Diagnostic Performance of Afirma and Interpace Diagnostics Genetic Testing in Indeterminate Thyroid Nodules: A Single Center Study

Simple Summary Molecular testing can stratify the risk of malignancy among indeterminate thyroid nodules and subsequently reduce the need for diagnostic surgery. We assessed the performance of the GSC + XA; ThyGeNEXT + ThyraMIR, and GSC + GEC molecular platforms to elucidate their diagnostic accuracy. All three platforms were reliable for ruling out thyroid cancer, with a tendency for the newest Afirma genetic testing platform (GSC + XA) to outperform the other platforms. Abstract Indeterminate thyroid nodules (ITN) represent 20–30% of biopsied nodules, with a 10–60% risk of malignancy. Molecular testing can stratify the risk of malignancy among ITNs, and subsequently reduce the need for unnecessary diagnostic surgery. We aimed to assess the performance of these molecular tests at a single institution. Patients with Bethesda III, IV, and V nodules with Afirma and Interpace Diagnostics genetic testing data from November 2013 to November 2021 were included. Three cohorts were formed, including GSC + XA, ThyGeNEXT + ThyraMIR, and GSC + GEC. Statistical analysis determined the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), diagnostic odds ratio (DOR), and accuracy of each type of testing. The PPV of nodules undergoing genetic testing by ThyGeNEXT + ThyraMIR (45.00%, 95%CI: 28.28–62.93%, p = 0.032) and GSC + XA (57.14%, 95%CI: 29.32–81.08%, p < 0.001) were superior to that of GEC + GSC (30.72%, 95%CI: 26.83–34.90%). The NPV was above 85% in all cohorts, suggesting overall suitable rule-out tests. The Afirma platform (GSC + XA) had the highest NPV at 96.97%. The overall accuracy for nodules undergoing ThyGeNEXT + ThyraMIR was 81.42% (95%CI: 73.01–88.11%, p < 0.001). A total of 230 patients underwent thyroidectomy, including less than 60% of each of the ThyGeNEXT + ThyraMIR and GSC + XA cohorts. Specifically, only 25% of patients in the GSC + XA cohort underwent surgery, considerably decreasing the rate of unnecessary surgical intervention. Sub-group analysis, including only patients with surgical pathology, found that PPV tended to be higher in the GSC + XA cohort, at 66.67% (95%CI: 37.28–87.06%), as compared to the ThyGeNEXT + ThyraMIR cohort, at 52.94% (95%CI: 35.25–69.92%). The Afirma genetic testing platform GSC + XA outperformed the other platforms with regards to both PPV and NPV and decreased the rate of surgery in patients with ITNs by 75%, significantly preventing unnecessary surgical intervention.


Introduction
Thyroid nodules are common, prevalent in up to two-thirds of the general population [1]. While most thyroid nodules are benign, they can represent concerning pathology which may significantly adversely impact patient survival. As a result, the accurate assessment of thyroid nodules is important.
Thyroid nodule assessment typically includes sampling by fine-needle aspiration (FNA). Current (2015) American Thyroid Association (ATA) guidelines recommend cytopathological evaluation with FNA biopsy of thyroid nodules that meet specific criteria, such as solid nodules greater than 1 cm, complex nodules greater than 1.5 cm, and thyroid nodules with suspicious features of malignancy on ultrasound [2]. FNA is a safe and minimally-invasive procedure, typically performed under sonographic guidance, allowing for cytopathological assessment of the aspirate. The management of thyroid nodules relies principally on a cytopathological diagnosis, which is reported using The Bethesda System for Reporting Thyroid Cytology (TBSRTC) [3]. Still, a considerable portion of thyroid nodules that undergo FNA are classified as indeterminate (20-25%), of which 10-60% of these are malignant [3].
Indeterminate thyroid nodules (ITNs) include Bethesda Class III, atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS), and Class IV, follicular neoplasm or suspicious for a follicular neoplasm (FN/SFN) [3]. Some literature includes Bethesda Class V, suspicious for malignancy, as well. Since these nodules are neither benign nor malignant on preoperative assessment, ITNs pose a challenge for clinical management [4]. Options for management include repeat FNA, molecular testing, patient monitoring, minimally-invasive ablative techniques, and hemithyroidectomy [3,4].
Advancement in thyroid tumorigenesis genetics has allowed the development of molecular testing to complement cytological diagnosis and improve the risk-based stratification of ITNs [5]. Molecular testing platforms are currently implicated to refine the preoperative diagnosis of ITNs by reporting a refined patient risk stratification, subsequently reducing the need for diagnostic surgery [5,6]. Two common molecular testing platforms include those developed by Afirma (San Francisco, CA, USA) and Interpace Diagnostics (Parsippany, NJ, USA), both of which have been described as good rule-out tests due to a high negative predictive value (NPV) [7]. The use of these genetic tests can assist clinicians and surgeons in patient risk stratification and management planning [7].
Interpace Diagnostics utilizes a genotyping panel, ThyGeNEXT, as well as a molecular test with a high NPV for thyroid cancer which utilizes a microRNA-based assay, ThyraMIR. MicroRNAs are small, noncoding ribonucleic acid molecules that play a role in regulating cellular function including cell cycle gene expression, proliferation, and survival [8]. Samples first undergo ThyGeNEXT evaluation, and if not determined to possess a strong/driver mutation, subsequently undergo further evaluation by ThyraMIR. "Strong" driver mutations include those with a strong probability of cancer, including the BRAF V600E mutation, TERT promoter mutations, and ALK mutations. The Interpace Diagnostics molecular testing platforms ThyGeNEXT and ThyraMIR possess a high NPV (95%) and are used clinically to rule-out thyroid cancer [9].
Veracyte Inc. (San Francisco, CA, USA) developed the Afirma Gene Expression Classifier (GEC), a microarray-based test with a proprietary algorithm that examines the mRNA expression of a panel of 167 genes. This molecular test was validated clinically in a blinded prospective multicenter trial including 265 thyroid nodules [10]. In 2018, Afirma developed and published its validation study with the improved Gene Sequencing Classifier (GSC) [11]. In addition to GSC evaluation, which measures RNA gene expression levels, Afirma developed the Xpression Atlas (XA) which can detect sequence variants, such as gene fusions, insertions, and deletions, as well as point mutations. The use of the latest Afirma technology demonstrated an impressive NPV of 96% [11]. The use of both the GSC and XA among ITNs provides valuable insight for clinical decision-making [12].
While several works have reported their experience with genetic testing platforms, there remain only a few works which investigate different genetic testing platforms from a single institution. For example, one work which analyzed three commercial testing platforms (Afirma, RosettaGX, and Interpace) reported only on a total of 70 nodules. The authors reported that all genetic testing platforms displayed high NPV [13]. Works which analyze only a single genetic testing platform are subject to inherent biases when compared to other single-institution works, considering the variation in cytopathologists' assessments (i.e., intra-rater as well as inter-rater agreement), as well as the prevalence of thyroid cancer and mutations, which fundamentally alter the positive predictive value (PPV) and the NPV of these molecular tests. Considering these limitations, we performed a comparative analysis of molecular performance tests at a single institution. Our primary aim in this study was to assess the diagnostic performance of the latest Afirma (GSC + XA) and Interpace Diagnostics (ThyGeNEXT + ThyraMIR) platforms.

Study Design
After obtaining institutional review board (IRB) approval from Tulane Medical Center, we reviewed all results of thyroid biopsies from patients with ITNs. All patients underwent FNA by a single, high-volume, fellowship-trained endocrine surgeon. Patients with ITNs according to a cytopathological analysis using Afirma and Interpace Diagnostics molecular testing data from November 2013 to November 2021 were included.

Ultrasound Assessment
All patients with thyroid nodules underwent comprehensive neck ultrasound evaluation. Ultrasound assessment was conducted using a 15-MHz linear transducer. Nodules were assessed by evaluating the three-dimensional diameter, composition, echogenicity, margins, echogenic foci, vascularity, and elastography of the nodules. Suspicious nodules underwent further evaluation with FNA.

Fine-Needle Aspiration
FNA biopsies were conducted utilizing a 10 mm 25-gauge needle with an uninterrupted real-time observation of the needle tip through each nodule via ultrasound guidance. The lower neck was prepared with povidone-iodine swabs and draped in a standard surgical fashion. A total of 2 mL of 1% lidocaine was injected as local anesthesia. Lidocaine cream (EMLA cream 5%) was used prior to injections. In the same FNA setting, samples were also collected for molecular testing in specified preservation tubes from Afirma or Interpace Diagnostics. Samples were subsequently sent for laboratory analysis to a cytopathologist at our institution. Consistent with most academic centers, further analysis by molecular testing was up to the discretion of the pathologist. At our practice, the vast majority of thyroid nodules which are not classified as benign (Bethesda I/III-VI) are sent for molecular testing. For the minority of patients who elect to undergo thyroidectomy, regardless of biopsy cytological analysis, genetic testing was not sent. The decision to send the biopsy specimen to Afirma or Interpace Diagnostics for evaluation by genetic testing was also up to the pathologist. The tendency to send nodules to Afirma for genetic testing can be explained by the fact that different providers at our institution each have their own preferences regarding their test of choice. This reflects the true nature of an endocrine practice. As a whole, however, pathologists did not have a preference for either genetic testing platform and sent specimens without any particular arrangement. The majority of the FNA were performed at our out-patient clinic, and a few were performed under general anesthesia in the operating room, per patient preference. FNA sites were matched to the respected nodules to the best of the pathologist's knowledge.

Afirma Molecular Testing
Following evaluation by a cytopathologist, Bethesda III, IV, and V nodules were sent for further molecular testing. Implementation of XA use at our institution began in May 2018, when all patients began receiving the GSC and its complementary XA. FNA samples sent for genetic testing prior to this time-point were subject to either GEC or GSC testing alone. Specimen collection and shipping were in accord with Afirma's recommendations. Afirma requires that additional aspirates be procured and submitted in special collection media and additionally requires that specimens be refrigerated and/or frozen prior to shipping [14].

Interpace Molecular Testing
Following evaluation by a cytopathologist, ITNs classified as Bethesda Class I, III, IV, and V nodules were sent for further molecular testing. Molecular testing was performed only for those patients with aspirates subject to the ThyGeNEXT panel, with its corresponding ThyraMIR panel. Patients receiving ThyGenX, which was introduced in 2014 and assesses for 100 genetic alterations across 8 different oncogenes [13], were excluded from this cohort, as there were only a few patients, and this technology is no longer in use. ThyGeNEXT was introduced in 2018 and includes additional markers. Specimen collection and shipping were in accord with Interpace Diagnostic's recommendations [15]. Since Interpace Diagnostics also offers the ability to test nondiagnostic FNA specimens, Bethesda I specimen samples were also shipped.
All pathology reports were read by a single pathology department before being sent out for genetic testing by Afirma or Interpace Diagnostics. The results of molecular testing were identified using the Veracyte physician portal for Afirma results or the Interpace Diagnostics physician portal for Interpace Diagnostics results. Afirma molecular testing included the GEC, GSC, and XA. Interpace diagnostics molecular testing included Thy-GeNEXT and ThyraMIR. Molecular testing results were then copied directly from the physician portals to a predesigned spreadsheet. Patients with insufficient genetic testing were removed. Electronic medical records were used to extract demographic information, nodule characteristics, and final pathology results associated with each thyroid nodule.

Cohorts
Considering that the primary aim of our study was to assess the diagnostic performance of the latest Afirma and Interpace Diagnostics platforms, we sub-grouped the study population by the molecular test that the patients received. Three cohorts were determined, including GSC + XA (the latest Afirma platform), ThyGeNEXT + ThyraMIR (a recent Interpace Diagnostics platform, only recently updated in 2022 with ThyraMIR v2 [16]), and GSC + GEC (Afirma testing platform prior to the introduction of the XA panel). The GSC has been proven superior to the GEC, including a higher benign-call rate (61.2% vs. 41.6%), and it was not further sub-stratified into individual cohorts, as this was not the primary concern of the study, and it has been reported on previously [11,17,18].

Statistical Analysis
IBM SPSS Statistics 27 was used for statistical analysis. A two-sided Chi-square test was used for analysis, and the p-values < 0.05 were set to be significant. Since clinician decision making and patient preference both play an important role in our practice, not all patients with ITNs underwent thyroidectomy. Therefore, we conducted an overall analysis including the whole study population using surgical pathology, if available, and conservative management follow-up decision making, if surgical pathology was not available. All patients not undergoing thyroidectomy were followed-up for at least 6 months to determine if there were any suspicious features or changes on ultrasound images. Typical criteria during the management of active surveillance patients, including nodule growth of more than 50% or suspicious lymph node involvement, were considered. Patient nodules not exhibiting these features were considered benign, while patients exhibiting these fea- tures were considered to have malignant nodules. While this reflects true clinical decision making, as well as the nature of an endocrine practice, we also conducted a sub-group analysis including only patients undergoing thyroidectomy who consequently have available surgical pathology to determine malignancy status. For Afirma, a benign genetic test was read as a negative test, while a suspicious test was read as a malignant test. For Interpace Diagnostics, low-risk was read as a benign test and an intermediate risk or high risk was read as a malignant test. The sensitivity, specificity, PPV, NPV, diagnostic odds ratio (DOR), positive likelihood ratio, negative likelihood ratio, and accuracy were determined. True positives were those nodules determined as suspicious by molecular testing and which were subsequently proven malignant by surgical pathology. True negatives were those nodules determined as benign by molecular testing and which were subsequently proven benign by surgical pathology. False positives were nodules determined as suspicious by molecular testing and which were subsequently proven benign by surgical pathology. False negatives were those nodules determined as benign by molecular testing and which were subsequently proven malignant by surgical pathology. Meta-disc version 4.0 was used for the analysis of the diagnostic accuracy measures and comparison between genetic testing platforms.

Study Population
A total of 517 FNA biopsies were performed. Of these, 109 were non-diagnostic and were sent for Interpace Diagnostics testing, which is beyond the scope of this study. This allowed for a total of 408 ITNs for our study population, including 113 in the ThyGeNEXT + ThyraMIR cohort, 255 in the GEC + GSC cohort, and 40 in the GSC + XA cohort ( Figure 1). clinical decision making, as well as the nature of an endocrine practice, we a sub-group analysis including only patients undergoing thyroidecto quently have available surgical pathology to determine malignancy statu benign genetic test was read as a negative test, while a suspicious test w lignant test. For Interpace Diagnostics, low-risk was read as a benign test diate risk or high risk was read as a malignant test. The sensitivity, specifi diagnostic odds ratio (DOR), positive likelihood ratio, negative likelihood racy were determined. True positives were those nodules determined a molecular testing and which were subsequently proven malignant by sur True negatives were those nodules determined as benign by molecular te were subsequently proven benign by surgical pathology. False positives w termined as suspicious by molecular testing and which were subsequentl by surgical pathology. False negatives were those nodules determined a lecular testing and which were subsequently proven malignant by sur Meta-disc version 4.0 was used for the analysis of the diagnostic accurac comparison between genetic testing platforms.

Sub-Group Analysis including Nodules Undergoing Surgery
A total of 230 patients underwent thyroidectomy and accordingly, had available surgical pathology. The surgical rate per cohort was significantly different (p < 0.001) by cohort. The rates of thyroidectomy in the ThyGeNEXT + ThyraMIR and GSC + XA cohorts were 36.3% and 25.0%, respectively ( Figure 3). Conversely, 70.2% of the GEC + GSC cohort underwent thyroidectomy. The overall rate of malignancy in this sub-group was 36.52% (N = 84/230), with similar rates of malignancy between the cohorts (p > 0.05). Considering the low conversion rate to thyroidectomy in the ThyGeNEXT + ThyraMIR and GSC + XA cohorts, the count of nodules with surgical pathology was 41 and 10, respectively. The GEC + GSC cohort included a total of 179 nodules.

Sub-Group Analysis including Nodules Undergoing Surgery
A total of 230 patients underwent thyroidectomy and accordingly, had available surgical pathology. The surgical rate per cohort was significantly different (p < 0.001) by cohort. The rates of thyroidectomy in the ThyGeNEXT + ThyraMIR and GSC + XA cohorts were 36.3% and 25.0%, respectively ( Figure 3). Conversely, 70.2% of the GEC + GSC cohort underwent thyroidectomy. The overall rate of malignancy in this sub-group was 36.52% (N = 84/230), with similar rates of malignancy between the cohorts (p > 0.05). Considering the low conversion rate to thyroidectomy in the ThyGeNEXT + ThyraMIR and GSC + XA cohorts, the count of nodules with surgical pathology was 41 and 10, respectively. The GEC + GSC cohort included a total of 179 nodules.

Discussion
Thyroid nodules are common and are being detected at increased rates over the past few decades [1,19]. Nodules represent a wide spectrum of pathology, including benign nodules, as well as high-risk carcinomas, which adversely affect patient survival. Standard first-line assessment of thyroid nodules includes preoperative ultrasound and, typically, FNA [2]. FNA cytological analysis heavily guides clinical judgment, although its diagnostic accuracy has not improved within recent decades [20]. Molecular testing enhances patient risk stratification. In our study, all three cohorts, including GSC + XA, ThyGeNEXT + ThyraMIR, and GSC + GEC, were effective in ruling-out thyroid cancer. Specifically, the use of ThyGeNEXT + ThyraMIR and GSC + XA decreased the surgical rate in patients with ITNs by more than 60%.
Surgery remains the mainstay for establishing a definitive histopathological diagnosis, especially among ITNs. Still, surgery carries the concerning risk of complications which impact patient quality of life, such as dysphonia and permanent hypothyroidism [21]. In a recent review, Schneider et al. reported that approximately 75% of ITNs undergoing thyroidectomy are benign on surgical pathology, suggesting unnecessary surgical intervention [22]. In their 60-patient ITN cohort study, Balentine et al. reported only a 6% rate of malignancy in surgical pathology [23]. Importantly, the authors noted that 47% of patients develop hypothyroidism following thyroidectomy [23]. Accordingly, the

Discussion
Thyroid nodules are common and are being detected at increased rates over the past few decades [1,19]. Nodules represent a wide spectrum of pathology, including benign nodules, as well as high-risk carcinomas, which adversely affect patient survival. Standard first-line assessment of thyroid nodules includes preoperative ultrasound and, typically, FNA [2]. FNA cytological analysis heavily guides clinical judgment, although its diagnostic accuracy has not improved within recent decades [20]. Molecular testing enhances patient risk stratification. In our study, all three cohorts, including GSC + XA, ThyGeNEXT + ThyraMIR, and GSC + GEC, were effective in ruling-out thyroid cancer. Specifically, the use of ThyGeNEXT + ThyraMIR and GSC + XA decreased the surgical rate in patients with ITNs by more than 60%.
Surgery remains the mainstay for establishing a definitive histopathological diagnosis, especially among ITNs. Still, surgery carries the concerning risk of complications which impact patient quality of life, such as dysphonia and permanent hypothyroidism [21]. In a recent review, Schneider et al. reported that approximately 75% of ITNs undergoing thyroidectomy are benign on surgical pathology, suggesting unnecessary surgical intervention [22]. In their 60-patient ITN cohort study, Balentine et al. reported only a 6% rate of malignancy in surgical pathology [23]. Importantly, the authors noted that 47% of patients develop hypothyroidism following thyroidectomy [23]. Accordingly, the importance of accurate pre-operative risk stratification by molecular testing cannot be understated.
Molecular tests are increasingly utilized as ancillary tools to avoid diagnostic surgery as an approach for cytologically ITNs. For example, Angell et al. reported on 600 nodules and found a surgery rate of 45.4% (221/486) and 28.1% (32/114) in the GEC and GSC cohorts, respectively [17]. While our study stratified cohorts slightly differently, we also found that the newer Afirma technology (GSC + XA) significantly decreased the rate of surgery in these patients. The authors reported a PPV of 33.9% (75/221) with the GEC and 50% (16/32) with the use of the GSC [17]. The PPV using ThyGeNEXT + ThyraMIR is reported to be 52% [9]. In our study population, we reported similar PPVs of 45% and 57.14% using ThyGeNEXT + ThyraMIR and GSC + XA, respectively. These values understandably increased to 52.94% and 66.67%, respectively, when sub-grouped by patients who underwent surgery. Our work suggests a similar PPV to that of a 2022 meta-analysis including 13 studies evaluating the diagnostic performance of the Afirma GSC, reporting a PPV of 64.9% [24]. The NPV of ThyGeNEXT + ThyraMIR and Afirma GSC are 95% and 96%, respectively [9]. We found similar NPVs, with 89.25% and 96.97%, in our study population which decreased to 62.50% and 80%, when including only patients who underwent surgery. The increase in PPV and decrease in NPV can be explained considering that patients with higher risks of malignancy (i.e., suspicious sonographic features or increases in nodule size) were more likely to undergo thyroidectomy. The increase in PPV and decrease in NPV can be explained by considering two factors. The first is that the prevalence of malignancy in our patient population is higher than that of the general population. We previously published these findings [25,26], reporting a prevalence of malignancy in ITNs which is at the upper-end and/or higher than ranges reported from other institutions (10-19% at Mount Sinai Beth Israel, 30-38% at Memorial Sloan Kettering Cancer Center) [27]. The second, which is related to the first, is that patients with higher risks of malignancy (i.e., suspicious sonographic features or increases in nodule size) were more likely to undergo thyroidectomy. Importantly, this allowed a malignancy rate of 36.52% in patients with ITNs, much higher than the 6% reported by Balentine et al. [23], demonstrating the importance regarding the selection of patients with suspicious nodules, and the obviation of surgery in patients who may not have required it. With NPVs of 89% and higher in nodules undergoing genetic testing with either of these platforms, our work corroborates the current literature suggesting both ThyGeNEXT + ThyraMIR and GSC + XA to be effective rule-out genetic tests. Specifically, the NPV of GSC + XA (96.97%) tended to be higher than that of ThyGeNEXT + ThyraMIR (89.25%) genetic testing. It is the opinion of the authors that genetic testing should be a first-line option for patients with ITNs.
FNA is widely performed to allow for a minimally-invasive mechanism of determining nodular cytology. A recent 2022 meta-analysis including 16,697 patients (36 studies) found that the sensitivity and specificity of FNA was 85.6% (95%CI:, 79.9-89.5%) and 71.4% (95%CI: 61.1-79.8%), respectively [20]. While the diagnostic accuracy of FNA is high, the authors noted that the accuracy of FNA has not increased over time, including studies published as far back as 1982 [20]. This primarily highlights the importance of accurate genetic testing, but also the importance of clinical decision making. For example, a patient with a Bethesda III nodule which is only 1 cm and without suspicious sonographic features may be suitable for non-surgical management and close patient monitoring. Accordingly, our population-level analysis determined by conservative management follow-up or final surgical pathology, if available, possesses merit, as it reflects a true endocrine practice.
Several reviews have discussed the strengths and weaknesses of each of these genetic testing platforms beyond their diagnostic accuracy [7,9]. While details of these discussions are beyond the scope of this work, including the number of genes assessed by gene expression analysis or the number of variants assessed, several factors are worth mentioning. Foremost, there are financial considerations regarding genetic testing. Since thyroidectomy is costly (considering operating room time, staffing fees, equipment, time away from work for the patient, complications, etc.), the use of molecular testing has been suggested to be a cost-effective management option [28,29], although other works still suggest the opposite [30,31]. Determining which genetic testing platform is the most cost-effective is complicated, given that some patients may require additional testing, which is purchased separately. For example, the cost of BRAF mutation genetic testing alone is USD 475 USD, while the MTC genetic testing alone is 975 USD from Afirma [32]. Additionally, different works suggest different rates of surgery following genetic testing, further blurring a transparent cost-effective analysis. In our study, the use of either ThyGeNEXT + Thyra-MIR or GSC + XA decreased the rate of thyroidectomy by more than 60%. In 2019, one study found that the use of GSC decreased the rate of thyroidectomy by 45% compared with that for GEC alone [18]. Specifically, our work found that Afirma's GSC + XA decreased the rate of surgery to 25%, which was more than that of Interpace Diagnostic's ThyGeNEXT + ThyraMIR (which decreased the rate to 36.3%). Recently, Sunoco et al. incorporated sonographic features in their cost-effective analysis and determined that molecular testing was only cost-effective for ITNs with either an intermediate or a low suspicion of malignancy [33]. Nicholson et al. also reported that molecular testing, specifically either GSC or ThyroSeq version 3, is considerably more cost-effective than diagnostic lobectomy for patients with ITNs who lack other indications for thyroidectomy [29]. Future works incorporating nodule sonographic features, similar to the work of Zanocco et al., as well as considering the different commercially available genetic tests, are warranted to better elucidate whether genetic testing is more cost-effective than diagnostic lobectomy and to facilitate the comparison of molecular platforms.
One advantage of the utilization of the Interpace Diagnostics genetic testing platform is their ease of specimen collection. Interpace Diagnostics does not require a dedicated FNA, but rather samples can be collected as cells on a direct smear or as cells/washout collected from an FNA pass [11]. Conversely, Afirma requires a dedicated two-pass FNA [9]. Therefore, the use of the Interpace Diagnostics platform may be more appropriate in patients who experience considerable discomfort during FNA biopsy.
There are limitations to our study. Foremost, the retrospective study design of the work is subject to inherent biases. Furthermore, the authors acknowledge that the populationlevel analysis including patients monitored for at least 6 months, but without surgical pathology, may artificially inflate the true negative rate. In consequence, the sub-group analysis was performed including only patients with surgical pathology. Another minor limitation of this study was our obviation of patients analyzed by the newest Interpace Diagnostics risk classifier, ThyraMIR v2 [16]. This technology was only recently introduced in 2022 and aims to further improve ITN risk stratification. One strength of this work is its investigation of different genetic testing platforms within the same institution.

Conclusions
ThyGeNEXT + ThyraMIR and GSC + XA are genetic tests with effective rule-out abilities. The use of these genetic testing platforms, specifically ThyGeNEXT + ThyraMIR and GSC + XA, decreased the surgical rate in patients with ITNs by more than 60%.