avidity is strongly dependent on tumor size due to the partial volume effect [30]. Yoon et al. reported in the study of DTC patients that SUVmax of the larger tumors (> 20 mm) was significantly higher than tumors with a smaller size (< 10 mm) [31]. We assumed that the partial volume effect may impact the SUVmax values of the smaller tumor (< 10 mm). When the analysis was performed only in tumors with larger size (> 10 mm), again no significant difference in SUVmax between BRAFV600E mutation and wild type patients was found.

The limitation of our study is the small number of patients in the group with negative BRAFV600E mutation in comparison with the majority with positive mutation which may have limited the statistical power of our study to depict the significant difference in FDG uptake between these two groups. Besides, the limited number of patients in the subgroup with metastatic disease did not allow us to investigate the relationship between BRAFV600E mutation and histopathologic types in these subcategories. As a result, SUVmax values could not be compared to enhance our understanding of the behavior of distant metastatic sites. All patients in our study were completely RAI- negative with varying levels of FDG avidity. However, in clinical practice, a mixed picture could also be seen where both RAI-avid and FDG-positive diseases coexist. Whether these patients still may benefit from RAI treatment or immediately need to be considered for other therapeutic approaches is unclear.

Conclusions

In patients with RAI-negative and suspected recurrence, FDG PET/CT detects the sites of metastatic disease in the majority of patients. This study suggests that aggressive histopathologic subtypes but not BRAFV600E mutation status are significantly associated with FDG-avidity of the recurrent or metastatic lesions in RAI-negative DTC.