Purging Grafts Via CD34+ Cell Selection May Show Benefits

NASHVILLE--Results of a trial of positive-selection purging inbreast cancer patients undergoing high-dose chemotherapy/autologousbone marrow or peripheral blood transplantation show that patientswho have no evidence of breast cancer in the graft after purginghave longer progression-free survival at a median follow-up of18 months than patients who have persistent evidence of tumor.Purging is performed by selection of CD34+ marrow or peripheralblood progenitor cells (PBPCs).

Speaking at a scientific session of the annual meeting of theAmerican Society of Hematology (ASH), Elizabeth J. Shpall, MD,described her study of 121 patients with advanced (47) or high-risk(74) breast cancer undergoing transplantation, as well as ongoingefforts at ex vivo expansion of CD34+ selected stem cells.

Patients were divided into several cohorts who received eitherCD34+ marrow, blood, or both. Of the 47 stage IV patients (whowere treated on phase I protocols), 13 were found to have no evidenceof cancer in their grafts after purging, and these patients hada significantly higher duration of disease-free survival thanthe 34 patients whose grafts still contained tumor following positiveselection, said Dr. Shpall, associate professor of medicine, Universityof Colorado Health Sciences Center, Denver.

In a third group of 55 patients in whom no tumor cells were everdetected in their grafts, disease-free survival times fell inbetween those of the other two groups and did not differ significantlyfrom either one.

In a multivariate analysis of all the patients (including the74 high-risk patients enrolled in phase II protocols), the graftsthat contained breast cancer both before and after positive selection(referred to as positive-to-positive) were used as the reference.

"The purification of the graft to negativity and enrollmenton phase II as opposed to phase I studies were the only two covariantsthat independently predicted for a significantly better disease-freesurvival," Dr. Shpall said. P values were identical for thetwo variables at .005. When these two variables were combined,P value was even more significant.

Factors found to have no significant effect on relapse-free survivalwere patient age; disease-free interval between diagnosis andtransplant; transplantation with CD34+ marrow, blood, or both;estrogen receptor status of the tumor; and administration of adjuvantchemotherapy, Dr. Shpall said.

She pointed out that the number of patients in the relevant cohort(positive- to-negative) was small, and the duration of follow-uphas been too short to make definitive statements about the efficacyof this purging procedure.

The Purging Method

CD34 is an "obvious target" for purging of marrow orblood in breast cancer patients, Dr. Shpall said, since CD34 antigendoes not appear to be expressed on the surface of breast cancercells.

In this study, bone marrow buffy coat cells or G-CSF (Neupogen)-mobilizedPBPCs from three leukaphereses were incubated with the biotinylatedanti-CD34 antibody 12.8. The cells were then applied to a columnof avidin-coated beads (Ceprate SC Stem Cell Concentration System,developed by Dr. Ronald Berenson and his colleagues at CellPro,Inc., Bothwell, Wash), and the CD34+ cells were isolated.

Using a quantitative immunohistochemical assay developed by Dr.Wilbur Franklin at the University of Colorado, every graft wastested before and after positive selection, to calculate the tumorburden for each graft. In those grafts in which tumor had beendetected before purging, CD34+ selection reduced the number oftumor cells by a mean of 2 logs (range, 1 to more than 4 logs).

No Engraftment Delays

Engraftment times of CD34+ selected stem cells were comparableto standard times produced with unpurified cells, Dr. Shpall said."For the two cohorts that received peripheral blood only,granulocytes engrafted within 10 days and platelets within 14days," she said. This is in sharp contrast to negative purgingtechniques, such as 4-HC, which may damage or deplete the normalmarrow and extend engraftment times to between 28 and 36 days,she noted.

In the two cohorts that received selected PBPCs only, an averageof 2.3 million CD34+ cells per kilogram of patient weight wereinfused, including a total of 9.8 million CFU-GMs (progenitorcells that include both myeloid and megakaryocyte colonies), Dr.Shpall said.

This number--9.8 million--was then chosen as a target for ex vivoexpansion studies at Colorado, since it was associated with rapidand complete engraftment in the vast majority of patients. Theexpansion procedure is being tested in an attempt to further depletetumor cells from grafts, since clearly many patients still havetumor cells after CD34+ selection, she said.

Over the course of a year of preclinical studies of various methodsof culturing G-CSF-mobilized CD34-selected peripheral blood, conductedby Malcolm Purdy, an optimized method of CFU-GM expansion wasdetermined.

This ex vivo expansion method was then tested in peripheral blooddonated by four breast cancer patients. "The number of CFU-GMsgenerated was five times higher than the target number of 9.8million, and we postulate that if infused, this product shouldproduce prompt engraftment," Dr. Shpall reported.

All four of the patients had tumor in their graft prior to expansion,and, following expansion, no tumor was detected in three patients,nor in an additional two patients who have donated blood for study.The patient who remained positive had only rare tumor cells.

In a separate talk at a symposium on purging sponsored by ActivatedCell Therapy, Inc. (Mountain View, Calif), Dr. Shpall said thata randomized trial would be the best way to show definitivelythat purging is beneficial. "Unfortunately, it would be difficultto conduct such a trial in our current climate," she said."The next best way to determine the efficacy of purging isto gene-mark the cells, and we plan to do that."

The Colorado group plans to mark the CD34+ selected cells thougha gene transfer method, "infuse them, track them, and showthat marked cells do relapse," Dr. Shpall said. Such studiescould be forthcoming in 5 years, she said, and may obviate theneed for large randomized clinical trials.