Four-Cycle High-Dose Therapy with Hematopoietic Support for Metastatic Breast Cancer: No Improvement in Outcomes Compared with Single-Course High-Dose Therapy

Four-Cycle High-Dose Therapy With Hematopoietic Support for Metastatic Breast Cancer: No Improvement in Outcomes Compared With Single-Course High-Dose Therapy

Wendy W. Hu, Robert S. Negrin, Keith Stockerl-Goldstein, Laura J. Johnston, Judith A. Shizuru, Ruby M. Wong, Nelson J. Chao,* Gwynn D. Long,* Robert H. Feiner,† Karl G. Blume

Division of Bone Marrow Transplantation, Department of Medicine, Stanford University Medical Center, Stanford, C a l i f o rn i a .

A d d ress correspondence to Wendy W. Hu, MD, Division of Bone Marrow Transplantation, Stanford University Medical Center, 300 Pasteur Drive, H1353, Stanford, CA 94305; email: [email protected] . e d u

(Received July 6, 1999; accepted November 18, 1999)

ABSTRACT Multiple-cycle high-dose therapy with autologous hematopoietic progenitor cell (AHPC) support has been used to deliver dose-intensive therapy. We have used this approach as well as single-cycle high-dose therapy in trea t i n g patients with metastatic breast cancer. We present the outcomes of multiple-cycle high-dose therapies and compare them with those resulting from single-course high-dose therapies perf o rmed at a single institution. Fifty-five patients received 4 cycles of intensive chemotherapy with AHPC support. Three multicycle regimens were sequentially applied. Twenty patients were enrolled to receive 4 cycles of high-dose mitoxantrone, thiotepa, and cyclophosphamide. Nineteen subsequent patients received this regimen modified by the incorporation of paclitaxel. Sixteen patients received 2 cycles of high-dose melphalan, thiotepa, and paclitaxel and 2 cycles of mitoxantrone, thiotepa, and paclitaxel. The results of all 3 multiple-cycle therapies are compared with those of 55 contemporaneous patients with metastatic breast cancer who received a single course of high-dose cyclophosphamide and thiotepa or cyclophosphamide, cisplatin, and BCNU (carmustine) with hematopoietic cell rescue. Multiple-cycle therapy was associated with more infectious complications, increased transfusion req u i r ements, and increased hospital admissions. However, there were no significant diffe r ences in outcomes between the groups. For 55 patients who rec e i v e d multiple-cycle therapy, the actuarial 3-year overall survival rate was 36% (95% confidence interval [CI] 23%-49%); fr eedom from prog r ession and event-free survival were both 15% (CI 5%-25%). The median time to disease pro- gr ession and median survival were 1.0 and 1.6 years, res p e c t i v e l y . For the 55 patients who underwent a single course of high-dose therapy, the 3-year overall survival was also 36% (CI 18%-54%), whereas freedom from prog re s s i o n and event-free survival were both 19% (CI 7%-31%). The median time to progression and median survival were 0.8 and 2.2 years, res p e c t i v e l y . Within the constraints of this patient population, the outcomes of 4 cycles of high- dose therapy with AHPC support were not superior to those resulting from single courses of high-dose therapy in Multiple-Cycle Therapy for Metastatic Breast Cancer tation continues to rep r esent an incurable condition, prov i d - men in treating malignancies. Applying their methodology ing justification for deferring cytotoxic treatment until ame- to regimens used in the treatment of metastatic breast can- lioration of symptoms becomes necessary. cer, they observed statistically significant linear correlations To improve survival outcomes for such patients, high- of SDI to overall response, complete remission (CR), and dose chemotherapy with autologous hematopoietic pro- median survival in 19 randomized trials comparing more genitor cell (AHPC) support has been increasingly used to dose-intensive regimens with those administered at conven- tr eat patients with metastatic breast cancer [8], resulting in 3- tional doses. Their results support the concept of dose year prog re s s i o n - f r ee survival rates of 15% to 20% [9]. Sur- intensity in the treatment of metastatic breast cancer. I n vival outcomes for patients who are able to achieve complete addition, they observed that a moderate increase in SDI was remission at the time of high-dose therapy are re p o rt e d l y n e c e s s a ry to demonstrate only a slight improvement in b e t t e r, with 27% to 37% of women achieving 3-year pro- median survival (eg, 0.5 SDI units correlated to 2 months). gre s s i o n - f r ee survival [7]. Given the more favorable risk/ben- This observation suggests that substantial increases in dose efit ratio for these patients, transplantation protocols have intensity may be necessary to demonstrate statistically signifi- re q u i red patients to achieve minimal disease (complete cant improvements in survival. Since the drugs effective in remission or very good partial remission) before high-dose br east cancer treatment generally cause myelosuppression as a th e r a p y . This strategy usually entails early use of cytored u c - dose-limiting toxicity, the use of AHPCs to escalate doses of tive therapy when tumor volume is low and could account multiple-cycle combination therapy administered in rel a t i v e l y for the improved survivals seen with high-dose therapy and sh o r t intervals rep r esents an attractive approach for the trea t - autografting [10]. Whether high-dose therapy will ultimately ment of patients with metastatic breast cancer. p rolong survival after diagnosis of metastatic breast cancer Accordingly, we and others have implemented multiple- remains to be determined from prospective randomized tri- cycle regimens supported with AHPCs and have demon- als. Such trials must adequately stratify for patient risk fac- strated the feasibility of this approach in treating a variety of tors and provide sufficient duration of follow-up evaluation. malignancies [20-27]. Our initially reported 4-cycle regimen Most high-dose therapy protocols have used a single of mitoxantrone, thiotepa, and cyclophosphamide (MTC) course of high-dose therapy with autograft support when was subsequently modified twice to escalate drug delivery minimal disease is achieved after conventional cytored u c t i v e and to ameliorate toxicity [21]. We hereby present our sin- chemotherapy [7,9]. This approach provides peak doses of gle-institution results for all 3 multiple-cycle high-dose reg- d rugs for “consolidation” of the clinical response achieved imens (n = 55) and compare them in a nonrandomized fash- with conventional-dose treatment, but does not increase the ion with those of contemporaneous patients (n = 55) who numbers of patients who become eligible for high-dose received a single course of high-dose therapy for tre a t i n g therapy. metastatic breast cancer. Another strategy to improve survival outcomes involves augmentation of dose intensity with multiple cycles of high-dose combination chemotherapy. Pre c l i n i c a l MATERIALS AND METHODS data and tumor modeling have supported impro v e d Patient Selection responses with therapies involving rapid administration of One hundred ten female patients with re c u rrent or chemotherapy [11,12], and clinical studies have under- newly diagnosed metastatic breast cancer were enrolled in s c o red the importance of maximal conventional dosing and consecutive protocols using high-dose therapy with AHPC scheduling in the achievement of event-free and overall support between November 1991 and November 1997. All s u rvivals [13,14]. It is conceivable that both dose escalation patients had been presented at a weekly new patient confer- and efficient dose delivery could improve outcomes in the ence, at which time clinical data were reviewed and protocol setting of re c u rrent disease. Furt h e rm o re, an approach that assignment was established. For all protocols, patient eligi- combines the two would not necessarily exclude patients bility criteria included age ≤65 years, Karnofsky perf o r- who could not achieve minimal disease with conventional mance status ≥80%, adequate organ function assessed by c y t o reduction or those with moderately bulky disease at pulmonary diffusion capacity of carbon monoxide at ≥60% d i a g n o s i s . predicted, exercise left ventricular ejection fraction of ≥0.50, Cloned growth factors, AHPC support, and impro v e- HIV seronegativity, and no central nervous system metasta- ments in supportive care have allowed for chemotherapy sis. The clinical protocols were approved and re v i e w e d W. W. Hu et al.

Table 1. Pretreatment Patient Characteristics*

Multiple Cycle (n = 55) Single-Cycle (n = 55) Median Age (Range) 44 (29-58) n (%) 46 (26-62) n (%) P value

Initial Diagnosis Estrogen receptor (ER)+/reported ER status 34/51 (62) 29/46 (53) Progesterone receptor (PR)+/reported PR status 32/49 (58) 24/41 (44) Pre-/postmenopausal/unknown 51/2/2 46/5/4 Tumor size T1 16 (29) 7 (13) T2 22 (40) 20 (36) T3 13 (24) 15 (27) T4 1 (2) 4 (7) Unknown 3 (5) 9 (16) No. lymph nodes 0 13 (24) 15 (27) 1-3 12 (22) 14 (25) 4-9 11 (20) 10 (18) 10-19 10 (18) 6 (11) ≥20 5 (9) 6 (11) Unknown 4 (7) 4 (7) Prior adjuvant therapy Radiation therapy 26 (48) 21 (38) Hormonal therapy 22 (40) 14 (25) Chemotherapy † Median 1 (0-2) Median 1 (0-2) .01 † 0 regimens 10 (18) 24 (44) 1 regimen 37 (67) 23 (42) 2 regimens 8 (15) 8 (15) Metastatic disease New diagnosis Stage IV disease 3 (5) 14 (25) Disease-free interval (range) Median 27 mo (1-96) Median 24 mo (1-120) .78 Disease sites Visceral 16 (29) 20 (36) Bone 34 (62) 32 (58) Soft tissue 36 (65) 38 (69) Minimal disease 10 (18) 10 (18) No. sites of disease Median 2 (1-5) Median 2 (1-7) .31 1 15 (27) 21 (38) 2 18 (33) 14 (28) 3 17 (31) 18 (33) ≥4 5 (9) 2 (4) Radiation therapy for Met Ds 17 (31) 14 (25) Hormonal therapy for Met Ds 16 (29) 18 (33) Cytoreductive regimens † Median 1 (0-3) Median 1 (1-4) .0004 † 0 15 (27) 0 (0) 1 32 (58) 40 (73) 2 7 (13) 11 (20) 3 1 (2) 3 (5) 4 0 (0) 1 (2) Total no. of prior regimens Median 2 (0-4) Median 2 (1-6) .27 0 1 (2) 1 17 (31) 15 (27) 2 27 (49) 25 (45) 3 9 (16) 13 (24) Multiple-Cycle Therapy for Metastatic Breast Cancer protocols did not receive cytoreductive therapy before treat- tient setting. Nineteen subsequent patients were enrolled to ment. All 55 patients treated with single-cycle high-dose a modification of the protocol using the same drugs but with therapy received 1 to 4 (median of 1) regimens of reinduc- dose intensification of thiotepa and the incorporation of tion (cytoreductive) therapy at conventional dose. The paclitaxel into the treatment schema (MTTC × 4). This pro- choice of therapy was based on prior treatment, and treat- tocol was amended to deliver 4 cycles of therapy every 21 ment was administered by the primary referring oncologist. days. Patients received MTTC therapy and AHPCs in the outpatient setting. The last group of 16 patients enrolled to Assessment of Patient Characteristics multiple-cycle therapy received 2 cycles of high-dose mel- Characteristics evaluated in the initial presentation of phalan, thiotepa, and paclitaxel, and 2 cycles of mitox- b reast cancer used standard TNM staging criteria [28]. an t r one, thiotepa, and paclitaxel (ATT × 2/MTT × 2) deliv- Patients who received surgical ovarian ablation were er ed in 21-day intervals and supported with AHPCs [27]. assessed as having received hormonal therapy. In the evalua- Fifty-five patients received a single course of high-dose tion of metastatic disease, disease-free interval was defin e d therapy with G-CSF–mobilized AHPC support. Two estab- as the length of time between completion of primary ther- lished regimens were sequentially used. Eleven patients apy and presentation of re c u rrent disease. The number of received high-dose cyclophosphamide and thiotepa (CT) sites involved by disease was evaluated as follows: disease to [29] and 44 subsequent patients received high-dose cyclophos- any 1 organ, such as lungs, liver, or bone was scored as 1 site phamide, cisplatin, and BCNU (CPB or STAMP I) [30]. each; disease to lymph node groups was scored by anatomic The patients who received CT and the initial 4 pa t i e n t s site. For example, involvement of both cervical and medi- who received CPB received high-dose therapy with AHPC astinal lymph node groups was scored as 2 sites. Pre t re a t- infusion and remained hospitalized until myeloid graft rec o v - ment patient characteristics are detailed in Table 1. er y. Subsequent patients who received CPB were discharge d on post-transplantation day +1 if symptoms could be con- Study Design/Treatment Plan t rolled with outpatient treatment and if clinically stable. Patients first underwent placement of a double-lumen These regimens are also listed in Table 2. tunneled catheter. Fifty-five patients were enrolled to receive 4-cycle high-dose therapy with AHPC support, as S u p p o rt i v e Care and Response Assessment represented schematically in Figure 1. This approach used a P rophylactic antibiotics, transfusion support, G-CSF single-agent “pulse” for mobilization of AHPCs followed by administration, and continuous bladder irrigation were 4 cycles of combination therapy. During this study period, instituted according to the current institutional standard at 3 multicycle regimens were sequentially developed. The the time of protocol enrollment [21,27,29-32]. treatment regimens have been previously described in detail Toxicity monitoring was graded using the Southwest [21,27]; their dosing and schedules are listed, with active Oncology Group common toxicity criteria. Patients were p rotocol dates, in Table 2. For mobilization of peripheral removed from the trial for any grade III-IV nonhematologic blood progenitor cells, 38 patients received etoposide (VP-16) toxicity or for refractory or progressive disease. 2 g/m2 over 4 hours, and 15 patients received cyclophos- Breast cancer lesions were assessed by physical examina- phamide 4 g/m2 over 2 hours. Granulocyte-colony stimulat- tion and imaging studies at the start of high-dose therapy. ing factor (G-CSF) at approximately 10 µg/kg per day was CR, partial response (PR), and very good partial re s p o n s e a d m i n i s t e red 24 hours after chemotherapy and continued (VGPR) were defined by standard criteria: disappearance of daily until the target number of mononuclear cells or all measurable tumor, >90% reduction in measurable CD34+ cells was collected by apheresis. Two patients did not lesions, and >50% reduction in the product of the bidimen- receive VP-16 or cyclophosphamide because AHPCs were sional measurements, re s p e c t i v e l y, for 4 weeks duration. p reviously collected by G-CSF alone. The AHPCs were Definitions for minimal response, stable disease, and pro- p rocessed and cry o p re s e rved as described previously [21]. g ressive disease included <50% reduction in disease, no Purging techniques were not used. change in tumor measurements, and >25% increase in Twenty patients were enrolled in the first multicycle pro- tumor size or the appearance of new lesions, respectively. tocol using 4 cycles of high-dose MTC delivered every 28 days. MTC therapy with AHPC infusion was administered Treatment After Recov e r y From High-Dose Therapy in hospital for the first 5 patients. Subsequent patients were After re c o v e ry from autografting, patients with tumors W. W. Hu et al.

Table 2. High-Dose Chemotherapy Regimens*

Day No. of Patients Total Regimen Dose

Four-Cycle HDC Regimens VP-16 Mobilization 2 g/m 2 20 1. MTC 3 4 (February 1992-May 1993) Mitoxantrone 6 mg/m 2 CI ϫ 3 d Ϫ5,Ϫ4,Ϫ3 72 mg/m 2 Thiotepa 50 mg/m 2 CI ϫ 2-3 d Ϫ5,Ϫ4, Ϫ3 400-600 mg/m 2 Cyclophosphamide 1.5-1.65 g/m 2 ϫ 3 d Ϫ5,Ϫ4,Ϫ3 18-19.8 g/m 2 2. MTTC ϫ 4 (June 1993-December 1994) 19 Mitoxantrone 9 mg/m 2 CI ϫ 2 d Ϫ4,Ϫ3 72 mg/m 2 Thiotepa 100 mg/m 2 CI ϫ 2 d Ϫ4, Ϫ3 800 mg/m 2 Cyclophosphamide 2.25 g/m 2 ϫ 2 d Ϫ4, Ϫ3 18 g/m 2 Paclitaxel 100-150 mg/m 2 ϫ 1 Ϫ2 400-600 mg/m 2 3. Cyclophosphamide Mobilization 4 g/m 2 ATT ϫ 2/MTT ϫ 2 (December 1994-June 1996) Melphalan 80 mg/m 2 ϫ 1 Ϫ4 16 160 mg/m 2 Thiotepa 150 mg/m 2 ϫ 2 d CI Ϫ4, Ϫ3 1200 mg/m 2 Paclitaxel 200 mg/m 2 ϫ 1 Ϫ2 800 mg/m 2 Mitoxantrone 15 mg/m 2 ϫ 2 d CI Ϫ4, Ϫ3 60 mg/m 2 Thiotepa 150 mg/m 2 ϫ 2 d CI Ϫ4,Ϫ3 Paclitaxel 200 mg/m 2 ϫ 1 Ϫ2 Single-Course Regimens 1. CT (November 1991-November 1994) Cyclophosphamide 1.5 g/m 2 ϫ 4 d CI Ϫ6 to Ϫ3 11 6 g/m 2† Thiotepa 200 mg/m 2 ϫ 4 d CI Ϫ6 to Ϫ3 800 mg/m 2 2. CPB (February 1995-November 1997) Cyclophosphamide 1.875 g/m 2 ϫ 3 d Ϫ5,Ϫ4, Ϫ3 44 5625 mg/m 2† Cisplatin 55 mg/m 2 ϫ 3 d CI Ϫ5,Ϫ4,Ϫ3 165 mg/m 2 BCNU 600 mg/m 2 ϫ 1 Ϫ2 600 mg/m 2

*ATT/MTT indicates melphalan, thiotepa, and paclitaxel/mitoxantrone, thiotepa, and paclitaxel; BCNU, carmustine; CI, continuous infusion; HDC, high-dose chemotherapy; CPB, cyclophosphamide, cisplatin, and BCNU; CT, cyclophosphamide and thiotepa; MTC, mitoxantrone, thiotepa, and cyclophosphamide; MTTC, MTC with dose intensification of thiotepa and addition of paclitaxel. †As previously reported in references 29 and 30. sites of bony or soft tissue disease was initiated 6 to 8 weeks of high-dose therapy. The group of patients who re c e i v e d after recovery from high-dose therapy. multiple-cycle and single-course high-dose therapy had similar characteristics, including disease-free intervals, median Statistical Anal y s i s number of sites of disease, numbers of patients who had P ro g re s s i o n - f ree intervals and survivals were measure d bone and soft tissue disease, and median number of total fr om the day of first autograft infusion until disease prog re s - prior regimens received by the time of high-dose therapy. sion or death. Descriptive statistics are rep o r ted as freq u e n - Equal numbers of patients had minimal disease at the time cies and medians. The method of Mann-Whitney rank sum of metastatic presentation. Prior anthracycline therapy was testing was used to determine diffe r ences in median values. a d m i n i s t e red to 47 patients who received multicycle re g i- Pe a r s o n ’ s chi-square method was used to test diffe r ences in mens and to 44 patients who received single-cycle therapy. distribution of disease status at time of high-dose therapy. Characteristics resulting in significant statistical diff e r- Ev e n t - f r ee, prog re s s i o n - f r ee, and overall survival curves were ences in distribution between the 2 groups include number calculated by means of the Kaplan-Meier pro d u c t - l i m i t of initial adjuvant regimens and number of cytore d u c t i v e Multiple-Cycle Therapy for Metastatic Breast Cancer

Table 3. Delivery of Multiple-Cycle Therapy*

No. Cycles Intervals Between Cycles (Median No. of Days) Delivered/No. Mobilization to ⌬t (Cycle 1 to 2) ⌬t (Cycle 2 to 3) ⌬t (Cycle 3 to 4) Regimens Possible Cycles (%) Cycle 1 (Range) (Range) (Range) (Range)

MTC 72/80 90 25 (18-49) 28 (23-58) 29 (25-56) 28 (21-47) MTTC 69/76 91 17 (13-47) 21 (20-35) 24 (21-46) 28 (21-39) ATT/MTT 52/64 81 23 (20-52) 23 (21-42) 25 (21-45) 25 (21-45)

*ATT/MTT indicates melphalan, thiotepa, and paclitaxel/mitoxantrone, thiotepa, and paclitaxel; MTC, mitoxantrone, thiotepa, and cyclophosphamide; MTTC, MTC with dose intensiﬁcation of thiotepa and addition of paclitaxel; ⌬t (cycle 1 to 2), time interval from 1 cycle to another of multicycle therapy.

mononuclear cells per kilogram per cycle. Two initial MTTC regimen, whereas the toxicities associated with patients enrolled in multicycle therapy received MTC with ATT/MTT are notable for infectious complications and a G-CSF alone to de t e r mine whether bone marrow rec o v e r y lack of significant effect on organ function. The single- could be achieved without AHPC support. Both patients course regimen toxicities with CPB are predominantly due required AHPC infusion 15 to 16 days after completion of to the well-known effect of interstitial pneumonitis associ- chemotherapy because of persistent profound neutro p e n i a ated with BCNU [34,35]. (absolute neutrophil count [ANC] <1 0 0 /µL), after which ANC re c o v e ry to ≥5 0 0 /µL re q u i red another 8 days. All S u r vival Data subsequent patients received AHPC support. The dose of The median follow-up is 2 years in the single-cycle AHPCs administered at each cycle of therapy and the g roup compared with 1.8 years in the multicycle gro u p . median re c o v e ry of ANC ≥5 0 0 /µL and platelet counts There was no signiﬁcant difference in overall survival, free- ≥ 2 0 , 0 0 0 /µL have been described in detail for MTC and dom from pro g ression, or event-free survival between the ATT/MTT [21]. For patients who received MTTC, the g roup of patients who received multiple-cycle high-dose median times to ANC and platelet re c o v e ry were 9 therapy and those who received single-course high-dose (range 8-14) and 10 (range 7-48) days, respectively. t h e r a p y. Figure 2 illustrates the Kaplan-Meier surv i v a l Patients treated with single-cycle therapy were re c o n s t i- curves comparing the outcomes of multicycle treatment and tuted with AHPCs at a dose range of 0.95 to 21.2 ϫ single-cycle treatment. For all 55 patients enrolled in multi- 1 08 mononuclear cells/kg. The median times to ANC and ple-cycle therapy, the actuarial 3-year overall survival was platelet recovery for patients who received CT were 10 and 36% (95% confidence interval [CI] 23%-49%), and fre e- 13 days, re s p e c t i v e l y. For patients who received CPB, the dom from pro g ression and event-free survival were both median CD34+ cell dose was 5.1 ϫ 106/kg (range 1.5-9.1 ϫ 15% (CI 5%-25%). The median time to disease progression 106/kg), which resulted in median days to ANC and platelet and median s u rvival were 1.0 and 1.6 years, re s p e c- recovery of 9 and 10.5, respectively. t i v e l y. For the 55 patients who received single-course high-dose therapy, the 3-year overall survival was also 36% Tr a n s f u s i o n s The total numbers of transfusion products required per patient enrolled to each protocol are delineated in Table 4. Table 4. Number of Tranfusion Products per Patient* The number of transfusions for patients who received multiple-cycle therapy include blood products re q u i red during apheresis after mobilization chemotherapy as well as during No. PRBC Transfusions No. Plt Transfusions recovery from each cycle of combination high-dose therapy. Median (Range) Median (Range) The fewer numbers of red cell and platelet pro d u c t s Multicycle Regimen required for patients who received CPB single-course ther- MTC 17.5 (5-38) 6 (2-29) + apy reflect the designation of CD34 cells rather than MTTC 21 (4-35) 11 (3-40) W. W. Hu et al.

Table 5. Toxicities of High-Dose Therapy*

Multiple-Cycle Regimens Single-Course Regimens MTC (No. Cycles) MTTC ATT/MTT Total CT CPB Total Toxicities (No. of Episodes) VP-16 (19) (72) VP-16 (19) (69) CY (15) (52) Episodes (11) (44) Episodes

Fever (no source) 10 20 1 15 2 7 55 2 9 11 GI toxicity 11 1 12 3 4 7 TPN 2 1 1 4 1 2 3 Pancreatitis 1 1 Hemorrhagic cystitis 2 2 5 9 2 2 Hospital admissions 2 37 2 23 3 18 85 11 † 5 16 Overnight observations 7 7 Congestive heart failure 1 4 5 2 2 Veno-occlusive disease of liver 1 2 3 1 1 Diffuse alveolar hemorrhage 2 2 Interstitial pneumonitis 1 5 2 6 1 24 25 Idiopathic thrombocytopenic purpura 1 1 1 1 Hemolytic uremic syndrome 1 1 Mortality 0 1 1 2 2 2 Bacteremia Gram-pos organisms 2 5 7 1 1 Gram-neg organisms 4 2 4 4 14 Fungemia 3 1 4 2 2 Viral reactivation (HSV, VZV) 1 9 6 5 21 5 5 Pneumonia Bacterial 2 2 1 1 2 Fungal 2 2 Viral (RSV, CMV) 1 1 2 Pneumocystis carinii 1 1 Miscellaneous Cellulitis: Xanthomonas 1 1 Catheter-related 1 1 2 Clostridium difﬁcile enterocolitis 1 3 4 Exudative pharyngitis 2 2 Sinusitis 1 1 Pilonidal cyst 1 1 Perirectal abscess 1 1 Urinary candidiasis 2 2

*ATT/MTT indicates melphalan, thiotepa, and paclitaxel/mitoxantrone, thiotepa, and paclitaxel; CPB, cyclophosphamide, cisplatin, and BCNU; CMV, cytomegalovirus; CT, cyclophosphamide and thiotepa; Cy, cyclophosphamide; GI, gastrointestinal; HSV, herpes simplex virus; MTC, mitoxantrone, thiotepa, and cyclophosphamide; MTTC, MTC with dose intensiﬁcation of thiotepa and addition of paclitaxel; RSV, respiratory syncytial virus; TPN, total parenteral nutrition; VZV, varicella zoster virus. †All patients hospitalized until autograft recovery.

(CI 18%-54%) while freedom from progression and event- t h e r a p y. The median follow-up is 3.1 years for multicycle f ree survival were both 19% (CI 7%-31%). The median patients and 2 years for single-cycle patients. Outcome time to disease progression and median survival was 0.8 and analysis of this subset for 3-year overall survival was 54% Multiple-Cycle Therapy for Metastatic Breast Cancer W. W. Hu et al.

Table 6. Regimen Dose Intensity and Calculation of Summation Dose Intensity*

Conventional Protocol Total Protocol Protocol Unit Dose Fractional Chemotherapy Dose Dose Dose Dose Intensity (DI) Intensity (UDI) UDI Drug (mg/m 2 per cycle) (mg/m 2 per cycle) (mg/m 2) (mg/m 2 per wk) (mg/m 2 per wk) (DI/UDI)

MTC Protocol (q 28 d over 16 wk) VP-16 2000 667 340 1.96 Mitoxantrone 12-14 18 72 4.5 5 0.9 Thiotepa 20-50 150-200 600-800 (37.5 to) 50 58 † 0.86 Cyclophosphamide 1400 4500-5000 18,000-20,000 (1125 to) 1250 650 1.92 Weighted SDI ‡ = 3.41 MTTC Protocol (q 21 d over 12 wk) VP-16 2000 667 340 1.96 Mitoxantrone 12-14 18 72 6 5 1.2 Thiotepa 20-50 200 800 66.7 58 1.15 Paclitaxel 135-175 100-150 400-600 (33 to) 50 50 1.0 Cyclophosphamide 1400 4500 18,000 1500 650 2.3 Weighted SDI = 4.91 ATT/MTT Protocol (q 21 d over 12 wk) Cyclophosphamide 4000 1333 650 2.05 Melphalan 25-40 80 160 13.3 9 1.47 Mitoxantrone 12-14 30 60 10 5 2.0 Thiotepa 20-50 300 1200 100 58 1.72 Paclitaxel 135-175 200 800 66.7 50 1.33 Weighted SDI = 4.24

*ATT/MTT indicates melphaln, thiotepa, and paclitaxel/mitoxantrone, thiotepa, and paclitaxel; MTC, mitoxantrone, thiotepa, and cyclophosphamide; MTTC, MTC with dose intensiﬁcation of thiotepa and addition of paclitaxel; SDI, summation dose intensity; VP-16, etoposide. †Thiotepa UDI (Hryniuk, personal communication). ‡Weighted average SDI calculated by contribution of initial single-agent mobilization followed 3 weeks later by combination chemotherapy. Sample calculation of weighted SDI: (1.96)(3 weeks) + (0.9 + 0.86 + 1.92)(16 weeks)/19 weeks = 3.41.

statistically inferior to the 13 patients who had achieved CR delayed in subsequent cycles because of the development of or VGPR (P = .001, P = .002, P = .039, respectively). o rgan toxicities (hemorrhagic cystitis, pneumonitis, card i a c For patients who did not achieve CR or VGPR, there dysfunction) or infections (Table 5). Nevertheless, 91% of w e re no signficant diff e rences in survival between the intended cycles were delivered with multiple delays. The 2 treatment groups. ATT/MTT regimen was slightly less dose-intensive com- p a red with MTTC and appeared to be more feasible, although only 81% of intended cycles were delivered. Fewer DISCUSSION re g i m e n - related toxicities occurred; however, infectious The feasibility of delivering 4 cycles of high-dose therapy complications continued to be prevalent [27]. with AHPC support has been previously rep o r ted for MTC As reflected by the multiple exposures to high-dose [21] and ATT/MTT [27] as well as for other re g i m e n s therapy and resultant episodes of pancytopenia, more [25,26,36,37]. As with conventional dose combinations, com- episodes of life-threatening organ toxicities and a range of parison of a reg i m e n ’ s dose intensity with that of another has infectious complications occurred in the group of patients been problematic. However, a single score for a particular reg - who received multiple-cycle therapy. These patients also imen can be calculated by applying the SDI method rep o rt e d re q u i red readmissions more frequently and received more Multiple-Cycle Therapy for Metastatic Breast Cancer tus at time of high-dose therapy. Because response to tamoxifen, and the surprisingly poor survivals re s u l t i n g c y t o reductive therapy and outcomes for patients with f rom “conventional” tre a t m e n t . metastatic breast cancer are significantly influenced by prior In reviewing the relative dose intensity of conventional t h e r a p y, meaningful interpretation of survival data may be chemotherapy regimens, Hryniuk et al. [19] found that the constrained owing to these diff e rences in distribution highest SD I originated from a study by Hortobagyi et al. [43], between the patient groups and the relatively limited num- whose high-dose FAC regimen yielded an SDI value of 3.24, ber of patients evaluable in subset analyses. slightly lower than our MT C regimen SDI of 3.41, suggest- For all patients, there are no significant diff e rences in ing that chemotherapy dose intensity for certain drugs may e v e n t - f ree survival, freedom from pro g ression, and overall be escalated sufficiently without the need for AHPC support. s u rvival between the 2 treatment approaches. Six patients Un f o rt u n a t e l y , the SDI model presumes equal efficacy from enrolled in multicycle treatment received 2 or fewer cycles later and earlier cycles of therapy and cannot address the effe c t s of high-dose therapy. The survival outcomes of the remain- of peak doses of high-dose chemotherapy administered as con- ing 49 patients who received 3 or 4 cycles of therapy are solidation after achievement of CR or VG P R as perfo rm e d similar to those of patients who received single-cycle ther- with single autografting. Our data also indicate that multiple- apy. Therefore, the majority of patients enrolled to multicy- cycle high-dose regimens do not appear to benefit patients cle therapy received adequate numbers of high-dose cycles who cannot achieve CR or VG P R with conventional therapy. to allow for evaluation of outcomes. A single course of h i g h - d o s e therapy can provide peak The 13 patients who achieved a CR or VGPR before doses of drug rather than dose-intensive drug delivery. multicycle high-dose therapy may have had a trend toward P re s u m a b l y, decreasing the number of cycles of high-dose superior freedom from pro g ression (P = .13). A review of therapy may allow for increases in peak drug dosing. Nev- the pre t reatment characteristics of these patients suggests e rtheless, it is unclear whether double [22,42,44-49] or that this group had features predicting more favorable out- triple [ 2 0 , 2 4 , 5 0 , 5 1 ] cycles of h i g h - d o s e therapy will re s u l t comes, including metastatic disease at 1 or 2 sites, and 1 to 2 in improved outcomes for patients with re c u rre nt disease, prior chemotherapy regimens before high-dose therapy. although the efficacy of such therapies would need to be The median freedom from disease pro g ression for this substantially superior to single autografts to justify the g roup was 2.3 years, compared with 0.9 years for the 31 potential pro b l e m s associated with multiple-cycle t h e r a p y. patients who received single autografts and demonstrated “High-dose sequential” [52-54] and “dose dense” [55] m o re heterogenous risk features. Nevertheless, the over- chemotherapy approaches using single agents at high dose all survival between the 2 groups was similar, with only a appear to yield promising results in the high-risk disease 2-month improvement in the multicycle gro u p . setting, but may be disappointing for re c u rrent disease due It is conceivable that if the 2 treatment arms were bal- to greater tumor heterogeneity and drug resistance [56,57]. anced with regard to CR/VGPR status, there may have been High-dose therapy with AHPC support offers signifi- a statistical difference in FFP and EFS. Because both groups cant tumor cytoreductive potential and may benefit patients were relatively balanced in the number of total prior thera- who achieve minimal disease before autografting. Neverthe- pies received, it is unclear whether the differences in num- less, emphasis should be placed on the viewpoint of trans- bers of adjuvant therapies or cytoreductive therapies could plantation technologies as part of multimodality tre a t m e n t have a significant impact on outcomes. The similar overall of malignant disease, rather than as an end therapy in itself, s u rvival between the 2 groups could also be accounted for particularly in the setting of recurrent disease. by the possible use of salvage therapies, including hormonal In summary, 4-cycle high-dose combination therapy and immune-based therapies in patients who suff e re d s u p p o rted with AHPCs is feasible, but is associated with a relapse after single-cycle high-dose therapy which may not p rolonged period of treatment, many episodes of pancy- have been feasible for patients with more toxicity who suf- topenia requiring more transfusion support, and a range of fered relapse after multicycle therapy. re g i m e n - related toxicities and infectious complications. The results of other multiple-cycle high-dose re g i- Within the constraints of this patient population, the out- mens in the treatment of patients with metastatic bre a s t comes of such an approach are no diff e rent from those cancer have been previously reviewed [38,39]. Diff e re n t resulting from a single course of high-dose therapy in the sequences of single-agent and combination regimens and t reatment of metastatic breast cancer. Other strategies are v a rying numbers of cycles of therapy preclude adequate clearly needed to treat this challenging disease. W. W. Hu et al.

5. Cody HS III. Sentinel lymph node mapping in breast cancer. 24. Rodenhuis S, Westermann A, Holtkamp MJ, et al. Feasibility of Oncology. 1999;13:25. multiple courses of high-dose cyclophosphamide, thiotepa, and carbo- 6. Hayes DF, Henderson IC, Shapiro CL. Treatment of metastatic platin for breast cancer or germ cell cancer. J Clin Oncol. 1996;14:1473. breast cancer: present and future prospects. Semin Oncol. 1995;22:5. 25. Shapiro CL, Ayash L, Webb IJ, et al. Repetitive cycles of 7. Peters WP, Baynes RD. Autologous hematopoietic cell transplan- cyclophosphamide, thiotepa, and carboplatin intensification with tation for breast cancer. In: Thomas ED, Blume KG, Forman SJ, eds. peripheral-blood progenitor cells and ﬁlgrastim in advanced breast can- Hematopoietic Cell Transplantation. Malden, MA: Blackwell Science; cer patients. J Clin Oncol. 1997;15:674. 1999:1029. 26. Schilder RJ, Shea TC. Multiple cycles of high-dose chemotherapy 8. Chlebowski RT, Lillington LM. A decade of breast cancer clinical for ovarian cancer. Semin Oncol. 1998;25:349. investigation: results as reported in the Program/Proceedings of the 27. Hu WW, Long GD, Stockerl-Goldstein KE, et al. A feasibility American Society of Clinical Oncology. J Clin Oncol. 1994;12:1789. study of multiple cycle high dose therapy with melphalan, thiotepa, and 9. Antman KH, Rowlings PA, Vaughan WP, et al. High-dose paclitaxel followed by mitoxantrone, thiotepa, and paclitaxel with autol- chemotherapy with autologous hematopoietic stem-cell support for ogous hematopoietic cell support for metastatic breast cancer. Clin Can - breast cancer in North America. J Clin Oncol. 1997;15:1870. cer Res. 1999;5:3411. 10. Rahman ZU, Frye DK, Buzdar AU, et al. Impact of selection 28. American Joint Committee on Cancer. In: Beahrs OH, Henson process on response rate and long-term survival of potential high-dose DE, Hutter DVP, Myers M, eds. Manual for Staging of Cancer. 3rd ed. chemotherapy candidates treated with standard-dose doxorubicin-con- Philadelphia: JB Lippincott; 1988. taining chemotherapy in patients with metastatic breast cancer. J Clin 29. Kennedy MJ, Beveridge RA, Rowley SD, Gordon GB, Pearlman Oncol. 1997;15:3171. ES, Macdonald JS. High-dose chemotherapy with reinfusion of purged 11. Coldman AJ, Goldie JH. Impact of dose-intense chemotherapy autologous bone marrow following dose-intense induction as initial on the development of permanent drug resistance. Semin Oncol. therapy for metastatic breast cancer. J Natl Cancer Inst. 1991;83:920. 1987;14:29. 30. Peters WP, Ross M, Vredenburgh JJ, et al. High-dose chemo- 12. Skipper H. Dose intensity versus total dose of chemotherapy: an therapy and autologous bone marrow support as consolidation after experimental basis. In: DeVita VJ, Hellman S, Rosenberg SA, eds. standard-dose adjuvant therapy for high-risk primary breast cancer. J Important Advances in Oncology. Philadelphia: JB Lippincott; 1993:43. Clin Oncol. 1993;11:1132. 13. Wood WC, Budman DR, Korzun AH, et al. Dose and dose inten- 31. Klumpp TR, Mangan KF, Goldberg SL, Pearlman ES, Macdonald sity of adjuvant chemotherapy for stage II, node-positive breast carci- J S. Granulocyte colony-stimulating factor accelerates neutrophil noma. N Engl J Med. 1994;330:1253. engraftment following peripheral-blood stem-cell transplantation: a 14. Budman DR, Berry DA, Cirrincione CT, et al. Dose and dose prospective, randomized trial. J Clin Oncol. 1995;13:1323. intensity as determinants of outcome in the adjuvant treatment of 32. Peters WP. High-dose chemotherapy with autologous bone mar- breast cancer. The Cancer and Leukemia Group B. J Natl Cancer Inst. row transplantation for the treatment of breast cancer: yes. I m p o r t a n t 1998;90:1205. Adv Oncol. 1995;215. 15. Hryniuk WM. The importance of dose intensity in the outcome of 33. Kaplan EL, Meier P. Nonparametric estimation from incomplete chemotherapy. In: DeVita VT, Hellman S, Rosenberg SA, eds. Impor - observations. J Am Stat Assoc. 1958;53:457. tant Advances in Oncology. Philadelphia: JB Lippincott; 1988:121. 34. Chap L, Shpiner R, Levine M, Norton L, Lill M, Glaspy J. Pul- 16. Gisselbrecht C. Chemotherapy dose intensity facilitated by use of monary toxicity of high-dose chemotherapy for breast cancer: a non- lenograstim: implications for quality of life and survival. Eur J Cancer. invasive approach to diagnosis and treatment. Bone Marrow Transplant. 1994;30A(suppl 3):S30. 1997;20:1063. 17. Lamar RE, Greco FA, Johnson DH, Murphy PB, Hainsworth JD. 35 . Wilczynski SW, Erasmus JJ, Petros WP, Vredenburgh JJ, Folz RJ. High-dose, brief duration, multiagent chemotherapy for metastatic Delayed pulmonary toxicity syndrome following high-dose chemotherapy breast cancer. Cancer. 1994;73:1842-1848. and bone marrow transplantation for breast cancer. Am J Respir Crit Care 18. Honkoop AH, Hoekman K, Wagstaff J, et al. D o s e - i n t e n s i v e Med. 19 9 8 ; 1 5 7 : 5 6 5 . chemotherapy with doxorubicin, cyclophosphamide and GM-CSF fails 36. Shea T, Graham M, Bernard S, et al. A clinical and pharmacoki- to improve survival of metastatic breast cancer patients. Ann Oncol. netic study of high-dose carboplatin, paclitaxel, granulocyte colony- 1996;7:35. stimulating factor, and peripheral blood stem cells in patients with 19. Hryniuk W, Frei R III, Wright FA. A single scale for comparing unresectable or metastatic cancer. Semin Oncol. 1995;22:80. dose-intensity of all chemotherapy regimens in breast cancer: summa- 37. Ford C, Spitzer G, Reilly W, Adkins D. A phase II study of repeti- tion dose-intensity. J Clin Oncol. 1998;16:3137. tive cycles of dose-intense carboplatin plus paclitaxel chemotherapy and Multiple-Cycle Therapy for Metastatic Breast Cancer

high-dose melphalan followed by cyclophosphamide, thiotepa, and car- support in patients with high-risk breast cancer. Semin Oncol. boplatin. J Clin Oncol. 1996;14:2984. 1 9 9 8 ; 2 5 : 7 . 42. Bezwoda WR, Seymour L, Dansey RD. High-dose chemotherapy 50. Honkoop AH, van der Wall E, Feller N, et al. Multiple cycles of with hematopoietic rescue as primary treatment for metastatic breast high-dose doxorubicin and cyclophosphamide with G-CSF mobilized cancer: a randomized trial. J Clin Oncol. 1 9 9 5 ; 1 3 : 2 4 8 3 . peripheral blood progenitor cell support in patients with metastatic 43. Hortobagyi GN, Bodey GP, Buzdar AU, et al. Evaluation of high- breast cancer. Ann Oncol. 1997;8:957. dose versus standard FAC chemotherapy for advanced breast cancer in 51. Huober J, Schneeweiss A, Hohaus S, et al. Tandem and triple protected environment units: a prospective randomized study. J Clin high-dose chemotherapy with autologous stem cell rescue in metastatic Oncol. 1987;5:354. breast cancer. J Cancer Res Clin Oncol. 1998;124:690. 4 4 . Yeung AW, Pang YK, Tsang YC, Wong SW. Double-cycle high- 52. Gianni A, Siena S, Bregna M, Di Nicola M, Orefice S. Effic a c y , dose chemotherapy with peripheral blood stem cells and hemato- toxicity, and applicability of high-dose sequential chemotherapy as poietic growth factor support in patients with a d v a n c e d solid tumor. A adjuvant treatment in operable breast cancer with 10 or more involved pilot study by the Hong Kong Biotherapy Group. C a n c e r . axillary nodes: five-year results. J Clin Oncol. 1 9 9 7 ; 1 5 : 2 3 1 2 . 1 9 9 4 ; 7 3 : 1 9 6 0 . 53. Viens P, Gravis G, Genre D, et al. High-dose sequential chemo- 45. M e i s e n b e r g BR, M i l l e r WE, M c M i l l a n R. Mobilized peripheral therapy with stem cell support for non-metastatic breast cancer. B o n e blood progenitor cells (PBPC) in support of tandem cycles of high- Marrow Transplant. 1997;20:199. dose c h e m o t h e r a p y (HDC). Bone Marrow Transplant. 1 9 9 6 ; 1 8 : 1 0 8 7 . 5 4 . Gianni AM, Bregni M, Siena S, et al. High-dose chemotherapy 4 6. Bitran JD, Samuels B, Klein L, et al. Tandem high-dose chemo- and autologous bone marrow transplantation compared with therapy supported by hematopoietic progenitor cells yields pro- MACOP-B in aggressive B-cell lymphoma. N Engl J Med. longed survival in stage IV breast cancer. Bone Marrow Transplant. 1 9 9 7 ; 3 3 6 : 1 2 9 0 . 1996;17:157. 55. Hudis C, Seidman A, Baselga J, et al. Sequential dose-dense dox- 4 7 . Lotz JP, Bouleuc C, Andre T, et al. Tandem high-dose chemo- orubicin, paclitaxel, and cyclophosphamide for resectable high-risk therapy with ifosfamide, carboplatin, and teniposide with autologous breast cancer: feasibility and efficacy. J Clin Oncol. 1999;17:93. bone marrow transplantation for the treatment of poor prognosis 5 6 . Fossati R, Confalonieri C, Torri V, et al. Cytotoxic and hor- common epithelial ovarian carcinoma. C a n c e r. 1996;77:2550. monal treatment for metastatic breast cancer: a systematic review of 48. Haas R, Schmid H, Hahn U, et al. Tandem high-dose therapy published randomized trials involving 31,510 women. J Clin Oncol with ifosfamide, epirubicin, carboplatin and peripheral blood stem cell 1 9 9 8 ; 1 6 : 3 4 3 9 . support is an effective adjuvant treatment for high-risk primary breast 57. Frei E III, Elias A, Wheeler C, Richardson P, Hryniuk W. The cancer. Eur J Cancer. 1997;33:372. relationship between high-dose treatment and combination chemo- 4 9 . H o h a u s S, Wallwiener D, Martin S, et al. Efficacy and toxicity therapy: the concept of summation dose intensity. Clin Cancer R e s . of sequential high-dose therapy with peripheral blood stem cell 1998;4:2027.