Inflammatory High-Risk Breast Cancer

Home , Busulfan, Thiotepa

Bone Marrow Transplantation (2000) 26, 51–59  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt High-dose busulfan, melphalan and thiotepa as consolidation for non- inﬂammatory high-risk breast cancer

F Gutierrez-Delgado, LA Holmberg, H Hooper, FR Appelbaum, RB Livingston, RT Maziarz, P Weiden, S Rivkin, P Montgomery, K Kawahara and W Bensinger

Fred Hutchinson Cancer Research, University of Washington and Puget Sound Oncology Consortium, Seattle, WA, USA

Summary: disease,3–7 and stage III disease8,9 have a probability of disease-free survival (DFS) at 5 years of 13 to 45% which The purpose of this study was to evaluate the toxicity falls to 10 to 30% at 10 years. and efficacy of high-dose busulfan, melphalan and thi- The experimental and clinical observations suggesting a otepa (Bu/Mel/TT) in patients with high-risk non- significant dose–response relationship in metastatic and pri- inflammatory breast cancer defined as stage II disease mary breast cancer10–14 have encouraged the evaluation of у10 lymph nodes (n = 52) or stage III (n = 69), and high-dose chemotherapy supported with autologous stem prognostic factors for treatment outcome. One hundred cell infusion (HDC/ASCI) as treatment for breast cancer. and twenty-one patients (median age, 46 years) were While a wide range of regimens and trials have evaluated treated with high-dose Bu (12 mg/kg), Mel (100 mg/m2) this approach in metastatic disease, relatively few trials and TT (500 mg/m2) (HDC) followed by autologous stem have investigated the therapeutic potential of HDC/ASCI cell infusion (ASCI). One hundred patients were in high-risk non-inflammatory breast cancer.15–22 The use of initially treated with surgery followed by standard adju- STAMP-I (cyclophosphamide, carmustine and cisplatin),15 vant chemotherapy prior to HDC/ASCI. Twenty-one STAMP-V (cyclophosphamide, thiotepa and patients with stage III disease had inoperable tumors at carboplatin),18–21 CAVP/CCVP (cyclophosphamide, etopo- diagnosis and were treated with neoadjuvant chemo- side and either cisplatin or doxorubicin)16 and CEP therapy and surgery before HDC/ASCI. Transplant- (cyclophosphamide, etoposide and cisplatin)22 has resulted related mortality was 6%. The probabilities of event- in event-free survival (EFS) probabilities ranging from 0.48 free survival (EFS) at 3 and 5 years (median follow-up to 0.72 at 3.5 years after diagnosis16,17,19,21 or from the first of 36 months) from transplant were, for all patients: day of adjuvant chemotherapy15,18,20,22 in patients with 0.62–0.60; stage II: 0.71–0.67: stage III: 0.55–0.55 (for high-risk non-inflammatory breast cancer. stage III adjuvant and neoadjuvant groups: 0.60–0.60 Studies carried out at the Fred Hutchinson Cancer and 0.42–0.42, respectively). Multivariate analysis did Research Center (FHCRC) have evaluated a regimen of not identify variables associated with poor outcome. The busulfan, melphalan and thiotepa, three drugs with efficacy of Bu/Mel/TT is similar to other HDC regimens observed activity in patients with hematologic malignancies reported for patients with high-risk non-inflammatory and patients with breast cancer. The rationale for use of breast cancer. Bu/Mel/TT has high activity in stage II this regimen is based on the recognized activity of mel- disease and a moderate benefit in stage III operable phalan and thiotepa as single agents in the treatment of tumors. Bone Marrow Transplantation (2000) 26, 51–59. breast cancer and the cytotoxic effect observed in meta- Keywords: high-dose chemotherapy; breast cancer; static breast cancer when high doses of busulfan are com- PBSC; busulfan; melphalan; thiotepa bined with cyclophosphamide.23–25 The busulfan, melphalan and thiotepa regimen has shown high activity in the treatment of patients with metastatic breast cancer.23–25 Bensinger et al25 reported estimated probabilities of EFS at Despite adjuvant chemotherapy at conventional doses, 1.5 years for patients with metastatic responsive and refrac- Ͼ patients with breast cancer who are stage II with 3 posi- tory disease as 0.53 and 0.24, respectively, with this regi- tive axillary lymph nodes or stage III at diagnosis have a men. The aim of this study was to evaluate the toxicity 1–9 high risk of relapse. At 5 years, the best reported relapse- and efficacy of this regimen in patients with high-risk non- free survival for patients with four to nine metastatic lymph inflammatory breast cancer, as well as to identify the prog- 1 2 nodes is 46–67%; at 10 years, 32–47%. Moreover, the nostic factors which could influence treatment outcome. risk of relapse increases with each additional metastatic lymph node found at surgery. Patients with stage II breast у cancer who have 10 lymph-nodes containing metastatic Patients and methods

After obtaining written informed consent, patients were Correspondence: Dr F Gutierrez-Delgado, Fred Hutchinson Cancer Research Center, 1100 Fairview Av North D5–390, PO Box 19024, treated at the FHCRC, the University of Washington or at Seattle, WA 98109–1024, USA hospitals affiliated with the Puget Sound Oncology Consor- Received 11 August 1999; accepted 10 April 2000 tium (PSOC) according to Institutional Review Board High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 52 approved protocols of the hospitals where the treatment was Table 1 Patient characteristics administered. Fifty patients were treated at the FHCRC and 71 in member hospitals of PSOC listed in the Appendix. Characteristics Stage The date of the last follow-up for patients in the present II III report was 21 July 1999. Between September 1993 and July 1997, 121 women n 52 69 Age, years, median (range) 47 (27–63) 46 (29–61) with high-risk non-inflammatory breast cancer, defined as Hormone receptor status (%) stage II у10 axillary lymph nodes (n = 52), stage IIIA ER+ PR+ 31 (60) 30 (43) (n = 62) or stage IIIB (n = 7),26 were treated with high-dose ER+ PR− 4 (8) 9 (13) busulfan, melphalan and thiotepa. Patients with stage IIIB ER−/PR− 4 (8) 4 (6) ER−/PR+ 12 (23) 20 (29) inflammatory carcinoma were excluded from this analysis. Unknown 1 (1) 6 (9) Eligibility criteria included age less than 65 years, adequate Number of metastatic lymph nodes, 13 (10–32) 10 (0–35) pulmonary, cardiac, hepatic and renal function and Karnof- median (range) sky score Ͼ70. Hepatitis B surface antigen (HbsAg) and 0–9 (%) 0 30 (44) human immunodeficiency virus serologically positive 10–14 (%) 30 (58) 19 (28) 15–19 (%) 15 (29) 9 (13) patients were excluded. Staging studies included computed Ͼ20 (%) 7 (13) 11 (16) tomography (CT) of chest and abdomen, bone scan and Tumor stage (%) bone marrow aspirations and biopsies. Cranial CT or mag- T1 17 (33) 1 (1) netic resonance imaging (MRI) of the head was performed T2 35 (67) 18 (26) T3 0 48 (70) if clinically indicated. Patients with stage II disease T4 (a) 0 2 (3) underwent modified radical mastectomy or lumpectomy Histology followed by standard-dose doxorubicin-based adjuvant Ductal 45 (87) 55 (80) chemotherapy prior to HDC/ASCI. Initial treatment for Lobular 5 (10) 12 (17) patients with stage III disease was determined by clinical Undifferentiated 2 (3) 2 (3) criteria of operable or inoperable tumors according to Haa- gensen’s classification.27 Patients with stage III disease (a) Extension to chest wall. diagnosed as having operable tumors were initially treated with surgery, followed by standard-dose doxorubicin-based Table 2 Therapy prior to HDC/ASCI adjuvant chemotherapy. Patients with stage III disease who had inoperable tumors received neoadjuvant chemotherapy Therapy Stage followed by surgery. In this study, patients with stage III II III disease were classified into two groups based on conventional chemotherapy administered following surgery n 52 69 (adjuvant chemotherapy group) or as initial treatment prior Surgery (%) Modified radical mastectomy 38 (73) 67 (97) to surgery (neoadjuvant chemotherapy group). Segmentectomy and node disection 14 (27) 2 (3) Chemotherapy (%) Doxorubicin-based chemotherapy 52 (100) 68 (99) Patient characteristics Number of cycles, median (range) 4 (3–6) 4 (2–13) Patient characteristics are shown in Table 1. Sixty and 71% Chemotherapy for stage III disease (%) Adjuvant chemotherapy 52 (100) 47 (68) of patients with stage II and stage III disease were younger Neoadjuvant chemotherapy 0 21 (31) than 50 years, respectively. The main differences between Chemotherapy regimen (%) patients with stage III disease beside the initial treatment CAF 30 (58) received were in the hormone receptor and lymph node AC 10 (19) 40 (59) Others: (Abeloff, A+Taxol, CMF)a 12 (23) 19 (28) status. Forty-eight (70%) patients with stage III disease had 9 (13) operable tumors and underwent modified radical mastec- Radiation therapy (%) tomy (n = 47) or lumpectomy (n = 1). Twenty-one (30%) Prior to HDC/ASCI 13 (25) 10 (14) patients were diagnosed as having inoperable tumors, and After HDC/ASCI 34 (65) 56 (81) after diagnostic biopsy, were treated with standard-dose doxorubicin-based induction chemotherapy. Twenty-four aAbeloff regimen13 = cyclophosphamide (C) + doxorubicin (A) + methotrexate (M) + vincristine + 5-fluorouracil (F). CMF for patients with stage (50%) patients of the adjuvant group were ER and PR hor- II disease. mone receptor positive compared to six (29%) patients of the neoadjuvant group, while 12 (20%) and eight (38%) patients were hormone receptor negative, respectively. The patients. One of the patients with stage III disease had a median number of metastatic lymph nodes for patients who history of idiopathic leukopenia. After mastectomy, she was received adjuvant chemotherapy after surgery was 11 not given adjuvant chemotherapy or radiotherapy and was (range 0–35) and five (range 1–18) for those treated with referred directly for HDC/ASCI using PBSC mobilized neoadjuvant chemotherapy prior to surgery. The median with recombinant human granulocyte colony-stimulating time between diagnosis and HDC/ASCI for the entire group factor (G-CSF) alone. Because this patient was diagnosed was 7 months (range 4–19). as having an operable tumor, she has been analyzed with Treatment prior to transplantation is shown in Table 2. the remaining 47 (68%) patients who received standard- Modified radical mastectomy was performed in 87% of dose doxorubicin-based adjuvant chemotherapy following

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 53 surgery. The combination of cyclophosphamide and adria- dose of Bu. Patients received Bu 1 mg/kg orally every 6 h mycin, with or without 5-fluorouracil, was the adjuvant on days −8, −7, and −6, for a cumulative dose of 12 mg/kg, chemotherapy regimen used in 77 and 87% of patients with melphalan 50 mg/m2/day intravenously on days −5 and −4, stage II and stage III disease, respectively. The median for a cumulative dose of 100 mg/m2 and thiotepa number of cycles administered was four, including those 250 mg/m2/day intravenously on days −3 and −2 for a total patients treated with neoadjuvant chemotherapy. Residual dose of 500 mg/m2. PBSC were thawed and infused 48 h primary tumor was found in 20 of 21 patients receiving after the last dose of thiotepa using previously published neoadjuvant chemotherapy. Six of the 21 patients received techniques.28,29 After the administration of HDC, all at least two cycles of consolidation chemotherapy after patients received prophylactic fluconazole. Prophylactic mastectomy with combinations of doxorubicin, taxol or cis- acyclovir was administered to patients serologically posi- platin. Fifteen patients with residual tumor after mastec- tive for Herpes simplex virus. Prophylactic antibiotics were tomy did not receive consolidation chemotherapy. Fourteen given when the absolute neutrophil count (ANC) was of these were referred directly for PBSC mobilization Ͻ0.5 × 109/l. G-CSF was administered at a dose of receiving one cycle of cyclophosphamide and taxol while 5 ␮g/kg/day after transplantation and was continued until one patient was mobilized with growth factor alone before engraftment at the discretion of the attending physicians HDC/ASCI therapy. Nineteen percent of all patients (n = 30). received radiation therapy prior to transplantation. Patients received HDC followed by infusion of autologous peripheral blood stem cells (PBSC) (n = 119), bone marrow and Post-transplant treatment = = PBSC (n 1) or syngeneic PBSC (n 1). Complete blood counts were measured daily until patients had a white blood cell count Ͼ2 × 109/l and self-supporting PBSC mobilization platelets Ͼ20 × 109/l. Local chest wall and axillary node radiotherapy was planned for all patients (n = 98) who did PBSC were collected after the administration of G-CSF not receive radiation prior to transplantation. Eight patients alone, or after the administration of chemotherapy using 2 2 did not receive post-transplant radiation therapy either due CY alone (4 g/m ); CY and taxol (T) (170–250 mg/m ) to regimen/transplant-related toxicity (stage II n = 4, stage (CT); CY and etoposide (E) 600 mg/m2 (CE); CE and = = = 2 III n 1) or refusal (stage II n 1, stage III n 2). Tamox- cisplatin (105 mg/m ) (CEP); doxorubicin and T; or ifen 20 mg p.o. daily for 5 years was given to all patients doxorubicin and CY (Table 3). PBSC collection and cryo- 28,29 with estrogen and/or progesterone receptor positive tumors. preservation was performed as previously described. Restaging with CT and bone scan or total body MRI was performed at 1-year post transplant, and thereafter accord- High-dose chemotherapy ing to clinical indications. Toxicity was evaluated during the first 100 days using No chemotherapy was given 4 weeks prior to the start of the Bearman scale.30 the conditioning regimen. Twenty-four hours prior to the administration of Bu, patients received phenytoin before the first dose of Bu and continuing until 24 h after the final Statistical analysis Univariate comparisons of proportions were made using a Table 3 Stem cell collection, infusion and engraftment after high- Fisher’s test. A Cox regression analysis was performed dose chemotherapy using the SAS system.31,32 Probabilities of survival and EFS were estimated from day 0 to the day of last contact and Stage from day 0 to the date of death or relapse censored by the II III date of last contact according to the method of Kaplan– Meier.33 Log-rank test was used to compare survival curves. n 52 69 PBSC mobilization (%) G-CSF alone 30 (58) 21 (31) Cy + taxol 17 (33) 33 (49) Others (Cy; T; CY+A; A+T; CEP; CE) 5 (9) 14 (20) Results Engraftment CD34 cells ×106/kg infused, PBSC mobilization and engraftment median 7.08 8.34 (range) (1.49–24.98) 2.26–51.63 Table 3 shows the proportion of patients who had PBSC CSF after transplant 9 (17%) 21 (30%) ANC Ͼ5 × 109/l 11 10 mobilized either with growth factor alone or with chemo- Median therapy, the numbers of stem cells infused after HDC and + (range) (9–16) (8–19) engraftment results. The median number of CD34 cells/kg Platelets Ͼ20 × 109/l infused was 7.59 × 106. Engraftment, defined as the first of Median of days 11 9 3 consecutive days on which the ANC was Ͼ0.5 × 109/l Range (7–46) (5–17) occurred in a median of 11 days (range 8–19). Platelet × 9 CSF = colony-stimulating factor; ANC = absolute neutrophil count; engraftment, defined as the first day with 20 10 /l unsup- CY = cyclophosphamide; T = taxol; A = doxorubicin; E = etoposide; ported for Ͼ7 days, was achieved in a median of 10 days P = cisplatin. (range 5–46).

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 54 Toxicity Table 5 Outcome in high-risk breast cancer patients treated with high- dose busulfan-melphalan thiotepa regimen supported with stem cell Regimen-related toxicity is shown in Table 4.30 The chemo- transplantation therapy regimen was generally well tolerated. Eighty-six percent of patients developed stomatitis grade II–III. Six- State II Stage III disease teen percent of the patients had gastrointestinal toxicity disease Adjuvant Neoadjuvant grade II–III, while grade II skin toxicity was observed in chemotherapy chemotherapy 10% of the patients. Hepatic toxicity grade II was observed n 52 48 21 in 2% of the patients. Pulmonary toxicity was the most Alive (%) 44 (85) 35 (73) 12 (57) frequent grade IV toxicity observed. There were five deaths In remission (%) 37 (71) 32 (67) 9 (43) due to regimen-related toxicity within 100 days of trans- Relapses (%) 11 (21) 13 (27) 12 (57) plantation. Three patients died on days 21, 34 and 61 sec- Deaths (%) ondary to pulmonary toxicity. One patient developed acute Disease progression 4 (8) 11 (23) 9 (43) Regimen-related toxicity 4 (8) 3 (6) 0 respiratory distress syndrome and in the other two patients pathology showed diffuse alveolar damage, mild interstitial inflammation, and alveolar cell dysplasia consistent with drug-induced injury. One patient died secondary to hepatic static recurrences 8–37 months after transplantation. Four veno-occlusive disease at day 37 after ASCI infusion. At patients with lung, pleura and bone metastases have died day 8 post transplant a 43-year-old patient with an unre- from disease progression in spite of salvage chemotherapy markable past medical history developed a sudden cardiac and radiation therapy. Two patients are still alive with arrest. The etiology, however, could not be determined disease with a follow-up of 34 and 42 months. from anatomic findings at autopsy. Since the onset of the terminal event was abrupt, cardiac toxicity was clinically Stage III: Forty-seven patients (68%) are alive between 21 suspected as the cause of death. Finally, two more patients and 63 months after transplantation with 41 (59%) in con- died on days 165 and 196 secondary to an idiopathic pneu- tinued remission. Twenty-five (36%) patients have monia syndrome and diffuse interstitial fibrosis, consistent relapsed, and 20 (29%) have died from disease progression. with Bu toxicity, respectively. Overall, transplant-related Three patients are not evaluable for relapse due to mortality was 6%. Twenty-five patients (20%) developed transplant-related death. Treatment response for patients documented bacterial or fungal infections, but there were with operable and inoperable tumors treated with adjuvant no deaths. and neoadjuvant chemotherapy, respectively, is shown in Table 5. Outcome (Table 5) Adjuvant chemotherapy group: Thirty-five of 48 (73%) Stage II: Forty-four of 52 patients (85%) are alive with 37 patients are alive, and 32 (67%) including the patient (71%) in continued remission at a median of 36 months treated with HDC followed by syngeneic transplant have no (range 18–64) follow-up. Four patients are not evaluable evidence of disease. Thirteen patients have relapsed. Three for relapse due to transplant-related death. Eleven patients patients relapsed with chest wall recurrences 2, 5 and 9 (21%) have relapsed. Three relapses were local or nodal months after transplantation, followed by visceral meta- recurrences at 6, 8 and 32 months after transplantation. stases and died. Ten patients have relapsed at distant sites These patients were treated with local excision and/or radi- after treatment with HDC/ASCI (median 21 months). In ation therapy and tamoxifen and have not progressed. Two five, bone metastases were the first metastatic site detected patients developed contralateral infiltrative ductal carci- between 9 and 34 months after transplantation. All five noma 9 and 24 months after HDC/ASCI therapy. They patients have died. Five patients developed visceral meta- were treated with mastectomy and adjuvant chemotherapy stases 6–34 months after HDC/ASCI therapy. Two patients and remain disease-free. Six patients have developed meta- with liver and lung metastases have received salvage chemotherapy and remain alive with disease after 42 and 51 months of follow-up since transplantation, respectively. Table 4 Regimen-related toxicity30 Neoadjuvant chemotherapy group: Twelve (57%) of 21 Category Toxicity grade patients with stage III disease who received neoadjuvant chemotherapy are alive, nine (43%) without evidence of I II III IV disease. Twelve patients (57%) relapsed and nine (43%) of Pulmonary (%) 6 (5) 3 (2) 0 3 (2) these died from disease progression. Three patients with Cardiac (%) 3 (2) 4 (3) 0 1 (Ͻ1) local recurrence in the chest wall, followed by distant recur- Skin (%) 37 (31) 12 (10) 0 0 rence 7, 8 and 9 months after HDC/ASCI treatment have Hepatic (%) 30 (25) 2 (2) 0 1 (Ͻ1) died. One patient developed bone metastases 13 months Ͻ Stomatitis (%) 15 (12) 103 (85) 1 ( 1) 0 after transplantation and died. In six patients, bone, liver, Gastrointestinal (%) 66 (55) 18 (15) 1 (Ͻ1) 0 Renal (%) 1 (Ͻ1) 5 (4) 0 0 lung and brain metastases were diagnosed 4–11 months Hemorrhagic cystitis (%) 5 (4) 3 (2) 3 (2) 0 after HDC/ASCI therapy. Except for two patients who developed lung and liver metastases 8 and 19 months after Values given are number of patients (%). transplantation, respectively and remain alive with disease

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 55 1 receptor status, and lymph node status on outcome were not signiﬁcant. Patients with stage III disease treated with 0.8 adjuvant chemotherapy tended to have better survival and EFS than patients who required neoadjuvant chemotherapy (0.60 vs 0.42) but this difference was not statistically 0.6 signiﬁcant (P = 0.9).

0.4 Multivariate analysis: Age, hormonal status, number of lymph nodes, tumor stage, stage of disease and patients 0.2 with stage III of disease classiﬁed according to the modality of conventional chemotherapy administered prior to trans-

0 plantation (adjuvant or neoadjuvant chemotherapy) were 01 2 3 4 56analyzed for the whole group for survival and EFS. A trend Years after transplantation for poor outcome was identiﬁed in the group of patients with stage III disease who initially were treated with neo- Figure 1 Kaplan–Meier event-free survival for stage II disease (ȦȦ) adjuvant chemotherapy (risk ratio 1.939, CI, 0.91–4.11 and stage III disease (——) breast cancer patients treated with high-dose = busulfan, melphalan and thiotepa followed by autologous stem cell P 0.08), but this did not reach statistical signiﬁcance. infusion.

Discussion with a follow-up of 30 and 20 months, the remaining patients have died. In one more patient, a contralateral duc- In the current study 121 patients with high-risk non- tal carcinoma was diagnosed 29 months after transplan- inflammatory breast cancer were treated with high-dose Bu, tation. She is without evidence of disease again after melphalan and thiotepa followed by ASCI. Transplant- treatment with surgery and chemotherapy. related mortality with this regimen was 6%, higher than STAMP-V,18–21 CEP22 and CAVP/CCVP,16 but lower than Survival: With a median follow-up of 36 months (18–64) STAMP-I.15 Seventy-eight (64%) patients remain alive in since transplantation, the 3-year probability of EFS from continued remission at a median follow-up of 3 years from time of transplant for all patients is 0.62 with projected EFS transplant (range 1.5–5.3). The probabilities of event-free at 5 years of 0.60. The Kaplan–Meier EFS estimates for survival at 3 and 5 years are, 0.62 and 0.60 which is stage II and stage III are 0.71 and 0.55, respectively comparable to other high-dose chemotherapy regimens (Figure 1). The probabilities of EFS at 3 and 5 years for administered in this setting. the 48 stage III patients who were treated with adjuvant Standard-dose chemotherapy has only a modest impact chemotherapy are 0.60 while for the 21 patients initially on DFS for patients with high-risk breast cancer. High- treated with neoadjuvant chemotherapy, they are 0.42. dose chemotherapy studies are designed to evaluate the (Figure 2). potential improvement in these survival figures. The role of HDC and stem cell transplantation has been evaluated Prognostic factor analysis more extensively in metastatic breast cancer, but few phase II trials have studied the therapeutic benefit of this Clinical variables were analyzed to determine effect on out- aggressive treatment strategy in patients with high-risk non- come. The influences of age, stage of disease, hormone inflammatory disease. Table 6 summarizes the principal features of the reported clinical trials which include 30 or more patients with stage II (у10 lymph nodes) and stage 1 III non-inflammatory breast cancer patients with a minimal follow-up of 2 years.15–22 Two treatment strategies 0.8 have been evaluated in these trials: HDC/ASCI either as consolidation therapy after initial conventional dose 15,16,18–22 17 0.6 therapy or as first-line therapy. As shown in Table 6, 616 high-risk non-inflammatory breast cancer patients with a median age of 44 years have been treated 0.4 in phase II trials. Five randomized phase II clinical trials which compared HDC/ASCI to standard-dose adjuvant 0.2 chemotherapy have been conducted in this setting.20,22,34–36 One of the five randomized trials reported have evaluated 36 0 HDC/ASCI as first-line therapy. However, an on-site 01 2 3 4 56review of that trial did not confirm the validity of the Years after transplantation study.37 Two published studies included 4120 and 39 patients.22 With a median follow-up from 24 to 78 months, Figure 2 Kaplan–Meier event-free survival analysis of patients with stage III disease breast cancer treated with HDC/ASCI according to initial the probabilities of EFS for all patients treated range treatment with surgery followed by adjuvant chemotherapy (ȦȦ) or neo- between 0.48 and 0.72. EFS estimates for stage II are 0.71 adjuvant chemotherapy (——) followed by surgery. and 0.52 to 0.57 for stage III patients, respectively. The

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 56 Table 6 Selected clinical trials in high-risk non-inﬂammatory breast cancer according to the number of patients evaluated

Author No. patients Age (median) Regimen Death All patients Stage II Stage III Follow-up years toxicity % (median S EFS S EFS S ESF months) %%%

FHCRC (2000) 121 46 Bu/Mel/TT 6 0.76 0.62 0.83 0.71 0.69 0.55 36c Schwartzberg (1998) 96 46 CTCb 0 0.77 0.61 0.86 0.71 0.70 0.52 53b Somlo (1997) 92 44 CAVP 1 NR 0.82 0.71 0.79 0.57 46a Peters (1993) 85 38 CBP 12 0.79 0.72 NR NR 36b Gianni (1997) 63 43 CMtxMit 2 0.70 0.57 NR NR 48a Rodenhuis (1998) 41 45 CTCb 0 0.75 0.54 NR NR 49b Thomas (1997) 40 43 CTCb 0 0.89 0.72 NR NR 35b Hortobagyi (2000) 39 45 CEP 3 0.58 0.48 NR NR 78b Garcia-Carbonero (1997) 39 43 CTCb ± Mit 3 0.85 0.71 NR NR 24a

NR = not reported; Bu = busulfan; Mel = melphalan; TT = thiotepa; C = cyclophosphamide; Cb = carboplatin; A = doxorubicin; V or E = etoposide; P = cisplatin; B = carmustine; Mtx = methotrexate; Mit = mitoxantrone. Median follow-up: afrom diagnosis; bfrom start induction chemotherapy; cfrom transplant.

Autologous Blood and Marrow Transplant Registry of the most common grade I–II toxicity and has been observed North America (ABMTR) has reported estimates of pro- in 80% to 100% of patients, while the main grade III/IV gression-free survival at 3.5 years for 750 patients with toxicities are gastrointestinal, hepatic, cardiac, and pulmon- stage II, and for 603 patients with stage III non-inflamma- ary ranging from 2% to 6%. In the current study, grade tory breast cancer of 0.65 and 0.60, respectively. The I–II gastrointestinal toxicity was the most frequent non- median age of patients reported was 44 years.38 hematological toxicity observed and there were five deaths The median follow-up reported for the majority of phase (4%) within 100 days secondary to pulmonary, hepatic and II trials in high-risk breast cancer is greater than 3 years cardiac toxicities. Two more deaths (2%) occurred on days (24–78 months) from diagnosis16,17,19,21 or from the first day 165 and 196. Thus, in spite of comparable efficacy of of adjuvant chemotherapy.15,18,20,22 An important difference Bu/Mel/TT with other regimens evaluated for high-risk in our study in comparison with those previously non inflammatory breast cancer, the 6% transplant-related reported15–22 is the method of calculating survival and EFS. mortality reported in this study is the main disadvantage Because this study is focused on the impact of transplan- of this regimen compared to STAMP-V or CAVP/CCVP. tation as consolidation of adjuvant treatment on disease, However, this death rate could be over-represented. A pre- rather than evaluating the influence on outcomes achieved treatment history analysis showed that four of seven by a specific adjuvant chemotherapy regimen followed by patients who died could be at highest risk for developing HDC/ASCI, the survival and EFS have been calculated transplant/regimen-related mortality. In the absence of a from the date of transplant, rather than from the date of remarkable past medical history, age may be an important diagnosis or start of adjuvant chemotherapy. This method risk factor for regimen-related toxicity. Two of the three excludes the period of time prior to transplantation and thus patients included in this study who were older than 60 years eliminates the bias generated when the time origin is calcu- died secondary to pulmonary toxicity and hepatic veno- lated from diagnosis/start of chemotherapy.39 This period occlusive disease. Two patients died after 100 days, one of time ranges from 3 to 6.5 months in the studies who was cytomegalovirus positive pretransplant and one reported.15,16,18–22 with a significant history of smoking who died of diffuse The best high-dose chemotherapy regimen for the treat- pulmonary fibrosis. ment of high-risk non-inflammatory disease is unknown, Since the efficacy of Bu/Mel/TT regimen is comparable but most current regimens are less toxic than the initial to other regimens evaluated, efforts to diminish the toxicity regimen of CY, cisplatin and carmustine (STAMP-I) which of this regimen are warranted. Studies to diminish hepatic produced a 12% mortality.15 With greater experience, mor- and pulmonary toxicities could be designed taking advan- tality from the STAMP I regimen has been reduced to 8%.34 tage of the availability of intravenous Bu, which could dim- The mortality of STAMP-V,40 a regimen based on CY, thi- inish interpatient variability of oral absorption. Bu i.v. otepa and carboplatin and used in the majority of patients administered by continuous infusion over 3–4 days to elim- with high-risk non-inflammatory breast cancer, is less than inate peak concentrations is another approach which could 1%.18–21 Other regimens have included the combination of be evaluated.43 Novel agents such as IL-11 or keratinocyte CY, etoposide, and either doxorubicin or cisplatin growth factor warrant further evaluation to reduce oral and (CAVP/CCVP)16 or (CEP)22 reporting a mortality between gastrointestinal toxicities. Moreover, if further follow-up 3%22 and Ͻ1%.16 The Bu/Mel/TT regimen reported here corroborates the effectiveness of Bu/Mel/TT regimen in was developed because of the demonstrated activity of the stage II disease, the aim of future studies should be to individual drugs in breast cancer and the ability to signifi- decrease toxicity without loss of effectiveness. cantly dose-escalate the drug in this regimen with the use Prognostic factors associated with an adverse outcome of stem cell support. Toxicity secondary to Bu/Mel/TT regi- in the high-risk non-inflammatory disease treated with men has been described previously.23–25,41,42 Mucositis is HDC/ASCI have been evaluated in two trials.16,21 Somlo et

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 57 al16 have identified progesterone receptor negativity and cancer is a heterogeneous disease, further studies are war- high-grade tumors in the high-risk non-inflammatory set- ranted to establish the role of HDC/ASCI in the treatment ting as adverse prognostic factors. By univariate analysis, of these patients. age Ͻ45, estrogen/progesterone negativity, N2 stage, and locally advanced breast cancer were identified as adverse prognostic factors in a series of 92 patients reported by Acknowledgements Schwartzberg et al.21 In that study all patients were treated with surgery followed by adjuvant chemotherapy prior to This study was supported by grants from the National Institutes transplantation. The 24 patients with stage III locally of Health CA 47748, CA 18221, CA 15704, CA 09515 and the advanced disease had the poorest survival, and the prob- Jose Carreras International Leukemia Foundation. FG-D is sup- ability of EFS at 4 years for this group of patients was ported by a grant from Instituto Mexicano del Seguro Social 0.37. The multivariate analysis confirmed that ER negative (IMSS), Mexico City, Mexico. patients and patients with advanced disease had a poor outcome. Our study included 69 patients with stage III disease, References 48 (70%) of whom were treated with surgery followed by adjuvant chemotherapy and 21 (30%) patients who had 1 Buzzoni R, Bonadonna G, Valagussa P, Zambetti M. Adjuvant inoperable tumors were treated initially with neoadjuvant chemotherapy with doxorubicin plus cyclophosphamide, chemotherapy followed by surgery. The probability of EFS methotrexate, and fluorouracil in the treatment of resectable at 3 and 5 years for the group of patients treated with adju- breast cancer with more than three positive axillary nodes. J vant chemotherapy was 0.60, which is superior to the 4- Clin Oncol 1991; 9: 2134–2140. year EFS of 0.37 reported by Schwartzberg et al.21 In our 2 Bonadonna G, Zambetti M, Valagussa P. Sequential or alter- series, the 21 patients with stage III disease who had inop- nating doxorubicin and CMF regimens in breast cancer with erable tumors and were treated with neoadjuvant chemo- more than three positive nodes. Ten-year results. JAMA 1995; therapy followed by surgery achieved an EFS at 3 and 5 273: 542–547. years of 0.42 and were identified as having the worst prog- 3 Jones SE, Moon TE, Bonadonna G et al. Comparison of dif- nosis of any group. The aggressive nature of the breast can- ferent trials of adjuvant chemotherapy in stage II breast cancer using a natural history data base. Am J Clin Oncol 1987; 10: cer in these patients was demonstrated by the high pro- 387–395. portion of patients (95%) who, in spite of the neoadjuvant 4 Carter CL, Allen C, Henson DE. Relation of tumor size, chemotherapy, had residual tumor and a median of five lymph node status, and survival in 24 740 breast cancer cases. metastatic lymph nodes at the time of mastectomy. The Cancer 1989; 63: 181–187. presence of residual disease after neoadjuvant chemo- 5 Davidson NE, Abeloff MD. Adjuvant systemic therapy in therapy has been confirmed as an extremely poor prognos- women with early-stage breast cancer at high risk for relapse. tic factor.44 However, in our study the univariate and the J Natl Cancer Inst 1992; 84: 301–305. multivariate analyses did not identify differences in the 6 Buzdar AU, Kau SW, Hortobagyi GN et al. Clinical course impact of HDC/ASCI consolidation on survival or EFS of patients with breast cancer with ten or more positive nodes when patients with stage III disease were analyzed accord- who were treated with doxorubicin-containing adjuvant therapy. Cancer 1992; 69: 448–452. ing to the initial treatment received before transplantation. 7 Bonadonna G, Valagussa P, Moliterni A et al. Adjuvant cyclo- The multivariate analysis showed a trend for adverse out- phosphamide, methotrexate, and fluorouracil in node-positive come of patients with stage III inoperable tumors treated breast cancer: the results of 20 years of follow-up. New Engl with neoadjuvant chemotherapy (risk ratio 1.939, CI 0.91– J Med 1995; 332: 901–906. 4.11 P = 0.08), but failed to identify other adverse prognos- 8 Piccart MJ, de Valeriola D, Paridaens R et al. Six-year results tic factors as reported by Somlo et al16 and Schwartzberg of a multimodality treatment strategy for locally advanced et al.21 This fact could be explained in part by smaller num- breast cancer. Cancer 1988; 62: 2501–2506. bers of patients with ER/PR negativity treated in our series. 9 Hortobagyi GN, Singletary SE, McNeese HD. Treatment of In both studies reported, the distribution of patients accord- locally advanced and inflamamtory breast cancer. In: Harris ing to hormonal status was well balanced. In Somlo’s JR, Lippman ME, Morrow M, Hellman S (eds). Diseases of the Breast. Lippincot-Raven: Philadelphia, 1996, pp 585–599. study,16 56% of patients were PR or ER positive while 21 10 Hryniuk W, Bush H. The importance of dose-intensity in Schwartzberg et al reported 55% and 40% of patients with chemotherapy of metastatic breast cancer. J Clin Oncol 1984; ER/PR positive and ER/PR negative, respectively. In con- 2: 1281–1288. trast, in the FHCRC/PSOC series, two-thirds (68%) of 11 Tannock IF, Boyd NF, De Boer G et al. A randomized trial patients were ER or PR positive and one-third (26%) were of two dose levels of cyclophosphamide, methotrexate, and ER and PR negative. fluorouracil chemotherapy for patients with metastatic breast The role of HDC/ASCI as a feasible modality of treat- cancer. J Clin Oncol 1988; 6: 1377–1387. ment for high-risk breast cancer patients has been ques- 12 Bonnadona G, Valagussa P. Dose-intense adjuvant treatment tioned in the context of negative results reported in two of high-risk breast cancer. J Natl Cancer Inst 1990; 82: 53–54. randomized trial, published in article form,20,22 and two pre- 13 Abeloff MD, Beveridge RA, Donehower RC et al. Sixteen- 34,35 week dose-intense chemotherapy in the adjuvant treatment of liminary reports published in abstract form comparing breast cancer. J Natl Cancer Inst 1990; 82: 570–574. HDC to a control group. These ongoing phase III ran- 14 Wood CW, Budman DR, Korzum AH et al. Dose and dose domized studies evaluating HDC/ASCI as consolidation intensity of adjuvant chemotherapy for stage II, node-positive will need further follow-up and definitive results will not breast carcinoma. New Engl J Med 1994; 330: 1253–1259. be available for at least 2–4 or more years. Since breast 15 Peters WP, Ross M, Vredenburgh JJ et al. High-dose chemo-

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 58 therapy and autologous bone marrow support as consolidation 30 Bearman S, Appelbaum F, Buckner C et al. Regimen-related after standard-dose adjuvant therapy for high-risk primary toxicity in patients undergoing bone marrow transplantation. breast cancer. J Clin Oncol 1993; 11: 1132–1143. J Clin Oncol 1988; 6: 1562–1568. 16 Somlo G, Doroshow J, Forman S et al. High-dose chemo- 31 Cox DR. Regression models and life tables. JR Stat Soc B therapy and stem cell rescue in the treatment of high-risk 1972; 34: 187–202. breast cancer: prognostic indicators of progression-free and 32 SAS Institute Inc. Technical report P-217, SAS/STAT overall survival. J Clin Oncol 1997; 15: 2882–2893. Software; the PHREG Procedure Version 6. SAS Institute Inc: 17 Gianni A, Siena S, Bregni M et al. Efficacy, toxicity and Cary, NC, 1991, p 63. applicability of high-dose sequential chemotherapy as adju- 33 Kaplan EL, Meier P. Nonparametric estimation from incom- vant treatment in operable breast cancer with 10 or more plete observations. J Am Stat Assoc 1958; 53: 457–481. involved axillary nodes; five-year results. J Clin Oncol 1997; 34 Peters WP, Rosner G, Vredenburgh JJ et al. A prospective, 15: 2312–2321. randomized comparison of two doses of combination alkylat- 18 Tomas JF, Perez-Carrion R, Escudero A et al. Results of a ing agents as consolidation after CAF in high-risk primary pilot study of 40 patients using high-dose therapy with hema- breast cancer involving ten or more axillary lymph nodes: pre- topoietic rescue after standard-dose adjuvant therapy for high- liminary results of CALGB 9082/SWOG 9114/NCIC MA-13. risk breast cancer. Bone Marrow Transplant 1997; 19: 661– Proc Am Soc Clin Oncol 1999; 18 (Abstr. 2). 670. 35 The Scandinavian Breast Cancer Study Group. Results from 19 Garcia-Carbonero R, Hidalgo M, Paz-Ares L et al. Patient a randomized adjuvant breast cancer study with high dose selection in high-dose chemotherapy trials: relevance in high- chemotherapy with CTCb supported by autologous bone mar- risk breast cancer. J Clin Oncol 1997; 15: 3178–3184. row stem cells vs dose escalated and tailored FEC therapy. 20 Rhodenhuis S, Richel D, van der Wall E et al. Randomised Proc Am Soc Clin Oncol 1999; 18 (Abstr. 3). trial of high-dose chemotherapy and haemopoietic progenitor- 36 Bezwoda WR. Randomised, controlled trial of high dose cell support in operable breast cancer with extensive axillary chemotherapy (HD-CNVp) vs standard dose (CAF) chemo- lymph-node involvement. Lancet 1998; 352: 515–521. therapy for high risk, surgically treated, primary breast cancer. 21 Schwartzberg L, Birch R, West W et al. Sequential treatment Proc Am Soc Clin Oncol 1999; 18 (Abstr. 4). including high-dose chemotherapy with peripheral blood stem 37 Weiss RB, Rifkin RM, Stewart FM et al. High-dose chemo- cell support in patients with high-risk stage II–III breast can- therapy for high-risk primary breast cancer: an on-site review cer. Am J Clin Oncol 1998; 21: 523–531. of the Bezwoda study. Lancet 2000; 355: 999–1003. 22 Hortobagyi GN, Buzdar AU, Theriault RL et al. Randomized 38 Antman K, Rowlings P, Vaughan W et al. High-dose chemo- trial of high-dose chemotherapy and blood cell autografts for therapy with autologous hematopoeitic stem-cell support for high-risk primary breast carcinoma. J Natl Cancer Inst 2000; breast cancer in North America. J Clin Oncol 1997; 15: 92: 225–233. 1870–1879. 23 Weaver C, Bensinger W, Appelbaum F et al. Phase I study 39 Anderson JR, Cain KC, Gelber RD, Gelman RS. Analysis and of high-dose busulfan, melphalan and thiotepa with autolog- interpretation of the comparison of survival by treatment out- ous stem cell support in patients with refractory malignancies. come variables in cancer clinical trials. Cancer Treat Rep Bone Marrow Transplant 1994; 14: 813–819. 1985; 69: 1139–1144. 24 Schiffman K, Bensinger W, Appelbaum F et al. Phase II of 40 Antman K, Ayash L, Elis A et al. A phase II study of high- high-dose busulfan, melphalan and thiotepa with autologous dose cyclophosphamide, thiotepa, and carboplatin with auto- peripheral blood stem cell support in patients with malignant logous marrow support in women with measurable advanced disease. Bone Marrow Transplant 1996; 17: 943–950. breast cancer responding to standard-dose therapy. J Clin 25 Bensinger W, Schiffman, Holmberg L et al. High-dose busul- Oncol 1992; 10: 102–110. fan, melphalan, thiotepa and peripheral blood stem cell 41 Schiffman K, Buckner C, Maziarz R et al. High-dose busul- infusion for the treatment of metastatic breast cancer. Bone fan, melphalan, and thiotepa followed by autologous periph- Marrow Transplant 1997; 19: 1183–1189. eral blood stem cell transplantation in patients with aggressive 26 American Joint Committee on Cancer. Breast. In: Fleming I, lymphoma or relapsed Hodgkin’s disease. Biol Blood Marrow Cooper J, Henson D et al (eds). AJCC Cancer Staging Man- Transplant 1997; 3: 261–266. ual, 5th edn. Lippincot-Raven: Philadelphia, 1997, pp 171– 42 Holmberg L, Demirer T, Rowley S et al. High-dose busulfan, 180. melphalan and thiotepa followed by autologous peripheral 27 Haagensen C. Clinical classification of the stage of advance- blood stem cell (PBSC) rescue in patients with advanced stage ment of breast carcinoma. In: Diseases of the Breast, 3rd edn. III/IV ovarian cancer. Bone Marrow Transplant 1998; 22: WB Saunders: Philadelphia, 1986, pp 851–863. 651–659. 28 Bensinger W, Singer J, Appelbaum F et al. Autologous trans- 43 Grochow LB. Busulfan disposition: the role of therapeutic plantation with peripheral blood mononuclear cells collected monitoring in bone marrow transplantation. Semin Oncol after administration of recombinant granulocyte stimulating 1993; 20 (Suppl. 4): 18–25. factor. Blood 1993; 81: 3158–3163. 44 Kuerer HM, Newman LA, Smith TL et al. Clinical course of 29 Rowley SD, Bensinger W, Gooley T et al. The effect of cell breast cancer patients with complete pathologic primary tumor concentration on bone marrow and peripheral blood stem cell and axillary lymph node response to doxorubicin-based neo- cryopreservation. Blood 1994; 83: 2731–2736. adjuvant chemotherapy. J Clin Oncol 1999; 17: 460–469.

Bone Marrow Transplantation High-dose chemotherapy for high-risk breast cancer F Gutierrez-Delgado et al 59 Appendix: Puget Sound Oncology Consortium

Hospitals and physicians participants

St Luke’s Medical Center/Mountain States Tumor Institute, Boise, ID, W Urba USA S Lewis W Kreisle P Montgomery Mercy Regional Medical Center/Cancer Care Consultants, Redding, N Zuckerman CA, USA M Gearn J Freeman R Dorn M Figueroa CG Carter Benefis West Hospital (former Columbus Hospital)/Big Sky Health Care, Great Falls, MT, USA Sacred Heart Medical Center/Rockwood Clinic, Spokane, WA, USA G Harrer Holy Family Hospital, Spokane, WA, USA T Warr J Wittenkeller K Lund Group Health Cooperative of Puget Sound, Seattle, WA, USA K Kraemer E Feldman R Gersh SK Callahan J Nichols S Ginsburg B Cutter In the past: D Dong T Wood H Stang The Queen’s Medical Center/The Queen’s Cancer Center, Honolulu, R Laugen HI, USA J Chestnut K Kawahara C Chong Capital Medical Center, Olympia, WA, USA K Loh D Osborn K Sumida P Robertson C Higuchi St Peter Medical Center/Memorial Clinic, Olympia, WA, USA St Francis Hospital, Honolulu, HI, USA J Lechner C Chong L Harper H Griffith Oregon Health Sciences University, Portland, OR, USA R Maziarz Swedish Medical Center/Tumor Institute, Seattle, WA, USA J Leis S Rivkin W Fleming E Ellis P Curtin H Kaplan D Stepan G Goodman C Marquez M Milder K Olson M Brunvand St Joseph Medical Center/Northwest Medical Specialties, Tacoma, L Johnson WA, USA G Jones J Klarnet F Senecal Providence Medical Center, Seattle, WA, USA T Baker R Bowden S Goldberg University of Washington Medical Center, Seattle, WA, USA M Fer R Livingston G Ellis Good Samaritan Medical Center/Rainier Oncology, Puyallup, WA, O Press USA S Petersdorf R Goldberg J Gralow R McCroskey C Higano R Ostenson J Thompson N Robinson S Liu L Bavisotto Providence Medical Center, Portland, OR, USA J Menashe B Dana Virginia Mason Medical Center/Cancer Clinic, Seattle, WA, USA P Kane P Weiden D Gruenberg A Jacobs K Olson R Rudolph E Soo D Aboulafia R Look V Picozzi K Hansen G Segal St Patrick Medical Center/Missoula Medical Oncology, Missoula, MT, FEy USA K Lanier W Nichols G Takahashi M Snyder

Bone Marrow Transplantation