- Top of page
- MATERIALS AND METHODS
This study aimed to demonstrate that sufficient Ph-negative blood progenitors could be collected following administration of glycosylated rhG-CSF (lenograstim) to patients with Philadelphia chromosome (Ph)-positive chronic myeloid leukaemia (CML) who responded to recombinant alpha-interferon (α-IFN) (Ph-positive marrow metaphases < 35%). 23 patients received lenograstim (150 μg/m2) once daily for a median of 9 d. Peak circulating numbers of white blood cells (36.4 × 109/l), CD34+ cells (24/μl) and colony-forming unit-granulocyte-macrophage (CFU-GM; 1346.5/ml) occurred at a median of day 8, day 8 and day 7, respectively. Two to six (median three) leukaphereses (LK) were performed from days 5 to 12. The median number of mononuclear cells (MNC), CD34+ cells and CFU-GM collected per patient was 7.35 × 108/kg, 2.72 × 106/kg and 10.23 × 104/kg, respectively. 22/23 patients had LK which contained either 104 CFU-GM/kg and/or 106 CD34+ cells/kg; 47 LK (from 20/22 patients) were evaluable for cytogenetics. The median percentage of Ph-positive cells was 0, and 43/47 LK (91%) contained < 35% Ph-positive cells; 25 (53%) were entirely negative. Sixteen of 20 evaluable patients had one or more LK with < 35% Ph-positive cells, and 12 had at least one 100% Ph-negative LK. Mobilization and collection of Ph-negative cells were not influenced by the dose or duration of α-IFN administration before (or during) lenograstim administration or by the quality of cytogenetic response (complete v major) during lenograstim administration. No significant side-effects were observed. Thus, lenograstim administration can result in satisfactory Ph-negative blood progenitor cell collection. Autologous transplantation of such cells may be used when indicated.
Chronic myeloid leukaemia (CML) is an acquired clonal haemopoietic stem cell disorder characterized by the presence of a chromosomal marker, the Philadelphia (Ph) chromosome, which results from a reciprocal translocation t(9,22) (q34;q11) and is present in all cells of the clone ( De Klein et al, 1982 ). This translocation transposes the c-abl proto-oncogene from its normal position on chromosome 9 to a new position on chromosome 22, in proximity to the BCR gene ( Groffen et al, 1984 ). The BCR-ABL hybrid gene or the generated fusion proteins can be detected and analysed using different molecular techniques.
Allogeneic bone marrow transplantation (alloBMT), which is the only curative treatment for CML, is restricted to a minority of patients who have both a human leucocyte antigen identical donor and are < 50 years of age ( Thomas et al, 1986 ). We, and others, have reported that recombinant alpha-interferon (α-IFN) can induce complete (cytogenetic) responses in about 30% of patients who, if not cured, may still have a significantly prolonged survival ( Montastruc et al, 1995 ; Kantarjian et al, 1995 ; The Italian Cooperative Study Group on Chronic Myeloid Leukemia, 1994). Thus, for the majority of patients who either cannot benefit from alloBMT or do not respond to IFN, alternative treatments such as autologous stem cell transplantation (ASCT) must be considered ( Reiffers et al, 1996 ; Goldman, 1990). Previous studies have indicated that ASCT must be performed during the chronic phase as the results observed in patients transplanted during transformation are very poor ( McGlave et al, 1994 ). The different sources of stem cells used to date do not seem to significantly influence the outcome following transplantation; there are no studies showing a significant difference between bone marrow and peripheral blood progenitor cell (PBPC) transplantation, or between purged or unpurged stem cell transplantation, on patient survival ( Reiffers et al, 1994 , 1996; McGlave et al, 1994 ). However, as for patients undergoing ASCT for acute leukaemia, it has been recently reported that, using gene marking techniques, leukaemic progenitor cells (i.e. Ph-positive or BCR-ABL-positive cells) could contribute to, or perhaps cause, relapse following transplantation ( Deisseroth et al, 1994 ). This observation has led investigators to experiment with different techniques of stem cell purging.
Existing ex-vivo purging techniques treat haemopoietic stem cells either with chemotherapeutic agents, e.g. cyclophosphamide derivatives or tyrosine kinase inhibitors ( Carlo-Stella et al, 1991 ), interferons ( McGlave et al, 1990 ), antisense oligonucleotides targeted to different genes (e.g. myc, myb or the BCR-ABL gene) ( De Fabritiis et al, 1995 ; Ratajczak et al, 1992 ), or expand the normal progenitor cell population using long-term cultures ( Barnett et al, 1994 ). Although these techniques have permitted efficient in vitro suppression of leukaemic cell growth, when applied clinically they do not appear to induce a longer duration of Ph-negativity than that observed without purging ( Carlo-Stella et al, 1991 ; McGlave et al, 1990 ; De Fabritiis et al, 1995 ; Ratajczak et al, 1992 ; Barnett et al, 1994 ). In vivo techniques, pioneered as early as 1978 ( Körbling et al, 1981 ), have developed significantly since growth factors (mainly granulocyte-colony stimulating factor, G-CSF) have become available. The Genova group has shown that PBPC mobilized with both intensive chemotherapy (similar to induction chemotherapy used for acute myeloid leukaemia) and G-CSF contain a high proportion of Ph-negative progenitor cells ( Carella et al, 1993 ); this proportion seems to be particularly high (reaching 100%) when mobilization is performed in patients during the early chronic phase.
As recombinant α-IFN can induce complete responses in patients with CML, it could therefore be used for in vivo purging. In a recent study ( Italian Cooperative Study Group, 1993), patients in whom α-IFN induced a cytogenetic response were proposed for bone marrow harvesting; in most cases a high proportion of Ph-negative cells was collected, but the number of progenitor cells was very low (because of the toxicity of α-IFN), indicating that this technique would be difficult to apply to autologous transplantation. Based upon these results, and the capacity of G-CSF to mobilize PBPC in healthy individuals and in patients with solid tumours or haematological malignancies, we conducted a phase II study to examine whether rhG-CSF (lenograstim 150 μg/m2/d for 6–12 d) could mobilize a sufficient amount of normal PBPC in patients with CML who respond following treatment with α-IFN (defined as < 35% Ph-positive metaphases in the marrow).
- Top of page
- MATERIALS AND METHODS
This study demonstrated that low doses of lenograstim (150 μg/m2/d, equivalent to 5 μg/kg/d), can mobilize haemopoietic progenitor cells into the blood of CML patients who have responded to IFN. Previous studies have indicated that G-CSF (lenograstim or filgrastim) can mobilize stem cells either in healthy volunteer donors ( Korbling et al, 1995 ) or in patients with haematological malignancies or solid tumours who are candidates for subsequent autologous transplantation ( Sheridan et al, 1992 ). In healthy donors the usual dose of G-CSF administered has ranged from 3 to 10 μg/kg/d for 4–8 d ( Korbling et al, 1995 ; Grigg et al, 1995 ). Peak CD34+ cells or CFU-GM usually occurs following 4–5 d administration of G-CSF, and in most cases, only one LK is needed to collect the target number of 3 × 106 CD34+ cells/kg or 5 × 104 CFU-GM/kg ( Korbling et al, 1995 ; Grigg et al, 1995 ). There is a trend for better collections with higher doses of G-CSF (≥10 μg/kg) ( Grigg et al, 1995 ). Not surprisingly, the results of blood cell mobilization and collection in healthy donors are more favourable than those observed in our study performed in α-IFN-treated patients with CML.
In patients with solid tumours or haematological malignancies who receive G-CSF alone as a mobilizing treatment before autologous transplantation, CD34+ cell and CFU-GM mobilization usually occurs following approximately 5 d treatment with G-CSF (usual dose 5 μg/kg/d), but seems to be less pronounced than in healthy individuals ( Sheridan et al, 1992 ; Bensinger et al, 1993 ). In some such patients, particularly those who have previously received intensive or prolonged chemotherapy, the standard growth factor dose of 3–5 μg/kg is insufficient to mobilize blood stem cells ( Tricot et al, 1995 ). For these patients it may be useful to increase the dose of growth factor ( Bensinger et al, 1993 ) or to combine growth factor with other cytokines.
Recombinant α-IFN has been shown have a mild inhibitory effect on haemopoiesis both in in vitro studies ( Neuman & Fauser, 1982) and in patients treated with α-IFN. The Italian Cooperative Study Group (1993) reported that, in patients with CML who had previously received α-IFN for several months, marrow stem cell collection was very difficult, even when α-IFN was discontinued several weeks before collection. Thus, when the present study commenced, it was not known whether G-CSF could mobilize progenitor cells into the blood of α-IFN-pretreated patients. Consequently, in order to facilitate blood stem cell mobilization, it was decided to discontinue α-IFN before administering lenograstim and to commence LK following 8–12 d of treatment with lenograstim. After the first patients had been treated, it was observed that WBC, CFU-GM and CD34+ cell mobilization was adequate to collect a sufficient number of progenitor cells. This favourable effect of lenograstim was subsequently confirmed in those patients who continued α-IFN treatment during lenograstim administration; in total, 22/23 patients enrolled in the study achieved successful mobilization. These results demonstrate that rhG-CSF is capable of reversing, at least partially, the haematological toxicity of long-term treatment with α-IFN. Whether this effect is dose-dependent remains unknown and could be addressed in a subsequent study, specifically in patients with CML who are at higher risk of poor mobilization, e.g. those patients who receive the highest cumulative doses of α-IFN.
The second objective of this study was to assess the capacity of rhG-CSF to preferentially mobilize Ph-negative blood cells in CML patients with a good response to α-IFN. Although some cell lines transfected with the BCR-ABL gene have become growth-factor independent, few studies have shown a differential effect of growth factors on normal versus leukaemic (CML) progenitors. Such an effect has been suggested with interleukin-1, interleukin-4 or some other cytokines, e.g. TNF-α, leukaemia-inhibiting-factor or macrophage-inflammatory-protein-1-α, but, to date, this has not led to significant clinical applications ( Estrov et al, 1991 , 1993; Piacibello et al, 1990 ; Chasty et al, 1995 ). We have also reported that, using a liquid culture system and a combination of three cytokines (interleukin-1, interleukin-3 plus stem cell factor), Ph-negative BCR-ABL-negative progenitors could be grown more easily than their Ph-positive counterparts, suggesting that these three cytokines preferentially stimulate normal progenitors ( Jazwiec et al, 1995 ). Few studies have compared the effect of G-CSF on normal and CML progenitors, but it is generally acknowledged that G-CSF can promote CFU-GM growth from both normal and CML mononuclear cells; G-CSF is also produced by maturing neoplastic cells from patients with CML, and could subsequently contribute to the expansion of leukaemic mature cells ( Klein et al, 1990 ). Bedi et al (1994 ) have shown that growth factors (and G-CSF) can favour terminal differentiation of CML CD34+ cells more importantly than that of normal counterparts. Giralt et al (1993 ) have reported that in some patients with CML who relapse following AlloBMT, G-CSF re-induced complete remission and it was hypothesized that this effect resulted from preferential stimulation of Ph-negative cells. In the present study some results suggest that — at least in some patients — lenograstim preferentially mobilized Ph-negative BCR-ABL cells. Similar results were reported by Carreras et al (1997 ) who obtained Ph-negative LK in one patient who had minimal cytogenetic response during treatment with α-IFN. These experimental and clinical observations suggest that, at least in some patients with CML, normal (i.e. Ph-negative, BCR-ABL-negative) progenitors are more sensitive to stimulation by rhG-CSF than leukaemic progenitors. However, this needs to be confirmed in subsequent studies.
The Genoa group, and others, have reported that in CML patients either at diagnosis or who did not respond to α-IFN, it was possible to collect Ph-negative progenitor cells after mobilization with both chemotherapy and G-CSF ( Carella et al, 1993 , 1998). In our study performed in CML patients with a good response to α-IFN, the proportion of patients who could achieve Ph-negative LK, after mobilization with G-CSF alone, was higher than that observed by Carella et al (1998 ) in patients treated with chemotherapy and G-CSF. Moreover, the procedure was very safe and could be performed on an out-patient basis; however, patients mobilized by chemotherapy and G-CSF have severe myelosuppression lasting 2–3 weeks and which requires hospitalization. Therefore, we think that for CML patients who have responded to α-IFN, a low-dose of G-CSF, without chemotherapy, is sufficient to mobilize normal progenitor cells. However, this observation needs to be confirmed in a larger group of patients. In addition, other points such as the optimal dose of G-CSF and, more importantly, the utility of these Ph-negative progenitors to reconstitute long-term Ph-negative haemopoiesis when transplanted to patients with CML, also remain to be investigated. This point could not be answered in the present study as the 23 patients included had a good response to α-IFN and thus no indications for ASCT. As no significant adverse effects were experienced by the patients in the present study, these remaining questions can be addressed safely in future studies.