Accuracy of Preoperative Diagnosis of Pediatric Neck Masses

Abstract

Objective: Diagnosing pediatric neck masses is challenging, often requiring open biopsy. Epidemiology and diagnostic tests utility have been described, but accuracy of clinicians’ judgment has not. We will describe/analyze preoperative diagnostic accuracy and identify the relative contributions from clinical, radiological, and pathological data. Method: Retrospective review of pediatric patients’ charts with an undiagnosed neck mass who underwent open biopsy from January 1986 to December 2011. 248 charts were reviewed, with approximately 30 more to review prior to presentation. Data collection identified preoperative accuracy and relative contributions of clinical clues, imaging studies, and FNA. Results: Preoperative accuracy varied for each category of diagnoses. Congenital masses were diagnosed with 76% (n = 98) accuracy. Nodal inflammatory masses were correct in 55% (n = 85) of cases, whereas non-nodal inflammatory were correct in only 13% (n = 8). Benign tumors were diagnosed in 71% (n = 17), yet only 64% (n = 14) in malignant cases. For miscellaneous diagnoses, accuracy was 35% (n = 26). Chi-squared test indicates accuracy of diagnosis is significantly different between categories. Thirty-eight percent of patients underwent preoperative CT scan, 11% had ultrasound, and 12% had an FNA of the suspicious mass. Sixteen percent of patients had no ancillary tests and all had correct diagnoses. Conclusion: Clinical and radiographic clues often establish accurate diagnoses in pediatric neck masses. Most inaccuracies occurred in masses with nonspecific clinical clues as in inflammatory lymphadenopathy and malignancies. Delaying diagnosis of malignancies is unacceptable; therefore, a certain percentage of unnecessary open biopsies should be accepted despite the most judicious preoperative workup.