Abstracts - Orals, Poster Previews, and Posters
SAT 312-345-Biomarkers and Hormone-Dependent Cancers
Hall D-F, Tumor Biology (San Diego Convention Center)
Poster Board SAT-344
Current clinical management of thyroid cancer entails the cytopathological evaluation of ultrasound-guided fine needle aspirations (FNAs) from solid thyroid nodules larger than 1 cm. For nodules with indeterminate cytology, approximately 60% of malignant cases (1,2) or 50% of benign cases (3) can be identified using independent molecular tests with high positive or negative predictive value (PPV or NPV). We investigated the potential utility of miRNA biomarkers to overcome current limitations and further increase the preoperative diagnostic yield of molecular cytology. Methods:
miRNA expression was measured in surgically resected thyroid lesions and preoperative FNAs by reverse transcription-quantitative PCR using the miRCURY LNA Universal RT microRNA PCR system. Qualitative analyses for
17 distinct oncogenic gene alter
ations in the BRAF
genes were performed with the miRInform
Thyroid testing service. Results:
A comprehensive set of 254 unique thyroid lesions was used to 1/ identify the most significant miRNA candidates differentially expressed between various papillary or follicular carcinomas and benign conditions such as adenoma, diffuse hyperplasia or chronic thyroiditis, and 2/ develop robust linear classification models that could discriminate between benign and malignant lesions with a sensitivity or specificity >90%. A multi-step, 10-miRNA algorithm was further optimized in the surgical training set and in 235 preoperative FNAs by targeting a type I error <5% compatible with high sensitivity/specificity and actionable PPV/NPV when added to gene mutation testing (mutation- or miRNA-positive results both scored as positive). In an independent cross-sectional cohort study on 109 nodules with AUS/FLUS or FN/SFN cytology collected at 12 distinct endocrinology practices across the United States, mutations were detected in 68.6% of nodules with a malignant histology outcome. Among mutation-negative specimens, the optimal miRNA classifier identified 63.6% of the malignant cases and 98.4% of the benign cases. The diagnostic sensitivity and specificity of the combined algorithm was 88.6% (95% confidence intervals: 73.3-96.8%) and 85.1% (95% confidence intervals: 75.0-92.3%), respectively. Importantly, the NPV (95.6% at 24% prevalence) was similar to the NPV previously reported for the Afirma gene expression classifier (94.3% at 24% prevalence) (3) with a 65% increase in true benign call rate. Conclusions:
Multicategorical testing for DNA, mRNA and miRNA in preoperative thyroid nodule FNAs has high PPV/NPV and improves the diagnostic yield of molecular cytology. Independently of variations in cancer prevalence in distinct pathology practices, this novel testing algorithm can identify more benign or malignant nodules and may further decrease the number of unnecessary surgeries and 2-step total thyroidectomies.
(1) Nikiforov et al. 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 2011;96:3390-7 (2) Beaudenon-Huibregtse et al. Centralized molecular testing for oncogenic gene mutations complements the local cytopathological diagnosis of thyroid nodules. Thyroid 2014;24:1479-87 (3) Alexander et al. Preoperative diagnosis of benign thyroid nodules with indeterminate cytology. N Engl J Med 2012;367:705-15.
Disclosure: EL: Consultant, Interpace Diagnostics, Former employee, Asuragen. SB: Consultant, Interpace Diagnostics, Former employee, Asuragen. DW: Employee, Asuragen. TJG: Consultant, Interpace Diagnostics, Clinical Advisory Board member, Asuragen.
*Please take note of The Endocrine Society's News Embargo Policy at https://www.endocrine.org/news-room/endo-annual-meeting