Bone Marrow Transplantation, (1999) 23, 427–435  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Ifosfamide in combination with or : regimens which effectively mobilize peripheral blood progenitor cells while demonstrating anti-tumor activity in patients with metastatic breast

HM Prince1, J Gardyn1, MJ Millward1, D Rischin1, P Francis1, P Gates1, P Chapple1, M Quinn1, S Juneja1, M Wolf1, EH Januszewicz1, G Richardson2, J Scarlett2, P Briggs2, M Brettell1 and GC Toner1

1Blood and Marrow Transplant Service, Division of Haematology and Medical Oncology, Peter MacCallum Cancer Institute; and 2Department of Medical Oncology and Clinical Haematology, Monash Medical Centre, Melbourne, Victoria, Australia

Summary: with PBPCs. Rapid neutrophil and platelet recovery has been observed. Ifosfamide with G-CSF in combination For patients with metastatic (MBC) who with doxorubicin or paclitaxel achieves effective mobil- undergo high-dose therapy with autologous peripheral ization of PBPC and anti-tumor activity with minimal blood progenitor cell (PBPC) transplantation, an toxicity. important prerequisite is a mobilization regimen that Keywords: ifosfamide; paclitaxel; doxorubicin; mobiliz- efficiently mobilizes PBPCs while producing an effective ation; stem cells anti-tumor effect. We prospectively evaluated ifosfam- ide-based for mobilization efficiency, tox- icity and disease response in 37 patients. Patients received two cycles of the ifosfamide-based regimen; The value of dose-intensive chemotherapy in metastatic ifosfamide (5 g/m2 with conventional-dose cycle and breast cancer (MBC) continues to be debated1–4 and high- 6 g/m2 with mobilization cycle) with either 50 mg/m2 dose chemotherapy with autologous bone marrow (ABM) doxorubicin (if limited prior and/or pro- or peripheral blood progenitor cell (PBPC) transplantation gression more than 12 months after an anthracycline- continues to be investigated.5 A single small randomized based regimen) or 175 mg/m2 paclitaxel. For the mobil- study examining the outcome of patients with MBC treated ization cycle, all patients received additional G-CSF with high-dose chemotherapy and ABM or PBPC support (10 ␮g/kg SC, daily) commencing 24 h after completion demonstrated superior response rates, complete responses of chemotherapy. The target yield was Ͼ6 × 106 CD34+ (CR) and overall survival (OS) for patients treated in the cells/kg, sufficient to support the subsequent three cycles high-dose arm.6 Furthermore, in heavily pre-treated of high-dose therapy. The mobilization therapy was well patients, CR rates of 20–64% have been reported using a tolerated and the peak days for peripheral blood (PB) single cycle of high-dose chemotherapy with PBPC trans- CD34+ cells were days 10–13 with no significant differ- plantation.7 Despite these promising results, only a minority ences in the PB CD34+ cells mobilization kinetics (approximately 10–25%) of these highly selected groups of between the ifosfamide-doxorubicin vs ifosfamide-pacli- women with MBC experience prolonged progression-free taxel regimens. The median PBPC CD34+ cell content survival following high-dose therapy.5 ranged from 2.9 to 4.0 × 106/kg per day during days 9– We were therefore interested in examining the toler- 14. After a median of 3 (range 1–5) collection days, the ability and efficacy of a novel approach utilizing repetitive median total CD34+ cell, CFU-GM and MNC for all 44 high-dose combination chemotherapy for patients with individual sets of collections was 9.2 × 106/kg (range MBC. We initiated a phase I/II study in heavily pre-treated 0.16–54.9), 37 × 104/kg (range 5.7–247) and 7.3 × 108/kg patients with MBC using three cycles of high-dose ifosfam- (range 2.1–26.1), respectively. The PBPC target yield ide, and paclitaxel, with each cycle supported by was achieved in 35 of the 37 patients. The overall PBPCs. An important prerequisite for this therapy, was a response rate for the 31 evaluable patients was 68% reliable method of obtaining sufficient PBPCs to ensure with 10% having progressive disease. Thirty-three rapid and durable engraftment. Furthermore, an effective patients have subsequently received high-dose therapy mobilization regimen needed not only to mobilize adequate consisting of three planned cycles of high-dose ifosfam- PBPCs, but also potentially achieve an anti-tumor effect to ide, thiotepa and paclitaxel with each cycle supported maintain patients in at least a stable disease state at the time of commencing the high-dose phase of therapy. We selected ifosfamide as the ‘back-bone’ for our mobil- Correspondence: Dr HM Prince, Division of Haematology and Medical Oncology, Peter MacCallum Cancer Institute, Locked Bag 1, A’Beckett ization regimen for a number of reasons. Firstly, ifosfamide St, Melbourne 3000, Victoria, Australia can be incorporated into chemotherapy regimens to mobil- Received 22 May 1998; accepted 16 October 1998 ize PBPCs.8–14 Secondly, ifosfamide has some potential Ifosfamide mobilization in breast cancer HM Prince et al 428 Table 1 Patient characteristics (n = 37)

All Ifosfamide-doxorubicin Ifosfamide-paclitaxel G-CSF

Mobilization regimen (patients) 37 24 13 Disease status pre-ifosfamide-based chemotherapy Sensitive 15 15 0 Stable 2 2 0 Resistant 15a 411 Not evaluable 5 3 2 1st mobilisation regimen 37 22 12 3 2nd mobilisation regimen 7 4 2 1 Total 44 26 14 4

Median age (range) 44 (23–58) 44 (23–56) 44 (38–58) 43 (29–50)

Locally advanced 3 Metastatic 34 nodal 17 liver 13 bone 17 marrow 7 lung 12 brain 1 skin 2 peritoneum 2 ovary 2 adrenal 1 Median No. met sites 2 (1–4) Median No. of prior regimens 2 (1–3) 1 (1–3) 2 (1–3) 3 (1–3) Median No. of prior chemotherapy cycles 7 (1–15) 6 (1–15) 8 (1–15) 9 (7–15) No. of patients with prior 30 (81%) 15 (63%) 13 (100%) 4 (100%) No. of patients with prior 11 (30%) 2 (8%) 7 (54%) 2 (50%) No. of patients with prior RT for bone mets 5 (14%) 1 (4%) 7 (15%) 2 (50%)

aFour patients with refractory disease were not evaluable for response to ifosfamide-based chemotherapy (see text). RT = radiotherapy.

advantages over cyclophosphamide. Pre-clinical studies gressed after a recent anthracycline. Therefore, in these have demonstrated that the unique structural properties of patients paclitaxel was substituted for doxorubicin. Further- ifosfamide may have a beneficial anti-tumor activity; ifosfa- more, paclitaxel can achieve 20–30% objective responses mide is an analogue of cyclophosphamide with translo- in anthracycline-resistant patients,26,27 with evidence that it cation of a chlorethyl group which provides a more effec- may enhance PBPC yields.28–30 tive DNA cross-linking distance between two independent Here, we report on ifosfamide with doxorubicin or pacli- functional alkylating moieties.15,16 Pre-clinical studies dem- taxel for mobilization of PBPCs in heavily pre-treated onstrate differences in alkylation kinetics between the two patients with MBC. We sought to determine if these ifosfa- active metabolites, phosphoramide mustard and isofosfora- mide combinations were tolerable and effective regimens in mide mustard. This should result in longer survival of iso- such patients, while offering a reliable method for obtaining fosforamide mustard within the cell cytoplasm and an adequate PBPCs to support multiple cycles of high-dose increased probability of reaching the nucleus and alkylating therapy. DNA.17 These findings have been supported by the demon- stration that ifosfamide has superior anti-tumor activity in 18,19 some cell lines. Thirdly, in clinical studies, ifosfamide Materials and methods has demonstrated efficacy in advanced breast carci- 16,20–23 noma. Although to date there is no conclusive evi- Patient characteristics dence that ifosfamide is superior to cyclophosphamide in the treatment of breast cancer, taken together, we believed Patients with histologically and/or cytologically proven it rational to use ifosfamide-based mobilization in our metastatic or locally advanced breast cancer were prospec- cohort of heavily pre-treated patients. tively entered into a phase I/II utilizing three We selected either doxorubicin or paclitaxel in combi- cycles of high-dose ifosfamide, thiotepa and paclitaxel (see nation with ifosfamide depending on prior anthracycline below) with each cycle supported with PBPCs. Sufficient exposure. Doxorubicin-containing regimens are well docu- PBPCs to support three cycles of high-dose therapy were mented as achieving effective mobilization of PBPCs in collected prior to any of the cycles of high-dose therapy. breast cancer.24,25 However, we anticipated that a number Patients were excluded if they had a history of ventricular of patients in this prospective study may have been heavily arrhythmia, myocardial infarction 6 months prior to pre-treated and at risk of cardiac toxicity, and/or had pro- enrollment, congestive heart failure, sensitivity to any of Ifosfamide mobilization in breast cancer HM Prince et al 429 the study drugs, were pregnant or lactating, had an ECOG Disease assessment performance status Ͼ2, life expectancy of less than 2 months, absolute neutrophil count (ANC) Ͻ1.5 × 109/l, pla- Patients had disease assessment prior to their first cycle of telet count Ͻ100 × 109/l, bilirubin Ͼ1.5 times upper limit ifosfamide-based chemotherapy and prior to their first cycle of normal, aspartate transaminase Ͼ2 times upper limit of of high-dose therapy. Disease status was determined by normal, albumin Ͻ34 g/l or creatinine Ͼ1.5 times upper complete physical examination, chest X-ray, computerised limit of normal. Written informed consent was obtained tomography (CT) of abdomen and thorax, isotope bone from all patients and the protocol was approved by the scan, CA 15–3, and bone marrow biopsy (repeated only if Institutional Ethics Committee of Peter MacCallum originally positive). As a means of attempting to identify Cancer Institute. a patient’s potential response to ifosfamide-based chemo- therapy, patients were classified as sensitive or resistant to prior chemotherapy. Patients with a complete response Chemotherapy (CR) or partial response (PR) were defined as having Patients entered on the study received two cycles of ifosfa- chemotherapy-sensitive disease. Patients with progressive mide-based chemotherapy, the mobilization cycle was or stable disease following the last cycle of chemotherapy administered using a higher dose of ifosfamide with G- were defined as having resistant disease. Patients were eval- CSF. The was determined by prior uable for response to subsequent ifosfamide-based chemo- cumulative anthracycline exposure or time between the last therapy if they had bi-dimensionally measurable disease. anthracycline treatment to disease progression; patients Patients were not evaluable for response if complete surgi- with a cumulative dose of doxorubicin exceeding cal resection of tumor was performed prior to ifosfamide- 250 mg/m2 (or equivalent) or those who had progressed based chemotherapy or if bone lesions detectable by radio- within 12 months of receiving an anthracycline-based nucleotide bone scans were the sole evidence of disease. regimen, received a mobilization regimen of ifosfamide- Response to subsequent ifosfamide-based therapy was paclitaxel. All other patients received an ifosfamide- defined as follows: CR was defined as the total disappear- doxorubicin regimen. ance of all reversible clinical evidence of disease. Partial For the first cycle of chemotherapy patients received response was defined as at least a 50% reduction in the size ifosfamide (Holoxan; Asta Medica Australasia Pty Ltd, of all measurable tumor areas as measured by the perpen- Sydney, Australia) at a dose of 5 g/m2 as a 24 h continuous dicular diameters of the greatest length and width. Because infusion (day 1). (Uromitexan; Asta Medica Aus- patients proceeded immediately to high-dose therapy, tralasia Pty Ltd) was administered at 1 g/m2 intravenous response could not be followed for 4 weeks, nor could bolus over 30 min prior to the ifosfamide infusion, followed response duration be evaluated. Progressive disease (PD) by 5 g/m2 (combined with the ifosfamide) and administered was defined as the appearance of a new lesion, or an over 24 h. Following completion of the ifosfamide/mesna increase of 25% in the sum of areas of lesions within any infusion, mesna (1 g/m2) was infused over 6 h or 2 g/m2 of organ site. Stable disease (SD) was defined as disease that oral mesna administered in two divided doses. Doxorubicin fails to qualify for either a response or progressive disease. at 50 mg/m2 was administered intravenously on day 1 prior to the ifosfamide. Paclitaxel (Anzatax; FH Faulding Pharm- Apheresis procedure aceuticals, Adelaide, Australia) was administered as a 3 h infusion of 175 mg/m2 on day 2, at least 3 h following Patients had a routine full blood examination on day 8 fol- completion of ifosfamide. No sedative hypnotics were lowing mobilization. Daily CD34+ cell enumeration of the administered during the ifosfamide infusion. No hematopo- peripheral blood (PB) was commenced when the absolute ietic growth factors were given following this phase of neutrophil count (ANC) reached 0.5 × 109/l following the chemotherapy. nadir. Apheresis was initiated when the PB CD34+ cell On approximately day 21, patients received mobilization count exceeded 0.5 × 107/l.31 By employing this trigger chemotherapy consisting of the same drugs except that the point for initiating apheresis, patients were harvested 8–16 ifosfamide dose was increased to 6 g/m2 and administered days following mobilization. For patients mobilized with with mesna (1.2 g/m2) bolus followed by 6 g/m2 24 h G-CSF alone (n = 4), apheresis was commenced on day 5. infusion and 1.2 g/m2 post infusion. Doxorubicin and pacli- Apheresis was performed using a Haemonetics V50, taxel doses were the same. Granulocyte colony-stimulating CS3000plus (Baxter Healthcare Corp, Deerfield, IL, USA) factor (G-CSF: Neupogen; Amgen Australasia Pty Ltd) was or Spectra (Cobe, Denver, CO, USA) cell separators. Blood commenced 24 h after completion of mobilization chemo- was collected via a central venous access device or from therapy at 10 ␮g/kg subcutaneously daily and continued peripheral veins with 1.5–2.5 times the patient’s estimated until the last day of apheresis. Four patients received G- blood volume processed. Apheresis was continued until CSF (10 ␮g/kg) alone for mobilization of PBPCs. In three sufficient cells to support three separate cycles of high-dose of these patients, G-CSF mobilization was chosen first therapy, had been collected, ie a minimum target of because of prior extensive chemotherapy exposure. In each 6 × 106/kg (= 3 cycles of high-dose therapy, each with a case it was unsuccessful and was followed by an ifosfam- minimum of 2 × 106/kg CD34+ cells). To ensure uniformity, ide-based mobilization regimen. In one patient mobilization each collection was divided into three separate bags, so that with G-CSF alone was utilized after an inadequate collec- each re-infusion contained PBPCs collected from each day tion following the ifosfamide-doxorubicin mobilization of apheresis. Repeat PBPC collection was performed if the regimen. initial set of collections was deemed inadequate (ie initial Ifosfamide mobilization in breast cancer HM Prince et al 430 100.00

10.00 /l 7 10 × 1.00 PB CD34

0.10

0.01 D8 D9 D10 D11 D12 D13 D14 D15 D16 Days after mobilization D8 D9 D10 D11 D12 D13 D14 D15 D16 n 628323425113 6 4 min 0.01 0.01 0.01 0.01 0.21 0.23 2.2 0.63 0.47 median 0.01 0.46 5.6 5.5 10.0 3.4 6.1 1.8 1.1 max 8.0 17 48 21 21 51 73 7.3 7.6 mean 2.3 4.0 9.7 7.4 9.8 9.1 27 2.5 2.6

Figure 1 The PB CD34+ cell counts (log scale) following mobilization with the ifosfamide-based regimens (n = 40). Median values represented by horizontal lines.

total collection Ͻ6 × 106/kg). In patients with a borderline served PBPC were thawed and infused on day 0. All collection according to CD34+ cell criteria (ie 4.5– patients received G-CSF (5 ␮g/kg/day) subcutaneously 6.0 × 106/kg), an autograft was deemed acceptable if the from day 1 until the ANC was Ͼ1.5 × 109/l for 3 total colony-forming units granulocyte–macrophage (CFU- consecutive days. All patients received prophylactic GM) exceeded a total of 20 × 104/kg. ciprofloxacin, aciclovir and ranitidine.

Cell enumeration, processing and cryopreservation Statistical analysis Determination of PB and PBPC cell counts including WCC, Spearman correlation was used to compare PB CD34 and + MNC and CD34 cells has been previously described.31 The PBPC CD34 values. The Mann–Whitney U test was used + autograft content of MNC, CD34 cells and CFU-GM (per to compare the influence of the ifosfamide-doxorubicin and kg body weight) was determined using the patient’s actual ifosfamide-paclitaxel mobilization regimens on the PBPC body weight. CD34, CFU-GM and MNC, and ANC and platelet recovery As described, to ensure uniformity, the product from following high-dose therapy. Fisher’s exact test was used each day of collection was equally divided into three separ- to compare the doxorubicin- and paclitaxel-containing regi- ate bags, cryopreserved in 10% DMSO and stored in vapor- mens on response rate. All results are expressed as two- phase liquid nitrogen. Each re-infusion was equivalent and sided P values. Statistical analysis was performed using contained PBPCs collected from each day of apheresis. GraphPad Prism Ver 2.01 and GraphPad StatMate, Ver 1.00 for Windows 3.1. (GraphPad Software, San Diego, High-dose therapy CA, USA). Prior to mobilization, patients were enrolled in a phase I/II 2 study of high-dose ifosfamide (7.5–12.5 g/m /cycle), thi- Results otepa (200–350 mg/m2/cycle) and conventional-dose pacli- 2 32 taxel (175 mg/m /cycle). High-dose therapy was adminis- Patient characteristics tered according to the following schedule; on day −4 patients received mesna (20% w/w of the ifosfamide dose) Patient characteristics are summarized in Table 1. Thirty- over 30 min followed immediately by a 24 h infusion of six females and one male with the median age of 44 (range ifosfamide with an equivalent dose of mesna. Immediately 23–58 years) underwent PBPC collection. The majority of following the ifosfamide/mesna infusion, mesna (20% w/w patients had metastatic disease (n = 34) and three had of the ifosfamide dose) was administered intravenously locally advanced disease. Most patients were heavily pre- over 8 h. Thiotepa was administered in equally divided treated and 15 were resistant to conventional-dose chemo- doses on days −4to−2, inclusive. Paclitaxel (175 mg/m2) therapy, the majority of whom received the paclitaxel- was administered as a 3 h infusion on day −2. Cryopre- containing regimen (Table 1). Ifosfamide mobilization in breast cancer HM Prince et al 431 a 100.00

10.00 /l 7 10 × 1.00 PB CD34

0.10

0.01 D8 D9 D10 D11 D12 D13 D14 D15 D16 Days after mobilization D8 D9 D10 D11 D12 D13 D14 D15 D16 n 2182023211125 4 min 0.01 0.01 0.01 0.01 0.21 0.23 6.08 0.63 0.47 median 0.01 0.10 4.77 7.57 12.1 3.4 39.6 1.85 1.14 max 0.01 16.5 48.3 20.5 20.9 50.7 73.2 7.27 7.6 mean 0.01 4.0 9.7 8.1 10.4 9.11 39 2.68 2.58

b 100.00

10.00 /l 7 10 × 1.00 PB CD34

0.10

0.01 D8 D9 D10 D11 D12 D13 D14 D15 D16 Days after mobilization D8 D9 D10 D11 D12 D13 D14 D15 D16 n 41012114 1 1 0 min 0.01 0.01 0.50 2.17 3.40 2.20 1.70 median 3.03 2.39 6.18 5.27 6.91 2.20 1.70 max 7.98 13.5 30.0 12.8 11.2 2.20 1.70 mean 3.51 4.89 9.72 5.99 7.10 2.20 1.70

Figure 2 The PB CD34+ cell counts (log scale) following mobilization with (a) the ifosfamide-doxorubicin regimen (n = 26) and (b) the ifosfamide- paclitaxel regimen (n = 14). Median values represented by horizontal lines.

Apheresis collections 8 in all cases) the PB CD34/l was determined. If the PB CD34+ cells exceeded 0.5 × 107/l, apheresis was performed. All 37 patients underwent mobilization with an ifosfamide Indeed, we have previously reported the value of this PB based-regimen. Of these, seven patients had an additional CD34+ cell threshold; the strong correlation between PB mobilization and collection with either the same regimen + + CD34 cell and the PBPC CD34 cell content was con- or G-CSF alone (Table 1). Consequently, the total data set firmed in this study (Spearman correlation; r = 0.8854, includes the results of 44 mobilizations and collections with P Ͻ 0.0001).31 The peak days for PB CD34+ cells were 133 individual apheresis procedures. 9 days 10–12 (Figure 1). There was no significant difference Once the PB ANC exceeded 0.5 × 10 /l (on or after day + in the PB CD34 cell mobilization kinetics between ifos- Ifosfamide mobilization in breast cancer HM Prince et al 432 100.00

10.00 /kg 6 10

× 1.00 PB CD34 0.10

0.01 D8 D9 D10 D11 D12 D13 D14 D15 D16 Days after mobilization D8 D9 D10 D11 D12 D13 D14 D15 D16 n 2142531241036 4 min 1.54 0.01 0.28 0.55 0.01 0.72 0.61 0.58 0.15 median 2.98 3.7 3.6 4.0 4.0 2.9 3.0 0.76 1.1 max 4.41 12.0 18.0 12.0 16.0 25.0 23.0 5.9 4.5 mean 2.98 4.6 5.4 4.6 5.2 6.1 8.7 1.7 1.7 Figure 3 The peripheral blood progenitor cell (PBPC) CD34+ cell content (log scale) following mobilization with the ifosfamide-based regimen (n = 40). Median values represented by horizontal lines.

famide-doxorubicin vs ifosfamide-paclitaxel (day 9, second collection was unlikely to succeed if the first collec- P = 0.1191; day 10, P = 0.4715; day 11, P = 0.8540; day tion had failed (data not shown). 12, P = 0.3940) (Figure 2). Using the above criteria, the median day to initiate PBPC Toxicity and responses to mobilization chemotherapy harvesting was day 10 (range 8–13; s.d. = 1.15). The median PBPC CD34+ cell content ranged from 2.9 to 4.0 The mobilization therapy was generally well tolerated. One x106/kg per day during days 9–14 (Figure 3). Following a patient developed a grade II acute brain syndrome 12 h median of 3 days (range 1–5) collection for each patient after commencing ifosfamide. A CT scan of brain was nor- following each mobilization, the median total CD34+ cell, mal and the patient received three doses of CFU-GM and MNC for all 44 individual sets of collections with progressive resolution of symptoms over 48 h. No was 9.2 × 106/kg (range 0.16–54.9), 37 × 104/kg (range patient developed hemorrhagic cystitis. 5.7–247) and 7.3 × 106/kg (range 2.1–26.1), respectively. Hematopoietic parameters following mobilization are There were no significant differences in total PBPC auto- detailed in Table 2. The neutrophil nadir was lower in graft yields between the mobilization regimens of ifos- + famide-doxorubicin vs ifosfamide-paclitaxel; CD34 cells Table 2 Haematopoietic toxicity (P = 0.059), CFU-GM (P = 0.156) and MNC (P = 0.550). All patients were planned to receive subsequent repeti- Ifosfamide- Ifosfamide- P value tive high-dose ifosfamide, thiotepa and paclitaxel. As doxorubicin paclitaxel described, each of three high-dose cycles was followed by PBPC support each with a target CD34+ cell autograft count Mobilization phase a 9 × 6 Nadir ANC (×10 /l) 0.2 (0.01–1.89) 0.88 (0.11–1.67) 0.0320 of 2 10 /kg, thus the goal of the mobilization prior to this b + Nadir platelets 56 (11–139) 108 (30–308) 0.0062 high-dose phase was to achieve a CD34 cell yield of (×109/l) greater than 6 × 106/kg which was divided equally into three separate bags. Utilizing an ifosfamide-based mobiliz- Transplant phase (n = 84 ation regimen we were able to achieve this CD34+ cell tar- cycles) ANC Ͼ0.5 × 109/l 10d (7–19) 10d (8–26) 0.1345 get in 33 of the 37 patients (89%). Of the four patients ANC Ͼ1.0 × 109/l 10d (8–20) 11d (9–28) 0.0830 failing to achieve this threshold, two of these patients had Platelets Ͼ20 × 109/l 13d (9–25) 14d (10–38) 0.0862 a CFU-GM yield of Ͼ20 × 104/kg (CD34 × 106/kg were 3.7 Platelets Ͼ50 × 109/l 18d (10–33) 22d (11–40) 0.0062 and 5.1) and subsequently underwent autotransplant. The Platelets Ͼ100 × 109/l 24d (10–152) 30d (16–124) 0.0840 two remaining patients had poor CD34+ cell and CFU-GM yields and did not undergo autotransplant. aFull blood counts were performed on day1 and on and after day 8 rou- tinely. Nadir values that may have occurred prior to day 8 cannot be rep- Seven patients underwent a second mobilization and col- resented therefore these results may not reflect the true nadir counts. lection because of an inadequate primary collection. The bPlatelet nadirs universally occurred during apheresis. Two patients yield of these second collections was generally poor and a developed platelets Ͻ20 × 109/l and neither required platelet transfusions. Ifosfamide mobilization in breast cancer HM Prince et al 433 patients mobilized with the ifosfamide-doxorubicin regi- plantation as part of the phase I/II study.32 Twenty-one of men, however, as blood counts were performed on or after these patients were mobilized with ifosfamide-doxorubicin day 8 only, this may not reflect the true nadir in all cases. and 12 with ifosfamide-paclitaxel. Of the four patients not No patients developed febrile neutropenia. Platelet nadirs undergoing high-dose therapy and PBPC transplantation, universally occurred during the period patients were two did not obtain a sufficient autograft to proceed to trans- undergoing apheresis and were also lower in the ifosfam- plant and two were excluded because of progressive disease ide-doxorubicin group. Only two patients (both in the ifos- associated with deteriorating performance status. The famide-doxorubicin group) developed thrombocytopenia median time from initial diagnosis of breast cancer to trans- below 20 × 109/l and neither required platelet transfusions. plant was 16 months (range 2.3–134.4) and the median time No patients required red blood cell transfusions. from mobilization chemotherapy to transplant was 33 days Thirty-one patients were evaluable for response with bi- (18–100 days). Of the 33 patients undergoing repetitive dimensionally measurable disease. Four patients with high-dose therapy and transplant, 84 cycles of high-dose refractory disease were not evaluable at time of entry into treatment were delivered. For all 84 cycles, the median time the study; three patients had had locally advanced disease, to ANC Ͼ0.5 × 109/l and 1.0 × 109/l was 10 days (range failed to achieve a response to conventional-dose therapy 7–26) and 10 days (range 8–28), respectively. The median and subsequently underwent complete surgical resection. A time to platelets greater than 20 × 109/l, 50 × 109/l and further patient had a soft tissue and bone recurrence, failed 100 × 109/l was 13 days (range 9–28), 19 days (range 10– to respond to an anthracycline and received local RT. The 40) and 25 days (range 10–152), respectively. The overall response rate following the mobilization chemo- recoveries according to mobilization regimen are detailed therapy for the 37 patients (including six patients with non- in Table 2. evaluable disease) was 57%, with an additional 19% of patients having stable disease (Table 3). The response rate for the 31 evaluable patients was 68%, 23% having stable Discussion disease and 10%, progressive disease. Two patients were excluded from subsequent high-dose therapy because of We demonstrate that the mobilization regimens reported progressive disease and deteriorating performance status here have effective anti-tumor activity while mobilizing (brain and liver metastases). All 15 patients with chemo- sufficient PBPCs in the majority of patients. Our prospec- therapy-responsive disease prior to the mobilization chemo- tive study demonstrates that 33 of the 37 (89%) patients therapy had a further response (three CRs, 12 PRs). Of the obtained sufficient PBPCs while maintaining at least stable 15 patients who had disease resistant to conventional-dose enough disease to undergo high-dose therapy with sub- chemotherapy prior to entry into the study (four patients sequent rapid platelet and neutrophil recovery. with non-evaluable disease), there was an 18% (2/11) PR Although ifosfamide has previously been incorporated rate with six (55%) having stable disease and three (27%) into chemotherapy regimens to mobilize PBPCs,8–14 these progressive disease. There was a statistically significant dif- studies have not been restricted to patients with breast can- ference in response rates between the ifosfamide-doxorub- cer,8–11,14 and those with breast cancer have generally been icin (n = 24) and the ifosfamide-paclitaxel groups (n = 13) minimally pre-treated and not had metastatic or relapsed (75% vs 23%; P = 0.0046). The result was similar when disease.12,13 The choice of the most appropriate mobiliz- only evaluable patients (n = 31) were examined ation regimen for patients with previously treated disease (P = 0.0152). However, the pre-treatment characteristics in is further limited by prior cyclophosphamide and/or anthra- the paclitaxel group were substantially worse (Table 1). cycline exposure. Based on the findings reported here, we believe that the combination of ifosfamide with either High-dose therapy doxorubicin or paclitaxel is an appropriate strategy in such patients. Nonetheless, 27% of patients who were refractory Of the 37 patients undergoing PBPC collections, 33 patients to prior conventional-dose therapy remained resistant to the subsequently received high-dose therapy and PBPC trans- ifosfamide-based regimen. The ifosfamide-paclitaxel and ifosfamide-doxorubicin Table 3 Response rate for all patients following mobilization chemo- regimens appeared at least equally efficacious in mobilizing therapy (n = 37) PBPCs, with equivalent CD34+ cells mobilized into the PB, similar mobilization kinetics and equivalent yields in the All (%) Ifosfamide- Ifosfamide- PBPC collections. There was a trend toward better CD34+ doxorubicin (%) paclitaxel (%) cell collections in the ifosfamide-doxorubicin group (P = 0.059) and this appears to have translated into a trend n 37 24 13 Response rates toward slower neutrophil and platelet recovery following CR 4 (11%) 4 (17%) 0 (0%) high-dose therapy (Table 2). However, the paclitaxel group PR 17 (46%) 14 (58%) 3 (23%) was more heavily pre-treated with worse disease character- SD 7 (19%) 4 (17%) 3 (23%) istics and may reflect the trend toward poorer mobilization. PD 3 (8%) 0 (0%) 3 (23%) In this study the median time to initiate collection was NE 6 (16%) 2 (8%) 4 (31%) RR (total) 21/37 (57%) 18/24 (75%) 3/13 (23%) day 10 after mobilization, however, this varied by 2 days RR (evaluable) 21/31 (68%) 18/22 (82%) 3/9 (33%) either way. We have previously demonstrated the value of a morning PB CD34 count in predicting PBPC yield,31 and SD = stable disease; PD = progressive disease; NE = not evaluable. the results presented here demonstrate that by using an Ifosfamide mobilization in breast cancer HM Prince et al 434 appropriate PB CD34+ cell trigger to initiate collection, the chemotherapy dose escalation in patients with haematological timing of harvesting can be optimized with approximately malignancies and solid tumors. J Clin Oncol 1997; 15: 3 × 106 CD34 cells/kg collected each day. Six patients 2981–2995. required a second cycle of mobilisation and collection. The 4 Canellos GP. Selection bias in trials of transplantation for yield of these second collections was generally poor, parti- metastatic breast cancer: have we picked the apple before it was ripe? J Clin Oncol 1997; 15: 3169–3170. cularly if the first collection had failed. Similar findings in 5 Antman KH, Rowlings PA, Vaughan WP et al. High-dose myeloma patients undergoing second collection has pre- chemotherapy with autologous hematopoietic stem-cell sup- 33 viously been demonstrated. port for breast cancer in North America. J Clin Oncol 1997; The overall response rate for all patients was 57% with 15: 1870–1879. an additional 19% of patients having stable disease. Three 6 Bezwoda WR, Seymour L, Dansey RD. High-dose chemo- patients developed progressive disease during the mobiliz- therapy with hematopoietic rescue as primary treatment for ation phase. Furthermore, if only evaluable patients were metastatic breast cancer: a randomized trial. J Clin Oncol analyzed (ie bi-dimensionally measurable disease), 68% of 1995; 13: 2483–2489. such patients responded. Indeed, our cohort had poor dis- 7 Sledge GW, Antman KH. Progress in chemotherapy for meta- ease characteristics; patients had a median of two prior regi- static breast cancer. Semin Oncol 1996; 19: 317–332. 8 Brugger W, Bross KJ, Glatt M et al. Mobilization of tumor mens, 15 having refractory disease, 81% had received prior cells and hematopoietic progenitor cells into peripheral blood cyclophosphamide, and 30% patients had received prior of patients with solid tumors. Blood 1994; 83: 636–640. therapy. These response rates are comparable to 9 Brugger W, Bross K, Frisch J et al. Mobilization of peripheral other studies utilizing ifosfamide in combination chemo- blood progenitor cells by seqential administration of interleu- therapy in MBC.16,20–23 Although there was a better kin-3 and granulocyte–macrophage colony-stimulating factor response in the ifosfamide-doxorubicin group (75% vs following polychemotherapy with , ifosfamide and 23%; P = 0.0046), the pre-treatment characteristics in the . Blood 1992; 79: 1193–1200. paclitaxel group were worse. 10 Brugger W, Birken R, Bertz H et al. Peripheral blood progeni- One important aspect of this study was to determine if tor cells mobilized by chemotherapy plus granulocyte-colony ifosfamide (6 g/m2) could be delivered safely as a 24 h stimulating factor accelerate both neutrophil and platelet recovery after high-dose VP16, ifosfamide and cisplatin. Br J infusion. A relatively short duration of infusion was Haematol 1993; 84: 402–407. required; the postulate being that a short exposure (as com- 11 Socinski A, Elias A, Schipper L et al. Granulocyte–macro- pared to the frequently used ifosfamide administration over phage colony stimulating factor expands the circulating 3–5 days in divided doses) would allow more synchronous, haemopoietic progenitor cell compartment in man. Lancet and possibly more efficacious, mobilization of PBPCs. 1988; ii: 1194–1198. Indeed, this regimen was associated with tolerable side- 12 Haas R, Schmid H, Hahn U et al. Tandem high-dose therapy effects. No red cell or platelet transfusions were required with ifosfamide, , and periphral blood following mobilization and there were no episodes of stem cell support is an effective adjuvant treatment for high- febrile neutropenia. Given the of a 3 h risk primary breast cancer. Eur J Cancer 1997; 33: 372–378. paclitaxel infusion, it was not surprising that neutrophil and 13 Schulze R, Schulze M, Wischnik A et al. Tumor cell contami- nation of peripheral stem cell transplants and bone marrow in platelet nadirs following mobilization were lower in the high-risk breast cancer patients. Bone Marrow Transplant ifosfamide-doxorubicin group. There were no episodes 1997; 19: 1223–1228. of hemorrhagic cystitis although one patient developed 14 McQuaker IG, Haynes AP, Stainer C et al. Stem cell mobiliz- reversible neurotoxicity. ation in resistant or relapsed : superior yield of pro- We conclude that ifosfamide-based chemotherapy with genitor cells following a salvage regimen comprising ifos- G-CSF in combination with doxorubicin or paclitaxel phamide, etoposide and epirubicin compared to intermediate- achieves effective mobilization of PBPC and anti-tumor dose cyclophosphamide. Br J Haematol 1997; 98: 228–233. activity with acceptable toxicity. 15 Druckrey H. Krebs-experimentelle ursachen forschung und chemothrapie. Z Krebsgeschehen 1973; 5: 73–79. 16 Hortobagyi GN. Activity of ifosfamide in breast cancer. Semin Oncol 1992; 6 (Suppl. 12): 36–42. Acknowledgements 17 Boal JH, Williamson M, Boyd VL et al. 31P NMR studies of the kinetics of bisalkylation by isophosphoramide mustard: We wish to thank Asta Medica (Sydney, Australia) and FH Fauld- comparisons with phosphoramide mustard. J Med Chem 1989; ing (Adelaide, Australia) for their generous support of this study, 32: 1768–1773. the research nurses, apheresis nurses and nursing staff on the Hae- 18 Goldin A. Ifosfamide in experimental tumor systems. Semin matology and Day Wards at Peter MacCallum Cancer Institute Oncol 1982; 9 (Suppl. 1): 14–23. for commitment, dedication and expert patient care. 19 Berger DP, Fiebig HH, Winterhalter BR et al. Preclinical phase II study of ifosfamide in human tumour xenografts in vivo. Cancer Chemother Pharmacol 1990; 26 (Suppl.): S7– References S11. 20 Millward M, Lind M, Gumbrell L et al. Results of chemo- 1 Hyrniuk W, Bush H. The importance of dose intensity in therapy using ifosfamide with doxorubicin in advanced breast chemotherapy of metastatic breast cancer. J Clin Oncol 1984; cancer. Breast Cancer Res Treat 1994; 29: 271–277. 2: 1281–1288. 21 Lind M, Gumbrell L, Cantwell B et al. The use of granulocyte 2 Siu LL, Tannock IF. Chemotherapy dose escalation: case colony-stimulating factor to deliver four cycles of ifosfamide unproven. J Clin Oncol 1997; 15: 2765–2768. and epirubicin every 14 days in women with advanced or 3 Savarese DMF, Hsieh C, Stewart FM. Clinical impact of metastatic breast cancer. Br J Cancer 1995; 71: 610–613. Ifosfamide mobilization in breast cancer HM Prince et al 435 22 Paridaens R, Focan C, Michel J et al. Experience of the tions in patients with breast cancer receiving granulocyte col- Belgian society of medical oncology with single adminis- ony-stimulating factor. J Clin Oncol 1997; 15: 684–690. tration 5 g/m2 ifosfamide with mesna as second- or third-line 29 Raptis G, Tiersten A, Fennelly D et al. The addition of pacli- therapy in advanced breast cancer. Cancer Chemother Pharm- taxel to high dose cyclophosphamide + filgrastim markedly acol 1990; 26 (Suppl.): S63–S65. enhances the mobilization of CD34+ peripheral blood progeni- 23 Overmeyer BA. Ifosfamide in the treatment of breast cancer. tor cells in patients with responding metastatic breast cancer Semin Oncol 1996; 23 (Suppl. 6): 38–41. undergoing high dose chemotherapy. Proc Am Soc Clin Oncol 24 Prince HM, Imrie K, Keating A et al. Cyclophosphamide, 1995; 14: 313 No. 919. adriamycin, 5 flurouracil and G-CSF mobilization achieves 30 Weaver CH, Schwartzberg L, Birch R et al. Collection of per- high peripheral blood progenitor cell yield which can support ipheral blood stem cells following administration of paclitaxel, multiple cycles of high-dose chemotherapy. Blood 1995; 86 cyclophosphamide, and filgrastim in patients with breast and (Suppl. 1): 404a No. 1602. . Biol Blood Marrow Trans 1997; 3: 83–90. 25 Viens P, Gravis G, Genre D et al. High-dose sequential 31 Chapple P, Prince HM, Quinn M et al. Peripheral blood chemotherapy with stem cell support for non-metstatic breast CD34-positive cell count reliably predicts autograft yield. cancer. Bone Marrow Transplant 1997; 20: 199–203. Bone Marrow Transplant 1998; 22: 125–130. 26 Seidman A, Reichman BS, Crown JPA et al. Paclitaxel as 32 Prince HM, Millward MJ, Rischin D et al. Repetitive high- second and subsequent therapy for metastatic breast cancer: dose therapy with ifosfamide, thiotepa and paclitaxel with per- activity independent of prior anthracycline response. J Clin ipheral blood progenitor cell and filgrastim support for meta- Oncol 1995; 13: 1152–1159. static and locally advanced breast cancer: results of a phase I 27 Seidman AD, Hudis C, Norton L. Memorial Sloan-Kettering Center experience with paclitaxel in the treatment of breast study. Ann Oncol 1999 (in press). cancer: from advanced disease to adjuvant therapy. Semin 33 Prince HM, Imrie K, Sutherland DR et al. Peripheral blood Oncol 1995; 22 (Suppl. 8): 3–8. progenitor cell collections in multiple myeloma: predictors 28 Demirer T, Buckner CD, Storer B et al. Effect of different and management of inadequate collections. Br J Haematol chemotherapy regimens on peripheral-blood stem-cell collec- 1996; 93: 142–145.