ThyroSeq v2 Testing: Impact on Cytologic Diagnosis,Management,and Cost of Care in Patients with Thyroid Nodule

Abstract

Background:

Novel molecular tests (MTs), such as ThyroSeq, may improve the management of thyroid nodules with indeterminate cytologic diagnoses; however, the impact of these tests on cost and outcome of management is unknown. Here, we evaluated the impact of ThyroSeq testing on the cytopathologic diagnosis, management, and cost of care in patients with thyroid nodules.

Methods:

In a retrospective study, using actual patient cohorts, the outcome and cost of management of patients with thyroid nodules seen before the introduction of ThyroSeq v2 at our institution (standard of care [StC] cohort) were compared with those seen after the introduction of this test (MT cohort).

Results:

A total of 773 consecutive patients entered the study (393 StC, 380 MT). The incidence of cytologically benign nodules decreased from 71.0% (StC) to 53.2% (MT) and those of atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS) and follicular neoplasm/suspicious for follicular neoplasm (FN/SFN) increased from 8.9% (StC) to 21.3% (MT) and from 3.1% (StC) to 6.3% (MT), respectively. The overall rate of surgery did not change significantly (23.4% in StC vs. 23.2% in MT). Among patients who underwent surgery, the rate of overtreatment (surgeries performed on histologic benign nodules without clinical indication: compressive symptoms, hyperthyroidism resistant to medication, and nodule size >4 cm) slightly decreased from 18.8% (StC) to 16.7% (MT). The rate of malignancy decreased from 45.5% (StC) to 37.9% (MT) in AUS/FLUS and increased from 40.0% to 53.8% in FN/SFN. However, the overall rate of malignancy remained equal (47.8% in StC vs. 47.7% in MT). The average cost of care per patient in the AUS/FLUS and FN/SFN categories increased from $6,566 (StC) to $8,444 (MT) and from $9,313 (StC) to $10,253 (MT), respectively. Similarly, the overall average cost of care of a patient who underwent thyroid fine-needle aspiration increased from $3,088 (StC) to $4,282 (MT). Finally, the average cost per thyroid cancer detected increased from $26,312 (StC) to $38,746 (MT).

Conclusions:

Introduction of ThyroSeq v2 resulted in a shift toward indeterminate cytology results. The institutional rate of surgery, overtreatment, and malignancy did not change significantly. Lack of decrease in the rate of surgery along with the additional cost of ThyroSeq v2 increased the overall cost of care of patients including those with indeterminate cytology results.

Introduction

Fine-needle aspiration (FNA) biopsy is the gold standard for the evaluation of thyroid nodules (1). The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) is the standardized template used for FNA interpretation (2). TBSRTC recommends six diagnostic categories including: (I) nondiagnostic, (II) benign, (III) atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS), (IV) follicular neoplasm/suspicious for follicular neoplasm (FN/SFN), (V) suspicious for malignancy, and (VI) malignant (2). Up to 20–30% of FNA cytology samples will result in an “indeterminate” cytologic diagnosis (3). Therefore, more definitive distinction between benign and malignant nodules is important so that unnecessary surgeries can be avoided for patients with benign nodules. Molecular testing (MT) of FNAs with indeterminate cytology helps risk stratify these nodules (4 –8) and reduces the rate of surgery (9 –11). However, these tests are costly and it is important to evaluate their cost impact on the management of thyroid nodules.

Several studies have examined the economic and clinical impact of MT on thyroid nodule FNA, many of which predicted reduction in the rate of surgery. However, controversy still exists regarding the cost-effectiveness of these tests (12 –16). While some authors have shown these tests to be cost saving (14 –16), some have debated the economic efficiency of these tests due to the low specificity or relatively high price of the tests (12,13). The majority of these studies have been performed on hypothetical models (13 –17), and only two on actual cohorts (12,18). It is important to note that previous model-based approaches in patients with thyroid nodules did not evaluate the impact of multiple factors on management and cost, including patient preference and presence of multiple nodules and their association with the risk of malignancy. Another important factor that was not considered in those models is the potential impact of routine use of MT on the cytologic diagnosis and consequently clinical management and overall cost of care.

The ThyroSeq v2 test, a multigene next-generation sequencing panel, developed by Nikiforov et al., has been shown to provide high sensitivity and specificity for cancer detection in thyroid nodules with AUS/FLUS (Bethesda III) or FN/SFN (Bethesda IV) cytologic diagnoses (9,10). The authors suggested that this test improves clinical management in these two categories and allows clinicians to avoid unnecessary surgeries. To date, the cost impact of ThyroSeq testing on the management of patients with thyroid nodule with indeterminate thyroid cytology has only been studied in a small cohort of 51 patients (18). The increasing use of this test merits further investigation of its impact on the cytologic diagnosis, clinical management, and cost of care in patients with thyroid nodules.

In the present study, we used actual cohorts of patients with thyroid nodules to evaluate the impact of routine use of ThyroSeq v2 testing on the (i) cytology diagnosis; (ii) average cost of care including initial diagnosis, management, and follow-up over the course of two years; (iii) rate of surgery (including lobectomy and total thyroidectomy); (iv) rate of overtreatment defined as surgical procedure performed on histologically benign nodules without any clinical indication (including compressive symptoms, hyperthyroidism resistant to medical therapy, and nodule size >4 cm) and only based on cytology and/or ThyroSeq results; and (v) rate of malignancy in each category of the Bethesda System and also among the entire population of patients who underwent FNA and also overall institutional rate of malignancy. We included all the six Bethesda categories in our study to be able to evaluate the impact of ThyroSeq on cytology diagnosis and also on the overall average cost of care across the entire population who underwent FNA.

Methods

In February 2015, a reflex ThyroSeq v2 test was introduced to all thyroid FNAs performed at our institution with a diagnosis of AUS/FLUS and FN/SFN. Before performing ThyroSeq testing on these materials, FNA slides underwent a second review at a reference laboratory, CBLpath (Rye Brook, NY). With institutional review board approval, all consecutive in-house thyroid FNAs performed from January 2014 to February 2016 were retrieved from the pathology department database. Patients were divided into two cohorts:

(i) standard of care (StC): Initial FNA was performed before the implementation of ThyroSeq testing in the management of thyroid nodules at our institution and no MT was performed on the aspiration material; and

(ii) MT: Initial FNA was performed after the implementation of the reflex ThyroSeq testing in the management of thyroid nodules at our institution.

The histologic diagnosis was correlated with the FNA interpretation.

We compared the two cohorts regarding the frequency of Bethesda cytologic categories, rate of surgery, type of procedure, and the rate of overtreatment, which was defined as surgical procedure performed on histologically benign nodules only based on cytology and/or ThyroSeq results and without any clinical indication (compressive symptoms, hyperthyroidism resistant to medical therapy, and nodule size >4 cm). We then calculated the rate of malignancy among patients who underwent surgery and also across the entire population who underwent FNA. Finally, we estimated the cost of care including the diagnosis, treatment, and follow-up over a period of two years and compared between the two cohorts. Costs included: FNA biopsy including initial and repeat FNAs; follow-up including office visits and ultrasound; surgical management including surgical office visit, surgical procedure, and anesthesiology; pathology including cytologic and histologic diagnosis; and ThyroSeq v2 testing. The second review of FNA by CBLpath did not have any influence on the management decision-making and its cost was not included in our analysis.

In a separate analysis, we excluded the cost of MT on nodules that were called benign on second review at CBL and then compared the overall cost between the two cohorts. Costs were derived from Medicare physician fee schedule 2015 (19), Health Care Cost and Utilization Project (20), institutional physician fee and hospital cost, and various literature sources. We calculated the total and average costs in each cohort and separately in each Bethesda category. The chi-squared test and the Mann–Whitney U-test were used to analyze categorical and continuous variables respectively; a value of p ≤ 0.05 was considered statistically significant. STATA/IC 13 (College Station, TX) was used to perform all the statistical analyses.

Results

A total of 773 consecutive patients (651 [84.2%] female and 122 [15.8%] male) with a median age of 55 (interquartile range 41–64) years were included in the study (Table 1). Of these, 393 (50.8%) were in the StC cohort and 380 (49.2%) in the MT cohort. Distribution of cytologic diagnoses according to the Bethesda system is shown in Table 1. The frequency of benign (Bethesda II) diagnosis decreased significantly after the introduction of ThyroSeq testing (71.0% in StC vs. 53.2% in MT; p < 0.001). Conversely, there was a significant increase in the AUS/FLUS category from 8.9% in the StC cohort to 21.3% in the MT cohort (p < 0.001), and in the FN/SFN category from 3.1% in the StC cohort to 6.3% in the MT cohort (p = 0.040). Frequency of Bethesda categories I, V, and VI did not differ significantly between the two cohorts.

Table 1.

Demographic Information and Distribution of Bethesda System Cytologic Diagnoses

Characteristic	StC, n = 393 (50.8%), n (%)	MT, n = 380 (49.2%), n (%)	p
Age, years, median (IQR)	54 (41–65)	55 (41–63)	0.672
Sex
Female	336 (85.5)	315 (82.9)	0.326
Male	57 (14.5)	65 (17.1)
TBSRTC distribution
ND	43 (10.9)	42 (11.1)	1.000
Benign	279 (71.0)	202 (53.2)	<0.001
AUS/FLUS	35 (8.9)	81 (21.3)	<0.001
FN/SFN	12 (3.1)	24 (6.3)	0.040
SM	11 (2.8)	13 (3.4)	0.681
M	13 (3.3)	18 (4.7)	0.361

AUS/FLUS, atypia of undetermined significance/follicular lesion of undetermined significance; FN/SFN, follicular neoplasm/suspicious for follicular neoplasm; IQR, interquartile range; M, malignant; MT, molecular test; ND, nondiagnostic; SM, suspicious for malignancy; StC, standard of care; TBSRTC, The Bethesda System for Reporting Thyroid Cytopathology.

The clinical management is summarized in Table 2. ThyroSeq testing was associated with a significant reduction in the rate of surgery in the AUS/FLUS category (68.2% in StC to 35.8% in MT) and in the FN/SFN category (83.3% in StC to 54.2% in MT). After combining Bethesda III and IV, the rate of surgery decreased from 68.1% in StC to 40% in MT (Table 6). However, despite the significant reductions in the rates of surgery in these two Bethesda categories, the overall institutional rate of surgery between the two cohorts did not change significantly (23.2% in MT vs. 23.4% in StC). The rate of diagnostic lobectomy in Bethesda category III slightly increased from 22.7% in StC to 27.6% in MT, and in Bethesda category IV decreased from 30% to 15%. Table 3 compares the rate of overtreatment in Bethesda III and IV between the two cohorts. The rate of overtreatment lobectomy among all lobectomies in the combined categories of III and IV decreased from 37.5% in StC to 30% in MT. Conversely, the rate of overtreatment total thyroidectomy increased from 12.5% to 15.2%. The overall rate of overtreatment in the combined categories of III and IV did not change significantly (18.8% in StC vs. 16.7% in MT).

Table 2.

Clinical Management Based on Bethesda System Cytologic Diagnoses

Initial FNA diagnosis (Bethesda system)	StC, n = 393 (50.8%)				MT, n = 380 (49.2%)
	Nonsurgical, n (%)	Surgical			Nonsurgical, n (%)	Surgical
	Nonsurgical, n (%)	Total, n (%)	Lobectomy, n (%)	TT ± ND, n (%)	Nonsurgical, n (%)	Total, n (%)	Lobectomy, n (%)	TT ± ND, n (%)
ND	38 (88.4)	5 (11.6)	1 (20)	4 (80)	37 (88.1)	5 (11.9)	1 (20)	4 (80)
Benign	248 (88.9)	31 (11.1)	12 (38.7)	19 (61.3)	188 (93.1)	14 (6.9)	5 (35.7)	9 (64.3)
AUS/FLUS	13 (37.1)	22 (68.2)	5 (22.7)	17 (77.3)	52 (64.2)	29 (35.8)	8 (27.6)	21 (74.1)
FN/SFN	2 (16.7)	10 (83.3)	3 (30)	7 (70)	11 (45.8)	13 (54.2)	2 (15.4)	11 (84.6)
SM	0	11 (100)	0	11 (100)	2 (15.4)	11 (84.6)	3 (27.3)	8 (72.7)
Malignant	0	13 (100)	0	13 (100)	2^a (11.1)	16 (88.9)	1 (6.3)	15 (93.7)
Total	301 (76.6)	92 (23.4)	21 (22.8)	71 (77.2)	292 (76.8)	88 (23.2)	20 (22.7)	68 (77.3)

Metastatic anaplastic thyroid cancer.

FNA, fine-needle aspiration; TT ± ND, total thyroidectomy ± neck dissection.

Table 3.

Rate of Overtreatment in Bethesda Categories III and IV

Initial FNA diagnosis (Bethesda system)	Lobectomy		TT ± ND		Total
Initial FNA diagnosis (Bethesda system)	StC, n (%)	MT, n (%)	StC, n (%)	MT, n (%)	StC, n (%)	MT, n (%)
AUS/FLUS	2 (40.0)	2 (25.0)	2 (11.8)	3 (14.3)	4 (18.2)	5 (17.2)
FN/SFN	1 (33.3)	0	1 (14.3)	2 (25.0)	2 (20.0)	2 (15.4)
Total	3 (37.5)	2 (20.0)	3 (12.5)	5 (15.2)	6 (18.8)	7 (16.7)

Table 4 shows the rate of malignancy. Among patients who underwent surgery, the rate of malignancy in patients with Bethesda III nodules decreased from 45.5% in the StC cohort to 37.9% in the MT cohort. However, the rate of malignancy in patients with Bethesda IV nodules significantly increased from 40.0% in the StC cohort to 53.8% in the MT cohort. The overall rate of malignancy in patients who underwent surgery in the MT cohort (47.7%) remained essentially the same as in the StC cohort (47.8%). The rate of malignancy among the patients who underwent FNA decreased from 28.6% (StC) to 13.6% (MT) in Bethesda III and from 33.3% (StC) to 29.2% (MT) in Bethesda IV, but, interestingly, the overall rate of malignancy did not change significantly between the StC (11.2%) and the MT (11.6%) cohorts.

Table 4.

Rate of Malignancy

FNA diagnosis	Rate of malignancy in patients who underwent surgery		Rate of malignancy among all patients who underwent FNA
FNA diagnosis	StC, n = 92, n (%)	MT, n = 88, n (%)	StC, n = 393, n (%)	MT, n = 380, n (%)
ND	1 (20)	1 (20)	1 (2.3)	1 (2.4)
Benign	6 (19.4)	0	6 (2.2)	0
AUS/FLUS	10 (45.5)	11 (37.9)	10 (28.6)	11 (13.6)
FN/SFN	4 (40.0)	7 (53.8)	4 (33.3)	7 (29.2)
SM	10 (90.9)	7 (63.6)	10 (90.9)	7 (53.8)
Malignant	13 (100)	16 (100)	13 (100)	18 (100)
Total	44 (47.8)	42 (47.7)	44 (11.2)	44 (11.6)

Cost of care in each Bethesda category and in the entire population of each cohort is represented in Table 5. The average cost of care per patient in Bethesda category III significantly increased from $6,566 in the StC cohort to $8,444 in the MT cohort. This increase is mainly due to a substantial additional cost of MT that was not offset by the reduction in the rate and cost of surgery in this category. We observed a similar but less significant trend in the average cost of care in the FN/SFN category ($9,313 in StC vs. $10,253 in MT). Table 6 shows the rate of surgery and average cost for Bethesda III and IV combined. Despite the decrease in the rate of surgery from 68.1% to 40%, the average cost of care per patient increased from $7,268 to $8,858.

Table 5.

Summary of Average Cost Per Patient and Total Cost in Each Category

Initial FNA diagnosis	StC, n = 392 (55.8%), average cost per patient (total in the category), USD	MT, n = 380 (49.2%), average cost per patient (total), USD
ND	2,343 (100,750), n = 43	2,404 (100,967), n = 42
Benign	1,643 (458,453), n = 279	1,339 (270,412), n = 202
AUS/FLUS	6,566 (229,825), n = 35	8,444 (683,992),^a n = 81
FN/SFN	9,313 (111,761), n = 12	10,253 (246,073),^b n = 24, n = 22
SM	12,783 (140,615), n = 11	8,944 (116,284), n = 13
Malignant	12,995 (168,940), n = 13	11,645 (209,612), n = 18
Total average per patient	3,088 (1,210,344)	4,282 (1,627,340) ^c

$7,044 ($570,592) after excluding cost of MT from nodules called benign on second review.

$9,728 ($233,473) after excluding cost of MT from nodules called benign on second review.

$3,951 (1,501,340) after excluding cost of MT from nodules called benign on second review.

The combined cost of Bethesda III and IV categories amounted to only 28.2% ($341,586) of the total cost in the StC cohort ($1,210,344) but 57.2% ($930,065) of the total cost in the MT cohort ($1,627,340). In other words, management of patients in Bethesda III and IV became more costly upon implementation of ThyroSeq. Furthermore, comparing the entire population of StC and MT cohorts, the overall average cost of management per patient with a thyroid nodule increased from $3,080 in the StC cohort to $4,282 in the MT cohort. As explained above, this increase is due to the additional cost of MT without any significant reduction in the overall institutional rate of surgery (23.4% in StC vs. 23.2% in MT).

Our institutional rate of Bethesda III and IV became relatively high after the implementation of ThyroSeq testing. According to our current data, 36% of our institutional Bethesda III and IV lesions were interpreted as being benign at CBL. In further analysis, we excluded the cost of MT from those nodules that were called benign on second review. We observed that the cost of care decreased in both Bethesda III and IV categories ($7,044 for Bethesda III, $9,728 for Bethesda IV, and $7,658 for combined Bethesda III and IV) but did not fall lower than the costs in StC ($6,566 for Bethesda III and $9,313 for Bethesda IV) (Tables 5 and 6, footnote). In further analysis, we estimated the cost per thyroid cancer detected in the two cohorts, which was markedly higher in MT ($26,312 in StC and $38,746 in MT).

Table 6.

Rate of Surgery and Average Cost Per Patient in Bethesda III and IV Categories Combined

	StC, n = 47, n (%)	MT, n = 105, n (%)
Rate of surgery	32 (68.1)	42 (40)
Cost, mean (total), USD	7,268 (341,586)	8,858 (930,065)^a

$7,658 ($804,065) after excluding MT cost from nodules called benign on second review.

Discussion

The current study is the largest actual patient cohort to evaluate the impact of ThyroSeq v2 test on the management and cost of care in patients with thyroid nodules. Our data show that the introduction of ThyroSeq v2 was associated with a significant increase in the rate at which thyroid nodules were called AUS/FLUS or FN/SFN. We also observed a significant reduction in the rate of surgery with the use of ThyroSeq in these two categories, but interestingly, the overall institutional rate of surgery did not differ between the two cohorts (23.4% in StC vs. 23.2% in MT). The rate of overtreatment among the patients who underwent surgery in Bethesda III, IV, or combined slightly decreased (18.8% in StC and 16.7% in MT in Bethesda III and IV combined). The average cost of care including the initial diagnosis and follow-up over the course of two years in patients with Bethesda III or IV thyroid nodules increased. Also, comparing the entire population of the two cohorts, the overall average cost of care per patient increased by $1,202 (from $3,080 in StC to $4,282 in MT).

Compared with previous studies that used hypothetical models to evaluate the impact of MT on cost of care in patients with thyroid nodules, the design of the current study is unique in two ways: (i) unlike the previous studies that were performed only on Bethesda III and IV lesions, we included all the Bethesda diagnostic categories. This provides several specific advantages including the possibility: (a) to determine the impact of MT on cytologic diagnoses as a potential factor that can affect management and cost of care; (b) to calculate the institutional rate of surgery and overall rate of malignancy among patients who underwent surgery; (c) to determine the rate of malignancy among the entire population who underwent FNA; and (ii) actual patient cohorts were used in our study. Using actual patient cohorts in the current study has significant advantages since it allowed for evaluation of multiple factors that have not been assessed in the previous model-based approaches. These factors include patient preference, malignancy risk associated with multiple nodules, and clinical factors that affect the requirements for surgery, including pressure symptoms, rapid growth of nodules, family history of cancer, and presence of comorbidities.

Impact on cytologic diagnosis and management

ThyroSeq v2 testing in our cohort was associated with a shift in the initial cytologic diagnosis of the thyroid nodule toward Bethesda III and IV categories that would have been called benign in the pre-ThyroSeq period. This increased the number of patients with lower risk of malignancy in these categories who would have never undergone surgery in the pre-ThyroSeq era. Consequently, the rate of surgery decreased in these two categories while the institutional rate of surgery did not change. Sacks et al. observed similar findings regarding the impact of the Afirma Gene Expression Classifier (GEC) on the initial cytologic diagnosis (21). Afirma GEC is a genetic test that has shown a high negative predictive value in the preoperative diagnosis of cytologically indeterminate nodules (11). The authors reported a significant increase in Bethesda III (10.7% vs. 13.4%) and Bethesda IV (1.8% vs. 2.9%) diagnoses associated with the introduction of the Afirma GEC. Similar to our findings, they showed that the use of this MT was not associated with a significant change in their institutional rate of thyroidectomies (37.7% pre-Afirma vs. 45.1% post-Afirma; p = 0.11).

Other studies, using hypothetical models, have reported discordant results regarding the impact of MT on the rate of diagnostic lobectomy, unnecessary surgeries, or surgery on histologically benign nodules. Labourier showed that GEC testing resulted in a decrease in the rate of unnecessary surgeries by 32% and concluded that this is the main cause of reducing the cost of care (16). Li et al. evaluated routine GEC use compared with standard practice over a 5-year timeline and showed 74% fewer surgeries for benign nodules resulting in savings of $1,453 per patient (14). Najafzadeh et al. also showed that using MT is associated with a decrease in the rate of unnecessary surgeries. Among 10,000 simulated patients, 4407 in the MT group versus 323 in the group without MT underwent unnecessary surgery (17). Yip et al. showed that implementation of a molecular panel resulted in an ∼1.9% reduction in the rate of diagnostic lobectomy (15).

In the study presented here, implementation of ThyroSeq v2 was not associated with a significant change in the rate of diagnostic lobectomy in the combined Bethesda III and IV categories (25.4% in StC vs. 25.2% in MT).

Impact on cost of care

Analysis of the cohort presented here shows that implementation of ThyroSeq v2 is associated with an increase in the average cost of care per patient both in the individual Bethesda category III ($1,878) and IV ($940) as well as in the overall population who underwent FNA ($1,202). Since the cost of surgery before and after the introduction of ThyroSeq v2 did not change significantly, the increased cost per patient is attributed to the additional cost of MT. Our cohort shows that part of this additional cost was unnecessary because of the shift of previously called benign FNAs in the pre-ThyroSeq era into the indeterminate category with the use of ThyroSeq.

There are two previous studies that partially or entirely incorporated actual patient cohorts to evaluate cost-effectiveness of MT in the workup of indeterminate thyroid nodules (12,18). Shapiro et al. (12), using a simulator and a real cohort comparator, showed that the MT group had a 13.1% reduction in the rate of patients undergoing surgery. The authors showed that GEC testing reduced the costs associated with ultrasound, FNA, and initial surgery. However, the cost of completion surgery increased. Overall, the cost of managing 1 nodule over 2 years increased from $7,697 in the StC group to $10,096 in the MT group. Interestingly, their results showed that a great majority of patients with indeterminate nodules in the StC group underwent surveillance with ultrasound or up-front total thyroidectomy rather than diagnostic lobectomy or repeat FNA as recommended by the American Thyroid Association guidelines, which suggest that the management decision for an indeterminate nodule is difficult to capture in an algorithmic approach. The simulations in their study predicted that MT was more costly than StC management despite a decrease in the proportion of patients undergoing surgery (12).

In another study, Rivas et al. used an actual cohort of 22 nodules with Bethesda category III and 27 nodules with Bethesda category IV and showed that ThyroSeq is a cost-saving approach compared with diagnostic thyroidectomy (18). The authors reported that the average cost per Bethesda III evaluated nodule was $14,669 using ThyroSeq compared with $23,338 for diagnostic thyroid surgery. In nodules with Bethesda category IV cytology, the authors showed that the average cost per nodule evaluated using ThyroSeq and diagnostic thyroidectomy was $14,641 and $24,345, respectively. They also showed that the cost per thyroid cancer case detected in Bethesda III was $62,233 for diagnostic thyroid surgery compared with $58,674 using ThyroSeq. The cost per thyroid cancer case detected in Bethesda IV was $31,721 when using ThyroSeq compared with $53,560 for diagnostic thyroidectomy (18).

Significantly different from the above findings, the results presented here show that the cost per thyroid cancer detected was $26,312 in the StC cohort and $38,746 in the MT cohort. The difference between the results of cohort by Rivas et al. (18) and the data presented here could be due to different institutional rate of indeterminate categories and also risk of malignancy in these categories, which can affect the cost of care (22). It is also noteworthy to mention that the costs reported in the cohort by Rivas et al. are significantly higher than that of the present and other studies (12,14 –16,22,23). This raises concerns regarding the generalizability of these findings.

Our study has several limitations. The cost of postoperative hypothyroidism and indirect costs such as lost time at work were not included in our study due to lack of data. However, because of equivalent rate of surgery and malignancy in the two arms of the cohort, this would have had a limited impact on the results. Another limitation is that the follow-up duration of only two years may not be long enough for an indolent case of thyroid cancer to declare itself. Also, our physician schedule fee estimates were derived from the Centers for Medicare and Medicaid Services reimbursement schedules, which vary between institutions and geographical regions (13). Therefore, it cannot be generalizable across all institutions. The high rate of indeterminate cytology in our institution also makes our results less generalizable. However, after excluding the cost of MT from indeterminate nodules that were called benign on second review, we did not observe a significant change in the results regarding the cost. Finally, our observations may not be relevant to ThyroSeq v3, an expanded version of the ThyroSeq v2 test, which includes recently discovered genetic markers related to thyroid cancer (24) or MT with other techniques, such as messenger RNA sequencing, microRNA analyses, and Afirma Genomic Sequencing Classifier and Expression Atlas (25). Since these advancements in MT could potentially improve the preoperative diagnosis of benign versus malignant nodules and therefore management and cost of care, we recommend performing similar studies, particularly multi-institutional trials, with these newer tests. We also recommend future studies to evaluate the impact of reduced costs of these tests on the cost of management.

Conclusions

Using an actual cohort of patients with thyroid nodules, we found that routinely available ThyroSeq v2 testing can shift the cytologic diagnoses toward indeterminate categories. However, this might vary among institutions and among cytopathologists. The importance of this factor in cost-effectiveness studies is underscored by the fact that the use of ThyroSeq v2 in our cohort resulted in a reduction in the rate of surgeries for Bethesda III and IV lesions without reducing the institutional rate of surgery, type of procedure, and cost of care. While one of the values of MT, other than establishing a preoperative diagnosis, is the reduction in the rate of unnecessary surgeries, the rate of unnecessary surgeries in our cohort did not significantly change with the use of ThyroSeq v2. Findings of this cohort also highlight the complexity of decision-making in indeterminate thyroid nodules, which can be influenced by multiple factors that should be addressed in simulation models, including patient preference, rate of surgery due to clinical symptoms, and rate of indeterminate cytologic diagnosis.

Footnotes

Acknowledgments

We acknowledge Dr. Karen Quillen, MD, MPH, for her input in preparing the article and Dr. Elizabeth N. Pearce, MD, for her input into the cost estimate of this study.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research is partially supported by Mallory Pathology Associates, Inc. Grant, Department of Pathology & Laboratory Medicine at Boston Medical Center.

References

Cooper

, Doherty

, Haugen

, Kloos

, Lee

, Mandel

, Mazzaferri

, McIver

, Pacini

, Schlumberger

, Sherman

, Steward

, Tuttle

. 2009. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19:1167–1214.

Ali

. 2011. Thyroid cytopathology: Bethesda and beyond. Acta Cytol, 55:4–12.

Cibas

, Ali

. 2009. The Bethesda system for reporting thyroid cytopathology. Am J Clin Pathol, 132:658–665.

Moses

, Weng

, Sansano

, Peng

, Khanafshar

, Ljung

, Duh

, Clark

, Kebebew

. 2010. Molecular testing for somatic mutations improves the accuracy of thyroid fine-needle aspiration biopsy. World J Surg, 34:2589–2594.

Cantara

, Capezzone

, Marchisotta

, Capuano

, Busonero

, Toti

, Di Santo

, Caruso

, Carli

, Brilli

, Montanaro

, Pacini

. 2010. Impact of proto-oncogene mutation detection in cytological specimens from thyroid nodules improves the diagnostic accuracy of cytology. J Clin Endocrinol Metab, 95:1365–1369.

Nikiforov

, Steward

, Robinson-Smith

, Haugen

, Klopper

, Zhu

, Fagin

, Falciglia

, Weber

, Nikiforova

. 2009. Molecular testing for mutations in improving the fine-needle aspiration diagnosis of thyroid nodules. J Clin Endocrinol Metab, 94:2092–2098.

Nikiforov

, Ohori

, Hodak

, Carty

, LeBeau

, Ferris

, Yip

, Seethala

, Tublin

, Stang

, Coyne

, Johnson

, Stewart

, Nikiforova

. 2011. Impact of mutational testing on the diagnosis and management of patients with cytologically indeterminate thyroid nodules: a prospective analysis of 1056 FNA samples. J Clin Endocrinol Metab, 96:3390–3397.

Yip

, Nikiforova

, Carty

, Yim

, Stang

, Tublin

, Lebeau

, Hodak

, Ogilvie

, Nikiforov

. 2009. Optimizing surgical treatment of papillary thyroid carcinoma associated with BRAF mutation. Surgery, 146:1215–1223.

Nikiforov

, Carty

, Chiosea

, Coyne

, Duvvuri

, Ferris

, Gooding

, LeBeau

, Ohori

, Seethala

, Tublin

, Yip

, Nikiforova

. 2015. Impact of the multi-gene ThyroSeq next-generation sequencing assay on cancer diagnosis in thyroid nodules with atypia of undetermined significance/follicular lesion of undetermined significance cytology. Thyroid, 25:1217–1223.

10.

Nikiforov

, Carty

, Chiosea

, Coyne

, Duvvuri

, Ferris

, Gooding

, Hodak

, LeBeau

, Ohori

, Seethala

, Tublin

, Yip

, Nikiforova

. 2014. Highly accurate diagnosis of cancer in thyroid nodules with follicular neoplasm/suspicious for a follicular neoplasm cytology by ThyroSeq v2 next-generation sequencing assay. Cancer, 120:3627–3634.

11.

Alexander

, Kennedy

, Baloch

, Cibas

, Chudova

, Diggans

, Friedman

, Kloos

, LiVolsi

, Mandel

, Raab

, Rosai

, Steward

, Walsh

, Wilde

, Zeiger

, Lanman

, Haugen

. 2012. Preoperative diagnosis of benign thyroid nodules with indeterminate cytology. N Engl J Med, 367:705–715.

12.

Shapiro

, Pharaon

, Kellermeyer

. 2017. Cost-effectiveness of Gene Expression Classifier testing of indeterminate thyroid nodules utilizing a real cohort comparator. Otolaryngology Head Neck Surg, 157:596–601.

13.

Lee

, How

, Tabah

, Mitmaker

. 2014. Cost-effectiveness of molecular testing for thyroid nodules with atypia of undetermined significance cytology. J Clin Endocrinol Metab, 99:2674–2682.

14.

, Robinson

, Anton

, Saldanha

, Ladenson

. 2011. Cost-effectiveness of a novel molecular test for cytologically indeterminate thyroid nodules. J Clin Endocrinol Metab, 96:E1719–E1726.

15.

Yip

, Farris

, Kabaker

, Hodak

, Nikiforova

, McCoy

, Stang

, Smith

, Nikiforov

, Carty

. 2012. Cost impact of molecular testing for indeterminate thyroid nodule fine-needle aspiration biopsies. J Clin Endocrinol Metab, 97:1905–1912.

16.

Labourier

. 2016. Utility and cost-effectiveness of molecular testing in thyroid nodules with indeterminate cytology. Clin Endocrinol, 85:624–631.

17.

Najafzadeh

, Marra

, Lynd

, Wiseman

. 2012. Cost-effectiveness of using a molecular diagnostic test to improve preoperative diagnosis of thyroid cancer. Value Health, 15:1005–1013.

18.

Rivas

, Nassar

, Zhang

, Casler

, Chindris

, Smallridge

, Bernet

. 2018. ThyroSeq((R))V2.0 molecular testing: a cost-effective approach for the evaluation of indeterminate thyroid nodules. Endocr Pract, 24:780–788.

19.

Center for Medicare and Medicaid Services. Physician fee schedule. Available at www.cms.gov/medicare/medicare-fee-for-service-payment/physicianfeesched (accessed September 2018 ).

20.

Agency for Healthcare and Research Quality. Healthcare cost and utilization project. Available at https://hcupnet.ahrq.gov/#setup (accessed September 2018 ).

21.

Sacks

, Bose

, Zumsteg

, Wong

, Shiao

, Braunstein

, Ho

. 2016. Impact of Afirma gene expression classifier on cytopathology diagnosis and rate of thyroidectomy. Cancer Cytopathol, 124:722–728.

22.

, Lam

, Levin

, Rao

, Sullivan

, Yeh

. 2016. Effect of malignancy rates on cost-effectiveness of routine gene expression classifier testing for indeterminate thyroid nodules. Surgery, 159:118–126.

23.

Lee

, Mitmaker

, Chabot

, Lee

, Kuo

. 2016. Cost-effectiveness of diagnostic lobectomy versus observation for thyroid nodules >4 cm. Thyroid, 26:271–279.

24.

Nikiforova

, Mercurio

, Wald

, Barbi de Moura

, Callenberg

, Santana-Santos

, Gooding

, Yip

, Ferris

, Nikiforov

. 2018. Analytical performance of the ThyroSeq v3 genomic classifier for cancer diagnosis in thyroid nodules. Cancer, 124:1682–1690.

25.

Patel

, Angell

, Babiarz

, Barth

, Blevins

, Duh

, Ghossein

, Harrell

, Huang

, Kennedy

, Kim

, Kloos

, LiVolsi

, Randolph

, Sadow

, Shanik

, Sosa

, Traweek

, Walsh

, Whitney

, Yeh

, Ladenson

. 2018. Performance of a genomic sequencing classifier for the preoperative diagnosis of cytologically indeterminate thyroid nodules. JAMA Surg, 153:817–824.