The impact of surgical radicality on outcome in childhood neuroblastoma.

AIM: Improvement of treatment results for neuroblastoma (NB) has been achieved during recent years, especially by intensifying therapy for advanced NB. Surgery, however, has not contributed very much to this progress and there is still controversy regarding the best approach for high-risk NB. We therefore attempted to find criteria for a differentiated strategy for tumour resection in NB.
METHODS: We retrospectively analysed the data of 2251 NB patients treated in the German Cooperative NB Studies NB79 - NB97 (1979 - 1999) including patients' age, tumour stage, MYCN oncogene status, surgical intervention, completeness of resection, surgical complications and outcome.
RESULTS: 1148 patients had a localised NB (stage 1 - 3), 878 had stage 4, and 225 had stage 4 S disease. 2112 patients underwent surgery. Tumour resection as the final result of primary or delayed operation was complete in 1403 (66.4 %), incomplete in 449 (21.3 %), and only a biopsy was carried out in 260 (12.3 %) cases. Complete resection was performed most often in localised NB (73.5 %), less often in stage 4 NB (59.2 %) and in stage 4 S (54.5 %). Clinically relevant complications occurred in 19.2 % of all operations independent of the time of resection during treatment, patients' age, tumour stage, tumour site and MYCN status. For 1787 patients from the studies NB79 - NB90 the probability of 5-year event-free survival (EFS) could be analysed and correlated with surgical radicality. In localised NB of patients aged > 1 year there was a significant difference in 5-year EFS between complete, incomplete, and no substantial resection ( P < 0.0001), while this was not the case in infants < 1 year. In the early study period (NB79 - NB85) patients with stage 4 NB had a significantly better 5-year EFS after complete excision of the primary tumour. In contrast, this difference was not confirmed during the later period (NB90) with intensified drug therapy. Timing of surgery did not influence prognosis. For infants with stage 4 S NB there was no difference in outcome after complete or incomplete resection of the primary tumour. MYCN status did not alter the significant relevance of tumour resection for outcome in localised NB. In contrast to MYCN non-amplified stage 4 NB, however, stage 4 NB with MYCN amplification had a significantly better outcome if the primary tumour could be completely or incompletely resected, compared to a biopsy only.
CONCLUSIONS: Radical tumour resections with the risk of surgical complications are only justified in children > 1 year with a localised NB. In stage 4 NB, the primary tumour may be resected during intensive multimodal treatment without risky complications. Complete or incomplete resection of the primary tumour improves prognosis only in MYCN-amplified stage 4 NB. Stage 4 S NB with non-amplified MYCN are subject to spontaneous regression; a resection after chemotherapy may be indicated in cases of non-sufficient regression or growth of the primary tumour.