High-Dose Busulfan, Melphalan and Thiotepa Followed by Autologous Peripheral Blood Stem Cell (PBSC) Rescue in Patients with Advanced Stage III/IV Ovarian Cancer

Home , Busulfan

Bone Marrow Transplantation, (1998) 22, 651–659  1998 Stockton Press All rights reserved 0268–3369/98 $12.00 http://www.stockton-press.co.uk/bmt High-dose busulfan, melphalan and thiotepa followed by autologous peripheral blood stem cell (PBSC) rescue in patients with advanced stage III/IV ovarian cancer

LA Holmberg1,2,3, T Demirer1, S Rowley1,2, CD Buckner1,2,3,4, G Goodman3, R Maziarz3, J Klarnet3, N Zuckerman3, G Harrer3, R McCloskey3, R Gersh3, R Goldberg3, W Nichols3, A Jacobs3, P Weiden3, P Montgomery3, S Rivkin3, FR Appelbaum1,2,3 and WI Bensinger1,2,3

1Fred Hutchinson Cancer Research Center, Seattle; 2University of Washington School of Medicine, Seattle; 3Puget Sound Oncology Consortium, Seattle, WA, USA

Summary: 130 000 women in the USA will be diagnosed with ovarian cancer and more than 75 000 will die from their disease.2 The purpose of this study was to evaluate the efficacy At diagnosis, 75 to 85% of patients have advanced stage of high-dose chemotherapy (HDC) with busulfan, mel- III or IV disease,3 with a 5-year survival with conventional phalan and thiotepa (BUMELTT) followed by autolog- therapy of only 15–25% for stage III and less than 5% for ous PBSC infusion in treating patients with advanced stage IV disease.4 Clearly, there is a need for more effective ovarian cancer. Thirty-one patients, 18 with stage therapy for this disease. III/IIIc and 13 with stage IV ovarian cancer, were Early trials of high-dose chemotherapy (HDC) with auto- treated with BU (12 mg/kg), MEL (100 mg/m2) and TT logous stem cell rescue demonstrated initial high response (500 mg/m2) and autologous PBSC rescue. Fifteen rates for patients with relapsed ovarian cancer both for sin- patients were in clinical complete remission (CR) at gle agents and chemotherapy combinations.5–14 Patients treatment; 11 had platinum-sensitive disease. Sixteen treated had usually failed at least two prior regimens and patients were not in CR; two had platinum-sensitive dis- had developed progressive disease either during or within ease. The probabilities of overall survival (OS), event- 6 months of achieving a remission with platinum therapy. free survival (EFS) and relapse (R) for all patients at Although initial response rates of 55–75% have been 18 months were 0.57, 0.30 and 0.63; for patients in CR, reported, remission durations were consistently short-lived, the rates were 0.87, 0.44 and 0.49 and for patients not usually only 5–7 months. in CR, 0.38, 0.13 and 0.81. Two patients (6.5%) died of Because long-term event-free survival has been seen in treatment-related causes. Among the 13 patients with only 5–25% of all ovarian cancer patients treated with HDC platinum-sensitive disease, all are still alive, with seven regimens, there is a need to improve the conditioning regi- having relapsed 129–1021 days after PBSC infusion. men. A conditioning regimen of busulfan, melphalan and OS, EFS and R were 1.00, 0.52 and 0.48. Of the 18 thiotepa (BUMELTT) was developed in a phase I clinical patients with platinum-resistant disease, four remain trial at Fred Hutchinson Cancer Research Center (FHCRC). alive (two in remission). Six patients did not respond These three drugs were selected because they can be admin- and eight relapsed from days 104–429. The OS, EFS istered in an outpatient setting and they have demonstrable and R were 0.33, 0.11 and 0.78. We conclude that activity against adenocarcinoma, including breast15–20 and BUMELTT is well tolerated in patients with advanced ovarian cancer.7,9,10,15,18,20,21 Finally, their dose-limiting ovarian cancer and results are equivalent to other toxicity is marrow ablation which is easily resolved by the published HDC regimens. infusion of PBSC. In the phase I clinical trial, the maximum Keywords: busulfan; melphalan; thiotepa; PBSC; high- tolerated dose of TT was 500 mg/m2, when given with a dose chemotherapy; ovarian cancer fixed dose of 12 mg/kg BU and 100 mg/m2 MEL.20 This report presents the results of BUMELTT followed by autologous PBSC rescue in treating patients with advanced stage III/IV ovarian cancer. Ovarian cancer is the principal cause of gynecological death in the USA and ranks fifth in causes of death from 1 cancer in women. Over the next 5 years, more than Patients and methods

Between 9 February 1994 and 2 April 1996, 31 patients Correspondence: Dr LA Holmberg, Fred Hutchinson Cancer Research with advanced ovarian cancer (15 patients in clinical CR Center, 1100 Fairview Ave N, AC-133, PO Box 19024, Seattle, WA and 16 patients with persistent disease) were treated with 98109-1024, USA 4Current address: Response Oncology, 600 Broadway, Seattle, WA BUMELTT, followed by autologous PBSC infusion. 98122, USA Patients were eligible for this protocol if they had a bili- Received 15 December 1997; accepted 11 May 1998 rubin Ͻ2 mg/dl, a creatinine clearance 50 mg/min and a Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 652 Karnofsky score у70. Ten patients were treated at FHCRC histologically by damaged endothelial cells of the terminal or Oregon Health Sciences Center, Portland, OR, USA and hepatic venules, dilation of the sinusoids and necrosis of 21 patients were treated at seven participating community hepatocytes.22 Idiopathic pneumonia syndrome (IPS) was cancer centers under the auspices of the Puget Sound defined as pulmonary infiltrates not associated with pul- Oncology Consortium. Signed consent was obtained from monary edema or an identifiable infectious agent.23 The all patients to be treated on the FHCRC or Puget Sound incidence of infection is excluded from this grading system Oncology Consortium protocol which was approved by the and is reported separately. Complete toxicity grading was Institutional Review Board of the hospital where the available on all 31 patients. therapy was administered. Collection and cryopreservation of PBSC Definitions PBSC were collected after the administration of granulo- Patients who relapsed within 6 months of completing plati- cyte colony-stimulating factor (G-CSF) alone (n = 3) or num-based chemotherapy or whose disease progressed on following intermediate doses of chemotherapy and G-CSF platinum-based therapy were classified as platinum-resist- (n = 28). Intermediate-dose chemotherapy consisted of ant. Patients who relapsed longer than 6 months from cyclophosphamide (4 g/m2) (C) (n = 2); C and taxol (170– completion of platinum-based therapy were classified as 250 mg/m2) (CT) (n = 15); C and etoposide (600 mg/m2) platinum-sensitive.11 (CE) (n = 6) or CE and cisplatinum (105 mg/m2) (CEP) (n Patients were staged non-surgically prior to the adminis- = 5). The median days of collection of PBSC were 2 days tration of chemotherapy for mobilization of PBSC. Com- (range 1–10 days). The median number of CD34+ cells/kg plete baseline staging included CT of the abdomen and pel- collected was 14.16 × 106/kg (range 2.85–94.66). Tech- vis, chest X-ray, bone marrow aspirate/biopsy and CA125 niques for collecting PBSC, cryopreserving, thawing and blood test. For purposes of measuring response to HDC, reinfusing PBSC have been previously reported.24,25 Nine- restaging was carried out within 6 weeks of HDC. teen patients proceeded immediately to HDC after collec- Patients with clinical CR had no evidence of disease by tion of PBSC. The other patients received additional physical examination, radiographic evaluation or an elev- chemotherapy (adriamycin/taxol, one cycle and VP16, one ated CA125 blood test. Fifteen patients were in clinical CR cycle (n = 1), cytoxan/VP16, one cycle (n = 2), entering HDC and consequently were not evaluated for cytoxan/carboplatinum (CBDCA)/VP16, two cycles (n = initial response to therapy but were evaluated for survival 1), taxol, two cycles (n = 2), cytoxan/VP16/cisplatinum and time to relapse. Sixteen patients had clinical evidence (CDDP), one cycle (n = 3), cytoxan/taxol, two cycles (n = of disease at HDC and these patients were available for 2), taxol/carboplatinum, two cycles (n = 1)) before HDC. evaluation for initial response to the HDC. Bulky disease у at HDC was defined as 1 cm of measurable tumor at HDC treatment regimen and PBSC infusion transplant. Duration of response was calculated from day of PBSC Patients were treated with phenytoin beginning 24 h prior infusion (day 0). Responses for patients with detectable and to the first dose of BU and continuing until 24 h after the measurable disease were defined as CR if there was a com- last dose. All patients received BU at 1 mg/kg/dose orally plete disappearance of all tumor and normalization of every 6 h for 12 doses (total dose of 12 mg/kg) on days Ϫ8, CA125 for a minimum of 30 days. Partial remission (PR) Ϫ7 and Ϫ6. MEL at 50 mg/m2 was given intravenously on was defined as a reduction of 50% or more in all measur- days Ϫ5 and Ϫ4, for a total dose of 100 mg/m2.TTata able disease including CA125 levels for at least 30 days. dose of 250 mg/m2 was administered intravenously on days Responses of less than 50% or progression of disease were Ϫ3 and Ϫ2, for a total dose of 500 mg/m2. PBSC were considered as no response (NR). All patients were restaged thawed and infused 36 h after the last dose of TT on day 0. at sites of prior disease at 60–85 days and yearly after HDC. Neutrophil engraftment was defined as the first day on Supportive care which the absolute neutrophil count (ANC) exceeded 0.5 × 109/l following the nadir (for 3 consecutive days). Following the administration of the HDC treatment regi- Platelet transfusion independence was defined as the first men, patients received prophylactic antibiotics when their day on which the platelet count exceeded 20 × 109/l without absolute neutrophil count (ANC) was Ͻ0.5 × 109 cells/l. transfusion for 7 consecutive days. Patients who were positive by serology for herpes simplex virus prior to HDC received prophylactic acyclovoir. All patients received anti-fungal prophylaxis with fluconazole Regimen-related toxicity grading from day Ϫ8 to day +75 post-PBSC infusion. PCP prophy- Regimen-related toxicities (RRT) were graded to day 100 laxis was administered through conditioning, discontinued following PBSC infusion using a previously described 48 h before PBSC infusion and resumed after evidence of grading system.20,21 Toxicities of the oral cavity, skin and engraftment to day +180 following PBSC infusion. Eight renal, hepatic, cardiac, pulmonary, gastrointestinal, uro- patients also received G-CSF at 5 ␮g/kg/day beginning logical and neurological systems were assessed. Veno- after the infusion of PBSC and continuing until engraftment occlusive disease of the liver (VOD) was defined clinically of neutrophils. Patients received irradiated blood products by hepatomegaly, liver tenderness, weight gain Ͼ2% of from the appropriate CMV status donors according to the baseline and an elevated serum bilirubin Ͼ2 mg/dl, and guidelines of their respective institution. Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 653 Statistics Table 1 Patient characteristics (n = 31)

Actuarial probabilities of OS and EFS were computed Number (%) according to the method of Kaplan and Meier.26 Survival was calculated from the time of PBSC infusion until death At initial diagnosis or date of last contact. The end-points for EFS were death Stage I 1 (3.2) or relapse. Relapse (R) was estimated by the method of Stage II 3 (9.7) cumulative incidence.27 Patients not achieving CR or PR Stage III/IIIc 21 (67.7) Stage IV 6 (19.4) were considered to have relapsed on day 1 following PBSC infusion. Patients who achieved a PR after HDC were con- Pathology at diagnosis Serous adenocarcinoma 19 (61.3) sidered to have relapsed at the first documentation of pro- Clear Cell 1 (3.2) gression of disease. Patients who died within 100 days after Adenocarcinoma NOS 4 (12.9) PBSC infusion were categorized as dying from transplant- Endometroid 4 (12.9) related toxicities. Patients who died after day 100 with evi- Mixed epithelial 1 (3.2) dence of recurrent or persistent disease were categorized as Undifferentiated carcinoma 2 (6.5) dying from relapse, irrespective of actual cause of death. Initial surgery The date of last contact with surviving patients was 8 optimal 17 (54.8) August 1997. The median follow-up of all patients who suboptimal 14 (45.2) survived beyond day 100 following PBSC infusion was Response to initial induction therapy 564 days. Induction failure 4 (12.9) PR 14 (45.2) CR 13 (41.9) Second surgical look 22 (70.9) Results Pathological CR 8 Microscopic residual disease 4 Macroscopic residual disease 6 Patient characteristics (Table 1) Debulking complete or optimal 4 Of the 31 patients, one patient had stage I disease at diag- Prior cycles of chemotherapy Ͻ10 14 (45.2) nosis, three had stage II, 21 had stage III/IIIc and six had 10–20 15 (48.4) stage IV. The predominant pathology of these patients at Ͼ20 2 (6.5) diagnosis was serous adenocarcinoma. The median number Prior chemotherapy regimens of cycles of conventional chemotherapy these patients 1 2 (6.5) received prior to mobilization and collection of PBSC was 2 8 (25.8) 11 (range 6–36 cycles). The median number of different у3 21 (67.7) regimens prior to mobilization was three (range 1–6). One Intraperitoneal chemotherapy or abdominal radiation patient also received abdominal radiation and nine patients Yes 10 (32.3) received intraperitoneal chemotherapy. The median time Surgical debulking immediately before HDC 6 (19.4) from initial diagnosis to HDC was 640 days (range 237– Median age at HDC, years 47 2205 days) and the median age at HDC was 47 years (range (range) (39–63) 39–63 years). At HDC, 18 patients had stage III/IIIc and 13 had stage IV disease. Median time from diagnosis to HDC, days 640 (range) (237–2205) Of the 15 patients with no clinical evidence of disease at HDC; seven were in first clinical CR, seven in second Highest stage prior to HDC Stage III/IIIc 18 (58.1) clinical CR and one in third clinical CR. The treatment Stage IV 13 (41.9) histories of the seven patients in first clinical CR are out- = lined in Table 2. Three of the seven patients received intra- No clinical evidence of disease at HDC: n 15 First CR 7 (22.6) peritoneal therapy. The number of different regimens these Second CR 7 (22.6) first clinical CR patients received was from two to four, Third CR 1 (3.2) including mobilization chemotherapy and the number of Persistent disease at HDC (n = 16) cycles of chemotherapy was from six to 12. Only one Primary 4 (12.9) patient (No. 6) had documented pathological complete First relapse 11 (35.5) remission at second look after intraperitoneal therapy. One Third relapse 1 (3.2) patient (No. 2) was treated with five cycles of conventional Platinum-sensitive disease 13 (41.9) taxol/CBDCA and the mobilization chemotherapy regimen. Platinum-resistant disease 18 (58.1) Eleven patients out of the 15 had platinum-sensitive disease at the time of HDC. Of the 16 patients with persistent dis- Bulky disease at HDC 9 (29.0) ease at HDC: four had primary persistent disease, 11 were in first relapse and one was in third relapse. Only two had platinum-sensitive disease. Nine patients had bulky disease at HDC. One patient with bulky disease died of RRT. Four patients had no response to HDC. Three patients entered clinical CR after Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 654 Table 2 Chronological treatment history of first clinical CR patients

Patient No. Treatment prior HDC (BUMELTT) Stage at TX

1 Suboptimal surgical debulking IV CDDP/taxol (4 cycles) Cytoxan/taxol (2 cycles) CBDCA/cytoxan/VP16 (2 cycles) 2 Suboptimal surgical debulking III Taxol/CBDCA (3 cycles) Cytoxan/taxol (2 cycle) Taxol (2 cycles) 3 Suboptimal surgical debulking IIIc Taxol/CDDP (6 cycles) 2nd surgical look (macroscopic disease) VP16/CBDCA (1 cycle) Cytoxan/taxol (1 cycle) VP16/CBDCA (2 cycles) 4 Optimal surgical debulking IIIc CDDP/VP16 (6 cycles) 2nd surgical look ( microscopic disease) Ip CDDP/i.v. adriamycin (2 cycles) 5 Suboptimal surgical debulking IIIc CDDP/VP16 (6 cycles) 2nd surgical look (suboptimal tumor debulking) Taxol (5 cycles) Cytoxan/taxol (1 cycle) 6 Optimal tumor debulking IIIc CBDCA(2 cycles) i.v. taxol/Ip CDDp (6 cycles) 2nd surgical look (pathological complete remission) CE (1 cycle) 7 Suboptimal surgical debulking IIIc CBDCA (2 cycles) Ip CDDP/i.v. taxol (6 cycles) 2nd surgical look (bulky disease/optimal tumor debulking) Cytoxan/VP16 (1 cycle)

HDC but relapsed at day 107, 160 and 220. One patient Regimen-related toxicities (RRT) and transplant-related had a PR to HDC and her disease progressed on day 429. mortality

Grade 1–2 toxicities were common. All patients developed Compliance with HDC grade 2 stomatitis, 6.5% developed grade 1–2 cardiac toxicity, 12.9% grade 1–2 pulmonary toxicity, 77.4% grade 1– All patients received the prescribed doses of the HDC 2 colitis, 45.2% grade 1–2 skin toxicity, 6.5% grade 1–2 regimen, BUMELTT. neurological toxicity, 6.5% grade 1–2 bladder toxicity, 3.2% grade 1–2 renal insufficiency and 29.0% grade 1–2 hepatic toxicity. PBSC infusion and engraftment Grade 3–4 toxicities occurred in four patients (12.9%). One patient developed grade 4 veno-occlusive disease The median number of PBSC infused was 9.41 ϫ 106 (VOD), two patients developed grade 3 colitis and one CD34 cells/kg (range 2.85–57.69). All patients engrafted patient developed not only grade 4 pulmonary toxicity but following HDC. An ANC Ͼ0.5 ϫ 109 cells/l was reached also grade 3 stomatitis and grade 3 skin toxicity. Two of at a median of 11 days (range 8–15 days) and an unsup- the patients (6.5% of all patients) died from RRT: one death ported platelet count Ͼ20 ϫ 109/l at a median of 10 days was due to pulmonary complications and occurred in an (range 6–45 days) after the infusion of PBSC. Transfusion individual who had no evidence of disease at the initiation data were available on 29 of the 31 patients. A median of of HDC and the other death was due to VOD and occurred 12 units (ie two transfusions of six units of random pooled in a patient with persistent disease in the liver at the donor) platelets (range 0–64) and a median of four units of initiation of HDC. Both deaths occurred in patients with packed red blood cells (range 0–8) were transfused. platinum-resistant disease. Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 655 In addition, seven patients (22.6%) developed positive a bacterial or fungal blood cultures, but none died of 1.0 infection. 0.8 (n = 15) Response 0.6 Sixteen patients were not evaluable for the initial response 0.4 to HDC due to early death (n = 2) or lack of detectable disease at initiation of the HDC regimen (n = 14). Six out 0.2 of the 15 patients (40.0%) with persistent disease at the Probability of survival initiation of HDC had no response to the conditioning regi- 0 men. Five patients (33.3%) achieved a CR, of whom two 01234 had platinum-sensitive disease. Four out of 15 patients had Years after transplant a PR (26.7%). The total CR and PR response to BUMELTT b in evaluated patients was 60%. 1.0 The actuarial probabilities of OS, EFS and R for all patients at 18 months were 0.57, 0.30 and 0.63 (Figure 1a 0.8 and b). Fifteen patients were treated with HDC at a time of clini- 0.6 cal complete remission. The actuarial probabilities of OS, (n = 15) EFS and R for these patients at 18 months were 0.87, 0.44 0.4 and 0.49 (Figure 2a and b). One patient died of pulmonary complications on day 34. Seven patients remain alive and 0.2 in remission. Five patients relapsed at days 129, 133, 155, 348 and 614 after HDC and are still alive. Two patients 0 01234 who relapsed at 160 and 379 days succumbed to their Probability of event-free survival disease on days 171 and 875, respectively. Of the seven Years after transplant patients who were in first clinical CR at the time of HDC, Figure 2 The OS (a) and EFS (b) rates of ovarian cancer patients who were in clinical complete remission when treated with BUMELTT. a 1.0 five remain alive and in remission at 316–1135 days. Of the seven patients in 2nd clinical CR, only one patient 0.8 remains alive and in remission. The one patient treated with HDC in 3rd clinical CR remains alive and in remission. 0.6 Sixteen patients with persistent disease were treated with (n = 31) BUMELTT. The actuarial probabilities of OS, EFS and R 0.4 in these patients at 18 months were 0.38, 0.13 and 0.81 (Figure 3a and b). One patient died of RRT secondary to 0.2 VOD on day 56. The other patients died from relapse on Probability of survival days 139, 250, 320, 359, 418, 429, 422 and 520 days fol- 0 lowing PBSC infusion. Only one patient with platinum- 01234resistant disease and in first relapse at HDC remains alive Years after transplant and in remission at 579 days. This patient alone received b taxol/adriamycin and high-dose etoposide therapy prior to 1.0 HDC. Of the 13 patients with platinum-sensitive disease, none 0.8 died of RRT. All patients remain alive and seven patients have relapsed at 129–1021 days. Four of the six patients

0.6 that remain alive without evidence of disease were transplanted in 1st clinical CR. The OS, EFS and R rates for these patients at 18 months were 1.00, 0.52 and 0.48 0.4 (Figure 4a and b). (n = 31) Of the 18 patients with platinum-resistant disease, two 0.2 died of RRT. Six patients did not respond to the HDC. Eight additional patients have relapsed between days 104– 0 429. Four patients remain alive and two of these four Probability of event-free survival 01234(transplanted in 1st relapse and 2nd CR) are still in Years after transplant remission at 579 and 736 days after HDC. The OS, EFS Figure 1 The overall survival (OS) (a) and event-free survival (EFS) (b) and R rates at 18 months were 0.33, 0.11 and 0.78 (Figure rates of all stage III/IV ovarian cancer patients treated with BUMELTT. 5a and b). Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 656 a a 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4 (n = 18) (n = 16) 0.2 0.2 Probability of survival Probability of survival

0 0 01230123 b Years after transplant b Years after transplant 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 (n = 18) 0.2 (n = 16)

0 0 Probability of event-free survival 0123

Probability of event-free survival 0123 Years after transplant Years after transplant Figure 5 The OS (a) and EFS (b) rates of ovarian cancer patients with Figure 3 The OS (a) and EFS (b) rates of ovarian cancer patients who platinum-resistant disease and treated with BUMELTT. had persistent disease when treated with BUMELTT.

Discussion

a A regimen of HDC with BUMELTT followed by autolog- 1.0 ous PBSC rescue was administered to 31 patients with (n = 13) advanced stage III/IV ovarian cancer. Death due to compli- 0.8 cations of the HDC occurred in 6.5% of the patients. Not surprisingly the overall and event-free survivals were 0.6 superior for patients transplanted in clinical CR who had mainly platinum-sensitive disease. Although these results 0.4 are encouraging, follow-up is still relatively short. The ulti- mate number of patients who will be long-term disease-free 0.2

Probability of survival survivors remains to be determined. Overall, the results of this study with BUMELTT in 0 patients transplanted with relapsed or persistent disease are 01234similar to the results that have been published for other b Years after transplant regimens used to treat relapsed or persistent advanced 1.0 ovarian cancer.5–14 A number of different HDC regimens including BUMELTT can lead initially to high overall CR 0.8 and PR rates in patients with detectable ovarian cancer. Sixty percent of the patients initially evaluated responded 0.6 to the BUMELTT regimen. With other regimens, the initial 5–14 (n = 13) response rate has been reported to be 55–75%. How- 0.4 ever, relapse remains a major problem in all of the studies. Long-term disease-free survival is seen in only 5–25% of 0.2 all patients treated. In the BUMELTT regimen, the EFS rate for all patients was 30%, for patients with persistent 0 disease 13% and for patients with platinum-resistant disease 01234 Probability of event-free survival 11%. In 1992, an update was published of the transplant Years after transplant experience for ovarian cancer at 11 different centers.10 Figure 4 The OS (a) and EFS (b) rates of ovarian cancer patients with Patients with persistent or relapsed ovarian cancer were platinum-sensitive disease and treated with BUMELTT. offered a single high-dose regimen. Sixty-one patients had Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 657 platinum-resistant disease. Overall, 85% of the patients ach- achieved a PR. Three patients (23%) remained without ieved a clinical response, with 34% achieving a complete evidence of disease at 108–443 days. response (CR). Of the 37 patients with platinum-sensitive Viens et al30 reported 35 patients with ovarian cancer disease, 87% achieved a clinical response, with 73% a CR. treated with bone marrow transplantation as part of their However, for the whole group, the median event-free sur- initial therapy. After initial surgery, 30 patients with stage vival lasted only 6 months, with a 14% disease-free sur- III and five with stage IV received a median of six cycles of vival rate at 1 year. Recently, Stiff et al14 have published a standard cisplatinum-based regimen. Following primary the outcome of 100 consecutive patients transplanted for therapy, three had progressive disease. The rest underwent persistent or relapsed ovarian cancer using carboplatinum, second-look laparotomy to document tumor response to mitoxantrone and cyclophosphamide with or without cyclo- initial therapy. Nine patients were in CR, 10 had Ͻ2cm sporine (n = 70), melphalan and mitoxantrone with or with- residual disease, and eight had Ͼ2 cm residual disease. out taxol (n = 25) or other regimens (n = 5). Overall, 66% Patients then received MEL (140–240 mg/m2) followed by of the patients had platinum-resistant disease, 61% had autologous bone marrow transplantation. The overall sur- у1 cm tumor burden and 70% had received two or more vival was 47% at 54 months. The initial response rate was chemotherapy regimens prior to HDC. The median pro- 75% in the 12 patients evaluated. Of the nine individuals gression-free survival and OS times were 7 and 13 months who at second-look laparotomy were in CR, six remain for all patients and 19 and 30 months, respectively, in plati- NED at 9+ to 32+ months. Of the 10 individuals with Ͻ2 num-sensitive patients with р1 cm of disease at HDC. A cm residual disease, six are alive and NED at 13+ to 54+ stepwise Cox proportional hazards model identified tumor months. bulk and cisplatin sensitivity as the best predictors of Recently, Legros et al28 published the outcome of 53 progression-free survival. stage III or IV ovarian cancer patients treated with surgery As of 6 January 1995, the American Blood and Marrow followed by cisplatin combination chemotherapy. All Transplant Registry had 249 patients with ovarian cancer underwent second-look surgery. HDC was then adminis- treated at 51 centers. Seventeen percent of the patients were tered. Twenty-three patients received melphalan alone, and treated with HDC in remission and 83% still had evidence 30 received carboplatinum/cyclosphamide. After HDC, of persistent disease. The 2-year probability of survival was autologous PBSC were infused. One patient died of RRT. 33% for the patients with persistent disease at HDC and With a median follow-up of 81.5 months, the 5-year overall 68% for the patients transplanted in clinical CR.28 These survival rate of all of these 53 treated patients was 59.9% results are similar to our results where the probabilities of and the disease-free survival rate was 23.6%. In the 19 survival at 18 months for patients in clinical CR when patients with pathological CR at second-look surgery, the treated with BUMELTT was 0.87 and for patients not in 5-year survival rate was 74.2% and the DFS rate was CR 0.38 (Figures 2 and 3). 32.8%. Overall, the 5-year survival of these 53 patients It is clear that a single course of high-dose chemotherapy treated with HDC appears superior to the 15–25% survival alone as used in a number of different regimens will not that is reported with conventional therapy.31,32 Of interest, cure the majority of patients with platinum-resistant or per- HDC appeared to improve survival by delaying relapse sistent advanced disease. It is possible though, that those without inducing chemoresistance to second-line cisplatin patients with platinum-sensitive disease and clinical CR therapy. The 38 patients who relapsed after HDC had a prior to HDC may still benefit from being treated with an good response to cisplatin-based therapy with an overall HDC regimen as their overall survival may exceed that seen survival rate after relapse of 17.3 months. with conventional chemotherapy alone. In this current report, we did not treat patients who ful- Therefore, new approaches are needed to treat patients filled the above definition of first clinical complete with advanced ovarian cancer. One approach is to offer remission after six cycles of platinum combination therapy. HDC to patients in first clinical CR after surgical debulking Six of the seven patients that were in first clinical CR and completion of a maximum of six cycles of before BUMELTT took 7–12 cycles of chemotherapy and platinum/taxol standard chemotherapy. HDC may be more two to four different regimens to achieve a CR (Table 2). effective if utilized earlier in the course of ovarian cancer The trials reported by Menichella et al,29 Viens et al30 before the emergence of chemotherapy-resistant disease and Legros et al28 suggest that HDC can be successfully and at a time of minimal tumor burden. This is suggested incorporated into primary therapeutic strategies for treating by the outcome of the platinum-sensitive patients in our ovarian cancer. Consequently, large prospective ran- study who have an EFS rate of 0.52 at 18 months and by domized trials of HDC as consolidation for primary induc- the fact that five out of seven patients who were in 1st tion regimens are needed to ascertain whether they will clinical CR at the time of HDC remain alive and in improve the poor 5-year survival rates for patients with remission. stage III and IV ovarian cancer. To date, there are few studies of HDC followed by stem The effectiveness of treating advanced ovarian cancer cell rescue as front-line therapy for ovarian cancer. Menich- with HDC remains undefined. Patients with platinum-sensi- ella et al29 treated 13 patients who had Ͼ0.5 cm of tumor tive disease appear most likely to benefit initially from a after staging laparotomy. These patients initially received HDC regimen such as BUMELTT. However, the duration two courses of standard-dose cisplatinum and cyclophos- of this benefit remains to be determined. Future studies are phamide with PBSC collection. They then received a plati- now needed to direct HDC toward stage III/IV ovarian can- num-based HDC regimen. After second-look laparotomy, cer patients in first clinical remission following standard there were four patients who achieved a CR and five who platinum-based regimens to ascertain whether any impact Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 658 can be achieved with this approach on the long-term dis- 10 Stiff P, McKenzie RS, Alberts DS et al. Bone marrow trans- ease-free survival of these patients. New and novel HDC plantation for ovarian carcinoma in the United States: a survey regimens are needed to treat patients with persistent and of active programs. In: Dicke KA, Keating A (eds). Proc Sixth platinum-resistant disease. From a number of published International ABMT Symposium. Cancer Treatment Research studies including the above study using BUMELTT, it is Education Fund: Arlington, 1993, pp 192–225. 11 Stiff P, Bayer R, Camarda M et al. A phase II trial of high obvious that a single HDC regimen fails to cure the dose mitoxantrone, carboplatin and cyclophosphamide with majority of these patients. As a consequence, we are offer- autologous bone marrow rescue for recurrent epithelial ing these patients immunotherapy with IL-2 incubated ovarian carcinoma: an analysis of risk factor for clinical out- PBSC and escalating doses of sequential IL-2 after com- come. Gynecol Oncol 1995; 57: 278–285. pletion of the BUMELTT regimen in a phase I clinical trial. 12 Shpall EJ, Clarke-Pearson D, Soper JT et al. High-dose We have also begun evaluating the response of patients alkylating agents chemotherapy with autologous bone marrow with persistent disease to two tandem transplants using support in patients with stage III/IV epithelial ovarian cancer. mitoxantrone/TT followed by PBSC rescue, then MEL fol- Gynecol Oncol 1990; 38: 386–391. lowed by PBSC rescue and finally 12 weeks of IL-2 immu- 13 Benedetti-Panici P, Greggi S, Scambia G et al. High-dose notherapy. Whether the addition of immunotherapy or tan- chemotherapy with autologous peripheral stem cell support in advanced ovarian cancer. Ann Med 1995; 27: 133–138. dem cycles of HDC will impact favorably on the outcome 14 Stiff PJ, Bayer R, Kerger C et al. High-dose chemotherapy of these latter patients remains to be seen. A valid argument with autologous transplantation for persistent/relapsed ovarian can still be made for pursuing treatment with a simple sin- cancer: a multivariate analysis of survival for 100 consecu- gle HDC regimen only in patients with platinum-sensitive tively treated patients. J Clin Oncol 1997; 15: 1309–1317. disease. 15 Corringham R, Gilmore M, Prentice H et al. High-dose melphalan with autologous bone marrow transplant: treatment of poor prognosis tumor. Cancer 1993; 52: 1783–1787. Acknowledgements 16 Marinchi D, Piana L, Blauner D et al. Phase I–II studies of high-dose alkylating agents in poor risk patients. In: Dicke K, Spitzer T, Jaggenth S (eds). ABMT: Proc Third International This work was supported by grant CA 47748 from the National Symposium. University of Texas MD Anderson Hospital: Cancer Institute and 255902 from NIH. We express our gratitude Houston, 1987, pp 475–480. to all the members and staff of the Puget Sound Oncology Consor- 17 Knight WA, Page CP, Kuhn JG et al. High-dose L-PAM and tium who enrolled their patients on study and provided us with autologous marrow infusion for refractory solid tumors. Proc follow-up. Am Soc Clin Oncol 1994; 3: 150 (Abstr.). 18 Lazarus HM, Reed MD, Spitzer TR et al. High-dose intra- venous thiotepa and cryopreserved autologous bone marrow References transplant for therapy of refractory cancer. Cancer Treat Rep 1987; 71: 689–695. 1 Parker SL, Tong T, Bolden S, Wingo PA. Cancer Statistics, 19 Demirer T, Buckner CD, Appelbaum FR et al. High-dose bus- 1997. Ca Cancer J Clin 1997; 47: 5–27. ulfan and cyclophosphamide followed by autologous trans- 2 Wingo PA, Tong T, Bolder S. Cancer Statistics, 1995. Ca plantation in patients with advanced breast cancer. Bone Mar- Cancer J Clin 1995; 45: 8–20. row Transplant 1996; 17: 769–774. 3 Thigpen JT, Blessing JA, Vance RB, Lambuth BW. Chemo- 20 Weaver CH, Bensinger WI, Appelbaum F et al. Phase I study therapy in ovarian carcinoma: present role and future pros- of high-dose busulfan, melphalan, thiotepa with autologous pects. Semin Oncol 1989; 16: 58–65. stem cell support in patients with refractory malignancies. 4 Quazi F, McGuire WP. The treatment of epithelial ovarian Bone Marrow Ttransplant 1994; 14: 813–819. cancer. Cancer J Clin 1995; 45: 88–101. 21 Schiffman KS, Bensinger WI, Appelbaum FR et al. Phase II 5 Lotz JP, Machover D, Malassagne B et al. Tandem high-dose study of high-dose busulfan, melphalan and thiotepa with (HD) chemotherapy with VM-26, ifosfamide (IFM) and car- autologous peripheral blood stem cell support inpatients with boplatinum (CBDCA) with autologous bone marrow trans- malignant disease. Bone Marrow Transplant 1996; 17: 943– plantation (ABMT) for patients with stage IIIc–IV ovarian 950. cancer. Proc Am Soc Clin Oncol 1993; 12: 257 (Abstr.). 22 Shuhart MC, McDonald GB. Gastrointestinal and hepatic 6 Mulder PO, Willemse B, Aalders JG et al. High-dose chemo- complications, in Forman SJ, Blume KG, Thomas ED (eds). therapy with autologous bone marrow transplantation in Bone Marrow Transplantation. Blackwell Scientific Publi- patients with refractory ovarian cancer. Eur J Cancer Clin cations: Boston, MA, 1994, pp 454–481. Oncol 1989; 25: 645–649. 23 Crawford SW. Critical care and respiratory failure. In: Thomas 7 Dauplat J, Legras M, Condat P et al. High-dose melphalan and ED, Forman SJ, Blume K (eds). Bone Marrow Transplan- autologous bone marrow support for the treatment of ovarian tation. Blackwell Scientific Publications: Boston, MA, 1994, carcinoma with positive second-look operation. Gynecol pp 513–526. Oncol 1989; 34: 294–298. 24 Bensinger WI, Singer J, Appelbaum F et al. Autologous trans- 8 Mulder PO, Sleijfer DT, Willemese PH et al. High-dose cyclo- plantation with peripheral blood mononuclear cells collected phosphamide or melphalan with escalating doses of mitoxan- after administration of recombinant granulocyte stimulating trone and autologous bone marrow transplant for refractory factor. Blood 1993; 81: 3158–3163. solid tumors. Cancer Res 1989; 49: 4654–4658. 25 Rowley SD, Bensinger WI, Gooley TA et al. The effect of 9 Viens P, Maraninchi D. High-dose chemotherapy and autolog- cell concentration on bone marrow and peripheral blood stem ous marrow transplantation for common epithelial ovarian cells cryopreservation. Blood 1994; 83: 2731–2736. cancer. In: Armitage JO, Antman KH (eds). High-Dose Can- 26 Kaplan EL, Meier P. Nonparametric estimation from incom- cer Therapy. Williams and Walkins: Baltimore, 1992, pp plete observations. J Am Stat Assoc 1958; 53: 457–488. 729–734. 27 Pepe MS, Longton G, Pettinger M et al. Summarizing data Busulfan, melphalan, thiotepa for ovarian cancer LA Holmberg et al 659 on survival, relapse and chronic graft-versus-host disease after phalan and autologous marrow rescue in advanced epithelial bone marrow transplantation: motivation for and description carcinoma: a retrospective analysis of 35 patients with of new methods. Br J Haematol 1993; 83: 602–607. advanced ovarian adenocarcinoma treated in France. Bone 28 Legros M, Dauplat J, Fleury J et al. High-dose chemotherapy Marrow Transplant 1990; 5: 227–233. with hematopoietic rescue in patients with stage III to IV 31 McGuire WP, Hoskins WJ, Brady MF et al. Cyclophospham- ovarian cancer: long-term results. J Clin Oncol 1997; 15: ide and cisplatin vs paclitaxel and cisplatin in patients with 1302–1308. stage III and stage IV ovarian cancer. New Engl J Med 1996; 29 Menichella G, Pierelli L, Foddai ML et al. Autologous blood 334: 1–6. stem cell harvest and transplantation in patients with advanced 32 Ozols RF. Treatment of ovarian cancer: current status. Semin ovarian cancer. Br J Haematol 1991; 79: 444–450. Oncol 1994; 21: 1–9. 30 Viens P, Maraninchi D, Giacchetti S et al. High-dose mel-

Appendix

Collaborating community institutions ACOS designation: (1) Comprehensive Cancer Center St Lukes Regional Medical Center, Boise, Idaho (1) (2) Teaching Center Hospital Cancer Program Swedish Hospital Medical Center, Seattle, WA (2) (3) Comprehensive Community Cancer Center Virginia Mason, Seattle, WA (1) (4) Community Cancer Center St Joseph Hospital, Tacoma, WA (1) Community Medical Center, Missoula, MT (4) Good Samaritan Hospital, Puyallup, WA (4) Sacred Heart Hospital, Spokane, WA (1)