Intensive sequential chemotherapy with repeated blood stem-cell support for untreated poor-prognosis non-Hodgkin's lymphoma.

PURPOSE: To demonstrate the feasibility and efficacy of six ambulatory high-dose sequential chemotherapy courses that include three intensified cycles supported by stem-cell infusion in high-risk and high-intermediate-risk untreated non-Hodgkin's lymphoma (NHL) patients. PATIENTS AND METHODS: A pilot nonrandomized study included 20 untreated patients aged less than 60 years with aggressive histologically identified NHL and two or three adverse-prognosis criteria (International Index). Patients received an ambulatory regimen with high-dose chemotherapy supported by granulocyte colony-stimulating factor (G-CSF) and repeated peripheral-blood stem-cell (PBSC) infusion. The median age was 39 years (range, 20 to 59), with 13 men and seven women. Chemotherapy consisted of one cycle every 21 days for a total of six cycles. The first three cycles (A1, A2, and A3) consisted of cyclophosphamide (Cy) 3,000 mg/m2, doxorubicin (Doxo) 75 mg/m2, and vincristine 2 mg (plus corticosteroids). The last three cycles (B4, B5, and B6) consisted of the same drug combination plus etoposide 300 mg/m2 and cisplatin 100 mg/m2. For an expected duration of 18 weeks, the projected dose-intensity was 25 mg/m2/wk for Doxo and 1,000 mg/m2/wk for Cy. G-CSF 300 micrograms was administered from day 6 following each cycle until neutrophil reconstitution. Two aphereses were performed at approximately day 13 after each A cycle, and PBSCs were injected at day 4 of each B cycle. Radiotherapy on tumor masses > or = 5 cm was scheduled after completion of the last cycle. RESULTS: The median duration of grade 4 neutropenia was 1 day (range, 0 to 7) for each A cycle and 4 days (range, 1 to 10) for each B cycle (P = .02). The median duration of grade 4 thrombopenia was 0 days (range, 0 to 8) for each A cycle and 6 days (range, 1 to 21) for each B cycle (P < .001). Hospitalization for febrile neutropenia was required for 18% and 44% of patients during cycles A and B, respectively (P < .01). Only three patients did not complete the protocol: one due to emergency surgery after cycle B4, one who died after cycle B5 from interstitial pneumonia, and one with delayed hematologic reconstitution after cycle B4. Chemotherapy delivery was optimal (median actual relative dose-intensity, 97%; range, 66 to 100). The median total dose administered over 18 weeks was 18,000 mg Cy (range, 12,000 to 18,000), 450 mg Doxo (range, 300 to 450), 900 mg etoposide (range, 300 to 900), and 300 mg cisplatin (range, 100 to 300). Evaluation of response after six courses showed 13 complete remissions ([CRs] 65%), four partial remissions (PRs), two nonresponses (NRs), and one toxic death. With a median follow-up period of 25 months (range, 16 to 43), 15 patients are alive, with 12 in continuous first CR; five patients relapsed (four of four PRs and one of 13 CRs). Two-year survival and failure-free survival (FFS) rates are 73% and 56%, respectively. The disease-free survival (DFS) rate for the CRs is 86%. CONCLUSION: PBSC support contributes to the feasibility of first-line, very-high-dose, ambulatory chemotherapy delivery in poor-risk NHL and is associated with a high rate of remission and FFS.

Early administration of recombinant erythropoietin improves hemoglobin recovery after reduced intensity conditioned allogeneic stem cell transplantation.

The use of recombinant human erythropoietin (rHuEPO) has been controversial after myeloablative allogeneic Stem cell transplantation (allo-SCT). Reduced intensity conditioning regimens (RIC) offer a novel approach that might translate into a different profile of erythropoietic recovery. We treated 20 consecutive patients with rHuEPO early after matched sibling RIC allo-SCT. Conditioning included fludarabine, busulfan and antithymocyte globulin. EPO treatment was analyzed in terms of toxicity, impact on the frequency of Red blood cell transfusions (RBCT) and kinetics of Hemoglobin recovery within the 60 days post-allo-SCT. Results were compared with 27 matched patients who did not receive rHuEPO. In the first 2 months after allo-SCT all patients receiving rHuEPO (100%) achieved an Hb level > 11 g/dl at a median of 30 (15-35) days post-allo-SCT, as compared to only 63% of the patients not receiving rHuEPO (P = 0.007) at a median of 35 (20-55) days (P = 0.03). A total of 70% (95% CI, 50-90) of rHuEPO patients maintained an Hb over 11 g/dl in the second month as compared to only 19% (95% CI, 4-34) in the other group (P = 0.0004). For patients receiving RBCT, the use of rHuEPO was associated with a trend towards reduced RBCT requirements. This pilot study suggests a potential benefit of early administration of rHuEPO after RIC allo-SCT on early erythropoietic recovery.

Production of ex vivo expanded hematopoietic cells and progenitors in a closed bioreactor, starting with a small volume marrow collection: A feasibility study in patients with poor-risk breast cancer and receiving high-doses of cyclophosphamide.

We report a clinical pilot study conducted in 6 women with poor-prognosis breast cancer. The goal was to evaluate the feasibility and safety of producing hematopoietic progenitors and cells from a small marrow sample, for clinical use after high-dose cyclophosphamide. A small volume marrow collection was obtained, using local anesthesia and conscious sedation, before the first of two chemotherapy cycles. Cells were cryopreserved, and later thawed to inoculate two Aastrom Biosciences Inc Replicell bioreactors, on time to reinfuse ex vivo expanded cells after the second chemotherapy cycle. Patients recovered neutrophils and platelets at similar times after the first and second chemotherapy cycles, and showed comparable clinical events. This pilot study prepares future randomized trials, designed to evaluate clinical benefits associated with the use of ex vivo expanded cells in the setting of multicycle high-dose chemotherapy.

Expression of the p53 gene in Hodgkin's disease: dissociation between immunohistochemistry and clinicopathological data.

Expression of the p53 protein has been detected recently by immunohistochemistry in Hodgkin's disease (HD), but the relationship between p53 expression and the prognosis and clinicopathological heterogeneity of HD is still unclear. To address these questions we investigated 49 cases of HD for p53 expression by immunohistochemistry using the DO1 monoclonal antibody (MAb) on paraffin sections. Thirty-five cases were simultaneously tested with the 1801 MAb on frozen sections. Thirty-seven of 49 cases (75%) were DO1 positive while 14 of 35 (40%) were PAb 1801 positive. Both MAbs gave a nuclear staining restricted to Reed Sternberg cells (RSCs) and variants and distributed among the three HD subtypes analyzed (ie, nodular lymphocyte predominant, nodular sclerosing, and mixed cellularity). The percentage of positive neoplastic cells in each case was heterogeneous, ranging from almost 100% to less than 5%. In 39 patients for whom clinical data were available statistical analysis did not show any significant correlation between p53-positive immunostaining and clinical staging, B symptoms, probability of relapse, or disease-free survival. We conclude that p53 expression is a common event in HD regardless of histological subtyping, but does not bear any pejorative significance.

Delayed administration of granulocyte colony-stimulating factor after autologous bone marrow transplantation: effect on granulocyte recovery.

Recombinant granulocyte colony-stimulating factor (rhG-CSF) has been shown to hasten granulocyte recovery after autologous BMT. In current protocols, rhG-CSF treatment starts 1 day after BM reinfusion. Our study retrospectively examined the effects on haematological recovery of a day 6 delayed administration. Seventy-eight patients receiving autologous BMT for malignant lymphoma (21 non-Hodgkin's lymphoma and 9 Hodgkin's disease) or solid tumors (33 breast carcinoma and 5 ovarian carcinoma) were split up into three study groups. Two groups receiving a 5 micrograms/kg/day of rhG-CSF starting either 1 day (day +1 group, n = 25 patients) or 6 days (day +6 group, n = 24 patients) after BM reinfusion were compared with 29 historical control patients. Granulocyte recovery to 0.5 x 10(9)/l was 12 days in day +6 and day +1 groups versus 16 days in control group (p < 0.005) without any difference in other hematological parameters, infectious complications or length of hospitalisation between the three groups. The day +6 administration allows elimination of a median of 7 days rhG-CSF. It has been concluded that the day +6 administration gives the same clinical benefit as day +1 administration with consequent cost reductions.

Prevention of acute GVHD by in vivo use of anti-interleukin-2 receptor monoclonal antibody (33B3.1): a feasibility trial in 15 patients.

Results of allogeneic bone marrow transplantation (BMT) are still impaired due mainly to acute graft-versus-host disease (GVHD). Successful T cell depletion may abolish GVHD but causes increased rates of rejection and relapse. In vivo blocking of the CD25 receptor on T cells induces efficient and selective immunosuppression. We present results of a pilot trial studying the use of a CD25 MoAb (33B3.1) to prevent acute GVHD after allogeneic BMT. Fifteen patients were included in the study; 14 had a fully HLA matched sibling donor. In association with short methotrexate and cyclosporin A post-graft immunosuppression, the patients received 10 mg of 33B3.1 monoclonal antibody daily according to three successive schedules to study toxicity on marrow graft. No major adverse clinical experiences occurred. Engraftment was not delayed; after 4 months, full chimerism was documented in 14/15 studied patients. No human anti-rat antibody was detected during treatment period. No GVHD occurred during the period of antibody administration. Two patients relapsed; seven are alive and well with a median follow-up of 650 days. This study justifies a prospective controlled study to determine the real impact of 33B3.1 on GVHD prophylaxis.

High-dose sequential chemotherapy with stem cell support for non-metastatic breast cancer.

The importance of dose-intensity has been suggested in breast cancer. The aim of this study was to evaluate the feasibility of a high-dose intensity doxorubicin-cyclophosphamide regimen with supporting G-CSF and blood stem cells. Twenty-five patients with non-metastatic breast cancer received four cycles of doxorubicin (75 mg/m2) and cyclophosphamide (3000 mg/m2) at 3 week intervals. Apheresis was performed after the first cycle and if necessary after the second cycle. Stem cells were reinfused after the third and fourth cycles. G-CSF was started on day 3 of each cycle (5 microg/kg/day) and was stopped the day before the last apheresis or when absolute neutrophil count was above 0.5 x 10(9)/l. Median received dose-intensity was respectively 25 mg/m2/week (range 22-26) and 1000 mg/m2/week (range 904-1065) for doxorubicin and cyclophosphamide. Grade IV thrombocytopenia occurred in 8% of cycles. Two patients needed platelets and 12 red cell transfusion. Fifteen patients were readmitted for a median duration of 4 days (range 1-7). We have established a safe, outpatient, high-dose intensity doxorubicin-cyclophosphamide regimen with supporting G-CSF and blood stem cells which can be submitted for comparison with the current standards.

Overexpression of erb B2 remains a major risk factor in non-metastatic breast cancers treated with high-dose alkylating agents and autologous stem cell transplantation.

The importance of dose intensity has been strongly emphasized in high-risk breast cancer. Overexpression of erb B2 is clearly correlated with an overall poor prognosis which could be limited in patients receiving intensive chemotherapy with alkylating agents and autologous stem cell transplants (SST). Thirty-five patients with high-risk non-metastatic breast cancer (>4 involved lymph nodes), treated with high-dose chemotherapy (HDC) followed by SST were analyzed. All were previously treated by four cycles of standard-dose anthracycline or anthracene dione. Nine had erb B2 overexpression. Minimum follow-up duration was 41 months (median 68 months). At 5 years, the actuarial relapse-free survival is 57.4% and actuarial overall survival 67.4%. Patients with overexpression of erb B2 had significantly lower disease-free survivals (P: 0.021) and overall survivals (P: 0.001). On multivariate analysis, erb B2 overexpression appeared to be the single independent poor prognosis factor for relapse (RR 3.25, range 1.12 to 9.45) and overall (RR 5.28, range 1.74 to 16.03) survival. These results suggest that poor prognosis of erb B2 overexpression is unchanged after HDC with alkylating agents but a possible benefit may exist in these patients with the additional monoclonal antibody, herceptin.

A pilot study of busulfan and melphalan as preparatory regimen prior to allogeneic bone marrow transplantation in refractory or relapsed hematological malignancies.

In this pilot study; we assessed the immunosuppressive and the antileukemic potential of a combination of busulfan and melphalan prior to allogeneic BMT in 25 adult patients with refractory or relapsed hematological malignancies. Twelve patients were transplanted for acute myeloid leukemia (relapse: five patients; primary refractory: four patients; second remission: two patients), two patients for primary refractory acute lymphoblastic leukemia, nine patients for chronic myelogenous leukemia (accelerated phase: six patients; blastic phase: three patients) and two patients for primary refractory lymphoma. All received an unmanipulated marrow from HLA-identical siblings. All patients but one engrafted (median time to ANC > or = 0.5 x 10(9)/l = 17 days, to platelets > or = 50 x 10(9)/l = 29 days). Full chimerism was documented in the seven evaluable patients. The probability for developing acute GVHD was 58%. Complete remission was obtained in 17/18 measurable patients. With a 42 month median follow-up, eight patients are alive in unmaintained remission. The 4-year probabilities for relapse, survival, and DFS are respectively: 42%, 35%, and 31%. These results show that the combination of busulfan and melphalan ensures an effective immunosuppression allowing long-term engraftment. This regimen can provide long-term disease-free survival in patients with high-risk disease and thus represents an interesting alternative to the CY and/or TBI-containing regimens.

Clinical and economic comparison of lenograstim-primed blood cells (BC) and bone marrow (BM) allogeneic transplantation.

The study presented is a clinical and economic comparison of bone marrow (BM) and blood cells (BC) allogeneic transplantation. We performed a case-control study to compare 17 patients receiving allogeneic BC transplant in a pilot study to an historical group of 17 patients allografted with BM. We evaluated the clinical outcomes and the direct medical costs of transplantation from conditioning regimen until day 100 by detailed observation of patients' medical records. Patients in the BC group received a median of 8 x 10(6)/kg CD34+ cells (1.58-29.1) and 266 x 10(6)/kg CD3+ cells (128-469). All patients had neutrophil engraftment with a median of 14 days in the BC group vs 19 days in the BM group (P < 0.05). The Kaplan-Meier estimation of the median number of days to a platelet count of > 25 x 10(9)/l, independent of platelet transfusion, was significantly shorter in the BC group (15 (9-74)) compared with the BM group (25 (15-45)). Acute graft-versus-host disease (AGVHD) of grade > or = 2 was not significantly different between the two groups. Patients treated with BC presented a US$16,134 decrease in the cost of the first 100 days (29%, P = 0.006). Our comparison suggested that platelet reconstitution and total costs were in favor of the BC group.

Modulations of dose intensity of doxorubicin and cyclophosphamide in association with G-CSF and peripheral blood stem cells in adjuvant chemotherapy for breast cancer: comparative evaluation of completion and safety of three intensive regimens.

The aim of this study was to evaluate and to compare in terms of toxicity the modulations of dose intensity of cyclophosphamide and doxorubicin in adjuvant chemotherapy for high-risk breast cancer. Four cycles of sequential high-dose chemotherapy with doxorubicin and cyclophosphamide (AC), supported with G-CSF and peripheral blood stem cells (PBSC) were administered to 81 women. Three successive cohorts were studied: doxorubicin (75 mg/m(2)) + cyclophosphamide (3000 mg/m(2)) every 21 days (group 1), doxorubicin (75 mg/m(2)) + cyclophosphamide (3000 mg/m(2)) every 15 days (group 2), and doxorubicin (75 mg/m(2)) + cyclophosphamide (6000 mg/m(2)) every 21 days (group 3). Seventy-five patients received four cycles of treatment with a total of 310 cycles administered. The received dose intensity of doxorubicin was higher in group 2 and that of cyclophosphamide was lower in group 1 than in the other two groups. Hematological and extra-hematological toxicities, as well as the number and duration of hospitalizations for toxicity, were significantly higher in group 3. We conclude that the group 3 regimen is associated with toxicities comparable to autologous transplantation. Increasing dose intensity of doxorubicin and cyclophosphamide is feasible in an outpatient setting and safe in groups 1 and 2 with the support of hematopoietic factor and PBSC.

Simultaneous occurrence of kaposi's sarcoma and chronic myelogenous leukemia.

We report here a 75-year-old man from South France who developed Kaposi's Sarcoma (KS) 5 months after diagnosis of Philadelphia-chromosome positive chronic myelogenous leukemia (CML). He was found positive for HHV-8 by PCR, negative for both HIV 1 and HIV 2 by serology, and had a normal CD4/CD8 ratio. Favourable evolution of both CML and KS has been obtained with vinblastine and interferon alpha treatment. The patient is currently alive in complete remission of SK and major cytogenetic remission of CML with a 48 month follow-up. Since no immune deficiency could be documented in the patient, this rare observation suggests that CML may have triggered the onset of SK through cytokine release.

Persistence of BCR/ABL mRNA-expressing bone-marrow cells in patients with chronic myelogenous leukemia in complete cytogenetic remission induced by interferon-alpha therapy.

Complete hematologic and cytogenetic responses can be obtained with interferon-alpha (IFN-alpha) in 15-25% of the patients with chronic myelogenous leukemia (CML). In these patients, reverse-transcription polymerase chain reaction (RT-PCR) can be used to evaluate minimal residual disease. We studied 12 patients who remained Philadelphia-negative for a median period of 21 months on IFN-alpha therapy. Using RT-PCR, the specific transcript was found in all bone marrow (BM) samples. Ten patients still exhibiting a persistent residual clone remained in cytogenetic remission for a median period of 14 months. As we observed a dissociation between bcr-abl expression in BM and peripheral blood (PB) cells, and given the known fluctuations of the bcr-abl PCR results, we suggest that PB negative results should be confirmed on BM specimens. Alternatively, it remains to be demonstrated whether longitudinal monitoring of residual disease would benefit from quantitative PCR or double fluorescence in situ hybridization.

Second neoplasms following high-dose chemotherapy and autologous stem cell transplantation for malignant lymphomas: a report of six cases in a cohort of 171 patients from a single institution.

High dose chemotherapy with autologous stem cell transplantation (ASCT) is increasingly used in the treatment of patients with lymphoma. As previously shown with conventional treatments, second neoplasms are emerging as a long term complication of the procedure. In this study, we investigate the incidence of second neoplasm in a cohort of 171 patients treated with BEAM or BEAC regimens for Hodgkin's disease (n = 62) and non-Hodgkin's lymphomas (n = 109) followed up for a median of 52 months post ASCT. Six patients developed six second malignancies 12 to 105 months after ASCT: fibrolamellar carcinoma of the liver, malignant fibrous histiocytoma, pancreatic carcinoma, squamous cell carcinoma of the lung, invasive carcinoma of the vulva and acute myelogenous leukemia. The cumulative actuarial risk for developing second malignancy is 16.7% (95% confidence interval: 5.9-39.3%) 13 years after transplant. The age-adjusted incidence of cancer in the study group is 4.1 times higher than that of primary cancer in the general population. These data confirm that ASCT recipients are at increased risk of later malignancies. This complication adds significant morbidity and mortality to the transplant process and therefore, needs to be taken into account in long term evaluation of new strategies which involve early intensification in the treatment of lymphomas.

Impact of recombinant human granulocyte colony stimulating factor on dose intensity and toxicity of three cycles of methotrexate, vinblastine, doxorubicin and cisplatin in patients with previously untreated urothelial bladder carcinoma.

This single arm, open labeled, non randomized study was aimed to evaluate the toxicity of 3 cycles of MVAC (methotrexate, vinblastine, doxorubicin and cisplatin) with rhG-CSF (5 micrograms/kg/day from day 3 to day 14), on 14 patients with previously untreated infiltrating bladder carcinoma, 42 cycles were administered. Chemotherapy toxicity was very low, with 7% of neutropenia grade 3 or 4.4% of thrombocytopenia grade 2, no mucositis above grade 2 and no nadir sepsis. Bone pain related to rhG-CSF occurred in 14% of cycles. 88% of the cycles were given at full dose without any delay and mean relative dose intensity was 96.4% (RDI was 100% for 9 patients). One patient achieved a complete pathological response (cystectomy: 1) and 6 clinical responses with negative transurethral resection. Addition of rhG-CSF to MVAC chemotherapy allows a high dose intensity of MVAC with very low toxicity over 3 cycles. This association should be compared to standard MVAC or intensified regimens to evaluate efficacy, toxicity, and cost effectiveness.

Benefits of granulocyte-colony-stimulating factor after stem cell transfusion in intensive sequential chemotherapy for breast cancer.

The aim of this study was to evaluate the clinical and economic benefit of filgrastim given with intensive sequential chemotherapy. Women with poor-prognosis breast cancer received four cycles of high-dose cyclophosphamide (3 g/m2) and doxorubicin (75 mg/m2), followed by filgrastim 5 microg/kg/dy, stem cell collection after the cycle 1, and stem cell infusion after cycle 3 and cycle 4. The first cohort received filgrastim after the fourth cycle but the second cohort did not.Thirty three patients were included in the first cohort and 13 in the second. The results indicate that the duration of grade IV neutropenia was shorter in the group given filgrastim as was the median time to recover an absolute neutrophil count (ANC) > 1.0 x 10(9)/L. The rate and duration of the rehospitalizations were higher in the group not receiving filgrastim. We found that costs such as drugs and hospitalizations were significantly higher (p = 0.032 and p = 0.049) in the non-filgrastim-treated group. Using ANC > 1.0 x 10(9)/L as an intermediary efficiency criterion it was more cost effective to give filgrastim. It can be concluded from this study that filgrastim can decrease the duration of grade IV neutropenia in patients receiving intensive sequential chemotherapy. This, in turn, reduces the cost of hospitalization. However, in our study, this reduction of neutropenia did not have any impact on further therapy.

Phase II study of a combination of cisplatin, all-trans-retinoic acid and interferon-alpha in squamous cell carcinoma: clinical results and pharmacokinetics.

UNLABELLED: Preclinical and clinical data suggest a certain antitumor efficacy for combination of retinoids, cytokines and cytotoxic compounds. PATIENTS AND METHODS: Between November 1994 and October 1996, 38 patients with advanced squamous cell carcinoma were enrolled in a phase II study to investigate an association of low-dose all-trans-retinoic acid (tRA) (40 mg/m2/day, 84 days), interferon-alpha (IFN-alpha) (6.10(6) UI/day, 84 days s.c.) and cisplatin (40 mg/m2/day, day 1, 28 and 56, i.v.). A Pharmacokinetic evaluation was performed on 12 patients. RESULTS: The incidence of grade 3/4 hematologic toxicities was moderate (< 20%). Extra hematological toxicities were frequent but easily manageable and not life threatening. However, treatment delivery was poor since only 6 patients (16%) received full therapy. Seven objective responses were observed (21%), suggesting some degree of synergism between tRA, IFN-alpha and cisplatin. Interestingly, continuous tRA treatment in combination with IFN-alpha and cisplatin did not induce a significant decrease in plasma levels, as had been previously described. CONCLUSIONS: Regarding the short median response duration and the frequency of toxic events, this regimen should no longer be recommended in pretreated patients with advanced disease. However, the consistent response rate reported here may warrant further investigation in an early setting.

Outpatient sequential high dose alkylation with stem cell support for patients with advanced breast cancer: a phase I-II study.

We evaluated the feasibility of administering, in an out-patient setting, a sequential high dose alkylating regimen with hematopoietic growth factor (HGF) and stem cell support to patients with advanced breast cancer. Peripheral blood stem cells (PBSC) were previously collected after chemotherapy and HGF. Two consecutive cycles of alkylating agents were planned: Thiotepa (T) then, 15 days later, BCNU (B). Three dose levels of each agent were administered in cohorts of consecutive patients: 400, 500 and 600 mg/m2 respectively. HGF and reinfusion of PBSC followed both cycles. Toxicity and response were evaluated according to the WHO recommendations. From April 1996 to August 1988, 30 women were enrolled: 8 in the first, 12 in the second and 10 in the third dose level. In all cases, B was administered after T with a median delay of 25 days because of grade 3/4 hematological toxicity. 4 patients did not receive B because of previous lung radiotherapy, persistent tricytopenia or insufficient PBSC collection. 19 patients with measurable lesions were considered for response. The objective response rate was 48% (11% CR, 37% PR). We recommended T and B at a dose of 600 mg/m2 to conduct a phase II study in metastatic breast cancer and even to administer B before T.

[Intensive chemotherapy with autologous stem cell transplantation in ovarian cancers: analysis of 67 patients treated at the Paoli-Calmettes Institute and a review of the literature]. / Chimiothérapie intensive avec autogreffe de cellules souches hématopoïétiques dans le cancer de l'ovaire: analyse de 67 patientes traitées à l'institut Paoli-Calmettes et revue de la littérature.

Despite important initial chemosensitivity, advanced ovarian cancer has a bad prognosis with a median survival of 20 to 30 months. These results might be better with intensive chemotherapy. We analysed 67 patients treated by intensive chemotherapy with autologous stem cell transplantation for advanced ovarian cancer at Institute Paoli-Calmettes between 1980 and 1994. Population was divided in two groups: salvage group (n = 30) for initial chemotherapy-refractory patients and consolidation group (n = 37) for sensitive patients. Several successive conditioning regimens were used, all based on alkylating agents. Principal toxicities were severe aplasia and mucositis. Four patients died from toxicity related to infection during strong immunosuppression. In salvage group, 9 out of 21 evaluable patients responded (43%), but duration of responses was short (median range of 5 months) and 2-year overall survival rate was 8% after transplantation. In consolidation group, 19 patients are alive and 15 are without disease progression with a median follow-up of 42 months (17, 161) after diagnosis. Five-year disease-free survival rate is 28% (median range of 35 months) and 5-year overall survival rate is 48% (median range of 41 months). Intensification does not seem to be long term beneficial for initial chemotherapy refractory patients, despite objective responses rate better than classical treatment. On the other hand, results seem better than conventional treatments in case of chemosensitive disease and should be confirmed prospectively in larger cohort of patients. Moreover, other research directions are open like intensification supported by hematopoietic growth-factors and peripheral stem cells, definition of best conditioning regimen, use of taxanes, and intensification in first line chemotherapy after initial surgery.