New Consecutive High-Dose Chemotherapy Modality with Fractionated Blood Stem Cell Support in the Treatment of High-Risk Pediatric Solid Tumors: a Feasibility Study

Home , Ranimustine

New consecutive high-dose chemotherapy modality with fractionated blood stem cell support in the treatment of high-risk pediatric solid tumors: a feasibility study

T Kajiume1,2, Y Kawano1, Y Takaue1, T Abe1, T Watanabe1, Y Okamoto1, A Makimoto1, K Suenaga1, H Suzuya1, J Sato1, A Yokobayashi2, T Hashimoto3, K Yoshida3, H Ishibashi3, H Takehara3, S Tashiro3 and Y Kuroda1

Departments of 1Pediatrics and 3Surgery I, University of Tokushima, Kuramoto-cho, Tokushima; and 2Department of Pediatrics, Kawasaki Medical School, Kurashiki, Japan

Summary: Keywords: solid tumor; children; transplantation; CD34+ cells For the treatment of childhood solid tumors, we performed a pilot feasibility study of consecutive high-dose therapies, in which each course was followed by transplantation with granulocyte colony-stimulating factor- Despite the fact that childhood solid tumors are generally mobilized peripheral blood cells which had been separ- chemosensitive and that multimodality therapy has signifi- ated into CD34-positive and -negative fractions by an cantly improved the outcome, subgroups continue to do Isolex system (Baxter). Positive selection of CD34؉ cells poorly. Poor prognostic features include metastasis and a has been associated with inevitable cell loss. To over- bulky primary tumor. Long-term survival is extremely rare -come this loss, CD34؉ cells that had migrated into the for patients who have primary refractory or relapsed dis negative fraction were saved and used for the first trans- ease. These problems may be solved through the use of plant, which was followed by a second transplant after high-dose chemotherapy, especially in patients with chemo- a 3- to 5-month interval. In this phase I feasibility study, sensitive tumors, since the effectiveness of anticancer ther- the results in six children were evaluated for safety and apy depends on both the dose intensity of chemotherapy engraftment. Multi-drug cytoreductive regimens using and the timing of drug administration.1 Dose intensity can ranimustine (MCNU), melphalan, thiotepa, carboplatin, be improved either by increasing the dosage of the drug cyclophosphamide or VP-16 were comparable between given at the same intervals, by reducing the delay between the two transplant procedures in terms of their inten- each cycle, or by using stem cell transplantation support. -sity. The number of CD34؉ cells in the ‘CD34(؉) frac- At present, the efficacy of autologous bone marrow trans tion’ was 3.31 × 106/kg (0.63–4.3 × 106/kg), while this plantation (ABMT) is unresolved for most pediatric number in the ‘CD34(؊) fraction’ could not be evalu- tumors,2 although its superiority over regular chemotherapy ated correctly due their scarcity (Ͻ0.1%). The median has been suggested in some patient subgroups.3 Autologous numbers of infused MNC and CFU-GM were, respect- transplants are associated with the inherent disadvantage ively, 4.2 ؋ 106/kg and 1.75 ؋ 105/kg in the CD34(؉) of cancer cell contamination in the graft. To improve the fraction, and 4.8 ؋ 108/kg and 3.35 ؋ 105/kg in the therapeutic efficacy by removing contaminating cancer CD34(؊) fraction. The number of days required to achi- cells from autografts, positive selection of blood CD34+ eve an ANC Ͼ0.5 ؋ 109/l and a platelet count cells has been widely used. However, major drawbacks of Ͼ20 ؋ 109/l and Ͼ50 ؋ 109/l were, respectively, 14.5, this strategy include a variable period of cytopenia due to .and 19.5 in the first transplant with CD34؊ cells, inevitable cell loss and a high cost 15.0 and 13.5, 18.0 and 25.0 in the second transplant with To improve the outcome in children with solid tumors CD34؉ cells, with no essential difference between the associated with poor prognostic factors with intensifying two treatments. Although the small number of patients, chemotherapy, we initiated a strategy of administering mul- the variation in clinical status and treatment, and the tiple ‘mini-transplants’ given in rapid succession, each with short follow-up invalidate any evaluation of the thera- the support of peripheral blood stem cells (PBSC). peutic benefit of this strategy, the encouraging results In particular, we evaluated a new consecutive-autograft support the feasibility of this strategy, which enables an concept, in which cells that had been recovered in the escalation of dose intensity with an improved CD34-negative fraction are used in the initial transplant. cost/benefit ratio. The basic idea for this study was that the intensity of treatment can be increased by the use of cells in the CD34- negative fraction, which still contains a large number of + Correspondence: Dr Y Takaue, Department of Pediatrics, University CD34 cells. To overcome the possible infusion of malig- Hospital of Tokushima, Kuramoto-cho 3, Tokushima 770, Japan nant cells, patients then received a second and final trans- + Received 19 May 1997; accepted 25 August 1997 plant with CD34 cells that had been indirectly depleted of New consecutive high-dose therapy modality T Kajiume et al 148 malignant cells. New methods in cancer therapy should be Collection and isolation of blood cells critically evaluated before they are incorporated into routine practice. In this respect, the end-point of evaluation in this In two patients, aphereses were performed in the recovery pilot feasibility study in six children with high-risk solid phase of consolidation chemotherapy with 50–200 ␮g/m2 tumors was safety and the recovery of hematopoiesis. We of recombinant human granulocyte colony-stimulating fac- believe that the results of this small study may contribute tor (G-CSF; filgrastim, Kirin Brewery Co, Tokyo, Japan). to the development of strategies for clinical application of The remaining patients were mobilized by G-CSF alone (10 enriched CD34+ cells. ␮g/kg/day for 5 days). Details of the procedure for col- lecting blood cells have been described elsewhere.4–7 Briefly, a Fenwal CS-3000 Plus cell separator (Baxter Limited, Deerfield, IL, USA) was operated using a small- Materials and methods volume collection chamber (SVCC) and a granulocyte chamber to process 100–350 ml of blood per kg body Patients weight. Cells were recovered in 50-ml suspension. Cells were then isolated with the Isolex 50 system Six patients (one of each with ovarian tumor, testicular (Baxter Limited) according to the manufacturer’s instruc- tumor, retinoblastoma, rhabdomyosarcoma, brain tumor tions. Briefly, cells were incubated with 0.5% human ␥- and Wilms’ tumor) were enrolled into this study with the globulin (Polyglobin N; Bayer Japan, Tokyo, Japan) con- written consent of the patients or their parents. This study taining PBS for 15 min to block intact Fc-receptors on the was approved by the Institutional Review Board. There cell surface. After incubation, cells were washed with PBS were four males and two females, and the patients ranged containing 1% HSA and incubated with anti-CD34 mono- in age from 1 to 16 years (median, 9.5 years). All of the clonal antibody (9C5; Baxter) at a cell/antibody ratio of patients received surgery, and their characteristics and a 106/0.5 g. After 30 min of incubation at 4°C with gentle detailed history of previous therapy are summarized in rotation (4/min), cells were washed three times with PBS Table 1. Associated poor-prognosis factors included mul- and resuspended in IMDM with 1% HSA at 1–2 × 107/ml. tiple metastases (n = 3), recurrent tumor (n = 3) and wide The anti-CD34 antibody-coated cells were incubated with local invasion (n = 1). All of these patients were considered sheep anti-mouse IgG1-coated microspheres (Dynabeads; to be eligible for this experimental consecutive high-dose Dynal Corporation, Oslo, Norway) for 30 min with rotation, therapy protocol. and cells which were bound to the beads were then col-

Table 1 Patient characteristics

Case Age/Sex Diagnosis Risk factors/Previous therapy Conditioning regimen Interval (months) No. First transplant Second transplant Dx (recurrence) 1st/2nd with CD34(Ϫ) fr with CD34(ϩ) fr to PBSCT transplant

1 10/F Ovarian tumor Recurrent abdominal CBDCA (1.6 g/m2) CBDCA (1.2 g/m2)9 3 (immature teratoma) dissemination, multiple VP-16 (0.6 g/m2) VP-16 (0.6 g/m2) pulmonary metastases/CBDCA, CY (50 mg/kg) L-PAM (70 mg/m2) VP-16, pirarubicin 2 16/M Testicular tumor Stage IV, inoperable wide CBDCA (1.6 g/m2) CBDCA (1.2 g/m2)6 4 (yolk sac tumor) invasion refractory to therapy VP-16 (1.2 g/m2) VP-16 (0.6 g/m2) with VCR, CBDCA and CY (50 mg/kg) L-PAM (140 mg/m2) pirarubicin 3 11/F Bilateral Recurrent metastatic tumor 6 CBDCA (1.6 g/m2) CBDCA (1.2 g/m2)3 3 retinoblastoma years after surgery and VP-16 (1.2 g/m2) VP-16 (0.6 g/m2) radiation therapy CY (50 mg/kg) L-PAM (210 mg/m2) 4 1/M Rhabdomyosarcoma Multiple pulmonary CBDCA (1.2 g/m2) CBDCA (1.2 g/m2)5 5 (shoulder, embryonal metastases/actinomycin-D, VP-16 (0.6 g/m2) VP-16 (0.6 g/m2) type) cisplatinum, VP-16, CY and L-PAM (120 mg/m2) Thiopeta (250 pirarubicin THP-ADR (50 mg/m2) mg/m2) 5 5/M Brain tumor (yolk Recurrent tumor after CBDCA (1.6 g/m2) CBDCA (1.2 g/m2)3 4 sac tumor) surgery/CBDCA and VP-16 VP-16 (1.2 g/m2) VP-16 (1.2 g/m2) IFO (6 g/m2) MCNU (220 mg/m2) 6 9/M Wilms’ tumor Multiple pulmonary and brain L-PAM (210 g/m2) CBDCA (1.2 g/m2)8 5 (nephroblastoma, metastases/actinomycin-D, VP-16 (0.6 g/m2) favorable) VCR, doxorubicin, cisplatinum L-PAM (210 mg/m2)

M = male; F = female; Dx = diagnosis; CBDCA = carboplatinum; VP-16 = etoposide; CY = cyclosphosphamide; L-PAM = melphalan; ADR = doxorubicin; IFO = ifosphamide; MCNU = ranimustine. New consecutive high-dose therapy modality T Kajiume et al 149 lected with a magnet using an Isolex system. Finally, Dyna- were used in the initial transplant. This was followed by bead-bound cells were incubated with 200 U/ml of chymo- two to three courses of regular-dose chemotherapy for papain (Chymocell-R; Baxter) for 15 min at 37°C. After further in vivo purging without rescue, before the final incubation, CD34+ cells were isolated from the beads using transplant with purified CD34+ cells. The detailed trans- a magnet. The purity of the CD34+ cell fraction was 79%, plantation procedure has been reported elsewhere.6–8 Dose with a recovery rate of 21%. Cells were mixed slowly with intensities of cytoreductive regimens incorporating ranimu- an equal volume of a freezing solution containing 8% stine (MCNU), melphalan, thiotepa, carboplatin, cyclopho- human albumin, 12% HES, and 10% DMSO to give final sphamide or VP-16 were considered to be comparable in concentrations of 6% HES and 5% DMSO.8 Both CD34- each course of high-dose therapy, as summarized in Table positive and -negative cells were transferred to 5-ml poly- 1. No patient received total body irradiation. Thirty-six propylene tubes and then placed directly in an electric hours after completion of the cytoreductive regimen, the freezer that maintained a temperature of Ϫ135°C (Sanyo cells were rapidly thawed at 37°C and promptly infused Electric Co, Tokyo, Japan). The cells were stored in the into patients through a central venous catheter without same freezer until use. additional post-thaw washing. Platelet transfusion was given as indicated to maintain the platelet count Ͼ20 × 109/l. According to the treatment plan, all patients Hematopoietic progenitor assay were scheduled for at least two cycles of disease-oriented Details of the progenitor assay have been published else- conventional-dose chemotherapy prior to the second trans- where.9 Briefly, cells (0.5–2.0 × 105/ml) were incubated in plant. All patients who had resectable residual disease with methylcellulose culture supplemented with 20% fetal bov- either normal or only minimally elevated tumor markers ine serum, 450 ␮g/ml human transferrin, 20 ng/ml interleu- after high-dose therapy were selected to undergo kin-3, 20 ng/ml stem cell factor, 20 ng/ml G-CSF and 2 additional surgery. units/ml erythropoietin. Quadruplicate cultures were incubated for 14 days under humid conditions with 5% O2,5% CO2 and 90% N2, and scored for colony-forming unit Results granulocyte/macrophage (CFU-GM). Procedure-related complications and engraftment Flow cytometric analysis of cells All six patients completed the consecutive-PBSCT, with an Aliquots (0.3–0.5 ml) of cell suspension (3 × 106 cells/ml) interval ranging from 3 to 5 months. Procedure-related were mixed with 1.5 ml RPMI-1640 supplemented with complications were transient and clinically negligible. 10% FBS and stored at 4°C for flow-cytometry analysis Engraftment data are summarized in Table 2. The mean performed within 48 h. One hundred microliters of cell sus- number of CD34+ cells in the ‘CD34(+) fraction’ was pension were then dispensed into test tubes (Falcon 2052) 3.31 × 106/kg (0.63–4.3 × 106/kg), while this number in the for staining and for a control. Staining was performed by ‘CD34(−) fraction’ could not be evaluated correctly due adding FITC-conjugated CD33 antibody and PE-conju- their scarcity (Ͻ0.1%) and estimated values are shown in gated CD34 antibody (Immunotech, Marseille, France, Table 2. The median numbers of infused MNC and CFU- IOM34) for 30 min at 4°C in the dark. Anti-mouse IgG1 GM were, respectively, 4.2 × 106/kg and 1.75 × 105/kg conjugated with FITC or PE was used as a control. Samples (0.31–4.6 × 105/kg) in the CD34(+) fraction, and were analyzed by a FACScan flow cytometer (Becton 4.8 × 108/kg and 3.35 × 105/kg (0.34–9.0 × 105/kg) in the Dickinson, San Jose, CA, USA), which was calibrated CD34(−) fraction. using caliBRITE beads, and compensated using the After the first transplant, one patient was given scheduled CD33/CD34 fluorescent information from non-coexpress- conventional dose consolidation chemotherapy before ing controls stained with Becton Dickinson reagents. After achieving a platelet count of 50 × 109/l, and was removed function was verified, the samples were drawn into the flow from the comparative analysis of platelet engraftment speed cytometer using FSC and SSC as gating parameters and for the first part of the transplant. In our historical data of each group of 2 × 104 cells was analyzed. The flow cyto- five conventional transplants for solid tumor with unmani- metric data were analyzed using a gated analysis via a set pulated PBSC, the number of days required to achieve a of SSC-FL2 parameters for the CD34+ cells to calculate the ANC Ͼ0.5 × 109/l and a platelet count Ͼ20 × 109/l and percentage of positive cells. Cell populations gated with a Ͼ50 × 109/l were, respectively, 12.0 (8–15), 20.0 (10–36), FL1-FL2 parameter were used in the analysis of lympho- and 37.0 (10–46). These numbers were, respectively, 14.5, cyte subsets. If the CD34+ population detected in 2 × 104 15.0, and 19.5 in the first transplants with the CD34-nega- cells fell short of 1 × 103 cells, the count was extended to tive fraction, and 13.5, 18.0 and 25.0 in the second trans- a maximum of 5 × 104 cells to collect a sufficient number plants with the CD34-positive fraction (Table 2). Although of CD34+ cells for evaluation. transplantation with the CD34-negative fraction tended to be associated with delayed recovery, there was no statisti- cally significant difference in the recovery speed between Transplant procedure the two types of transplantation. In a previous study, we demonstrated that cells that had Although an evaluation of the therapeutic outcome is not been recovered in the CD34-negative fraction still con- the major purpose of this study, which has only a short tained a large number of CFU-GM.10 Hence, these cells follow-up period, four of the six patients achieved complete New consecutive high-dose therapy modality T Kajiume et al 150 Table 2 Infused cells and engraftment

Case No. (/kg) infused cells transplanted with No. of days to engraft No. CD34(Ϫ) fr CD34(+) fr First transplant with CD34(Ϫ) fr Second transplant with CD34(+)fr

CD34(+)a CFU-GM CD34(+) CFU-GM ANC Ͼ 0.5 Plt Ͼ 20 Plt Ͼ 50 ANC Ͼ 0.5 Plt Ͼ 20 Plt Ͼ 50 (×106)(×105)(×106)(×105)

1 2 3.9 4.3 2.4 12 12 14 14 16 18 2 10 9.0 1.8 1.8 14 12 15 13 16 30 3 2.4 5.0 3.9 4.6 17 11 11 10 20 20 4 0.5 0.5 0.6 0.5 16 24 32 12 12 19 5 6.6 1.4 3.3 0.9 12 14 20 17 29 34 6 2.3 0.34 5.9 0.3 15 43 — 42 51 51 Mean 4.0 3.4 3.3 1.8 15 13 15 14 18 25 Ϯ 3.6 3.4 1.9 1.6

aEstimated number of CD34+ cells in the negative fraction is shown.

remission, as evidenced by complete disappearance of isolation of CD34+ cells, this has been associated with measurable tumor, and are currently surviving despite their inevitable cell loss. To minimize the risks and hazards of initial poor risk. Two patients (cases 2 and 5) did not therapy and to improve the therapeutic ratio of stem cell respond to therapy. therapy, we developed a consecutive-autograft concept, in which we recovered cells in the CD34-negative blood cell fraction and used them in the initial transplant. After posi- Discussion tive selection, some CD34bright cells may still migrate into the CD34-negative fraction, as do CD34dull progenitor cells. Dose intensification and the sequential use of agents to To confirm this, we showed that the CD34-negative fraction overcome drug resistance may benefit some children with still contains a large number of progenitor cells (CFU-GM) high-risk solid tumors. The size of the cancer cell mass which can support the supply of mature blood cells for might be reduced by repeated, closely timed courses of some time. Finally, to eliminate the possibility of infusing high-dose chemotherapy. Although PBSC transplantation malignant cells, patients then received a second and final has become an established option for the treatment of vari- transplant with CD34+ cells, which were indirectly depleted ous malignant disorders, the therapeutic efficacy of this pro- of malignant cells. cedure has not been conclusively determined in the treat- In addressing the utility of new therapeutic strategies, ment of childhood solid tumors.6,11–14 For the treatment of any comparative analysis must address the clinical end- solid tumors, it is likely that multiple intermediate-dose points of prolongation of survival and improved ultimate therapy delivers a higher dose rate with an improved safety cure rate, safety of the procedure and cost-effectiveness. margin than does a single high-dose therapy. To address We demonstrated that toxicity was acceptable with regards this point, we have developed a PBSCT program for pedi- to disease prognosis, and the cost-effectiveness of this cell atric patients.15 fractionation approach was self-evident. The speed of hem- Factors which limit the application of a multiple trans- atopoiesis was essentially the same between the two phases plantation strategy include patient tolerance and the fact of the transplants, with no difference from the histological that an adequate amount of stem cells to support multiple data of PBSCT with unmanipulated cells.6 The current courses of transplant may not be available in some patients. results confirm the efficacy of high-dose therapy as either A double harvesting procedure has the obvious disadvan- the first or subsequent salvage treatment in children with tage of delayed engraftment, primarily due to disturbance solid tumors who do not become disease-free after conven- of the stem cell pool and the marrow microenvironment by tional-dose chemotherapy, but who do respond to the the preceding transplant procedure. It has been established reported protocol. that PBSC grafts are as durable as marrow grafts to fully In conclusion, our procedure may make higher doses of support life-long hematopoiesis, and these are increasingly chemotherapy, as well as chemotherapy at more frequent used for rescue after high-dose therapy. Most importantly, intervals, more feasible, by ameliorating cytopenia, with an the graft size can be increased through the use of PBSC improved cost/benefit ratio and therapeutic margin. rather than bone marrow, which makes a double-transplant Reduction of the duration of aplasia may allow a wider procedure more feasible. The lower concentration of cancer application of multiple-cycle high-dose therapy, which may cells in PBSC graft than in marrow harvest products may be of particular importance in certain malignant disorders. be an additional advantage of the use of PBSC. Recognizing that cancer cells are relatively enriched in the Nevertheless, there is some evidence that contamination CD34-negative fraction, our next approach will include an of grafts with tumor cells may contribute to recurrence additional purification of this fraction for negative depletion upon reinfusion.16 Although the effective depletion of can- of cancer cells using disease-specific monoclonal anti- cer cells from a PBSC graft is possible through the positive bodies. New consecutive high-dose therapy modality T Kajiume et al 151 References engraftment after peripheral blood stem cell autografts cryopreserved by controlled vs uncontrolled-rate method. Bone 1 Hryniuk W, Bush H. The importance of dose intensity in Marrow Transplant 1994; 13: 801–804. chemotherapy of metastatic breast cancer. J Clin Oncol 1984; 9 Abe T, Takaue Y, Kawano Y, Kuroda Y. Effect of recombi- 2: 1281–1288. nant erythropoietin in interaction with stromal factors on cord 2 Stram DO, Matthay KK, O’Leary M et al. Consolidation blood hematopoiesis. Blood 1996; 87: 3212–3217. chemoradiotherapy and autologous bone marrow transplan- 10 Kawano Y, Takaue Y, Abe T et al. New dose-intensity strat- tation versus continued chemotherapy for metastatic neuro- egy: tandem autografts with purified blood CD34(Ϫ) and blastoma: a report of two concurrent Children’s Cancer Group CD34(+) cells. Exp Hematol 1996; 24: 1135 (Abstr.). Studies. J Clin Oncol 1996; 14: 2417–2426. 11 Kanold J, Berger M, Rapatel C et al. CD34+ cell immunose- 3 Shuster JJ. The role of autologous bone marrow transplan- lection from G-CSF-alone-primed peripheral blood in children tation in advanced neuroblastoma. J Clin Oncol 1996; 14: with low body mass. Br J Haematol 1995; 91: 431–433. 2413–2414. 12 Diaz MA, Villa M, Alegre A et al. Collection and transplan- 4 Takaue Y, Kawano Y, Abe T et al. Collection and transplant tation of peripheral blood progenitor cells mobilized by G- of peripheral blood stem cells in very small children weighing CSF alone in children with malignancies. Br J Haematol 1996; 20 kg or less. Blood 1995; 86: 372–380. 94: 148–154. 5 Takaue Y, Kawano Y, Kuroda Y. Application of recombinant 13 Diaz MA, Alegre A, Villa M et al. Pediatric experience with granulocyte colony-stimulating factor in peripheral blood stem- autologous peripheral blood progenitor cell transplantation: cell transplantation: a pediatric experience. In: Levit DJ, Mert- influence of CD34+ cell dose in engraftment kinetics. Bone elsmann R (eds). Hematopoietic Stem Cells: Biology and Thera- Marrow Transplant 1996; 18: 699–703. peutic Applications. Marcel Dekker: New York, 1995, pp 611– 14 Jones N, Williams D, Broadbent V et al. High-dose melphalan 630. followed by autograft employing non-cryopreserved periph- 6 Eguchi H, Takaue Y. Peripheral blood stem cell autografts in eral blood progenitor cells in children. Eur J Cancer 1996; the treatment of pediatric solid tumors. In: Dicke KA, Keating 32: 1938–1942. A (eds). Autologous Marrow and Blood Transplantation. The 15 Takaue Y. Peripheral blood stem cell autografts for the treat- Cancer Treatment Research and Educational Institute: Arling- ment of childhood cancer: a review of Japanese experience. J ton, 1995, pp 597–606. Hematother 1993; 2: 513–518. 7 Takaue Y, Kawano Y, Watanabe A et al. Transplantation with 16 Brenner MK, Rill DR, Noen RC et al. Gene-marking to trace purified or unmanipulated mobilized blood stem cells in chil- origin of relapse after autologous bone marrow transplan- dren. In: Ikehara S, Takaku F, Good RA (eds). Bone Marrow tation. Lancet 1993; 341: 85–86. Transplantation: Basic and Clinical Studies. Springer-Verlag: New York, 1996, pp 246–249. 8 Takaue Y, Abe T, Kawano Y et al. Comparative analysis of