A Safe and Active Regimen for Tumor Cytoreduction and Stem Cell Mobilization in Metastatic Breast Cancer

Home , Aclarubicin

Bone Marrow Transplantation, (2000) 25, 123–130  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt Dose-intense paclitaxel, etoposide and cyclophosphamide: a safe and active regimen for tumor cytoreduction and stem cell mobilization in metastatic breast cancer

S Bilgrami, JM Feingold, RD Bona, RL Edwards, AM Khan, F Rodriguez-Pinero, IA Khan, D Kazierad, J Clive and PJ Tutschka

Bone Marrow Transplant Program, University of Connecticut Health Center, Farmington, CT, USA

Summary: Keywords: stem cell mobilization; tumor cytoreduction; breast cancer; cyclophosphamide; etoposide; paclitaxel Patients with metastatic breast cancer in complete remission are the ones most likely to have an improved outcome with subsequent high-dose chemotherapy and autologous peripheral blood stem cell transplantation Metastatic carcinoma of the breast is a fatal malignancy (HDC-PBSCT). Peripheral blood stem cells are usually with a median survival of about 2 years when treated with procured following mobilization with single agent chemo-hormonal therapy. High-dose chemotherapy with chemotherapy and colony-stimulating factor support. autologous peripheral blood stem cell transplantation We utilized a dose-intense regimen of paclitaxel (HDC/PBSCT) has been performed increasingly over the 200 mg/m2 i.v., etoposide 60 mg/kg i.v., and cyclophos- past decade in an effort to impact positively on the usually phamide 3 g/m2 i.v. (TEC) followed by daily adminis- dismal outcome of metastatic breast cancer. Unfortunately, tration of granulocyte colony-stimulating factor. The there is a dearth of multicenter controlled trials comparing aim was not only to mobilize stem cells but also to achi- HDC/PBSCT to conventional chemotherapy. However, one eve optimal tumor cytoreduction prior to HDC/PBSCT. single institution study revealed a survival advantage in One hundred consecutive patients with metastatic patients treated with HDC/PBSCT1 whereas a recent multi- breast cancer received 257 cycles of TEC between institutional trial failed to demonstrate any such benefit.2 March 1994 and June 1997, with the aim of collecting In addition, a recent multi-institutional analysis of North -؋ 106 CD34-positive cells/kg usually following the American patients undergoing autotransplantation for meta 5 second cycle of chemotherapy. Patient characteristics static breast cancer demonstrated a 4-year progression-free included a median age of 45 years, a median of two survival of 32% among patients who had achieved com- organ systems involved by disease, a median of two plete response to conventional chemotherapy given prior to prior chemotherapy regimens and eight prior chemo- HDC/PBSCT. The progression-free survival was 13% therapy cycles, and a median interval of 8 months from among partial responders, and only 7% among non- diagnosis of metastases to first cycle of TEC. There were responders.3 Therefore, this suggests that one should 61 febrile episodes during neutropenia and 13 of these attempt to induce a complete remission prior to were associated with bacteremia or fungemia. Mortality HDC/PBSCT. An ideal regimen for this purpose should not rate was 1%. An adequate number of stem cells was only be highly active against breast cancer but should also collected in 90% of patients. The overall response rate be capable of mobilizing peripheral blood stem cells of the tumor was 58.8% with 23.7% complete (PBSC) for harvesting. The current study analyzes the responders among 97 evaluable patients. Multivariate results of an intense chemotherapy regimen consisting of analysis demonstrated chemosensitivity to the most paclitaxel, etoposide and cyclophosphamide (TEC) admin- recent standard chemotherapy regimen administered istered to 100 consecutive women with metastatic breast for metastatic disease, an ECOG performance score of cancer who were being considered for HDC/PBSCT. The 0 as opposed to 1, 2 or 3, and involvement by disease aim of this study was to determine the stem cell mobilizing of only one organ system as significant variables for ability, anti-tumor effect, and toxicity of TEC in patients achieving a complete remission with TEC. This novel with metastatic breast cancer. dose-intense regimen was safe and well tolerated, highly active against metastatic breast cancer, and capable of excellent stem cell mobilization. Bone Marrow Transplan- Patients and methods tation (2000) 25, 123–130. Patients

Correspondence: Dr S Bilgrami, MC-1315, University of Connecticut Patients with metastatic carcinoma of the breast were Health Center, 263 Farmington Avenue, Farmington, CT 06030, USA referred to the University of Connecticut Health Center, Received 10 May 1999; accepted 15 September 1999 Farmington, CT, for HDC/PBSCT. Every individual was TEC for stem cell mobilization in breast cancer S Bilgrami et al 124 required to give informed consent prior to the adminis- dexamethasone; and ondansetron 0.15 mg/kg i.v. every 6 h. tration of stem cell mobilizing chemotherapy and granulo- The 6th and 7th doses of dexamethasone administered at cyte colony-stimulating factor (G-CSF), and collection of hours 36 and 42 respectively, were 20 mg each instead of PBSC. Between March 1994 and June 1997, 100 consecu- 4 mg each. Diphenhydramine 50 mg i.v. and ranitidine tive patients with metastatic breast cancer received a novel 50 mg i.v. were administered 30 min prior to the com- chemotherapy regimen with the aim of collecting PBSC. mencement of the paclitaxel infusion. MESNA 12 mg/kg The protocol was approved by the institutional review i.v. was given at hours 27, 30, 33, 36, 39 and 42. board of the University of Connecticut Health Center. Eligi- Antibiotic prophylaxis consisted of ciprofloxacin 500 mg bility criteria included an age р65 years, an ECOG per- orally twice daily from day 4 until recovery of the white formance score of 0, 1, 2 or 3, an absolute neutrophil count blood cell (WBC) count (ANC Ͼ1 ϫ 109/l) or fever requir- (ANC) у1.5 ϫ 109/l, a platelet count у100 ϫ 109/l, hep- ing the initiation of parenteral antibiotics. Prophylaxis atic transaminases less than twice normal, a creatinine against gram-positive organisms was provided by either clearance у60 ml/min, a left ventricular ejection fraction ampicillin 250 mg orally four times daily starting at day 6 у50%, and a diffusion lung capacity у50% of the or clarithromycin 250 mg orally twice daily from day 6 predicted value. until recovery of the WBC count or initiation of parenteral antibiotics for fever. Neutropenic fever was treated with Chemotherapy hospitalization and broad-spectrum antibiotic therapy. Other supportive measures included ranitidine 150 mg Patients received combination chemotherapy supported by orally twice daily from day 4 until recovery of the WBC G-CSF to facilitate PBSC mobilization and harvesting as count and prednisone 40 mg orally twice daily from day 8 well as to optimize tumor cytoreduction. The regimen con- until recovery of the WBC count. Packed red blood cell sisted of etoposide 60 mg/kg by continuous i.v. infusion (PRBC) transfusions were administered when the hemoglo- from hours 0 to 18, cyclophosphamide 1.5 g/m2 i.v. from bin was р8 g/dl and platelet products were transfused for hours 24 to 25 and 30 to 31, and paclitaxel 200 mg/m2 i.v. a platelet count р20 ϫ 109/l. All individuals were seen in from hour 48 to 51. Granulocyte colony-stimulating factor the outpatient clinic on alternate days from day 6 until 5–10 ␮g/kg/day i.v. was initiated from hour 60 and con- recovery of the WBC count. tinued until an ANC of greater than 5 ϫ 109/l was reached or until completion of PBSC collection (Table 1). All Peripheral blood stem cell collection patients were hospitalized from hour 0 until completion of the first dose of G-CSF (3 days). Peripheral blood stem cells were enumerated by determin- High-dose chemotherapy began 1 month after the ing the percentage of CD34-positive mononuclear cells on initiation of the final cycle of TEC and consisted of busul- the day that the total WBC count exceeded 1 ϫ 109/l fol- fan 4 mg/kg/day orally for 4 days (day −10 to day −7), lowing the nadir from chemotherapy. If the circulating thiotepa 125–175 mg/m2/day i.v. for 4 days (day −6 to day CD34-positive cell count was greater than 1%, stem cell −3), and carboplatin 200–250 mg/m2/day i.v. for 4 days leukapheresis was conducted with a continuous flow cell (day −6 to day −3). Stem cells were infused on day 0. Gra- separator (Cobe Spectra, Cobe CBT, Lakewood, CO, USA) nulocyte colony-stimulating factor 5 ␮g/kg/day was admin- processing 10–20 l of blood per day at flow rates of 50– istered i.v. from day +1 until recovery of the neutrophil 80 ml/min. Stem cell leukapheresis was attempted follow- count following aplasia. ing both the first as well as the second cycle of TEC in the initial group of patients. However, if an adequate number Supportive care of stem cells was collected after the second cycle of TEC, any stem cells collected and cryopreserved following the Supportive measures during the administration of chemo- first course were not utilized for PBSCT for fear of a therapy included an intensive anti-emetic regimen con- greater likelihood of contamination with tumor cells. From sisting of dexamethasone 10 mg i.v. 30 min prior to the December 1994 onwards, 10 ml of bone marrow obtained initiation of etoposide and then 4 mg i.v. every 6 h until prior to the initiation of TEC, and 10 ml of the PBSC pro- hour 60; lorazepam 0.5–1.0 mg i.v. every 6 h along with duct collected on the first day of leukapheresis following each course of TEC was submitted for detection of micro- metastases by an immunocytochemical stain (BIS Labora- Table 1 Chemotherapy regimen tories, Reseda, CA, USA). Patients who mobilized well (Ͼ5% CD34-positive cells in the circulation) following the Hour Therapy first cycle of chemotherapy also mobilized adequately Ͼ 0–18 Etoposide 60 mg/kg ideal body weight by ( 1%) following the second course. Therefore, stem cell continuous i.v. infusion collection was deferred to the second course of chemo- 24–25 and 30–31 Cyclophosphamide 1.5 g/m2 (actual body therapy if mobilization of CD34-positive cells in the circu- weight) in 250 ml of 5% dextrose i.v. lation exceeded 5% following the first cycle of TEC in the 24, 27, 30, 33, 36, 39, 42 Mesna 12 mg/kg actual body weight i.v. final third group of patients. The aim in all individuals was 48–51 Paclitaxel 200 mg/m2 (actual body weight) by continuous i.v. infusion to re-infuse only those stem cells which had not only been 60 Granulocyte colony-stimulating factor collected after the second cycle of TEC but which had also 5 ␮g/kg actual body weight/day i.v. failed to demonstrate any evidence of micro-metastatic disease. Stem cell collection was performed daily until a target

Bone Marrow Transplantation TEC for stem cell mobilization in breast cancer S Bilgrami et al 125 of 5 ϫ 106 CD34-positive cells/kg body weight was using the median test. Finally, the joint effect of a combi- achieved, if possible. nation of variables as a set of ‘risk factors’ for not achiev- The CD34-positive cell count was determined from a ing a CR following TEC was determined using a logistic whole blood specimen prior to leukapheresis and from the regression model. leukapheresis product by a method described previously.4 Mononuclear cells were separated from other cellular elements in the blood through the use of Ficoll–Paque Results (Pharmacia Biotech, Uppsala, Sweden). Contaminating red blood cells were lysed by ammonium chloride. The Patient characteristics resulting cell suspension was stained with anti-CD34 mono- clonal antibody conjugated with fluorescein isothiocyanate The characteristics of patients enrolled in this study are (CD34 FITC Anti-HPCA-2) (Becton Dickinson, San Jose, presented in Table 2. A total of 100 patients with metastatic breast cancer received 257 courses of TEC. Their median CA, USA). The product underwent flow cytometric analysis on a FACS Calibur (Becton Dickinson) The number of age was 45 years and the median interval from diagnosis colony-forming units granulocyte–macrophage (CFU-GM) of breast cancer to the first cycle of TEC was 30 months. The ECOG performance score was 0 in 58 patients, 1 in after 14 days of culture of the PBSC product was determined by modification of a previously described method.5 29 patients, 2 in seven patients and 3 in six individuals. Ninety patients had been diagnosed originally with less than stage 4 disease. The interval from diagnosis of breast Staging evaluation and definitions of response cancer to the first cycle of TEC was less than 2 years in 77 of these patients, and 2 or more years in 13 individuals. Sites of disease were determined clinically, with com- Ten other patients presented with de novo metastatic breast puterized axial tomography (CAT) of the chest and abdo- men, and with whole body radionuclide skeletal scintigra- cancer. Seventy-one out of the 90 patients who presented with localized breast cancer were treated with adjuvant phy prior to the initiation of TEC chemotherapy and chemotherapy, 36 of these with a doxorubicin-containing following the completion of all courses of chemotherapy. A CAT scan of the brain was also performed prior to the regimen. Furthermore, 71 of these 90 patients also received a standard chemotherapy regimen once metastatic disease initiation of the first cycle of chemotherapy. Complete had been identified. The remaining 19 individuals with remission or response (CR) was defined as an absence of measurable disease and normalization of a previously metastatic disease all of whom had previously received adjuvant chemotherapy were treated either with radiation abnormal bone scan. Partial remission or response (PR) was therapy (n = 8) or surgery (n = 4), or did not receive any defined as у50% reduction in the size of bidimensionally measurable disease and/or a decrease in intensity or number treatment following the detection of metastases prior to the first cycle of TEC. Nine out of 10 patients with de novo of lesions seen on bone scan. Any response less than a PR metastatic disease received at least one standard chemo- was categorized as no response (NR). Progressive disease (PD) was defined as worsening of disease by clinical exam- therapy regimen prior to the first course of TEC. Therefore, all except one patient had already received standard chemo- ination and/or radiologic evaluation. Chemosensitivity or chemoresistance of metastatic breast cancer was ascertained therapy prior to entering this program. Doxorubicin- by its response (CR and PR vs NR and PD) to any standard containing regimens had been administered to 57 patients, taxane-containing regimens to one patient, and both chemotherapy regimen administered for the treatment of metastatic disease prior to the first cycle of TEC. Standard doxorubicin-containing and taxane-containing regimens to chemotherapy was defined as any regimen utilized com- 32 patients. Nine individuals received standard chemotherapy regimens which contained neither anthracyclines monly in the management of breast cancer. Toxicity was evaluated according to the World Health Organization guidelines.6 Table 2 Patient characteristics

Statistical analysis Number (range) Several variables were examined as potential determinants Patients 100 of achieving CR after TEC chemotherapy. These included Courses of TEC 257 age р or Ͼ than median (45 years), hormone receptor Median number of cycles of TEC/patient 3 (1–4) status, response to the most recent standard chemotherapy Subsequent stem cell transplantation 78 regimen administered for metastatic disease prior to TEC, Median age in years 45 (25–63) Median interval from diagnosis of breast cancer to 30 (2–175) intervals from diagnosis of breast cancer and diagnosis of TEC in months metastases to ﬁrst cycle of TEC (р of Ͼ than median Median interval from diagnosis of metastases to TEC 7 (1–153) interval), the number of prior chemotherapy regimens (р1 in months or Ͼ1), the number of organ systems involved (1 or Ͼ1), Prior radiation therapy 62 prior use of doxorubicin as part adjuvant chemotherapy, Median prior chemotherapy regimens 1 (0–4) Median prior chemotherapy cycles 8 (0–30) presence of visceral disease, and an ECOG performance Prior hormonal therapy 44 score of 0 vs 1, 2 or 3. Contingency table (␹2) analyses Number with visceral metastases 65 were utilized to estimate the statistical signiﬁcance of each Median number of organ systems involved 2 (1–6) discrete variable. Continuous variables were compared

Bone Marrow Transplantation TEC for stem cell mobilization in breast cancer S Bilgrami et al 126 nor taxanes. Hormonal therapy was administered to a total chemotherapy. Seven individuals did not receive any of 44 patients, in 18 each as part of either adjuvant treat- further TEC because of PD, intolerance of chemotherapy, ment or treatment of metastatic disease, and in eight indi- or early death. The median circulating CD34-positive cell viduals as part of both adjuvant therapy as well as treatment count was 3.6%. Leukapheresis was performed in 89 for metastatic disease. Thirty-seven of these individuals patients who had more than 1% circulating CD34-positive demonstrated hormone resistance. Two patients were intol- cells. A median of 4.7 ϫ 106/kg CD34-positive cells was erant of hormonal therapy and could not be evaluated for collected with a median of two leukapheresis procedures. response. Twenty-nine patients received radiation as part Less than 2 ϫ 106 CD34-positive cells/kg could be col- of adjuvant therapy, 27 for the treatment of metastatic dis- lected from only two individuals. The median CFU-GM ease, and 10 patients received radiation as part of both adju- colony count was 61/105 cells. Micro-metastases were vant therapy as well as for the treatment of metastatic dis- identified in two of 65 submitted samples. ease. Sites of metastatic disease included the skeletal When the yield of CD34-positive cells collected follow- system (61%), lungs (38%), lymph nodes (32%), liver ing both courses of TEC chemotherapy was combined (31%), skin (27%), pleura (13%), bone marrow (9%), per- using products which were devoid of micro-metastases, a itoneum (2%) and pericardium (2%). Visceral metastases total of 90 individuals achieved the target of 5 ϫ 106/kg were noted in 65 patient (57 had liver and/or lung or greater. The remaining nine evaluable patients failed to metastases) whereas the skin and/or lymph nodes were the mobilize an adequate number of PBSC. only sites of metastases in 18 individuals. The skeletal system was the only site of involvement in 17 patients. More Response than one organ system was involved with metastatic disease in most (67%) patients. Micro-metastases were detected in One individual suffered early mortality and could not be 10 out of 57 patients in whom samples of bone marrow evaluated for response to TEC. Furthermore, two other obtained prior to the first cycle of TEC were submitted for patients were also non-evaluable because both entered the immunocytochemical analysis. program in CR and maintained their remission status following TEC. Therefore, 97 patients were evaluable for Peripheral blood stem cell collection response to TEC (Table 4). The number of patients who achieved CR or PR following the completion of TEC was Four patients failed to mobilize PBSC (р0.1% CD 34-posi- 23 (23.7%) and 34 (35.1%), respectively. Nineteen patients tive cells) and could not undergo stem cell collection fol- (19.6%) had NR and 21 individuals (21.6%) demonstrated lowing the first cycle of TEC. An additional patient died PD. Seventy-eight patients underwent subsequent prior to recovery of her WBC count. In the remaining 95 HDC/PBSCT whereas the remaining 21 patients were patients, adequate (у1%) stem cell mobilization was deemed ineligible (NR or PD, 19 patients, intolerance of achieved in the peripheral blood on the day of WBC recov- chemotherapy, two patients). Eighty patients had received ery. Leukapheresis was performed in 63 of these 95 patients at least one standard chemotherapy regimen for the treat- following the first course of TEC and was deferred to the ment of metastatic disease prior to referral to our program. second cycle of chemotherapy in the remaining 32 patients. Seventy-one patients began TEC immediately after the The median yield of CD34-positive cells following the first completion of a standard chemotherapy regimen. Disease cycle of TEC was 6.2 ϫ 106/kg after a median of one leu- status in these 71 patients prior to the initiation of TEC was kapheresis (Table 3). In 16 patients, fewer than 2 ϫ 106 CR in two patients, PR in 31 patients, NR in 15 patients and CD34-positive cells/kg could be harvested. The median fre- PD in 23 patients. Nine other patients received their last quency of CFU-GM was 68.5/105 cells. Micro-metastases cycle of standard chemotherapy more than 6 months prior were detected in four of 54 samples of PBSC submitted for to the first course of TEC. The status of their disease upon analysis following the first cycle of chemotherapy. completion of their last standard chemotherapy regimen Ninety-three patients received a second course of TEC was CR in one patient, PR in five patients, NR in one patient and PD in two patients. All of these nine individuals developed PD before their first cycle of TEC. Therefore, Table 3 Stem cell leukapheresis in the entire group of 80 patients who had received at least one standard chemotherapy regimen for the treatment of Number (range) metastatic disease, 41 (16 patients with NR and 25 patients with PD) had failed to respond to the latest regimen Cycle 1 Cycle 2 (chemoresistant group) whereas 39 patients (three CR and

Number of patients evaluable 99 93 Median peripheral blood CD34+ % 5.7 (0–56.3) 3.6 (0–48.6) a Inadequate stem cell mobilization 4 4 Table 4 Response status following TEC Eligible for leukapheresis (patients) 95 89 Leukapheresis conducted (patients) 63 89 Complete response 23 Median CD34+ cells × 106/kg 6.2 (0–23.9) 4.7 (0.1–39.2) Partial response 34 collected No response 19 Median CFU-GM colony count/105 68 (6–Ͼ300) 61 (17–Ͼ300) Progressive disease 21 cells Overall response rate 59% Median days of leukapheresis 1 (1–3) 2 (1–6) aNinety-seven were evaluable for response.

Bone Marrow Transplantation TEC for stem cell mobilization in breast cancer S Bilgrami et al 127 36 PR) had responded to chemotherapy suggesting ‘chemo patients upon completion of TEC was as follows: CR in 24 sensitive disease.’ After excluding the three patients who patients; PR in 32 patients; NR in 16 patients; and PD in were not evaluable for response to TEC, the ‘chemosensiti- six patients. Restaging evaluation was performed 2 months ve’ group demonstrated a greater response to TEC chemo- following PBSCT and demonstrated CR in 18 patients, PR therapy. Following TEC, the ‘chemosensitive’ group had a in 17 patients, NR in 19 patients, and PD in 22 patients. CR rate of 40.54% (15 out of 37 patients) while the Two patients could not be evaluated for response because ‘chemoresistant’ group had a CR rate of 7.50% (three out of early non-relapse mortality. Furthermore, 17 out of the of 40 patients), P value Ͻ0.001 (Table 5). In the group of 24 patients transplanted in CR maintained their remission 20 patients who did not receive any standard chemotherapy status and were also not evaluable for response. In the for the treatment of metastatic disease, six individuals achi- remaining 59 patients, there was only one new complete eved a CR including the solitary patient who had not responder and 17 partial responders for an over-all response received any standard chemotherapy (adjuvant or for rate of 30.5% following HDC-PBSCT. metastases) prior to TEC. A number of variables was investigated in an attempt to predict which patients would enter CR at the completion of TEC chemotherapy (Table 5). Univariate analysis ident- Toxicity ified four significant variables for a patient to achieve CR. These included previously documented chemosensitivity Ͻ A total of 61 (23.7%) febrile episodes were observed fol- (P 0.001), involvement of only one organ system lowing 257 courses of TEC. These patients required (P ϭ 0.002), an ECOG performance score of 0 (P Ͻ 0.001) р readmission to the hospital as well as initiation of broad- and a short interval ( 8 months) from diagnosis of meta- spectrum antibiotic therapy. Bacteremia was documented static disease to first course of TEC (P ϭ 0.038). Response in only 10 instances (4%). Organisms causing bacteremia was also analyzed using a logistic regression model. Candi- included Streptococcus viridans (n = 3), Staphylococcus date predictors included interval from diagnosis of meta- epidermidis (n = 2), Eschericia coli (n = 1), Proteus mira- static disease to first cycle of TEC, number of organ bilis (n = 1), Xanthamonas maltophilia (n = 1), Acineto- systems involved, ECOG performance score and chemosen- = = sitivity of the tumor. The final model included the number bacter species (n 1), and Staphylococcus aureus (n 1). of organ systems involved, ECOG performance score and Three episodes of fungemia with Candida tropicalis were chemosensitivity terms. Involvement of two or more organ also observed. One patient died from hemodynamic col- systems increased the chances of less than a CR by 7.2- lapse during neutropenic fever following the first cycle of fold. Chemosensitivity reduced the chances of less than a TEC. However, no organisms could be isolated from this CR by nearly 92.6%. The correct classification rate for this patient and permission for autopsy was not granted. There model was 84.8%. were eight other readmissions for either vomiting, dehy- The median time to engraftment in the 78 patients who dration or catheter-related thrombosis. The total number of underwent HDC-PBSCT was day +9 (range +7to+12) fol- patients readmitted was 50. The median duration of lowing stem cell infusion. The median duration of neutro- readmission following TEC was 1.5 days yielding a total penia was 7 days (range 4 to 11 days). An unsupported of 4.5 days of hospitalization (3 additional days for admin- platelet count of 20 ϫ 109/l was observed on the same day istration of chemotherapy). as neutrophil recovery. The disease status of these 78 The incidence of grade 4 leukopenia was 100%. In fact, a total WBC count of Ͻ0.1 ϫ 109/l was noted after every cycle of TEC chemotherapy. The median interval from Table 5 Variables associated with complete response after TEC initiation of chemotherapy to recovery of ANC Ͼ1 ϫ 109/l P value was 14 days. The median duration of neutropenia was 6 days. Grade 3 thrombocytopenia was seen following 38 Univariate analysis cycles (14.8%) of chemotherapy whereas grade 4 toxicity Age in years р45 vs Ͼ45 0.993 was observed during the remaining courses. There was not Hormone receptor status positive vs negative 0.338 a single instance of grade 4 anemia. The median number Response to standard salvage chemotherapy sensitive vs Ͻ0.001 resistant of platelet products transfused per course of chemotherapy Interval from diagnosis to TEC in months р30 vs Ͼ30 0.081 was one. The median number of PRBC transfusions per Interval from metastases to TEC in months р7 vs Ͼ7 0.038 cycle of chemotherapy was also one. Red cell transfusions Number of prior chemotherapy regimens р1 vs Ͼ1 0.192 were not required during 101 cycles of TEC chemo- Number of organ systems involved 1 vs Ͼ1 0.002 Adjuvant doxorubicin vs no adjuvant doxorubicin 0.638 therapy (Table 6). ECOG performance score 0 vs 1, 2 and 3 Ͻ0.001 Grade 3 alopecia was noted in all evaluable patients. Visceral disease vs no visceral disease 0.063 Grade 3 stomatitis, nausea and vomiting, and diarrhea were seen during 26, 40 and 56 courses of chemotherapy, Multivariate analysis Chemosensitivity to standard chemotherapy 0.004 respectively. Only one patient developed hyperbilirubine- Involvement of a single organ system 0.009 mia during the course of neutropenic fever, which resulted ECOG performance score of 0 0.018 in mortality. There were 62 instances of grade 1 peripheral Shorter interval from metastases to TEC NS neuropathy. Mortality rate as a direct consequence of TEC was 1% per patient and 0.4% per cycle of chemotherapy. NS, not significant.

Bone Marrow Transplantation TEC for stem cell mobilization in breast cancer S Bilgrami et al 128 Table 6 Hematologic and infectious complications the completion of chemotherapy until recovery of the WBC count and/or collection of PBSC. Hematologic complications Either CFU-GM assays or peripheral blood CD34-posi- Initiation of chemotherapy to recovery of counts in 14 (11–19) tive cell counts can be used to ascertain the adequacy of days Duration of neutropenia in days 6 (4–12) stem cell collection. CD34-positive cell counts are being Grade 4 leukopenia 100% utilized increasingly because their assessment can be per- Grade 3 or 4 thrombocytopenia 100% formed rapidly and the results are readily reproducible. A Median platelet products transfused/cycle 1 (0–8) minimum of 0.5–2.0 ϫ 106 CD34-positive cells/kg has Median PRBC transfusions/cycle 1 (0–8) been postulated as necessary for subsequent hematopoietic 32,33 Infectious complications reconstitution. Furthermore, faster recovery of counts Number of chemotherapy cycles 257 is not observed after PBSCT if Ͼ8 ϫ 106 CD34-positive Febrile episodes Ͼ38°C 61 (24%) cells/kg are infused vs 5–8 ϫ 106 cells/kg.13 Risk factors Episodes of bacteremia 10 (4%) for poor stem cell mobilization include bone marrow Episodes of fungemia 3 (1%) involvement by malignancy, a slow rise of the WBC count following chemotherapy, numerous prior chemotherapy PRBC, packed red blood cells. cycles and regimens, an interval of less than 50 days from the most recent chemotherapy, prior use of radiation ther- Discussion apy, increasing age, stem cell mobilization with chemotherapy alone, diagnosis other than breast cancer especially The concentration of stem cells in the peripheral blood is acute myelogenous leukemia, metastatic breast cancer as only 1–10% that in the bone marrow.7 However, the num- opposed to locally advanced breast cancer, platelet count ber of circulating stem cells as determined by assays for Ͻ150 ϫ 109/l, and prior use of melphalan, BCNU or mito- CFU-GM, blood mononuclear cells or CD34-stained cells mycin-C.16,25,29,31,32,34,35 Ninety percent of our patients increases up to 20-fold during recovery from pancytopenia mobilized a sufficient number of PBSC despite the co-exist- following myelosuppresive chemotherapy.8,9 Numerous ence of numerous risk factors for poor mobilization strategies have been devised to increase the number of described above. The characteristics of those nine patients PBSC.10–14 Chemotherapy alone was utilized initially for who failed to mobilize the targeted number of PBSC are PBSC mobilization. The doses of chemotherapy adminis- outlined in Table 7. As is evident, these characteristics were tered for this purpose have been intense and have resulted not significantly different from the entire patient popu- in severe myelosuppression. The most commonly utilized lation (Table 2). agent has been cyclophosphamide (4–7 g/m2 i.v.) which has Peripheral blood stem cell harvesting has been performed yielded a 5- to 30-fold increase in the number of CD34- successfully following disease-oriented chemotherapy with positive cells or CFU-GM in the circulation at a median of CSF support.31,36,37 However, it has also been demonstrated 14–16 days following chemotherapy.15–17 The use of com- that an increase in dose intensity is associated with bination chemotherapy as opposed to single agent chemo- improved stem cell mobilization and yield.31 We chose a therapy enhances stem cell mobilization.18 The addition of regimen in which all three chemotherapeutic agents were paclitaxel or etoposide to cyclophosphamide has yielded active against breast cancer and two (paclitaxel and better results than cyclophosphamide alone.19 The emerg- cyclophosphamide) were moderately to highly active.38 The ence of colony-stimulating factors (CSF) for clinical use rationale was to utilize this dose-intense regimen not just has also been associated with improved stem cell mobiliz- for stem cell mobilization but also as disease-oriented ther- ation. The use of granulocyte–macrophage colony-stimulat- apy. Furthermore, several cycles of chemotherapy were ing factor (GM-CSF) as a single agent increases circulating administered to achieve maximal tumor cytoreduction prior CFU-GM by 8.5- to 18-fold.20,21 Similarly, G-CSF alone to HDC/PBSCT. Additionally, whenever possible, only enhances the number of circulating CFU-GM 58-fold.22 A stem cells collected following the second cycle of TEC comparison of G-CSF and GM-CSF demonstrated better were used for transplantation in order to reduce the results with G-CSF as a single agent.23 The addition of a possibility of tumor cell contamination of harvested PBSC. CSF to cyclophosphamide results in a 62.5- to 159-fold Patients already in CR at the time of HDC/PBSCT are increase in the number of circulating CFU-GM compared the ones most likely to have prolonged progression-free to a 30-fold elevation with cyclophosphamide alone.17,20,24 survival following PBSCT.3 The probability of achieving Furthermore, CSFs shorten the median time to reach peak a CR with first-line conventional chemotherapy for newly values of PBSC following chemotherapy.25 Additionally, it has been suggested that the use of CSF alone may fail to recruit the most primitive hematopoietic progenitors.26 Table 7 Characteristics of patients who did not mobilize stem cells Therefore, a combination of multi-agent chemotherapy and adequately CSF appears to be most effective in mobilizing PBSC. Fre- quently used chemotherapeutic agents include cyclophos- Number 9 phamide, etoposide, paclitaxel, ara-C, ifosfamide, cisplatin Median age in years (range) 49 (31–63) 18,19,25,27–31 Number of prior chemotherapy regimens 2 (1–4) and anthracyclines. In this study, we combined Number of prior chemotherapy cycles 8 (4–30) cyclophophamide with paclitaxel and etoposide because all Number with skeletal/marrow metastases (%) 7 (77%) three are excellent stem cell mobilizing agents. Granulocyte Number receiving radiation therapy previously (%) 5 (55%) colony-stimulating factor was administered daily following

Bone Marrow Transplantation TEC for stem cell mobilization in breast cancer S Bilgrami et al 129 diagnosed metastatic breast cancer is 15% (range 5–25%).38 erythropoietic differentiation in vitro: definition of three However, the chances of obtaining a CR are reduced by eythroid colony responses. Blood 1977; 49: 855–864. half with each subsequent disease progression.38 Almost all 6 Miller AB, Hoogstraten B, Staguet M, Winkler A. Reporting of our patients had already received one or two prior results of cancer treatment. Cancer 1981; 47: 207–214. chemotherapy regimens. Therefore, the CR rate of 23.7% 7 McCredie K, Hersh E, Friereich E. Cells capable of colony formation in the peripheral blood of man. Science 1971; 171: achieved with TEC in this group of patients compares fav- 293–294. orably with the results obtained with standard second-line 8 Richman CM, Weiner RS, Yankee RA. Increase in circulating chemotherapy regimens utilized in the treatment of meta- stem cells following chemotherapy in man. Blood 1976; 47: static breast cancer. Responsiveness to prior standard 1031–1039. chemotherapy, an ECOG performance score of 0 and 9 To LB, Haylock DN, Kimber RJ, Juttner CA. High levels of involvement by disease of only a single organ system was circulating hemopoietic stem cells in very early remission associated with a higher probability of attaining a CR from acute non-lymphoblastic leukemia and their collection after TEC. and preservation. Br J Haematol 1984; 58: 399–410. In conclusion, we report a novel, dose-intense combi- 10 Henon PR, Butturini A, Gale RP. Blood-derived hemato- nation of paclitaxel, etoposide and cyclophosphamide sup- poietic cell transplants: blood to blood? Lancet 1991; 337: ported by G-CSF which was utilized not just for stem cell 961–963. 11 Lowry PA, Tabbara IA. Peripheral hematopoietic stem cell mobilization but also for optimal tumor cytoreduction prior transplantation: current concepts. Exp Hematol 1992; 20: to HDC/PBSCT. The regimen was safe and well-tolerated 937–942. with a mortality rate of 1%. However, patients had to be 12 Gillespie TW, Hillyer CD. Peripheral blood progenitor cells followed closely during pancytopenia because of grade 4 for marrow reconstitution: mobilization and collection leukopenia and thrombocytopenia. Although the current strategies. Transfusion 1996; 36: 611–624. study did not compare TEC to standard disease-oriented 13 To LB, Haylock DN, Simmons PJ, Juttner CA. The biology chemotherapy in metastatic breast cancer, the CR rate of and clinical uses of blood stem cells. Blood 1997; 89: over 23% in this population of heavily pre-treated patients 2233–2258. was quite encouraging. The regimen was also effective in 14 Moog R, Muller N. Technical aspects and performance in col- mobilizing PBSC in the vast majority of individuals. lecting peripheral blood progenitor cells. Ann Hematol 1998; Finally, such a regimen could be utilized in the future to 77: 143–147. + select patients with metastatic breast cancer most likely to 15 Siena S, Bregni M, Brando M et al. Circulating CD34 hema- benefit from HDC/PBSCT such as those with a limited topoietic stem cells in the peripheral blood of high-dose cyclophosphamide-treated patients: enhancement by intravenous amount of chemosensitive disease who respond completely recombinant human granulocyte–macrophage colony-stimulat- to TEC. Furthermore, HDC/PBSCT induced only one ing factor. Blood 1989; 74: 1905–1914. additional CR beyond those achieved with TEC and failed 16 To LB, Shepperd KM, Haylock DN et al. Single high doses to maintain TEC-induced complete remissions in seven of of cyclophosphamide enable the collection of high numbers 24 patients. Therefore, it will also be interesting to ascertain of hemopoietic stem cells from the peripheral blood. Exp whether a dose-intense regimen such as TEC could obviate Hematol 1990; 18: 442–447. the need for HDC/PBSCT and its attendant complications 17 Tarella C, Ferrero D, Bregni M et al. Peripheral blood expan- in patients with metastatic breast cancer. sion of early progenitor cells after high-dose cyclophosphamide and rhGM-CSF. Eur J Cancer 1991; 27: 22–27. 18 Demirer T, Buckner CD, Storer B et al. Effect of different References chemotherapy regimens on peripheral-blood stem-cell collections in patients with breast cancer receiving granulocyte col- 1 Bezwoda WR, Seymour L, Dansey RD. High-dose chemo- ony-stimulating factor. J Clin Oncol 1997; 15: 684–690. therapy with hematopoietic rescue as primary treatment for 19 Demirer T, Rowley S. Buckner CD et al. Peripheral-blood metastatic breast cancer: a randomized trial. J Clin Oncol stem-cell collections after paclitaxel, cyclophosphamide, and 1995; 13: 2483–2489. recombinant human granulocyte colony-stimulating factor in 2 Stadtmauer EA, O’Neill A, Goldstein LJ et al. Phase III ran- patients with breast and ovarian cancer. J Clin Oncol 1995; domized trial of high dose chemotherapy (HDC) and stem cell 13: 1714–1719. support (SCT) shows no difference in overall survival or sev- 20 Socinski MA, Cannistra SA, Elias A et al. Granulocyte– ere toxicity compared to maintenance chemotherapy with macrophage colony stimulating factor expands the circulating cyclophosphamide, methotrexate and 5-fluorouracil (CMF) for hemopoietic progenitor cell compartment in man. Lancet women with metastatic breast cancer who are responding to 1988; I: 1194–1198. conventional induction chemotherapy: the Philadelphia 21 Haas R, Ho AD, Bredthauer U et al. Successful autologous Intergroup Study (PBT-1). Proc Am Soc Clin Oncol 1999; transplantation of blood stem cells mobilized with recombi- 18: 1a. nant human granulocyte–macrophage colony stimulating fac- 3 Antman KH, Rowlings PA, Vaughan WP et al. High-dose tor. Exp Hematol 1990; 18: 94–98. chemotherapy with autologous hematopoietic stem-cell sup- 22 Sheridan WP, Begley CG, Juttner CA et al. Effect of port for breast cancer in North America. J Clin Oncol 1997; peripheral-blood progenitor cells mobilized by filgrastim 15: 1870–1879. (G-CSF) on platelet recovery after high-dose chemotherapy. 4 Siena S, Bregni M, Belli N et al. Flow cytometry for clinical Lancet 1992; 339: 640–644. estimation of circulating hematopoietic progenitors for auto- 23 Peters WP, Rosner G, Ross M et al. Comparative effects of logous transplantation in cancer patients. Blood 1991; 77: granulocyte–macrophage colony-stimulating factor (GM-CSF) 400–409. and granulocyte colony-stimulating factor (G-CSF) on prim- 5 Gregory CJ, Eaves AC. Human marrow cells capable of ing peripheral blood progenitor cells for use with autologous

Bone Marrow Transplantation TEC for stem cell mobilization in breast cancer S Bilgrami et al 130 bone marrow after high-dose chemotherapy. Blood 1993; 81: blood stem cell mobilization from dose-intensive cyclophos- 1709–1719. phamide, etoposide, cisplatin plus G-CSF than from less inten- 24 Rosenfeld CS, Shadduck PK, Zeigler ZR et al. Cyclophos- sive chemotherapy regimens. Bone Marrow Transplant 1999; phamide mobilized peripheral blood stem cells in patients with 23: 111–117. lymphoid malignancies. Bone Marrow Transplant 1995; 15: 32 Bensinger W, Appelbaum F, Rowley S et al. Factors that 433–438. influence collection and engraftment of autologous peripheral- 25 Brugger W, Bross K, Frisch J et al. Mobilization of peripheral blood stem cells. J Clin Oncol 1995; 13: 2547–2555. blood progenitor cells by sequential administration of interleu- 33 Weaver CH, Hazelton B, Birch R et al. An analysis of kin-3 and granulocyte–macrophage colony-stimulating factor engraftment kinetics as a function of the CD34 content of peri- following polychemotherapy with etoposide, ifosfamide, and pheral blood progenitor cell collections in 692 patients after cisplatin. Blood 1992; 79: 1193–1200. the administration of myeloblative chemotherapy. Blood 1995; 26 Schneider JG, Crown JP, Wasserheit C et al. Factors affecting 86: 3961–3969. the mobilization of primitive and committed hematopoietic 34 Olivieri A, Offidani M, Ciniero L et al. Optimization of the progenitors into the peripheral blood of cancer patients. Bone yield of PBSC for autotransplantation mobilized by high-dose Marrow Transplant 1994; 14: 877–884. chemotherapy and G-CSF: proposal for a mathematical model. 27 Schimazaki C, Oku N, Ashihara E et al. Collection of peri- Bone Marrow Transplant 1994; 14: 273–278. pheral blood stem cells mobilized by high-dose Ara-C plus 35 Bensinger WI, Longin K, Appelbaum F et al. Peripheral blood VP-16 or aclarubicin followed by recominant human granulo- stem cells (PBSCs) collected after recombinant granulocyte cyte-colony stimulating factor. Bone Marrow Transplant colony stimulating factor (rhG-CSF): an analysis of factors 1992; 10: 341–346. correlating with the tempo of engraftment after transplan- 28 Menichella G, Pierelli L, Scambia G et al. Low-dose cyclo- tation. Br J Haematol 1994; 87: 825–831. phosphamide in combination with cisplatin or epirubicin plus 36 Elias AD, Ayash L, Anderson KC et al. Mobilization of peri- rhG-CSF allows adequate collection of PBSC for autotrans- pheral blood progenitor cells by chemotherapy and granulo- plantation during adjuvant therapy for high-risk cancer. Bone cyte–macrophage colony-stimulating factor for hematologic Marrow Transplant 1994; 14: 907–912. support after high-dose intensification for breast cancer. Blood 29 Weaver CH, Schwartzberg LS, Birch R et al. Collection of 1992; 79: 3036–3044. peripheral blood progenitor cells after the administration of 37 Kohno A, Takeyama K, Narabayashi M et al. Low-dose gra- cyclophosphamide, etoposide, and granulocyte-colony-stimul- nulocyte colony-stimulating factor enables the efficient collec- ating factor: an analysis of 497 patients. Transfusion 1997; tion of peripheral blood stem cells after disease-oriented, con- 37: 896–903. ventional-dose chemotherapy for breast cancer, malignant 30 Schwartzberg LS, Weaver CH, Birch R et al. A randomized lymphoma and germ cell tumor. Bone Marrow Transplant trial of two doses of cyclophosphamide with etoposide and G- 1995; 15: 49–54. CSF for mobilization of peripheral blood stem cells in 318 38 Harris JR, Morrow M, Norton L. Malignant tumors of the bre- patients with stage II–III breast cancer. J Hematother 1998; ast. In: DeVita VT Jr, Hellman S, Rosenberg SA (eds). Can- 7: 141–150. cer. Principles and Practice of Oncology. Lippincott-Raven: 31 Steward DA, Guo D, Morris D et al. Superior autologous Philadelphia, 1997, pp 1557–1616.

Bone Marrow Transplantation