Predictive factors of malignancy in pediatric thyroid nodules

Rashmi Roy, Guennadi Kouniavsky, Eric Schneider, John D. Allendorf, John A. Chabot, Paul Logerfo, Alan P.B. Dackiw, Paul Colombani, Martha A. Zeiger, James A. Lee

Research output: Contribution to journalArticlepeer-review

27 Scopus citations

Abstract

Background: Studies suggest that while most pediatric thyroid nodules are benign, there is a higher rate of malignancy than in adults. We investigate clinical factors that may predict malignancy in pediatric thyroid nodules. Methods: A retrospective review of 207 pediatric thyroidectomies was conducted over 15 years at 2 tertiary hospitals. Analyses examined predictive values of 16 clinicopathologic factors associated with cancer. Positive predictive values (PPVs) of fine-needle aspiration biopsy specimens (FNABs) were analyzed independently. Results: Malignancy occurred in 41% of patients. After excluding missing data, malignancy was more likely with family history of thyroid cancer (34.2% vs 17.7%; P =.111), palpable lymphadenopathy (34.2% vs 2.9%; P =.001), and hypoechoic nodules (52.2% vs 19.2%; P =.016). Palpable lymphadenopathy indicated greater than 2-fold increased risk for malignancy (relative risk, 2.18; 95% confidence interval, 1.56-3.05). PPVs of FNAB results were 0.94 for malignancy, 0.63 for suspicious for malignancy, and 0.55 for indeterminate lesions. PPV for benign FNAB to be benign on final pathology was 0.71. Conclusion: While malignancy is associated with family history of thyroid cancer and hypoechoic lesions, palpable lymphadenopathy had the greatest risk. When compared to adults, a benign FNAB in children is not as accurate and the likelihood that an indeterminate nodule is cancer is greater.

Original languageEnglish (US)
Pages (from-to)1228-1233
Number of pages6
JournalSurgery
Volume150
Issue number6
DOIs
StatePublished - Dec 2011

ASJC Scopus subject areas

  • Surgery

Fingerprint

Dive into the research topics of 'Predictive factors of malignancy in pediatric thyroid nodules'. Together they form a unique fingerprint.

Cite this