Collection and Use of Peripheral Blood Stem Cells in Young Children with Refractory Solid Tumors

Collection and use of peripheral blood stem cells in young children with refractory solid tumors

V Shen, C Woodbury, R Killen, C Van de Ven, L Sender and MS Cairo

Division of Pediatric Hematology/Oncology, Children’s Hospital of Orange County, Orange, CA, USA

Summary: dose-intensive chemotherapy compared with less intensive ‘conventional’ chemotherapy in a variety of solid tumors, Fifteen children 4 years of age or under (8–46 months), including breast cancer, lymphoma, ovarian cancer, and weight 7.8 to 17 kg, underwent 44 peripheral blood stem sarcomas.2–5 cell (PBSC) collections. Diagnoses included PNET/ Hematologic toxicity is a common dose-limiting finding medulloblastoma (five), neuroblastoma (five), and following dose escalation. Recombinant hematopoietic others (five). PBSCs were collected following G- growth factors and several interleukins (IL-3, IL-6, IL-11) CSF/GM-CSF or chemotherapy plus G-CSF/GM-CSF have been used to enhance hematopoietic recovery follow- mobilization. All PBSC collections were well tolerated. ing non-myeloablative dose-intensive chemotherapy. To -The average yield per collection was 6.80 ؋ 108 mono- overcome prolonged and profound life-threatening neutro nuclear cells/kg (1.1–30 ؋ 108/kg) or 57.60 ؋ 106 penia and thrombocytopenia following myeloablative or CD34+/kg (1.37 to 480 ؋ 106/kg). Eight patients submyeloablative chemotherapy, hematopoietic stem cells underwent stem cell transplantation following mye- have been employed to stimulate hematological reconsti- loablative chemotherapy. Six of the eight children who tution. Peripheral blood stem cells (PBSC) have been received PBSC following myeloablative therapy also increasingly applied in place of bone marrow to reconstitute received autologous bone marrow (0.7 to 3.6 ؋ 108 hematopoiesis.6–8 The major advantage of using PBSCs MNC/kg). One heavily pretreated patient experienced over bone marrow is the rapid hematopoietic reconstitution delayed hematologic reconstitution, while the remaining through maturation of more committed progenitor cells.9 seven patients had a median ANC recovery to Early hematological recovery results in shorter duration of -Ͼ0.5 ؋ 103/␮l by day ؉10 (9–11 days) and platelets antibiotic use, fewer blood product transfusions, early hos Ͼ50 ؋ 103/␮l by day ؉15 (12–17 days). Seven patients pital discharge, and significant cost saving. In addition, the received PBSCs following repetitive submyeloablative collection process for PBSCs is less invasive than bone chemotherapy (ICE: ifosfamide 1.8 g/m2/day, etoposide marrow harvest. PBSC collection is also a preferred source mg/m2/day ؋ 5, carboplatin 400 mg/m2/day ؋ 2) or of stem cells in patients who have had prior pelvic 100 other similar combination chemotherapy. Median days irradiation.10 -to recover ANC у1 ؋ 103/␮l and platelets у100 ؋ The safety and efficacy of PBSC collection is well estab ␮l in children receiving ICE ؉ PBSCs were 10 and lished in adults and older children.11 However, experience/103 14 days, respectively, compared with 16 and 22 days in regarding the feasibility, safety, and efficacy of the pro- children receiving ICE ؉ G-CSF in historical controls. cedure in infants and young children is still limited.12–17 In conclusion, collection and use of PBSCs to support We report our experience in collection and use of PBSCs either myeloablative chemotherapy or multicycle sub- in young children with refractory solid tumors to support myeloablative chemotherapy is well tolerated and may myeloablative chemotherapy and multicycle submyelo- enhance hematological recovery in young children and ablative therapy. infants. Keywords: peripheral blood stem cell collection; hematological recovery; infant Methods

PBSC mobilization Intensification of chemotherapy is a strategy frequently Patients of Children’s Hospital of Orange County who were 1 used in treating recurrent or refractory malignancies. identified as candidates for high-dose chemotherapy (either Improved clinical outcomes have been demonstrated using myeloablative or submyeloablative) for the treatment of their malignancies were eligible for PBSC collection. Among 36 patients who had PBSC collections between Correspondence: Dr V Shen, Division of Hematology/Oncology, Chil- 1993 and 1995, 15 children were 4 years old or less and dren’s Hospital of Orange County, 455 S Main Street, Orange, CA constitute this report. Thirteen patients received PBSC 92668, USA 2 Presented in part at the 37th Annual Meeting of The American Society mobilization with cyclophosphamide (4.0 g/m ) or ifosfam- 2 of Hematology, December 1995, Seattle, WA, USA ide (9.0 g/m )-based myelosuppressive chemotherapy and Received 29 January 1996; accepted 22 October 1996 hematopoietic growth factor (G-CSF 5.0 ␮g/kg/day or GM- Use of PBSC in children V Shen et al 198 CSF 250 ␮g/m2/day). Two patients were mobilized with High-dose chemotherapy and PBSC infusions hematopoietic growth factor alone (patients Nos 8 and 13) Eight patients received PBSC infusions following myelo- for 5 days prior to PBSC harvest. ablative chemotherapy. Six also received autologous bone marrow. Myeloablative chemotherapy regimens included: PBSC harvest thiotepa (900 mg/m2), etoposide (750 mg/m2)witheither BCNU (600 mg/m2) (two patients) or carboplatin PBSC collections were performed when WBC recovered to (1500 mg/m2) (four patients); and thiotepa (900 mg/m2), eto- Ͼ1000/mm3 following the nadir from the myelosuppressive poside (1800 mg/m2), and cyclophosphamide (200 mg/kg) chemotherapy or after 5 days of hematopoietic growth fac- (two patients). Eight patients received PBSCs following tor mobilization. Eleven patients had a pediatric MedComp multicycle submyeloablative chemotherapy with one of the hemodialysis catheter (Medical Components, Harleysville, following regimens: ifosfamide (900 mg/m2/day), etoposide PA, USA) inserted at the time of diagnosis. The other four (100 mg/m2/day for 5 days), carboplatin (400 mg/m2/day for patients required a temporary hemodialysis catheter, placed 2 days) (ICE regimen, four patients, seven courses); car- under conscious sedation on the pediatric ward, before boplatin (500 mg/m2/day for 2 days) and etoposide PBSC collection. After informed consent was obtained, (150 mg/m2/day for 3 days) (two patients, three courses); patients underwent leukapheresis through a COBE Spectra carboplatin (400 mg/m2 × 1 day) and cyclophosphamide Blood Component Separator (COBE, Lakewood, CO, (2 g/m2 × 2 days) (two patients, two courses); carboplatin USA). Anticoagulant-citrate-dextrose solution USP (ACD- (400 mg/m2/day for 2 days), cyclophosphamide Formula A) (Baxter Healthcare, Fenwal Division, Deer- (2 g/m2/day × 2 days), and etoposide (250 mg/m2/day × 3 field, IL, USA) was used as the anticoagulant during the days) (one patient, two courses). One patient received procedure and calcium gluconate (American Reagents, sequential high-dose chemotherapy consisting of one cycle Shirley, NY, USA) (50–100 mg/kg) was infused during the of carboplatin (1950 mg/m2) followed by one cycle of etopo- procedure to prevent citrate toxicity and maximize the cit- side (3000 mg/m2), then a third cycle with cyclophospham- rate-to-whole-blood ratio. For children weighing less than ide (3600 mg/m2 and etoposide 450 mg/m2). 25 kg, the extracorporeal circuit was preprimed with 300 ml PBSCs were reinfused 24–48 h after completion of of irradiated modified whole blood (HCT of 45–55%) to chemotherapy. For patients receiving submyeloablative avoid acute blood volume loss. Total running time per day chemotherapy, PBSC infusions were performed in an out- was targeted at 3 h or three times the blood volume pro- patient setting. Patients were premedicated with acetamin- cessed during 1993 and 1994, and was increased in 1995 ophen (10 mg/kg), diphenhydramine (1 mg/kg), ondanse- to 5–6 h or until six times the blood volume processed. tron (0.15 mg/kg), mannital (0.5 g/kg) and hydrocortisone Mild sedatives such as chloral hydrate or pentobarbital (1 mg/kg) prior to PBSC reinfusion. PBSCs were thawed were used to avoid excessive movement. Vital signs and in a 37°C waterbath at the bedside and were infused via blood pressures were closely monitored during the pro- an intravenous catheter by gravity without filter. Patients cedures. PBSC collections were continued until a minimum received 4 h intravenous hydration at 125 ml/m2/h follow- of 6 × 108/kg of mononuclear cells (MNC) or 5 × 106/kg of + ing PBSC infusion. CD34 cells were collected. Higher numbers of PBSCs A minimum of 5 × 106/kg of CD34+ cells or 6 × 108/kg were collected if the patient was to receive multicycle dose- of mononuclear cells were infused following myeloablative intensified chemotherapy with PBSC support. Complete chemotherapy, and a minimum of 2.5 × 106/kg of CD34+ blood counts and electrolytes were checked before and after cells were given following submyeloablative chemo- the procedure. Patients received packed red blood cells or therapy. Patients also received hematopoietic growth fac- platelet transfusions prior to leukapheresis if the hematocrit tors (rhG-CSF or rhGM-CSF) following PBSC infusion was less than 26% or platelet count less than 50 000/␮lto until stable myeloid recovery was achieved (usually when ensure the safety of the procedure. absolute neutrophil count (ANC) у5000/␮l). CBCs were After stem cell collection, the apheresis product was cen- monitored daily following myeloablative chemotherapy and trifuged at 1500 r.p.m. for 15 min to remove the plasma three times a week after submyeloablative chemotherapy and then resuspended with 5A McCoy’s modified media ++ ++ until ANC recovered to greater than 1000/␮l and platelet (Ca ,Mg free) (Sigma, St Louis, MO, USA) sup- count greater than 50 000/␮l. plemented with 10% autologous platelet-poor plasma to make the final cell concentration of 1–1.5 × 108 cells/ml. One aliquot of cell suspension was collected to check for + Results mononuclear cell count and CD34 cell count. Cryoprotect- ant (DMSO 10% V/V; Tera Pharmaceuticals, Buena Park, Patient characteristics CA, USA) was then added and the final product was trans- ferred into freezing bags and cryopreserved in a controlled- Fifteen children, nine boys and six girls, ages ranging from rate freezer and stored at vapor phase in a liquid nitrogen 8 months to 46 months (median 30 months), weight 7.8 kg freezer. CD34+ cell determination was performed by flow to 17 kg with a median of 12.9 kg, underwent a total of cytometry using phycoerythrin (PE)-conjugated CD34 anti- 44 PBSC collections (Table 1). Disease categories included body and fluorescein-isothiocyanate (FITC)-conjugated neuroblastoma (five patients), PNET (three patients), CD45 antibody (Becton Dickinson, San Jose, CA, USA). medulloblastoma (two patients), malignant glioma Total nucleated cells were gated using a CD45 marker and (two patients), recurrent Wilms’ tumor (two patients), and the CD34 percentage was then calculated. choroid plexus carcinoma (one patient). The interval Use of PBSC in children V Shen et al 199 Table 1 Patient characteristics and the results of peripheral blood stem cell collection

Patient Age at Wt Diagnosis Prior Prior Chemo Day of Growth Blood vol MNC yield % CD34 in CD34 No. PBSCH (kg) chemo XRT priming harvest factor processed (×108/kg) total yield (months) from start prime (ml) nucleated (×106/kg) of chemo cells

1 8 8 Neuroblastoma VAC ×2 N VAC 14 GM-CSF 4310 5.03 1.05 5.42 15 4500 6.00 0.73 5.46 16 4500 6.06 0.71 5.62 2 9 9.4 Brain stem glioma CPM, VP N CPM, 21 G-CSF 2200 12 NA NA Carbo VP, Carbo 22 1650 4.6 3 10 9.6 Anaplastic VICE ×2 N VICE 18 G-CSF 1450 2.8 NA NA ganglioglioma 21 2410 3.0 22 2508 4.7 23 2063 7.0 4 11 7.8 Choroid plexus CPM, VP, N CPM, VP 21 1100 1.2 NA NA carcinoma V CDDP, V 24 1777 4.8 CDDP ×3 25 1854 1.4 5 15 10 PNET ICE ×1 N ICE 14 G-CSF 4898 4.3 1.18 6.4 15 4905 1.2 2.63 3.5 16 5045 14.2 2.42 47.4 6 18 10.4 Neuroblastoma VAC ×1 N VAC 20 GM-CSF 6614 13.9 1.36 28.2 7 19 12.9 PNET ICE ×1 N ICE 17 G-CSF 2898 19.8 5.13 206 18 3400 22.4 7.59 210 19 3080 15.8 6.94 121 8 26 13.6 Wilms’ tumor CPM, Y No NA G-CSF 6693 6.5 0.22 3.9 Carbo, NA 6603 6.1 0.63 4.4 VP ×6 9 34 15.6 PNET VICE ×1 N VICE 21 G-CSF 2300 1.8 NA NA 22 2300 1.2 23 2498 1.3 24 2569 1.1 VICE ×2 VICE 19 2243 2.6 20 3315 1.5 10 34 13 Neuroblastoma CDDP, N CDDP, VP 30 GM-CSF 4458 5.8 NA NA Doxo Doxo, CPM 31 4097 1.4 CPM, VP ×4 11 40 17 Medulloblastoma ICE ×1 N ICE 20 G-CSF 2900 11.4 4.81 6.5 21 6000 30.0 1.61 71.6 22 2799 16.6 1.64 29.7 12 41 15.6 Neuroblastoma CPM, N CPM, Carbo 21 G-CSF 4263 15.7 2.75 82.3 Carbo Carbo, VP 22 3633 5.5 2.26 15.3 21 G-CSF 2943 2 1.29 6.3 22 2270 1.6 1.29 4.4 13 41 11 Medulloblastoma VICE ×5 N No NA GM-CSF 6300 3.9 0.25 1.37 NA 6301 3.4 0.65 2.77 NA 6300 3.3 0.52 1.97 14 44 16.4 Wilms’ tumor V, ACT-D N CPM, VP 19 G-CSF 2198 2.2 NA NA CPM, VP 20 3515 2.5 21 3587 1.5 15 46 13.4 Neuroblastoma CPM, N CPM, Carbo 11 G-CSF 4286 6.1 2.34 34 Carbo 13 5274 14.3 1.09 480

Median 26 12.9 21 3358 4.65 1.32 6.45 Mean 26 12.2 20 3655 6.80 2.13 57.67 Range 8–46 7.8–17 11–31 1100–6693 1.1–30.0 0.22–7.59 1.37–480

ICE = ifosfamide, carboplatin, etoposide; V = vincristine; CPM = cyclophosphamide; VP = etoposide; Carbo = carboplatin; Doxo = doxorubicin; ACT- D = dactinomycin; NA = not available. Use of PBSC in children V Shen et al 200 between diagnosis and PBSC collection was 2 months in 10 unsupported platelet count у50 000/␮l on day 201 after patients, 4–8 months in the other four patients. One patient PBSC infusion); the remaining seven patients had ANC (patient No. 8) had abdominal radiation prior to PBSC har- recovery to greater than 500/␮l in 9–11 days (median 10 vest; the remaining 14 patients did not receive any radiation days) following stem cell infusion. Unsupported platelet therapy prior to PBSC harvest. counts equal to or greater than 50 × 103/␮l occurred on day 12 to day 18 with a median of 15 days (Table 2). The rate of platelet recovery did not correlate with the number of Peripheral blood stem cell collection mononuclear cell or CD34 cells from bone marrow or PBSC collections were started when patients showed evi- PBSCs infused. The median discharge date was 27 days dence of hematologic recovery from myelosuppressive from transplant. chemotherapy. Median WBC at the time of PBSC collec- Nine patients received a total of 18 courses of submye- tion was 14.7 × 103/␮l (range 1.3–88.2 × 103/␮l) and loablative chemotherapy with PBSC support for cytoreduc- median ANC was 10.1 × 103/␮l. The collection time ranged tion. Patient No. 13 died shortly after PBSC infusion from from 1.9 to 6 h per apheresis. The inlet/AC ratios were a rapid progressive leptomeningeal disease and was not between 12:1 and 14:1. The whole blood flow rate ranged evaluable to hematologic recovery. Three patients (Nos 5, from 9 cm3/min to 21 cm3/min as determined by the 7 and 9) went on to subsequent stem cell transplants. patient’s total blood volume. Total blood processed per leu- Median days to ANC Ͼ1000/␮l was 11 days (range 6–17 kapheresis ranged from 1100 to 6693 ml, which was equiv- days) and platelet count Ͼ100 × 103/␮l was 17 days (range alent to 2- to 8-fold of blood volume. The average yield per 11–30 days) from the day of PBSC infusion. Subsequent collection was 6.80 Ϯ 1.00 × 108/kg of mononuclear cells chemotherapy was given 15–33 days following PBSC (range 1.1–30 × 108/kg). CD34+ cells per collection ranged infusion (median 22 days) (Table 3). Four patients received from 1.37 to 480 × 106/kg with a median 6.45 × 106/kg and a total of seven courses of ICE chemotherapy (ifosfamide an average of 57.6 Ϯ 22 × 106/kg per collection. 1.8 g/m2/day × 5 days, etoposide 100 mg/m2/day × 5 days and carboplatin 400 mg/m2/day × 2 days) with PBSC support. Median ANC recovery to у1000/␮l was 10 days and Side-effects from PBSC harvest and reinfusion median platelet recovery to у100 000/␮l was 14 days. The Modified whole blood was used in priming the extra-cor- median dose itensity administered was 2800 mg/m2/week poreal circuit in all 44 leukaphereses. Hematocrit remained for ifosfamide, 156 mg/m2/week for etoposide, and stable after the procedure. No hemodynamic instability was 249 mg/m2/week for carboplatin. Only one patient (patient noted during PBSC collection. Since all patients received No. 1) did not receive PBSC etoposide, and calcium gluconate infusion during PBSC collection, no 249 mg/m2/week for carboplatin. Only one patient (patient clinical or laboratory evidence of hypocalcemia was No. 1) did not receive PBSC infusion because tumor cell observed. Transient mild hypokalemia (K+ Ͻ3.5 mEq/l contamination was found in the PBSC product by immuno- Ͼ3.0 mEq/l) was noted immediately following apheresis cytochemical studies. and usually self-corrected by the next day. A significant drop of platelets occurred in all patients secondary to trap- ping of platelets in the PBSC product. The average drop Discussion of platelets after PBSC harvest was 59.8 Ϯ 10.0 × 103/␮l or 50 Ϯ 3.7%. In the past, the collection of PBSCs in infants and young Nausea and vomiting were common during the PBSC children was hampered by technical difficulties such as dif- reinfusion and occurred in 75% of patients. Two earlier ficult venous access and lack of cooperation of patients. patients developed transient hypoxia while receiving a large Young children have a small blood volume to tolerate the volume of PBSCs and bone marrow in a short period of redistribution of blood into the extra-corporeal circuit of time. They responded to diuretics and oxygen. Since our the apheresis apparatus and they are more susceptible to procedure has been changed to dividing stem cells into two citrate toxicity. Lasky et al12 first reported their experience separate infusions if the total stem cell volume was greater in the collection and use of peripheral blood stem cells in than 20 ml/kg, there has been no incidence of hypoxia with three children. They primed the Fenwall CS3000 separator this new procedure. Transient hemoglobinuria occurred with blood reconstituted with fresh frozen plasma/albumin after PBSC infusion in all patients. and packed red blood cells. Heparin was added to ACD-A to increase the whole blood flow rate. One patient developed citrate toxicity with tingling and facial numb- Hematopoietic recovery following high-dose ness. Administration of calcium gluconate infusion during chemotherapy and PBSC reinfusion PBSC collection can alleviate the citrate toxicity and max- Eight patients received myeloablative chemotherapy as imize blood flow rate.15–17 This is particularly important in consolidation therapy for treatment of their tumors. Two young children who cannot communicate well about the patients received PBSCs alone, six patients received PBSCs symptoms of hypocalcemia. and autologous bone marrow following myeloablative Venous access is usually poor in young children. Com- chemotherapy. One patient (No. 8) who had PBSCs har- monly used silastic catheters (Hickman or broviac) do not vested 5 months from the diagnosis and received abdominal have enough resistance to withstand the pressure of the radiation therapy within 1 month of PBSC collection had blood outflow. Femoral catheters or radial artery catheters a delayed hematologic recovery (ANC у500/␮l on day 53, have been utilized for PBSC collection.15–17 We have Use of PBSC in children V Shen et al 201 Table 2 Hematological reconstitution following myeloablative chemotherapy and stem cell transplantation

Patient No. Chemo Prior Bone marrow PBSC PBSC Days to Days to Days to XRT (108 MNC/kg) (108 MNC/kg) (106 CD34/kg) ANC Ͼ500/␮l Plt Ͼ50 × 103/␮l discharge

2 TT-VP-BCNU N 2.4 8.0 NA 9 15 21 3 TT-VP-BCNU N 1.9 11 NA 11 12 30 4 TT-VP-Carbo N 2.6 7.4 NA 9 13 27 5 TT-VP-Carbo N — 5.9 19.8 11 16 21 7 TT-VP-Carbo N 0.7 7.9 60.5 10 17 27 8 TT-VP-CPM Y — 12.1 7.9 53 201 80 9 TT-VP-Carbo N 2.8 6.4 NA 10 15 28 14 TT-VP-CPM N 3.6 6.2 NA 10 18 20

Median 2.5 7.7 19.8 10 15.5 27 Mean 2.33 8.1 60.5 10 38 32 Range 0.7–3.6 5.9–12.1 7.9–60.5 9–53 12–201 20–80

NA = not available; TT-VP-BCNU = thiotepa 900 mg/m2, etoposide 750 mg/m2 BCNU 600 mg/m2; TT-VP-Carbo = thiotepa 900 mg/m2, etoposide 750 mg/m2 and carboplatin 1500 mg/m2; TT-VP-CPM = thiotepa 900 mg/m2, etoposide 1800 mg/m2, cyclophosphamide 200 mg/kg.

Table 3 Hematologic recovery following submyeloablative chemotherapy and PBSC support

Patient Chemo PBSC PBSC Days to Days to Days to Plt Days to Plt Days to next No. (108 MNC/kg) (106 CD34/kg) ANC Ͼ500/␮l ANC Ͼ1000/␮l Ͼ50 × 103/␮l Ͼ100 × 103/␮l chemo given

5 ICE 1.45 3.21 8 8 9 12 16 ICE 2.08 3.06 8 8 11 13 15 6 CCE 1.39 2.83 15 17 15 19 23 CCE 1.46 2.95 15 17 16 19 25 7 ICE 3.3 34.2 8 10 12 14 18 ICE 3.3 34.2 10 11 11 11 15 NA 11 11 15 17 18 6.2 ءCarbo 9 NA 6 6 10 13 b 6.0 ءEtoposide VP-CPM 5.5 NA 10 11 12 16 33a 10 ICE 7.2 NA 15 17 16 19 b 11 ICE 7.6 4.4 8 9 11 14 15 ICE 5.5 9.9 9 10 17 22 23 12 CE 5.2 8.2 11 13 18 22 22 CC 5.2 8.2 9 10 NE NE 22 CE 5.2 8.2 9 9 26 30 27 15 CE 0.8 4.6 10 11 19 19 24 CC 0.8 4.5 10 11 18 20 25

Median 5.2 4.6 10 11 15 17 22 Mean 4.0 9.9 10 11 15 17 21 Range 0.8–7.6 2.83–34.2 6–15 6–17 9–26 11–30 15–33 aChemotherapy delayed secondary to bacterial sepsis. bPatient went on to surgery. ICE = ifosfamide 1.8 g/m2/d × 5 d, etoposide 100 mg/m2/d × 5 d, carboplatin 400 mg/m2/d × 2 d; CCE = cyclophosphamide 2 g/m2/d × 2 d, carboplatin 400 mg/m2/d × 2 d, etoposide 250 mg/m2/d × 3d;CE=carboplatin 500 mg/m2/d × 2 d, etoposide 150 mg/m2/d × 3d;CC=carboplatin 400 mg/m2, cyclo- etoposide 3000 mg/m2; VP-CPM = etoposide 450 mg/m2, cyclophosphamide = ءCarboplatin 1950 mg/m2; etoposide = ءphosphamide 4000 mg/m2; Carbo 3600 mg/m2;NA=CD34 determination not done; NE = not evaluable, patient developed gross hematuria and received platelet transfusions to keep platelet count above 100 000/␮l. employed a pediatric size MedComp hemodialysis catheter In the initial Lasky et al12 study, no mobilization was for PBSC collection with success. The MedComp catheter utilized before PBSC collection. The PBSC yield was rela- can last for many months after placement and hence is a tively low: 4.8 to 6.4 × 108 MNC/kg were collected during more convenient device for patients requiring multiple six leukapheresis procedures per patient. Mobilization of PBSC harvests, high-dose multi-agent chemotherapy and the stem cells with myelosuppressive chemotherapy with vigorous supportive care. or without hematopoietic growth factors greatly facilitates To improve the yield of collections, we found that PBSC collection and enhances the hematological recovery administering mild sedatives reduced excessive movement after high-dose chemotherapy.6,7,18,19 Takaue et al16 in young children and improved the blood flow rate during reported their experience with PBSC collection in 61 chil- apheresis. The longer collection time per pheresis also dren. All patients received myelosuppressive chemotherapy improved PBSC yield, decreased the number of procedure prior to apheresis. In 18 children 4 years or under, the total days, and was still well tolerated by our young patients. yield was 111 Ϯ 97 × 104 CFU-GM/kg from a mean of 3.0 Use of PBSC in children V Shen et al 202 aphereses, which was significantly better than older children undergoing high-dose therapy. Platelet recovery was dren. Kanold et al14 found no difference in the number of correlated with the number of CFU-Meg/kg infused.20 In PBSCs collected or in the hematological reconstitution fol- our study, the hematological reconstitution was excellent in lowing PBSC transplant between patients mobilized with seven out of eight patients following myeloablative chemo- G-CSF alone and patients mobilized with chemotherapy therapy. Since the majority of these patients had brain plus G-CSF or GM-CSF. The median total mononuclear tumors and severe mucositis, we maintained their platelet cells collected in their report was 14 × 108/kg (range 6– counts above 50 × 103/␮l in the early transplant course. In 83 × 108/kg).14 Demeócq et al15 reported CFU-GM col- six patients with brain tumors (patient Nos 2, 3, 4, 5, 7, 9), lected in children receiving hematopoietic growth factor the median days to ANC Ͼ500/␮l and untransfused platelet was 7.8-fold higher than in children not receiving growth Ͼ50000/␮l were 10 and 15 days. One heavily pretreated factors. He also found the cumulative intensity of prior patient had a severe delay in hematologic recovery. His treatment received is an essential factor in the number of PBSCs were collected 6 months following diagnosis and PBSCs collected. Although various assays were employed he had undergone prior abdominal radiation therapy. It is in different reports, Leibundgut et al20 demonstrated a sig- advisable that a backup PBSC or bone marrow should be nificant correlation between CFU-GM/kg with MNC/kg, collected in heavily pretreated patients before administering CD34+ cells/kg, and CD34+ 33− cells/kg in apheresis pro- myeloablative therapy. ducts in children. The yield of PBSC collection in our Single-course, high-dose chemotherapy with autologous experience compared favorably with other reports in pedi- stem cell transplantation has been demonstrated to be cap- atric patients (Table 4). Sufficient cells for transplant use able of achieving high rates of response in patients in whom can be collected from a single apheresis. CD34+ cell assays conventional chemotherapy has previously failed. These provide a rapid and reliable assessment of the quantity of responses, however, are frequently of short duration. Cura- stem cells collected.20 WBC counts at the time of apheresis, tive chemotherapy requires the administration of repeated prior chemotherapy, and radiation therapy affect PBSC courses of effective chemotherapy regimens. A combi- yields.21–23 Heavily pretreated patients with limited bone nation of ICE chemotherapy (ifosfamide 1.8 g/m2/day × 5 marrow reserve may not have effective mobilization. It is days, etoposide 100 mg/m2/day × 5 days, and carboplatin important to identify the high-risk patient population who 400 mg/m2/day × 2 days) with rhG-CSF support has been will benefit from high-dose therapy at the time of diagnosis used in children with recurrent or refractory solid tumors. and to plan PBSC harvest early in their course of treatment. The overall response rate (CR + PR) was 51%. However, After PBSC infusion following high-dose therapy, hema- the treatment was associated with a very high incidence of tological recovery is correlated with the number of CD34+ grade III/IV hematological toxicities. The median cells or CFU-GM/kg. Kawano et al13 demonstrated that recoveries of ANC to у1000/␮l and platelets у100 000/␮l patients who received Ͼ1 × 105 CFU-GM/kg had an ANC were 16 and 22 days from the completion of chemo- recovery Ͼ0.5 × 103/␮l between day 6 and 15 (mean therapy.24 Four patients reported here received a total of 10.5 Ϯ 2.5 days) and the platelet count reached seven courses of ICE chemotherapy with PBSC support. Ͼ50 × 103/␮l between days 9 and 89 (mean 24 Ϯ 21 days). Median ANC recovery to у1000/␮l was 10 days and The ANC recovery in patients receiving Ͻ1 × 105 CFU- median platelet recovery to у100 000/␮l was 14 days. The GM/kg was between day 13 and 39 (22.0 Ϯ 8.4 days); five use of PBSCs may facilitate the hematological recovery and of 10 patients who had received Ͻ1 × 105 CFU-GM/kg enable repeated cycles of high-dose chemotherapy to be failed to recover platelets within 8 months.13 Leibundgut et delivered at short intervals for initial cytoreduction.25,26 al20 found a significant correlation between the number of Several hematopoietic growth factors with thrombopoietic MNC/kg, CFU-GM/kg, CD34+ cells/kg and CD34+ CD33− activities have been cloned and are currently being investi- cells/kg reinfused and the time to myeloid recovery in chil- gated in clinical trials. Large clinical trials are needed to

Table 4 PBSC collection and hematological recovery following PBSC in young children

Author Patient Age range Weight Mobilization Average No. Total yield Following myeloablative (Ref.) No. (median) (median) of aphereses (median) chemotherapy

Days to Days to ANC Ͼ500/␮l Plt Ͼ20 000/␮l (median) (median)

Takaue16 15 7 mo–3 yrs 7–15 kg Chemo 3 1.5–16 × 108 MNC/kg 9–22 12–132 (2 yrs) (11 kg) (8 × 108) (12) (55) Demeócq15 8 0.5–3 yrs 6.8–14 kg Chemo Ϯ G-CSF/ 3 3.2–398 × 104/kg 11–44 11–48 (2.5 yrs) (12.4 kg) GM-CSF CFU-GM (18) (27) (58 × 104/kg) Shen21; 15 8 mo–3 yrs 7.8–17 kg G-CSF/ 3 6.2–58 × 108 MNC/kg 9–53 12–201 this study (26 mo) (12.9 kg) GM-CSF Ϯ chemo (16.6 × 108) (10) (15.5) Takaue17 38 6–100 mos 7–20 kg Chemo Ϯ G-CSF 1.2 0.5–28 × 108 MNC/kg 6–15 9–46 (2.7 × 108) (10) (14) Use of PBSC in children V Shen et al 203 compare the cost, safety, and the rate of hematological 7 Hohaus S, Goldschmidt H, Ehrhardt R, Haas R. Successful recovery following high-dose chemotherapy with either autografting following myeloablative conditioning therapy PBSC rescue or a combination of granulopoietic and throm- with blood stem cells mobilized by chemotherapy plus rhG- bopoietic growth factors. CSF. Exp Hematol 1993; 21: 508–514. In tumors with a high propensity for bone marrow met- 8 Schwartzberg L, Birch R, Blanco R et al. Rapid and sustained hematopoietic reconstitution by peripheral blood stem cell astasis, tumor cells may be mobilized into peripheral blood 27,28 infusion alone following high-dose chemotherapy. Bone Mar- circulation. Five patients with stage IV neuroblastoma row Transplant 1993; 11: 369–374. and bone marrow disease had PBSC harvest after the first 9 Beyer J, Schwella N, Zingsem J et al. Hematopoietic rescue course of chemotherapy. PBSC product was reinfused fol- after high-dose chemotherapy using autologous peripheral- lowing subsequent chemotherapy. Although immunostain- blood progenitor cells or bone marrow: a randomized compari- ing of PBSC product was negative for tumor cells, three son. J Clin Oncol 1995; 13: 1328–1335. patients developed progressive disease shortly after demon- 10 Haas R, Hohaus S, Egerer G et al. Recombinant human gra- stration of initial tumor response. One patient relapsed in nulocyte–macrophage colony-stimulating factor (rhGM-CSF) the bone marrow and the other two patients developed new subsequent to chemotherapy improves collection of blood bone lesions. In another study, Rosenthal et al29 demon- stem cells for autografting in patients not eligible for bone + marrow harvest. Bone Marrow Transplant 1992; 9: 459–465. strated that use of positive selected CD34 cells from PBSC 11 Inwards D, Kessinger A. Peripheral blood stem cell transplan- product may avoid tumor cell contamination. tation: historical perspective, current status, and prospects for In conclusion, PBSC collection is safe and feasible in the future. Trans Med Rev 1992; VI: 183–190. infants and young children. PBSCs can be used to reconsti- 12 Lasky L, Fox S, Smith J, Bostrom B. Collection and use of tute hematopoiesis following myeloablative chemotherapy peripheral blood stem cells in very small children. Bone Mar- and to support multicycle submyeloablative chemotherapy row Transplant 1991; 7: 281–284. in hopes for improved tumor response. Future directions of 13 Kawano Y, Takaue Y, Watanabe T et al. Effects of progenitor PBSC technology include positive selection of CD34+ cells cell dose and preleukapheresis use of human recombinant gra- to avoid tumor cell contamination and ex vivo expansion nulocyte colony-stimulating factor on the recovery of hemato- of CD34+ population to obviate the need for repetitive poiesis after blood stem cell autografting in children. Exp 30 Hematol 1993; 21: 103–108. leukapheresis. 14 Kanold J, Rapatel C, Berger M et al. Use of G-CSF alone to mobilize peripheral blood stem cells for collection from children. Br J Haematol 1994; 88: 633–635. Acknowledgements 15 Demeócq F, Kanold J, Chassagne J et al. Successful blood stem cell collection and transplant in children weighing less We would like to thank Linda Rahl for her editorial assistance in than 25 kg. Bone Marrow Transplant 1994; 13: 43–50. the preparation of this manuscript. This work was supported by 16 Takaue Y, Watanabe T, Abe T et al. Experience with periph- grants from the Pediatric Cancer Research Foundation and the eral blood stem cell collection for autografts in children with Walden W and Jean Young Shaw Foundation. active cancer. Bone Marrow Transplant 1992; 10: 241–248. 17 Takaue Y, Kawano Y, Abe T et al. Collection and transplantation of peripheral blood stem cells in very small children weighing 20 kg or less. Blood 1995; 86: 372–380. References 18 To LB, Shepperd KM, Haylock DN et al. Single high doses of cyclophosphamide enable the collection of high numbers 1 Frei E III. Pharmacologic strategies for high-dose chemo- of hemapoietic stem cells from the peripheral blood. Exp Hem- therapy. In: Armitage J, Antman K (eds). High-dose Cancer atol 1990; 18: 442–447. Therapy. Pharmacology, Hematopoietins, Stem Cells. Willi- 19 Socinski M, Elias A, Schnipper L et al. Granulocyte–macro- ams & Wilkins: Baltimore, 1992, pp 3–13. phage colony stimulating factor expands the circulating 2 Antman KH. Dose-intensive therapy in breast cancer. In: Arm- haemopoietic progenitor cell compartment in man. Lancet itage J, Antman K (eds). High-dose Cancer Therapy. Pharma- 1988; 1: 1194–1198. cology, Hematopoietins, Stem Cells. Williams & Wilkins: Bal- 20 Leibundgut K, von Rohr A, Brulhart K et al. The number timore, 1992, pp 701–718. of circulating CD34+ blood cells predicts the colony-forming 3 Philip T, Guglielmi C, Chauvin F et al. Autologous bone mar- capacity of leukapheresis products in children. Bone Marrow row transplantation (ABMT) versus conventional chemo- Transplant 1995; 15: 25–31. therapy (DHAP) in relapsed non-Hodgkins lymphoma (NHL): 21 Shen V, Bennetts G, Woodbury C et al. The safety and effi- final analysis of the Parma randomized study (216 patients). cacy of repetitive peripheral stem cell support in children Proc Am Soc Clin Oncol 1995; 14: 390 (Abstr.). receiving submyeloablative chemotherapy. Proc Am Assoc 4 Fennelly D, Schneider J. Role of chemotherapy dose intensi- Cancer Res 1994; 35: 240 (Abstr.). fication in the treatment of advanced ovarian cancer. Oncology 22 Haas R, Mohle R, Fruhauf S et al. Patient characteristics asso- 1995; 9: 911–921. ciated with successful mobilizing and autografting of periph- 5 Stram DO, Matthay KK, O’Leary MC et al. Autologous bone eral blood progenitor cells in malignant lymphoma. Blood marrow transplantation vs continued chemotherapy for stage 1994; 83: 3787–3794. IV neuroblastoma: a non-randomized Childrens Cancer Group 23 Dreger P, Kloss M, Petersen B et al. Autologous progenitor study. Proc Am Soc Clin Oncol 1995; 14: 448 (Abstr.). cell transplantation: prior exposure to stem cell-toxic drugs 6 Sheridan WP, Begley CG, To LB et al. Phase II study of determines yield and engraftment of peripheral blood progeni- autologous filgrastim (G-CSF)-mobilized peripheral blood tor cell but not of bone marrow grafts. Blood 1995; 86: progenitor cells to restore hemopoiesis after high-dose chemo- 3970–3978. therapy for lymphoid malignancies. Bone Marrow Transplant 24 Cairo MS, Shen W-P, Miser J et al. A randomized trial of two 1994; 14: 105–111. doses of G-CSF (5.0 vs 10.0 ␮g/kg/d) following ifosfamide, Use of PBSC in children V Shen et al 204 carboplatin, and etoposide (ICE) chemotherapy in children 27 Brugger W, Bross KJ, Glatt M et al. Mobilization of tumor with recurrent solid tumors: significant clinical activity but no cells and hematopoietic progenitor cells into peripheral blood improvement in hematopoietic recovery with increased dose of patients with solid tumors. Blood 1994; 83: 636–640. of G-CSF. Proc Am Soc Clin Oncol 1995; 14: 255 (Abstr.). 28 Moss TJ, Cairo M, Santana VM et al. Clonogenicity of circul- 25 Tepler I, Cannistra SA, Frei E III et al. Use of peripheral- ating neuroblastoma cells: implications regarding peripheral blood progenitor cells abrogates the myelotoxicity of repeti- blood stem cell transplantation. Blood 1994; 83: 3085–3089. tive outpatient high-dose carboplatin and cyclophosphamide 29 Rosenthal J, Sender L, Ellis R et al. A pilot study utilizing chemotherapy. J Clin Oncol 1993; 11: 1583–1591. positive stem cell (CD34+) selection and verapamil to modu- 26 Shea TC, Mason JR, Storniolo AM et al. Sequential cycles of late MDR prior to ABMT in advanced neuroblastoma. Exp high-dose carboplatin administered with recombinant human Hematol 1994; 22: 766 (Abstr.). granulocyte–macrophage colony-stimulating factor and 30 Brugger W, Heimfeld S, Berenson RJ et al. Reconstitution repeated infusions of autologous peripheral-blood progenitor of hematopoiesis after high-dose chemotherapy by autologous cells: a novel and effective method for delivering multiple progenitor cells generated ex vivo. New Engl J Med 1995; 333: courses of dose-intensive therapy. J Clin Oncol 1992; 10: 283–287. 464–473.