Importance of CD34+ cell subsets in autologous PBSC transplantation: the mulhouse experience using CD34+CD38- cells as predictive tool for hematopoietic engraftment.

Over the years, various biological parameters have been proposed for predicting rapidity and long term maintenance of hematopoietic engraftment after peripheral blood stem cell transplantation (PBSCT). Determination of the graft content in CFU-GM was the only one available until the end of the eighties. But, for technical reasons, and also because it does not actually evaluate the self-renewal potential of the cell products reinfused, it has now been commonly replaced by the determination of CD34+ cell amounts, which are known to contain the pluripotent hematopoietic stem cells. However, a frequent discrepancy still exists between the number of CD34+ cells reinfused and the engraftment efficiency. We have recently demonstrated a higher accuracy of the numbers of CD34+38- cells contained in graft products to predict rapidity of trilineage engraftment, which has further been confirmed by other investigators. Furthermore, we and others, have proposed a threshold dose of 5 x 10(4) CD34+38- cells/kg b.w. below which the trilineage engraftment kinetics are significantly slower and unpredictible. This "cut-off" value also appears to be a realistic clinical tool to decide if hematopoietic growth factor(s) must be administered or not after PBSCT. Indeed, when for example, rh-G-CSF administration after transplant of CD34+38- amounts < 5 x 10(4) kg has indisputable positive effects on the rapidity of neutrophil engraftment, length of hospitalization and posttransplant costs, enough to make it fully justified in this situation, it is absolutely not the case when it is administered after reinfusion of CD34+38- cell amounts > 5 x 10(4) /kg. In this case, posttransplant rh-G-CSF administration could even result in a decrease in stem cells with self-renewal potential of the graft, which should still raise more concerns for its indiscriminate and costly use.

Impeded normal hematopoiesis in bone marrow of patients with multiple myeloma.

Insufficient output of mature blood cells frequently accompanies the typical impairments of late B cell development in multiple myeloma (MM). In a large group of previously untreated patients, bone marrow samples were analyzed and showed a general decrease of mononuclear cell (MNC) content. Colony growth of granulo-monocytic progenitors in short-term assays is compromised in a substantial number of patients at partly severe degrees, who at the same time show higher plasma cell content and belong to clinically more severe groups; the other patients show normal in vitro growth, contain less plasmocytes in the marrow and belong to varying degrees of aggressiveness. Thus a heterogeneity of the disease is emerging on the level of bone marrow cells which matches with high aggressiveness of the disease in one type. It can be speculated that in this type there are different underlying mutational events compared to the others: besides the characteristic changes in B cell differentiation, here the cellular defects have an impact on normal granulo-monocytic (and other) progenitor recruitment, which is absent in the other cases.

Positive selection of CD34+ cells: a short review of the immunoadsorption methods currently available for experimental and clinical use. Report on the "2nd European Workshop on stem Cell Methodology," Mulhouse, France, May 3-7, 1993.

At the "2nd European Workshop on Stem Cell Methodology," held in Mulhouse, France, on May 3-7, 1993, part of the meeting was dedicated to the positive selection of CD34+ cells. All devices that are currently in use, or will be available in the near future, were explained and practically demonstrated using human cell populations by scientists involved in their development. In this paper, a review of these methods is given in the form of a short description, together with the data presented in Mulhouse and available from the literature.

CD34+ cells and CD34+CD38- subset from mobilized blood show different patterns of adhesion molecules compared to those from steady-state blood, bone marrow, and cord blood.

As suggested previously, a down-regulation of some cellular adhesion molecules (CAMs) on CD34(+) hematopoietic progenitor cells (HPC) may contribute to their egress from bone marrow (BM) to peripheral blood (PB) by decreasing their adhesion to BM stromal cells. Besides counting the percentage of CAM-positive cells, we decided to define clearly the antigen density (AgD) of the CAM on mobilized- and steady-state CD34(+) HPC using QIFIKIT calibration beads. Five sources of cells were compared: PB and BM from normal donors (nPB, nBM) cord blood (CB), mobilized PB obtained from leukapheresis products (LKP), and mobilized BM (mBM) samples. In our study the CAM-AgD was the lowest on CD34(+) cells in LKP which, on the contrary, contained the highest percentage of CD117(+), CD54(+), CD58(+) cell subsets. As for CB, a greater proportion of CD44(+) and CD62L(+) cells was observed in LKP than in other products. The LKP-CD34(+) cell population contained a greater percentage of CD11a(+) cells when compared to mBM, but the lowest percentage of CD49d(+) and CD49e(+) cells when compared to all products. The proportion of the CD34(+)CD38(-) immature subset expressing CD11a, CD44, CD54, or CD62L was greater in LKP than in mBM; the CD62L-AgD was higher in LKP than in mBM. This quantitative analysis clearly showed a downregulation of all CAM on LKP-CD34(+). The CD44, CD62L, CD11a, and CD54 AgD decrease appears to be specifically involved in the egress of the CD34(+) subsets into PB. The control of antigen density of these adhesion molecules is likely to be clinically important for effective mobilization of HPC as well as for rapid engraftment following HPC transplant.

Ambiguous patterns in cellular differentiation in a case of "M3 variant" leukemia.

We report a case of acute non lymphoblastic leukemia in which clinical and cytological patterns corresponded closely to the M3 variant as defined in the FAB classification, although we did not find the characteristic t (15; 17) chromosomal translocation. However, cytochemistry, DNA content studies and immunophenotyping showed unusual patterns suggesting a monocytic differentiation of most of the blast cells, while a smaller population of blasts showed in contrast typical markers of granulocytic lineage.

Mature accessory cells influence long-term growth of human hematopoietic progenitors on a murine stromal cell feeder layer.

A recently described long-term culture system for early human progenitor cells was established with the murine preadipocyte stromal line FBMD-1 grown in 96-well plates; cobblestone areas formed by inoculated hematopoietic cells are determined in a limiting dilution setting after five weeks' culture. To compare the capacity of cobblestone-area-forming cell (CAFC) formation by bone marrow and leukapheresis products in this system, mononuclear cells (MNC) of both origins were cultured. As related to CD34+ cell content, CAFC yields after five weeks' culture were in the same range in bone marrow and leukapheresis stemming from patients with efficient mobilization of hematopoietic cells. In purified CD34+ cell fractions, the CAFC yield per inoculated cell number was considerably higher than in MNC; however, if the CAFC number was related to the inoculated CD34+ cell number in MNC and after purification, the yield was four to eight times decreased in purified fractions. Addition of the mature cells brought the CAFC yield back up to the numbers obtained in the unseparated MNC fraction. By contrast, slightly more advanced progenitors per CAFC were found in cultures of purified hematopoietic cells from both origins than in whole MNC. The results suggest that mature human accessory cells give noticeable support to recruitment of early progenitors on this feeder but lead to lower yield of GM progenitors.

Mobilization of peripheral blood stem cells with chemotherapy and cytokines in multiple myeloma.

In an attempt to offset the impaired hematopoietic progenitors' mobilization and collection which are frequently encountered in multiple myeloma (MM), we have started a pilot study to evaluate the ability of a combination of high-dose melphalan (HDM) and sequential s.c. administration of recombinant human interleukin 3 (rhIL-3) and rh-granulocyte colony-stimulating factor (G-CSF) to mobilize blood cells (BC) in MM patients. Two different schedules for administration were successively tested. Schedule A consisted of IL-3 (5 micrograms/kg/d) from day 7 to day 11 after HDM followed by G-CSF (5 micrograms/kg/d) from day 12 to day 20. Under schedule B, HDM was followed by IL-3 alone at the same dosage from day 1 to day 3, IL-3 and G-CSF (idem) from day 4 to day 7 and G-CSF alone from day 8 until completion of apheresis. Two patients (one previously untreated, one having received prior chemotherapy for one year) underwent schedule A; three patients (one previously untreated, two pretreated) underwent schedule B. The post-HDM aplasia was not shortened in schedule A patients in comparison to what we usually observed following HDM alone (25 days) correlated with a very moderate two- to three-fold CD34+ cell increase. Only one patient was further transplanted with apheresis products: the post-transplant granulocyte recovery was slower than usual (16 days versus 12 days) while platelet count never recovered over 20 x 10(9)/l. In contrast, the post-HDM aplasia was significantly shortened in two of the schedule B patients (3 to 10 days) and was followed by a 25- to 165-fold increase in CD34+ cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Comparative analysis of class I, II and III epitope-detecting CD34 monoclonal antibodies by quantitative flow cytometry.

UNLABELLED: The aim of this study was to compare different CD34 monoclonal antibodies (MAbs) belonging to three different classes: MY10 class I, QBend10 class II, a mixture of three selected MAbs class I and II designated as CD34 Pool, and 8G12 class III. Bone marrow (BM) samples from 13 healthy donors were analyzed for: 1) percentage of CD34+ cells, 2) quantitative expression of CD34 epitopes (antigen's density - AgD) using a quantitative indirect immunofluorescence (QIFI) test, 3) study of CD34+ cell subsets defined by CD34 and CD38 coexpression. 8G12 MAb showed the highest reactivity with regard to the percentage of detected CD34+ cells and AgD on these cells. A nearly identical percentage of CD34+ cells was detected with CD34 Pool, but with a lower AgD. With QBend10, the percentage of CD34 expressing cells was insignificantly decreased and the AgD was slightly lower. The expression of the MY10 epitope was the lowest and was detected on the lowest number of CD34+ cells. Concerning CD34 and CD38 coexpressing subset, we observed that 8G12 class III MAb detected CD34loCD38dim cells with comparable efficiency with MY10 class I MAb, but with significantly higher level than QBend10 class II and CD34 Pool class I+II MAbs. The CD34hiCD38dim subset was detected with the same efficiency by QBend10, CD34 Pool or 8G12 MAbs but with significantly higher frequency than MY10 MAb. IN CONCLUSION: class II and III MAbs appear preferable for flow cytometric quantification of CD34+ cells; for CD34+ cell subsets determination class III MAbs should be suitable.

Difficulties in collecting peripheral blood stem cells in chronic myeloid leukemia related to persistence of the leukemic cell clone.

Eight chronic myeloid leukemia (CML) patients ineligible for allogeneic bone marrow transplantation (BMT) were intensively treated by a myeloablative chemotherapy identical to the treatment that we use in acute myeloid leukemia (AML). The objectives of such an intensive treatment were both to reduce the size of the leukemia stem cell mass as much as possible and subsequently to allow a better mobilization of the residual Ph-negative (Ph-) stem cells. Cytogenetic analyses were systematically performed on blood-derived stem cells collected at the hematopoietic recovery phase following post-chemotherapy aplasia. The length of aplasia did not correlate with the evolutive stage of the disease, but was negatively correlated with the total colony forming units-granulocyte macrophage (CFU-GM) amounts collected. The cytogenetic abnormality remained present in most cases in all metaphases counted in leukapheresis products. Three patients were transplanted with these leukapheresis products. One died due to sepsis before engraftment; the two others engrafted very slowly, while Ph-positive (Ph+) cells were found at post-transplant controls. These disappointing results suggest that the myeloablative chemotherapy used in this study has not resulted in satisfactory advantages for the proliferation of residual normal stem cells over the expansion of the Ph+ clone.

Human accessory cells have a humoral bystander effect on CAFC growing on murine feeder.

Human early hematopoietic progenitors from bone marrow (BM) and leukapheresis products (LP) are highly proliferative in presence of accessory cells in standard culture on the murine FBMD-1 cell feeder with weekly addition of human interleukin-3 (HuIL-3) and granulocyte-colony stimulating factor (HuG-CSF). If however purified CD34+ cells are cultured under otherwise identical conditions, cobblestone areas (CAFC) formed by the same number of target cells are diminished by more than 1 log, as we showed previously. This suggests that mature cells are involved in growth of early progenitors. To determine whether this bystander effect is mediated by soluble growth factors, or by direct cell-to-cell contact with early progenitors, we stimulated mature plastic adherent cells separately and tested the resulting conditioned supernatant (ACS) on CAFC and colony-forming unit-granulocyte-macrophage (CFU-GM) production. In ACS-complemented standard cultures of purified CD34+ cells, the yield of CAFC was up to 1 log higher if compared to parallel cultures without ACS. Likewise, the CFU-GM production was enhanced in presence of ACS, especially in the adherent fraction of the culture. When CD34+ cell cultures were performed with ACS but without added interleukin-3 (IL-3) and granulocyte colony-stimulating factor (G-CSF), CAFC production was in the same range as if these growth factors were added alone. Addition of anti-G-CSF antibody (Ab) to ACS decreased CAFC recruitment significantly, whereas anti-IL-3 Ab had no significant effect. These findings suggest that ACS complemented with IL-3 and G-CSF replaces the accessory cells largely; this is not only due to presence of G-CSF, because ACS in combination with recombinant growth factors mounts CAFC yield higher than saturating amounts of growth factors alone do. There must be further synergizing soluble factors in the supernatant.

Isolation and identification of two CD34+ cell subpopulations from normal human peripheral blood.

Circulating CD34+ progenitors were separated from normal human peripheral blood on the basis of size and density by counterflow centrifugal elutriation (CCE). The CD34+ cells, 0.15% of peripheral blood mononuclear cells, were heterogeneous with respect to their elutriation characteristics, mainly size and density. The least mature CD34+ cells, characterized by lack of CD38 antigen, were predominantly found in the small lymphoid cell fraction. In fractions containing larger and denser cells (large lymphocytes, monocytes, and granulocytes), CD38 was increasingly expressed on the CD34+ cells, as were lineage commitment markers CD10 (B lymphoid), CD33 (myeloid), CD13 (myelomonocytic) and CD71 (erythroid) antigens. The smaller and less dense CD34+ cells expressed CD34 antigen brightly while the larger and denser CD34+ cells expressed it dimly. The smaller and less dense CD34+ high cells failed to establish colony growth in short-term culture while the larger and denser CD34+low cells gave rise to high counts of colony forming units-granulocyte macrophage (CFU-GM). Physical separation on the basis of size and density by CCE differentiates between two main classes of steady-state CD34+ cells from normal human peripheral blood. The smaller and less dense CD34+high cells correspond to the earliest progenitors that express differentiation markers poorly but CD34 antigen brightly, do not give rise to short-term colony growth in vitro, and thus represent indirect evidence for pluripotent hematopoietic stem cells (PHSC). The larger and denser CD34+low cells are the more mature progenitor cells, already committed to myeloid, lymphoid or erythroid differentiation but only dimly expressing CD34 antigen, and these cells were responsible for short-term colony growth in vitro.

Primordial role of CD34+ 38- cells in early and late trilineage haemopoietic engraftment after autologous blood cell transplantation.

In order to better define which cell subset contained in graft products might be the most predictive of haemopoietic recovery following autologous blood cell transplantation (ABCT), the relationships between the amounts of reinfused mononuclear cells (MNC), CFU-GM, total CD34+ cells and their CD33 and CD38 subsets. and the successive stages of trilineage engraftment kinetics, were studied in 45 cancer patients, using the Spearman correlation test, a linear regression model and a log-inverse model. No relationship was found between the infused numbers of MNC, CD33+ and CD33- subsets observed and the numbers of days to reach predetermined absolute neutrophil (ANC), platelet and reticulocyte counts. The infused numbers of CFU-GM, CD34+ and CD34+ 38+ cells correlated inconstantly with haemopoietic recovery parameters. The strongest and the most constant correlations were significantly observed between the infused numbers of CD34+ 38- cells and each trilineage engraftment parameter. The log-inverse model determined a threshold dose of 0.05 x 10(6) (= 5 x 10(4)) CD34+ 38- cells/kg, below which the trilineage engraftment kinetics were significantly slower and unpredictable. Post-transplant TBI-conditioning regimens increased the low cell dose-related delay of engraftment kinetics whereas post-transplant administration of haemopoietic growth factors (HGF) seemed to abrogate this delay. This would justify clinical use of HGF only in patients transplanted with CD34+ 38- cell amounts lower than the proposed threshold value. This study suggests that the CD34+ 38- subpopulation, although essentially participating in late complete haemopoietic recovery, is also composed of committed progenitor cells involved in early trilineage engraftment.

Peripheral blood and BM CD34+ CD38- cells show better resistance to cryopreservation than CD34+ CD38+ cells in autologous stem cell transplantation.

BACKGROUND: We and others have shown a critical role for CD34+ CD38- cells in hematopoietic recovery after autologous stem cell transplantation (ASCT), in particular for platelet reconstitution. Thus a routine assessment of CD34+ CD38- cells in freezing-thawing procedures for autografting could represent an important tool for predicting poor engraftment. METHODS: To compare the impact of cryopreservation on CD34+ CD38+ and CD34+ CD38- hematopoietic stem cell subsets, 193 autograft products collected in 84 patients with malignancies were assessed before controlled-rate cryopreservation in 10% DMSO and after thawing for autografting. RESULTS: Cell counts after thawing were significantly different from the pre-freezing counts for total CD34+ (P<0.0001) and CD34+ CD38+ (P<0.0001) cells, but not for CD34+ CD38- cells (P=0.252). Median losses for CD34+, CD34+ CD38+ and CD34+ CD38- cells were, respectively, 11.8%, 11.4% and 0.0%. The magnitude of fresh/post-thawing percentage cell variation was significantly different when comparing between the CD34+ CD38+ and CD34+ CD38- cell subsets (P<0.001). Moreover, CD34+ CD38- cells exhibited recovery values > or =100% in 85/160 graft products, compared with 51/193 in CD34+ CD38+ cells (P<0.0001). Also, recovery values > or =90% were significantly better in the CD34+ CD38- (98/160 grafts) than in the CD34+ CD38+ subsets (89/193 grafts) (P<0.01). DISCUSSION: In this work we have demonstrated that CD34+ cells that do not express the CD38 Ag show a significantly better resistance to cryopreservation. This could represent another example of the particular ability of less committed progenitor cells to overcome environmental injuries. Moreover, we consider routine assessment of CD34+ CD38- cells before freezing as clinically relevant, but post-thawing controls may be avoided because of their good resistance to freezing.

Role of the CD34+ 38- cells in posttransplant hematopoietic recovery.

Using three different statistical tests in parallel, we showed in a preliminary study that neither mononuclear cells, CD34+ 33+ or 33- cells, nor CD34+ 38+ cells significantly correlated with engraftment kinetics following autologous blood cell transplantation (ABCT). We additionally demonstrated here, in a series of patients suffering from malignant diseases, that the graft content in CD34+ 38- cells is individually a more sensitive indicator of the earliest, as well as the latest post-ABCT trilineage hematopoietic recovery than the colony-forming units-granulocyte-macrophage and even the total CD34+ cell content. This suggests that the CD34+ 38- cell population is itself subdivided into two more subsets, one being already lineage-committed and responsible for short-term engraftment, the other containing only very primitive hematopoietic cells responsible for sustained engraftment. Strong arguments favor the probability that these subsets correspond to HLA-DR+ and DR cells, respectively. We also defined an optimal threshold value of 0.05 x 10(6) CD34+ 38- cells/kg of the patient's body weight (b.w.) above which a rapid and sustained trilineage engraftment safely occurs. In fact, infusion of lower numbers of cells seems to have a more significant impact on long-term compared to short-term neutrophil recovery and on platelet kinetics engraftment. We additionally looked for the eventual influence on engraftment time of the type of disease, and of post-ABCT administration of hematopoietic growth factors (HGF). When the type of disease appeared to have no influence on the engraftment time, posttransplant HGF administration significantly reduced the time to trilineage engraftment in patients transplanted with < 0.05 x 10(6) CD34+ 38- cells, thus justifying it in case of reinfusion of low numbers of CD34+ 38- cells. On the other hand, the administration of HGF after infusion of more than 0.05 x 10(6) CD34+ 38- cells/kg b.w. did not hasten more, or only very little, the engraftment time, thus becoming not only unprofitable for the patients but costly as well.