PubMed2002
PURPOSE: To review patterns of failure in infants with neuroblastoma and determine late toxicity and efficacy of radiotherapy (RT) on locoregional control. MATERIALS AND METHODS: From 1955 to 1998, 53 children (35 males and 18 females) <1 year old with neuroblastoma were seen at our institution. Twenty-one (40%) were </=1 month of age (neonates). Seven congenital anomalies were seen in 4 children (atrial septal defect, pulmonary valve stenosis, and absent corpus callosum were seen in 1 patient each; the other child had atrial and ventricular septal defect, aortic arch hypoplasia, and mitral valve stenosis). Primary tumor was located in the adrenal gland in 26 (49%), abdomen/nonadrenal in 14 (26%), thorax in 9 (17%), neck in 2 (4%), and pelvis in 2 (4%). All infants were retrospectively staged according to the International Neuroblastoma Staging System (INSS); 8 had Stage 1, 7 Stage 2A, 6 Stage 2B, 15 Stage 3, 6 Stage 4, and 11 Stage 4S. All patients, except 11 with Stage 4S and 4 with Stage 4, had resection of the primary tumor. Sixteen infants had pathologic involvement of regional lymph nodes (LN+). Twenty patients received RT. Kilovoltage beams were used in 10, megavoltage photons in 9, and electrons in 1 child. Fifteen received RT to the primary site and regional nodes (postoperative 8, preoperative 7). Postoperative doses ranged from 15 to 25 Gy whereas preoperative doses ranged from 12 to 31 Gy using a median fraction size of 1.5 Gy. Chemotherapy was employed in 22 of 53 patients (42%) with the most common agents being cyclophosphamide in 22 and doxorubicin in 11. Median follow-up was 13.1 years (range, 2-41 years). RESULTS: Tumor control: The 5-year overall and freedom from progression rates were 79 +/- 9% and 81 +/- 11%; the locoregional control rate was 88 +/- 9%. INSS Stage was a prognostic factor for overall survival (p = 0.03) and freedom from progression (p = 0.035). Gender, age (</=1 month vs. >1 month), and primary site were not found to impact on survival or progression. None of the Stage 1, 2A, or 2B patients recurred. One of 15 Stage 3 and 5 of 6 Stage 4 children recurred (6 distant metastases, 4 local failure). Four of 6 (67%) LN+ patients treated with locoregional RT and 8 of 10 (80%) LN+ patients treated without RT were locally controlled. There was no isolated locoregional relapse. Two Stage 4S patients died of respiratory compromise secondary to hepatomegaly. RT toxicity: For the 20 infants who received RT, 13 are alive with long-term follow-up ranging from 9.3 to 41 years, median 23 years. The 10 and 15-year musculoskeletal toxicity rates were 38.5% and 47.3% for those receiving RT and 3.3% for no RT (p = 0.02, log-rank test). Five of 6 infants <6 months of age and 1 of 7 >/=6 months developed musculoskeletal toxicity. Musculoskeletal effects were seen in 6 RT patients and included bony hypoplasia in 6, scoliosis in 5, soft tissue hypoplasia in 3, slipped capital femoral epiphysis in 2, kyphosis in 1, and osteochondroma in 1. Three required orthopedic intervention, all receiving >/=20 Gy. One child developed bowel obstruction at 21 months and another developed a leiomyosarcoma in the treatment field 34 years after RT. CONCLUSION: Our study shows that most LN+ infants achieve locoregional control without RT. Infants <6 months receiving RT were the most susceptible to musculoskeletal abnormalities. Further studies are needed to determine if cardiovascular anomalies are more frequently seen in children with neuroblastoma.