Followed by Autologous Hematopoietic Stem Cell Re

Home , Lonidamine

Phase II study of high-dose cyclophosphamide, etoposide, and carboplatin (CEC) followed by autologous hematopoietic stem cell rescue in women with metastatic or high-risk non-metastatic breast cancer: multivariate analysis of factors affecting survival and engraftment

TR Klumpp, SL Goldberg, AJ Magdalinski and KF Mangan

Temple University Bone Marrow Transplant Program, Division of Medical Oncology, Department of Medicine, Temple University School of Medicine, Philadelphia, PA, USA

Summary: Breast cancer remains an important source of morbidity and Seventy women with high-risk stage II (n = 10), IIIA mortality among women, with an estimated 46 000 deaths 1 (n = 12), IIIB (n = 11), or IV (n = 37) breast cancer annually in the United States alone. Despite extensive received cyclophosphamide 6000 mg/m2, etoposide research, recently published studies continue to report zero 2400 mg/m2, and carboplatin 1200 mg/m2 followed by or near-zero long-term progression-free survival rates 2–7 infusion of autologous hematopoietic stem cells (AHSC). among women with metastatic disease and less than 50% Women with high-risk stage II disease had eight or long-term survival among women with locoregional dis- 8,9 more involved axillary lymph nodes (n = 9) or axillary ease. For example, a recently published 20-year follow- 8 and breast relapse following lumpectomy, chemo- up analysis of a major adjuvant chemotherapy trial therapy, and radiation therapy (n = 1). Women with reported 37% long-term progression-free survival among measurable stage III or stage IV disease were required women with one to three involved axillary lymph nodes, to demonstrate complete or partial response to conven- 26% progression-free survival among women with four to tional-dose chemotherapy prior to transplant. The over- 10 axillary nodes involved, and no long-term progression- all (complete plus partial) response rate for the 31 free survivors among women with more than 10 involved patients not in complete remission at the time of trans- nodes. Similarly, a recently reported large chemotherapy plant was 55%. With a median follow-up of 545 days, trial in women with previously untreated metastatic disease the 2-year actuarial progression-free survival rates for reported a 0% survival rate at less than 5 years.7 patients with stage II, IIIA, IIIB and IV are 86, 75, 42 Disappointing long-term results with conventional-dose and 13%, respectively. Factors independently predictive therapy, as well as both controlled and uncontrolled studies of longer progression-free survival by multivariate suggesting a chemotherapeutic dose–response relationship analysis included lower stage disease, status of disease for breast cancer within conventional dose ranges7–9 have at transplant (in CR vs not in CR), and positive estrogen led to the introduction of high-dose therapy (HDC) fol- receptor status. Factors predictive of more rapid neu- lowed by rescue with autologous blood or marrow-derived trophil engraftment by multivariate analysis included hematopoietic stem cells (AHSC) in women with metastatic post-transplant administration of hematopoietic growth or high-risk non-metastatic breast cancer.10–33 factors, greater number of infused CFU-GM, mobiliz- We elected to study the combination of high-dose cyclo- ation with G-CSF or cyclophosphamide/G-CSF (vs phosphamide, etoposide, and carboplatin (CEC) for a phase mobilization with GM-CSF or no mobilization), and II study in this patient population for several reasons: (1) lower stage disease. Only one patient (1.4%) died prior The efficacy of high-dose alkylating agent-based combi- to day 100 from any cause. High-dose cyclophospham- nations in inducing high rates of response in individuals ide, etoposide, and carboplatin followed by infusion of with breast cancer is well-established.10–33 (2) The synergy AHSC constitutes an active and well-tolerated regimen of the platinum–etoposide combination is well-established in the treatment of women with high-risk non-meta- in the treatment of a variety of other tumor types.34–36 (3) static or metastatic breast cancer. At least one group has documented the efficacy of the plati- Keywords: breast cancer; hematopoietic stem cell trans- num–etoposide combination in conventional doses in the plantation; bone marrow transplantation; cyclophospha- treatment of women with metastatic breast cancer.37–39 (4) mide; etoposide; carboplatin Our recently published phase I/II trial40 had established maximum-tolerated doses as well as excellent tolerability and efficacy of a similar combination (CEP, ie high-dose Correspondence: Dr TR Klumpp, Bone Marrow Transplant Program, Temple University Cancer Center, 3322 N. Broad Street, Philadelphia, PA cyclophosphamide, etoposide, and cisplatin) in patients 19140, USA with relapsed or refractory lymphoma. Ifosfamide-contain- Received 12 February 1997; accepted 1 May 1997 ing high-dose regimens were not in widespread use at the CEC for breast cancer TR Klumpp et al 274 time that the study was being designed,10 and were not con- Table 1 Patient characteristics sidered for use in the present trial. n (%)

Patients and methods Age at transplant (years) 20–29 1 (1.4) 30–39 21 (30.0) Eligibility 40–49 34 (48.6) 50–59 12 (17.1) Eligibility requirements included a histologically proven over 60 2 (2.9) diagnosis of adenocarcinoma of the female breast, age less Stage of disease than or equal to 65 years, left ventricular ejection fraction II 10 (14.3) IIIA 12 (17.1) greater than 50%, serum bilirubin less than 2 mg/dl, serum IIIB 11 (15.7) transaminases less than two times the upper limit of normal, IV 37 (52.9) serum creatinine less than 1.8 mg/dl, negative HIV status, No. of cycles of chemotherapy prior to transplant less than 50% marrow replacement with tumor, and willing- Zero 5 (7.1) ness and ability to provide verbal and written informed con- 1–3 13 (18.6) 4–6 23 (32.9) sent. Women with stage II disease with eight or more 7–9 16 (22.9) involved axillary lymph nodes, stage III disease, or stage 10+ 13 (18.6) IV disease were eligible; disease was staged according to Estrogen receptor status standard American Joint Commission on Cancer Staging Negative 24 (34.3) 41 Positive 43 (61.4) criteria. For the purposes of this study, patients with Unknown 3 (4.3) locally relapsed disease were not considered to have stage Progesterone receptor status IV disease unless at least one distant metastatic site was Negative 31 (44.3) also involved. Examination of the bone marrow for tumor Positive 37 (52.9) was performed primarily by routine light microscopy. Unknown 2 (2.9) Menopausal status at diagnosis Patients with measurable disease were required to have Premenopausal 53 (75.7) a partial or complete response to a minimum of three cycles Perimenopausal 6 (8.6) of conventional-dose chemotherapy prior to transplant; for Postmenopausal 11 (15.7) patients with disease involving only the bone cortices, Disease status at transplant In complete remission 39 (55.7) stabilization of bone scans over a minimum of three cycles Not in complete remission 31 (44.3) were considered eligible. Prior chest irradiation This protocol was reviewed and approved by the Temple No 42 (60.0) University Institutional Review Board (IRB). Yes 28 (40.0) Prior surgery Biopsy only 3 (4.3) Treatment protocol Lumpectomy, axillary node dissection 9 (12.9) Modified radical mastectomy, Autologous hematopoietic stem cells were obtained from axillary node dissection 57 (81.4) Simple mastectomy, the bone marrow and/or peripheral blood. In general, axillary node dissection 1 (1.4) patients treated in the early portion of the study received Source of hematopoietic stem cells both marrow and blood-derived stem cells whereas later in Peripheral blood alone 39 (55.7) the study (following the publication of additional studies Combined marrow and blood 30 (42.9) documenting the long-term engraftment potential of PBSC) Marrow alone 1 (1.4) Mobilization regimen patients received blood-derived stem cells only. Bone mar- G-CSF alone 50 (71.4) row harvesting was performed by multiple aspirates from Cyclophosphamide-G-CSF 12 (17.1) the posterior iliac crests. GM-CSF alone 7 (10.0) Three stem cell mobilization studies were approved by Nonea 1 (1.4) the Temple University IRB and utilized as accessory proto- a cols during the course of this study. Thus, blood-derived One patient (UPN 108) was rescued with bone marrow alone; this patient did not undergo mobilization. stem cells were obtained via multiple peripheral blood leu- kapheresis procedures following: (1) mobilization with G- CSF 10 ␮g/kg/day s.c. administered for 4 days; or (2) G- 12 h for six doses on day −6 to day −4 for a total dose of CSF preceded by the administration of cyclophosphamide 2400 mg/m2, and carboplatin 300 mg/m2 per day as 24-h 3000 mg/m2 i.v. × 1; or (3) GM-CSF as previously intravenous infusions beginning on day −6 to day −3 for a decribed42 (Table 1). One patient (UPN 108) underwent total dose of 1200 mg/m2. Dose adjustments based on ideal bone marrow harvesting alone and was thus not mobilized. body weight were made for overweight patients. DMSO- Leukaphereses, progenitor cell enumeration, and cryo- cryopreserved stem cells were infused beginning on day 0. preservation were performed as previously described.42 Collections greater than approximately 10 cc/kg were split The high-dose regimen consisted of cyclophosphamide into two aliquots and infused on day 0 and day +1. 1500 mg/m2 per day as a 1-h intravenous infusion daily for Decisions regarding pre- and/or post-transplant adminis- four days from day −6 to day −3 for a total dose of tration of radiation therapy were made on a patient-by- 6000 mg/m2, etoposide 400 mg/m2 as 1-h infusion every patient basis by the individual attendings based primarily CEC for breast cancer TR Klumpp et al 275 on the perceived risk of relapse in areas of known disease Statistical methods and the perceived ability of the bone marrow to tolerate radiation in the post-transplant period. All raw data were subjected to detailed computer-assisted error detection prior to the initiation of the statistical analysis. Data were analyzed according to the most recent follow-up as of 15 July 1996. Overall and progression-free Supportive care survival rates and neutrophil engraftment rates were calcu- 44 Patients were treated in a dedicated inpatient bone marrow lated according to the method of Kaplan and Meier. Uni- transplant unit in single-bed hospital rooms equipped with variate and multivariate associations involving the above variables were evaluated with the log rank test, and Cox laminar airflow (LAF) or high efficiency positive pressure 44 air (HEPA) filtration. The prophylactic antibiotic regimen proportional hazards regression, respectively. in all patients consisted of ciprofloxacin 500 mg p.o. twice Patients tabulated as having been in ‘complete remission a day administered from hospital admission to neutrophil at any time prior to transplant’ may or may not have been engraftment, acyclovir 800 mg p.o. twice a day adminis- in remission at the actual time of the transplant – as detailed tered from day −3 to day +100, and fluconazole 400 mg below, a separate variable to capture the latter data element p.o. daily from hospital admission to neutrophil was included and analyzed separately with respect to its engraftment. Initial coverage for neutropenic fevers con- effect on survival. sisted of ceftazidime 1 g i.v. q8hinconjunction with All P values are two-sided. All data entry, data checking, vancomycin 1 g i.v. q 12 h, with amphotericin B 0.3– and data analyses were performed with SAS statistical 0.8 mg/kg/day added for refractory fevers or deteriorating software (SAS Institute, Cary, NC, USA). clinical status. Patients received G-CSF 5 ␮g/kg/day beginning at day +1 as a short i.v. infusion and continued to the date of stable neutrophil engraftment, except for 15 patients Results who received no HGF in the post-transplant period due to participation in a previously published randomized trial,43 Patient characteristics and one patient (UPN 108) who received GM-CSF 7 ␮g/kg/day starting at day +1. Seventy patients were transplanted on this protocol between 26 November 1991 and 31 October 1995. The median age was 42 years (range 26–63). As of the closing date (15 July 1996), the median follow-up of surviving patients was 545 Evaluation of engraftment, response and toxicity days (range 278–1563 days). Patients had received a The date of neutrophil engraftment was taken as the date median of six cycles of conventional-dose chemotherapy on which the absolute peripheral blood neutrophil count prior to transplant (range 0–20). The median ECOG per- first recovered to a sustained count of 0.5 × 109/l or greater. formance status at transplant was 0 (range 0–1). Patients The date of platelet engraftment was taken as the first date with stage II disease had a median of 11 positive lymph on which the patient achieved a sustained unsupported peri- nodes (range 8–23 except for one patient (UPN 134) who pheral blood platelet count of 20 × 109/l. The number of had only two positive axillary lymph nodes at diagnosis but days required to achieve neutrophil and platelet was judged to have high-risk disease following the develop- engraftment were calculated based on the date on which ment of progressive disease in the breast and axilla shortly the infusion of stem cells began. following the completion of lumpectomy, axillary lymph Routine disease restaging was scheduled at 3, 6 and 12 node dissection, six cycles of conventional chemotherapy, months following transplant, and annually thereafter, and and 4680 cGy irradiation to the chest wall and axilla). consisted of a history and physical examination, a complete Among patients with stage IV disease, 62% had bone cor- blood chemistry panel, a CT scan of the chest, abdomen tex involvement, 22% had bone marrow involvement, 14% and pelvis, and a bone scan. Additional interim restaging had liver involvement, 32% had lung involvement, 24% took place as clinically indicated. Complete response was had mediastinal involvement, and 38% had involvement of defined as disappearance of all evidence of active disease. non-axillary lymph nodes. Other patient characteristics are For the purposes of this study, disappearance of all measur- shown in Table 1. able disease in conjunction with stable or resolving bone scans over a period of a minimum of 3 months from the Toxicity date of the transplant was also defined as a complete response. Partial reponse was defined as a 50% or greater As summarized in Table 2, commonly observed toxicities reduction in cross-sectional diameter of all measurable included neutropenia, thrombocytopenia, anemia, mucos- lesions. itis, nausea, vomiting, diarrhea, and skin rash. All patients Toxicities were graded according to standard Southwest developed at least one episode of febrile neutropenia requir- Oncology Group (SWOG) toxicity grading criteria except ing non-prophylactic antibiotics. for oral or esophageal mucositis which were graded as fol- One patient (UPN 147), a 58-year-old woman with stage lows: grade 1, mild erythema or discomfort; grade 2, mod- IIIA disease who was receiving low-dose steroids for mild erate discomfort; grade 3, patient required parenteral nar- pre-existing systemic lupus erythematosis at the time of the cotics for pain control; grade 4, grossly hemorrhagic transplant, died at day +53 following the development of mucositis. hepatic veno-occlusive disease, gram-positive bacteremia, CEC for breast cancer TR Klumpp et al 276 Table 2 Summary of observed toxicitiesa Table 3 Relationship between selected factors and rate of neutrophil engraftment Gradeb Days to 500 P valueb a 012345 neutrophils

+ c Cardiac 65 4 0 1 0 0 CD34 cells infused × 6 = Cystitis 68 1 0 1 0 0 Ͻ8 10 /kg (n 18) 11 (10–24) 0.5431 × 6 = Dermatologic 28 18 17 7 0 0 у8 10 /kg (n 20) 11 (9–23) c Diarrhea 4 37 28 1 0 0 CFU-GM infused × 4 = d Hemorrhage 36 26 4 3 1 0 Ͻ2.5 10 /kg (n 33) 12.0 (10–24) 0.0066 × 4 = Hepatic 58 5 3 3 1 0 у2.5 10 /kg (n 36) 10.5 (9–22) Infectionc 0006731 Growth factors post-transplant = d Mucositis 8 11 15 34 2 0 No (n 15) 14 (10–24) 0.0006 Yes (n = 55) 11 (9–24) Neutropenia 0 0 0 0 70 0 e Neurologic 68 0 0 1 1 0 Irradiation prior to transplant = Pulmonary 57 3 2 5 3 0 No (n 42) 11 (9–24) 0.3060 = Renal 65 4 1 0 0 0 Yes (n 28) 12 (9–22) Thrombocytopenia 0 0 1 1 68 0 Mobilization regimen = f Vomiting 2 37 27 4 0 0 CY-GCSF (n 12) 11 (9–15) G-CSF (n = 50) 11 (9–24) 0.0725d,g GM-CSF (n = 7) 14 (12–22) a Figures indicate the number of patients observed with the respective toxi- Mononuclear cells infusedc city grades. Ͻ3 × 108/kg (n = 31) 11 (9–24) 0.7736 b Graded according to standard SWOG toxicity criteria except for mucositis у3 × 108/kg (n = 39) 12 (9–24) as detailed in Patients and methods section. c c Nucleated cells infused Neutropenic fever requiring i.v. antibiotics scored as grade 3 infection by Ͻ8 × 108/kg (n = 32) 11 (9–24) 0.8814 SWOG definition. у8 × 108/kg (n = 38) 11 (9–24) No. of prior cycles of chemotherapy р6(n=41) 11 (9–22) 0.0289 septic shock and multisystem failure. No other patient died Ͼ6(n=29) 12 (9–24) of any cause prior to day 100, and all subsequent deaths Source of stem cells were attributable to progressive disease. Blood (n = 39) 11 (9–24) Marrow (n = 1) 15 (NA) 0.6048 Four additional patients (UPN 273, UPN 300, UPN 339 Combined (n = 30) 11 (9–22) and UPN 372) required transient ICU admission for, Stage of disease respectively, bacteremia due to vancomycin-resistant II (n = 10) 11 (9–12) enterococcus, immediate-onset hypotensive and hypoxic IIIA (n = 12) 11 (9–21) 0.0474d = reaction to a platelet transfusion, apparent septic shock IIIB (n 11) 12 (9–19) IV (n = 37) 12 (9–24) responding to antibiotics but with negative cultures, and transient respiratory failure of uncertain etiology despite aExpressed as median and range. extensive workup. Of note, all four of the latter patients bGlobal univariate P value by log-rank test. remain alive (two without evidence of disease) at day +599, cCells/kg based on ideal body weight. CD34 and CFU-GM data not avail- day +691, day +492 and day +442 post-transplant, respect- able for all patients. Analyses utilizing lower CD34 cutoffs were not stat- ively. istically feasible because fewer than 10% of patients received less than 5 million CD34/kg. One patient (UPN 146) suffered an easily reversed epi- dIndependently predictive of rate of neutrophil engraftment in multivariate sode of apparent anaphylaxis during the infusion of stem Cox proportional hazards model. cells and did not require transfer to the ICU. One patient eIrradiation to breast, chest wall, and/or axilla. (UPN 385) suffered transient hyponatremia with a sodium fCyclophosphamide 3000 mg/m2 followed by G-CSF 10 ␮g/kg/day. gWilcoxon rank-sum analysis revealed statistically significant acceleration of 115. One patient (UPN 274) developed progressive in neutrophil engraftment for both CY-GCF and G-CSF alone over multifocal neurologic deficits of uncertain etiology in con- GM-CSF. junction with rapid multifocal progression of her underly- ing cancer. Three patients (in addition to UPN 147, decribed above) transiently satisfied the Seattle criteria45 for poietic growth factors following transplant; (2) infusion of hepatic veno-occlusive disease but all recovered a higher number of CFU-GM/kg; (3) mobilization with G- spontaneously. CSF or cyclophosphamide plus G-CSF (vs mobilization with GM-CSF or no mobilization); and (4) lower stage disease (Table 3). Engraftment and transfusion support The median number of days required to attain an unsup- All 70 patients achieved sustained neutrophil engraftment. ported peripheral blood platelet count of 20 000/␮lor The median time to achieve 500 polys/␮l in the peripheral greater was 14 days (range 9–39 days); the rate of platelet blood was 11 days (range 9–24 days). Associations between engraftment was highly correlated with the rate of neutro- selected factors and the rate of neutrophil engraftment are phil engraftment (P Ͻ 0.001). The median number of plate- shown in Table 3. Multivariate Cox proportional hazards let transfusions administered was four (interquartile range analysis revealed that the following factors were indepen- 2–6). dently predictive of a shorter interval between transplant The median number of units of red cells transfused was and neutrophil engraftment: (1) administration of hemato- four (interquartile range 4–6). The median interval between CEC for breast cancer TR Klumpp et al 277 the transplant and the date of the last red cell transfusion Discussion was 10.5 days (range 2–41 days). This study demonstrates that the combination of high-dose cyclophosphamide, etoposide, and carboplatin (CEC) con- Response and survival stitutes an active and well-tolerated regimen in the treatment of women with high-risk non-metastatic or metastatic Fifty of the 70 patients (71%) achieved or remained in com- breast cancer. As noted, only one of the 70 patients (1.4%) plete remission for a minimum of 3 months following the died of any cause prior to day 100, and the 2-year actuarial transplant, whereas two patients (3%) achieved a partial progression-free survival rates among patients with high- remission, three patients (4%) were judged to have stable risk stage II, stage IIIA, stage IIIB and stage IV disease disease for a minimum of 3 months, and 14 patients (20%) were 86, 75, 42 and 13%, respectively. developed progressive disease within 3 months following The incidence of treatment-related mortality (1.4%) the transplant. One patient (UPN 147, 1.4%) was not evalu- observed in the present study is lower than has able for reponse due to death prior to day 100. Of the 31 been reported in many recently reported studies patients who were not in complete remission at the time of high-dose therapy followed by autologous stem cell of transplant, 15 (48%) achieved a complete remission a or bone marrow rescue in women with breast can- minimum of 3 months following transplant. cer.11,12,14,15,17,19,20,22,23,26,27,29,32,33 For example, in a recently The actuarial median overall survival of the study popu- reported phase II study27 of the popular10,17,26–28 cyclophos- lation as a whole was 719 days. As of the closing date phamide-cisplatin-carmustine (CCB) regimen, Peters et al27 the actuarial median overall survival had not been reached reported 10 treatment-related deaths among 102 women among stage II or stage IIIA patients, and was 1165 days with high-risk breast cancer. Although comparisons among and 575 days for stage IIIB patients and stage IV phase II trials must be made with caution, the low treat- patients, respectively. ment-related mortality observed in the present trial, along The 2-year actuarial progression-free survival rates for with the observation that the stage-specific response and patients with stage II, stage IIIA, stage IIIB, and stage IV progression-free survival rates observed in the present trial disease are 86, 75, 42 and 13%, respectively. Associations are comparable to those reported in other recently reported between other factors and the probability of progression- trials of HDC in women with metastatic or non-metastatic free survival are summarized in Table 4. Factors indepen- breast cancer,10–33 and the favorable efficacy and toler- dently predictive of progression-free survival included: (1) ability observed in several other recently reported trials of stage of disease prior to transplant; (2) status of disease (in high-dose CEC in the treatment of a variety of malig- CR vs not in CR) at the time of the transplant; and (3) nancies,46–52 suggest that a phase III trial comparing CCB estrogen receptor status (Table 4). with CEC may be warranted. Multivariate analysis of progression-free survival among Two recently published smaller studies also suggest that patients with stage IV disease revealed that estrogen recep- high-dose CEC has a faborable risk–benefit ratio in women tor status, number of sites of metastatic disease, and disease with breast cancer. In 1993, Kritz et al24 reported treating status at transplant were independently predictive of 18 patients with metastatic breast cancer with high-dose progression-free survival within this subset. CEC. Of note, eight of the 18 patients (44%) remained alive Nine patients had developed chest wall involvement and progression-free 10–18 months post-transplant, a figure prior to the transplant but had not developed distant meta- which compares favorably with other recently reported stases and were considered to have non-metastatic disease trials of high-dose therapy for metastatic disease.10–26,28–33 for the purposes of this study. The median progression-free More recently, Broun et al15 reported treating 28 women survival among these nine patients was 2.09 years. In con- with metastatic breast cancer with tandem autotransplan- trast, the 24 patients with non-metastatic disease without tation utilizing high-dose CEC. The progression-free sur- prior chest wall involvement had not reached their median vival rate was 32% with a median follow-up of 23 months, progression-free survival as of the closing date (P = 0.15). and only one patient (3.6%) died from treatment-related causes. Sites of progression Few published trials of HDC followed by autologous HSC rescue in patients with metastatic or high-risk non- The correlation between sites of involvement prior to trans- metastatic breast cancer have included formal analyses of plant and sites of progression following transplant was vari- factors predictive of survival or engraftment rate. As noted, able depending upon the site. For example, 15 of the 26 factors found to be independently predictive of progression- patients with progression involving the bone cortex (58%) free survival in the present study included stage of disease, had recognized bone cortex involvement prior to transplant, status of disease at transplant (in CR vs not in CR), and whereas only three of the 13 patients with progression estrogen receptor status. Recently, Ayash et al,12 in a study involving the liver (23%) had recognized liver involvement of 62 women with metastatic breast cancer undergoing prior to transplant. Seven of the 15 patients with pro- HDC followed by infusion of AHSC, reported that attain- gression involving the lung parenchyma (47%) had recog- ment of a CR with induction chemotherapy and involve- nized lung parenchymal involvement prior to transplant. ment of a single (as opposed to multiple) metastatic site There was no correlation between the administration of were independently predictive of prolonged progression- radiation therapy in the pre- or post-transplant period and free survival. the incidence of relapse in bone. In contrast to the study of Ayash et al,12 Dunphy et al,16 CEC for breast cancer TR Klumpp et al 278 Table 4 Relationship between selected factors and progression-free survival

Parameter Median PFSa (days) P valueb

Age at transplant Ͻ 45 years (n = 39) 763 у 45 years (n = 31) 279 0.0894 Bone cortex as sole site of metastatic diseasec No (n = 28) 221 Yes (n = 9) 202 0.1930 Bone cortex involvement at any time prior to transplantc No (n = 14) 220 Yes (n = 23) 216 0.8895 Bone marrow involvement at any time prior to transplantc No (n = 29) 215 Yes (n = 8) 223 0.9368 Disease status at transplant In CR (n = 39) 800 Not in CR (n = 31) 202 0.0001d ECOG performance status at transplant 0(n=67) 351 1(n=3) 230 0.4521 Estrogen receptor status Negative (n = 27)e 215 Positive (n = 43) 415 0.1600d Growth factor support post-transplant No (n = 15) 216 Yes (n = 55) 415 0.0881 In complete remission at any time prior to transplant No (n = 31) 173 Yes (n = 39) 518 0.0004 Interval between progression to stage IV and transplantc Ͻ 1 year (n = 24) 199 у 1 year (n = 13) 230 0.6094 Interval from diagnosis to transplant Ͻ 1 year (n = 34) NRf у 1 year (n = 36) 228 0.0001 Liver involvement at any time prior to transplantc No (n = 32) 221 Yes (n = 5) 120 0.7054 Lung parenchymal involvement at any time prior to transplantc No (n = 25) 225 Yes (n = 12) 105 0.1300 Mobilization regimen CY-Gg (n = 12) 234 Otherh (n = 58) 393 0.3772 Menopausal status at diagnosis Pre (n = 53) 393 Peri/Post (n = 17) 290 0.3553 Non-axillary lymph node(s) as sole site of metastatic diseasec No (n = 33) 202 Yes (n = 4) 252 0.9772 Nucleated cells infused Ͻ8 × 108/kg (n = 32) 415 у8 × 108/kg (n = 38) 314 0.7075 No. of cycles of chemotherapy to transplant Ͻ 6(n=29) NRf у 6(n=41) 238 0.0056 No. of relapses prior to transplant Zero (n = 35) NRf 1+ (n = 35) 230 0.0002 No. of sites of metastatic diseasec One (n = 17) 225 2+ (n = 20) 199 0.0816 Progesterone receptor status Negative (n = 33)i 339 Positive (n = 37) 406 0.4677 Radiation to chest wallj No (n = 24) 297 Yes (n = 46) 393 0.8677

(Continued) CEC for breast cancer TR Klumpp et al 279 Table 4 Continued

Ratio of actual to ideal body weight Ͻ1.3 (n = 41) 415 у1.3 (n = 29) 289 0.2477 Source of stem cells PBSC alone (n = 39) 290 PBSC + marrowk (n = 31) 351 0.6844 Stage of disease II (n = 10) 1011 IIIA (n = 12) NRf,l IIIB (n = 11) 518 IV (n = 37) 216 0.0001d Treated in LAFm No (n = 45) 415 Yes (n = 25) 339 0.3992 aActuarial median progression-free survival estimates calculated from the day of transplant to the day of progression as per the method of Kaplan and Meier. b Global univariate P values. cAnalyzed among stage IV patients only. dIndependently predictive of progression-free survival in multivariate Cox proportional hazards model. eIncludes three patients with equivocal or unknown estorgen receptor status. fActuarial median progression-free survival not reached. gCyclophosphamide 3000 mg/m2, followed by G-CSF 10 ␮g/kg/day. hIncludes one patient who underwent conventional bone marrow harvesting not preceded by growth factor administration. iIncludes two patients with unknown or equivocal progesterone receptor status. jIncludes prophylactic radiation to the ipsilateral breast or chest wall in the pre-transplant period or in the early post-transplant period prior to any observed relapse. kIncludes one patient who received bone marrow alone. lEight of the 10 stage II patients (80%) were alive and without evidence of disease at the closing date, vs nine of the 12 stage IIIA patients (75%). However, the actuarial median progression-free survival was longer with stage IIIA due to relapse of a single stage II patient (UPN 134) on the late (unstable) portion of the Kaplan–Meier curve (see Ref. 18). mPatients tabulated as treated in LAF were placed in a laminar airflow room equipped with a high-efficiency particular air (HEPA) filter on the day of the transplant and removed at the time of neutrophil engraftment. in a study of 80 women undergoing HDC followed by independent statistical significance for each of the above infusion of AHSC for metastatic breast cancer, reported factors except for stage of disease which was not studied. three entirely different factors as having independent prog- In summary, the present study suggests that high-dose nostic significance with regard to overall survival, ie liver CEC followed by infusion of AHSC constitutes an active involvement by tumor, soft tissue involvement by tumor, and well-tolerated regimen in women with high-risk non- and prior adjuvant chemotherapy. The present study metastatic or metastatic breast cancer; direct comparison appears to support the concept that the extent of pre-trans- against other commonly used regimens such as CCB may plant chemotherapy is associated with a less favorable be warranted. prognosis; as noted in Table 4, a strong univariate trend (P = 0.007) towards extended survival among patients exposed to fewer than six cycles of pre-transplant chemo- References therapy was observed. However, the possibility that more extensive prior treatment merely represents a marker for 1 Harris JR, Morrow M, Bonadonna G. Cancer of the breast. more aggressive disease remains to be ruled out. in: DeVita VT, Hellman S, Rosenberg SA (eds). Cancer: Prin- For the purposes of this study, patients with disease that ciples and Practice of Oncology, 4th edn. Lipincott: Philadel- phia, 1993, pp 1264–1332. had relapsed in the breast or chest wall were not considered 2 Aisner J, Cirrincione C, Perloff M et al. Combination chemo- to have stage IV disease unless distant metastatic sites were therapy for metastatic or recurrent carcinoma of the breast – present. Some authors have tabulated such patients as hav- a randomized phase III trial comparing CAF versus VATH ing stage IV disease. As such, the progression-free survival versus VATH alternating with CMFVP: Cancer and Leukemia rates calculated in our analysis may underestimate the effi- Group B Study 8281. J Clin Oncol 1995; 13: 1443–1452. cacy of CEC in comparison with some other series. 3 Calabresi F, Di Lauro L, Marolla P et al. Fluorouracil, doxoru- Factors independently predictive of the rate of neutrophil bicin, and cyclophosphamide versus fluorouracil, doxorubicin, engraftment in the present trial included the number of and cyclophosphamide plus lonidamine for the treatment of CFU-GM infused, the administration of post-transplant advanced breast cancer: a multicentric randomized clinical hematopoietic growth factors, the use of G-CSF or cyclo- study. Semin Oncol 1991; 18: 66–72. 4 Fumoleau P, Delgado FM, Delozier T et al. Phase II trial of phosphamide plus G-CSF (vs GM-CSF or no mobilization) weekly intravenous vinorelbine in first-line advanced breast for mobilization, and the stage of disease (Table 3). Inter- cancer chemotherapy. J Clin Oncol 1993; 11: 1245–1252. 43 estingly, our recently published multivariate analysis of 5 Gelmon KA, O’Reilly SE, Tolcher AW et al. Phase I/II trial of factors affecting rate of neutrophil engraftment following biweekly paclitaxel and cisplatin in the treatment of metastatic HDC in patients with a variety of malignancies revealed breast cancer. J Clin Oncol 1996; 14: 1185–1191. CEC for breast cancer TR Klumpp et al 280 6 Muss HB, Case LD, Richards F II et al. Interrupted versus of high-dose busulfan and CY followed by autologous BM continuous chemotherapy in patients with metastatic breast or peripheral blood stem cell transplantation in patients with cancer. New Engl J Med 1991; 325: 1342–1348. advanced chemosensitive breast cancer. Bone Marrow Trans- 7 Tannock IF, Boyd NF, DeBoer G et al. A randomized trial plant 1993; 11: 337–339. of two dose levels of cyclophosphamide, methotrexate, and 24 Kritz A, Crown JP, Motzer RJ et al. Beneficial impact of peri- fluorouracil chemotherapy for patients with metastatic breast pheral blood progenitor cells in patients with metastatic breast cancer. J Clin Oncol 1988; 6: 1377–1387. cancer treated with high-dose chemotherapy plus granulocyte– 8 Bonadonna G, Valagussa P, Moliterni A et al. Adjuvant cyclo- macrophage colony-stimulating factor. Cancer 1993; 71: phosphamide, methotrexate, and fluorouracil in node-positive 2515–2521. breast cancer. New Engl J Med 1995; 332: 901–906. 25 Patrone F, Ballestrero A, Ferrando F et al. Four-step high- 9 Wood WC, Budman DR, Korzun AH et al. Dose and dose dose sequential chemotherapy with double hematopoietic pro- intensity of adjuvant chemotherapy for stage II, node-positive genitor-cell rescue for metastatic breast cancer. J Clin Oncol breast carcinoma. New Engl J Med 1994; 330: 1253–1259. 1995; 13: 840–846. 10 Antman K, Gale RP. Advanced breast cancer: high-dose 26 Peters WP, Shpall EJ, Jones RB et al. High-dose combination chemotherapy and bone marrow autotransplants. Ann Intern alkylating agents with bone marrow support as initial treat- Med 1988; 108: 570–574. ment for metastatic breast cancer. J Clin Oncol 1988; 6: 11 Antman K, Ayash L, Elias A et al. A phase II study of high- 1368–1376. dose cyclophosphamide, thiotepa, and carboplatin with auto- 27 Peters WP, Ross M, Vredenburgh JJ et al. High-dose chemo- logous marrow support in women with measurable advanced therapy and autologous bone marrow support as consolidation breast cancer responding to standard-dose therapy. J Clin after standard-dose adjuvant therapy for high-risk primary Oncol 1992; 10: 102–110. breast cancer. J Clin Oncol 1993; 11: 1132–1143. 12 Ayash LJ, Wheeler C, Fairclough D et al. Prognostic factors 28 Shpall EJ, Jones RB, Bast RC et al. 4-Hydroperoxycyclophos- for prolonged progression-free survival with high-dose phamide purging of breast cancer from the mononuclear cell chemotherapy with autologous stem-cell support for advanced fraction of bone marrow in patients receiving high-dose breast cancer. J Clin Oncol 1995; 13: 2043–2049. chemotherapy and autologous marrow support: a phase I trial. 13 Bezwoda WR, Seymour L, Dansey RD. High-dose chemo- J Clin Oncol 1991; 9: 85–93. therapy with hematopoietic rescue as primary treatment for 29 Somlo G, Doroshow JH, Forman SJ et al. High-dose cisplatin, metastatic breast cancer: a randomized trial. J Clin Oncol etoposide, and cyclophosphamide with autologous stem cell 1995; 13: 2483–2489. reinfusion in patients with responsive metastatic or high-risk 14 Bowers C, Adkins D, Dunphy F et al. Dose escalation of primary breast cancer. Cancer 1994; 73: 125–134. mitoxantrone given with thiotepa and autologous bone marrow 30 Stadtmauer EA, Hodi FS, Sickles C et al. High-dose therapy transplantation for metastatic breast cancer. Bone Marrow and growth factor mobilized peripheral blood progenitor cell Transplant 1993; 12: 525–530. support for responding patients with metastatic breast cancer. 15 Broun ER, Sridhara R, Sledge GW et al. Tandem autotrans- Prog Clin Biol Res 1994; 389: 487–495. plantation for the treatment of metastatic breast cancer. J Clin 31 Spitzer TR, Cirenza E, McAfee S et al. Phase I–II trial of Oncol 1995; 13: 2050–2055. high-dose cyclophosphamide, carboplatin, and autologous 16 Dunphy FR, Spitzer G, Rossiter Fornoff JE et al. Factors pre- bone marrow or peripheral blood stem cell rescue. Bone Mar- dicting long-term survival for metastatic breast cancer patients row Transplant 1995; 15: 537–542. treated with high-dose chemotherapy and bone marrow sup- 32 Williams SF, Mick R, Desser R et al. High-dose consolidation port. Cancer 1994; 73: 2157–2167. 17 Eder JP, Antman K, Peters W et al. High-dose combination therapy with autologous stem cell rescue in stage IV breast alkylating agent chemotherapy with autologous bone marrow cancer. J Clin Oncol 1989; 7: 1824–1830. support for metastatic breast cancer. J Clin Oncol 1986; 4: 33 Vaughan WP, Reed EC, Edwards B et al. High-dose cyclo- 1592–1597. phosphamide, thiotepa and hydroxyurea with autologous hem- 18 Elias AD, Ayash L, Anderson KC et al. Mobilization of peri- atopoietic stem cell rescue: an effective consolidation chemo- pheral blood progenitor cells by chemotherapy and granulo- therapy regimen for early metastatic breast cancer. Bone cyte–macrophage colony-stimulating factor for hematologic Marrow Transplant 1994; 13: 619–624. support after high-dose intensification for breast cancer. Blood 34 Durand RF, Goldie JH. Interaction of etoposide and cisplatin 1992; 79: 3036–3044. in an in vitro tumor model. Cancer Treat Rep 1987; 71: 19 Fields KK, Perkins JP, Hiemenz JW et al. Intensive dose ifos- 673–679. famide, carboplatin, and etoposide followed by autologous 35 Bajorin DF, Sarosdy MF, Pfister DG et al. Randomized trial stem cell rescue: results of a phase I/II study in breast cancer of etoposide and cisplatin versus etoposide and carboplatin in patients. Surg Oncol 1993; 2: 87–95. patients with good-risk germ cell tumors: a multi-instutional 20 Ghalie R, Williams SF, Valentino LA et al. Tandem peripheral study. J Clin Oncol 1993; 11: 598–606. blood progenitor cell transplants as initial therapy for meta- 36 Loehrer PJ, Ansari R, Gonin R et al. Cisplatin plus etoposide static breast cancer. Biol Blood Marrow Transplant 1995; 1: with and without ifosfamide in extensive small-cell lung can- 40–46. cer: a Hoosier Oncology Group study. J Clin Oncol 1995; 13: 21 Holland HK, Dix SP, Geller RB et al. Minimal toxicity and 2594–2599. mortality in high-risk breast cancer patients receiving high- 37 Cocconi G, Tonato M, Di Costanzo F et al. Platinum and eto- dose cyclophosphamide, thiotepa, and carboplatin plus auto- poside in chemotherapy refractory metastatic breast cancer. A logous marrow/stem-cell transplantation and comprehensive phase II trial of the Italian Oncology Group for Clinical supportive care. J Clin Oncol 1996; 14: 1156–1164. Research. Eur J Cancer Clin Oncol 1986; 22: 761–764. 22 Kennedy MJ, Vogelsang GB, Beveridge RA et al. Phase I 38 Cocconi G, Lottici R, Bisagni G et al. Combination therapy trial of intravenous cyclosporine to induce graft-versus-host with platinum and etoposide of brain metastases from breast disease in women undergoing autologous bone marrow trans- carcinoma. Cancer Invest 1990; 8: 327–334. plantation for breast cancer. J Clin Oncol 1993; 11: 478–484. 39 Cocconi G, Bisagni G, Bacchi M et al. Cisplatin and etoposide 23 Klumpp TR, Mangan KF, Glenn LD et al. Phase II pilot study as first-line chemotherapy for metastatic breast carcinoma: a CEC for breast cancer TR Klumpp et al 281 prospective randomized trial of the Italian Oncology Group interferon alfa-2b for patients with lymphoma and Hodgkin’s for Clinical Research. J Clin Oncol 1991; 9: 664–669. disease. J Clin Oncol 1994; 12: 2423–2431. 40 Klumpp TR, Mangan KF, Glenn LD et al. Phase I/II study of 47 Tenny CM, Jacobs SA, Stoller RG et al. Ablative chemo- high-dose cyclophosphamide, etoposide, and cisplatin fol- therapy with cyclophosphamide, etoposide, cis-platinum or lowed by autologous bone marrow or peripheral blood stem carboplatinum and autologous bone marrow rescue in patients cell transplantation in patients with poor-prognosis Hodgkin’s with recurrent breast cancer. Proc Am Soc Clin Oncol 1990; disease or non-Hodgkin’s lymphoma. Bone Marrow Trans- 9: 46 (Abstr. 175). plant 1993; 12: 337–345. 48 Motzer RJ, Mazumdar M, Bosl GJ et al. High-dose carbopla- 41 Ravdin PM, Osborne CK. Breast cancer. In: Macdonald JS, tin, etoposide, and cyclophosphamide for patients with refrac- Haller DG, Mayer RJ (eds). Manual of Oncologic Thera- tory germ cell tumors: treatment results and prognostic factors peutics, 3rd edn. Lippincott: Philadelphia, 1995, pp 153–161. for survival and toxicity. J Clin Oncol 1996; 14: 1098–1105. 42 Mangan K, Mullaney M, Klumpp T et al. Mobilization of 49 Ascensao JL, Miller KB, Schenkein D et al. A new regimen peripheral blood stem cells by subcutaneous injections of yest- for relapsed lymphomas: high-dose carboplatin, VP-16, and derived granulocyte macrophage colony stimulating factor: a cyclophosphamide with hemopoietic rescue. Proc Am Soc phase I–II study. Stem Cells 1993; 11: 445–454. Clin Oncol 1991; 10: 284 (Abstr. 991). 43 Klumpp TR, Mangan KF, Goldberg SL et al. Granulocyte col- 50 Shea TC, Storniolo AM, Mason JR et al. A dose-escalation ony-stimulating factor accelerates neutrophil engraftment fol- study of carboplatin/cyclophosphamide/etoposide along with lowing peripheral-blood stem-cell transplantation: a prospec- autologous bone marrow or peripheral blood stem cell rescue. tive, randomized trial. J Clin Oncol 1995; 13: 1323–1327. Semin Oncol 1992; 19: 139–144. 44 The LIFETEST Procedure. In: SAS/STAT User’s Guide, Ver- 51 Saez RA, Slease RB, Strnad C et al. High-dose carboplatin, sion 6, 4th edn, vol. 2. SAS Institute, Inc: Cary, NC, 1990, etoposide and cyclophosphamide with autologous bone mar- pp 1027–1069. row transplantation for the treatment of advanced malig- 45 McDonald GB, Hinds MS, Fisher LD et al. Veno-occlusive nancies: a phase I study. Bone Marrow Transplant 1995; 16: disease of the liver and multiorgan failure after bone marrow 507–514. transplantation: a cohort study of 355 patients. Ann Intern Med 52 Ibrahim A, Zambon E, Bourhis JH et al. High-dose chemo- 1993; 118: 255–267. therapy with etoposide, cyclophosphamide and escalating dose 46 Schenkein DP, Dixon P, Desforges JF et al. Phase I/II study of of carboplatin followed by autologous bone marrow transplan- cyclophosphamide, carboplatin, and etoposide and autologous tation in cancer patients. A pilot study. Eur J Cancer 1993; hematopoietic stem cell transplantation with posttransplant 29A: 1398–1403.