Outcome of 18f-Fluorodeoxyglucose Positron Emission Tomography (PET)-Directed Resections in Differentiated Thyroid Carcinoma (DTC): Experience in a Developing Country

Abstract

The aim of this study was to assess the accuracy of 18F-fluorodeoxyglucose positron emission tomography (18-FDG-PET scan) in localizing the disease in differentiated thyroid carcinoma patients undergoing re-operations. This is a retrospective analysis of a prospectively maintained data (December 2007 to December 2016). The patients included had elevated serum thyroglobulin (Tg) levels and negative iodine uptake (TENIS) and planned for re-operation with one or more accessible site of metastasis detected on FDG-PET scan. Clinical details, FDG-PET/CT findings, operative findings, histology, pre-, and post-operative Tg levels were recorded. Thirty-two patients were included. The mean age of the patients was 46.8 ± 15.8 years (M:F = 1:1.6) and mean pre-operative Tg value was 247.6 ± 92.3 ng/ml. FDG-PET disclosed a total of 77 hot spots in these 32 patients, 56 of which were surgically explored and resections performed. Patient- and lesion-based positive predictive value (PPV) of FDG-PET in detecting recurrent/metastatic DTC lesions was 87.5 and 71%, respectively. Remaining cases had granulomatous or nonspecific inflammatory lesions. A total of 12.5% of recurrent DTC patients explored could achieve biochemical cure. All these had disease confined to neck. Remaining patients continued to have high serum Tg level, though it fell substantially in majority of patients. False positive scans are frequent in regions with high prevalence of inflammatory diseases. Hence, FDG-PET directed re-operations should be taken up judiciously.

Introduction

The follow-up of differentiated thyroid carcinoma (DTC) patients after treatment with surgery and radioactive-iodine utilizes serum thyroglobulin (Tg) and whole body radio-iodine (RAI) scan to detect recurrence. Tg serves as a very sensitive marker of recurrence, and when elevated, the recurrent site can usually be localized on RAI scan. But, in about 10–30% cases, one encounters a situation where despite elevated serum Tg, the RAI scan fails to show any lesion [1]. This condition of rising Tg level and negative RAI scan is commonly known by acronym of TENIS. 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) is now considered the investigation of choice to localize the recurrent/persistent DTC lesion [2–8]. The most recent meta-analysis on this subject has found the pooled sensitivity and specificity of FDG-PET in this situation to be 76.6% and 75.7%, respectively, and the one preceding this had reported both to be 84% [3, 4]. The addition of CT scan seems to enhance the diagnostic performance of PET further, although it was not statistically significant in the meta-analysis by Haslerud et al. [3]. However, one must keep it in mind that these results are mostly from studies that have used a combination of histology, cytology, and follow-up/imaging as gold standard. Uptake on FDG-PET is not specific for neoplastic process, and it is also observed in other inflammatory processes such as tuberculosis [9, 10]. This is particularly relevant in developing countries like ours where incidence of tuberculosis is very high, so can an uptake on FDG-PET be taken on face value? The aim of this study was to assess the accuracy of 18-FDG-PET scan in localizing the recurrent disease in DTC patients undergoing re-operations.

Material and Methods

This is a retrospective analysis of a prospectively maintained data. The criterion for inclusion in this study was patients having TENIS and planned for re-operation with one or more accessible site of metastasis amenable to resection. The patients included were operated between December 2007 and December 2016 for recurrent/persistent DTC. All patients had received at least one dose of RAI. Our protocol is to follow all the cases of DTC at 6 monthly intervals with serum Tg estimation, high-resolution neck ultrasonography (USG), and/or RAI scan. FDG-PET is performed when Tg is high and RAI scan does not localize the disease. All FDG-PET scans were performed after TSH stimulation with thyroid hormone withdrawal or recombinant thyroid stimulation hormone (rTSH). All patients underwent USG of the neck as per follow-up protocol, and those with suspicious findings underwent USG guided fine needle aspiration cytology (FNAC). Image guided biopsy was also performed from extra-cervical lesions wherever possible. Only those patients with distant metastases were explored who had disease in central neck. The intent was palliative to prevent future asphyxia. Standard surgical procedures were followed in each patient depending on the site of disease. Disease localized to a particular lymph node basin was resected using en-block lymph node dissection. All these patients received TSH suppressive dose of thyroxine in follow-up. Clinical details, pre-operative Tg level, FDG-PET/CT findings, operative findings, histology, and 6 months post-operative Tg levels were recorded.

Results

Thirty-two patients underwent re-operation for recurrent/persistent DTC in our department between the given time period. The mean age of the patients was 46.8 ± 15.8 years (M:F = 1:1.6). The median interval between the last surgery for DTC and PET directed surgery was 24.5 (11–204) months. Majority of these patients (56%) were operated at other centers, and staging information for most of them was not available. All but 3 patients had initial diagnosis of papillary thyroid carcinoma (PTC). The rest had follicular thyroid carcinoma (FTC) as referral diagnosis, though one of these did not have tissue or slides available for review and the re-operation histology was consistent with poorly differentiated thyroid carcinoma (PDTC). The mean dose of RAI administered prior to re-operation was 257.3 ± 182.3 mCi (50–800 mCi), and mean pre-operative Tg value was 247.6 ± 92.3 ng/ml. Table 1 summarizes the details of the patients. Pre-operative FNAC was successfully tried from 24 sites in 18 cases and was positive from 14 sites.

Table 1.

Summary of operative procedure and outcome

S No	Diagnosis	Age/sex	Pre-op. Tg	FDG uptake site	Sites explored	Histology$	Post-op Tg*
1	PTC	32/F	300.0	CC	CC	Positive	1.0
2	PTC	27/F	300.0	CC	CC	Positive	10.5
3	PTC	14/M	500.0	CC	CC	Negative^	90.0
4	PTC	55/F	149.0	CC	CC	Negative^	5.0
5	PTC	47/M	59.4	RL	RL	Positive	2.56
6	PTC	55/M	237.0	CC	CC	Negative^	300.0
7	PTC	67/F	190.0	CC	CC	Positive	1.0
8	PTC	69/F	300.0	RL	RL	Positive	2.85
9	PTC	23/F	300.0	RL	RL	Positive	27.0
10	PTC	49/F	341.3	RL	RL	Positive	300.0
11	PTC	47/M	78.0	CC	CC	Positive	31.0
12	FTC	35/F	343.0	Sternum	Sternum	Positive	10.0
13	PTC	54/M	248.0	RL, LL	RL, LL	Positive, granuloma	100.0
14	PTC	57/M	300.0	CC, LL	CC, LL	Negative, positive	11.5
15	PTC	47/M	174.0	CC, LL	CC, LL	Positive, positive	41.0
16	PTC	78/F	300.0	RB, RA	RB, RA	Other#, other#	300.0
17	PTC	62/F	300.0	CC, RL	CC, RL	Positive, positive	44.0
18	PTC	25/F	138.0	CC, RL, LL	CC, RL, LL	Positive, positive, positive	64.0
19	PTC	37/F	236.0	CC, RL, LL	CC, RL, LL	Positive, negative, positive	225.0
20	PTC	49/M	300.0	RL, LL, ML	RL, LL	Positive, negative	30.0
21	PTC	42/F	161.0	CC, LL, LA	CC, LL, LA	Positive, positive, granuloma	110.0
22	PTC	56/M	167.0	CC,RL, LA	CC, RL	Positive, negative	57.0
23	PDTC	32/F	300.0	CC, RL, Pharynx	CC, Pharynx	Positive, positive	300.0
24	PTC	56/M	228.0	CC, RL, LL	CC	Positive	0.2
25	PTC	30/F	214.0	CC, RL, ML	RL	Positive	100.0
26	PTC	60/M	300.0	RL, both lungs	RL	Positive	300.0
27	PTC	62/F	181.0	CC, LL, LA, ML	CC, LL, LA, ML	Negative, positive, negative, negative	0.99
28	PTC	32/F	300.0	CC, RA, LA, ML	RA, LA, Med	Positive, positive, positive	99.0
29	PTC	42/F	77.0	CC, RL, LL, R	RL, LL	Positive, negative	25.0
30	PTC	35/F	300.0	CC, RL, LL, RA, LA, ML	RL, LL, ML	Negative, positive, positive	300.0
31	PTC	46/F	300.0	Trachea, ML, LL, both lungs, left iliac bone	Trachea, LL, ML	Positive, positive, positive	300.0
32	FTC	76/M	300.0	RL, both lungs, ML	RL	Positive	220.0

^*Serum Thyroglobulin (Tg) level at 6-month follow-up (ng/ml)

^$Histology of individual explored sites: Positive confirmed DTC, Negative nonspecific inflammation/granuloma

^#Other occult breast cancer with axillary lymphadenopathy. This patient underwent breast conserving surgery (upper outer quadrantectomy and axillary clearance)

Twenty-one (65.6%) patients had FDG uptake confined to neck of which 7 had uptake confined to central compartment, whereas 4 (12.5%) in addition had mediastinal uptake and 6 (18.6%) distant uptake (with or without loco-regional uptake). All metastatic cases had serum Tg > 300 ng/ml. Twenty-one patients had central compartment disease, with or without lateral neck involvement. Of these, two patients were missed on USG (sensitivity 90.5%). Overall, 30 had disease focus in neck with or without mediastinal or distant disease, and neck lesions could be detected in all but two aforementioned patients (sensitivity 93%). FDG-PET disclosed a total of 77 hot spots (central compartment 21, lateral neck 29, mediastinum 7, axilla 9, lung 6, and 1 each in sternum, trachea and breast, pharynx and iliac bone) in 32 patients. Fifty-six out of these 77 sites were explored, and resections were performed. Seventy-one percent (n = 40) of those excised were proven to be metastatic differentiated thyroid carcinoma (DTC) lesions on histology. Overall, 75% operated sites contained malignant lesions including one patient who in addition to recurrent DTC had occult breast cancer with metastases to axillary lymph node. The histology of other 14 sites was reported as granulomatous lesion and other non-specified chronic inflammatory lesions in 5 and 20% cases, respectively. At least one DTC lesion was confirmed on histology in 29 out of 32 patients. No neoplastic lesion could be detected in 3 patients, one of these had a histology report of granuloma and the remaining two were described as having nonspecific inflammation. All three of these patients had uptake in CC and none of these patients underwent FNAC pre-operatively. Follow-up Tg value fell substantially in two of these (case no. 3 and 4, Table 1) but continued to be high in another (case no. 6).

Mean duration of follow-up after re-operation was 35.8 ± 26.0 months. Patient and lesion-wise positive predictive value of PET was 87.5% and 71.4%, respectively. Patients were followed with serial Tg measurements. Mean post-surgery (6 months) Tg levels was 35.8 ± 26.0 which was significantly lower (p < 0.001) than the mean pre-operative Tg level. Twelve percent of all (n = 4) patients achieved biochemical cure (stimulated serum Tg < 1 ng/ml) at 6 months post-surgery. Three of these had recurrence confined to neck and pre-operative Tg value were 300, 190, and 228 ng/ml, respectively, whereas one patient (Tg = 181 ng/dL) had FDG uptake outside the neck but only lateral compartment of neck had histologically confirmed recurrent DTC; remaining sites showed nonspecific inflammations. Pre-operative Tg did not correlate well with cure. Nine (n = 3), 21.8 (n = 7), 15.6 (n = 5), and 53.1% (n = 17) patients had pre-op Tg value of < 100, 100–200, 200–300, and > 300 ng/ml, respectively, and 0, 28.6 (2/7), 20 (1/5), and 5.9% (1/17) of these achieve biochemical cure. After 12 months of follow-up, Tg showed rising trend in most patients, and two of these were started on Sorafenib for progressive symptomatic disease.

Discussion

Our study shows that positive predictive value (PPV) of FDG-PET in detecting recurrent/metastatic DTC lesions was 71%, though patient-based PPV was 87.5%. Twenty-five percent of FDG avid lesions were non-neoplastic, and only 12.5% of surgically explored patients could achieve cure. Patients whose pre-operative Tg values were less than 200 ng/ml and who had disease confined to neck are more likely to achieve cure.

Three meta-analyses on the subject have been published so far, and all three of them have found FDG-PET/FDG-PET-CT to be a useful and sensitive diagnostic method for localization of recurrences in patients with TENIS [3, 4, 8]. The sensitivity in these meta-analyses ranged from 79.4 to 93% and specificity ranged from 79.4 to 84.7%. However, two limitations were highlighted from these meta-analyses, first as is true for all radionuclide-based imaging, there is difference between lesion-based and patient-based sensitivity and specificity, i.e., while some lesions are picked up but others may be missed in same patient and some of the localized lesions may not be neoplastic. Second limitation is that all studies did not confirm the findings of positive scan with tissue diagnosis. In fact, Caetano et al. found in their meta-analysis that only 5% of the 18 studies they included had exclusively used histology of the excised lesion as the gold standard [8].

The 87.5% PPV in our study is comparable with others. But we observed higher incidence of inflammatory lesions. This finding of our study points towards an important limitation that has been known but probably not emphasized enough. FDG-PET only identifies sites with high metabolic activity, and hence, sites with inflammation (infection/granulomatous inflammation/nonspecific inflammation) show an increased uptake and FDG-PET cannot always distinguish these sites reliably from malignant sites. This issue assumes a great importance in limiting the diagnostic capability of FDG-PET in developing countries where the prevalence of granulomatous and nonspecific inflammatory lymphadenopathy is high [9, 10]. Although we attempted to confirm the diagnosis of DTC recurrence by FNAC of all the FDG uptake sites, we intended to explore, but some of these sites have produced either an inconclusive report or difficult access, problems which are common with small lesions and can be attributed to a number of factors. The large difference between patient-wise and lesion-wise PPV in our study also underscores the importance of reporting lesion-wise sensitivity. Reporting just patient-wise sensitivity ignores the false positive lesions in a patient, and this not only overestimates the sensitivity but also leads to unnecessary surgeries/extra-surgical explorations.

Some of our patients with disease confined to neck (case no 6, 10, 21) did not show any appreciable decline in Tg level at 6-month follow-up, a finding suggestive of presence of a few lesions that remain undetected (false negative) on FDG-PET scan. This could be explained by “Flip-Flop-phenomenon” [6, 7]. Well-differentiated thyroid cancer cells take up iodine due to expression of sodium iodine symporter (NIS) but not glucose, while the less-differentiated cancer cells express receptors for glucose uptake but not NIS and hence cannot concentrate iodine and probably some lie in between which are missed on scans but continue to secrete Tg. FDG-PET positivity and number of positive lesion both have been inversely correlated with survival in these patients in many studies [11, 12]. Although 12% of our patients who underwent surgery with a curative intent achieved biochemical cure (Table 1), it did not last long, and majority continued to have persistent thyroglobulinemia. We could not analyze the survival because of the limited duration of follow-up.

The most pertinent question is the value of FDG-PET in TENIS particularly in a resource-constrained setting. Patients having very low Tg are likely to be missed on FDG-PET scanning whereas the patients likely to achieve cure would have disease limited to neck which could be easily picked up on USG. FDG-PET scores over USG in those likely to have distant metastases. Furthermore, it facilitates monitoring of systemic metastases and could help in prognostication [11–13]. As of now, we do not know for sure the cut-off Tg concentration where USG should exclusively be used and FDG-PET could be avoided. Our protocol is still evolving. But based on the observations of current and previous studies, we could propose a protocol for a resource-constrained setting. We suggest that in setting of TENIS, if Tg levels are low, then high-resolution neck USG should be employed first for detection of loco-regional disease. A CT thorax may be performed next if USG fails to pick up neck disease. FDG-PET should be reserved for those with non-localized disease. Patients with high Tg level, if planned for surgery, should have FDG-PET scanning to optimize the extent and intent of surgery (curative or palliative). The cut-off value for high and low Tg could be based on institutional data.

The strength of this study is that we analyzed the histology of each site of uptake on FDG-PET scan separately and then calculated the patient-wise and lesion-wise PPV in cases of TENIS. The limitation is its retrospective nature.

Conclusion

Our study revealed that false positive scans are frequent in regions with high prevalence of inflammatory diseases, and FDG-PET-directed re-operations may not contribute to long-term disease control in patients with TENIS.

Abbreviations

PTC

Papillary thyroid carcinoma

FTC

Follicular thyroid carcinoma

PDTC

Poorly differentiated thyroid carcinoma

CC

Central compartment

RL

Right lateral compartment

LL

Left lateral compartment

ML

Mediastinal lymph nodes

RB

Right breast

RA

Right axilla

LA

Left axilla

Author Contributions

Chandan Kumar Jha, Anjali Mishra - literature search, figures, study, design, data collection, data analysis, data interpretation, writing.

Prasanta Kumar Pradhan, Sanjay Gambhir, Gaurav Agarwal, Gyan Chand, Amit Agarwal, Saroj Kanta Mishra - revision and editing of manuscript.

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.

Ethical Approval

This retrospective study does not involve any experiment on human participation. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent

Informed written consent was obtained from the patient.

This is to declare that all authors have contributed to the study. No part of the manuscript has been sent for consideration elsewhere or published in any International or National journal.

The authors clearly certify that there is no aspect of plagiarism. In case of any dispute, the authors will be held fully responsible for the statement disclosed in the cover letter. The authors are also aware of the copyright rules and also declare that they will not reproduce any published text without due permission from the journal.