Mobilization Topotecan–Filgrastim Combination Is an Effective Regimen

Home , Topotecan

Bone Marrow Transplantation (2001) 28, 563–571  2001 Nature Publishing Group All rights reserved 0268–3369/01 $15.00 www.nature.com/bmt Mobilization Topotecan–ﬁlgrastim combination is an effective regimen for mobilizing peripheral blood stem cells

E-JA Yeoh1,7, JM Cunningham1,5,GCYee6, D Hunt2, JA Houston1, SL Richardson1, CF Stewart3, PJ Houghton4, LC Bowman1,5 and AJ Gajjar1,5

Departments of 1Hematology-Oncology, 2Biostatistics, 3Pharmaceutical Sciences, 4Molecular Pharmacology, St Jude Children’s Research Hospital, Memphis, TN, USA; 5Department of Pediatrics, College of Medicine, University of Tennessee Health Sciences Center, Memphis, TN, USA; 6Department of Pharmacy Practice, College of Pharmacy, University of Nebraska Medical Center, Omaha, NE, USA; and 7Department of Pediatrics, National University of Singapore, Singapore

Summary: malignancies.1–5 Improvements in supportive care and increasing experience with stem-cell rescue after myeloabl- We compared the efficacy, toxicity, and cost of topote- ative chemotherapy have reduced the morbidity of high- can–filgrastim and filgrastim alone for mobilizing dose chemotherapy, and this reduction in morbidity has peripheral blood stem cells (PBSCs) in 24 consecutive made such treatment applicable to a wide spectrum of pediatric patients with newly diagnosed medulloblas- diseases.6 Peripheral blood stem cells (PBSCs) are increas- toma. PBSCs were mobilized with an upfront window of ingly used for rescue because of their engraftment charac- topotecan–filgrastim for 11 high-risk patients (residual teristics and the ease with which they can be collected.7,8 tumor у1.5 cm2 after resection; metastases limited to Stem cells are usually mobilized into the peripheral neuraxis) and with filgrastim alone for 13 average-risk blood in one of three ways: with cytokine therapy alone, patients. All patients subsequently underwent craniospi- with myelosuppressive chemotherapy alone or with a com- nal irradiation and four courses of high-dose chemo- bination of both.9 The present standard of care uses filgras- therapy with stem cell rescue. Target yields of CD34+ tim (G-CSF) or sargramostim (GM-CSF) to mobilize -cells (у8 ؋ 106/kg) were obtained with only one apher- PBSCs, but these agents may not mobilize a sufficient num esis procedure for each of the 11 patients treated with ber of stem cells in as many as 67% of patients.10–12 For topotecan–filgrastim, but with a mean of 2.3 apheresis these patients, alternative mobilization regimens are procedures for only six (46%) of the 13 patients treated needed.10,13 The choice of regimen to enhance mobilization, with filgrastim alone (P = 0.0059). The median peak and particularly for children, is important because of the poten- median total yield of CD34+ cells were six-fold higher tial increased toxicity of chemotherapy-based mobilization for the topotecan–filgrastim group (328/␮l and 21.5 ؋ and because of the need for additional apheresis procedures 106/kg, respectively) than for the filgrastim group (54/␮l and marrow backups required with regimens using cyto- and 3.7 ؋ 106/kg, respectively). Mean times to neutro- kines alone. Therefore, it is imperative to evaluate the phil and platelet engraftment were similar. Myelosup- economic and therapeutic impact of any new mobilizing pression was the only grade 4 toxicity associated with regimen to determine its usefulness.14 topotecan–filgrastim mobilization and lasted a median Topotecan, a new topoisomerase I inhibitor with promi- of 5 days. Compared with filgrastim mobilization, topo- nent activity15 against many malignancies in pediatric and tecan–filgrastim mobilization resulted in a mean cost adult patients, is being evaluated in several new treatment saving of $3966 per patient. Topotecan–filgrastim is an regimens for various tumors.16–19 Our aim was to compare efficacious, minimally toxic, and cost-saving combi- the efficacy, toxicity, and relative costs of intravenously nation for PBSC mobilization. Bone Marrow Transplan- administered topotecan and filgrastim and of filgrastim tation (2001) 28, 563–571. alone in mobilizing PBSCs. This retrospective obser- Keywords: topotecan; peripheral blood stem cell; filgra- vational study was part of our institutional medullo- stim; high-dose chemotherapy; medulloblastoma; children blastoma/primitive neuroectodermal tumor (PNET) protocol.

High-dose chemotherapy is increasingly used as front-line Patients and methods therapy for adult and pediatric patients with high-risk Patients Correspondence: Dr A Gajjar, Department of Hematology-Oncology (Room 6024), St Jude Children’s Research Hospital, 332 North Lauder- Patient eligibility criteria and definitions of average-risk 20 dale, Memphis, TN 38105–2794, USA and high-risk disease have been previously reported. The Received and accepted 18 July 2001 protocol was approved by the Institutional Review Board. Topotecan as a PBSC mobilizer E-JA Yeoh et al 564 Signed informed consent was obtained from parents or anti-CD34 monoclonal antibody (CD34-PE; BD legal guardians, as appropriate, before protocol therapy Biosciences), PerCP-labeled anti-CD45 antibody (CD45- was initiated. PerCP; BD Biosciences), and fluorescein isothiocyanate (FITC)-labeled anti-CD3 antibody (CD3-FITC; DAKO ␮ Study design Corporation, Carpinteria, CA, USA). Briefly, 50 lof diluted whole blood from each patient was mixed with 50 Patients with average-risk disease received intravenously ␮l of Dulbecco’s phosphate-buffered saline (BioWhittaker, (i.v.) administered filgrastim (Neupogen; Amgen, Thousand Walkersville, MD, USA), 1% human serum albumin, 0.1% Oaks, CA, USA) at a dose of 10 ␮g/kg daily21 until the sodium azide, and 10 ␮l of each antiserum (CD45-PerCP, absolute peripheral CD34+ cell count was at least 20/␮l. CD3-FITC, and CD34-PE). A control sample was analyzed PBSC harvesting then began and continued until the target with PE-labeled and FITC-labeled mouse IgG1 and CD45- yield of 8 ϫ 106 CD34+ cells/kg was obtained or until the PerCP. The mixtures were incubated in the dark at room absolute peripheral CD34+ cell count dropped below 20/␮l. temperature for 10 min and erythrocytes were subsequently Because of patient tolerance, the maximum number of aph- lysed with FACSLyse (BD Biosciences). Each sample was eresis procedures per patient was prospectively limited to then washed twice before 105 to 106 cells from each sample three. The patients subsequently underwent craniospinal were analyzed. irradiation over a period of 6 weeks, followed by a 6-week rest period. Patients then received four courses of high-dose Peripheral and marrow stem-cell harvest chemotherapy; each course was followed by reinfusion of at least 2 ϫ 106 autologous CD34+ cells per kg (Figure 1).22 PBSCs were harvested by using COBE Spectra (COBE, Patients with high-risk disease were given two cycles of Lakewood, CO, USA) apheresis machines. The total blood outpatient chemotherapy consisting of topotecan (5.5 volume (each blood volume was assumed to be 75 ml/kg mg/m2 daily as an i.v. infusion for 5 days every 21 days). of body weight) from each patient was processed three The day 1 topotecan dosage was 5.5 mg/m2, and subsequent times. On the day before PBSC harvest (ie the day on which doses were adjusted to maintain the target topotecan lactone the absolute peripheral CD34+ cell count was predicted to area under the plasma concentration-time curve (AUC) be greater than 20/␮l), a 5-French single-lumen central between 120 and 160 ng/ml/h. Twenty-four hours after the venous catheter (Cook, Bloomington, IN, USA) was last dose of topotecan, patients received i.v. filgrastim (10 inserted into the femoral vein of those patients without ␮g/kg per day) until the absolute peripheral CD34+ cell adequate peripheral venous access. The catheter was left in concentration was at least 20/␮l. Daily peripheral CD34+ place for the duration of apheresis and was removed after cell counts began on day 1 of filgrastim administration for the last apheresis procedure of the mobilization cycle. the first four patients in the topotecan–filgrastim mobiliz- Anticoagulation was achieved by adding acid-citrate- ation. Because the initial four patients who received topote- dextrose (ACD); the ratio of blood to ACD was 15:1. Blood can and filgrastim did not have measurable CD34+ cell flow was dependent on the draw line and did not exceed counts before day 4, the CD34+ cell counts for the remain- 15 ml/kg/min. The maximum rate of ACD was 1 ing patients were obtained when the leukocyte count ml/kg/min. No calcium replacement was given unless exceeded 1.0 ϫ 109/l or after day 3 of filgrastim adminis- symptomatic hypocalcemia occurred. tration. Guidelines for harvesting PBSCs were similar to those of the average-risk group. After PBSCs were col- Patient monitoring lected, the patients in the high-risk group were given a second course of topotecan followed by craniospinal During mobilization, complete blood counts (CBCs) and irradiation, a 6-week rest period, and four courses of high- absolute peripheral blood CD34+ cell counts were perfor- dose chemotherapy and PBSC rescue that were identical to med daily. Patients were monitored for signs of toxicity as those given to patients with average-risk disease (Figure 1). defined by Version 2 of National Cancer Institute (NCI) Although the chemotherapy regimen was not myeloabl- common toxicity criteria. Blood chemistry values were ative, PBSC rescue was performed after each course of evaluated twice a week or when clinically appropriate. chemotherapy, so that the dose intensity of the regimen 20 could be maintained. Storage and reinfusion of stem cells In either group, if the target cell yield was not obtained, bone marrow was harvested from the iliac bone or PBSC Stem cells underwent no manipulation and were stored in a harvest was repeated, either before or after radiotherapy. final concentration of 10% dimethyl sulfoxide diluted with The target dose of marrow cells required for each cycle autologous plasma or 5% human serum albumin according of high-dose chemotherapy was 1 ϫ 108 nucleated cells to standard procedures. Stem cells were thawed at the bed- per kg. side, and the unmanipulated cells were transfused by slow intravenous push over a period of 10 to 30 min. Flow cytometric studies Supportive care The CD34+ cell count was measured by using the FACSCa- libur Flow Cytometry System (BD Biosciences, San Jose, After stem-cell infusion patients were discharged and CA, USA) and three-color direct immunofluorescence. The treated in the outpatient setting. Filgrastim (5 ␮g/kg) was following antibodies were used: phycoerythrin (PE)-labeled administered intravenously to all patients 24 h after rein-

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 565 Diagnosis of medulloblastoma with no metastases outside neuraxis

High-risk disease Standard-risk disease Metastases limited to neuraxis No metastases in neuraxis or and Residual tumor >1.5 cm2 Gross total resection or residual tumor <1.5 cm2

Topotecan cycle 1 (5.5 mg/m2 i.v. per Filgrastim (10 m g/kg i.v. per day) day over 4h ´ 5 days, adjusted on basis of AUC) followed by filgrastim (10 m g/kg i.v. per day)

PBSC harvest when absolute CD34+ cell count >20/l Target yield of CD34+ cells >8 ´ 106/k

Topotecan cycle 2

Craniospinal irradiation 3639 Gy Craniospinal irradiation 23.4 Gy 3-D conformal boost to tumor bed 55.8 Gy 3-D conformal boost to posterior fossa 36 Gy Over a period of 6 weeks 3-D conformal boost to tumor bed 55.8 Gy over a period of 6 weeks

Four courses of high-dose chemotherapy with stem-cell rescue

Cisplatin (75 mg/m2) on day 1 Vincristine (1.5 mg/m2) on days 1 and 6 Cyclophosphamide (2 g/m2) on days 3 and 4 PBSCs (2 ´ 106 CD34+ cells per kg) on day 6

Figure 1 Treatment scheme. fusion of peripheral or marrow stem cells. Filgrastim was To determine the topotecan systemic exposure in each given for at least 7 days or until the ANC was at least 2.0 patient, plasma samples were collected before topotecan ϫ 109/l for 2 consecutive days after the expected nadir. infusion and at serial times after the end of the infusion. Blood products were given to maintain normal platelet Topotecan lactone AUC was measured by using an iso- counts (at least 30 ϫ 109/l) and hematocrit concentrations cratic high-performance liquid chromatography assay with (at least 20%). If febrile neutropenia (temperature Ͼ38°C; ﬂuorescence detection.23 A two-compartment model was ANC Ͻ0.5 ϫ 109/l) developed, patients were admitted and ﬁtted to the topotecan lactone plasma and concentration- given broad-spectrum antibiotics intravenously. time data by using a maximum a posteriori Bayesian algor- ithm as implemented in ADAPT II.24 The model parameters Measurement of topotecan lactone AUC and dose for each patient were used to determine the appropriate adjustments topotecan dose to attain the target range. The topotecan dose for each patient was adjusted on the Cost analysis basis of plasma levels obtained during each course to attain We performed a cost-minimization analysis14,25 of the total a single-day topotecan lactone AUC of 140 Ϯ 20 ng/ml/h. cost of high-dose chemotherapy to determine whether the

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 566 higher initial cost of the topotecan–filgrastim mobilization 1996 to September 1999. Ten patients did not qualify for regimen (as compared with filgrastim alone) was offset by this mobilization study: seven had an ECOG performance cost savings brought about by reduced use of resources dur- status of 4 (six had posterior fossa syndrome; one was blind ing the mobilization and collection of PBSCs and the initial after surgery), and three required immediate radiotherapy post-transplantation hospitalization period. A provider (ie (one had spinal metastases and paraparesis and two had to hospital) perspective was used in the analysis. Resource begin radiotherapy to comply with protocol timing). A total units for important health care resources were collected by of 24 patients were enrolled prospectively on the protocol. medical chart review and were converted to costs. Drug The results of follow-up that are reported here include those cost was based on actual acquisition cost in 1999, nor- up to 31 October 1999. malized to an adult surface area of 1.73 m2 to facilitate Eleven patients with high-risk disease were enrolled on comparison with findings from adult studies. The cost of the topotecan–filgrastim arm; 13 patients with average-risk hospitalization and transfusions was based on actual costs disease were enrolled on the filgrastim arm (Table 1). One at St Jude Children’s Research Hospital in 1999. For PBSC patient on the filgrastim arm was subsequently found to and bone marrow harvests, we used the cost estimates from have medulloblastoma cells in the CSF. However, this the time and motion study of Glaspy et al.26 The analysis patient was included on the filgrastim arm because his was made by using an intent-to-treat method for each arm. PBSCs had already been mobilized with filgrastim.

Hematologic endpoints Topotecan levels and dosage The primary end point of the study was the mobilization + ϫ 6 of the target yield of CD34 cells (8 10 /kg). Failure to The median topotecan dose used in this study was 5.0 achieve the target yield significantly affected the adminis- mg/m2 per day (range, 2.0–8.0 mg/m2 per day). These doses tration of high-dose chemotherapy and constituted are higher than the published maximum tolerated dose for primary failure. this schedule (ie 2.0 mg/m2 per day),32 because we perfor- The secondary endpoints were the number of platelet med dosage adjustments for each patient to attain a topote- transfusions required and the time required for the can lactone AUC of 140 Ϯ 20 ng/ml/h. The median AUC engraftment of platelets and neutrophils. The time to plate- was 143.7 ng/ml/h (range, 30.1–367.2 ng/ml/h). let engraftment was defined as the number of days from the day of stem-cell infusion to the first day of a period of 7 consecutive days during which the platelet count was at least 20 ϫ 109/l without platelet support. The time to neu- PBSC mobilization by filgrastim trophil engraftment was defined as the number of days from the day of stem cell infusion to the first of 3 consecutive We attempted to mobilize the PBSCs of the 13 patients days after the chemotherapy-induced nadir during which with average-risk disease by using filgrastim alone; the tar- the ANC was at least 0.5 ϫ 109/l. get yield of CD34+ cells was obtained from only six patients (46%). The target yield was achieved with 1 day of PBSC harvest from only one patient, with 2 days of Statistical methods PBSC harvest from four patients, and with 3 days of PBSC 27 harvest from one patient. PBSCs failed to mobilize in four Fisher’s exact test was used as a test of equality of pro- + portions of patients in the treatment groups whose PBSCs of the seven remaining patients; the peak peripheral CD34 were mobilized. Mixed-model analyses of repeated meas- cell counts of two patients were less than 20/␮l, and the ures28 data estimated the absolute CD34+ cell count during counts of two other patients peaked on study day 4 and rapidly declined to less than 20/␮l on the following day. the period of filgrastim administration. The exact log-rank + test29 was used to determine whether the percentage of The peripheral CD34 cell counts of the other three patients + patients in whom neutrophil and platelet engraftment was were more than 20/␮l, but the target yield of CD34 cells achieved differed after each cycle of high-dose chemo- was not obtained after 2 to 3 days of harvest. Despite therapy. Kaplan–Meier estimates30 in each group (by cycle additional bone marrow and PBSC harvests, the total yield of high-dose chemotherapy) were obtained for the percent- of stem cells from three of these seven patients was insuf- ages of patients in whom neutrophil and platelet ficient. Therefore, the dose of chemotherapy during the last engraftment was achieved. The Wilcoxon rank-sum test31 course for one patient was reduced by 25%, and no stem was used to compare (by cycle) the number of platelet cell rescue was performed for these three patients. transfusions and the number of red blood cell transfusions The median day on which the first PBSCs could be har- between the two treatment groups. vested from the patients on the filgrastim arm (n = 6) was study day 6 (range, days 4 to 6). PBSC harvest was completed within 5 days (range, days 4 to 8). The median peak + Results count of peripheral CD34 cells for patients on the filgrastim arm was 54/␮l (range, 6–69/␮l). The median total yield + ϫ Patient characteristics of CD34 cells from the first PBSC harvest was 3.7 106/kg (range, 0–21.5 ϫ 106/kg), whereas the median yield Thirty-four patients with medulloblastoma were treated at of the second PBSC harvest was 3.2 ϫ 106/kg (range, 0– St Jude Children’s Research Hospital between October 5.0 ϫ 106/kg).

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 567 Table 1 Patient data

All patients Topotecan–ﬁlgrastim group Filgrastim group

Number of patients 24 11 13 Median age (years) 7.1 5.7 8.9 Median weight (kg) 20.6 20.2 21.1 Male:Female 17:7 9:2 8:5 Localized disease; n (%) 16 (67%) 4 (36%) 12 (92%) Residual disease Ͼ1.5 cm2; n (%) 4 (17%) 4 (36%) 0 (0%) Neuraxis-limited metastases; n (%) 8 (33%) 7 (64%) 1 (8%)a aCerebrospinal ﬂuid positive for medulloblastoma cells, but PBSCs had already been mobilized with ﬁlgrastim alone.

PBSC mobilization by topotecan–ﬁlgrastim Table 2 Toxicity associated with topotecan–ﬁlgrastim

In contrast to the low percentage of patients (46%) whose NCI common toxicity criteria Topotecan–filgrastim PBSCs were mobilized by treatment with filgrastim alone, (version 2) (n = 11)a 100% of patients who had high-risk disease and were treated with a combination of topotecan and filgrastim No. of patients Median days experienced successful PBSC mobilization (P = 0.0059). (range) The target yield of CD34+ cells in patients treated with topotecan and filgrastim was achieved with a single apher- Leukopenia, grade 4 5 5 (1–13) р × 9 esis procedure. The median day to apheresis was study day (WBC count, 1 10 /l) Neutropenia, grade 4 7 6 (2–8) 9 (range, day 7 to day 10). The median peak of peripheral р × 9 + (ANC 0.5 10 /l) blood CD34 cells was 328/␮l (range, 54 to 495/␮l). The Anemia grade 3 7a NA mean total yield of CD34+ cells 21.5 ϫ 106/kg (range, (Hb 65–79 g/dl) 12.9–72 ϫ 106/kg), which was six-fold greater than that of Thrombocytopenia, grade 4 1b Ͻ × 12 ϫ 6 (Platelet count 25 10 /l) patients treated with filgrastim alone (3.7 10 /kg). Infection, grade 2 2 Antibiotics 7 4 (3–6) Febrile neutropenia 4 4 (3–5) Engraftment Culture positive 1 3 By the end of October 1999, 13 patients in the filgrastim Admission Yes 7 group and nine patients in the topotecan–filgrastim group No. of days 4 (3–5) had completed a total of 72 courses of high-dose chemo- Fever grade 2 4 3.5 (3–5) therapy with PBSC rescue. PBSCs had been mobilized in (Temp Ͻ40°C) two patients on the topotecan–filgrastim arm, but these Days of filgrastim treatment 11 9 (7–10) patients had not yet undergone stem cell transplantation at = = = the time that this PBSC mobilization study was closed NCI National Cancer Institute; WBC white blood cell; ANC absolute neutrophil count; NA = not applicable; Hb = hemoglobin. for analysis. aFive patients received red blood cell transfusions (range, 1 to 2 U; The median times required for neutrophil engraftment median, 1 U). and for platelet engraftment in patients in the topotecan– bFour patients received platelet transfusions (range, 1 to 4 U; median, 1 U). filgrastim group after each course of chemotherapy were similar to those of patients in the filgrastim group (data not shown). The paper describing the overall feasibility of this protocol has recently been published.20

Toxicity associated with mobilization kopenia was 5 days (range, 1–13 days), whereas that of neutropenia was 6 days (range, 2–8 days). Febrile neutro- The administration of filgrastim was well tolerated without penia developed in four patients, and each of the four were any significant toxicity. Topotecan-induced grade 2 fever treated with intravenously administered antibiotics for a (temperature between 38.1°C and 40°C) occurred in four median of 4 days (range, 3–5 days). The results of blood patients (36%) during infusion (Table 2) and lasted an aver- cultures were positive for only one patient with febrile neu- age of 4 days (range, 3 to 5 days). Results of blood cultures tropenia; in the peripheral blood culture grew a mixture of were negative for all patients during these febrile periods Staphylococcus saprophyticus and Corynebacterium spp., attributed to topotecan administration. which were probably contaminants from the skin. No The only grade 4 toxicity in the topotecan–filgrastim patient treated with topotecan–filgrastim experienced grade group was related to myelosuppression; all episodes were 3 or 4 mucositis or grade 3 or 4 toxicity involving the liver, short and uncomplicated. Five patients (45%) had leukocyte kidney, or central nervous system (CNS). Moreover, we counts lower than 1.0 ϫ 109/l, and seven (64%) had ANCs observed no fatal complications that were related to topote- less than 0.5 ϫ 109/l. The median duration of grade 4 leu- can therapy.

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 568 Cost analysis ences in diagnosis)33,34 that make it difficult to draw valid conclusions about the efficacy of various mobilization regi- The cost of mobilizing PBSCs with topotecan–filgrastim mens. The only difference between the two groups in our was $17818 per patient, whereas that of mobilizing PBSCs study was the presence or absence of neuraxis dissemi- with filgrastim alone was $12710 per patient. The higher nation, which was unlikely to interfere with PBSC cost of topotecan–filgrastim mobilization is mainly attri- mobilization. buted to the high cost of topotecan. This higher cost is off- + The recommended number of CD34 cells for autologous set, however, by the decreased use of resources in the topo- transplantation has been increased from 2 ϫ 106/kg to at tecan–filgrastim group after high-dose chemotherapy and least 5 ϫ 106/kg.35,36 In recent studies33,35 in which the tar- stem-cell infusion: the result was a saving of $9074 per + get total yield of CD34 cells was at least 5 ϫ 106/kg, only patient after the stem cell infusion phase (Table 3). Thus, 33% to 50% of the patients treated with filgrastim alone the use of topotecan–filgrastim resulted in an overall saving for PBSC mobilization achieved the target yield. Similarly, of $3966 per patient. mobilization of stem cells for patients in our filgrastim group was poor (PBSCs were successfully mobilized in only 46% of patients) in part because our study design Discussion + required a high target yield of CD34 cells (at least 8 ϫ 106/kg). This requirement was necessary because stem cell In a homogenous population of chemotherapy-naive rescue was performed after each course of high-dose patients with normal marrow function, we directly com- + chemotherapy (CD34 cell dose given after each course, 2 pared the efficacy of topotecan–filgrastim with that of fil- ϫ 106/kg). This feature highlights the failure of filgrastim grastim alone in mobilizing PBSCs. Findings of many stem alone in mobilizing sufficient stem cells for as many as cell mobilization studies are confounded by various factors 67% of patients11,33,35 and the need of a more efficient (eg extensive prior therapy, marrow disease, and differ- mobilization regimen.10,13

Table 3 Results of cost analysis of resources used

Category Topotecan–ﬁlgrastim Filgrastim Difference ($) Mean period of use Total cost Mean period of use Total cost (days) ($) (days) ($)

First mobilization Topotecan administrationa 5.0 7183 0 0 Filgrastim administrationa,b 3.6 1269 2.4 912 First PBSC harvestc 1 6200 0.7 4340 Subsequent days of ﬁrst PBSC harvest 0 0 0.8 1600 Subtotal 14652 6852 +7800 Toxicity of mobilization Admissiona 2.8 2167 0 0 Antibioticsa 2.3 370 Blood productsa Red blood cell transfusion 0.5 120 Platelet transfusion 0.5 509 Subtotal 3166 +3166 Repeat PBSC harvestc Subsequent PBSC harvest 0 0 0.2 1240 Subsequent days of second PBSC harvest 0.1 200 Additional ﬁlgrastim used 0.8 288 Bone marrow harvestc 0 0 0.7 4130 Subtotal 5858 −5858 Hospitalization costs after high-dose chemotherapy Hospital days 13.4 11 508 17.2 14 750 Antibiotic therapy 11.8 1896 16.4 2638 Blood products Red blood cell transfusion 6.7 1600 9.3 2234 Platelet transfusion 8.4 8596 12.0 12 216 Filgrastim usage 55.8 19 634 58.2 20 470 Subtotal 43 234 52 308 −9074 Total 61 052 65 180 −3966

aWholesale price for medication or hospital bed charges at St Jude Children’s Research Hospital in 1999. bFilgrastim use beyond the ﬁve doses allocated in each cycle of PBSC harvest. cGlaspy et al.26 Bone Marrow Transplant 1999; 23 (Suppl. 2): S21–S27.

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 569 We used a higher dose of topotecan (median dose, 5.0 contribute to an increased risk of relapse after autologous mg/m2 per day) than has been previously reported in the stem cell transplantation.43 In peripheral blood, stem cells literature.15,16 We chose to use this pharmacokinetically are mobilized more rapidly (days 1 and 2 of recovery) than guided dosing approach for topotecan on the basis of our tumor cells (days 5 and 6).44 Kahn et al45 reported that in preclinical37 and clinical data.23 This topotecan dosage and a group of patients with breast cancer, 5.4% of PBSC pro- the resultant high level of systemic exposure to topotecan ducts obtained during the first apheresis procedure were lactone were remarkably well tolerated with acceptable contaminated with tumor cells; during the second apheresis myelotoxicity. Determining whether a lower dose of topote- procedure, 15.4% were contaminated; and during the fourth can is equally efficacious and reduces costs will require and subsequent procedures, 42% were contaminated. Thus, further study. it is postulated that to reduce the risk of contamination with Although chemotherapy-based mobilization regimens are tumor cells, it is crucial to obtain sufficient CD34+ cells often complicated by infection,9 the topotecan–filgrastim from early apheresis procedures alone. This criterion combination was well tolerated with manageable toxicity. appeared to be met when we used topotecan–filgrastim as a The only grade 4 toxicity was related to myelosuppression: mobilizing agent. Thus, topotecan–filgrastim administration the ANCs of 64% of our patients were less than 0.5 ϫ 109/l may result in grafts with less contamination by tumor cells. (median duration, 6 days). Febrile neutropenia occurred in Reduced hospitalization stay and the consequent use of only four patients (36%). This finding compares favorably fewer resources offset the additional cost of adding topote- with those of studies of high-dose cyclophosphamide38 and can to filgrastim mobilization. The mean saving for the of other chemotherapy-filgrastim mobilization regimens;39–41 topotecan–filgrastim group was $3966 per patient. This in those reports rates of febrile neutropenia ranged from finding supports the cost saving of the topotecan–filgrastim 20% to 100%. combination in PBSC mobilization, at least in the context Scheduling the optimal time for PBSC harvest is difficult of tandem transplantation. because the time at which the number of CD34+ cells peaks On the basis of our findings, we conclude that the combi- after chemotherapy-based mobilization is unpredictable.9 nation of topotecan and filgrastim is an efficacious, mini- This lack of predictability is especially problematic in busy mally toxic and cost-saving method for mobilizing PBSCs centers in which apheresis machines are used both for stem in pediatric patients with medulloblastoma enrolled on our cell mobilization and platelet apheresis. The use of the trial in which stem-cell rescue was required after each topotecan–filgrastim combination resulted in a predictable course of chemotherapy. This combination may warrant pattern of PBSC mobilization. Time to mobilization in the testing in patients with other disease entities. topotecan–filgrastim-treated patients (study days 7 to 10) was as predictable as that in patients treated with filgrastim alone (study days 4 to 6). Planning was easier because one Acknowledgements harvest was required to obtain target yields; in contrast, a mean of 2.3 apheresis procedures and bone marrow har- This work was supported by Cancer Center Support (CORE) grant vests were needed for those whose PBSCs failed to mobil- P30 CA 21765 and grant P01 CA 23099 from the National Cancer Institute; by grant NMRC 0082/95 from the National Medical ize after treatment with filgrastim alone. This finding Research Council (Singapore); and by the American Lebanese reflects the general clinical experience that mobilization Syrian Associated Charities (ALSAC). We thank Dr Julia Cay with filgrastim alone results in the need for multiple apher- Jones for editorial consultation and Patsy Burnside for typing the esis procedures and additional bone marrow harvesting to manuscript. We also acknowledge the National Medical Research obtain sufficient numbers of stem cells.13 The increasing Council of Singapore for the funding of Dr Yeoh at St Jude Chil- demand for greater numbers of stem cells for autologous dren’s Research Hospital. transplantation may require the development of newer methods for enhancing PBSC mobilization, such as myelosuppressive therapy with filgrastim, as in our study, or treat- References ment with a combination of cytokines. Stem cells mobilized by the topotecan–filgrastim combi- 1 Matthay K, Villablanca JG, Seeger RC et al. Treatment of high-risk neuroblastoma with intensive chemotherapy, radio- nation rapidly restored marrow function and allowed high- therapy, autologous bone marrow transplantation, and 13-cis- dose chemotherapy to be administered as scheduled. retinoic acid. Children’s Cancer Group. New Engl J Med Although the patients in the topotecan–filgrastim group 1999; 341: 1165–1173. received higher doses of spinal irradiation (36 to 39 Gy) 2 Motzer RJ, Mazumdar M, Bajorin DF et al. High-dose car- than those in the filgrastim group (23.4 Gy) before high- boplatin, etoposide, and cyclophosphamide with autologous dose chemotherapy, the mean numbers of transfusions of bone marrow transplantation in first-line therapy for patients platelets and red blood cells per patient were similar for with poor-risk germ cell tumors. J Clin Oncol 1997; 15: both groups, and platelet and neutrophil engraftment also 2546–2552. occurred at approximately the same time in both groups. 3 Legros M, Dauplat J, Fleury J et al. High-dose chemotherapy Our findings are supported by those of Ferari et al who with hematopoietic rescue in patients with stage III to IV ovarian cancer: long-term results. J Clin Oncol 1997; 15: showed that topotecan followed by filgrastim and erythro- 1302–1308. poietin treatment mobilized high-quality stem cells as docu- 4 Rodenhuis S, Richel DJ, van der Wall E et al. Randomised mented by flow-cytometric parameters with low apoptotic trial of high-dose chemotherapy and haematopoietic progeni- features.42 tor-cell support in operable breast cancer with extensive axil- Contamination of the graft by tumor cells is thought to lary lymph-node involvement. Lancet 1999; 352: 515–521.

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 570 5 Gianni AM, Bregni M, Siena S et al. High-dose chemotherapy 23 Stewart CF, Baker SD, Heideman RL et al. Clinical pharmac- and autologous bone marrow transplantation compared with odynamics of continuous infusion topotecan in children: sys- MACOP-B in aggressive B-cell lymphoma. New Engl J Med temic exposure predicts hematologic toxicity. J Clin Oncol 1997; 336: 1290–1297. 1994; 12: 1946–1954. 6 Blume KG, Thomas ED. A review of autologous hematopo- 24 D’Argenio DZ, Schumitzky A. A program package for simul- ietic cell transplantation. Biol Blood Marrow Transplant 2000; ation and parameter estimation in pharmacokinetic systems. 6:1–12. Comput Programs Biomed 1979; 9: 115–134. 7 Shea TC. Introduction: current issues in high-dose chemo- 25 Drummond ME, O’Brien B, Stoddart GL, Torrance GW. therapy and stem cell support. Bone Marrow Transplant 1999; Methods for the Economic Evaluation of Health Care Pro- 23 (Suppl. 2): S1-S5. grammes, 2nd edn. Oxford Medical Pub: New York, 1997. 8 Schmitz N, Linch DC, Dreger P et al. Randomized trial of 26 Glaspy JA. Economic considerations in the use of peripheral filgrastim-mobilised peripheral blood progenitor cell trans- blood progenitor cells to support high-dose chemotherapy. plantation versus autologous bone-marrow transplantation in Bone Marrow Transplant 1999; 23 (Suppl. 2): S21–S27. lymphoma patients. Lancet 1996; 347: 353–357. 27 Fisher RA. Statistical Methods for Research Workers. Oliver 9 To LB, Haylock DN, Simmons PJ, Juttner CA. The biology and Boyd: Edinburgh, 1925. and clinical uses of blood stem cells. Blood 1997; 89: 28 Laird NM, Ware JH. Random-effects models for longitudinal 2233–2258. data. Biometrics 1982; 38: 963–974. 10 Shpall EJ. The utilization of cytokines in stem cell mobiliz- 29 Mantel N, Haenszel W. Statistical aspects of the analysis of ation strategies. Bone Marrow Transplant 1999; 23 (Suppl. 2): data from retrospective studies of disease. J Natl Cancer Inst S13–S19. 1959; 22: 719–748. 11 Somlo G, Sniecinski I, ter Veer A et al. Recombinant human 30 Kaplan EL, Meier P. Nonparametric estimation from incom- thrombopoietin in combination with granulocyte colony-sti- plete observations. J Am Stat Assoc 1958; 53: 457–481. mulating factor enhances mobilization of peripheral blood pro- 31 Wilcoxon F. Individual comparisons by ranking methods. Bio- genitor cells, increases peripheral blood platelet concentration, metrics 1945; 1:80–83. and accelerates hematopoietic recovery following high-dose 32 Tubergen DG, Stewart CF, Pratt CB et al. Phase I trial and chemotherapy. Blood 1999; 93: 2798–2806. pharmacokinetic (PK) and pharmacodynamics (PD) study of 12 Akard LP, Thompson JM, Dugan MJ et al. Matched-pair topotecan using a five-day course in children with refractory analysis of hematopoietic progenitor cell mobilization using solid tumors: a Pediatric Oncology Group study. J Pediatr G-CSF vs cyclophosphamide, etoposide, and G-CSF: Hematol Oncol 1996; 18: 352–361. enhanced CD34+ cell collections are not necessarily cost- 33 Bensinger W, Appelbaum F, Rowley S et al. Factors that effective. Biol Blood Marrow Transplant 1999; 5: 379–385. influence collection and engraftment of autologous peripheral- 13 Akard L. Optimum methods to mobilize stem cells. J Clin blood stem cells. J Clin Oncol 1995; 13: 2547–2555. Oncol 2000; 18: 3063 (letter). 34 Haas R, Mo¨hle R, Fru¨hauf S et al. Patient characteristics asso- 14 Waters TM, Bennett CL, Vose JM. Economic analyses of new ciated with successful mobilizing and autografting of periph- technologies: the case of stem-cell transplantation (editorial). eral blood progenitor cells in malignant lymphoma. Blood J Clin Oncol 1997; 15:2–4. 1994; 83: 3787–3794. 15 Rodriguez-Galindo C, Radomski K, Stewart CF et al. Clinical 35 Glaspy JA, Shpall EJ, LeMaistre CF et al. Peripheral blood use of topoisomerase I inhibitors in anticancer treatment. Med progenitor cell mobilization using stem cell factor in combi- Pediatr Oncol 2000; 35: 385–402. nation with filgrastim in breast cancer patients. Blood 1997; 16 Nitschke R, Parkhurst J, Sullivan J et al. Topotecan in pedi- 90: 2939–2951. atric patients with recurrent and progressive solid tumors: a 36 Siena S, Schiavo R, Pedrazzoli P et al. Therapeutic relevance Pediatric Oncology Group phase II study. J Pediatr Hematol of CD34 cell dose in blood cell transplantation for cancer ther- Oncol 1998; 20: 315–318. apy. J Clin Oncol 2000; 18: 1360–1377. 17 ten Bokkel Huinink W, Gore M, Carmichael J et al. Topotecan 37 Friedman HS, Houghton PJ, Schold SC et al. Activity of 9- versus paclitaxel for the treatment of recurrent epithelial dimethylaminomethyl-10-hydroxycamptothecin against pedi- ovarian cancer. J Clin Oncol 1997; 15: 2183–2193. atric and adult central nervous system tumor xenografts. 18 von Pawel J, Schiller JH, Shepherd FA et al. Topotecan versus Cancer Chemother Pharmacol 1994; 34: 171–174. cyclophosphamide, doxorubicin, and vincristine for the treat- 38 Rowlings PA, Bayly JL, Rawling CM et al. A comparison of ment of recurrent small-cell lung cancer. J Clin Oncol 1999; peripheral blood stem cell mobilisation after chemotherapy 17: 658–667. with cyclophosphamide as a single agent in doses of 4 g/m2 19 Chang AY. The potential role of topotecan in the treatment or 7 g/m2 in patients with advanced cancer. Aust NZJ Med of advanced breast cancer. Semin Oncol 1997; 24 (6 Suppl. 1992; 22: 660–664. 20): S49–S54. 39 Copelan EA, Ceselski SK, Ezzone SA et al. Mobilization of 20 Strother D, Ashley D, Kellie SJ et al. Feasibility of four con- peripheral-blood progenitor cells with high-dose etoposide and secutive high-dose chemotherapy cycles with stem-cell rescue granulocyte colony-stimulating factor in patients with breast for patients with newly diagnosed medulloblastoma or suprat- cancer, non-Hodgkin’s lymphoma, and Hodgkin’s disease. J entorial primitive neuroectodermal tumor after craniospinal Clin Oncol 1997; 15: 759–765. radiotherapy: results of a collaborative study. J Clin Oncol 40 Kroger N, Zeller W, Fehse N et al. Mobilizing peripheral 2001; 19: 2696–2704. blood stem cells with high-dose G-CSF alone is as effective 21 Grigg AP, Roberts AW, Raunow H et al. Optimizing dose as with Dexa-BEAM plus G-CSF in lymphoma patients.BrJ and scheduling of filgrastim (granulocyte colony-stimulating Haematol 1998; 102: 1101–1106. factor) for mobilization and collection of peripheral blood pro- 41 Weaver CH, Schwartzberg LS, Zhen B et al. Mobilization of genitor cells in normal volunteers. Blood 1995; 86: 4437– peripheral blood stem cells with docetaxel and cyclophos- 4445. phamide (CY) in patients with metastatic breast cancer: a ran- 22 Bender JG, To LB, Williams S et al.Defining a therapeutic domized trial of 3 vs 4 g/m2 of CY. Bone Marrow Transplant dose of peripheral blood stem cells. J Hematother 1992; 1: 1999; 23: 421–425. 329–341. 42 Ferari S, Danova M, Porta C et al. Circulating progenitor cell

Bone Marrow Transplantation Topotecan as a PBSC mobilizer E-JA Yeoh et al 571 release and functional characterization after topotecan plus G- tiple myeloma after treatment with cyclophosphamide and gra- CSF and erythropoietin in small cell lung cancer patients. Int nulocyte–macrophage colony-stimulating factor. Blood 1996; J Oncol 1999; 15: 811–815. 87: 805–811. 43 Brenner MK, Rill DR, Moen RC et al. Gene-marking to trace 45 Kahn DG, Prilutskaya M, Cooper B et al. The relationships origin of relapse after autologous bone-marrow transplan- between the incidence of tumor contamination and number of tation. Lancet 1993; 341:85–86. phereses for stage IV breast cancer. Blood 1997; 90 (Suppl.): 44 Gazitt Y, Tian E, Barlogie B et al. Differential mobilization 565 (Abstr. 2514). of myeloma cells and normal hematopoietic stem cells in mul-

Bone Marrow Transplantation