Multiple Regression Analysis of Predictors of Bone Scintigraphy Positivity in Patients with Head and Neck Cancers

摘要

Purpose The value of bone scintigraphy (BS) in patients with head and neck cancers (HNCs) has been questioned, with conflicting findings regarding positivity yield with some reports recommending BS be omitted from work-up of HNC patients since it rarely yields positive results. This study aims to determine the positivity yield of BS in HNCs and to determine predictors for BS positivity to help tailor appropriate BS utilization. Methods BS studies of HNC patients were reviewed, the positivity yield was determined. Clinical predictors for BS positivity including age, sex, site of cancer, staging, histological grading were analyzed using univariable and multivariable logistic regression. Results Among the 259 BS studies included, 35 (13.5), 194 (74.9), and 30 (11.6) were positive, negative, and equivocal for bone metastasis, respectively. After exclusion of equivocal cases, 229 were analyzed in the regression models. Independent predictors of BS positivity include site of tumor at the nasopharynx (OR 4.37, 95 C.I. 1.04-18.41, p = 0.044), age less than 45 years (OR 3.01, 95 C.I. 1.24-7.33, p = 0.015), and presence of distant metastasis to other organs (OR 3.84, 95 C.I. 1.19-12.43, p = 0.025). Conclusions In contrast to several studies, bone metastasis as detected by BS was found in a relatively high proportion of patients with HNCs. Independent predictors of BS positivity include the age of less than 45 years, tumor site at the nasopharynx, and the presence of extraskeletal distant metastasis. BS could be useful in patients with these characteristics which enhance the pretest probability of bone metastasis.