CD34 selection as a stem cell purging strategy for neuroblastoma: Preclinical and clinical studies

Background. The suitability of CD34 selection for purging peripheral blood progenitor cells (PBPC) collected from patients with neuroblastoma (NB) has been called into question, largely because of reports of detection of low levels of CD34 on the surface of some NE cell lines and tumors. Procedure. We used three approaches to address the issue of purging of NE from stem ce ll specimens and possible labeling of NE: 1) Flow cytometric detection of C D34 on NE cell lines. We assessed CD34 expression using a panel of anti-CD3 4 monoclonal antibodies (MoAbs) including 9C5, 12.8, and QBend10 and showed no increase in labeling over secondary-only control. 2) Spiking experiment s with the Isolex 50 system. NE cell lines were used to contaminate aliquot s of PBPC collections, after which the products were purified using the Iso lex 50. Purging of NE was assessed by quantitative multiplex RT-PCR (TaqMan system) using a tumor-specific transcript, CAGE. We demonstrated >2 logs o f tu- mor cell depletion from these specimens. 3) Analysis of clinical spec imens. PBPC pre- and post-CD34 selection were analyzed from patients treate d on the CHP-594 transplant trial. Results. In nine specimens selected usin g the Ceprate LC CD34 selection system where tumor was detectable by immuno cytochemistry preselection, we observed >2.4 to >4.6 logs of NE purging aft er selection. We then analyzed 23 aliquots of PBPC infused into patients po st-CD34 selection and compared them to the product preselection; 20/23 spec imens showed depletion of NE, although some level of GAGE message was obser ved in most post-CD34 selection specimens. Conclusion. These data show that purging of NE from PBPC specimens using CD34 selection is feasible, yieldi ng infused products that are negative at the level of ICC but often positiv e at the level of RT-PCR. (C) 2000 Wiley-Liss, Inc.