Abnormal Rib Number in Childhood Malignancy

Abstract

Retrospective review. To determine if rib anomalies are present in pediatric malignancies in the United States. Scoliosis surgeons view radiographs of the entire spine, counting the number of ribs. A European study noted that rib anomalies were more common in certain malignancies. We wished to determine if this is also true in the United States. If so, the potential for screening, early detection of malignancy, and a better understanding of tumor biology is possible. A retrospective review of 218 children with malignancy and a control group of 200 children with polytrauma or suspected child abuse was performed. Chest radiographs were reviewed to determine the number of ribs, and the presence of rib anomalies. 24 ribs was considered normal, <24 or >24 was considered abnormal. P < 0.05 was considered significant. The average age was 6.8 +/- 5.5 years and number of ribs was 23.8 +/- 0.6. Rib number was normal in 86.8%. There were significant differences between the malignancy and control groups in age (control, 5.7 +/- 5.1 years; malignancy, 7.8 +/- 5.7 years, P = 0.00007), rib number (control, 23.9 +/- 0.5; malignancy, 23.7 +/- 0.7, P = 0.001), and normal/abnormal rib counts (control, 92% normal; malignancy, 82% normal, P = 0.003). In the malignant group, 50% had a lymphoproliferative malignancy, 33% a solid tumor, and 17.0% a neural tumor. Neural malignancies had a higher incidence of rib abnormalities compared with lymphoproliferative or solid malignancies (P = 0.01). Relative to the control group, those with a neural and lymphoproliferative malignancy were 6.23 (95% CI, 2.7-14.5) and 2.0 (95% CI, 1.0-4.1) times more likely to have an abnormal rib count. Homeobox genes, important in vertebral and rib sequencing, are abnormally expressed in many different malignancies. This association is a question of great interest. What is the potential for rib number being used as a predictor of childhood malignancy? Can these findings be expanded to adults? These questions require further research. The association noted in this study is interesting but should not yet be used to alarm parents regarding an increased risk of malignancy in their children.