Predicting 131I-avidity of metastases from differentiated thyroid cancer using 18F-FDG PET/CT in postoperative patients with elevated thyroglobulin

The quantitative relationship between iodine and glucose metabolism in metastases from differentiated thyroid cancer (DTC) remains unknown. Aim of the prospective study was to establish the value of 18F-FDG PET/CT in predicting 131I-avidity of metastases from DTC before the first radioiodine therapy. A total of 121 postoperative DTC patients with elevated stimulated serum thyroglobulin (ssTg) who underwent 131I adjuvant therapy or therapy after 18F-FDG PET/CT scan were enrolled. The Receiver operating characteristic curve was established to create an optimal cut-off point and evaluate the value of SUVmax for predicting 131I-avidity. In our study, the median SUVmax in 131I-nonavid metastatic target lesions was also significantly higher than that in 131I-avid metastatic target lesions (5.37 vs. 3.30; P = 0.000). At a cut-off value of 4.0 in SUVmax, the area under curve was 0.62 with the sensitivity, specificity, positive predictive value and negative predictive value of 75.3%, 56.7%, 76.1%, and 54.8%, respectively. These results suggest that 18F-FDG PET/CT may be of great value in identifying metastases in postoperative DTC patients with elevated ssTg before 131I administration, leading to an improved management of disease. 18F-FDG positive metastatic DTC with SUVmax of greater than 4.0 possesses higher probability of non-avidity to radioiodine.

molecular level and the lack of definitive evidence to avid stunning even using very low dose of 131 I according to strict imaging protocols have also been reported [7][8][9] .
Recently, as a non-invasive whole-body imaging technique, the value of 18 F-FDG PET/CT in DTC has been confirmed by several studies. 18 F-FDG PET/CT is especially sensitive and effective in detecting metastatic DTC lesions, especially in 131 I WBS-negative, Tg-positive patients after 131 I administration [10][11][12][13][14] . Besides, 18 F-FDG PET/ CT has also been demonstrated to aid stratify DTC patients [15][16][17] . But the experience on the application of 18 F-FDG PET/CT before 131 I administration in the identification of metastases in postoperative DTC patients is very limited 17,18 . Feine et al. demonstrated that radioiodine-avid thyroid cancer lesions usually be 18 F-FDG-nonavid, and vice versa 19 . However, metastatic DTC lesions with both uptake of 18 F-FDG and 131 I have also been found by previous studies [20][21][22] . Therefore, the relationship between radioiodine accumulation and 18 F-FDG metabolism in DTC have not been quantitatively assessed to date. And the value of maximum standardized uptake value (SUVmax) measured on 18 F-FDG PET/CT in predicting the status of 131 I uptake in metastases from DTC remains unknown.
Here, therefore, we carried out this dedicated prospective study to evaluate the role of 18 F-FDG PET/CT in identifying metastatic DTC in postoperative patients with elevated stimulated serum Tg (ssTg) before 131 I administration. The value of 18 F-FDG PET/CT in predicting the 131 I-avidity of metastatic DTC are qualitatively and quantitatively assessed as well.

Results
Characteristics of patients. One hundred and twenty-one consecutive DTC patients (15 with follicular and 106 with papillary carcinoma) constituted the study group including 47 (38.84%) males and 74 (61.16%) females. Eighty-five (70.25%) patients had undergone a total thyroidectomy and 36 (29.75%) patients had undergone a near total thyroidectomy before the enrollment. Mean age of patient at diagnosis was 45.3 ± 12.9 years. The mean interval between thyroidectomy and 131 I administration was 3.2 ± 1.4 months. Before 131 I administration, mean TSH and ssTg were 75.6 ± 32.7 mIU/L and 59.6 ± 49.2 ng/mL, respectively. Positive TgAb (>115 kIU/L) was found in nine patients. The median follow-up of all patients was 22.3 months (range, 10.7-34.5 months). A flow diagram is given in Fig. 1.

Efficacy of 18 F-FDG PET/CT in identifying metastases in postoperative DTC patients with elevated ssTg.
Final diagnostic criteria for metastases from DTC were based on pathological findings, serum Tg levels and other imaging techniques including CT, high-resolution ultrasonography, magnetic resonance imaging, and by correlation with clinical follow-up 23 .
In all the enrolled 121 postoperative DTC patients with elevated ssTg, final diagnosis of metastatic DTC was established in 104 (85.95%) patients. Seventeen (14.05%) subjects were excluded by negative imaging (ultrasonography and CT) and clinical follow up, including three patients (5 lesions located in lymph node) showing false-positive results in 18 F-FDG PET/CT. Of the 104 DTC patients with metastatic lesions, 92 (88.46%) cases were detected by 18 F-FDG PET/CT, and the remaining 12 (11.54%) patients showed negative 18 F-FDG PET/CT results. The sensitivity, specificity, negative predictive value, positive predictive value, and accuracy of 18   the serum TSH level, blood glucose level and uptake time of 18 F-FDG between the patients with 18 F-FDG-avid metastases and 18 F-FDG-nonavid patients were found.
The information obtained from 18 F-FDG PET/CT scan led to changes in management decision in 39 (32.23%) of the 121 patients. In detail, 18 F-FDG PET/CT showed residual cervical nodal metastases avid for 18 F-FDG with the smallest dimension ≥1 cm in 5 patients, who were then referred back to surgery for reoperative neck dissection prior to 131 I therapy; Besides, an activity of 5.55GBq was enhanced to 7.4 GBq in 34 patients in consideration with their lung and/or bone metastases determined by 18 F-FDG PET/CT just before 131 I administration. Analyses of factors potentially relative to 131 I uptake. In patient-based analyses, the clinical characteristics, including age (≤45 years or >45 years), gender (male or female), pathological type (PTC or FTC), qualitative FDG uptake (positive or negative), ssTg (10-100 ng/mL, 100-1000 ng/mL or >1000 ng/mL), TSH (30-60 mIU/L, 60-90 mIU/L, >90 mIU/L), and TgAb (≤115 kIU/L or >115 kIU/L) were analyzed as independent variables using χ 2 test. After univariate analysis between the groups with 131 I-avid and 131 I-nonavid metastases, significant factors related to the ability of 131 I uptake were age and TSH. After multivariate logistic regression analysis, only TSH remained significant for 131 I uptake (Table 2).

Comparison of findings in metastatic DTC patients between
In lesion-based analyses, 18 of 99 patients with metastases which did not meet the criteria of target lesion (RECIST 1.1) were excluded. Then, the size, site, qualitative (positive or negative) and quantitative data (SUVmax) of 18 F-FDG uptake of 160 target lesions in the remaining 81 patients were taken into account, with details in

Discussion
In the current prospective study, we initiated 18 F-FDG PET/CT scans just before 131 I administration in postoperative DTC patients with elevated ssTg, which was compared directly with the subsequent 131 I Rx-WBS. We observed significantly higher possibility of 18 F-FDG uptake in 131 I-nonavid DTC metastases than in 131 I-avid ones (95.3% vs. 76.1%), which is generally in accordance with the previously published data on 18 F-FDG PET/CT scans after 131 I administration 17,24 . Lesion-based analyses demonstrated that the percentage of positive 18 F-FDG uptake and the median SUVmax in 131 I-nonavid metastases were significantly higher than those in 131 I-avid metastases.
In the process of dedifferentiation of DTC, up-regulated glucose metabolism combined with down-regulated iodine metabolism coexist [25][26][27] . We have even shown down-regulated glucose metabolism combined with up-regulated iodine metabolism in papillary thyroid carcinoma cells treated with tyrosin kinase inhibitors by an recent in vitro study, which elucidated partially the underlying mechanism of such "flip-flop" phenomenon and lay a foundation of the present study 28 . Theoretically, the level of glucose metabolism can be both qualitatively   (Table 2), however, the status of 18 F-FDG uptake in 131 I-avid group was not significantly different from that in 131 I-nonavid group, indicating that only per-patient qualitative analysis of 18 F-FDG PET/CT is not sufficient to distinguish 131 I-nonavid metastases from those avid to 131 I. The status of 18 F-FDG uptake in different metastatic lesions could be different even in an individual, which can be explained by the multicentricity and polyclone of DTC [29][30][31] . Our findings revealed that lesion-based analyses and quantitatively assessing the data of 18 F-FDG PET/CT using SUVmax to predict 131 I-avidity for metastatic DTC would be more reliable than qualitative per-patient evaluation only. Moreover, in predicting the capability of DTC in accumulating 131 I, other isotopes ( 123 I and 124 I) and the topography of primary malignant thyroid nodule can be resorted [32][33][34] . However, difficult availability of such isotopes and the small sample size of enrolled patient with metastatic lesions represent major limitations. And since the value of molecular testing in guiding postoperative 131 I therapy has not be established well, no recommendation in this aspect can be provided at present 2 .
The level of serum Tg is related to the amount of neoplastic tissue in postoperative patients [35][36][37] . In those patients with elevated serum Tg (generally >10 ng/mL) and negative 131 I WBS, 18      In addition, we further investigated the potential patient-based factors related to 131 I uptake. After multivariate logistic analysis, only TSH remained significant for 131 I uptake, demonstrating that the TSH level is a significant factor that might influence the 131 I uptake of DTC metastases. The 131 I uptake by metastatic DTC might continue to increase as TSH rose to a higher level. Recently, one study has also confirmed that ssTg measured under a higher preablative TSH level might be more convincing as a prognostic marker for DTC 38 . In our study, 18 F-FDG PET/CT was carried out just one day before 131 I administration to maximize TSH stimulation without additional preparation and minimize inconvenience to patients.
We admit that there are several limitations in our study. First, in reflecting glucose metabolism, only SUVmax was focused on in the present study, other parameters such like SUV normalized to lean body mass, or body surface area, may be worthy of assessment in the coming studies. Meanwhile, many physiologic and technical variables may affect the outcome of SUVmax, resulting in difficulties of its reproducibility 39,40 . Second, our studies enrolled patients with metastases only in lymph node, lung and bone. So the predictive value of SUVmax in metastatic DTCs in other sites is still unknown. Third, the exact prognostic value, if any, of 18 F-FDG PET/CT via survival analyses of such indexes remains to be more clearly established.

Conclusion
Our study suggest that 18 F-FDG PET/CT can play vital role in identifying metastases in postoperative DTC patients with elevated ssTg (>10 ng/mL) before 131 I administration, leading to refined management of disease. 18 F-FDG positive metastatic DTC with SUVmax of greater than 4.0 possesses higher probability of non-avidity to radioiodine.

Materials and Methods
Study design and population. We prospectively enrolled postoperative DTC patients who received total or near total thyroidectomy by our general surgery and were suspected to have metastatic disease with elevated ssTg levels (>10 ng/mL). Central neck lymph node removal was conducted in all DTC patients without the history of other malignant tumors. All patients had undergone thyroxin withdrawl for 4 weeks before 131 I administration. The prescribed 131 I activity was either 5.5 GBq (150 mCi) for local metastases and suspected but unproven metastases (adjuvant therapy) or 7.4 GBq (200 mCi) for distant metastases. The findings of 18 F-FDG PET/CT scans and 131 I Rx-WBS were directly compared.
This study was approved by the ethics board of Shanghai Jiao Tong University Affiliated Sixth People's Hospital before its initiation. All participants were fully informed of details of the study with the information sheet and signed in the consent form prior to their inclusion in the study. We confirm that all methods were carried out in line with the relevant guidelines and regulations. tration in our department of nuclear medicine. All patients fasted for at least 6 h and the blood glucose level was less than 150 mg/dL (8.3 mmol/L) before intravenous injection of 18  Criteria of target lesion and final diagnoses. According to RECIST 1.1, we chose up to 5 lesions per patient and up to 2 per organ as target lesions in 18 F-FDG PET/CT images, and monitored in the subsequent post-therapeutic 131 I scans 41 . 18 F-FDG uptake in tumor was quantified as SUV max (the maximum activity concentration of 18 F-FDG divided by the injected dose and corrected for the body weight of the patient) 40 .
Statistical analyses. SPSS version 16.0 (SPSS, Inc. Chicago, IL, USA) were used for statistical analyses. The significance of categorical data were compared using Fisher exact tests and Chi-Square Tests. The nonparametric Wilcoxon rank sum test was applied to evaluate quantitative data when it was not normally distributed. The factors related to 131 I uptake were investigated using logistic regression analysis. Receiver-operating characteristic (ROC) curve analysis was used to obtain the cut-off value of SUVmax for differentiating 131 I uptake status in metastatic lesions. Sensitivity, specificity, positive predictive value and negative predictive value of the cutoff value of SUVmax were calculated. All P values reported were 2-sided, and a P value < 0.05 was considered to indicate statistical significant.