Abstract
Radioactive therapy with 131I (RAI) is commonly used during the management of patients with differentiated thyroid cancer (DTC). The aim of this study was to determine the clinical significance of discordant findings between pre-RAI whole body scan (WBS) with 123I and post-RAI WBS in the management of DTC. We retrospectively evaluated 342 individuals between 2002 and 2008 who had a diagnosis of DTC and underwent RAI. All had WBS one day before RAI and WBS one week after RAI. Patients were divided into 3 groups: 1) RAI-naive subjects without known distant metastatic disease (M1); 2) patients with history of prior RAI and persistent disease (except M1); and 3) patients with known M1. In Group 1 (n=311), 7% of patients (n=22) had discordant scans, but in only 4 of these cases did this represent true disease (3 unsuspected lung and 1 mediastinal node metastasis). In the remaining 18 patients, discordant findings corresponded to physiologic or other benign causes. In group 2 (n=23), 7 subjects (30%) had discordant findings and all of the discrepant sites consisted of loco-regional nodal disease in the neck/upper mediastinum (n=6) and M1 in lung (n=1). In group 3 (n=8), 5 patients (62%) showed discordant uptake in lung and bone which corresponded to the locations of known M1. A total of 12 patients with iodine-avid M1 were identified on post-RAI WBS (3.5% of entire cohort). Pre-RAI WBS was only concordant in 3 of these cases (25%). In conclusion, the significance of pre and post-RAI WBS is highly influenced by the clinical setting. Unsuspected distant metastatic disease is infrequent in RAI-naive patients without known M1, where most discordant findings are usually due to benign explanations, and represent false positive findings in this group. In contrast, in patients with history of previous RAI or known M1, discordant results likely correspond to true disease. In our study, pre-RAI scans showed a low yield to detect iodine-avid distant metastatic disease when compared to post-RAI scans.
Keywords: Pre-therapy 123I scan, post-therapy 131I scan, discordance, thyroid cancer
Introduction
Thyroid cancer is the most common malignancy of the endocrine glands. Differentiated thyroid cancers (DTC), the most frequent histological types, generally are associated with a good prognosis [1]. The treatment of DTC consists of total or partial thyroidectomy and in many cases, radioactive therapy with 131I (RAI) following surgery [2]. In many centers, pre- and post-RAI whole body scans (WBS) are performed routinely on patients undergoing RAI for DTC. After surgery and prior to RAI delivery a pre-RAI WBS is performed with either 123I or 131I to identify remnant thyroid tissue and residual local or distant metastatic thyroid cancer. Post-RAI WBS is primarily done to identify metastatic disease not seen by the pre-RAI scan.
A previous study using 131I for pre-planning purposes found that post-RAI scans can detect additional metastatic thyroid disease in 10% of patients compared to pre-RAI with 131I scans. This newly discovered disease affected clinical management in up to 15% of patients receiving RAI [3].
Use of 123I has several advantages over 131I, including a shorter half-life (13.3 hours vs. 8 days), lack of beta emission, and better image quality. However, 123I is more expensive and not widely available [4-7]. Studies comparing pre-RAI with 123I and post-RAI scans have reported diagnostic concordances ranging from 72% to 94% [8-10]. Nevertheless, previous studies have provided limited information about the clinical implications of these discordant scans [8,11].
The aim of the present study was to determine the clinical significance of discordant findings between pre-RAI WBS with 123I and post-RAI 131I WBS in patients with DTC.
Materials and methods
Study population
We retrospectively reviewed our database searching for patients of all ages (including 12 patients under 21 years) who received RAI at our institution using the following inclusion criteria: Individuals who underwent near-total or total thyroidectomy and received RAI for the management of DTC, and who underwent both a pre-RAI 123I and post-RAI 131I WBS performed at the Johns Hopkins Hospital between 2002 and 2008 (follow-up data was reviewed up to March of 2012). Patients who had surgery at outside institutions were considered as long as the tumor pathology was submitted for review at our institution. Only patients with complete images, laboratory data and pathology were included. Patients with prior RAI were also included.
The following patients were excluded from the study: Patients with thyroid lymphoma, poorly differentiated and/or medullary thyroid cancer. Individuals not treated at our institution, or those who underwent dosimetry studies or pre-RAI scans using 131I prior to RAI delivery were also excluded. Within the group with discordant scans patients with no follow-up data available were also excluded.
Patients who met our inclusion/exclusion criteria were subsequently divided into 3 main groups: 1) Patients without history of prior RAI (naive) therapy or known distant metastatic disease; 2) Patients with history of prior-RAI and evidence of persistence disease (except distant metastatic disease); and 3) patients who received RAI for treatment of known distant metastatic disease.
Pathology
Differentiated thyroid tumors were divided into 4 categories: 1) Papillary thyroid cancer (PTC) and its follicular variant, 2) Follicular thyroid cancer (FTC), 3) Hurthle cell carcinoma (HCC) and 4) Aggressive variants of PTC including tall cell, columnar cell, and diffuse sclerosing histologies.
Staging
Patients were classified into 5 groups (I, II, III, IVa and IVc) according to the staging system of the American Joint Committee on Cancer [12]. Patients with distant metastatic disease were classified as M1 instead of IVc in order to include them in the same group as the patients under the age of 45 with distant metastatic disease, who would otherwise be classified as stage II. For this study, final staging was made following the pre-RAI WBS and before delivery of RAI.
Patient preparation and procedure protocol
Patients’ TSH levels were stimulated by one of two methods: 1) thyroid hormone withdrawal, in which thyroxine (T4) was withheld for at least 4 weeks or triiodothyronine (T3) for 10 days before the pre-RAI WBS, or 2) use of recombinant human thyroid-stimulating hormone (rh-TSH, Thyrogen®, Genzyme, Cambridge, MA), 0.9 mg intramuscularly daily for two days. The pre-RAI 123I WBS was acquired one day after the second activity. Radioactive therapy with 131I occurred one day after the 123I WBS. Post-therapy scans were performed one week after RAI for both protocols (Figure 1) [13]. The method employed was based on the referring physician and patient’s preference choice, except in patients with evidence of metastatic thyroid cancer, in which case thyroid hormone with-drawal was preferred.
Figure 1.
Preparation methods and procedure protocols during therapy week (RAI=radioactive therapy with 131I, WBS=whole body scan).
Independent of the preparation method used, all patients underwent 123I WBS one day before RAI and 131I WBS one week after RAI as detailed in Figure 1. Serum for assays of TSH (usually on or the day before 123I WBS) and thyroglobulin (TG) were obtained during the procedure week. TG was taken together with serum TSH if patient has been withdrawn or 48-72 hrs after second rh-TSH injection. We used 3 different TG antibody assays during the study period, as following: Nichols Institute Diagnostics Advantage (≤0.3 IU/mL), Siemens Immulite (Quest Diagnostics) (≤20 IU/mL), and Beckman Coulter Access (≤2.2 IU/mL).
Acquisition protocols for pre-therapy 123I whole body scan
Patients received 1.5 mCi of 123I as a sodium iodide capsule orally if withdrawn from thyroid hormone, or 2.0 mCi if stimulated with rh-TSH. Prior to dosing, the capsules were measured using a thyroid phantom at a distance of 6 cm for 1 minute and the counts for each capsule recorded in a spreadsheet to estimate the 24 hour neck uptake. Imaging was performed on a Millennium VG Hawkeye system (GE Healthcare) equipped with SPECT/CT capability using a medium energy, parallel-hole collimator, with a 20% energy window centered at 159 keV.
At approximately 24 hours after 123I intake, an anterior neck image over the thyroid was obtained at 6 cm for 10 minutes before the WBS acquisition. Time and counts were recorded within a region of interest over the thyroid in order to calculate the 24 hour % neck uptake. WBS imaging was performed in the anterior and posterior projections at 5 cm/min.
Radioactive iodine therapy with 131I (RAI)
All patients were treated as outpatients, and the final 131I administered activity was estimated the day before RAI delivery based on staging and the pre-RAI WBS findings including the percentage of 123I uptake in the neck. There is no accepted consensus as to how much 131I activity a patient should receive as part of the management of thyroid cancer, but in general at our institution low-risk patients receive anywhere between 25-75 mCi, those with nodal disease 75-125 mCi, and subjects with history of prior RAI and persistent TG levels as well as those with higher staging usually receive 125 mCi or more.
Acquisition protocols for post-treatment 131I whole body scan
Approximately 7 days after RAI delivery, imaging was performed using the same gamma camera system with a high energy parallel hole collimator and a 20% energy window centered at 364 keV. Anterior and posterior WBS images were obtained at 5 cm/min. An anterior and posterior image of the neck was also obtained for 75K.
Comparison between pre- and post-RAI whole body scan
Both sets of original images were re-uploaded and interactively reviewed on a Xeleris workstation (GE Healthcare-Applications) by a board certified nuclear medicine physician (PB). The scans were considered to be completely discordant if the post-RAI scan demonstrated at least one new area of increased radiotracer activity at sites where no uptake was visualized on the corresponding pre-RAI scan.
When the post-RAI scan showed one or more additional areas of increased radiotracer activity at sites that had already demonstrated radiotracer uptake on the corresponding pre-RAI WBS they were considered partially discordant and not included in the discordant group. The additional information obtained from the ones with partial discordance was unlikely to have an impact on clinical staging or patient management since the findings were already available to the clinician from the pre-therapy scan prior to RAI delivery.
Completely discordant findings were analyzed on a site and patient-basis. Any extra uptake seen on the post-RAI WBS within the nasopharynx, salivary glands, gastrointestinal and/or urinary system was considered a result of normal radioiodine metabolism, secretion and/or excretion, and thus not considered discordant.
In patients with evidence of complete discordance between scans, further attempts to assess the significance and/or clinical implication(s) of the discordant findings were made by reviewing and analyzing further information including clinic notes, laboratory, operative and pathology reports, and cross-sectional multimodality images, according to the following criteria:
1) Nodal or metastatic disease was considered likely at the site of complete discordance if there was biopsy-proven disease, and/or ultrasonographic, CT, MRI, SPECT/CT, or PET/CT evidence of disease, and/or abnormal uptake in same location on subsequent post-RAI examinations.
2) Disease was unlikely or negative at the site of discordance if the above-mentioned evaluations were unrevealing and/or stimulated TG levels were below 2 ug/L at least 8 months post-RAI on follow-up [14,15].
No reverse discordance (findings seen on pre-RAI and missed on post-RAI WBS) was observed in this study.
Statistical analysis
Statistical analyses were performed using SPSS (version 21.0). Continuous variables were presented as mean±SD and median. An independent-measures t-test and Mann-Whitney U test (for non-parametric distributions) were used to assess differences between 2-tail subgroups. Categorical variables were compared between groups using chi-square (x2) tests and are presented as percentages. A P value <0.05 was considered statistically significant for all calculations.
Results
A total of 342 patients were included in the study. Group 1 (RAI-naive) was made up of 311 patients; group 2 (history of prior RAI) of 23 individuals; and group 3 (known distant metastatic disease) consisted of 8 patients.
Concordance/discordance of scans in RAI-naive patients without known distant metastatic disease prior to RAI delivery (group 1)
Complete discordance was observed in 7% (22/311) of patients within this group. Except for certain racial differences and a higher 131I activity received by patients with discordant scans (85±21 vs. 97±18 mCi, P=0.007), baseline characteristics were similar between individuals with concordant and discordant pre/post WBS in this group (Tables 1 and 2).
Table 1.
Baseline characteristics of patients with differentiated thyroid cancer
| Characteristics | No prior RAI or evidence of distant metastatic disease (Group 1) | History of prior RAI and/or known distant metastatic disease (Group 2/Group 3) | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| Concordant scans (n=289) | Discordant scans (n=22) | P | Concordant scans (n=19) | Discordant scans (n=12) | P | |
| Age (yrs) | 47±15 | 51±17 | 0.2 | 44±20 | 52±14 | 0.2 |
| Tumor size (cm) | 2.1±1.5 | 2.6±2.7 | 0.2 | 3.6±2.3 | 3.6±1.8 | 0.9 |
| 123I activity (mCi) | 1.6±0.2 | 1.7±0.3 | 0.3 | 1.5±0.2 | 1.6±0.1 | 0.3 |
| 123I neck uptake (%) | 1.6±2.8 | 0.9±0.9 | 0.2 | 0.36±0.58 | 0.26±0.52 | 0.6 |
| Thyroglobulin (ug/L), median | 3.1 | 2.2 | 0.4 | 10 | 126 | 0.007 |
| TSH (mIU/L) | 103±61 | 120±69 | 0.2 | 110±88 | 94±48 | 0.6 |
| RAI activity (mCi) | 85±21 | 97±18 | 0.007 | 130±52 | 197±20 | <0.0001 |
Values are mean±SD unless otherwise indicated. RAI=radioactive therapy with 131I.
Table 2.
Baseline characteristics of patients with thyroid cancer (continuation)
| Characteristics | No prior RAI or evidence of distant metastatic disease (Group 1) | History of prior RAI and/or known distant metastatic disease (Group 2/Group 3) | |||||
|---|---|---|---|---|---|---|---|
|
| |||||||
| Concordant scans (n=289) | Discordant scans (n=22) | P | Concordant scans (n=19) | Discordant scans (n=12) | P | ||
| Gender, n (%) | Female | 188 (65) | 13 (59) | 0.6 | 11 (58) | 5 (42) | 0.4 |
| Race, n (%) | White | 229 (79) | 13 (59) | 0.04 | 16 (84) | 7 (59) | 0.2 |
| Black | 37 (13) | 4 (18) | 2 (11) | 4 (33) | |||
| Other | 23 (8) | 5 (23) | 1 (5) | 1 (8) | |||
| Pathology, n (%) | PTC | 241 (83) | 19 (86) | 0.6 | 11 (58) | 9 (75) | 0.4 |
| FTC | 14 (5) | 2 (9) | 3 (16) | 0 | |||
| HCC | 15 (5) | 0 | 1 (5) | 0 | |||
| Other* | 19 (7) | 1 (5) | 4 (21) | 3 (25) | |||
| Staging, n (%) | Stage I | 194 (67) | 15 (68) | 0.7 | 9 (47) | 3 (25) | 0.07 |
| Stage II | 25 (9) | 1 (5) | 0 | 0 | |||
| Stage III | 57 (20) | 4 (18) | 6 (32) | 1 (8) | |||
| Stage IVa | 13 (4) | 2 (9) | 1 (5) | 3 (25) | |||
| M1 | 0 | 0 | 3 (16) | 5 (42) | |||
| Preparation method, n (%) | Hormone withdrawal | 187 (65) | 11 (50) | 0.2 | 18 (95) | 12 (100) | 0.4 |
| rh-TSH | 102 (35) | 11 (50) | 1 (5) | 0 | |||
PTC=papillary thyroid cancer, FTC=follicular thyroid cancer, HCC=Hurthle cell carcinoma, Other PTC variants, rh-TSH=recombinant human TSH, RAI=radioactive therapy with 131I.
The discordant sites were in order of frequency in the extra-thyroidal neck (n=9), lungs (n=7), mediastinum (n=4), and abdomen (n=2). Clinical follow-up information was available for all discordant patients with a median follow-up of 2.9 years (Table 3). These data demonstrated that in only 4 cases (18% of discordant cases and 1.3% of the total patients in group 1), the activity that localized at the discordant sites corresponded to persistent disease, including 3 patients with unsuspected metastatic lung disease shown on a chest CT and 1 patient who was diagnosed with metastatic PTC to a right level VI lymph node by fine needle biopsy during follow-up.
Table 3.
Baseline and follow-up evaluations of patients with completely discordant scans and without a history of prior RAI or known distant metastatic disease at the time of first study (Group 1)
| ID | Age (yrs) | Stage | 131I (mCi) | Location of discordance | Examination modality for follow-up | Findings | TG* (ug/L) | f/u time (yrs) | Evidence of disease on follow-up |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 39 | I | 103 | Bilateral lung | SPECT/CT, CT | No abnormality | U | 2.1 | Negative |
| 2 | 51 | I | 108 | Liver/hepatic flexure | US, CT, SPECT/CT | 5 mm liver cyst | U | 1.3 | Negative |
| 3 | 34 | I | 76 | Bilateral neck | U/S, CT, FNA | Reactive LN | U | 1.2 | Negative |
| 4 | 44 | I | 101 | Left upper neck region | SPECT/CT, U/S | Benign-appearing left submandibular LN | U | 6.9 | Negative |
| 5 | 18 | I | 75 | Anterior mediastinum | SPECT/CT | Thymic tissue uptake | U | 3.4 | Negative |
| 6 | 61 | I | 77 | Right neck | U/S | Benign-appearing LN | U | 7.9 | Negative |
| 7 | 77 | I | 78 | Anterior mediastinum | SPECT/CT | Anterior mediastinal density, likely thymus | U | 1.8 | Negative |
| 8 | 46 | I | 107 | Left upper abdomen | Stimulated TG | Felt to be benign | U | 4.6 | Negative |
| 9 | 50 | I | 102 | Lower neck | U/S, FNA | Negative, left level IV LN | U | 3.8 | Negative |
| 10 | 40 | I | 128 | Right posterior neck | CT, U/S | benign-appearing LN | U | 5.9 | Negative |
| 11 | 29 | I | 103 | Anterior mediastinum | Stimulated TG | Felt to be thymus | U | 1.7 | Negative |
| 12 | 41 | I | 100 | Anterior mediastinum | Stimulated TG | Felt to be thymus | U | 7.2 | Negative |
| 13 | 66 | I | 103 | Lower neck | Stimulated TG | U | 2.5 | Negative | |
| 14 | 35 | I | 77 | Left lung | SPECT/CT | No abnormality | U | 2.1 | Negative |
| 15 | 60 | I | 100 | Right lateral thorax | SPECT/CT | No abnormality | U | 4 | Negative |
| 16 | 45 | II | 74 | Right posterior thorax | SPECT/CT, CT | No abnormality | 0.4† | 4.5 | Negative |
| 17 | 84 | III | 75 | Bilateral lung | CT, SPECT/CT, 131I WBS | Multiple lung nodules bilaterally | 187.7 | 2 | Positive |
| 18 | 71 | III | 146 | Lower neck | CT, U/S, FNA | Metastatic PTC in right level VI LN | 1312 | 0.8 | Positive |
| 19 | 46 | III | 103 | Lower neck | U/S | benign-appearing LN | 0.3† | 6.8 | Negative |
| 20 | 85 | III | 102 | Lower neck | Stimulated TG | 2.5† | 1.5 | Likely negative | |
| 21 | 53 | IVA | 103 | Bilateral lung | CT | Multiple lung nodules bilaterally | 60.9 | 4.5 | Positive |
| 22 | 54 | IVA | 101 | Right lower lung | CT, excision biopsy | Single lung nodule. Biopsy metastatic PTC | 0.4 | 1.8 | Positive |
=follow-up stimulated serum thyroglobulin (TG) levels at least 8 months after RAI.
=presence of anti-TG antibodies.
U=undetectable stimulated TG levels. U/S=ultrasound. FNA=fine needle aspiration. LN=lymph nodes. 131I WBS=subsequent post-RAI scan. PTC=papillary thyroid cancer. f/u=follow-up.
No evidence of persistent disease was documented in the remaining 18 cases (82% of discordant scans) on subsequent evaluations (Table 3). In these cases the discordant findings were considered to correspond to physiologic uptake (Figure 2), benign conditions or spurious findings (Figure 3), as following:
Figure 2.
Patient (ID #5) shows a large area of increased radiotracer uptake in the mediastinal region on post-RAI scan (B) not seen on pre-RAI study (A). SPECT/CT (C, D) demonstrates radiotracer activity localizing to the anterior mediastinum, corresponding to thymic tissue on CT. Stimulated Thyroglobulin remained undetectable 3.4 years after initial RAI delivery consistent with physiologic 131I uptake in the thymus.
Figure 3.
Patient (ID #1) has 3 foci of 131I increased uptake in the right and left lower lungs in the post-RAI scan (B and D) with no 123I uptake on pre-RAI study (A) and no CT abnormality (C). Stimulated Thyroglobulin remained undetectable 2.1 years after initial therapy.
Physiologic uptake: thymus (ID #5, 7, 11, 12) and bowel (ID #8). TG levels were undetectable on follow-up.
Benign uptake/conditions: liver cyst (ID #2), reactive lymph nodes on biopsy (ID #3, 9) and benign-appearing lymph node findings on ultrasound or CT (ID #4, 6, 10, 19). TG levels were undetectable on follow-up, except for case #19.
Spurious uptake/findings: bilateral lung (ID #1), left lung (ID #14), right lateral thorax (ID #15), and right posterior thorax (ID #16). No abnormality was detected on SPECT/CT and CT scans. TG levels were undetectable on follow-up, except for case #16.
There were 2 patients with lower neck 131I uptake wherein follow-up TG became undetectable in one case (ID #13) and persisted only mildly detectable on the other one (ID #20), both cases were considered negative for thyroid cancer.
The 4 patients with confirmed disease at the discordant sites had a significantly higher presenting stage of disease (III and IVa vs. I, II, III; P=0.002) and stimulated TG (median 270 vs. 1.3 ug/L; P=0.002) prior to RAI delivery compared to those with no evidence of disease on follow-up evaluations.
Concordance/discordance of scans in patients with history of prior RAI therapy and evidence of persistent disease, except distant metastasis (group 2)
In this group, 23 patients had history of prior RAI therapy and evidence of persistent disease (except M1) by elevated stimulated TG and/or cross-sectional imaging. Discordant findings were seen in 7 (30% of patients with prior RAI) of these patients, and consisted of new uptake in the neck or upper mediastinum that represented loco-regional nodal disease in the neck (n=5) and a left mediastinal paratracheal node (n=1). The last patient (stage IVa) showed discordant uptake in the posterior thorax, and cross-sectional imaging demonstrated unsuspected distant metastasis at the right transverse process of 10th thoracic vertebra (Table 4).
Table 4.
Baseline and follow-up evaluations of patients with completely discordant scans and history of prior RAI (Group 2) or known distant metastatic disease at the time of first study (Group 3)
| ID | Age (yrs) | Stage | 131I (mCi) | Location of discordance | Examination modality for follow-up | Findings | TG* (ug/L) | f/u time (yrs) | Evidence of disease on follow-up |
|---|---|---|---|---|---|---|---|---|---|
| 23 | 37 | I | 206 | Left neck | U/S | 3 suspicious left neck LN | 5.8 | 9.2 | Positive |
| 24 | 43 | I | 201 | Left neck | PET/CT, excision biopsy | Metastatic PTC involving left parapharyngeal LN | 45.3 | 0.9 | Positive |
| 25 | 33 | I | 203 | Left neck | U/S | Suspicious left level IIB LN | 44.9 | 0.8 | Positive |
| 26 | 66 | III | 209 | Right lower neck | SPECT/CT, U/S, CT | Suspicious left level IV LN | 1274 | 3.8 | Positive |
| 27 | 48 | IVA | 206 | Upper mediastinum | SPECT/CT | Left paratracheal LN | 12.5 | 2.9 | Positive |
| 28 | 68 | IVA | 205 | Lower neck/upper mediastinum | U/S, CT | Cluster of 3 suspicious level IV LN | 26.3 | 2.7 | Positive |
| 29 | 59 | IVA | 206 | Posterior thorax | SPECT/CT, CT, PET/CT | Expansile lytic metastasis at right T10 transverse process | 305.6 | 4.2 | Positive |
| 30 | 69 | M1 | 209 | Right occipital bone | MRI, PET, 131I WBS, bone scan | Cystic rim enhancing mass in right occipital bone on MRI | 1266 | 4.8 | Positive |
| 31 | 49 | M1 | 205 | Left femur | Plain films, PET/CT, excision biopsy | Metastatic PTC to femur | 216.2 | 4.9 | Positive |
| 32 | 68 | M1 | 157 | Bilateral lung | SPECT/CT, CT | Bilateral pulmonary nodules | 126.5 | 0.9 | Positive |
| 33 | 40 | M1 | 152 | Bilateral lung | CT, 131I WBS | Multiple lung nodules bilaterally | 596.1 | 8 | Positive |
| 34 | 51 | M1 | 203 | Posterior chest | CT, PET/CT, biopsy | pleura-based mass in posterior thorax | 361.2 | 3.4 | Positive |
=follow-up stimulated serum thyroglobulin levels at least 8 months after RAI.
=Individuals without history of prior RAI.
M1=Subjects with known distant metastatic disease. U/S=ultrasound. LN=lymph nodes. 131I WBS=subsequent post-RAI scan. PTC=papillary thyroid cancer. T10=thoracic vertebrae 10. f/u=follow-up. Red text indicates a case where M1 was unknown by referring physician.
Patients with discordant findings (n=7) received a significantly greater 131I activity (205±3 vs. 122±48 mCi; P<0.0001) and had higher stimulated TG levels (median 102 vs. 5.7 ug/L; P=0.003) prior to repeat RAI than individuals without discordance (n=16).
Concordance/discordance of scans in patients with known distant metastatic disease (group 3)
A total of 8 patients had known M1 in the lung (n=4), bone (n=3) and pleura (n=1) prior to RAI, but only 3 of them were identified by the pre-therapy 123I scan (2 lung and 1 bone). The remaining 5 patients (62.5% of known M1) showed discordant uptake in the bilateral lung (n=2) (Figure 4), occipital bone, left femur and posterior thorax (pleura-based mass), which corresponded to the locations of known M1 (Table 4). There was no significant difference in the 131I administered activity to M1 patients with and without discordant scans (172±59 vs. 185±28 mCi, P=0.7).
Figure 4.
Patient (ID #33) has evidence of diffuse lung uptake on post-RAI scan (B) not seen on pre-RAI study (A). Chest CT one month prior to RAI delivery showed scatter lung nodules suspicious for metastasis (C). Chest CT seven years following therapy revealed innumerable pulmonary nodules consistent with widespread metastatic lung disease (D).
Detection of Iodine-avid distant metastatic disease by pre and post-RAI scans
A total of 12 patients with M1 were identified by post-RAI scan in this cohort (3.5%). This includes 8 patients of group 3 who were known M1 prior to RAI delivery, 3 new patients discovered during group 1 evaluations, and one additional patient unrevealed by post-therapy scan of group 2. Pre-RAI WBS only detected 3 (25%) of these M1 findings.
Discussion
The main findings of this study are the following: 1) the detection likelihood of loco-regional or distant metastatic disease by a post-RAI scan not visualized by a pre-RAI study is greatly influenced by the presenting stage of disease. It can be as low as 1.3% in RAI-naive patients without known M1 on presentation, 30% in patients with history of prior RAI delivery and evidence of persistent disease (except M1), and as high as 63% in patients with known M1; 2) the majority (82%) of the additional findings observed on post-RAI WBS performed with higher activities in RAI-naive patients without known M1 were physiologic or otherwise spurious findings; and 3) pre-RAI WBS only detected 25% of all iodine-avid M1 lesions subsequently identified on post-RAI scans.
The frequency and clinical significance of discordant scans is to a great extent affected by the clinical scenario. In RAI-naive patients without clinical evidence of M1 (91% of studied patients) the pre- and post-RAI WBS discordance was 7%. However, only 18% of these discordant cases (1.3% of total studied patients within this population) resulted in loco-regional (n=1) or distant metastatic disease (n=3). The disease in these four patients had not been suspected by the referring physician, but was confirmed by following examinations. As expected, those patients with confirmed disease on follow-up had a significantly higher presenting disease stage and stimulated TG prior to RAI delivery compared to those with no evidence of disease on subsequent evaluations.
In contrast, 82% of discordant scans within this group were considered negative for disease after further evaluation with different cross-sectional imaging modalities, stimulated TG, and/or tissue biopsy, which revealed no pathologic abnormalities. In this regard, there are well known physiologic as well as pathologic states associated with radioiodine uptake not necessarily related to DTC. Inflammatory lung tissue and bronchiectasis can take up radioiodine, which leads to false positive findings on post-RAI WBS [16-18]. The thymus is also known for concentrating iodide, similar to the salivary glands, stomach, kidneys, breasts, sweat glands, choroid plexus and hair follicles, which may be falsely interpreted as nodal disease in the mediastinum on planar images [16,17,19-21]. Likewise, uptake in the esophageal region due to swallowed saliva and/or anatomical abnormalities (strictures, scarring or hiatal hernias) can be confused with disease [17,22]. This demonstrates that most often completely discordant findings are due to physiologic or other benign causes in RAI-naive patients without known M1.
Discordance increased to 30% in patients with a history of previous RAI and evidence of persistent disease found by different tests (except M1) before RAI delivery. In our series, all discordant findings represented loco-regional disease in the neck and upper mediastinum, and M1 in one case (T10 transverse process), indicating an absence of false positive findings within this higher risk group.
Discordance between pre- and post-RAI WBS peaked to 63% in patients with known M1. In this subgroup, discordant sites on post-RAI scans corresponded to the locations of known M1 missed by pre-RAI WBS. Moreover, pre-RAI WBS only detected 25% of all M1 patients identified by post-RAI scans in all studied patients. These findings point out to a significant limitation of this pre-RAI scan for detection of iodineavid M1.
Potential reasons for this discrepancy between pre- and post-RAI scans for disease detection include differences in imaging time. The pre-RAI scan was acquired 24 hrs after 123I dosing in contrast to the one-week delay in the acquisition following 131I therapy. A longer delay increases the target-to-background ratio and may improve lesion identification. Previous work has shown a higher detection of iodine-avid lung or bone metastasis on post-therapy scans at 7 vs. 3 days [23]. Gerard et al evaluated the diagnostic accuracy of pre-RAI scans at 4, 24 and 48 hrs after 123I dosing and observed that 48-hr imaging is feasible and may improve detection of weakly avid tumor seen only on post-RAI scans [5].
We found that patients with discordant findings received significantly higher 131I activity in groups 1 and 2 (but not in 3), suggesting that higher activities of 131I may be required for identification of loco-regional and distant metastatic disease. In this respect, expression of the human sodium iodide symporter (hNIS) system is considered a required step for radioiodine uptake in normal and malignant thyroid cells [24]. Studies have documented that lymph node and distant metastatic tissue exhibit significantly lower or absence of expression of hNIS than the primary thyroid tumor, suggesting a mechanistic explanation for the lack of iodine-avidity of some thyroid cancer [25,26]. It is conceivable that the reduced hNIS expression of some tumor cells may limit their visualization when using the low radioiodine activities in pre-RAI scans.
Using varying definitions of pre- and post-RAI scan discordance, a number of studies have previously attempted to evaluate its frequency and clinical significance in the management of DTC. The results of these studies vary widely [8-10]. Alzahrani et al, reported higher discordance rates in patients undergoing a second cycle of RAI (18%) than in those receiving their first RAI (6%) [10]. Urhan et al, compared patients with either a positive or negative pre-RAI 123I vs. post-RAI scans and found higher discordance in the latter group of patients (13 vs. 28%) [9]. They also reported that 21 of 292 patients with thyroid cancer (7% of study population) exhibited completely discordant findings in the lungs, bone, mediastinum, thyroid bed and neck comparing pre- and post-RAI scans [9], although the authors did not establish how many of them were due to thyroid cancer. Donahue et al observed that 19 of 108 RAI-naive patients with thyroid cancer had totally discordant scans, with 11 (10% of total group) showing clinical upstaging and 6 patients (5.5% of the total study group) showing scintigraphic patterns suggestive of distant metastatic disease [8].
Similar to previous studies, the authors did not confirm the veracity of the discordant findings with additional diagnostic modalities. It is worth-mentioning again that we did not consider the findings discordant where the post-RAI scan showed one or more additional areas of increased radiotracer activity at sites already demonstrating radiotracer uptake on the corresponding pre-RAI WBS. These findings were already available to the clinician prior to RAI, and the clinical information obtained from this group was unlikely to have any impact on clinical staging or patient management.
Another important observation is the fact that discordance rates between pre- and post-RAI scans in group 2 (30%) and group 3 (63%) were quite high in our study when compared to previous studies (6~28%). This difference is likely due to selection bias, and dilutional effect, such as when occurs by including patients with low- and high-risk characteristics in the same group. For instance, the overall discordance rate when combining groups 1-3 was only 10% (34/342) in our study, which falls well within the reported range in the literature.
We used SPECT/CT in 38% of the discordant cases (13/34). This technique played an important role for the detection of metastases not seen on planar WBS (nodal, lung, pleural, mediastinal, and bone metastases), and for precise localization of physiologic (thymus), benign (lymph nodes, liver cyst) and spurious findings. Previous studies have found SPECT/CT to be a useful tool to characterize atypical or cryptic findings on WBS by differentiating thyroid remnant or cancer from physiologic activity or non-thyroid pathology [16].
There are important clinical considerations derived from our study: 1) our results suggest that post-RAI WBS may not need to be routinely performed in RAI-naive patients without clinical evidence of M1 as the likelihood of unmasking metastatic disease by the post-RAI scan is approximately 1 in 100 according to our results, and discordant findings are more likely to represent false positive findings. Instead, they should be obtained on a case-by-case basis, based on the clinical suspicion for occult M1; 2) our study also indicates that in patients with clinical evidence of persistent disease after RAI and/or M1status, the post-RAI scan appears to yield greater iodine-avid disease detection than the pre-RAI study. This raises the question of the role of these pre-RAI-planning scans in this clinical scenario. This is especially relevant in patients with suspected or known M1 for whom pre-RAI scans are often used in the decision-making for delivery of higher RAI doses; 3) Additional research is required to assess the use of a higher 123I activity (5.0 vs. 2.0 mCi) and image acquisition delay to 48 hrs (instead of 24 hr) with pre-RAI WBS in higher-risk DTC patients, which may result in higher target to background ratios and enhanced disease detection, and 4) SPECT/CT should be considered in discordant cases to exclude metastatic DTC, especially in higher risk patients such as those with history of previous RAI.
Finally, it is worth mentioning potential limitations of this work, including its retrospective design, as well as the fact that our findings are limited to a single tertiary referring institution and may not be applicable to other centers. Another limitation is the heterogeneity of the 3 groups with a limited number of DTC patients in group 2 (n=23) and group 3 (n=8) compared to group 1 (n=311).
In summary, the significance of pre- and post-therapy scans during the management of patients with thyroid cancer is highly influenced by the clinical setting. Unsuspected distant metastatic disease is infrequent in RAI-naive patients and most discordant findings are usually due to physiologic or other benign conditions, and represent false positive findings. In contrast, in higher risk-patients, including those with previous RAI and/or known distant metastatic disease, discordant results likely correspond to true disease.
Acknowledgment
We thank Judy Buchanan from the Division of Nuclear Medicine and Lori Sokoll from the Division of Endocrinology and Metabolism at Johns Hopkins for their assistance in the preparation of this manuscript.
References
- 1.Cooper DS, Doherty GM, Haugen BR, Kloos RT, Lee SL, Mandel SJ, Mazzaferri EL, McIver B, Pacini F, Schlumberger M, Sherman SI, Steward DL, Tuttle RM. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid. 2009;19:1167–1214. doi: 10.1089/thy.2009.0110. [DOI] [PubMed] [Google Scholar]
- 2.Haymart MR, Banerjee M, Stewart AK, Koenig RJ, Birkmeyer JD, Griggs JJ. Use of radioactive iodine for thyroid cancer. JAMA. 2011;306:721–728. doi: 10.1001/jama.2011.1139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sherman SI, Tielens ET, Sostre S, Wharam MD Jr, Ladenson PW. Clinical utility of post-treatment radioiodine scans in the management of patients with thyroid carcinoma. J Clin Endocrinol Metab. 1994;78:629–634. doi: 10.1210/jcem.78.3.8126134. [DOI] [PubMed] [Google Scholar]
- 4.Mandel SJ, Shankar LK, Benard F, Yamamoto A, Alavi A. Superiority of iodine-123 compared with iodine-131 scanning for thyroid remnants in patients with differentiated thyroid cancer. Clin Nucl Med. 2001;26:6–9. doi: 10.1097/00003072-200101000-00002. [DOI] [PubMed] [Google Scholar]
- 5.Gerard SK, Cavalieri RR. I-123 diagnostic thyroid tumor whole-body scanning with imaging at 6, 24, and 48 hours. Clin Nucl Med. 2002;27:1–8. doi: 10.1097/00003072-200201000-00001. [DOI] [PubMed] [Google Scholar]
- 6.Siddiqi A, Foley RR, Britton KE, Sibtain A, Plowman PN, Grossman AB, Monson JP, Besser GM. The role of 123I-diagnostic imaging in the follow-up of patients with differentiated thyroid carcinoma as compared to 131I-scanning: avoidance of negative therapeutic uptake due to stunning. Clin Endocrinol (Oxf) 2001;55:515–521. doi: 10.1046/j.1365-2265.2001.01376.x. [DOI] [PubMed] [Google Scholar]
- 7.Hilditch TE, Dempsey MF, Bolster AA, McMenemin RM, Reed NS. Self-stunning in thyroid ablation: evidence from comparative studies of diagnostic 131I and 123I. Eur J Nucl Med Mol Imaging. 2002;29:783–788. doi: 10.1007/s00259-002-0785-6. [DOI] [PubMed] [Google Scholar]
- 8.Donahue KP, Shah NP, Lee SL, Oates ME. Initial staging of differentiated thyroid carcinoma: continued utility of posttherapy 131I whole-body scintigraphy. Radiology. 2008;246:887–894. doi: 10.1148/radiol.2463061328. [DOI] [PubMed] [Google Scholar]
- 9.Urhan M, Dadparvar S, Mavi A, Houseni M, Chamroonrat W, Alavi A, Mandel SJ. Iodine-123 as a diagnostic imaging agent in differentiated thyroid carcinoma: a comparison with iodine-131 post-treatment scanning and serum thyroglobulin measurement. Eur J Nucl Med Mol Imaging. 2007;34:1012–1017. doi: 10.1007/s00259-006-0341-x. [DOI] [PubMed] [Google Scholar]
- 10.Alzahrani AS, Bakheet S, Al Mandil M, Al-Hajjaj A, Almahfouz A, Al Haj A. 123I isotope as a diagnostic agent in the follow-up of patients with differentiated thyroid cancer: comparison with post 131I therapy whole body scanning. J Clin Endocrinol Metab. 2001;86:5294–5300. doi: 10.1210/jcem.86.11.8030. [DOI] [PubMed] [Google Scholar]
- 11.Iwano S, Kato K, Nihashi T, Ito S, Tachi Y, Naganawa S. Comparisons of I-123 diagnostic and I-131 post-treatment scans for detecting residual thyroid tissue and metastases of differentiated thyroid cancer. Ann Nucl Med. 2009;23:777–782. doi: 10.1007/s12149-009-0303-z. [DOI] [PubMed] [Google Scholar]
- 12.Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, editors. AJCC Cancer Staging Manual. Springer; 2010. [Google Scholar]
- 13.Bravo P, Ewertz M, Wahl R, Cooper D, Ladenson P. Comparison of a four-day recombinant thyrotropin regimen with thyroid hormone withdrawal to prepare for postoperative radioiodine ablation of thyroid remnants in patients with thyroid carcinoma. J Nucl Med. 2010;51(Suppl):311. [Google Scholar]
- 14.Pacini F, Ladenson PW, Schlumberger M, Driedger A, Luster M, Kloos RT, Sherman S, Haugen B, Corone C, Molinaro E, Elisei R, Ceccarelli C, Pinchera A, Wahl RL, Leboulleux S, Ricard M, Yoo J, Busaidy NL, Delpassand E, Hanscheid H, Felbinger R, Lassmann M, Reiners C. Radioiodine ablation of thyroid remnants after preparation with recombinant human thyrotropin in differentiated thyroid carcinoma: results of an international, randomized, controlled study. J Clin Endocrinol Metab. 2006;91:926–932. doi: 10.1210/jc.2005-1651. [DOI] [PubMed] [Google Scholar]
- 15.Mallick U, Harmer C, Yap B, Wadsley J, Clarke S, Moss L, Nicol A, Clark PM, Farnell K, McCready R, Smellie J, Franklyn JA, John R, Nutting CM, Newbold K, Lemon C, Gerrard G, Abdel-Hamid A, Hardman J, Macias E, Roques T, Whitaker S, Vijayan R, Alvarez P, Beare S, Forsyth S, Kadalayil L, Hackshaw A. Ablation with low-dose radioiodine and thyrotropin alfa in thyroid cancer. N Engl J Med. 2012;366:1674–1685. doi: 10.1056/NEJMoa1109589. [DOI] [PubMed] [Google Scholar]
- 16.Blum M, Tiu S, Chu M, Goel S, Friedman K. I-131 SPECT/CT elucidates cryptic findings on planar whole-body scans and can reduce needless therapy with I-131 in post-thyroidectomy thyroid cancer patients. Thyroid. 2011;21:1235–1247. doi: 10.1089/thy.2011.0010. [DOI] [PubMed] [Google Scholar]
- 17.Carlisle MR, Lu C, McDougall IR. The interpretation of 131I scans in the evaluation of thyroid cancer, with an emphasis on false positive findings. Nucl Med Commun. 2003;24:715–735. doi: 10.1097/00006231-200306000-00015. [DOI] [PubMed] [Google Scholar]
- 18.Jong I, Taubman K, Schlicht S. Bronchiectasis simulating pulmonary metastases on iodine-131 scintigraphy in well-differentiated thyroid carcinoma. Clin Nucl Med. 2005;30:688–689. doi: 10.1097/01.rlu.0000178242.69075.64. [DOI] [PubMed] [Google Scholar]
- 19.Wong KK, Avram AM. Posttherapy I-131 thymic uptake demonstrated with SPECT/CT in a young girl with papillary thyroid carcinoma. Thyroid. 2008;18:919–920. doi: 10.1089/thy.2007.0394. [DOI] [PubMed] [Google Scholar]
- 20.Muratet JP, Giraud P. Thymus accumulation of I-131 after therapeutic dose for thyroid carcinoma. Clin Nucl Med. 1996;21:736–737. doi: 10.1097/00003072-199609000-00016. [DOI] [PubMed] [Google Scholar]
- 21.Oh JR, Ahn BC. False-positive uptake on radioiodine whole-body scintigraphy: physiologic and pathologic variants unrelated to thyroid cancer. Am J Nucl Med Mol Imaging. 2012;2:362–385. [PMC free article] [PubMed] [Google Scholar]
- 22.Schuster DM, Alazraki N. Esophageal scarring causing false-positive uptake on I-131 whole-body imaging. Clin Nucl Med. 1998;23:334. doi: 10.1097/00003072-199805000-00020. [DOI] [PubMed] [Google Scholar]
- 23.Chong A, Song HC, Min JJ, Jeong SY, Ha JM, Kim J, Yoo SU, Oh JR, Bom HS. Improved Detection of Lung or Bone Metastases with an I-131 Whole Body Scan on the 7th Day After High-Dose I-131 Therapy in Patients with Thyroid Cancer. Nuclear Medicine and Molecular Imaging. 2010;44:273–281. doi: 10.1007/s13139-010-0051-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Filetti S, Bidart JM, Arturi F, Caillou B, Russo D, Schlumberger M. Sodium/iodide symporter: a key transport system in thyroid cancer cell metabolism. Eur J Endocrinol. 1999;141:443–457. doi: 10.1530/eje.0.1410443. [DOI] [PubMed] [Google Scholar]
- 25.Park HJ, Kim JY, Park KY, Gong G, Hong SJ, Ahn IM. Expressions of human sodium iodide symporter mRNA in primary and metastatic papillary thyroid carcinomas. Thyroid. 2000;10:211–217. doi: 10.1089/thy.2000.10.211. [DOI] [PubMed] [Google Scholar]
- 26.Lin JD, Chan EC, Chao TC, Chen KT, Hsueh C, Ho YS, Weng HF. Expression of sodium iodide symporter in metastatic and follicular human thyroid tissues. Ann Oncol. 2000;11:625–629. doi: 10.1023/a:1008340402577. [DOI] [PubMed] [Google Scholar]