Clinical Endocrinology (2014)
doi: 10.1111/cen.12419
ORIGINAL ARTICLE
The prevalence of thyroglossal tract thyroid tissue on SPECT/CT following 131I ablation therapy after total thyroidectomy for thyroid cancer Thomas W. Barber*, Martin H. Cherk*,†, Duncan J. Topliss†,‡, Jonathan W. Serpell§,¶, Kenneth S. K. Yap*, Michael Bailey** and Victor Kalff*,† *Department of Nuclear Medicine and PET Centre, †Department of Medicine, Monash University, ‡Department of Endocrinology & Diabetes, §Department of Breast and Endocrine Surgery, ¶Department of Surgery, Monash University, and **Department of Epidemiology and Preventive Medicine, Monash University, The Alfred Hospital, Melbourne, Vic., Australia
Abstract Objective To determine the prevalence of thyroglossal tract thyroid tissue on SPECT/CT and to assess the contribution of this tissue to total neck radioactive iodine (RAI) activity in patients given 131I ablation therapy after total thyroidectomy for thyroid cancer. Patients and Methods Eighty-three consecutive patients with thyroid cancer treated with total thyroidectomy underwent whole-body planar and SPECT/CT imaging of the neck following initial RAI ablation. On SPECT/CT, thyroglossal tract thyroid tissue was defined as RAI in the anterior neck, superior to the thyroid bed in close proximity to the midline without evidence of localization to lymph nodes. Quantification was performed using region of interest analysis on planar imaging following localization on SPECT/CT. SPECT/CT, and planar images were classified by two reviewers as positive, negative or equivocal with interobserver agreement quantified using a Kappa score. Disagreement was resolved using a third reviewer. Results Thyroglossal tract thyroid tissue was present in 39/83 (47%; 95%CI: 36–58%) patients on SPECT/CT. In these 39 patients, this tissue contributed to a significant amount of total neck activity (median = 50%; IQR 19–74%). Interobserver agreement for the presence of thyroglossal tract thyroid tissue was substantial on SPECT/CT (Kappa = 073) and fair on planar imaging (Kappa = 031). Conclusion Thyroglossal tract thyroid tissue was present in one half of our study population and contributed to a significant amount of total neck RAI activity. Given the high prevalence of this tissue, our results suggest that total neck RAI activity on planar imaging may not be suitable to assess the completeness of thyroid bed surgery.
Correspondence: Thomas Barber, Department of Nuclear Medicine and PET Centre, The Alfred Hospital, Commercial Road, Melbourne 3004, Australia. Tel.: +613 9076 2432; Fax: +613 9076 2599; E-mail: [email protected] © 2014 John Wiley & Sons Ltd
(Received 15 October 2013; returned for revision 27 November 2013; finally revised 16 January 2014; accepted 26 January 2014)
Introduction Remnants of thyroid tissue may be present at any point along the embryological path of thyroid descent from the base of the tongue to the thyroid bed.1 The pyramidal lobe can be thought of as the most caudal remnant of this thyroglossal tract but thyroid tissue may also be found more superiorly forming sites of ectopic thyroid tissue.1 During total thyroidectomy for thyroid cancer, the pyramidal lobe is typically excised to the level of the hyoid bone; however, some thyroglossal tract thyroid tissue may be separate from the thyroid gland, lie outside the operative field or even be occult at surgery.2–4 The prevalence and quantity of this residual thyroglossal tract thyroid tissue are poorly defined, and its presence may have implications in patients undergoing treatment with radioactive iodine (RAI). Following RAI therapy, planar imaging is typically performed and is highly sensitive for the detection of iodine avid thyroid tissue. Planar imaging, however, is limited in anatomical detail, and it is often difficult to determine whether neck activity localizes to thyroid bed remnant, lymph nodes or to thyroglossal tract thyroid tissue. Single photon emission computed tomography fused with computed tomography (SPECT/CT) improves anatomical localization and provides incremental information over planar imaging, particularly in the assessment of nodal metastatic disease.5–13 It is therefore expected that SPECT/CT will also improve delineation of thyroglossal tract thyroid tissue from activity in the thyroid bed. This may be important as the total amount of RAI activity in the neck on planar imaging is often used as a surrogate marker for the completeness of initial surgery.14 Furthermore, in institutions where RAI scanning is performed prior to RAI therapy, more accurate interpretation of neck activity may assist in defining whether the primary goal of RAI therapy is remnant ablation, adjuvant therapy or treatment for persistent disease.
1
2 T. W. Barber et al. Therefore, the aims of this study are (1) to determine the prevalence of thyroglossal tract thyroid tissue on SPECT/CT and (2) to assess the contribution of this tissue to total neck RAI activity in patients treated with 131I after total thyroidectomy for thyroid cancer.
Materials and methods Study population Eighty-three consecutive patients (55 female and 28 male) with thyroid cancer were referred to our institution between June 2012 and July 2013 for initial RAI ablation therapy following total thyroidectomy. All but one patient were prepared with recombinant human thyroid stimulating hormone, and the one remaining patient was prepared with thyroid hormone withdrawal. Patients received either low dose (1110 MBq, 30 mCi) or high dose (4255 MBq, 115 mCi) 131I therapy as determined by the treating clinician based on the histological/surgical findings. Disease stage was classified according to the American Joint Committee on Cancer (AJCC) staging system (7th edition)15, and disease risk was assessed using the American Thyroid Association (ATA) guidelines.16 The study was approved by our institutional ethics committee.
Image analysis On SPECT/CT, thyroglossal tract thyroid tissue was defined as RAI activity in the anterior neck, superior to the thyroid bed in close proximity to the midline (≤1 cm) without evidence of localization to lymph nodes on the coregistered CT. On planar imaging, thyroglossal tract thyroid tissue was defined as a linear/ oval pattern of RAI activity in the midline of the neck superior to the thyroid bed. Two nuclear medicine specialists independently and sequentially reviewed the planar and SPECT/CT images and classified the images as positive, negative or equivocal for thyroglossal tract thyroid tissue. Reviewers were blinded to prior interpretations and results of prior imaging studies. When there was disagreement between reviewers on SPECT/CT, a third reviewer was used to form a consensus regarding the presence of thyroglossal tract thyroid tissue. Quantification of thyroglossal tract thyroid tissue and total neck RAI activity was performed using region of interest (ROI) analysis on planar imaging after localization of RAI activity on SPECT/CT. The percentage of activity in the thyroglossal tract thyroid tissue ROI compared with the total neck ROI was calculated following correction for background activity using a ROI over the thigh (Fig. 1). Thyroid Stimulating Hormone (TSH) and thyroglobulin assays
Image acquisition All patients underwent planar and SPECT/CT imaging 48 h following RAI therapy. All images were performed on a General Electric (GE) Discovery 670 SPECT/CT system (Milwaukee, WI, USA) using high energy general purpose collimators with a 364 keV photopeak and 20% energy window. SPECT/CT images were acquired of the head and neck using 60 steps at 20 s per step with a 128 9 128 matrix. The CT scan was acquired using parameters of 120 kVp and 40–110 mAs with a 512 9 512 matrix and 25-mm slices. SPECT data were processed using GE’s Volumetrix MI Evolution for Oncology programme on a Xeleris 3 workstation (Milwaukee, WI, USA) and incorporating iterative reconstruction, attenuation correction, scatter correction and resolution recovery software. Planar images were acquired from the vertex of the skull to the knees at a speed of 17 cm/min with a 256 9 1024 matrix.
(a)
(b)
TSH was assayed by the Abbott Architect chemiluminescent microparticle immunoassay with a functional sensitivity of
doi: 10.1111/cen.12419
ORIGINAL ARTICLE
The prevalence of thyroglossal tract thyroid tissue on SPECT/CT following 131I ablation therapy after total thyroidectomy for thyroid cancer Thomas W. Barber*, Martin H. Cherk*,†, Duncan J. Topliss†,‡, Jonathan W. Serpell§,¶, Kenneth S. K. Yap*, Michael Bailey** and Victor Kalff*,† *Department of Nuclear Medicine and PET Centre, †Department of Medicine, Monash University, ‡Department of Endocrinology & Diabetes, §Department of Breast and Endocrine Surgery, ¶Department of Surgery, Monash University, and **Department of Epidemiology and Preventive Medicine, Monash University, The Alfred Hospital, Melbourne, Vic., Australia
Abstract Objective To determine the prevalence of thyroglossal tract thyroid tissue on SPECT/CT and to assess the contribution of this tissue to total neck radioactive iodine (RAI) activity in patients given 131I ablation therapy after total thyroidectomy for thyroid cancer. Patients and Methods Eighty-three consecutive patients with thyroid cancer treated with total thyroidectomy underwent whole-body planar and SPECT/CT imaging of the neck following initial RAI ablation. On SPECT/CT, thyroglossal tract thyroid tissue was defined as RAI in the anterior neck, superior to the thyroid bed in close proximity to the midline without evidence of localization to lymph nodes. Quantification was performed using region of interest analysis on planar imaging following localization on SPECT/CT. SPECT/CT, and planar images were classified by two reviewers as positive, negative or equivocal with interobserver agreement quantified using a Kappa score. Disagreement was resolved using a third reviewer. Results Thyroglossal tract thyroid tissue was present in 39/83 (47%; 95%CI: 36–58%) patients on SPECT/CT. In these 39 patients, this tissue contributed to a significant amount of total neck activity (median = 50%; IQR 19–74%). Interobserver agreement for the presence of thyroglossal tract thyroid tissue was substantial on SPECT/CT (Kappa = 073) and fair on planar imaging (Kappa = 031). Conclusion Thyroglossal tract thyroid tissue was present in one half of our study population and contributed to a significant amount of total neck RAI activity. Given the high prevalence of this tissue, our results suggest that total neck RAI activity on planar imaging may not be suitable to assess the completeness of thyroid bed surgery.
Correspondence: Thomas Barber, Department of Nuclear Medicine and PET Centre, The Alfred Hospital, Commercial Road, Melbourne 3004, Australia. Tel.: +613 9076 2432; Fax: +613 9076 2599; E-mail: [email protected] © 2014 John Wiley & Sons Ltd
(Received 15 October 2013; returned for revision 27 November 2013; finally revised 16 January 2014; accepted 26 January 2014)
Introduction Remnants of thyroid tissue may be present at any point along the embryological path of thyroid descent from the base of the tongue to the thyroid bed.1 The pyramidal lobe can be thought of as the most caudal remnant of this thyroglossal tract but thyroid tissue may also be found more superiorly forming sites of ectopic thyroid tissue.1 During total thyroidectomy for thyroid cancer, the pyramidal lobe is typically excised to the level of the hyoid bone; however, some thyroglossal tract thyroid tissue may be separate from the thyroid gland, lie outside the operative field or even be occult at surgery.2–4 The prevalence and quantity of this residual thyroglossal tract thyroid tissue are poorly defined, and its presence may have implications in patients undergoing treatment with radioactive iodine (RAI). Following RAI therapy, planar imaging is typically performed and is highly sensitive for the detection of iodine avid thyroid tissue. Planar imaging, however, is limited in anatomical detail, and it is often difficult to determine whether neck activity localizes to thyroid bed remnant, lymph nodes or to thyroglossal tract thyroid tissue. Single photon emission computed tomography fused with computed tomography (SPECT/CT) improves anatomical localization and provides incremental information over planar imaging, particularly in the assessment of nodal metastatic disease.5–13 It is therefore expected that SPECT/CT will also improve delineation of thyroglossal tract thyroid tissue from activity in the thyroid bed. This may be important as the total amount of RAI activity in the neck on planar imaging is often used as a surrogate marker for the completeness of initial surgery.14 Furthermore, in institutions where RAI scanning is performed prior to RAI therapy, more accurate interpretation of neck activity may assist in defining whether the primary goal of RAI therapy is remnant ablation, adjuvant therapy or treatment for persistent disease.
1
2 T. W. Barber et al. Therefore, the aims of this study are (1) to determine the prevalence of thyroglossal tract thyroid tissue on SPECT/CT and (2) to assess the contribution of this tissue to total neck RAI activity in patients treated with 131I after total thyroidectomy for thyroid cancer.
Materials and methods Study population Eighty-three consecutive patients (55 female and 28 male) with thyroid cancer were referred to our institution between June 2012 and July 2013 for initial RAI ablation therapy following total thyroidectomy. All but one patient were prepared with recombinant human thyroid stimulating hormone, and the one remaining patient was prepared with thyroid hormone withdrawal. Patients received either low dose (1110 MBq, 30 mCi) or high dose (4255 MBq, 115 mCi) 131I therapy as determined by the treating clinician based on the histological/surgical findings. Disease stage was classified according to the American Joint Committee on Cancer (AJCC) staging system (7th edition)15, and disease risk was assessed using the American Thyroid Association (ATA) guidelines.16 The study was approved by our institutional ethics committee.
Image analysis On SPECT/CT, thyroglossal tract thyroid tissue was defined as RAI activity in the anterior neck, superior to the thyroid bed in close proximity to the midline (≤1 cm) without evidence of localization to lymph nodes on the coregistered CT. On planar imaging, thyroglossal tract thyroid tissue was defined as a linear/ oval pattern of RAI activity in the midline of the neck superior to the thyroid bed. Two nuclear medicine specialists independently and sequentially reviewed the planar and SPECT/CT images and classified the images as positive, negative or equivocal for thyroglossal tract thyroid tissue. Reviewers were blinded to prior interpretations and results of prior imaging studies. When there was disagreement between reviewers on SPECT/CT, a third reviewer was used to form a consensus regarding the presence of thyroglossal tract thyroid tissue. Quantification of thyroglossal tract thyroid tissue and total neck RAI activity was performed using region of interest (ROI) analysis on planar imaging after localization of RAI activity on SPECT/CT. The percentage of activity in the thyroglossal tract thyroid tissue ROI compared with the total neck ROI was calculated following correction for background activity using a ROI over the thigh (Fig. 1). Thyroid Stimulating Hormone (TSH) and thyroglobulin assays
Image acquisition All patients underwent planar and SPECT/CT imaging 48 h following RAI therapy. All images were performed on a General Electric (GE) Discovery 670 SPECT/CT system (Milwaukee, WI, USA) using high energy general purpose collimators with a 364 keV photopeak and 20% energy window. SPECT/CT images were acquired of the head and neck using 60 steps at 20 s per step with a 128 9 128 matrix. The CT scan was acquired using parameters of 120 kVp and 40–110 mAs with a 512 9 512 matrix and 25-mm slices. SPECT data were processed using GE’s Volumetrix MI Evolution for Oncology programme on a Xeleris 3 workstation (Milwaukee, WI, USA) and incorporating iterative reconstruction, attenuation correction, scatter correction and resolution recovery software. Planar images were acquired from the vertex of the skull to the knees at a speed of 17 cm/min with a 256 9 1024 matrix.
(a)
(b)
TSH was assayed by the Abbott Architect chemiluminescent microparticle immunoassay with a functional sensitivity of
