Bone Marrow Transplantation (2000) 25, 1137–1140  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt High-dose ifosfamide and with filgrastim for stem cell mobilization in patients with advanced ovarian

ML Donato1, D Gershenson2, C Ippoliti3, JT Wharton2, RC Bast Jr4, A Aleman1, P Anderlini1, JG Gajewski1, S Giralt1, J Molldrem1, N Ueno1, J Lauppe1, M Korbling1, J Boyer1, D Bodurka-Bevers2, M Bevers2, T Burke2, R Freedman2, C Levenback2, J Wolf2 and RE Champlin1

Departments of 1Blood and Marrow Transplantation, 2Gynecologic Oncology, 4Clinical Investigation and 3Division of Pharmacy, The University of Texas MD Anderson Cancer Center, Houston, Texas, USA

Summary: 33% 5-year survival rate with this treatment approach. Transplantation of autologous PBSC is generally associated High-dose combined with autologous with more rapid hematological recovery than transplan- peripheral blood stem cell transplantation has shown tation of bone marrow.3–5 Many studies have indicated that promise as treatment for recurrent or persistent epi- hematopoietic recovery is related to the number of CD34+ thelial . We evaluated the stem cell mobi- cells infused and suggest a target of 2–5 × 106 CD34+ lization regimen of high-dose ifosfamide plus etoposide cells/kg body weight for a single myeloablative intensifi- in 32 patients with epithelial ovarian cancer, who had cation cycle.6,7 a positive second-look laparatomy or recurrent disease. PBSCs are commonly harvested during cytokine treat- Ifosfamide was given at 10 g/m2 by continuous i.v. from ment6 at the time of recovery of peripheral blood counts days 1 to 3. Etoposide was given at 150 mg/m2 every from chemotherapy.8 The ideal stem cell mobilization regi- 12 h for six doses on days 1–3. Filgrastim was given at men allows successful stem cell harvest with a minimum 10 ␮g/kg/d s.c. from day 5 until the completion of per- number of aphereses, is well tolerated, and produces effec- ipheral blood stem cell harvest. Fourteen of 32 patients tive antitumor activity. , in doses rang- had measurable or evaluable disease before mobiliz- ing from 3 to 7 g/m2 is the agent most commonly used for ation therapy and were assessed for response. In nine stem cell mobilization.9–11 In recent years, investigators (64%) of the 14 patients, treatment response was dem- have combined cyclophosphamide with with onstrated, and these patients received a second cycle of promising results.12 In this study, we evaluated a regimen mobilization therapy. The target CD34+ cell dose of high-dose ifosfamide plus etoposide with filgrastim for (Ͼ8 × 106 cells/kg) was achieved with a median of one PBSC mobilization in patients with advanced ovarian can- apheresis (range 1–5). A median of 25.1 (range 8.0– cer. The aim was to assess the stem cell yield, hematologic 122.5) × 106 CD34+ cells/kg body weight was collected. and non-hematologic toxicities, and overall antitumor Non-hematologic toxicity was limited to grade 2 renal activity produced by this regimen. dysfunction in one patient and grade 2 hepatic dysfunc- tion in three patients. In this patient group, high-dose ifosfamide plus etoposide with filgrastim support was Patients and methods well tolerated, lead to successful stem cell harvest and had antitumor activity. Bone Marrow Transplantation Patients (2000) 25, 1137–1140. Between October 1997 and March 1999, we studied 32 Keywords: ovarian cancer; stem cell collection; ifosfamide; etoposide patients with epithelial ovarian cancer who were under- going stem cell mobilization and apheresis (Table 1). Most of the patients were enrolled in a phase I trial of high-dose

Patients with recurrent or persistent epithelial ovarian can- Table 1 Patient demographics (n = 32) cer have a poor prognosis and there are few long-term sur- vivors. In recent studies,1,2 treatment with high-dose Age median 49 years (range 28–64 chemotherapy and peripheral blood stem cell (PBSC) or years) bone marrow transplantation has shown promising results Prior chemotherapy regimens median 2 (range 1–4) 2 Disease status prior to in selected patients. In a study by Stiff et al, patients with ifosfamide/etoposide recurrent, platinum-sensitive disease had approximately a Progressive disease 10 Stable disease 3 Partial remission 8 Correspondence: Dr ML Donato, Department of Blood and Marrow Clinical complete remission 11 Transplantation, Box 24, The University of Texas MD Anderson Cancer Number of patients with prior 2 Center, 1515 Holcombe Blvd, Houston, Texas, 77030, USA PBSC transplants Received 14 October 1999; accepted 22 February 2000 Ifosfamide/etoposide regimen for PBSC mobilization ML Donato et al 1138 chemotherapy and autologous stem cell transplantation for a day. Patients without measurable or evaluable disease, the treatment of ovarian cancer. Patients were eligible for including those with clinical CR at trial enrollment, or our study if they had invasive epithelial ovarian cancer with patients who did not demonstrate at least a PR to recurrent disease or a positive second-look laparotomy. ifosfamide/etoposide went on to receive the high-dose Prior to enrollment in the study, all patients with recurrent chemotherapy arm of the overall treatment plan. Patients disease were reinduced with standard-dose chemotherapy whose disease progressed were taken off the study. until maximal response was achieved. Response to reinduc- tion chemotherapy was not required. However, in all patients, the largest tumor mass could not exceed 2 cm. Fol- lowing standard-dose salvage chemotherapy or second-look Stem cell collection and CD34+ cell–dose analysis surgery and immediately prior to initiation of the ifosfamide/etoposide regimen, 10 patients had progressive For all patients, apheresis was performed using the COBE disease, three had stable disease, eight were in partial Spectra Version 4.7 cell separator (COBE BCT, Lakewood, remission, and 11 were in complete clinical remission (CR). CO, USA). If the patients platelet count was Ͻ30 × 109/l, Of the 11 patients in clinical CR, six had had a positive twice their blood volume was processed. Venous access second-look laparatomy and five were in second or greater was established via a Mahurkar dual-lumen hemodialysis CR following standard-dose salvage chemotherapy. Two of catheter (Quinton Instrument, Bothell, WA, USA) and anti- the 32 patients had received prior stem cell transplantation. coagulant citrate dextrose (ACD-A) solution was used to The median number of prior chemotherapy regimens per prevent clotting. The CD34+ cell content of the collection patient was two (range 1–4). All patients had received at product was measured immediately after apheresis. The least six cycles of platinum and paclitaxel. cells were subsequently cryopreserved using programmed freezing. The apheresis products and the CD34+ cell con- Treatment plan centration were analyzed by two-color flow cytometry. A Hewlett-Packard 340 computer system and LYSYS-II For PBSC mobilization and pretransplant cytoreduction, patients received ifosfamide at 10 g/m2 (total dose) by con- software (Becton Dickinson Immunocytometry, San Jose, tinuous i.v. over 72 h (days 1–3) and etoposide at CA, USA) were used for acquisition of total events and 2 analysis of the list-mode data. From a dot plot of forward 150 mg/m in 2 h i.v. infusions every 12 h during the same ° 72 h period (for a total of six doses). Filgrastim was given and 90 light scatter, a gate was established to include all at 5 ␮g/kg twice a day s.c. from day 5 until completion of lymphocytes and monocytes, excluding granulocytes. 2 Quadrant statistics were obtained from a two-axis dot plot stem cell harvest. was given at 2 g/m by i.v. + + infusion before and at the completion of the ifosfamide. (CD34 and CDw90) to obtain the percentage of CD34 2 cells within the gated population. The absolute number of Mesna also was given at 8 g/m by i.v. infusion concomi- + tant with the ifosfamide infusion (Table 2). Patients were CD34 cells was obtained by multiplying the number of also given oral prophylactic antibiotics consisting of peni- total nucleated cells by the percentage of gated events and cillin, norfloxacin, valacyclovir, and fluconazole starting on the percentage of positive events within the gate. day 5. Patients were hospitalized during the administration of the mobilization regimen and discharged after the final dose of mesna. Patients who demonstrated a clinical response to Statistical methods and definition of response this treatment could receive a second cycle of ifosfamide/etoposide followed by filgrastim 5 ␮g/kg once All analyses were performed using Statistica for Windows Release 5.1 (StatSoft, Tulsa, OK, USA). Patients were con- sidered to have measurable disease if they had tumor that Table 2 Mobilization regimena could be measured in two dimensions by imaging scans or physical examination. Patients were considered to have Drug Dose Administration Duration evaluable disease if they had no measurable disease on

2 scans or physical examination but had an elevated serum Ifosfamide 10 g/m (total continuous days 1–3 Ͼ dose) infusion CA-125 level ( 40 U/ml) in two samples. Response to Etoposide 150 mg/m2 × 6 2 h infusion, days 1–3 treatment for patients with measurable disease was deter- every 12 h mined according to the World Health Organization G-CSF 5 ␮g/kg subcutaneously, day 5 until 13 every 12 h completion of stem response criteria. Response to treatment for patients with cell harvest evaluable disease was based on CA-125 serum levels, as Mesna 2 g/m2 intravenously before first dose and described by Rustin et al14 and Bridgewater et al15: (1) A after final dose of 50% response is declared if there is a 50% decrease in CA- ifosfamide 8 g/m2 (total continuous days 1–3 125 serum levels. There must be two initial samples, and dose) infusionb the sample showing a 50% decrease must be confirmed by a fourth sample. (2) A 75% response is declared if there aHydration during ifosfamide included D5W (2000 ml) with sodium has been a serial decrease of Ͼ75% in CA-125 serum levels acetate (300 mEq) by continuous infusion over 24 h × 3. over three samples. bMesna 8 g/m2 was given concomitant to ifosfamide.

Bone Marrow Transplantation Ifosfamide/etoposide regimen for PBSC mobilization ML Donato et al 1139 Results graphy or magnetic resonance imaging of the abdomen and pelvis, and by physical examination. Response was docu- Patients mented in nine (64%) of the 14 patients. Three achieved a clinical CR (21.4%), four achieved a partial response A total of 32 patients was enrolled in this study, and a total (28.5%), and two had a Ͼ75% response based on CA-125 of 41 cycles of ifosfamide/etoposide was given. All patients criteria (both patients had normalization of serum CA-125 were assessed for stem cell yield and regimen toxicity. levels). Seven of the nine responders had been refractory to Fourteen patients had measurable (12 patients) or evaluable standard-dose reinduction chemotherapy. Of the five non- (two patients) disease and, thus, were assessed for response. responders, four had stable disease or minor response Thirty-one of the 32 patients went on to receive high-dose (28.5%), and in one patient, the disease progressed (7%). chemotherapy with stem cell infusion as planned. Only one patient had disease progression after mobilization therapy; this patient was taken off the study. Discussion

Apheresis data Ifosfamide, a cyclophosphamide analogue, has been pre- viously evaluated at modest doses as treatment for patients Stem cell collection was initiated a median of 14 days with epithelial ovarian cancer. As a single agent in the sal- (range 11–21 days) from the start of chemotherapy vage setting, ifosfamide is associated with a response rate (Table 3). A median of 1 day (range 1–5 days) was needed 17,18 + у × 6 of 10–20%. In other tumor types such as and to reach the target CD34 cell count of 8 10 /kg, sarcoma, ifosfamide has been shown to have a significant enough for two intensification cycles. Pre-apheresis per- dose–response relationship. The experience with high-dose iphal blood CD34+ cell count was used to help guide the × 6 ifosfamide comes mostly from studies done in patients with start of apheresis. A median of 25.1 (range 8.0–122.5) 10 soft tissue or bone sarcomas where doses of up to 14 g/m2 CD34+ cells/kg body weight was collected, with a mean of with filgrastim have been used successfully without the 16.0 × 106 CD34+ cells/kg per apheresis. need for PBSC support.19 Etoposide is a glycosidic derivative of epidophyllotoxin Toxicity data previously shown to be associated with a dose-related perturbation in ovarian cancer cell lines.20 Prolonged The ifosfamide/etoposide regimen was well tolerated. exposure to oral etoposide has demonstrated significant Using the National Cancer Institute Common Toxicity cri- activity.21 Intravenous etoposide has been studied as a sin- teria,16 grade 2 nausea/vomiting was observed in three gle agent with response rates lower than 25%. These studies patients. Grade 2 nephrotoxicity presenting as a transient used overall low doses of etoposide: 300 mg/m2 was given rise in creatinine level was seen in one patient. Three in the largest Gynecologic Oncology Group study.22–24 patients had a transient rise in their liver enzymes (grade Given the potential dose–response relationship for both 2). There was no neurotoxicity. Nine (22%) episodes of ifosfamide and etoposide, we evaluated the antitumor effect neutropenic fever occurred during the 41 cycles of mobiliz- that occurs when these drugs are combined and used at their ation therapy. The median time to recovery of an absolute Ͼ × 9 maximal tolerated doses with growth factor support. nucleated cell count of 1 10 /l was 13 days (range 11– A few studies have demonstrated the feasibility of har- 26 days). The median time to recovery of platelet counts Ͼ × 9 Ͼ × 9 vesting PBSCs upon recovery from an ifosfamide-based of 20 10 /l and 50 10 /l was 12 days (range 0–21 regimen. Baars et al25 have demonstrated that standard-dose days) and 15 days (range 0–37 days), respectively. A ifosfamide (3–4 g/m2) in combination with etoposide can median of 0 (range 0–3) platelet transfusions was required. be successfully used as a mobilization regimen in patients A median of 0 (range 0–2) PBSC transfusions was required with lymphoma or solid tumors. Our aims in the present (Table 3). Thirty-seven percent of the cycles required study were to cytoreduce patients prior to high-dose chemo- patients to receive packed red blood cell transfusions and, therapy and to successfully harvest PBSC. The combination in 44% of the cycles, platelet transfusions were required. of ifosfamide 10 g/m2 and etoposide 900 mg/m2 was asso- ciated with mild non-hematologic toxicity. Only one transi- Tumor response ent renal toxicity was observed in the group studied, and no neurotoxicity was observed. Myelosuppression was sig- Fourteen (44%) of the 32 patients had measurable or evalu- nificant, as expected with a mobilization regimen. How- able disease and, thus, were assessed for response. Patients ever, most of the patients did not require blood product were assessed by serum CA-125 level, by computed tomo- transfusions and were ready for PBSC harvest within 2 weeks of mobilization chemotherapy. All patients achieved Table 3 Apheresis data the targeted stem cell dose, with a median of one apheresis per patient. Number of aphereses required median 1 (range 1–5) The overall response rate of 64% in mostly platinum- Time from start of mobilization median 14 days (range 11–21 days) refractory patients compares favorably with other second- chemo to apheresis Number of CD34+ cells/kg median 25.1 × 106 (range 8–122 line regimens and suggests a dose–response relationship for × 106) these drugs. Commonly used second-line chemotherapeutic CD34+ cells/kg/day of collection mean 16 × 106 agents such as liposomal doxorubicin26 or topotecan27 offer expected responses of 13–26%.

Bone Marrow Transplantation Ifosfamide/etoposide regimen for PBSC mobilization ML Donato et al 1140 In conclusion, high-dose ifosfamide plus etoposide with cell collections after paclitaxel, cyclophosphamide, and filgrastim was associated with acceptable non-hematologic recombinant human granulocyte colony-stimulating factor in toxicity and led to successful PBSC harvest in the patients patients with breast and ovarian cancer. J Clin Oncol 1995: studied. For patients in whom an aggressive treatment 3: 1714–1719. approach is sought, the ifosfamide/etoposide regimen might 13 Miller AB, Hoogstraten B, Straquet M et al. Reporting results of cancer treatment. Cancer 1981; 47: 207–214. be an effective alternative to single-agent salvage therapy. 14 Rustin GJ, Nestrop AE, McClean P et al. Defining response However, further evaluation of this treatment option of ovarian carcinoma to initial chemotherapy according to through controlled trials is needed. serum CA-125. J Clin Oncol 1996; 14: 1545–1551. 15 Bridgewater JA, Nelstrop AE, Rustin GJ et al. Comparison of standard and CA-125 response criteria in patients with epi- References thelial ovarian cancer treatedwith platinum or paclitaxel. J Clin Oncol 1999; 17: 501–508. 1 Legros M, Dauplat J, Fleury H. High-dose chemotherapy with 16 National Cancer Institute. Guidelines for Reporting of Adverse hematopoietic rescue in patients with stage III and IV ovarian Drug Reactions. Division of Cancer Treatment, National Can- cancer: long term results. J Clin Oncol 1997; 15: 1302–1308. cer Institute: Bethesda, MD, 1988. 2 Stiff PJ, Bayer R, Kerger C. High-dose chemotherapy with 17 Sørensen P, Pfeiffer P, Bertelsen K. A phase II trial of autologous transplantation for persistent/relapsed ovarian can- ifosfamide/mesna as salvage therapy in patients with ovarian cer: a multivariate analysis of survival for 100 consecutively cancer refractory to or relapsing after prior platinum-contain- treated patients. J Clin Oncol 1997; 15: 1291–1293. ing chemotherapy. Gynecol Oncol 1995: 56: 75–78. 3 Indovina A, Majolino I, Buscemi F. Engraftment kinetics and 18 Markman M, Hakes T, Reichman B. Ifosfamide and mesna in long term stability of hematopoiesis following autografting of previously treated advanced epithelial ovarian cancer: activity peripheral blood stem cells. Haemotologica 1995; 80: 115– in platinum-resistant disease. J Clin Oncol 1989; 7: 1672– 122. 1676. 4 Langenmayer I, Weaver C, Buckner CD. Engraftment of 19 Benjamin RS, Legha SS, Patel SR: Single-agent ifosfamide patients with lymphoid malignancies transplanted with auto- studies in sarcomas of soft tissue and bone: the MD Anderson logous bone marrow, peripheral blood stem cells or both. Bone experience. Cancer Chem Pharmacol 1993; 31(Suppl. 2): Marrow Transplant 1995; 15: 241–246. 174–179. 5 To LB, Roberts MM, Haylock DN. Comparison of hematolog- 20 Ohishi Y, Fujiwara K, Kohno I. Effect of the exposure dose ical recovery times and supportive care requirements of auto- of etoposide on the cell growth and cell kinetics of human logous recovery phase peripheral blood transplants, autolog- ovarian cancer cells. Cancer Chemother Pharmacol 1996; 38: ous bone marrow transplants and allogeneic bone marrow 141–148. transplants. Bone Marrow Transplant 1992: 9: 277–284. 21 Rose PG, Blessing JA, Mayer AR. Prolonged oral etoposide 6 van der Wall E, Richel DJ, Holtkamp MJ. Bone marrow as second-line therapy for platinum-resistant and platinum- reconstitution after high-dose chemotherapy and autologous sensitive ovarian carcinoma: a Gynecologic Oncology Group peripheral blood progenitor cell transplantation: effect of graft study. J Clin Oncol 1998; 16: 405–410. size. Ann Oncol 1994: 5: 795–802. 22 Slayton RE, Creasman WT, Pretty W. Phase II trial of VP- 7 Weaver CH, Hazelton B, Birch R et al. An analysis of 16–213 in the treatment of advanced squamous cell carcinoma engraftment kinetics as a function of the CD34 content of per- of the cervix and adenocarcinoma of the ovary: a Gynecologic ipheral blood progenitor cell collections in 692 patients after Oncology Group study. Cancer Treat Rep 1979; 63: 2089– the administration of myeloblative chemotherapy. Blood 1995: 2092. 86: 3961–3969. 23 Maskens AP, Armand J-P, Lacave AJ. Phase II 8 Lickliter JD, Begley CG, Boyd AW. Combined chemotherapy of VP-16–213 in ovarian cancer. Cancer Treat Rep 1981; 65: and granulocyte colony-stimulating factor (G-CSF) mobilize 329–330. large numbers of peripheral blood progenitor cells in treated 24 Edmonson JH, Decker DG, Malkasian GD. Phase II evalu- patients. Leuk Lymphoma 1994; 15: 91–97. ation of VP-16–213 (NSC-141540) in patients with advanced 9 Goldschmidt H, Hegenbart U, Haas R. Mobilization of periph- ovarian carcinoma resistant to alkylating agents. Gynecol eral blood progenitor cells with high-dose cyclophosphamide Oncol 1978; 6: 7–9. (4 or 7 gm/m2) and granulocyte colony stimulation factor in 25 Baars JW, Holtkamp MJ, Nooyew WJ. Mobilization of blood patients with multiple myeloma. Bone Marrow Transplant progenitor cells with ifosfamide and etoposide (VP-16) in 1996: 17: 691–697. combination with recombinant human G-CSF (Filgastrim) in 10 Haynes AP, Hunter A, McQuaker G. Engraftment character- patients with malignant or solid tumors. istics of peripheral blood stem cells mobilized with cyclophos- Anticancer Res 1996; 16: 3089–3095. phamide and the delayed addition of G-CSF. Bone Marrow 26 Muggia FM, Hainsworth JD, Jeffers S. Phase II study of lipo- Transplant 1995; 16: 359–363. somal in refractory ovarian cancer: antitumor 11 Weaver A, Wrigley E, Watson A. A study of ovarian cancer activity and toxicity modification by liposomal encapsulation. patients treated with dose-intensive chemotherapy supported J Clin Oncol 1997; 15: 987–993. with peripheral blood progenitor cells mobilised by G-CSF 27 ten Bokkell Huinink W, Gore M, Carmicheal J. and cyclophosphamide. Br J Cancer 1996; 74: 1821–1827. versus paclitaxel for the treatment of recurrent epithelial 12 Demirer T, Rowley S, Buckner CD. Peripheral-blood stem- ovarian cancer. J Clin Oncol 1997; 15: 2183–2193.

Bone Marrow Transplantation