Development of a Low-cost NGS Test for the Evaluation of Thyroid Nodules

Abstract

Ultrasound-guided fine needle aspiration cytology (FNAC) is the preferred method of identifying malignancy in palpable thyroid nodules using the Bethesda reporting system. However, in around 30–40% of FNACs (Bethesda categories III, IV, and V), the results are indeterminate and surgery is required to confirm malignancy. Out of those who undergo surgery, only 10–40% of patients in these categories are found to have malignancies, thus proving surgery to be unnecessary for some patients or to be incomplete in others. While molecular testing on thyroid FNAC material is part of the American Thyroid Association (ATA) guidelines in evaluating thyroid nodules, it is currently unavailable in India due to cost constraints. In this study, we prospectively collected FNAC samples from sixty-nine patients who presented with palpable thyroid nodules. We designed a cost-effective next-generation sequencing (NGS) test to query multiple variants in the DNA and RNA isolated from the fine needle aspirate. The identification of oncogenic variants was considered to be indicative of malignancy, and confirmed by surgical histopathology. The panel showed an overall sensitivity of 81.25% and a specificity of 100%, while in the case of Bethesda categories III, IV, and V, the sensitivity was higher (87.5%) and the specificity was established at 100%. The panel could thereby serve as a rule-in test for the diagnosis of thyroid cancer and therefore help identify patients who require surgery, especially in the indeterminate Bethesda categories III, IV, and V.

Electronic supplementary material

The online version of this article (10.1007/s13193-019-01000-w) contains supplementary material, which is available to authorized users.

Introduction

Thyroid cancer is the most common endocrine cancer with incidence rates increasing steadily over the years [1]. Most thyroid tumors originate from thyroid follicular epithelial cells, known as thyroid follicular cell–derived tumors, and these are further subdivided into differentiated thyroid carcinoma (DTC), poorly differentiated carcinoma (PDC), and anaplastic carcinoma (ATC). Papillary thyroid carcinomas (PTC) and follicular thyroid carcinomas (FTC) fall under DTC, with the most common one being PTC, which accounts for 80–90% of all thyroid cancers [2].

Most often, patients present with an asymptomatic nodule in their neck. Physicians evaluate these nodules through ultrasound (US)-guided fine needle aspiration (FNA) biopsies [2]. Aspirates from the biopsy are sent for cytological evaluation. The Bethesda reporting system is used to assign samples into six diagnostic categories and thus guide clinical management of these cases [3]. The categorization system in Bethesda has a high accuracy in the following categories:

1. Categories I and II—classified as benign

2. Category VI—classified as malignant

However, categories III—atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS); IV—follicular or oncocytic (Hürthle cell) neoplasm/suspicious for a follicular or oncocytic (Hürthle cell) neoplasm (FN/SFN); and V—suspicious for malignant cells (SUSP) account for 10–30% of the samples [3]. Diagnostic hemithyroidectomies are still customarily performed to obtain a definite histological diagnosis. With a benign histopathological result, the patient is exposed to unnecessary surgical risks while in the case of malignant lesions, a second-stage completion thyroidectomy is often indicated, which is associated with additional costs and higher risks of surgical complications [4]. Therefore, a sensitive, specific, and less-invasive method to accurately evaluate thyroid nodules is the need of the hour.

Molecular markers have shown to help with decision-making of ruling in or ruling out malignancies in these indeterminate categories. Identification of driver mutations in samples is an indicator of malignancy and therefore guides subsequent decisions on thyroidectomies. Over the last decade, new biomarkers indicative of thyroid carcinoma have been identified. The most common genetic alteration in thyroid cancer occurs in B-type Raf Kinase (BRAF) gene, due to a transversion T1799A in exon 15 resulting in a V600E amino acid substitution. This mutation is present in almost 45% of all PTCs—it is an indication that the disease is aggressive, and therefore the patient has a poor survival outcome [5–7]. The RAS genes (HRAS, KRAS, and NRAS) are commonly found mutated in both FTCs and PTCs. Point mutations in the RAS genes usually occur in codons 12, 13, and 61 [5, 8]. The oncogenic fusion PAX8-PPAR훾 is found in one-third of FTC cases and in some follicular variant of PTC (FVPTC) cases, whereas RET-PTC fusions are identified in 10–20% of PTCs [5, 9]. These are the two most commonly observed translocations in thyroid cancer. In the American Thyroid Association (ATA) guidelines [10], the potentially strong diagnostic impact of molecular testing is explicitly unfolded, focusing on BRAF testing and the commercially available tests: the seven-gene mutation panel miRInform® thyroid (Asuragen, Inc., Austin, TX), the more recent ThyroSeq v3, and the Afirma® gene expression classifier (GEC) (Veracyte, Inc., San Francisco, CA). The ATA recommends considerate application of one of these molecular tests for Bethesda category III and IV nodules, provided that the result could change the treatment strategy [10]. In this paper, we describe the clinical utility of these molecular markers in assessing the malignancy status of thyroid nodules in the Indian cohort, thus aiding clinical management of indeterminate Bethesda categories.

Materials and Methods

Patient Recruitment

Ethics approval was obtained from ethics committees at both Mazumdar Shaw Cancer Centre (MSCC) and Strand Life Sciences Pvt. Ltd. Samples were collected with patients’ informed consent. Patients (n = 69) with palpable thyroid nodules were enrolled into the study from MSCC prospectively (Supplemental Table 1).

Sample Collection

Fine needle aspiration on the thyroid nodules in patients was performed under ultrasound guidance by an experienced endocrinologist as per standard operating procedure. Aspirate from the first pass was fixed on slides and used for cytological evaluation. The rest of the aspirate or the needle wash was collected in approximately 400 μl of RNAlater^TM Stabilization Solution (AM7021, Thermo Fisher Scientific, Waltham, MA, USA). The results from the cytological evaluation were blinded to the molecular profiling laboratory. The results of the mutation testing were not used to guide the management of the thyroid nodules.

Custom Panel Design

The panel consisted of primer pairs that spanned hotspot regions in four genes implicated in thyroid cancer, namely point mutations in codons 12, 13, 61, and 146 of HRAS, KRAS, NRAS, and codon 600 in BRAF. These loci were tested for in the DNA fraction of the fine needle aspiration cytology (FNAC) samples. RET/PTC1, RET/PTC2, and RET/PTC3 and PAX8-PPAR훾 rearrangements were detected from the RNA fraction of each sample along with the housekeeping gene, GAPDH.

Nucleic Acid Isolation

Both DNA and RNA were isolated from the FNAC samples using the Qiagen AllPrep DNA/RNA FFPE Kit (80234, Qiagen, Hilden, Germany) as per the manufacturer’s instructions and were quantified using Qubit® 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA). RNA was further converted to cDNA using High-Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (4374966, Applied Biosystems, Foster City, CA, USA). The cDNA was quantitated with reference to GAPDH, a housekeeping gene, using quantitative PCR (qPCR).

Genetic Profiling of FNAC Samples

Ten nanograms of DNA and 5 ng cDNA were used as input for multiplex PCR. Individual DNA and RNA multiplex products were subsequently pooled and taken for library preparation. Libraries were prepared using the standard protocol of KAPA HyperPlus Kit (Kapa Biosystems, Wilmington, MA, USA). As a QC check post-library preparation, the libraries were quantified on Qubit® 2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA) and also run on the Agilent 2200 TapeStation System (Agilent Technologies, Santa Clara, CA, USA). Uniquely indexed samples were pooled together and were sequenced on the MiSeq sequencer (Illumina, San Diego, CA, USA). Somatic mutations and translocations in samples were identified and annotated using Strand’s proprietary analysis tool, StrandNGS (ver. 2.6) [11].

Results

Patient Demographics and Study Details

Sixty-nine patients who underwent thyroid nodule FNAC were recruited from MSCC. A summary of the demographics is presented in Table 1 with details in Supplemental Table S1. Samples from seven patients were used for protocol optimization while nine FNA biopsies were not processed after extraction due to insufficient amount of DNA and RNA. An overview of the study is depicted in Fig. 1.

Table 1.

Patient demographics summary

Patients enrolled
Total patients	69
Age (years)
Mean (SD)	47.86 (15.81)
Gender (n, %)
Female	50 (72.47)
Male	19 (27.53)
Bethesda classification (n, %)
I	2 (2.90)
II	18 (26.09)
III	23 (33.33)
IV	8 (11.59)
V	5 (7.25)
VI	13 (18.84)

Fig. 1.

Clinical study design. Sixty-nine patients were enrolled prospectively across all thyroid subtypes and all Bethesda categories. FNAC samples were collected from all the patients. Seven samples were used in the standardization of the panel while nine were excluded due to sample insufficiency. The remaining fifty-three were sequenced and their mutation status was correlated with histopathological findings

Molecular Profiling Results and Histopathological Findings

Fifty-three FNAC samples were sequenced. In the samples, 60% were in Bethesda categories III (20%), IV (35%), and V (5%). We aimed for an average coverage of 200× in the targeted regions in the panel. Mutations were identified in twenty-five of the fifty-three samples. The details are listed in Supplemental Table S2. However, only thirteen patients in the aforementioned twenty-five had corresponding surgical histopathology data. Seven were confirmed to have PTC. BRAF p.V600E was identified in six of these cases whereas one other sample had an NRAS p.Q61R. Four cases were characterized as FTC—a majority of mutations were identified in the RAS genes. Two other cases of the twenty-five were medullary thyroid carcinoma (MTC) and FVPTC, respectively. In the remaining twelve samples, all the variants detected in the biopsies were from the RAS genes. Interestingly, four of these samples were in category II whereas the others were category III. Of the samples with confirmed surgical histopathology, 60% were in the indeterminate categories. Interestingly, four of them turned out to be benign whereas the remaining were malignant. No mutations were present in the benign samples.

Panel Performance

FNA biopsies with matched surgically-resected thyroid biopsies were used to evaluate the performance of the panel, since histopathological reports of surgical biopsies serve as a gold standard for confirming malignancy. Twenty patients of the fifty-three whose samples were profiled on next-generation sequencing (NGS) underwent surgery. Overall, the sensitivity of the panel was established as 81.25% and specificity at 100%. In the case of Bethesda categories III, IV, and V, the sensitivity was higher (87.5%) and the specificity was established at 100%. The details are provided in Table 2 and Supplemental Table S3. The correlation between the results from cytology, NGS, and surgical histopathology in these twenty samples is presented in Fig. 2.

Table 2.

Evaluation of panel performance on FNAC samples with confirmed surgical histopathological status

I. Overall
No. of patients operated upon	20
No. of true positives	13
No. of false positives	0
No. of true negatives	4
No. of false negatives	3
Sensitivity (%)	81.25
Specificity (%)	100
Positive predictive value (%)	100
Negative predictive value (%)	57.14
II. Bethesda categories III, IV, and V
No. of patients operated upon	12
No. of true positives	7
No. of false positives	0
No. of true negatives	4
No. of false negatives	1
Sensitivity (%)	87.5
Specificity (%)	100
Positive predictive value (%)	100
Negative predictive value (%)	80

Fig. 2.

Correlation between the three methods used in our study to diagnose malignancy in thyroid nodules, namely, cytology, NGS, and histopathology

Discussion

The panel is designed with a clear intention—molecular profiling of thyroid cancer to aid decisions on surgery in cases of indeterminate cytology. The panel is useful in identifying mutations to guide surgical decisions, as in the case of aggressive subtypes of PTC which commonly harbor BRAF p.V600E. However, one case of PTC carried an NRAS p.Q61R mutation which would have been missed in the absence of a panel-based test. We also identified mutations in the different RAS genes in four of the five FTC cases and a rare but oncogenic mutation in the BRAF gene in the remaining one. The atypical BRAF mutation would not be detected in a standard-of-care test. Given that mutations were identified in 81.25% of the surgically confirmed PTC and FTC cases using this test, we demonstrate the clinical utility of the panel in a majority of the thyroid cancer subtypes. Molecular marker–based NGS panels could be important tools to be used in conjunction with the standard-of-care tests. Future work will aim to increase the clinical utility of the panel across all thyroid subtypes by including additional markers. In our study, eight of the eighteen indeterminate cases with no matched surgical histopathology data showed oncogenic mutations, thereby suggesting a malignant phenotype. Clinical intervention, in the form of either aggressive monitoring or thyroidectomy, could be appropriate for these cases.

Despite a small sample size of twelve patients with matched FNA-surgical biopsies in the indeterminate categories, the panel shows 100% specificity and 87.5% sensitivity in this cohort. It is important to note that in the four cases confirmed to be benign by surgical histopathology, no pathological mutations were found. A positive predictive value (PPV) of 100% and a negative predictive value (NPV) of 80% in the indeterminate categories are encouraging. Arguably, the false negatives are FVPTC and MTC cases, which the current panel was not designed to cover. The specificity and %PPV remain at 100% overall while the sensitivity drops to 81.25% owing to the thyroid cancer subtypes which are outside the scope of the panel. Expansion of the panel with additional hotspots may address these issues. Overall, the test is a good candidate as a rule-in test for thyroidectomy in Bethesda categories III, IV, and V. In addition, the test may also be useful in profiling both common and rare subtypes of thyroid cancer for variants of diagnostic significance in a cost-effective manner. The laboratory-developed test (LDT) was designed as a multiplex PCR method coupled with massively parallel sequencing. The cost of library preparation and multiplex PCR is an order of magnitude cheaper than standard commercial custom amplicon panels and definitely much more economical than surgery and its monetary and non-monetary burden.

The higher expected rates of malignancy in the indeterminate category are not a surprise and in fact reflect a consistent practice pattern in our center [12]. Multiple variables including the size of the nodule, sonographic features, growth over time, age and gender, patient preference, and the conversation with the cytopathologist all go in the decision-making process of the sending of an indeterminate nodule for surgery. In the current study, the mutation analysis was not tested to play a role in this decision-making process. Our aim is to carry out future studies with this panel to see how the results of the mutation testing aid this decision-making process. It is important to note that a thyroidectomy comes with a significant cost associated, both monetary and patients’ quality of life. A low-cost NGS screening test paired with conventional tests makes sense in the Indian setting.

Compliance with Ethical Standards

Conflict of Interest

DV, AS, MS, AH, SS, UB, VV, and VG are employees of Strand Life Sciences Pvt. Ltd.