Detection of neuroblastoma in bone marrow by immunocytology: is a single marrow aspirate adequate?

Abstract

Except at diagnosis and relapse, when gross disease is present, histologic evaluation is less sensitive than immunocytology (IC) of bone marrow for detecting metastatic neuroblastoma. We examined whether the chance of a positive IC from a single marrow site was comparable to an IC of pooled marrow from multiple sites. We carried out 47 marrow examinations on 29 patients with high-risk neuroblastoma on therapy. Each examination consisted of histologic evaluation of four aspirates and two biopsies (six sites), IC of a 2.5-5-mL heparinized sample from a single site (the right posterior iliac crest; IC-RPIC), as well as IC of 8-10 mL of heparinized marrow pooled from bilateral anterior and bilateral posterior iliac crests (IC-pooled). IC was performed using a panel of monoclonal antibodies specific for ganglioside GD2. The number of GD2-positive tumor cells detected by IC-pooled had a strong linear correlation with that by IC-RPIC (R = 0.91), although IC-pooled detected an average of 3.3 times more GD2-positive cells. Of 47 examinations, 15 tested positive by histology (6 sites), 20 by IC- pooled, and 17 by IC-RPIC. Among 29 patients, the level of agreement between IC-RPIC and IC-pooled was generally good (kappa statistic > or = 0.72), giving a false negative rate of < or = 30%. Compared to conventional histologic examinations, immunocytology of a single marrow aspirate generally agrees with that of marrow pooled from six sites. However, the false negative rate may be too high in the setting of examination prior to bone marrow or stem cell harvest.