The group randomized to the RNA test underwent testing with the Afirma genomic sequencing classifier; the DNA-RNA group underwent testing with the ThyroSeq v3 multigene genomic classifier.
Expected positive predictive value curves of first-generation and second-generation RNA molecular tests and DNA-RNA molecular tests based on observed specificities and sensitivities over the range of possible prevalence of cancer or noninvasive follicular thyroid neoplasm with papillarylike features (NIFTP). Given a prevalence of cancer or NIFTP of 19.6% observed in the present study cohort, the RNA test compared with its previous version demonstrated a significant improvement in positive predictive value (37.5% vs 53.5%). The dotted dark blue line indicates the DNA-RNA test (ThyroSeq v3 multigene genomic classifier); dotted orange line, the previous version of the DNA-RNA test (ThyroSeq v2 next-generation sequencing); solid dark blue line, the RNA test (Afirma genomic sequencing classifier); and solid orange line, the previous version of the RNA test (Afirma gene expression classifier).
eTable. SAS code for analyses
Data Sharing Statement.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Livhits MJ, Zhu CY, Kuo EJ, et al. Effectiveness of Molecular Testing Techniques for Diagnosis of Indeterminate Thyroid Nodules: A Randomized Clinical Trial. JAMA Oncol. 2021;7(1):70–77. doi:10.1001/jamaoncol.2020.5935
Does an RNA test or a DNA-RNA test offer superior performance in estimating the risk of malignancy of thyroid nodules with indeterminate cytology?
In this randomized clinical trial of 346 patients with 372 indeterminate thyroid nodules, the RNA test and the DNA-RNA test demonstrated no statistically significant difference in sensitivity (100% vs 97%, respectively) and specificity (80% vs 85%, respectively).
The molecular testing techniques assessed showed no statistically significant difference in diagnostic performance and allowed 49% of patients with indeterminate thyroid nodules to avoid diagnostic surgery.
Approximately 20% of thyroid nodules display indeterminate cytology. Molecular testing can refine the risk of malignancy and reduce the need for diagnostic hemithyroidectomy.
To compare the diagnostic performance between an RNA test (Afirma genomic sequencing classifier) and DNA-RNA test (ThyroSeq v3 multigene genomic classifier).
Design, Setting, and Participants
This parallel randomized clinical trial of monthly block randomization included patients in the UCLA Health system who underwent thyroid biopsy from August 2017 to January 2020 with indeterminate cytology (Bethesda System for Reporting Thyroid Cytopathology category III or IV).
Molecular testing with the RNA test or DNA-RNA test.
Main Outcomes and Measures
Diagnostic test performance of the RNA test compared with the DNA-RNA test. The secondary outcome was comparison of test performance with prior versions of the molecular tests.
Of 2368 patients, 397 were eligible for inclusion based on indeterminate cytology, and 346 (median [interquartile range] age, 55 [44-67] years; 266 [76.9%] women) were randomized to 1 of the 2 tests. In the total cohort assessed for eligibility, 3140 thyroid nodules were assessed, and 427 (13.6%) nodules were cytologically indeterminate. The prevalence of malignancy was 20% among indeterminate nodules. The benign call rate was 53% (95% CI, 47%-61%) for the RNA test and 61% (95% CI, 53%-68%) for the DNA-RNA test. The specificities of the RNA test and DNA-RNA test were 80% (95% CI, 72%-86%) and 85% (95% CI, 77%-91%), respectively (P = .33); the positive predictive values (PPV) of the RNA test and DNA-RNA test were 53% (95% CI, 40%-67%) and 63% (95% CI, 48%-77%), respectively (P = .33). The RNA test exhibited a higher PPV compared with the prior test version (Afirma gene expression classifier) (54% [95% CI, 40%-67%] vs 38% [95% CI, 27%-48%]; P = .01). The DNA-RNA test had no statistically significant difference in PPV compared with its prior version (ThyroSeq v2 next-generation sequencing) (63% [95% CI, 48%-77%] vs 58% [95% CI, 43%-73%]; P = .75). Diagnostic thyroidectomy was avoided in 87 (51%) patients tested with the RNA test and 83 (49%) patients tested with the DNA-RNA test. Surveillance ultrasonography was available for 90 nodules, of which 85 (94%) remained stable over a median of 12 months follow-up.
Conclusions and Relevance
Both the RNA test and DNA-RNA test displayed high specificity and allowed 49% of patients with indeterminate nodules to avoid diagnostic surgery. Although previous trials demonstrated that the prior version of the DNA-RNA test was more specific than the prior version of the RNA test, the current molecular test techniques have no statistically significant difference in performance.
ClinicalTrials.gov Identifier: NCT02681328
Each year, more than 600 000 thyroid fine-needle aspiration (FNA) biopsies are performed in the United States.1,2 Indeterminate cytology, characterized by the presence of cytologic or architectural atypia without overt nuclear features of papillary thyroid cancer (PTC), is found in 20% of cases.3 Nodules with indeterminate cytology have a 10% to 40% risk of malignancy.4 The usual management for indeterminate thyroid nodules has expanded from either repeat FNA or diagnostic hemithyroidectomy to now include molecular testing.
Approximately one-third of indeterminate thyroid nodules currently undergo molecular testing.5,6 In 90% of patients with benign molecular test results, nodules are managed nonoperatively so more than 25 000 patients per year can avoid diagnostic surgery.7,8 Molecular testing techniques for the diagnosis of indeterminate thyroid nodules are primarily based on either analysis of RNA-based gene expression or detection of somatic mutations.9 The efficacy of analyzing messenger RNA expression using machine learning was demonstrated in 2012 by an RNA expression–based test (Afirma gene expression classifier [Veracyte]), which measured expression of 167 genes by microarray to achieve a diagnostic sensitivity of 92% and specificity of 52%.10 The latest version of the RNA test (Afirma genomic sequencing classifier [Veracyte]) uses next-generation messenger RNA sequencing and is reported to have similarly high sensitivity and improved specificity compared with the prior version.9
DNA-based and RNA-based next-generation sequencing is used to test for specific oncogenic mutations and detect 5 classes of molecular alterations: point mutations, insertions or deletions, gene fusions, copy number alterations, and gene-expression alterations. A previous test panel (ThyroSeq v2 next-generation sequencing [CBLPath]) included 14 genes and 42 gene fusions, and displayed a diagnostic sensitivity of 70% to 96% and specificity up to 77% to 98%.10-18 The current version of the DNA-RNA test (ThyroSeq v3 multigene genomic classifier [CBLPath]) has been expanded to 112 thyroid cancer–related genes and gene-expression alterations.19 The underlying technology of the 2 clinically validated modern molecular diagnostic techniques differs in important ways: (1) RNA-based next-generation sequencing vs DNA-based and RNA-based next-generation sequencing, and (2) use of a machine-learning algorithm yielding a binary result (benign or suspicious) vs detection and reporting of specific mutations with a corresponding numeric risk of malignancy.20,21
A randomized clinical trial compared the diagnostic performance of the prior RNA test and prior DNA-RNA test.22 That DNA-RNA test demonstrated higher specificity compared with the RNA test (66% vs 91%; P = .002). To our knowledge, no studies have compared the diagnostic performance of the current molecular test techniques. This prospective randomized study was performed to compare the test performance of the newer RNA test and DNA-RNA test, and to determine how much the newer versions improved test performance relative to the older platforms.
All patients who underwent thyroid FNA within the UCLA Health system from August 1, 2017, to November 30, 2019, were eligible for enrollment in this randomized clinical trial. Only patients with at least 1 cytologically indeterminate nodule (Bethesda System for Reporting Thyroid Cytopathology category III: atypia of undetermined significance or follicular lesion of undetermined significance; or Bethesda category IV: follicular neoplasm or suspicious for follicular neoplasm) were included for randomization. Additional inclusion and exclusion criteria is available in the study protocol (Supplement 1). This study was approved by the University of California, Los Angeles Institutional Review Board with a waiver for patient informed consent. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.
Radiologists, endocrinologists, and endocrine surgeons performed FNA biopsies across 9 clinical sites. During all FNAs, an additional sample was collected, preserved, and reflexively sent for molecular testing if cytopathology was indeterminate. The FNA specimens were analyzed centrally by a core group of 6 dedicated head and neck cytopathologists (including coauthor J. R.), each with an average of 15 years of experience. Cytologically indeterminate specimens were block randomized by month to undergo testing with the current RNA test and DNA-RNA test. Nodules were excluded if the clinician had preference for one test over the other (ie, nonrandomization), if molecular testing was not performed, or if the patient declined participation. Patients with a history of thyroid cancer or a concurrent biopsy of a separate nodule or lymph node demonstrating thyroid cancer were also excluded.
The pragmatic study design allowed clinicians to determine the management based on clinical judgment and incorporating molecular test results. Histopathology of surgical specimens was reviewed by expert thyroid pathologists who were not blinded to molecular test results. Equivocal cases were reviewed by a second pathologist or discussed in a multidisciplinary case conference. Patients managed nonoperatively were followed with ultrasonography surveillance every 6 to 12 months. Nodule growth was considered significant if there was a greater than 20% increase in 2 nodule dimensions with a minimum 2-mm growth or greater than 50% increase in volume.23 Clinicians subsequently made recommendations for continued surveillance, repeat FNA, or thyroid surgery.
Characteristics analyzed included age, sex, largest nodule dimension measured by ultrasonography, Bethesda category, presence of Hürthle cell predominance on FNA cytology (defined as majority to an exclusive population of cells exhibiting Hürthloid features),24 and pathology reports following surgical resection. Thyroid microcarcinomas (<1 cm) were only considered malignant if located in the same quadrant as the biopsied nodule.
The primary outcomes were the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the RNA test compared with the DNA-RNA test using surgical histopathologic results as the reference standard. Prevalence of malignancy, sensitivity, and NPV were determined with the assumption that nonoperatively managed nodules with negative molecular test results are benign. The results are reported with noninvasive follicular thyroid neoplasm with papillarylike features (NIFTP) grouped with cancer because it represents a premalignant entity that should be managed surgically.25 An analysis considering NIFTP as benign is provided in the eTable in Supplement 2.
Secondary outcomes included diagnostic performance of each test among Hürthle cell–predominant nodules and comparison of the RNA test and DNA-RNA test with their prior versions. Investigation of prior version performance occurred from May 2016 to July 2017 using the same methodology of patient randomization and histopathologic evaluation within the same health care system.22
Sensitivity, specificity, PPV, and NPV of each test were calculated with 95% Wilson confidence intervals (eTable in Supplement 2). Performance comparisons between the RNA test and DNA-RNA test were completed using the χ2 test. A sample of 143 nodules per group would detect a difference in specificity between the RNA test and DNA-RNA test (based on their reported specificities of 68% and 82%, respectively) with 80% power at the 5% level.9,19 Assuming a prevalence of malignancy of 15% and a 10% loss to follow-up, the sample-size calculation was 174 nodules per group. Performance comparisons between the RNA test and DNA-RNA test and their previous versions were similarly calculated using the prevalence of malignancy observed in the current study. Prior version performance characteristics were updated to include the results of surgeries that were performed since prior publication. Expected PPV curves were plotted using observed sensitivities and specificities across the full range of potential prevalence of cancer or NIFTP. P values less than .05 were considered statistically significant. Analyses were performed using SAS, version 9.4 (SAS Institute).
Of 2368 patients assessed for eligibility, a total of 3140 nodules underwent thyroid FNA: 2305 (73.4%) were benign, 427 (13.6%) were indeterminate, 47 (1.5%) were suspicious for malignancy, 217 (6.9%) were malignant, and 144 (4.6%) were nondiagnostic. Of the 427 cytologically indeterminate nodules, 358 (83.8%) nodules were classified as Bethesda category III and 69 (16.2%) nodules as Bethesda category IV. Fifty-five nodules were excluded primarily because of concurrent biopsy of an additional nodule with malignant cytology or clinician preference for nonrandomization (Figure 1). Of 372 cytologically indeterminate nodules in 346 patients included for analysis, 201 (54.0%) were randomized to undergo the RNA test and 171 (46.0%) were randomized to undergo the DNA-RNA test. The baseline characteristics of the groups were similar (Table 1). The median age was 55 (interquartile range [IQR], 44-67) years, and 266 (76.9%) patients were female. The median nodule size was 2.0 (IQR, 1.3-3.0) cm.
The RNA test demonstrated benign results in 107 (53.2%) nodules and suspicious results in 73 (36.3%) nodules. Nineteen (9.5%) samples were insufficient for molecular analysis, and 2 (1.0%) samples were suspicious for parathyroid tissue. Of the benign nodules tested through RNA test, 12 (11.2%) were surgically resected with benign histopathologic results. Of the nodules with suspicious RNA test results, 58 (79.5%) were resected. Histopathologic results revealed NIFTP in 10 (17.2%) nodules and malignancy in 21 (36.2%) nodules (Table 2). Three nodules tested positive for the BRAF malignancy classifier and were diagnosed as classic PTC on histopathologic report.
The DNA-RNA test demonstrated negative results in 103 (60.2%) nodules and positive results in 60 (35.1%) nodules. Seven (4.1%) samples were insufficient for molecular analysis, and 1 (0.6%) sample was suspicious for parathyroid tissue. Of nodules with negative results from the DNA-RNA test, 11 (10.7%) were surgically resected. Ten nodules had benign histopathologic results, while 1 nodule revealed a minimally invasive Hürthle cell carcinoma with capsular invasion only. In this nodule, the initial FNA was paucicellular but was deemed sufficient for DNA-RNA test interpretation. It was resected after 1 year of observation owing to physician preference.
Of the nodules with positive results from the DNA-RNA test, 49 (81.7%) were resected. Histopathologic results revealed NIFTP in 11 (22.4%) nodules and malignancy in 20 (40.8%) nodules (Table 2). Five nodules with BRAF V600E mutations were all PTC. Two nodules with combination TERT or TP53 mutations were minimally invasive follicular carcinoma without angioinvasion and multifocal oncocytic variant PTC, respectively. Nodules with isolated RAS-like mutations were benign in 29.1% (7 nodules), NIFTP in 37.5% (9 nodules), and malignant in 33.3% (8 nodules).
The prevalence of cancer or NIFTP was 19.6% (62 of 317 nodules) among all surgically resected indeterminate nodules and nonoperatively managed nodules with negative molecular testing results. The sensitivities for the RNA test and DNA-RNA test were 100.0% (95% CI, 88.8%-100.0%) and 96.9% (95% CI, 83.8%-99.9%), respectively (P > .99) (Table 3). The specificities for the RNA test and DNA-RNA test were 79.6% (95% CI, 71.7%-86.1%) and 84.8% (95% CI, 77.0%-90.7%), respectively (P = .32). The PPV for the RNA test and DNA-RNA test were 53.5% (95% CI, 39.9%-66.7%) and 63.3% (95% CI, 48.3%-76.6%), respectively (P = .33). Performance characteristics considering NIFTP as benign are reported in the eTable in Supplement 2. A total of 168 patients (48.6% of all patients with indeterminate nodules) with negative molecular testing results avoided diagnostic surgery.
The PPV of the RNA test was higher compared with the prior test version (53.5% [95% CI, 39.9%-66.7%] vs 37.5% [95% CI, 27.0%-48.1%]; P = .01; Figure 2). The RNA test had improved specificity compared with its prior version (79.6% [95% CI, 71.7%-86.1%] vs 65.2% [95% CI, 49.8%-78.7%]; P = .07). The PPV of the DNA-RNA test and its prior version showed no statistically significant difference (63.3% [95% CI, 48.3%-76.6%] vs 57.8% [95% CI, 42.6%-72.9%]; P = .75). The specificity of the DNA-RNA test also showed no statistically significant difference to that of its prior version (84.8% [95% CI, 77.0%-90.7%] vs 85.5% [95% CI, 75.0%-92.8%]; P > .99). The benign call rates of the RNA test and its prior version had no statistically significant differences (53.2% [95% CI, 46.8%-60.6%] vs 42.9% [95% CI, 53.0%-68.0%]; P = .17); however, the benign call rate of the DNA-RNA test was lower than that of its prior version (60.2% vs 77.2%; P = .01).
The rate of Hürthle cell cytology was 13.7% (51 nodules), including 28 (54.9%) nodules tested with the RNA test and 23 (45.1%) nodules tested with the DNA-RNA test. The RNA test demonstrated benign results in 20 nodules (benign call rate, 71.4%), suspected malignancy in 6 nodules (21.4%), and insufficient data in 2 nodules (7.1%). Three nodules that tested benign with the RNA test were resected with benign histopathologic results, and 5 nodules with suspicious RNA test results were resected with 1 malignant result (classic PTC). The DNA-RNA test demonstrated negative results in 11 nodules (benign call rate, 47.8%), positive results in 11 nodules (47.9%), and insufficient data in 1 nodule (4.3%). Four nodules with negative DNA-RNA test results were resected with benign histopathologic results, while 10 nodules with positive DNA-RNA test results were resected with 2 malignant results (1 oncocytic PTC and 1 minimally invasive Hürthle cell carcinoma). The RNA test had a sensitivity of 100.0%, specificity of 83.3% (95% CI, 62.6%-95.3%), PPV of 20.0% (95% CI, 0.5%-71.6%), and NPV of 100.0%. The DNA-RNA test had a sensitivity of 100.0%, specificity of 57.9% (95% CI, 33.5%-79.8%), PPV of 20.0% (95% CI, 2.5%-55.6%), and NPV of 100.0%.
A total of 187 indeterminate nodules with negative molecular test results (RNA test, 94; DNA-RNA test, 93) were observed. Surveillance ultrasonography was performed for 90 nodules over a median follow-up of 12.2 (IQR, 8.1-17.9) months. Only 5 nodules exhibited significant growth. Four patients (1 tested with the RNA test and 3 tested with DNA-RNA test) underwent surgery after a median surveillance period of 13.2 (IQR, 10.7-16.0) months. Histopathologic results revealed 3 nodules as benign and 1 nodule tested with the DNA-RNA test as the minimally invasive Hürthle cell carcinoma previously described.
This pragmatic randomized clinical trial showed no statistically significant difference in the performance of the RNA test and DNA-RNA test for predicting malignancy in cytologically indeterminate thyroid nodules. The tests demonstrated high sensitivity (97%-100%) and reasonably high specificity (80%-85%). The benign call rates had no statistically significant differences (53% for the RNA test vs 61% for the DNA-RNA test). Only 11% of patients with benign molecular test results underwent surgery, translating into a total of 168 of 346 patients (48.6%) who avoided diagnostic surgery. With a prevalence of malignancy of 20% among indeterminate nodules in the present cohort, the 2 molecular tests had no statistically significant difference in PPV (55% for the RNA test vs 63% for the DNA-RNA test).
Both the RNA test and DNA-RNA test are the most highly validated current molecular tests for the diagnosis of indeterminate thyroid nodules. The RNA test platform evolved from microarray analysis of messenger RNA expression to next-generation sequencing of the RNA transcriptome with the addition of several cassettes (aimed to identify specific pathologic entities) outside of the core machine-learning algorithm.9 These advancements improved specificity, especially in the diagnosis of Hürthle cell aspirates, while maintaining high sensitivity. The DNA-RNA test began as a 7-gene panel examining the mutations most commonly associated with thyroid cancer. These included 4 point mutations (BRAF, HRAS, KRAS, and NRAS) and 3 gene rearrangements (RET/PTC1, RET/PTC3, and PAX8/PPARγ). The previous version of the DNA-RNA test used next-generation sequencing to examine an expanded panel of 14 genes analyzed for point mutations and 42 types of gene fusions, with a small RNA expression component used to characterize cell lineage. The DNA-RNA test examines 112 thyroid cancer–related genes and includes a substantial RNA expression component. The first version of the DNA-RNA test functioned as a rule-in test because of its high PPV but lacked sensitivity owing to the limited number of genes assessed. The progressively expanded gene panel has increased the sensitivity of the DNA-RNA test.
The specificity of the RNA test was higher in the present study compared with the prior blinded validation study of RNA expression–based testing performed on 190 thyroid nodules with indeterminate cytology.9 With a 25% prevalence of malignancy, the RNA test was reported to have a sensitivity of 91% and specificity of 68%. The present DNA-RNA test results are consistent with the prior prospective, blinded multicenter study including 257 indeterminate nodules. With a relatively high prevalence of malignancy of 28%, the DNA-RNA test had a sensitivity of 94% and specificity of 82%.19 Both of the prior studies included operative management of all nodules with indeterminate cytology, while only 11% of nodules with benign molecular testing were resected in this study, revealing 1 false-negative result.
The dependence of test performance on prevalence of malignancy limits comparison of molecular test techniques across studies. The present study allows head-to-head comparison of test performance in a single institution with a homogeneous prevalence of malignancy and centralized, high-volume cytopathology interpretation. A prior randomized clinical trial22 compared the test performance of the earlier test versions, in which we reported a lower specificity of RNA testing compared with DNA-RNA testing (66% vs 91%). The current findings are consistent with other reports demonstrating an increase in the benign call rate and specificity of the RNA test (benign call rate, 63%-89%; specificity, 92%-100%) as compared with its prior version (benign call rate, 16%-26%; specificity, 33%-43%) for Hürthle cell nodules.26-28
Although the prior DNA-RNA test had the potential advantage of increased specificity compared with the RNA test, the sensitivity was not well established owing to the absence of a clinical validation study with all nodules undergoing surgery with histopathologic evaluation. The relatively low rate of positive mutations identified raised the possibility of false-negative cases.11,22,29 The DNA-RNA test examines an expanded panel of molecular alterations, resulting in an expected decrease in the benign call rate from 77% to 61% with maintenance of high specificity greater than 80%. The present findings are consistent with prior reports of the DNA-RNA test performance and support its use as a rule-out test for malignancy.19,30,31
The expanded gene mutation panel in the DNA-RNA test has enhanced use to diagnose and prognosticate thyroid cancer based on the specific molecular alteration(s) detected. High-risk mutations including TERT, TP53, and BRAF V600E confer an almost 100% risk of malignancy and are more commonly associated with high-risk histopathologic features, increased risk of recurrence, and disease-specific mortality.32-36 Thyroid nodules with isolated RAS mutations are generally either benign, NIFTP, or low-risk cancers such as minimally invasive follicular thyroid cancer or follicular variant of PTC.11,37 In a recent study of 80 RAS-positive indeterminate thyroid nodules, 59% were benign, 13% were NIFTP, and 29% were low-risk cancers on histopathologic reports.38 The present results are consistent with prior reports in that all nodules with high-risk combination or BRAF V600E mutations were malignant, while approximately one-third of nodules with isolated RAS mutations were benign, NIFTP, and malignant, respectively.
This study is limited by nonoperative management for most indeterminate nodules with benign molecular test results. Lack of histopathologic confirmation may prevent identification of false-negative cases, which would decrease the sensitivity and NPV of the tests. However, prior validation studies,9,19 including operative management of all indeterminate nodules, reported high sensitivity (91% for RNA testing vs 94% for DNA-RNA testing) and very low false-negative rates (2% for RNA testing vs 3% for DNA-RNA testing), which supports the high sensitivity (100% for RNA testing vs 97% for DNA-RNA testing) found in this study. In addition, we randomized patients rather than performing both molecular tests in all samples. Although having both molecular test results for each sample would allow for a direct comparison of test performance, such a study design was cost prohibitive. Pathologists were not blinded to molecular test results, which may have biased the histopathologic diagnosis. Finally, test performance is dependent on the prevalence of malignancy, which may affect the PPV and cost-effectiveness of molecular testing in different practice settings.
This direct comparison of an RNA test and DNA-RNA test showed no statistically significant difference in sensitivity and specificity for diagnosis of malignancy in cytologically indeterminate thyroid nodules. In this pragmatic study allowing for clinical decision-making, half of the patients with indeterminate cytology avoided diagnostic surgery on the basis of a benign molecular test result. In light of these findings, the choice of molecular test may hinge on factors other than diagnostic performance, such as cost, processing time, sample inadequacy rate, and information regarding specific mutations that may guide future treatment.
Accepted for Publication: September 15, 2020.
Published Online: December 10, 2020. doi:10.1001/jamaoncol.2020.5935
Corresponding Author: Masha J. Livhits, MD, Section of Endocrine Surgery, David Geffen School of Medicine at UCLA, 10833 Le Conte Ave, 72-228 CHS, Los Angeles, CA 90095 (email@example.com).
Author Contributions: Dr Livhits had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Livhits, Kuo, Rao, Gofnung, Smooke-Praw, Yeh.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Livhits, Zhu, Nguyen, Kim.
Critical revision of the manuscript for important intellectual content: Livhits, Zhu, Kuo, Kim, Tseng, Leung, Rao, Levin, Douek, Beckett, Cheung, Gofnung, Smooke-Praw, Yeh.
Statistical analysis: Zhu, Kuo, Nguyen, Kim, Tseng.
Obtained funding: Yeh.
Administrative, technical, or material support: Zhu, Kuo, Nguyen, Rao, Levin, Beckett, Cheung, Gofnung, Smooke-Praw, Yeh.
Supervision: Livhits, Zhu, Cheung, Gofnung, Smooke-Praw, Yeh.
Conflict of Interest Disclosures: Dr Rao reports receiving personal fees from AstraZeneca and other support from Oncogenesis, 3J Biotech, and Zhiwei. No other disclosures were reported.
Funding/Support: This study was supported by the Viola G. Hyde Scholarship Fund Research Award (Dr Livhits), the H & H Lee Surgical Research Grant (Drs Kuo and Zhu), and the Garry Shandling estate (Dr Yeh).
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Data Sharing Statement: See Supplement 3.
Additional Contributions: We thank UCLA Health and our community clinics for their collaboration in the coordination of the clinical trial. No compensation was received for their contributions.