LYMPHOBLAST MORPHOLOGY IN PREDICTING LEUKEMIC MENINGEAL RELAPSE WITH LOW CHAMBER COUNT AND LYMPHOBLASTS

The diagnostic criteria for meningeal relapse (MR) of acute lymphoblas tic leukemia (ALL) are a cerebrospinal fluid (CSF) chamber count of mo re than five leukocytes per microliter and a cytomorphological evaluat ion revealing lymphoblasts. A dilemma arises when confronted with a pa tient with a low CSF white blood cell (WBC) chamber count and lymphobl asts. We utilized a scoring system to review lymphoblast morphology in 12 such patients. A cell was defined as a lymphoblast if it could not be easily categorized as a lymphocyte, monocyte, histiocyte, or granu locyte. Each lymphoblast was scored on four parameters: presence of nu cleoli, homogeneous distribution of chromatin, nucleocytoplasmic ratio greater than 75%, and nuclear irregularity. Cells were scored without knowledge of the patients' out come. Seven patients eventually develo ped MR by current criteria and five patients never relapsed. The mean lymphoblast scores for patients that did and did not relapse were 2.35 and 1.53, respectively (P<.001). The percent of cells scored as lymph oblasts was also significantly higher in patients that relapsed, 36.9% vs. 19.4% (P=.01). Our study shows that careful cytomorphologic analy sis can predict which patients with low chamber counts and ''blasts'' on cytocentrifuge examination will progress to meningeal relapse. We r ecommend reviewing the definition of MR and using a scoring system whe n confronted with blasts in a low chamber count cerebrospinal fluid sp ecimen. (C) 1997 Wiley-Liss, Inc.