Lymph node sampling in localized neuroblastoma: a Pediatric Oncology Group study.

BACKGROUND/PURPOSE: Lymph node (LN) sampling was required by the Pediatric Oncology Group (POG) staging for neuroblastoma and currently is required as a part of the International Neuroblastoma Staging System (INSS). This retrospective study of planned lymph node sampling in patients with localized neuroblastoma was carried out with the intent of assisting surgeons in carrying out this procedure. The report documents the POG experience where LN, both uninvolved and involved with tumor, were found based on site of primary.

METHODS

From 391 patients with localized neuroblastoma of the abdomen, chest, and neck, 238 patients had LN sampling at the primary operation, and these patients constitute the major part of the study. In addition, 89 patients had a carefully documented search for LN, and 64 had neither search nor biopsy. The operative note, pathology report, and surgical study sheet were used in the 238 patients based on the site of the primary tumor to determine which nodal groups or basins underwent biopsy, and in which groups tumor was found.

RESULTS

The pattern of drainage, based on the primary site of abdominal tumors, favored an arterial rather than venous pathway. Primary tumors and metastatic LN were more numerous on the left side. The abdominal drainage followed three pathways: (1) infrarenal tumors from the left and midline were associated with paraaortic LN; (2) right infrarenal tumors were associated with LN in the paracaval basin; (3) with suprarenal primaries and with both adrenals, the superior mesenteric-portal-celiac basins were most productive for nodal sampling. Tumor was found most frequently in the left adrenal-renal basin and in the paraaortic basin. The actual number of LN sampled in a single case varied from 1 to 19 LN, with a mean number of LN based on stage and primary from one to seven LN. The tumor spread in LN was consistent with a "watershed" course, but this was not statistically significant. Patients for whom LN were sought had a better outcome, contrasting with the patients in whom LN were not sought or in whom nodal sampling was not possible.

CONCLUSIONS

The experience in this study is consistent with previous descriptions of the lymphatic drainage of the retroperitoneal area. Delineation of the various basins as they relate to the site of the primary tumor should assist the surgeon in lymph node sampling. The role of LN involvement still remains unclear in the light of current studies of biological factors and histopathology as determinants of "risk groups." It is hoped that this study will enable ongoing and future studies to clarify this problem. The adult experience with breast cancer and with melanoma has indicated a continued importance of anatomic factors (including LN status) along with biological factors.