TRANSLATIONAL AND CLINICAL RESEARCH

Transcription

1 TRANSLATIONAL AND CLINICAL RESEARCH Long-Term Follow-Up of Patients with Non-Hodgkin Lymphoma Following Myeloablative Therapy and Autologous Transplantation of CD34 -Selected Peripheral Blood Progenitor Cells MATHIAS WITZENS-HARIG, a CONNY HEILMANN, a MANFRED HENSEL, a MARTIN KORNACKER, a AXEL BENNER, b RAINER HAAS, a STEFAN FRUEHAUF, a ANTHONY D. HO a a Department of Hematology and Oncology, University Hospital Heidelberg and b Central Unit Biostatistics, German Cancer Research Center, Heidelberg, Germany Key Words. CD34 -selected PBPC Long-term follow-up Graft engineering Lymphoma High-dose chemotherapy ABSTRACT Graft engineering by CD34 selection of peripheral blood progenitor cells (PBPC) has been used in non-hodgkin lymphoma (NHL) with the aim to reduce relapse related to tumor cells within the graft. From September 1995 to January 2000, 39 patients with newly diagnosed (n 31) or relapsed (n 8) NHL were treated in our institution with myeloablative therapy followed by CD34 selected autologous PBPC transplantation. Thirty-one patients were diagnosed with follicular lymphoma, and eight patients with mantle-cell lymphoma. All patients had advanced disease (26% of patients stage III and 74% stage IV, Ann Arbor classification). Induction therapy resulted in a complete remission in 17 patients and a partial remission in 22 patients. PBPC were mobilized after cytotoxic chemotherapy with granulocyte colony-stimulating factor support. CD34 selection was performed using immunomagnetic beads (Baxter Isolex 300SA or 300i Magnetic Cell Separation System). Most patients (85%) received total body irradiation and high-dose cyclophosphamide as myeloablative regimen. Twelve patients also received rituximab 375 mg/m 2 before radiation and before the start of the cyclophosphamide treatment. The mean CD34 cell number for transplantation was CD34 cells/kg of body weight. Platelet recovery (>20,000/ l median on day 13) and leukocyte recovery (>1,000/ l median on day 12) were within expected range. The estimated median follow-up was 47 months. The probabilities of freedom from progression, overall survival, and event-free survival 4 years after transplantation were 96%, 90%, and 87%, respectively, for patients with follicular lymphoma and 42%, 63%, and 33%, respectively, for patients with mantle-cell lymphoma. Risk factors for relapse were age and extranodal manifestation of disease. The rate of lethal infections in the 12-month follow-up period was 8%. We conclude that CD34 selection of autologous transplants following myeloablative therapy is feasible and results in long-term remission in the majority of patients, but the procedure is probably related to a higher rate of lethal infections. STEM CELLS 2007;25: INTRODUCTION Follicular lymphoma (FL) and mantle-cell lymphoma (MCL) are hematologic B-cell malignancies that were considered to be not curable using conventional cytotoxic chemotherapy. Although standard therapy with CHOP (cyclophosphamide 750 mg/m 2 per day, doxorubicin 50 mg/m 2 per day, vincristine 1.4 mg/m 2 per day, and prednisone 100 mg per day) can induce remissions in the majority of patients, median survival is only 8 years in FL and only 2 3 years in MCL. Higher doses of chemotherapy followed by autologous transplantation have been attempted to improve therapeutic outcome. The use of peripheral blood progenitor cells (PBPC) as the source of hematopoietic stem cells has been evaluated as an alternative approach to improve upon the efficacy of autologous transplantation in patients with lymphoma, leukemia, and multiple myeloma [1 8]. In randomized studies, event-free survival (EFS) was prolonged in patients with FL and MCL who received myeloablative therapy and PBPC transplantation as primary treatment [9, 10]. Marrow transplantation was shown to improve response rates, disease-free survival, and overall survival (OS) in a randomized study when compared with standard chemotherapy [11]. This kind of treatment has also been widely used for patients with non-hodgkin lymphoma (NHL) and poor prognosis [12]. One concern in the context of autologous transplantation is the risk of reinfusing contaminating NHL cells, which may contribute to relapse [13, 14]. Therefore, immunomagnetic purging techniques that allow positive selection of CD34 hematopoietic progenitor cells and the reduction of potentially contaminating cells have been applied [15 19]. One important method for positive in vitro selection of CD34 cells, which allows the clinical isolation of CD34 to a very high purity, is magnetically activated cell sorting [18, 20]. In this study, CD34 cells selected by immunomagnetic beads that were consequently detached from the cells were used for autografting of 39 patients with NHL. The primary objective of the study was to evaluate the safety of the transplantation with selected CD34 cells as assessed by the time to hematopoietic reconstitution and the incidence of adverse events. The secondary objectives were to evaluate clinical response rate, Correspondence: Mathias Witzens-Harig, M.D., University of Heidelberg Hospital Internal Medicine V, INF 410 Heidelberg Germany. Telephone: ; Fax: ; mathias.witzens-harig@med.uni-heidelberg.de Received December 7, 2005; accepted for publication September 8, AlphaMed Press /2007/$20.00/0 doi: /stemcells STEM CELLS 2007;25:

2 Witzens-Harig, Heilmann, Hensel et al. 229 disease-free survival, and OS. We found transplantation of CD34 -selected grafts following myeloablative therapy to be feasible and to result in long-term remission in the majority of patients. PATIENTS AND METHODS Between September 1996 and January 2000, 39 patients with NHL were treated with high-dose therapy (HDT) and autologous transplantation using CD34-selected PBPC. The study was conducted according to the guidelines of the local ethics committee, and informed consent was obtained from all patients. Patient characteristics are shown in Table 1. Twenty patients were men, and 19 patients were women. The median age was 48 years (range, years). Before transplantation, 31 patients were in first remission, and 8 patients were in second remission. FL was present in 31 patients, and MCL in 8. All patients were in advanced stage of disease, with 10 patients with stage III and 29 in stage IV (modified Ann Arbor classification). The prognostic parameter lactate dehydrogenase (LDH) was elevated in five patients. The Follicular Lymphoma International Prognostic Index (FLIPI) was determined for the 31 patients with FL. Thirty-two percent had a score of 1 (low), 48% a score of 2 (intermediate), and 19% a score of 3 or 4 (high). In 29 patients, there was an extranodal disease manifestation, with bone marrow as the most frequent site (Table 2). Thirty-one patients received first-line treatment, and eight patients were treated in relapse. The median duration of the first remission in these patients was 19 months (range, 5 77 months). Treatment history of the patients in relapse is shown in Table 3. Study Treatment Thirty-four patients received induction therapy followed by mobilization therapy and leukapheresis. Five patients in relapse received mobilization therapy without previous induction therapy. The results of the induction therapy are shown in Tables 4 and 5. PBPC mobilization was performed by using the HAM (cytosinarabinoside 2 g/m 2 every 12 hours days 1 and 2 and mitoxantrone 10 mg/m 2 days 2 and 3) [21] or D-BEAM (dexamethasone 24 mg per day, days 1 7, melphalan 30 mg/m 2 day 2, carmustine 60 mg/m 2 day 3, etoposide 75 mg/m 2 per day, and cytosinarabinoside 200 mg/m 2 per day days 4 7) protocol, followed by administration of granulocyte colony-stimulating factor (G-CSF). Sixteen patients who received HAM were additionally treated with rituximab as in vivo purging. Stem cell harvest was performed between 10 and 17 days after the last day of mobilization therapy. Leukapheresis Autologous PBPC were obtained by continuous-flow leukapheresis using a Fenwal CS 3000 (Baxter, Munich, Germany, baxter.de) or, since 1998, also by a Spectra device (Cobe Laboratories, Lakewood, CA, Blood volume processed per run was l. Flow rate was ml/minute (Fenwal CS 3000) or ml/minute (Spectra device). The collection goal was to obtain a transplantation with a minimum of CD34 cells/kg of body weight. Thirteen patients required one leukapheresis, 24 patients two leukaphereses, 1 patient four leukaphereses, and 1 patient six leukaphereses to reach the collection goal. CD34 selection was performed in all patients; 15 patients obtained an additional nonselected transplant as back-up. Back-up leukapheresis products were cryopreserved in 10% dimethyl sulfoxide by controlled-rate freezing and were stored in the vapor phase of liquid nitrogen. CD34 Selection Immunomagnetic CD34 selection was performed with the Baxter Isolex 300 SA or the 300i Magnetic Cell Separation System (Baxter Immunotherapy, Irvine, CA) according to the manufacturer s protocol. The CD34-selected PBPC were then cryopreserved in 10% dimethyl sulfoxide by controlled-rate freezing and were stored in the vapor phase of liquid nitrogen. CD34 content was controlled by flow cytometry. In three patients, an additional CD19 depletion was performed. HDT Thirty-three patients were conditioned with a combination of 14.4 Gy of fractionated total body irradiation (TBI; over 4 days) and cyclophosphamide 200 mg/kg (50 mg/kg per day over 4 days). Twelve patients also received rituximab (375 mg/m 2 ) before radiation and before the start of the cyclophosphamide treatment. Six patients were conditioned with the BEAM (carmustin 300 mg/m 2 per day on day 1, cytosinarabinoside 200 mg/m 2 2 per day on days 2 5, etoposide 100 mg/m 2 2 per day on days 2 5, and melphalan 140 mg/m 2 per day on day 6) protocol. All patients received a transplant with CD34 -selected PBPC. Hematologic restitution was defined as platelets above 20 cells per nanoliter and leukocytes above 1 cell per nanoliter. In our center, differential blood counts were not available on public holidays and weekends on a routine basis. To monitor reconstitution more closely, we therefore chose the absolute leukocyte count as lead parameter. Patients with fever ( 38.5 C) were treated with ceftazidim. If the fever persisted for more than 72 hours, vancomycin and amphotericin B were added. Red cells were transfused at a hemoglobin level below 8 g/dl and according to clinical requirement. Platelets were given if the platelet count was below 20 cells per nanoliter. Only three patients required G-CSF for leukocyte recovery. Treatment After HDT Thirteen patients had bulky disease and were irradiated at a median of 3 months (range, 1 8 months) after HDT with a median dose of 30 Gy (range, Gy). Three patients received interferon maintenance therapy; however, in two of these patients the treatment had to be stopped because of pancytopenia. Statistical Analysis EFS, PFS, and OS were calculated with the SAS software (SAS Institute, Cary, NC, Univariate analysis was calculated with Cox regression analysis with Firth s correction. RESULTS PBPC Transplantation and Hematologic Reconstitution After mobilization therapy with HAM or D-BEAM, all patients were in remission (complete, n 17; partial, n 22). In 26 (67%) of 39 patients additional leukaphereses were performed because of the cell selection process. Of the 39 patients, all CD34 cells (100%) recovered in 6 patients, whereas CD34 cell losses of up to 81% were observed in 32 patients (Table 2). The purity of CD34 -selected products was a median of 98.35% (range, 59.2% %; Table 2). The purity of the CD34 - selected products was excellent with the Baxter device and, thus, comparable with recently used devices [20]. Following high-dose conditioning therapy, the patients received a CD34 -selected transplant that contained a median of CD34 cells per kg of body weight (range, cells per kg of body weight) and a median of mononuclear cells (MNC)/kg of body weight (range, MNC/kg of body weight). Platelets were reconstituted at a median of 13 days after transplantation; leukocyte recovery occurred a median of 12 days after transplantation, and two patients experienced graft failure. One of these patients received a nonselected back-up transplant 14 days after the first transplantation. The other patient died 57 days after transplantation as a consequence of multiple organ failure. The patients experienced a median of 3

3 230 CD34 Selected Transplantation in NHL Patients Table 1. Clinical characteristics Patient No. Age (years) Sex Disease Time from diagnosis to TPX (months) Stage at diagnosis scd25 elevated at diagnosis FLIPI score at diagnosis Result of induction Mobilization regimen CD34 cells/kg Viability bw 10 6 Conditioning regimen Days to reconstitution Platelets >20/nl Leucocytes >1/nl Outcome Outcome (months from TPX) Relapse (months after TPX) 1 45 Female FL 7 IV A 1 2 PR HAM, Ritux TBI, Cp, Ritux Alive 39 No 2 49 Female FL 15 III B 0 2 CR HAM TBI, Cp Alive 59 No 3 33 Male FL 6 IV A 0 2 CR HAM TBI, Cp Alive 57 No 4 37 Female FL 8 III A 0 1 CR HAM TBI, Cp Alive 60 No 5 37 Female FL 6 IV A 1 3 CR HAM TBI, Cp Alive 63 No 6 51 Female FL 9 III A 1 2 PR HAM, Ritux TBI, Cp, Ritux Alive 39 No 7 58 Female FL 11 IV B 1 3 PR HAM, Ritux TBI, Cp, Ritux 1 9 Died of respiratory 10 No failure 8 32 Female FL 17 III A 0 1 CR HAM TBI, Cp Alive 47 No 9 39 Male FL 5 IV A 0 2 CR HAM TBI, Cp Alive 38 No Male FL 8 IV B 0 2 PR HAM TBI, Cp Died of septic multiple 2 No organ failure Male FL 8 IV A 1 1 PR HAM, Ritux TBI, Cp, Ritux Alive 46 No Male FL 10 III A 1 1 PR HAM TBI, Cp Alive 45 No Male FL 54 IV B 1 3 CR D-BEAM, Ritux ,2 BEAM Alive 39 No Female FL 13 IV B 0 4 CR HAM, Ritux TBI, Cp, Ritux Alive 30 No Male FL 7 III A 1 1 CR HAM TBI, Cp Alive 62 No Male FL 10 III B 1 3 PR HAM, Ritux BEAM 8 12 Alive 30 No Female FL 9 IV A 1 2 CR HAM TBI, Cp Alive 67 No Female FL 23 IV A 1 2 PR HAM, Ritux TBI, Cp, Ritux Alive 44 No Male FL 8 IV B 1 2 PR HAM, Ritux TBI, Cp, Ritux Alive 36 No Male FL 10 IV A 0 2 PR HAM TBI, Cp Alive 62 No Female FL 9 IV A 1 1 CR HAM, Ritux ,3 TBI, Cp, Ritux Alive 41 No Female FL 82 IV A 0 2 CR HAM TBI, Cp Alive 46 No Male FL 12 IV A 1 2 PR HAM TBI, Cp 13 9 Alive 49 No Female FL 57 III A 0 1 CR HAM BEAM Alive 55 Yes (13) Male FL 9 IV A 0 2 CR HAM TBI, Cp Alive 62 No Male FL 42 III A 0 1 PR HAM TBI, Cp Alive 62 No Female FL 31 IV A 0 1 PR HAM TBI, Cp Alive 58 Yes (57) Female FL 46 IV A 0 2 CR HAM, Ritux TBI, Cp, Ritux Died of septic multiple organ failure Male FL 81 III A 0 1 PR D-BEAM BEAM Alive 54 No Male FL 9 IV A 1 2 PR HAM, Ritux TBI, Cp, Ritux Alive 46 No Female FL 9 IV B 1 3 PR HAM, Ritux TBI, Cp, Ritux Alive 47 No Female MCL 7 IV B 1 1 (IPI) CR HAM TBI, Cp Alive 52 Yes (29) Female MCL 28 IV A 1 1 (IPI) PR HAM, Ritux d-beam Alive 29 No Female MCL 81 IV B 1 2 (IPI) PR HAM TBI, Cp 9 13 Died of infection 15 No Male MCL 11 IV B 1 2 (IPI) PR HAM TBI, Cp Alive 55 Yes (1.25) Male MCL 9 IV A 0 1 (IPI) CR HAM TBI, Cp 6 12 Died of relapse 60 Yes (57) Male MCL 29 IV A 1 1 (IPI) CR HAM, Ritux TBI, Cp 9 9 Died of cardiorespiratory failure 1,5 No 26 Yes (12) Male MCL 4 IV B 1 3 (IPI) PR HAM, Ritux 17.6 NA BEAM 11 9 Died of cerebral hemorrhage 9 Yes (4) Male MCL 86 IV A 1 1 (IPI) PR HAM, Ritux TBI, Cp, Ritux Alive 45 No Abbreviations: bw, body weight; Cp, cyclophosphamide; CR, complete remission; FL, follicular lymphoma; FLIPI, Follicular Lymphoma International Prognostic Index; MCL, mantle-cell lymphoma; NA, not available; PR, partial remission; Ritux, rituximab; TBI, total-body irradiation; TPX, transplantation.

4 Witzens-Harig, Heilmann, Hensel et al. 231 Table 2. Graft characteristics Patient No. Leukapheresis (No.) Disease CD34 Before selection After selection Transplantation CD / MNC 10 9 / CD34 CD19 MNC 10 9 / CD / CD34 recovery CD / CD34 cells transplanted CD34 viability 1 2 FL FL FL NA NA FL FL FL FL NA NA FL , FL , FL NA NA FL FL FL FL FL NA NA FL FL NA NA FL FL FL , FL FL , FL FL FL FL FL FL FL , FL NA NA FL MCL MCL , MCL MCL MCL MCL MCL NA NA NA NA NA NA NA NA NA 17.6 NA 39 2 MCL When CD34 recoveries 100% were calculated, these figures were set to 100%. Patients 1, 13, and 39 were additionally CD19 -selected. Abbreviations: bw, body weight; FL, follicular lymphoma; MCL, mantle-cell lymphoma; MNC, mononuclear cells; NA, not available. Table 3. Extranodal disease Location Patient No. Bone marrow 27 Spleen 9 Liver 2 Kidney 1 Pleura 1 Lung 1 days with fever (range, 0 11 days). Patients with fever were treated with intravenous antibiotics for a median of 12 days. There were a median of six platelet transfusions and four erythrocyte transfusions in each patient. Parenteral nutrition was required by most patients and lasted a median of 10 days. Median duration of hospital stay was 16 days (range, days; Table 6). Platelet and leukocyte recovery were dependent on the number of CD34 cells transplanted (p.0001 and.0006, respectively; Fig. 1A, 1B). Patients who received a transplant containing more than CD34 cells per kg of body weight had a faster platelet and leukocyte reconstitution than patients who received transplants with less than CD34 cells per kg of body weight. Table 4. Previous therapies in patients who experienced relapse Therapy protocol Result Duration of remission (months) 6 CHOP CR 5 4 COP PR 5 4 CHOP CR 8 3 COP, 6 CHOP CR 19 6 CHOEP CR 19 6 CHOP CR 32 8 CHOP, bulk radiation 40 Gy CR 33 3 COP-BLAM, bulk radiation 40 Gy CR 77 Abbreviations: CHOEP, cyclophosphamide, doxorubicin, vincristine, prednisone, and etoposide; CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone; COP, cyclophosphamide, vincristine, and prednisone, partial remission; COP-BLAM, cyclophosphamide, doxorubicin, vincristine, prednisone, procarbazin, and bleomycin; CR, complete remission. Antibiotic Treatment The median duration of antibiotic treatment after transplantation was 12 days (range, 0 53 days). Age was a significant risk factor for long antibiotic treatment (p.04). The number of transfused stem cells also influenced the duration of the antibiotic treatment (p.07). The more PBSC the patients received, the less antibiotic treatment was required (p.07; not shown).

5 232 CD34 Selected Transplantation in NHL Patients Table 5. Results of induction chemotherapy No. of patients Protocol Total CR PR NC CHOP CHOEP MCP Abbreviations: CHOEP, cyclophosphamide, doxorubicin, vincristine, prednisone, and etoposide; CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone; CR, complete remission; MCP, mitoxantrone, chlorambucile, prednisone; NC, no change; PR, partial remission. Table 6. Transplantation and supportive therapy Median Range Autograft (CD34 cells/kg body weight 10 6 ) Days until platelet recovery ( /l) Days until leukocyte recovery ( /l) Intravenous antibiotics No. of red cell transfusions No. of platelet transfusions Duration of fever after transplantation (days) Duration of parenteral nutrition (days) Duration of hospital stay (days) Extranodal disease, elevated LDH, elevated scd25 ( 900 U/ml) [22], number of chemotherapy cycles, first-line versus second-line therapy, and rituximab therapy (yes vs. no) were not significant risk factors. Survival The probabilities of freedom from progression (FFP), OS, and EFS 4 years after transplantation were 96%, 90%, and 87%, respectively, for patients with FL and 42%, 63%, and 33%, respectively for patients with MCL. (Figs. 2A and 2B and 3A and 3B). The median follow-up time was 47 months. Risk factors for unfavorable EFS were age (p.07) and extranodal disease (p.07). Number of transplanted CD34 cells, elevated LDH, elevated scd25, number of chemotherapy cycles, first-line versus second-line therapy, and rituximab therapy (yes vs. no) were not significant risk factors. In the total group of 39 patients, there were seven deaths during the follow-up period (Table 7). In the small group of patients who were treated during relapse (n 8), there were three deaths. In the group of patients receiving first-line treatment (n 31), there were four deaths. Two deaths occurred within 100 days after transplantation, both due to septic multiple organ failure. One of these patients had experienced graft failure before. Taken together, the rate of death due to infection within the first year in this patient cohort is approximately 8%. Two additional patients died in the first year after transplantation, one because of cerebral mass hemorrhage and one because of respiratory failure. In years 2, 3, and 6 of follow-up there were three deaths (one infection, one respiratory failure, one relapse). All deaths occurred in patients with extranodal disease. Figure 1. Leukocyte and platelet reconstitution, transplanted CD34 cell number, and extranodal disease. (A): Leukocyte reconstitution and transplanted CD34 cell number. The red line represents patients (pts) who received transplants with more than cells per kg of body weight; the blue line represents pts who received transplants with less than cells per kg of body weight (p.0006) (B): Platelet reconstitution and transplanted CD34 cell number. The red line represents pts with transplants with more than cells per kg of body weight; the blue line represents pts with transplants with less than cells/kg body weight (p.0001). Pts who received transplants in second remission had a slower platelet reconstitution than did pts in first remission (p.01; data not shown). Pts with extranodal disease had a slower platelet reconstitution than did pts without extranodal disease (p.008). This is probably attributable to the fact that nearly all pts with extranodal disease in our cohort had bone marrow infiltration. The number of pts was too low for multivariate analysis. (C): Pts with extranodal disease had a slower platelet reconstitution than did pts without extranodal disease (p.008). Relapse After transplantation, there were relapses in 8 of 39 patients. Median time between transplantation and relapse was 21 months (range, 39 days to 59 months). The relapsed patients were treated heterogeneously. Two patients received RTE chemotherapy (rituximab 375 mg/m 2 day 1, endoxan 700 mg/m 2 days 2 4, and docetaxel 75 mg/m 2 day 2), one patient received rituximab monotherapy, two patients received an allogeneic transplantation, and one patient received a second autologous transplantation. DISCUSSION Here we report the results of a clinical trial in which CD34- selected PBPC were transplanted following HDT in 39 patients with NHL. The probabilities of FFP, OS, and EFS 4 years after transplantation were 96%, 90%, and 87%, respectively, for patients with FL and 42%, 63%, and 33%, respectively, for patients with MCL. Therefore, we conclude that transplantation

6 Witzens-Harig, Heilmann, Hensel et al. 233 Table 7. Deaths after transplantation Time of death Death causes 45 days Septic multiple organ failure 57 days Septic multiple organ failure 9 months Cerebral hemorrhage 10 months Respiratory failure 16 months Infection 26 months Cardiorespiratory failure 61 months Relapse Figure 2. EFS. (A): Probability for EFS 4 years after transplantation was 87.1% in the group of patients with FL. (B): Probability for EFS 4 years after transplantation was 33.3% in the group of patients with MCL. Abbreviations: EFS, event-free survival; FL, follicular lymphoma; MCL, mantle-cell lymphoma. Figure 3. Overall survival (OS) (A): Probability for OS 4 years after transplantation was 90.3% in the group of patients with FL. (B): Probability for overall survival 4 years after transplantation was 62.5% in the group of patients with MCL. Abbreviations: FL, follicular lymphoma; MCL, mantle-cell lymphoma. of CD34 -selected PBPC is feasible and results in long-term clinical remission in the majority of patients. A matched-pair analysis of the patients with FL in this study with a patient group that was treated with HDT and unselected transplantation showed no difference in OS (M. Kornacker et al., submitted for publication). Previously, Horning et al. treated patients with relapsed FL with a related protocol [23]. In that study, 37 patients were treated with CVP (cyclophosphamide, vincristine, and prednisone); HDT with TBI, etoposide, and cyclophosphamide; and a bone marrow transplantation with purged bone marrow. B-cell depletion was performed with negative selection with anti-cd9, -CD10, -CD19 and -CD20 antibodies. OS and FFP were 92% and 76%, respectively, after 5 years and 86% and 70%, respectively, after 10 years. Patients in the Horning et al. study were younger (median age, 37 years) than in our study (median age, 48 years), and all patients in the Horning et al. study were treated in first remission. Hematologic reconstitution in our study was within expected range. After 12 days, there was a leukocyte recovery ( 1 cell per nanoliter); after 13 days, platelets reconstituted ( 20 cells per nanoliter). Earlier studies have shown that CD34 cells per kg of body weight is required for a rapid and sustained hematologic recovery [24, 25]. Later studies have shown that platelet recovery can be further accelerated when more than CD34 cells per kg of body weight is transplanted [18, 26, 27]. These observations are in accordance with this data set where transfusion of more than CD34 cells per kg of body weight resulted in a significantly accelerated platelet and leukocyte reconstitution. Risk factors for a prolonged platelet recovery in our study were a low CD34 cell number, extranodal manifestation of disease (p.008), and treatment in relapse (p.01). For OS, age and extranodal manifestation of disease were the most important risk factors. Due to the low number of patients in this study, this risk factor did not reach statistical significance (p.12 and.21, respectively). Interestingly, the risk factors we have identified are in accordance with the results of a large international cooperative analysis including 1,795 patients [28]. In this analysis, five prognostic factors contribute to the FLIPI. The prognostic factors are age 60 years, Ann Arbor stage III-IV, Hb 12 g/dl, elevated serum LDH level, and more than four nodal manifestations of disease. In the total group, there were seven deaths during the follow-up period. Two deaths occurred within 100 days after transplantation, both due to septic multiple organ failure. One of these patients had experienced graft failure before. Taken together, the rate of deaths due to infection within the first year in this group of patients is approximately 8%. This rate is higher than that reported in previous studies using autografting with unselected transplants [29]. CD34 selection can be achieved with avidinbiotin immunoadsorption or, as in our study, with immunomagnetic techniques. In a study with 37 multiple myeloma patients, immunoadsorption resulted in a tumor cell reduction of log in the transplant [30]. Another purging strategy uses monoclonal antibodies against B-cell-associated epitopes (CD19, CD20, CD22, CD23). The combination of CD34 selection and antibody depletion resulted in a tumor cell depletion of 3.5 log [31]. As mentioned earlier, this study was initiated at a time when the anti-cd20 antibody rituximab was not available. Following the addition of rituximab to cytotoxic chemotherapy, it became apparent that this combination is highly effective in depleting B cells from the peripheral blood during G-CSF-enhanced leukocyte recovery. This in vivo purging capacity is best reflected by

7 234 CD34 Selected Transplantation in NHL Patients the polymerase chain reaction negativity that can be achieved in patients originally harboring t(14;18)-positive lymphoma cells. A first study was published by Gianni et al. showing that in patients with low-grade lymphoma, the addition of two cycles of rituximab to the mobilization therapy resulted in a reduced tumor cell contamination of the transplant [32]. These results were confirmed in further studies [33 35]. Purging of the stem cell transplant may be of potential benefit because, in FL, tumor cells frequently contaminate bone marrow [14, 36, 37] and PBPC transplants [38], and may contribute to relapse [13, 14]. However in vitro purging techniques, as applied in this study, are time-consuming and expensive. The cost of CD34 cell selection is approximately 8,500 Euros. The majority of our patients (67%) needed at least one additional leukapheresis because of the cell selection process. The cost per leukapheresis adds 1,500 Euros. These leukaphereses are also considered an additional patient strain. Another problem of these techniques seems to be the relatively high rate of lethal infections occurring in our study, a rate of 8%. Treatment with the anti-cd20 antibody rituximab before stem cell collection can be easily performed and serves both as an anti-lymphoma treatment and an in vivo purging and can now be considered the purging method of choice. ACKNOWLEDGMENTS M.W.-H. and C.H. contributed equally to this work. R.H. is currently affiliated with the Department of Hematology and Oncology and Clinical Immunology, University Hospital Düsseldorf, Düsseldorf, Germany. DISCLOSURES The author indicates no potential conflicts of interest. REFERENCES 1 Korbling M, Dorken B, Ho AD et al. Autologous transplantation of blood-derived hemopoietic stem cells after myeloablative therapy in a patient with Burkitt s lymphoma. Blood 1986;67: Haas R, Ho AD, Bredthauer U et al. Successful autologous transplantation of blood stem cells mobilized with recombinant human granulocytemacrophage colony-stimulating factor. Exp Hematol 1990;18: Ho AD, Haas R, Korbling M et al. Utilization of recombinant human GM-CSF to enhance peripheral progenitor cell yield for autologous transplantation. Bone Marrow Transplant 1991;7(suppl 1): Goldschmidt H, Hegenbart U, Haas R et al. Mobilization of peripheral blood progenitor cells with high-dose cyclophosphamide (4 or 7 g/m2) and granulocyte colony-stimulating factor in patients with multiple myeloma. Bone Marrow Transplant 1996;17: Marit G, Faberes C, Pico JL et al. Autologous peripheral-blood progenitor-cell support following high-dosechemotherapy or chemoradiotherapy in patients with high-risk multiple myeloma. J Clin Oncol 1996;14: Schlenk RF, Dohner H, Pforsich M et al. Successful collection of peripheral blood progenitor cells in patients with acute myeloid leukaemia following early consolidation therapy with granulocyte colonystimulating factor-supported high-dose cytarabine and mitoxantrone. Br J Haematol 1997;99: Fruehauf S, Schmitt K, Veldwijk MR et al. Peripheral blood progenitor cell (PBPC) counts during steady-state haemopoiesis enable the estimation of the yield of mobilized PBPC after granulocyte colony-stimulating factor supported cytotoxic chemotherapy: An update on 100 patients. Br J Haematol 1999;105: Gregory SA, Trumper L. Chemotherapy dose intensity in non-hodgkin s lymphoma: Is dose intensity an emerging paradigm for better outcomes? Ann Oncol 2005;16: Lenz G, Dreyling M, Schiegnitz E et al. Myeloablative radiochemotherapy followed by autologous stem cell transplantation in first remission prolongs progression-free survival in follicular lymphoma: Results of a prospective, randomized trial of the German Low-Grade Lymphoma Study Group. Blood 2004;104: Dreyling M, Lenz G, Hoster E et al. Early consolidation by myeloablative radiochemotherapy followed by autologous stem cell transplantation in first remission significantly prolongs progression-free survival in mantle-cell lymphoma: results of a prospective randomized trial of the European MCL Network. Blood 2005;105: Philip T, Guglielmi C, Hagenbeek A et al. Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive non-hodgkin s lymphoma. N Engl J Med 1995; 333: Haioun C, Lepage E, Gisselbrecht C et al. Survival benefit of high-dose therapy in poor-risk aggressive non-hodgkin s lymphoma: final analysis of the prospective LNH87 2 protocol A Groupe d Etude des Lymphomes de l Adulte study. J Clin Oncol 2000;18: Brenner M, Rill D, Moen R et al. Gene-marking to trace origin of relapse after autologous bone-marrow transplantation. Lancet 1993;341: Sharp JG, et al. Significance of detection of occult non-hodgkin s lymphoma in histologically uninvolved bone marrow by a culture technique. Blood 1992;79: Law P, Ishizawa L, van de Ven C et al. Immunomagnetic positive selection and colony culture of CD34 cells from blood. J Hematother 1993;2: Fruehauf S, Haas R, Zeller WJ et al. CD34 selection for purging in multiple myeloma and analysis of CD34 B cell precursors. STEM CELLS 1994;12: Voso MT, Hohaus S, Moos M et al. Lack of t(14;18) polymerase chain reaction-positive cells in highly purified CD34 cells and their CD19 subsets in patients with follicular lymphoma. Blood 1997;89: Voso MT, Hohaus S, Moos M et al. Autografting with CD34 peripheral blood stem cells: Retained engraftment capability and reduced tumour cell content. Br J Haematol 1999;104: Vescio R, Schiller G, Stewart AK et al. Multicenter phase III trial to evaluate CD34( ) selected versus unselected autologous peripheral blood progenitor cell transplantation in multiple myeloma. Blood 1999; 93: Miltenyi S, Muller W, Weichel W et al. High gradient magnetic cell separation with MACS. Cytometry 1990;11: Ho AD, del Valle F, Ruckle H et al. Mitoxantrone and high-dose cytarabine as salvage therapy for refractory non-hodgkin s lymphoma. Cancer 1989;64: Ho AD, Maruyama M, Maghazachi A et al. Soluble CD4, soluble CD8, soluble CD25, lymphopoieitic recovery and endogenous cytokines after high-dose therapy and blood stem cell transplantation. Blood 1994;84: Horning SJ, Negrin RS, Hoppe RT et al. High-dose therapy and autologous bone marrow transplantation for follicular lymphoma In first or partial remission: Results of a phase II clinical trial. Blood 2001;97: Haas R, Möhle R, Fruehauf S et al. Patient characteristics associated with successful mobilizing and autografting of peripheral blood progenitor cells in malignant lymphoma. Blood 1994;83: Bensinger W, Appelbaum F, Rowley S et al. Factors that influence collection and engraftment of autologous peripheral blood stem cells. J Clin Oncol 1995;13: Siena S, Schiavo R, Pedrazzoli P et al. (2000) Therapeutic relevance of CD34 cell dose in blood cell transplantation for cancer therapy. J Clin Oncol 2000;18: Blystad AK, Delabie J, Kvaloy S et al. (2004) Infused CD34 cell dose, but not tumour cell content of peripheral blood progenitor cell grafts, predicts clinical outcome in patient with diffuse large-cell lymphoma and follicular lymphoma grade 3 treated with high-dose therapy. Br J Haematol 2004;125: Solal-Celigny P, Roy P, Colombat P et al. Follicular lymphoma international prognostic index. Blood 2004;104: Bensinger WI, Rowley SD, Demirer T et al. High-dose therapy followed by autologous hematopoietic stem-cell infusion for patients with multiple myeloma. J Clin Oncol 1996;14: Schiller G, Vescio R, Freytes C et al. Transplantation of peripheral blood progenitor cells after high-dose chemotherapy for patients with advanced multiple myeloma. Blood 1995;86: Mohr M, Dalmis F, Hilgenfeld E et al. Simultaneous immunomagnetic CD34 cell selection in peripheral blood progenitor cell samples of patients suffering from B-cell non-hodgkin s lymphomas. Clinical Cancer Res 2001;7: Gianni AM, Magni M, Nicola DI et al. In-vivo purging of circulating CD34 progenitor cells in low-grade lymphoma with rituximab and high-dose therapy. Blood 1998;92:481.

8 Witzens-Harig, Heilmann, Hensel et al Belhadj K, Delfau-Larue MH, Elgnaoui T et al. Efficiency of in vivo purging with rituximab prior to autologous peripheral blood progenitor cell transplantation in B-cell non-hodgkin s lymphoma: a single institution study. Ann Oncol 2004;15: Magni M, Di Nicola M, Devizzi L et al. Successful in vivo purging of CD34-containing peripheral blood harvests in mantle cell and indolent lymphoma: evidence for a role of both chemotherapy and rituximab infusion. Blood 2000;96: Voso MT, Pantel G, Weis M et al. In vivo depletion of B cells using a combination of high-dose cytosine arabinoside/mitoxantrone and rituximab for autografting in patients with non-hodgkin s lymphoma. Br J Haematol 2000;109: Gribben J, Freedman A, Woo S et al. All advanced stage non- Hodgkin s lymphomas with a polymerase chain reaction amplifiable breakpoint of bcl-2 have residual cells containing the bcl-2 rearrangement at evaluation and after treatment. Blood. 1991;78: McQuaker I, Haynes A, Anderson S et al. Engraftment and molecular monitoring of CD34 peripheral-blood stem-cell transplants for follicular lymphoma: a pilot study. J Clin Oncol 1997;15: Leonard BM, Hetu F, Busque L et al. Lymphoma cell burden in progenitor cell grafts measured by competitive polymerase chain reaction: Less than one log difference between bone marrow and peripheral blood sources. Blood 1998;91: