Superior Autologous Blood Stem Cell Mobilization from Dose-Intensive Cyclophosphamide, Etoposide, Cisplatin Plus G-CSF Than from Less Intensive Chemotherapy Regimens

Home , Melphalan

Bone Marrow Transplantation, (1999) 23, 111–117  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Superior autologous blood stem cell mobilization from dose-intensive cyclophosphamide, etoposide, cisplatin plus G-CSF than from less intensive chemotherapy regimens

DA Stewart1, D Guo2, D Morris1, M-C Poon3, BA Ruether3, AR Jones3, J Klassen4, I Auer5, J Luider5, A Chaudhry1,6, C Brown6 and JA Russell6

Departments of 1Medical Oncology, 2Epidemiology, 3Hematology, 4Apheresis, 5Flow Cytometry and 6Alberta Bone Marrow Transplant Program, Tom Baker Cancer Centre, Foothills Hospital, and University of Calgary, Calgary, Alberta, Canada

Summary: increasing for a variety of cancers, not all patients mobilize sufficient autologous blood stem cells (ABSC) to receive The study purpose was to determine if G-CSF plus dose- this therapy safely.1 An additional concern, albeit still of intensive cyclophosphamide 5.25 g/m2, etoposide controversial significance, is tumor contamination of the 1.05 g/m2 and cisplatin 105 mg/m2 (DICEP) results in autograft. Current in vitro purging techniques unfortunately superior autologous blood stem cell mobilization result in significant ABSC loss from the autograft, and only (BSCM) than less intensive chemotherapy. From Janu- treat the graft, not the patient.1,2 Tumor cytoreduction in ary 1993 until May 1997, 152 consecutive patients with the patient is as important as in the autograft since survival non-Hodgkin’s lymphoma (n = 55), breast cancer (n = following HDCT and ASCT correlates with tumor bulk.3 47), Hodgkin’s disease (n = 14), multiple myeloma (n = We began using dose-intensive cyclophosphamide, eto- 9), AML (n = 9), or other cancers (n = 18) initially poside and cisplatin (DICEP) chemotherapy in 1995 to treat underwent BSCM by one of three methods: Group 1: G- multiple myeloma and poor prognosis (primary refractory, CSF alone ؋ 4 days (n = 30). Group 2: disease-oriented bulky and/or marrow-positive relapsed) lymphoma patients chemotherapy, dosed to avoid blood transfusions, fol- prior to ABSC collection. Initially, we used DICEP for anti- lowed by G-CSF starting day 7 or 8, and apheresis day tumor therapy, but quickly found it mobilized stem cells as 13 or 14 (n = 82). Group 3: DICEP days 1–3, G-CSF well. DICEP is a well studied, intensive, but non-myeloabl- starting day 14, and apheresis planned day 19, 20 or ative regimen with acceptable treatment-related morbidity 21 (n = 40). A multivariate analysis was performed to and mortality rates.4–6 It is very closely related to a common determine which factors independently predicted ‘high-dose’ regimen used with ASCT for Hodgkin’s and BSCM. The median peripheral blood CD34+ (PB non-Hodgkin’s lymphoma called CBV (cyclophosphamide CD34+) cell count the morning of apheresis linearly cor- 6 g/m2, BCNU 300 mg/m2 and etoposide 750 mg/m2).3,7 related with the number of CD34+ cells removed per DICEP plus G-CSF has previously been reported to mobil- litre of apheresis that day. The median PB CD34+ cell ize blood stem cells.8–10 Demirer and colleagues found C count and median CD34+ cells ؋ 106 removed per litre 4 g/m2, E 600 mg/m2 and P 105 mg/m2 to be superior to of apheresis were highest for Group 3, intermediate for cyclophosphamide 4 g/m2 + G-CSF (CD34+ yield 3.32 vs Group 2, and lowest for Group 1. By multivariate analy- 1.88 ϫ 106/kg) in a retrospective analysis.9 Jennis and col- sis, mobilization group (3 Ͼ 2 Ͼ 1), disease other than leagues reported on 80 patients who underwent C 5 g/m2, AML, no prior melphalan or mitomycin-C, and less E 1500 mg/m2, and P 120 mg/m2 plus G-CSF.10 In a than two prior chemotherapy regimens predicted better median of two apheresis procedures this regimen yielded BSCM. Out of 15 Group 3 patients who had infiltrated excellent quantities of CD34+ cells (median of marrows, 11 had no detectable cancer in marrow and 15.8 ϫ 106/kg) with efficient hematopoietic engraftment. apheresis products after DICEP. These data suggest The doses we chose were higher than those studied by that DICEP results in superior BSCM than less inten- Schwartzberg et al8 and Demirer et al9 in an attempt to sive chemotherapy regimens. improve anti-cancer effect, but slightly lower than those Keywords: hematopoietic stem cell transplantation; hem- reported by other investigators4–6,10 in order to avoid throm- atopoietic stem cells; DICEP bocytopenia at the time of apheresis. We observed that DICEP and G-CSF seemed to mobilize ABSC more effec- tively than our previous mobilization regimens. The pur- poses of this retrospective study were: (1) to determine if Although the use of high-dose chemotherapy (HDCT) and DICEP and G-CSF could mobilize larger numbers of autologous blood stem cell transplantation (ASCT) is ABSC than less intensive chemotherapy and G-CSF; (2) to determine what other factors were associated with the Correspondence: Dr D Stewart, Tom Baker Cancer Centre, 1331-29 Street ability to mobilize ABSC; and (3) to determine how fre- NW, Calgary, Alberta, Canada T2N 4N2 quently DICEP could result in the collection of a tumor- Received 28 February 1998; accepted 24 August 1998 free autograft. Stem cell mobilization with DICEP DA Stewart et al 112 Patients and methods 350 mg/m2, and cisplatin 35 mg/m2 days 1–3, G-CSF 300 ␮g(Ͻ70 kg) or 480 ␮g(Ͼ70 kg) s.c. starting day 14, Patients and had apheresis scheduled for day 19, 20 or 21. The median dose of G-CSF was 6.0 ␮g/kg (range 5.2–9.6) for All 152 patients studied signed informed consent prior to Group 1 patients, 5.9 ␮g/kg for Group 2 patients (range autologous blood stem cell mobilization (BSCM) and 4.3–6.9), and 5.2 ␮g/kg for Group 3 patients (range 4.3– ASCT. Sufficient data existed concerning DICEP to justify 6.7). its use outside of another phase II clinical trial. This issue + + Peripheral blood CD34 (PB CD34 ) cell counts were was discussed with our Ethics Committee. Data were pro- monitored each morning starting 1 day prior to the planned spectively collected on these DICEP patients as a quality apheresis day. For Group 2 and 3 patients, apheresis took assurance project. place when the PB CD34+ count was Ͼ10 ϫ 106/l, and the Autologous BSCM and transplantation was first perfor- Ͼ ϫ 9 med at our center in 1993. Data on the initial 152 consecu- WBC was 5 10 /l. All patients underwent large volume tive patients who underwent BSCM between January 1993 apheresis via a double lumen central venous apheresis cath- eter inserted on a short-term basis, specifically for stem cell and May 1997 were analyzed. Patient characteristics are + summarized in Table 1. These patients were divided into collection. The apheresis was monitored by CD34 counts every 4 litres until target CD34+ yields were obtained. Aph- one of three groups depending on whether their initial + method of BSCM was G-CSF alone (Group 1, n = 30), eresis was discontinued for: (1) target CD34 yield, (2) platelet counts dropping below 40 ϫ 109/l, (3) complications moderately intensive, disease-specific chemotherapy plus + G-CSF (Group 2, n = 82), or DICEP plus G-CSF (Group such as bleeding or citrate toxicity, or (4) decreasing CD34 3, n = 40). The three groups differed in gender, diagnosis, cell collection per litre apheresis during the procedure to and frequency of tumor infiltration of bone marrow (all P levels deemed unsatisfactory to continue. Apheresis was Ͻ 0.001). These differences reflect changes in our practice performed with a continuous flow cell separator (Cobe- pattern over time. For example, G-CSF alone was used for Spectra, Cobe Canada, Scarborough, Ontario) processing BSCM during 1993 and 1994, but chemotherapy plus G- 5–30 l of blood/day at flow rates of 50–80 ml/min without CSF has more commonly been used since 1995. Also in significant differences in technique between the three 1995 we started performing ASCT for breast cancer. groups. The procedure took place approximately 24 h after the last dose of G-CSF. Autologous blood stem cell mobilization and collection Group 1 patients received G-CSF 300 ␮g(Ͻ60 kg) or 480 ␮g(Ͼ60 kg) s.c. daily for 4 days. The chemotherapy for Blood stem cell cryopreservation Group 2 patients was dosed to produce short-duration neutropenia and thrombocytopenia which would not require Blood stem cells were volume adjusted to a final concen- platelet transfusion (see Table 2). Before combining Group tration of Ͻ3 ϫ 108 mononuclear cells/ml. An equal vol- 2 patients together, we analyzed them by chemotherapy ume of tissue culture medium TC 199, 5% albumin and regimen and found no evidence that any one regimen was 20% dimethylsulfoxide (DMSO) was mixed with the apher- significantly better or worse at mobilizing blood stem cells. esis product to yield a final concentration of 10% DMSO. Group 2 patients received daily G-CSF 300 ␮g(Ͻ70 kg) The final suspension was transferred into freezing bags and or 480 ␮g(Ͼ70 kg) s.c. starting day 7 or 8, and underwent frozen to Ϫ90°C using a Kryo 10 Model 10–16 programm- apheresis at day 13 or 14. Group 3 patients (n = 40) able controlled-rate freezer, and stored in liquid nitrogen at received daily cyclophosphamide 1.75 g/m2, etoposide Ϫ196°C following standard methods.

Table 1 Patient characteristics by mobilization regimen

Group 1 Group 2 Group 3 P value (n = 30) (n = 82) (n = 40)

Median age (range) in years 35 (14–61) 42 (18–65) 44 (22–61) 0.122 Male:Female (% male) 22:8 (73%) 22:60 (27%) 29:11 (73%) Ͻ0.001 Median weight in kg (range) 76.5 (43–93) 71 (44–109) 73 (47–110) 0.499 Diagnosis: Ͻ0.001 NHL 12 17 26 HD 3 5 6 MM 1 2 6 Stage II/III breast cancer 0 29 0 Stage IV breast cancer 0 18 0 AML 4 5 0 Other 10 6 2 Disease positive marrow 1 (3%) 5 (6%) 14 (35%) Ͻ0.001 Prior melphalan or mitomycin-C 1 (3%) 8 (10%) 2 (5%) 0.53 2+ prior chemotherapy regimens 17 (57%) 31 (38%) 12 (30%) 0.068 Prior radiotherapy 8 (27%) 24 (29%) 16 (40%) 0.40 Stem cell mobilization with DICEP DA Stewart et al 113 Table 2 Group 2 BSCM regimens

BSCM regimen Disease status Patient number

Cyclophosphamide 2 g/m2, BrCa 34 Doxorubicin 50 mg/m2 and 5-Fluorouracil 500 mg/m2 Cyclophosphamide 2 g/m2 BrCa and non-bulky NHL/HD who had prior doxorubicin, 28 AML, MM, OvCa Etoposide 450 mg/m2, Relapsed non-bulky, marrow-negative NHL 6 Ifosfamide 6 g/m2, Cisplatin 105 mg/m2 Cyclophosphamide 2 g/m2 High-risk NHL ﬁrst remission 5 Doxorubicin 50 mg/m2 and Vincristine 2 mg, Prednisone 100 mg/days Vinblastine 12 mg/m2, Relapsed or refractory GCT 4 Ifosfamide 6 g/m2, Cisplatin 105 mg/m2 High-dose cytarabine AML ﬁrst remission 2 Taxol 200 mg/m2 BrCa 1 Taxotere 100 mg/m2 BrCa 1 Doxorubicin 75 mg/m2 DSRCT 1

AML = acute myeloid leukemia; BrCa = breast cancer; DSRCT = desmoplastic small round cell tumor; GCT = germ cell tumor; HD = Hodgkin’s disease; MM = multiple myeloma; NHL = non-Hodgkin’s lymphoma; OvCa = ovary cancer.

Flow cytometric quantitation of CD34+ cells and assay these nine patients, tumor contamination of the apheresis for tumor cells product was evaluated prior to CD34+ selection. For daily monitoring of PB CD34+ cell counts, 100 ␮lof whole blood were incubated with CD45-FITC (KC56; Autologous stem cell transplantation Coulter, Hialeah, FL, USA) and CD34-PE (HPCA-2; Becton Dickinson, San Jose, CA, USA) for 10 min at room The high-dose chemotherapy regimens varied according to temperature. The samples were lysed using the Multi-Qprep disease. Seventy-five percent of patients received either system (Coulter) and 100 ␮l of Flow Count beads (Coulter) melphalan alone, melphalan plus TBI, or cyclophospham- were added to each tube. Samples were analyzed immedi- ide, mitoxantrone and vinblastine. Daily G-CSF 300 Ͻ Ͼ ␮ + ately. CD34+ cell counts were monitored during apheresis ( 70 kg) or 480 ( 70 kg) g/day s.c. from day 7 until Ͼ by incubating the antibodies listed above with 10 ␮l of well ANC 1.5 was used for consecutive patients starting in mixed product from the collection bag for 10 min at room 1995. Transfusion support following DICEP and ASCT Ͻ temperature. Red cells were lysed with 1 ml of Ortho-mune consisted of 2 units of irradiated RBCs for Hb 80 g/l, Lysing Reagent (Ortho Diagnostic Systems, Raritan, NJ, and 6 units irradiated random donor platelets for platelet Ͻ ϫ 9 USA). Flow cytometric analysis was performed on an counts 20 10 /l. EPICS XL-MCL (Coulter) using a modified protocol.11,12 The status of the bone marrow was determined by com- Statistics bining cytomorphologic and histologic features of the marrow aspirates and trephine biopsies, respectively. In Comparisons between the three mobilization groups: The addition, samples of the marrow aspirate as well as apher- Kruskal–Wallis rank sum test was used to test differences esis product were analyzed to look for minimal residual in patient age, weight, PB CD34+ count on the morning of disease using multiparameter flow cytometry in cases of the first apheresis, the total volume of apheresis per patient, NHL and myeloma. The detection of any number of cells CD34+ cell yield/l apheresis, and the total number of CD34+ showing immunophenotypic evidence of light chain restric- cells collected per kilogram. The ␹2 test was used to test tion or expressing an abnormal immunophenotype was con- differences in patient gender and the number of patients sidered to represent involvement by cancer. Samples dem- who had Ͼ2 ϫ 106 CD34+ cells/ kg collected in a single onstrating immunophenotypic evidence of clonal disease apheresis. Fisher’s exact test was used to test differences were further investigated and compared to the immuno- in patient characteristics (diagnosis, tumor positive marrow, phenotype of the respective diagnostic tissue to determine prior melphalan or mitomycin-C, two plus prior chemo- if they represented the same proliferative process. Nine therapy regimens), and the number of patients who had autografts were processed by positive CD34 selection inadequate BSCM for apheresis and ASCT. (Ceprate System, CellPRO, Bothell, WA, USA). These nine patients were either in BSCM Group 3 (n = 6) or Group 2 Analysis of BSCM (PB CD34+ counts on first apheresis (n = 3), and five had detectable cancer in their marrows. For morning): To stabilize variances and result in a nearly nor- Stem cell mobilization with DICEP DA Stewart et al 114 mal distribution, logarithmic transformation was made for PB CD34+ counts on the first apheresis morning. Univariate analysis was conducted first, using analysis of variance in conjunction with simple regression, for the following 100 potential predictors of BSCM: (1) BSCM group; (2) 0–1 80 vs 2+ prior chemotherapy regimens; (3) prior melphalan or

mitomycin-C; (4) marrow inﬁltration with tumor; (5) gen- 60 der; (6) age; (7) diagnosis; (8) prior radiotherapy. Multiple regression was then used to study the effect of a predictor 40 with other variables being simultaneously controlled. CD34 cells/kg collected 20 Analysis of time to engraftment by CD34+ cell dose: A multivariate analysis was performed to determine which of 0 + the following factors predicted engraftment: CD34 cell 0 500 1000 1500 2000 dose/kg, mobilization group, high-dose conditioning regimen and disease. PB CD34 count Figure 2 Relationship between PB CD34+ cell count and cD34+ cells collected per kg body weight on the ﬁrst apheresis day (r = 0.78, r2 = Results 0.6175).

Relationship between PB CD34+ count and CD34+ yields per litre apheresis cedures per patient was 1.58 for Group 1 (range 1–3), 1.17 for Group 2 (1–4) and 1.05 for Group 3 (1–2), with a Direct linear relationships were found between the morning median sum total volume of apheresis of 25.1, 15.4, and PB CD34+ cell count and the CD34+ cell yield per litre on 11.3 litres for Groups 1, 2 and 3, respectively. the first day of apheresis (Figure 1), and between the PB + + By univariate analysis, factors predictive of superior CD34 count and CD34 cells collected/kg that day + BSCM (defined as high PB CD34 count on the first apher- (Figure 2). esis morning) for all 152 patients were: (1) diagnosis other than AML (P = 0.0002), (2) mobilization group (Group 3 CD34+ cell mobilization better than 2 better than 1, P Ͻ 0.0001), (3) no prior treatment with melphalan or mitomycin-C (P Ͻ 0.0001), and Table 3 summarizes the results of BSCM and apheresis (4) less than two prior chemotherapy regimens (P = 0.031). according to the method of mobilization. Patients who had These four factors continued to be significantly associated insufficient BSCM for apheresis and HDCT/ASCT were with BSCM by multivariate analysis. Age, gender, prior excluded from analyzing the number of apheresis pro- radiotherapy, and marrow infiltration with tumor were not cedures, volume of apheresis, CD34+ cells collected/l aph- found to be related to BSCM by either univariate or multi- eresis, and total CD34+ cells collected/kg. PB CD34+ Ͼ ϫ 6 + variate analysis. Even after controlling for diagnosis, prior counts, frequency of collecting 2 10 CD34 cells/kg melphalan or mitomycin-C, and number of prior chemo- in a single apheresis, CD34+ cell yields/l of apheresis, and + therapy regimens in multivariate analysis, mobilization total CD34 cells collected/kg were all highest for Group Group 3 yielded a PB CD34+ count of 6.4-fold greater 3 patients, intermediate for Group 2 patients, and lowest (95% CI = 3.5–11.7) than Group 1, and 3.1-fold greater for Group 1 patients. The mean number of apheresis pro- (95% CI = 1.8–5.4) than Group 2. The results of this multivariate analysis could be criticized because of patient heterogeneity among the three mobilization groups. Therefore, the data were re-analyzed only = 600 including patients with multiple myeloma (n 9), Hodg- kin’s disease (n = 14), and non-Hodgkin’s lymphoma (n = 55) since these diagnoses were represented in each of the three mobilization groups. Following multivariate analysis 400 of these 78 patients, prior melphalan (P = 0.007), BSCM Group 3 vs 2(P = 0.002) and Group 2 vs 1(P Ͻ 0.001) continued to independently predict BSCM, while two or cells/l apheresis

+ 200 more prior chemotherapy regimens was no longer significant (P = 0.15). CD34 Additional support for superior BSCM with DICEP com- 0 pared to less intensive chemotherapy + G-CSF is the fact that three of five patients who failed to mobilize sufficient 0 500 1000 1500 2000 CD34+ cells for apheresis and ASCT with a Group 2 regi- PB CD34 count men (PB CD34+ counts 10 ϫ 106/l), had better BSCM with Figure 1 Relationship between PB CD34+ cell count and CD34+ cell DICEP + G-CSF, and were then able to undergo HDCT + yield per litre on the first apheresis day (r = 0.88, r2 = 0.7793). and ASCT. The PB CD34 counts ( ϫ 106/l) on the first Stem cell mobilization with DICEP DA Stewart et al 115 Table 3 CD34+ mobilization by regimen

Measure of BSCM Group 1 Group 2 Group 3 P value (n = 30) (n = 82) (n = 40)

Median PB CD34+ count ×106/l on the morning of the 21.5 65 278 Ͻ0.001 ﬁrst apheresis (range) (3–206) (2–371) (0–1991) Number of patients in whom a single apheresis collected 13 (43%) 68 (83%) 38 (95%) Ͻ0.001 Ͼ2 × 106 CD34 cells/kg Number patients who had inadequate BSCM for 3 (10%) 10 (12%) 2 (5%) 0.573 apheresis and ASCT Median CD34+ cells (×106) per litre apheresed on ﬁrst 11.3 36.8 132.4 Ͻ0.001 apheresis day (range) (1.4–52.4) (1.4–167.9) (7.6–563.3) Median sum total volume of apheresis per patient (litres) 25.1 15.4 11.3 Ͻ0.001 (8.5–64.0) (5.3–68.4) (5.4–32.3) Median total CD34+ cells collected × 106/kg (range) 3.3 6.9 18.3 Ͻ0.001 (0.6–7.1) (2.0–28.0) (2.6–127.0)

apheresis morning increased from 5 to 11, 7 to 19, and 6 febrile neutropenia. After the administration of DICEP, to 43 respectively, for these three patients following BSCM patients who had a spouse or other caregiver at home were with Group 2 and then Group 3 regimens. This occurred frequently followed on an ambulatory basis every 1–2 days, despite our general experience that a second attempt at and readmitted to an inpatient bed only if they developed BSCM using the same regimen within 1–2 months of the febrile neutropenia. With the first day of DICEP chemo- first attempt typically results in poorer BSCM on the therapy called day 1, the median day of neutrophil recovery second occasion. Ͼ0.5 ϫ 109/l was day 17 (range 15–23) and median day of platelet recovery Ͼ20 ϫ 109/l was also day 17 (range 14– 33). A median of 2 units RBC (0–16 units) and 2 days of Tumor cell contamination of autograft platelet transfusion (1–8 days) were required for DICEP. Eleven of the 15 (73%) Group 3 patients (NHL = 8, MM Patients underwent apheresis post-DICEP either on day 16 = 6, HD = 1) who had cancer cells detected in marrow (n = 1), 19 (n = 9), 20 (n = 17), 21 (n = 9), 22 (n = 1) or samples prior to DICEP, had negative marrows and apher- day 23 (n = 1). Twenty-nine patients (73%) experienced esis products post-DICEP. One additional patient who had febrile neutropenia (eight patients (20%) had positive blood high-risk NHL in first partial remission following four cultures). Mucositis and other extramedullary toxicity were courses of CHOP, was found to have an apheresis product generally mild. Only one patient experienced a life-threat- positive for lymphoma after BSCM with a Group 2 regimen ening complication requiring intensive care support. After (cyclophosphamide 2 g/m2, adriamycin 50 mg/m2 and apheresis on day 21, this patient developed ARDS requiring vincristine 2 mg day 1, and prednisone 100 mg days 1–5). mechanical ventilation. No infectious cause was found and This patient then underwent BSCM with DICEP + G-CSF the patient made a complete recovery. The patient had a which resulted in a marrow and apheresis product free of prior episode of ARDS following a laparotomy 1 year earl- lymphoma by morphology and flow cytometry. ier. No Group 3 patient experienced a treatment-related death; following DICEP or HDCT/ASCT. The median + length of hospital stay for DICEP was 14.5 days (range Influence of CD34 cell dose on hematopoietic recovery post-ASCT 4–51). By multivariate analysis, only CD34+ cell dose/kg was independently predictive of time to neutrophil and platelet Discussion engraftment; mobilization group was not. Group 3 patients who underwent ASCT (n = 37) had a median CD34+ cell One important result of this study is the finding that PB dose infused of 16.0 ϫ 106/kg (range 1.8–127.0), and CD34+ counts correlate very well with the CD34+ cell experienced median engraftment times of 11 days to ANC yields per litre of apheresis. Several factors can affect the Ͼ 0.5 ϫ 109/l (range 8–15), 11 days to platelets CD34+ yield per litre of apheresis in addition to the degree Ͼ20 ϫ 109/l (range 9–38), and 17 days to platelets of BSCM including timing of apheresis relative to BSCM, Ͼ100 ϫ 109/l (range 10–381+). operator- and machine-dependent factors, unstable inter- face, and total duration and volume of apheresis. These fac- Toxicity of DICEP + G-CSF tors, as well as variability in flow cytometric determination of CD34+ counts, may contribute to an imperfect corre- Patients in Group 3 experienced much more treatment- lation between PB CD34+ counts and apheresis CD34+ related toxicity than patients in Groups 1 or 2. All Group yields. The PB CD34+ count is probably, therefore, a more 3 patients required blood transfusions and hospitalization reliable indicator of BSCM than apheresis CD34+ yields. either for administration of DICEP or for treatment of CD34+ cell dose/kg body weight is even more unreliable Stem cell mobilization with DICEP DA Stewart et al 116 as a marker of BSCM due to the added variable of patient made in this study regarding chemotherapy intensity since body weight. For these reasons, PB CD34+ counts were the three groups received similar G-CSF dose and duration. used as the marker for BSCM in this study. This correlation If starting G-CSF earlier indeed results in superior BSCM, between PB CD34+ cell count and apheresis CD34+ cell then the significant difference found between Group 1 and yield has also been found by other investigators,13,14 and the other two groups may be conservative. suggests that PB CD34+ counts at defined times post- As expected, the time to hematopoietic engraftment post- chemotherapy or post-cytokine can be used as a measure ASCT depended mostly upon CD34+ cell dose/kg rather of BSCM. This would simplify design and conduct of than upon mobilization regimen. Group 3 patients had rapid future BSCM studies, and potentially allow variation in the engraftment post-ASCT, commensurate with their infused method of stem cell collection by individual centers partici- CD34+ cell dose. Even patients who received huge doses pating in these trials. This statement is not to discount the of CD34+ cells (up to 127 ϫ 106/kg) still required approxi- clinical importance of the number of CD34+ cells collected, mately 10 days to engraft neutrophils and platelets. This but is simply to identify the most direct endpoint when leads one to be skeptical that ex vivo CD34+ expansion up comparing different methods of BSCM. to this level will abolish pancytopenia unless the cells have We have not routinely measured CFU-GM, CFU- significantly different phenotypic and biologic properties GEMM, or BFU-E on stem cell products for the past few than unmanipulated blood stem cells. years because we found it did not add clinically useful We do not recommend BSCM with DICEP for all information to the CD34+ counts. Others have also reported patients. It is associated with substantial toxicity and cost, that CD34 yield alone is sufficient to determine adequacy and its use depends more on its effectiveness as anti-cancer of ABSC products.15,16 therapy than its effectiveness in BSCM. Nevertheless, for The results of this study also suggest that DICEP plus G- selected patients it may prove to be useful cancer therapy, CSF results in superior BSCM compared to less intensive and at the same time mobilize ABSC very well. We have chemotherapy plus G-CSF, or G-CSF alone. The dose– found promising early results with DICEP followed by response effect of chemotherapy on stem cell mobilization high-dose melphalan and ASCT for refractory and bulky has been reported previously.1,9,17,18 This knowledge is relapsed lymphoma patients.22 These results need to be con- potentially important only to those patients who may not firmed in a randomized trial. The three drugs in DICEP mobilize stem cells well with less intensive regimens, or to are active against a wide variety of tumor types including those who might benefit from the anti-tumor effects of Hodgkin’s and non-Hodgkin’s lymphoma, multiple mye- DICEP. Regarding the former group of patients, we ident- loma, ovarian cancer, breast cancer, testicular cancer and ified three factors which were associated with poor BSCM; small round cell sarcomas. Its efficacy could be studied pre- diagnosis of AML, two or more prior chemotherapy regi- ASCT in these cancers as well. Hopefully, new hematopo- mens, and prior treatment with melphalan or mitomycin-C. ietic growth factors can prevent chemotherapy-induced All three of these factors relate to marrow damage from myelosuppression as well as mobilize ABSC more effec- chemotherapy and possibly disease. There was a trend for tively than currently available agents.23,24 DICEP combined prior radiotherapy to be associated with poor BSCM. It is with these new growth factors may not only increase the possible that a larger sample size, and analysis by percent- safety and feasibility of this effective cancer therapy, but age of marrow radiated, would have resulted in a negative may result in such effective BSCM that the collection of association between radiotherapy and BSCM. Due to miss- large numbers of ABSC may be accomplished by simple ing data, we could not analyze other factors which may phlebotomy rather than apheresis. have affected BSCM such as the use of repetitive cycles of G-CSF with recent chemotherapy, interval from most recent chemotherapy to BSCM, cumulative duration of Acknowledgements uninterrupted chemotherapy prior to BSCM, time from diagnosis to BSCM, and blood counts and marrow cellu- We thank J McLaughlin for data management and Dr Peter larity pre-BSCM.19,20 Most of these additional factors also Forsyth for manuscript review. relate to treatment-related marrow damage. It can be con- cluded that numerous factors affect BSCM, and must be controlled in the design or analysis of BSCM trials. References The dose and duration of G-CSF used in these patients may be criticized. We did not start G-CSF the day after 1 To LB, Haylock DN, Simmons PJ, Juttner CA. The biology mobilization chemotherapy, mainly for reasons of cost. We and clinical uses of blood stem cells. Blood 1997; 89: also reasoned that if 4 days of G-CSF was sufficient for 2233–2258. BSCM in the absence of chemotherapy, then a similarly 2 Gulati SC, Romero CE, Ciavarella D. Is bone marrow purging short duration of G-CSF administration would probably be proving to be of value? Oncology 1994; 8: 19–24. sufficient when combined with chemotherapy. Also, the 3 Stahel RA, Jost LM, Pichert G, Widmer L. High-dose chemo- dose of G-CSF (approximately 4–8 ␮g/kg/day) may have therapy and autologous bone marrow transplantation for malignant lymphomas. Cancer Treatment Rev 1995; 21: 3–32. not been optimal since others have reported a dose-response 4 Dunphy FR, Spitzer G, Fornoff JE et al. Factors predicting ␮ 21 of G-CSF on BSCM up to 32 g/kg/day. We cannot com- long-term survival for metastatic breast cancer patients treated ment as to whether the groups may have had better BSCM with high-dose chemotherapy and bone marrow support. if G-CSF was started earlier after chemotherapy and/or at a Cancer 1994; 73: 2157–2167. higher dose. This, however, does not affect the comparisons 5 Huan SD, Yau JC, Dunphy FR et al. Impact of autologous Stem cell mobilization with DICEP DA Stewart et al 117 bone marrow infusion on hematopoietic recovery after high- 15 Hazelton B, Birch R, Schwartzberg L et al. CD34 yield alone dose cyclophosphamide, etoposide and cisplatin. J Clin Oncol is sufficient to determine adequacy of peripheral blood pro- 1991; 9: 1609–1617. genitor cell (PBPC) products. Proc Ann Meet Assoc Cancer 6 Livingston RB. High-dose cancer therapy without stem cell Res 1994; 35: a1223. support-human solid tumors. In: Armitage JO, Antman KH 16 Weaver CH, Hazelton B, Birch R et al. An analysis of (eds). High-Dose Cancer Therapy: Pharmacology, Hematopo- engraftment kinetics as a function of the CD34 content of per- ietins, Stem Cells. Williams and Wilkins: Baltimore, Mary- ipheral blood progenitor cell collections in 692 patients after land, 1992, pp 195–210. the administration of myeloablative chemotherapy. Blood 7 Bierman PJ, Bagin RG, Jagannath S et al. High dose chemo- 1995; 86: 3961–3969. therapy followed by autologous hematopoietic rescue in 17 Reiffers J, Bernard PH, Marit G et al. Collection of blood- Hodgkin’s disease: long term follow-up in 128 patients. Ann derived haemopoietic stem cells and applications for autolog- Oncol 1993; 4: 767–773. ous transplantation. Bone Marrow Transplant 1986; 1 (Suppl. 8 Schwartzberg L, West W, Tauer K et al. Cyclophosphamide, 1): 371–372. etoposide, cisplatin plus G-CSF (CEP+G) for peripheral blood 18 Schwartzberg L, Birch R, Hazelton B et al. Peripheral blood stem cell (PBSC) mobilization. Proc Ann Meet Am Soc Oncol stem cell mobilization by chemotherapy with and without 1992; 11: a1405. recombinant human granulocyte colony-stimulating factor. J 9 Demirer T, Buckner CD, Storer B et al. The effect of different Hematother 1992; 1: 317–327. chemotherapy regimens on peripheral blood stem cell mobiliz- 19 Ketterer N, Salles G, Tremisi P et al. Analysis of factors ation yield. Proc Ann Meet Am Soc Clin Oncol 1996; 15: influencing 300 peripheral blood progenitor cell (PBPC) col- a1007. lections in 273 patients treated for lymphoproliferative dis- 10 Jennis A, Pecora AL, Isaacs RE et al. Dose-intensive cyclo- eases. Blood 1997; 90 (Suppl. 1): 213a (Abstr. 939). 20 Bensinger W, Appelbaum F, Rowley S et al. Factors that phosphamide, etoposide, and platinum (DICEP) as a influence collection and engraftment of autologous peripheral priming/mobilization regimen for peripheral blood stem cell blood stem cells. J Clin Oncol 1995; 13: 2547–2555. (PBSC) harvesting prior to high dose chemotherapy (HDC). 21 Stiff P, Malhotra D, Bayer R et al. High dose G-CSF improves Blood 1996; 88 (Suppl. 1): 404a (Abstr. 1605). stem cell mobilization and collection compared to standard 11 Sutherland DR, Keating A, Nayar R et al. Sensitive detection + doses in patients with ovarian cancer which leads to a decrease and enumeration of CD34 cells in peripheral and cord blood in delayed platelet engraftment following stem cell transplants. by flow cytometry. Exp Hematol 1994; 22: 1003–1010. Blood 1997; 90 (Suppl. 1): 591a (Abstr. 2629). 12 Chin-Yee, Keeney M, Anderson L et al. Current status of + 22 Stewart DA, McLaughlin J, Chaudhry A et al. High dose CD34 cell analysis by flow cytometry: The ISHAGE Guide- sequential therapy (HDST) for very poor prognosis lym- lines. Clin Immunol News 1997; 17: 21–29. phoma. Ann Proc Am Soc Clin Oncol 1997; 16: 23a (Abstr. 13 Schots R, Van Riet I, Damiaens S et al. The absolute number 80). + of circulating CD34 cells predicts the number of hematopo- 23 DiPersio JF, Abbound CN, Winter JN et al. Phase I–II study ietic stem cells that can be collected by apheresis. Bone Mar- of mobilization of PBPC by SC-68420 in patients with breast row Transplant 1996; 17: 509–515. cancer or lymphoma. Blood 1997; 90 (Suppl. 1): 97a (Abstr. 14 Schwella N, Beyer J, Schwaner I et al. Impact of preleukaph- 426). eresis cell counts on collection results and correlation of pro- 24 Shpall EJ, Wheeler CA, Turner SA et al. A randomized phase genitor-cell dose with engraftment after high-dose chemo- III study of PBPC mobilization by stem cell factor (SCF, therapy in patients with germ cell cancer. J Clin Oncol 1996; STEMGEN) and Filgrastim in patients with high risk breast 14: 1114–1121. cancer. Blood 1997; 90 (Suppl. 1): 591a (Abstr. 2627).