Miami Thyroid Oncology Consortium Projects
The Miami Thyroid Oncology Consortium Project (MTOCP) is a translational research cooperative group supported by the Miami Cancer Research Center (MCRC). We carry on more than 80-year legacy of molecular theranostics started with Dr. Saul Hertz that have changed the way thyroid cancer was treated. More recently, molecular oncology has entered into a new paradigm with next generation sequencing (NGS) DNA sequencing technology. Miami Thyroid Oncology Consortium (MTOC) is an academic alliance, primarily including endocrinologists and surgical endocrinologists, formed with the common goal of assembling their intellectual resources to improve patient care through scientific collaboration and translational research. The participation to the consortium is accessible to all physicians involved or interested in the care of patients with thyroid disorders and cancer. The MTOC was established in 2015 and has implemented two projects since then. The first project (MTOCP-1) was a phase -2 theranostic clinical trial exploring the clinical value and dosimetric utility of I-124 PET/CT in thyroid cancer. The second project (MTOCP-2) was a phase-4 theranostic clinical registry investigating the diagnostic and theranostic value of Thyroseq, a NGS-based genomic profiling platform, in thyroid nodules and thyroid cancer.
I-124 PET/CT in Patients with Differentiated Thyroid Cancer: Clinical and Quantitative Image Analysis
The study was designed as a prospective phase II diagnostic trial with the objectives to determine the imaging characteristics and clinical feasibility of 124I PET/CT imaging for determination of extent of disease and evaluation of RAI kinetics in its physiologic and neoplastic distribution in patients with differentiated thyroid cancer (DTC). Patients with confirmed differentiated (both well-differentiated and poorly differentiated) thyroid cancers were studied. Patients who were newly diagnosed, as well as those who had known or suspected recurrent/metastatic disease, were eligible for the trial. The inclusion criteria for the study included a histological confirmation of DTC and a clinical indication for RAI imaging (detection of known or suspected postoperative residual thyroid bed or nodal disease, extent-of-disease evaluation in known recurrent/metastatic disease, suspicious nodule/mass detected by physical exam, imaging study or fine-needle aspiration, recurrent/metastatic disease suspected by elevated thyroglobulin).
The administered activity for I-124 was 2mCi by oral administration in liquid form. The basic imaging protocol involved a 5 time-point (2-4 h, 24±6 h, 48±6 h, 72±6 h, 96 ± 6h post-administration) whole body PET/CT imaging schedule. The patients were prepared for RAI imaging/dosimetry either by withholding suppressive T4 for an adequate length of time (to achieve a TSH level of >50 at the time of imaging) or by administering recombinant human TSH (rhTSH) (two consecutive daily doses of 0.9 mg IM, in the days preceding the RAI administration). Images were acquired on a Siemens unit with standard settings. All patients who had I-124 imaging, subsequently underwent RAI treatment with I-131 sodium iodide, with administered activities in the range of 100mCi to 300mCi. Post-treatment scans were obtained 5 to 7 days after RAI treatment. Anterior and posterior whole body scans as well as static antero-posterior and oblique neck images were acquired. The localization of I-124 in known/suspected lesions including cervical and remote metastatic sites was documented. I-124 images were compared to post-treatment I-131 images. Comparisons were performed on a by-patient and by-lesion basis. All images were reviewed and analyzed by two experienced nuclear medicine physicians. Quantitative image analysis was performed using semiautomatic region of interest (ROI) methodology. The total functional volume (ml), activity per functional volume (uCi/ml), and cumulated activity (uCi-hr) for remnants, salivary glands and nodal metastases were calculated. The I-124 images were also compared to F-18 FDG images that were acquired prior to RAI treatment in all patients. F-18 FDG PET/CT imaging was performed as part of a comprehensive extent of disease evaluation and not for the purpose of this study per se. Relative sensitivity determination for I-124 PET/CT vs Post-treatment planar I-131 imaging: Comparative image analysis was performed on by-patient, and by-lesion basis. For the purposes of by-lesion analysis, any distinct uptake noted on I-124 PET/CT or post-treatment I-131 planar images was considered “Positive Reference.” A positive reference implies presence of a tumor/remnant with RAI uptake. The true positive (TP) and false negative (FN) designations, and the sensitivity calculation for I-124 and I-131 imaging were performed based on the “positive reference.” The sites of physiologic uptake were carefully identified. A physiologic uptake was not considered as false positive (FP). A true negative (TN) designation was used when both I-124 and I-131 images were negative.
The overall by-lesion detection sensitivity for post-treatment I-131 planar images and pretreatment I-124 PET/CT images were 72% and 92% respectively. By-patient analysis indicated that remnant uptake was demonstrated in all of these patients on both I-124 and I-131 imaging studies. A total of 11 distinct foci of remnant uptake were identified. I-124 distinctly defined remnant uptake in right lobe, left lobe, and isthmus/pyramidal lobe anatomic sites. I-124 was positive in 11/11 (100%). I-131 revealed 9/11 (82%) distinct remnant foci. The two missed foci of remnant uptake by I-131 were in the trajectory of pyramidal lobe in the midline. FDG was negative in all remnant tissue, and none of the thyroid remnants were visually detected as a soft tissue abnormality on CT. The sequential I-124 images consistently demonstrated the maximum remnant activity to occur at 24 hours. After the peak activity was reached, the clearance was monoexponential. The maximum remnant activity ranged from 1.2 to 215.9 uCi with the total functional remnant volume (the total number of voxels within the remnant ROI) ranging from 1 to 60 mL. The activity per volume of remnant tissue ranged from 0.036 to 11.265 uCi/mL. The total cumulated activity within the remnant ranged from 68 to 12757.3 uCi/hr. There were 19 distinct foci of uptake identified as nodal metastasis. I-124 was positive in 16/19 (84%). I-131 revealed 9/19 (47%) distinct nodal uptake. The three negative nodes by I-124 were also negative by I-131 but positive on FDG (Iodine-refractory nodal disease). Nodal metastatic disease demonstrated a pattern of uptake that was significantly different than the thyroid remnant or physiologic salivary gland activity. A protracted retention was identified as a characteristic pattern for metastatic nodal disease. Despite an early period of fairly rapid uptake of radioiodine during the first 4 hours, there was prolonged radioiodine accumulation and retention in nodal disease over time with most lesions becoming visually detectable at 48-72 hours. There were five cases of metastatic lung disease (one micronodular, four macronodular). One case was negative on both I-124 and I-131, but was positive on FDG (Iodine-refractory disease). I-124 was positive in 1/4 (25%), I-131 post-treatment scan was positive in 4/4 (100%) cases. There was only one patient with abdominal disease. This was a very unusual case that presented with metastatic abdominal disease and no primary was identified in the total thyroidectomy specimen. The disease was discovered at an exploratory laparotomy and confirmed by H&E and IHC (for thyroglobulin and TTF-1 staining). A subsequent FDG study showed hepatic, mesenteric nodal and peritoneal disease. I-124 demonstrated positive uptake in all abdominal lesions, however I-131 was only positive in the hepatic disease. In conclusion, the data indicated that I-124 PET/CT imaging was clearly superior, providing exquisite details in terms of location and laterality of the remnant tissue as well as nodal and remote metastases.
Genomic Profiling of Nodular Thyroid Disease and Thyroid Cancer
This was an open ended prospective registry. The patients who were diagnosed with thyroid nodules underwent a complete clinical and US evaluation. Thyroid nodule biopsy indication and FNA vs core biopsy choices were made entirely on clinical grounds by the managing physicians. Thyroid nodule biopsies were performed at a participating medical or surgical endocrinology office. Following standard cytologic examination, molecular testing using ThyroSeq was performed. The biopsy results were categorized according to Bethesda system. Category I (non-diagnostic), Category II (benign cytology), categories III, IV and V (the indeterminate group), and category VI (malignant cytology) all underwent ThyroSeq analysis. The FNA results and ThyroSeq results of all registry patients were collected. Decisions concerning operative management were made entirely on clinical grounds by the managing physicians. The routine histopathology was processed and reported as per institutional protocols. An expert review was obtained in selected cases where the findings were equivocal. Molecular testing of surgical specimens was performed as indicated. Patients were eligible for registry enrollment if they were undergoing work up for a thyroid nodule. Ultrasound evaluation was done per the established practice protocols. Endocrinologists, surgeons, or radiologists performed US evaluations. A standard nodule US map and descriptions provided by the investigator were used. The decision to biopsy was based on the physician’s clinical assessment of the thyroid nodule. Samples were sent to the reference lab. Sample collection media, tubes, and shipping materials were provided by the reference lab. The managing physicians had the choice to elect performing a core biopsy with or without an FNA as they found appropriate clinically. Decisions to perform surgical intervention were based on a multidisciplinary clinical assessment. Surgical specimens were processed by the facility’s pathology department via routine histopathology. Select surgical specimens were submitted to University of Pittsburgh Medical Center (UPMC) for a second review and molecular testing with ThyroSeq. Operative reports and pathology reports were submitted to the data center for data entry. The US map was completed with information recorded regarding the nodule size, nodule location, and a description of nodule characteristics. Cytology results were recorded, including Bethesda category. Any further distinctions about the site of biopsy were recorded, including if it was performed in a recurrent or metastatic setting, or involving lymph nodes or other sites outside the thyroid. The ThyroSeq results were recorded for all cytology specimens. If operative intervention was performed, the date and type of surgery performed were recorded. Surgical pathology and surgical ThyroSeq results were collected and recorded if available. The project introduced a new concept named “Positive Predictive Value” for performance evaluation. Traditional definitions of False Negative (FN), True Negative (TN), False Positive (FP) and True Positive (TP) relating to nodular thyroid disease are limited due to the narrow use of end results as either cancer or no cancer. In the theranostics context, the definitions of FN, TN, FP and TP are more complex. False Negative (FN) and True Negative (TN): A neoplastic surgical pathology (adenoma, NIFTP, carcinoma) and a Negative Thyroseq were designated as FN. A non-neoplastic surgical pathology and/or Bethesda 2 category cytology and Negative Thyroseq was designated as TN. A Negative Predictive Value was then determined by True Negative / [True negative + False negative] equation. False Positive (FP) and True Positive (TP): False Positive (FP) is a null concept. This term relates to the null set in mathematics, a set that is negligible in some sense. There are no false positive results regarding genomic changes because everything means something. Therefore all positive results are True Positive (TP). We replaced the positive predictive value with Positive Prognostic Value. When a ThyroSeq result was reported to be positive, an interpretation of the positive findings was included in the report to guide the managing physician.
The registry collected data between January 1, 2017 and December 31, 2017. ThyroSeq genomic profiling was performed on all cytology specimens including benign, indeterminate, and malignant. A total of 202 patients with 227 nodules had complete cytology and ThyroSeq paired information. The negative predictive value of Thyroseq was found to be 97% and the positive prognostic value of the test was 95%. There were a total of 68 genomic changes from all cytology specimens including mutations, copy number alterations (CNA), fusions, and abnormal gene expression profiles (GEP) found in 56 nodules. In the Bethesda 2 category, there were 27 nodules with positive Thyroseq results, and 3 of them had more than one genomic change. There were TSHR/GNAS mutations found in 9 out of 27 (33.3%) of the nodules. No surgical confirmation was obtained for these nodules. There were RAS mutations detected in 9 out of 27 (33.3%) nodules. Surgical pathology was obtained in 6 of the 9 RAS-positive nodules revealing 2 PTC follicular variant, 1 PTC classical variant, 2 NIFTP, and 1 hyperplasia (clonality not assessed). The remaining patients with benign nodules and positive ThyroSeq results are under continued clinical observation. In the Bethesda 3 category, there were 20 nodules with positive ThyroSeq results, and 3 of them had more than one genomic change. There were RAS mutations detected in 15 out of 20 (75%) of these nodules. Surgical pathology was obtained in 15 out of the 20 nodules revealing 5 PTC follicular variant, 2 PTC classical variant, 1 follicular carcinoma, 5 follicular adenomas, 1 NIFTP, and 1 hyperplasia (clonality not assessed). In this indeterminate category there were 20 nodules with negative Thyroseq. Surgical pathology was obtained in 5 out of the 20 nodules revealing 1 follicular adenoma, 1 PTC follicular variant, 2 hyperplasia, and 1 chronic thyroiditis with multifocal papillary microcarcinoma. There was 1 nodule with cytology showing suspicious for follicular neoplasm, and ThyroSeq was positive for 3 genomic changes including HRAS mutation, CNA, and GEP. Surgical pathology was obtained from thyroid lobectomy, which showed noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP). ThyroSeq on the surgical specimen also showed the HRAS mutation. In the Bethesda 6 category there were 7 nodules with positive Thyroseq results, and 3 of them had more than one genomic change. One malignant nodule that had a negative Thyroseq proved to be a lymphoma. BRAF mutations were detected in 5 out of 7 of these nodules (71.4%). Surgical pathology was obtained in all 7 nodules revealing 1 PTC follicular variant and 6 PTC classical variant. There were 59 positive Thyroseq results, and 56 were considered true positive. There were 3 false positive results, where nodules had positive ThyroSeq but surgical pathology showed hyperplasia. The positive prognostic value in a theranostics context, was 94.9%. There were 132 true negatives and 4 false negatives Thyroseq results, translating into a Negative Predictive Value of 97.1%. There were 7 nodules (3.1%) with BRAF mutations out of the entire cohort. Out of these, 2 (28.5%) of the nodules had Bethesda 3 cytology and 5 (71.4%) had Bethesda 6 cytology. Surgical pathology was obtained for 6 of these nodules, revealing 2 (33.3%) PTC follicular variant and 4 (66.6%) PTC classic variant. ThyroSeq was obtained on 2 of the surgical specimens, which both confirmed the BRAF mutations. All 7 of the BRAF mutations identified by ThyroSeq on cytology specimens showed point mutations of p. V600E, c. 1799T>A. Out of these 7 nodules, 4 of them were also positive for GEP. There were 25 nodules (11.0%) out of the cohort with positive ThyroSeq for RAS mutations. Out of these, 9 (36%) had H-RAS mutations and 16 (64%) had N-RAS mutations. No K-RAS mutations were identified on cytology. Surgical pathology was obtained on 18 (72%) of the RAS positive nodules, revealing 4 (22.2%) NIFTP, 5 (27.7%) PTC follicular variant, 2 (11.1%) PTC classic variant, 4 (22.2%) follicular adenoma, 1 (5.55%) follicular carcinoma, and 2 (11.1%) nodular hyperplasia. ThyroSeq was obtained on 10 (55.5%) of the surgical specimens, and 9 of these were concordant with the cytology ThyroSeq. In 1 nodule for which surgical pathology showed NIFTP, cytology ThyroSeq was positive for HRAS and surgical ThyroSeq was positive for KRAS. Out of all RAS mutations, 21 nodules (84%) had point mutations of p.Q61R, c.182A>G and 4 nodules (16%) had point mutations of p.Q61K, c.181C>A. The Q61R point mutations were found in 7 H-RAS mutations and 14 N-RAS mutations. The Q61K point mutations were found in 2 H-RAS mutations and 2 N-RAS mutations. Out of all 25 RAS positive nodules, 1 was also positive for CNA, 2 were positive for GEP, and 1 was positive for CNA and GEP.
The knowledge of the genomic profile of thyroid nodules and cancer has proven to be a useful tool for the clinician to tailor the patient’s management strategy. While already established as a clear step in the work up of indeterminate nodules, our study suggested that genomic profiling likely has a broader role in the management of thyroid nodular disease. Theranostically, ThyroSeq genomic profiling is beneficial for the medical and surgical management of thyroid nodules regardless of the Bethesda category. We continue to interrogate the intricate details of certain genomic alteratins to better risk stratify patients and guide clinical decision-making.