Leukemia (1998) 12, 1699–1707  1998 Stockton Press All rights reserved 0887-6924/98 $12.00 http://www.stockton-press.co.uk/leu Autologous hematopoietic stem cell transplantation as salvage treatment for advanced B cell chronic lymphocytic leukemia L Sutton1, K Maloum2, H Gonzalez1, H Zouabi1, N Azar3, C Boccaccio3, F Charlotte4, J-M Cosset5, J Gabarre1, V Leblond1, H Merle-Beral2 and J-L Binet1,2

Departments of 1Clinical Hematology, 2Biological Hematology, 3Hemobiology, and 4Pathology, Hoˆpital Pitie´-Salpeˆtrie`re; and 5Department of Radiotherapy, Institut Curie, Paris, France

Given the generally poor outcome of advanced B cell chronic known radiosensitivity of CLL lymphocytes.13 The aim of the lymphocytic leukemia, experimental approaches are warranted, study was to determine whether an adequate response for especially for younger patients in whom classical treatments have failed. We therefore conducted a prospective single- hematopoietic stem cell collection could be obtained and to center study, using polychemotherapy (ESHAP) to prepare study the outcome of autologous transplantation in patients patients for hematopoietic stem cell collection and autologous with advanced CLL. Despite the limited number of patients, stem cell transplantation as consolidation therapy. Twenty this study shows that it can be impossible to collect adequate patients entered the study. An adequate response to ESHAP stem cells for autografting in multitreated patients, and argues was obtained in 13 patients, and sufficient stem cells for graft- in favor of stem cell harvest immediately after the first ing were obtained in eight of the 12 patients who underwent the collection procedure. Six of these grafted patients are alive remission is obtained. in complete clinical remission a median of 30 months after transplantation. It should be noted that we were only able to graft 40% of the patients enrolled in this study, either because a new remission could not be obtained or because not enough Patients and methods hematopoietic stem cells could be collected. This argues for stem cell collection as soon as a first remission is obtained, even if the autograft is done later in the course of the disease. Patient eligibility Keywords: B-CLL; advanced disease; autologous transplantation; hematopoietic stem cell B-CLL was diagnosed on the basis of classical criteria, using blood cell counts, cytology, immunophenotyping and bone Introduction marrow cell morphology. Patients with B-CLL were eligible for the study if they had received at least two different chemo- Hematopoietic stem cell transplantation has been widely used therapy regimens and were refractory to or had relapsed after in treating chronic lymphocytic leukemia (CLL), and is cur- the last treatment. Binet’s staging and new cytologic, pheno- rently one of the only approaches yielding immunophenotypic typic and bone marrow biopsy studies were performed at and molecular remission.1–3 However, the use of stem cell enrolment. Tumor burden was assessed by physical examin- transplantation as first-line treatment is debatable, as fludara- ation and by chest and abdominal-pelvic computed tomogra- bine and CHOP give 80% and 77% of responses, respectively, phy. From September 1992 to January 1997, all consecutive in previously untreated patients and a survival time tending patients aged up to 66 years, and those under 50 lacking an towards 10 years in patients entering clinical remission.4–7 HLA-identical sibling, were offered the protocol. Informed However, standard treatments are usually only palliative, and consent was obtained from all the patients. The treatment pro- curative treatments for younger patients are clearly needed.8 tocol was reviewed by the scientific committee of the French The prognosis is worse when the disease relapses or does not Cooperative Group on Chronic Lymphocytic Leukemia and respond to first-line therapies. Fludarabine only induces short- approved by the local Institutional Review Board. lived remission in such patients; in addition, the disease is rapidly life-threatening when it is resistant to this second-line therapy.4,8 Consequently, younger patients who relapse or are refractory after several lines of treatment need a salvage pro- Treatment plan cedure. We conducted a prospective study aimed at inducing remission with the combination of , , high- dose and high-dose methylprednisolone (ESHAP), Patients were first treated with ESHAP every 4 as originally proposed by Velasquez et al,9 as it has given weeks, combining intravenous infusions of methylpredniso- 2 interesting results in relapsing and refractory lymphoma. The lone (500 mg/day) for 5 days, etoposide (40 mg/m /day) and 2 need for myeloablative treatment for consolidation was sug- cisplatin (25 mg/m /day as a 24-h continuous infusion) for 4 2 gested by results in other lymphoproliferative disease.10–12 days, and cytarabine (2 g/m at the end of the last infusion of 9 Autologous hematopoietic stem cell transplantation (ASCT) cisplatin). The response to this treatment was studied after was thus planned for patients lacking an HLA-identical sibling the first three cycles and patients were moved towards hema- and those over 50 years. Total body irradiation was used as topoietic stem cell (HSC) collection procedures if they quali- a part of the myeloablative conditioning regimen, given the fied. Additional cycles of ESHAP could be performed to improve the response before collection, and also after collec- tion pending ASCT. Trimethoprim/sulphamethoxazole (160 mg/800 mg per day, three times a week) was adminis- Correspondence: L Sutton, Service d’He´matologie Clinique, Hoˆpital Pitie´-Salpeˆtrie`re, 47 boulevard de l’Hoˆpital, 75651 Paris cedex 13, tered from the start of the first ESHAP cycle through the ASCT France; Fax: 33 1 42 16 02 49 procedure as prophylaxis against Pneumocystis carinii Received 8 June 1998; accepted 6 August 1998 infection. ASCT in advanced B-CLL L Sutton et al 1700 Cell collection and processing 109 lymphocytes per liter of blood, and normal BM (except for a few nodules: CRnod). Phenotypic remission was defined Patients were eligible for HSC collection when they fulfilled when less than 5% of PBLy were CD5-CD19 double-positive the criteria of partial response (PR) and in addition had less in which case a molecular analysis of immunoglobulin gene than 4 × 109/1 peripheral blood lymphocytes (PBLy), irrespec- rearrangement was performed to detect minimal residual dis- tive of the PBLy phenotype and the bone marrow pattern of ease.1 The persistence of remission was assessed every 6 infiltration (see below).14 months or more frequently if clinically indicated. Hematologic Bone marrow (BM) and/or peripheral blood stem cells recovery after ASCT was based on daily blood counts, and (PBSC) were used as appropriate. BM was harvested from the defined as a neutrophil count of 0.5 × 109/1 on 2 consecutive iliac crests and treated in vitro with anti-CD19 and anti-CD20 days and a platelet count of 50 × 109/1 without transfusion. monoclonal antibodies plus rabbit complement.15 PBSC were The number of packed red cell and platelet transfusions was mobilized with cyclophosphamide (60 mg/kg/day intra- recorded.17 venously on 2 consecutive days) plus filgrastim (5 ␮g/kg/day) subcutaneously until neutrophil recovery and completion of harvesting; or with filgrastim plus molgramostim (5 ␮g/kg/day Histopathologic studies each); or with filgrastim alone (10 ␮g/kg/day subcutaneously for 6 days), with collection on days 5, 6 and 7.16 Mononuclear BM aspirates and posterior iliac crest biopsies were reviewed cells (MNC) were analyzed for CD19-positive and CD5-CD19 jointly at the Hematobiology and Hematopathology Depart- double-positive cells, CD34-positive cells and granulocyte– ments of Hoˆpital Pitie´-Salpeˆtrie`re. BM biopsies were exam- macrophage colony-forming units (CFU-GM) in the PBSC or ined for cellularity, lymphocyte percentages and the intertra- post-purge BM, and the cells were then cryopreserved.15,16 becular pattern of BM infiltration, which was noted as normal CD19-positive cells were also evaluated in pre-purge BM. (no evidence of lymphoproliferation), nodular (nodules of Patients were eligible for ASCT when CFU-GM and/or CD34- small mature lymphocytes lacking clear germinal centers), positive cell counts reached 5 × 104/kg and 1 × 106/kg, interstitial (replacement of normal hematopoietic tissue by respectively, in BM, or 15 × 104/kg and 2 × 106/kg, respect- small mature lymphocytes infiltrating through fat, with no dis- ively, in PBSC, in a single sample, as currently recommended tortion of the marrow architecture), mixed (both nodular and in our institution. interstitial involvement), or diffuse (disruption of marrow architecture by small mature lymphocytes).18,19

Preparatory regimen and HSC infusion Immunophenotyping Fractionated total body irradiation (TBI), (3.3 Gy every 24 h for 3 days), was first administered at a mean dose rate below Immunophenotyping was performed on peripheral blood, BM 0.25 Gy/min, using a linear accelerator; the total dose to the harvested before and after purging, and PBSC, by flow cyto- lungs was limited to 8 to 9 Gy. TBI was followed by cyclopho- metry with single or dual-color staining on a FACStar device sphamide (60 mg/kg/day for 2 days). After 2 days the cryopre- (Becton Dickinson, San Jose, CA, USA). The following mono- served HSC were rapidly thawed and infused via a central clonal antibodies were always used: CD19 (B4), CD20 (B1), venous line. Filgrastim (5 ␮g/kg/day) was administered from CD10 (J5), and CD23 (B6) from Coultronics (Margency, the day after graft reinfusion until neutrophil recovery. Patients France); and CD5 (OK23), anti-kappa, anti-lambda and anti- were treated in a laminar-flow protected environment until IgM from Dako (Trappes, France).20,21 Residual disease was granulocyte recovery. They received oral prophylactic anti- defined by coexpression of CD5 on CD19-positive B lympho- biotics for gut decontamination. All blood products were fil- cytes.22 When more than 5% of the total lymphocyte popu- tered and irradiated. Trimethoprim/sulfamethoxazole was lation coexpressed CD19 and CD5, residual disease was diag- interrupted at the outset of the conditioning regimen and nosed.1 Purging efficiency was evaluated by the percentage reintroduced 1 month after ASCT for 6 months. of CD19-positive cells before and after in vitro treatment. The percentage of contaminating CD5-CD19 double-positive cells and the quantity of CD34-positive cells among MNC were Response evaluation calculated before freezing.

Responses were scored according to the NCI Working Group for CLL criteria, which defined complete remission (CR), par- Immunoglobulin gene rearrangements tial response (PR), stable disease (SD) and progressive disease (PD), with the addition of the nodular remission (CRnod) Complementary determining region 3 (CDR3) PCR amplifi- group based on the presence of residual lymphoid nodules cation was first performed by using consensus framework 3 in BM, as described by Robertson et al.14,17,18 Patients were (FW3) and JH regions as 5Ј and 3Ј primers, respectively, as monitored at the Hematology Department of Hoˆpital Pitie´-Sal- previously described.22 PCR products from a normal poly- peˆtrie`re and the disease was staged after three cycles of clonal population of B lymphocytes should appear as a smear ESHAP, at HSC collection, at ASCT, and within 6 months after on the ethidium bromide-stained acrylamide gel (defined as the graft, by physical examination, chest and abdominal-pel- CDR3 molecular remission). Amplified DNA from cells con- vic computed tomography, complete blood counts and per- taining an abnormal B cell population formed a single tight ipheral blood lymphocyte (PBLy) counts, PBLy immunophen- band in the size range of 100–130 base pairs. This method is otyping, and bone marrow aspiration and biopsy. We thus capable of detecting tumor cell proportion exceeding 1%.22 assessed clinical complete remission (CR) as the absence of In addition, each patient’s CDR3 sequence was determined detectable disease, together with more than 110 g/l hemoglo- and a CDR3 clone-specific probe was synthesized. Using the bin, 1011 platelets, 1.5 × 109 neutrophils and fewer than 4 × patient’s previously determined leukemic cell immunoglob- ASCT in advanced B-CLL L Sutton et al 1701 ulin VH family as 5Ј primer and the clone-specific CDR3 cycle. UPN 9 died of disseminated herpes virus type 1 infec- probe in 3Ј, specific PCR amplification was performed (in case tion and disease progression at 4 months. UPN 10 had M. of a CDR3 molecular remission) and allowed the detection of tuberculosis meningitis that resolved on specific antibiotic about one clonal cell in 105 cells.23,24 A negative result therapy; HSC collection was possible but further treatments defined clone-specific molecular remission. had to be delayed and she died of disease progression at 12 months. The hematologic toxicity of chemotherapy was diffi- cult to distinguish from cytopenia related to the disease in Statistical methods resistant patients, but we observed a gradual recovery of blood counts over successive cycles in responding patients (data Survival data were updated on January 1998. Kaplan–Meier not shown). overall survival and relapse-free survival curves were calcu- lated from the dates of inclusion and ASCT, respectively.25 The end-point taken for relapse-free survival was the date of HSC collection and ASCT clinical relapse. BM was harvested once from seven patients and twice from two patients. Seven BM graft were purged (Table 3). The Results amount of CD34 cells or CFU-GM was too small for grafting in four cases (UPN 11, 16, 17 and 19). Further PBSC collec- Patients tions were only successful in UPN 11. UPN 17 underwent a total of six collections (two BM, four PBSC), all of which were From September 1992 to January 1997, 20 consecutive unsuccessful. Three patients (UPN 7, 10 and 20) had PBSC patients with repeatedly relapsing or refractory CLL were collection alone, which was successful in UPN 7 and UPN enroled in this prospective single-center trial. There were 15 20. UPN 16 and UPN 17 could not be grafted and were later male and five female patients. All the patients had CD5-posi- treated with the CPM/Famp combination when their disease tive B-CLL diagnosed on morphologic and phenotypic progressed.27 The percentage of contaminating CD5-CD19 grounds, except for UPN 4, whose leukemic cells were CD5- positive cells in the graft, the number of CD34 cells recovered, negative at diagnosis. The median time from diagnosis to and CFU-GM growth are shown in Table 3. In total, sufficient inclusion in the trial was 4.5 years (range 0.7 to 9 years). At stem cells for autografting were only obtained in eight of the inclusion the median age was 55 years (range 38 to 66 years). 12 patients who underwent the procedure. We were able to Bone marrow morphology showed a diffuse pattern of graft these eight patients, after a median of two more cycles involvement in 11 patients, mixed in four, interstitial in one of ESHAP (range 1–3). The median time from inclusion in the and nodular in four; 11 patients were in Binet’s stage B and trial to the ASCT procedure was 8 months (range 5.5–18 nine in Binet’s stage C, with a high tumor burden. All the months), only one patient being grafted after more than 9 patients had received at least two different regimens; all had months (Table 4). Four patients (UPN 1, 4, 13 and 14) were received a purine analog (mean 5.6 cycles, range 2–14) and grafted with purged BM; three patients (UPN 7, 11 and 20) 18 had received a doxorubicin-containing regimen (mean 7.2 received unselected PBSC; and UPN 18 received both purged cycles, range 2–12) (Table 1). and unpurged BM. We used the same conditioning regimen for the eight patients, and G-CSF (5 ␮g/kg) was routinely added on day 1 after ASCT. The median time required for Response to ESHAP (Table 2) peripheral blood neutrophil recovery was 15 days (range 11– 28 days). Platelet recovery took 40 days (range 18–60 days), Complete clinical responses (CR) and partial responses (PR) with the exception of UPN 4, who did not recover before were obtained in 10 and three patients, respectively, after a relapsing. Except for UPN 4, the patients received an average median of four cycles of ESHAP (range 3 to 6). However, in of 6.6 packed red cell transfusions (range 2–13) and eight plat- the 10 patients who entered complete remission after ESHAP, elet transfusions from single donors (range 6–16). None had the malignant clone remained detectable in both PCR life-threatening complications after ASCT. methods. One (UPN 6) of these 13 patients died of bleeding while in CR, 5 months after entering the study. Among the five patients who did not respond to ESHAP (UPN 3, 5, 9, 12 Outcome of ASCT and 15), four died of disease progression (associated with an infection in one) at 3, 4, 12 and 16 months, and one patient All eight grafted patients were in CR after ESHAP chemo- (UPN 12) is alive at 40 months after 4-Gy hemibody radio- therapy, ie before the graft. However, only three had a normal therapy followed by fludarabine (Famp) plus cyclophospham- BM morphology, one had a phenotypic remission (Ͻ5% CD5- ide (CPM).26,27 The last two patients (UPN 2 and UPN 8) were CD19 double-positive blood lymphocytes) and none had not assessable for the response because one died in aplasia CDR3 molecular remission (Tables 2 and 5). Within 6 months of intra-alveolar hemorrhage after the second cycle of ESHAP after ASCT, BM remission was obtained in all the patients (one and the other died of sepsis, at 2 months and 1 month, nodular CR), and all the patients but one (UPN 18) entered respectively. Four patients had severe infections: UPN 7, who phenotypic remission in blood (Table 4). Furthermore, three experienced severe and unusual cytopenia after each cycle of patients (UPN 7, 13 and 14) entered CDR3 molecular ESHAP, had Listeria monocytogenes meningitis occurring after remission. UPN 7 was also in clone-specific molecular the fourth cycle of ESHAP; it was cured by antibiotics but remission at 6 months, but not at 24 months. Another patient delayed the next two cycles; HSC collection and ASCT were (UPN 1) later entered both CDR3 and clone-specific molecu- delayed too at 14 and 18 months from inclusion, respectively. lar remissions, which persisted 42 months after ASCT. Two UPN 8, who responded to the first cycle of ESHAP, died at 1 patients (UPN 4 and 11) never entered CDR3 molecular month of Candida septicemia, in aplasia after the second remission and had a clinical relapse at 8 and 19 months, ASCT in advanced B-CLL L Sutton et al 1702 Table 1 Patient’s status at inclusion in the study

Patient (age years) Time from dg to Treatments before inclusion Binet’s stage BM involvement (% inclusion (years) (No of cycles) of Ly)

UPN 1 0.7 Famp (3) B nodular (58) CAP (2) (31) UPN 2 2 CAP (6) C diffuse (62) mCAP (5) (NE) Famp (3) UPN 3 9 Clb (6 y) C diffuse (66) CHOP (6) (NE) Famp (6) clb pdn (12) UPN 4 8 COP (11) B mixed (55) Clb (5y) (74) Famp (6) clb pdn (12) UPN 5 6 CHOP (6) C diffuse (45) Clb (3 y) (90) COP (3) Famp (3) UPN 6 6 Clb (2 y) C diffuse (54) Famp (12) (88) CAP (3) UPN 7 6 CHOP (12) C diffuse (56) MIME (5) (80) Famp (2) UPN 8 4.5 Clb (2 y) C mixed (59) COP (6) (NE) CAP (3) Famp (12) 2cda (2) UPN 9 5 Clb (3 y) C diffuse (55) CHOP (8) (100) Famp (3) 2cda (1) UPN 10 7 Clb (4 y) C diffuse (46) Famp (7) (99) Clb. pdn (5) 2cda (2) UPN 11 7.5 Clb (6 y) B mixed (56) Clb pdn (4) (58) CHOP (3) Famp (6) UPN 12 4 CHOP (6) B diffuse (66) Clb pdn (4) (88) Famp (6) UPN 13 8 CHOP (12) B interst (54) Famp (6) (20) UPN 14 4.5 CHOP (12) B diffuse (57) Famp (6) (76) UPN 15 1 CHOP (6) B diffuse (45) Famp (3) (91) UPN 16 3 CHOP (8) C nodular (48) Famp (6) (39) UPN 17 4.5 CAP (6) B nodular (51) mCAP (6) (33) Famp (6) UPN 18 8.5 2 cda (3) B nodular (63) CAP (6) (10) UPN 19 1 CHOP (6) B diffuse (50) Famp (4) (80) UPN 20 1 Famp (3) B mixed (38) CHOP (4) (NE) CHOP* (3)

BM, bone marrow; CAP, cyclophosphamide-doxorubicin (50 mg/m2) – prednisone; 2cda, 2-chloro-deoxy-adenosine; CHOP, cyclophos- phamide-doxorubicin (25 mg/m2) vincristine-prednisone; CHOP*, CHOP with doxorubicin (50 mg/m2); clb, chlorambucil; COP, CHOP with- out doxorubicin; dg, diagnosis; Famp, fludarabin; interst, interstitial; Ly, lymphocytes; mCAP, miniCAP (doxorubicine 25 mg/m2); MIME, guanylhydrazone dichlorohydrate-ifosfamide-methotrexate-etoposide; pdn, prednisone; NE, not evaluated. ASCT in advanced B-CLL L Sutton et al 1703 Table 2 Response to ESHAP treatment

Patient No. cycles to Clin status BM involvement Follow-up from Clinical status at follow-up response (% of Ly) inclusion (Ͼ, alive) (cause of death) (months)

UPN 1 3 CR none Ͼ63 CR (8) UPN 2 2 NE ND 2 NE (intra-alv Hrg) UPN 3 2 prog ND 3 prog (dis) UPN 4 4 CR nodular 18 Rel (dis) (11) UPN 5 9 prog diffuse 16 prog (dis) (100) UPN 6 5 CR nodular 5 CR (Hrg) (NE) UPN 7 6 CR none Ͼ53 CR (37) UPN 8 2 NE ND 1 NE (sepsis) UPN 9 4 prog ND 4 prog (sepsis) UPN 10 3 PR diffuse 12 prog (dis) (81) UPN 11 4 CR nodular Ͼ43 Rel (12) UPN 12 5 prog diffuse Ͼ40 prog (NE) UPN 13 3 CR nodular Ͼ37 CR (15) UPN 14 3 CR nodular Ͼ31 CR (20) UPN 15 4 prog diffuse 12 prog (dis) (92) UPN 16 5 PR nodular Ͼ30 prog (16) UPN 17 4 PR mixed Ͼ27 prog (NE) UPN 18 3 CR none Ͼ24 CR (4) UPN 19 5 CR nodular Ͼ17 CR (10) UPN 20 5 CR nodular Ͼ10 Rel (40) clin, clinical; CR, clinical complete remission; dis, disease; Hrg, hemorrhage; intra-alv Hrg, intra-alveolar hemorrhage; ND, not done; PR, partial remission; prog, progression; Rel, relapse. respectively, after the blood phenotype had again become results of ESHAP, the difficulty of collecting HSC in this series, positive. UPN 4 died at 11 months from progression, and UPN and the good results of autografting in general. 11, who entered a second CR with the Famp and CPM combi- We used ESHAP to induce remission in such patients.9 Our nation, was alive 35 months after ASCT.27 It is too early to results are comparable to those obtained with fludarabine by assess the molecular status of UPN 20. The median follow-up Keating et al,4 as we obtained a 65% response rate among of the 11 survivors in 36 months after inclusion in the trial. our 20 heavily pretreated patients. Four of the five non- The Kaplan–Meier (KM) overall survival rate is 53% ± 22% responders died of disease progression within 16 months, con- (95% CI), 36 months after inclusion in the study (Figure 1). firming the poor outcome of refractory CLL.6 In addition, The KM relapse-free survival rate is 69% ± 36%, 30 months another three patients (15%) died of toxicity (one in response after ASCT (Figure 2). and two before evaluation). All but one of the 20 patients had previously been treated with fludarabine and four (20%) had life-threatening opportunistic infections. We gave all our Discussion patients trimethoprim/sulfamethoxazole prophylaxis against Pneumocystis carinii and observed no cases of this infection. The lack of curative therapy for CLL, coupled with the poten- This marked toxicity was partly offset by the fact that these tially indolent course of the disease and the advanced age of patients had highly advanced disease and were heavily pre- many patients make symptom relief and prolongation of sur- treated. Forthcoming strategies using other drugs in combi- vival reasonable therapeutic goals in most patients requiring nation with fludarabine, or/and monoclonal antibodies against therapy. However, experimental approaches aimed at achiev- B cells (eg Campath-1H or anti-CD20) should increase the ing a cure are warranted for younger patients with poor-risk response rate in advanced B-CLL relative to ESHAP.8,30–32 factors.28,29 Moreover, patients in failure or relapse after first- HSC collection failed in four patients. This failure is prob- line therapy have a very short survival time, with a median of ably due to the cumulative marrow toxicity of previous treat- 29 or 9 months according to whether or not they respond to ments. The difficulty of collecting HSC after more than two a second treatment with fludarabine.4 Here we discuss the lines of treatment illustrates the need to perform the collection ASCT in advanced B-CLL L Sutton et al 1704 Table 3 HSC collection procedure when response to ESHAP

Patient Type of HSCC Mobilization procedure Ex vivo B cell % Frozen CD5 Frozen CFU-GM Frozen CD34 depletion CD19 cells ×104/kg ×106/kg

UPN 1 BM — + 0 27.5 1.3 UPN 4 BM — + 014 1 UPN 7 PBSC CPM + G.CSF 6 19.5 2.25 UPN 10 PBSC CPM + G.CSF 65 0.3 0.8 UPN 11 BM — + 2 4.6 ND PBSC G.CSF 30 45 1.5 UPN 13 BM — + 011 ND UPN 14 BM — + 0 14.7 1.5 UPN 16 BM — 12 3.5 ND PBSC G.CSF 4 1.6 0.4 UPN 17 BM — + 4 2.5 0.7 BM — 35 0.9 0.25 PBSC CPM + G.CSF 25 2.8 1 PBSC G.CSF + GM.CSF 22 4.2 1.3 PBSC G.CSF 30 1.25 0.4 UPN 18 BM — + 0 2 1.8 BM 9 1.5 1.8 UPN 19 BM — ND 1.4 ND PBSC CPM + G.CSF ND 3 0.9 UPN 20 PBSC CPM + G.CSF 1 28.5 5.6

CFU-GM, colony-forming unit granulocyte–macrophage; CPM, cyclophosphamide (120 mg/kg); G-CSF, granulocyte colony-stimulating- factor; GM-CSF, granulocyte–macrophage colony-stimulating-factor; HSCC, hematopoı¨etic stem cell collection; MNC, mononuclear cells; PBSC, peripheral blood stem cells.

Table 4 Follow-up post-ASCT

Patient Time from No. ESHAP from Status at 6 BM involvement PB Ly at 6 % db pos Ly Relapse Last follow-up inclusion to HSCC to ASCT months at 6 months (%) months CD5 CD19 (months) from ASCT ASCT (months) (CD19 CD23) (Ͼ, alive)

UPN 1 7 3 CR none (4) 0.7 3 (4) no Ͼ55 UPN 4 7 2 CR none (NE) 1.2 2 (1) 8 11 UPN 7 18 1 CR none (22) 1.8 2 (3) no Ͼ35 UPN 11 8 2 CR none (7) 1.9 4 (3) 19 Ͼ35 UPN 13 5.5 2 CR none (8) 1.4 1 (1) no Ͼ31 UPN 14 7 2 CR nodular (NE) 0.85 3 (3) no Ͼ28 UPN 18 8 3 CR none (10) 0.9 7 (4) no Ͼ16 UPN 20a 9 1 CR none (NE) 2.4 2.5 (2) no Ͼ2

aEvaluation at 2 months. ASCT, autologous stem cell transplantation; PB Ly, peripheral blood lymphocyte; db pos Ly, double positive lymphocyte.

in first remission and either to await disease progression any time after ASCT regularly led to relapse. Ten of the 11 before autografting, or to consolidate the good results of flud- patients reported by Khouri et al1 were in clinical CR 1 month arabine by immediate autografting. In our study only eight after ASCT. Only three remained in CR between 4 and 29 patients were able to undergo ASCT. As the median age of months; three died of Richter’s syndrome and two from tox- our patients was high, we chose to reduce the fractionated TBI icity; three were still alive but in relapse or partial response dose to 10 Gy in every case, principally to avoid pulmonary at 17, 15 and 2 months. We also discovered a case of Richter’s complications. All eight grafted patients were in clinical CR 6 syndrome on a post-mortem analysis of lymph nodes in the months after ASCT. Two relapsed and the other six remain in patient who relapsed 8 months after ASCT. It is important to CR a median of 29.5 months (range 2–55 months) after ASCT. note that no deaths were attributed to the autograft procedure Four are in CDR3 molecular remission and two entered clone- itself. This good tolerability confirms previous reports and specific molecular remission (Table 5). The two relapsing contrasts with the high rate of transplant-related mortality in patients never entered molecular remission. This suggests that allografting.33 UPN 18 is likely to relapse, as his PBLy phenotype remains A prospective study is required to determine the value of abnormal at 13 months. Other teams have reported larger ser- BM purging in vitro. In our series, the outcome of the patients ies of autografted patients with similar good results. Rabinowe seemed to be related more to the quality of the response et al2 obtained clinical CR at 3 months in 83% of 12 patients, before the graft than to the type of reinjected HSC. In the series which persisted for between 6 and 31 months in five cases. of Provan et al,3 all BM grafts were purged in vitro using a In Provan’s series of 21 patients,3 80% of the grafted patients panel of anti-B monoclonal antibodies and rabbit comp- remained in PCR-negative remission after between 7 months lement, but the post-autograft remission status did not depend and more than 5 years; the absence of molecular remission at on the efficiency of the purge, as assessed by immunoglobulin ASCT in advanced B-CLL L Sutton et al

Overall su.-vival 1705

Probahilit~· N = 20

0.9

0.8

0,7

0,6 ·

0,5

0,4

O.J .

0,2

0, 1 ·

12 18 24 JO 36 42 48 54 60 66 Time (111011ths)

Figure 1 Kaplan–Meier estimate of overall survival in the 20 patients, from inclusion in the trial.

Rela1•se-free survivnl Probal1illt~- N = 8

0,9 I ' 0,8

0,7 I I I '

0,6

0,5

0,4

O,J

0,2

0, 1

0 0 12 18 24 JO 36 42 48 60 Time (1111111l11 s)

Figure 2 Kaplan–Meier estimate of relapse-free survival in the eight autografted patients, from the date of transplantation. gene rearrangement PCR analysis. The study reported by nowe2 and Provan,3 but their patients seemed to have a Khouri et al1 does not support the use of BM purging either. shorter disease history before the graft. Anyway, ASCT is the The good results reported by Provan et al,3 which are con- only strategy (except for allogeneic marrow transplantation) firmed in our series, raise the possibility that both teams selec- giving a high rate of molecular remission in CLL. ted good-prognosis patients for ASCT, whose outcome would This is the only prospective study of ASCT in 20 consecutive have been favorable anyway. Effectively, in our study, seven patients with CLL based on an intention-to-treat analysis. It of the eight autografted patients remained in Binet’s stage B stresses the difficulty of HSC collection and ASCT in such at inclusion, as opposed to only four of the 12 non-grafted heavily pretreated patients, as only 40% underwent both pro- patients. However, irrespective of Binet’s status, all four cedures. As only stem cell grafting can yield molecular patients who remained alive after responding to ESHAP but remissions, HSC collection and ASCT should perhaps be done failing HSC collection had further disease progression, com- earlier in the disease. In our opinion, patients under 65 years pared to only two of the eight grafted patients. Furthermore, who have a first response to conventional treatment should the median follow-up of our autografted patients is more than undergo stem cell collection and should be randomized to 29 months, whereas heavily treated refractory patients usually ASCT or classical maintenance therapy in a prospective trial. die within a year.4 Comparable survival was reported by Rabi- In case of relapse or refractory disease after a first-line treat- ASCT in advanced B-CLL L Sutton et al 1706 Table 5 Peripheral blood biological remission study post-ASCT

patient : BM / :treated :PBSC : BM 55 ' UPNI 8 0u ' ~/ i ' UPN4 W "' "' '•; w d ' UPN7 '/I i D R CJ ' Ii, .~ UPNJJ /I i ' D • I I R UPNJJ ! o; ••D w u B UPNJ4 • : •; i •i UPNIB* ! •; i • • • r UPN20 /Di ro••

0 1 2 3 4 5 6 7 8 9 10 1112131415 1617 18 19 2021222324 2526 27 2629 3031323334 35 3637 3639 40 4142

mon1hs post ASCT

0 negative phenotyping • positive phenotyping r) negative CDR3 PCR e positive CDR3 PCR 6 negative clone specific PCR .a. positive clone specific PCR both treated and not BM were reinjected for this patient R, relapse d, death

ment, HSC collection, rapidly followed by ASCT should be Leukaemia in Adults. Prognostic factors in chronic lymphocytic proposed next to second-line chemotherapy. Such a trial leukaemia: the importance of age, sex and response to treatment could also assess the potential benefit of BM purging or non- in survival. Br J Haematol 1989; 72: 141–149. 6 Keating MJ, O’Brien S, Robertson L, Huh Y, Kantarjian H, Plunkett leukemic PBSC selection. W. Chronic lymphocytic leukemia – correlation of response and survival. Leuk Lymphoma 1993, 2: 167–175. 7 Leporrier M, Chevret S, Cazin B, Feugier P, Rapp MJ, Boudjerra Acknowledgements N, Autrand C, Dreyfus B, Desablens B, Dighiero G, Travade P, Chastang C, Binet JL. Randomised comparison of fludarabin, CAP We thank David Young for checking the English and Florence and ChOP, in 695 previously untreated stage B and C chronic lymphocytic leukemia. Early stopping of the CAP accrual. Blood Laffere for typing the manuscript. 1997; 90: 529a, Abstr. 2357. 8 Keating MJ, Lerner S, O’Brien S. Long-term results with fludarabine in CLL. VII International Workshop on CLL. Crete, Greece, 1997, References p 13, Abstr. 13. 9 Velasquez WS, McLaughlin P, Tucker S, Hagemeister FB, Swan F, 1 Khouri IF, Keating MJ, Vriesendorp HM, Reading CL, Przepiorka Rodriguez MA, Romaguera J, Rubinstein E, Cabanillas F. ESHAP – D, Huh YO, Andersson BS, Van Besien KW, Mehra RC, Giralt SA, an effective in refractory and relapsing Ippoliti C, Marshall M, Thomas MW, O’Brien S, Robertson LE, lymphoma: a 4-year follow-up study. J Clin Oncol 1994; 12: Deisseroth AB, Champlin RE. Autologous and allogeneic bone 1169–1176. marrow transplantation for chronic lymphocytic leukemia: pre- 10 Philips GL, Herzig TH, Lasarus HM, Fay W, Wolf SN, Mill WB, Lin liminary results. J Clin Oncol 1994; 12: 748–758. HN, Thomas PR, Glasgow GP, Shina DC, Herzig GP. Treatment of 2 Rabinowe SN, Soiffer RJ, Gribben JG, Daley H, Freedman AS, resistant malignant lymphoma with cyclophosphamide, total body Daley J, Pesek K, Neuberg D, Pinkus G, Leavitt PR, Spector NA, irradiation and transplantation of cryopreserved autologous mar- Grossbard ML, Anderson K, Robertson MJ, Mauch P, Chayt-Mar- row. New Engl J Med 1984; 310: 1557–1561. cus K, Ritz J, Nadler LM. Autologous and allogeneic bone marrow 11 Philip T, Armitage JO, Spitzer G, Chauvin F, Jagannath S, Cahn transplantation for poor prognosis patients with B cell chronic lym- JY, Colombat P, Goldstone AH, Gorin NC, Flesh M, Laporte JP, phocytic leukemia. Blood 1993; 82: 1366–1376. Maraninchi D, Pico J, Bosly A, Anderson C, Schots R, Biron P, 3 Provan D, Bartlett-Pandite L, Zwicky C, Neuberg D, Maddocks A, Cabanillas F, Dicke K. High-dose therapy and autologous bone Corradini P, Soiffer R, Ritz J, Nadler LM, Gribben JG. Eradication marrow transplantation after failure of conventional chemotherapy of polymerase chain reaction-detectable chronic lymphocytic leu- in adults with intermediate-grade or high-grade non-Hodgkin’s kemia cells is associated with improved outcome after bone mar- lymphoma. New Engl J Med 1987; 316: 1493–1498. row transplantation. Blood 1996; 88: 2228–2235. 12 Philip T, Chauvin F, Armitage J, Bron D, Hagenbeek A, Biron P, 4 Keating MJ, O’Brien S, Kantarjian H, Plunkett W, Estey E, Koller Spitzer G, Velasquez W, Weisenburger DD, Fernandez-Ranada J, C, Beran M, Freireich EJ. Long-term follow-up of patients with Somers R, Rizzoli V, Harousseau JL, Sotto JJ, Cahn JY, Guilhot F, chronic lymphocytic leukemia treated with fludarabine as a single Biggs J, Sonneveld P, Misset JL, Manna A, Jagannath S, Guglielmi agent. Blood 1993; 81: 2878–2884. G, Chevreau C, Delmer A, Santini G, Coiffier B. Parma inter- 5 Catovsky D, Fooks J, Richards S for the MRC Working Party on national protocol: pilot study of DHAP followed by involved-field ASCT in advanced B-CLL L Sutton et al 1707 radiotherapy and BEAC with autologous bone marrow transplan- and bone marrow remission using CD5–CD19 markers and PCR tation. Blood 1991; 77: 1587–1592. study of gene rearrangements. Leuk Lymphoma 1992; 7: 195–204. 13 Foon KA, Rai KR, Gale RP. Chronic lymphocytic leukemia: new 23 Maloum K, Pritsch O, Dighiero G. Minimal residual disease detec- insights into biology and therapy. Ann Intern Med 1990; 113: tion in B cell malignancies by assessing IgH rearrangement. Hem- 525–539. atol Cell Ther 1997; 39: 1–6. 14 Cheson BD, Bennett JM, Rai KR, Grever MR, Kay NE, Schiffer CA, 24 Maloum K, Magnac C, Sutton L, Leblond V, Cazin B, Boccaccio Oken MM, Keating MJ, Boldt DH, Kempin SJ, Foon KA. Guidelines C, Davi F, Binet JL, Merle-Be´ral H, Dighiero G. Minimal residual for clinical protocols for chronic lymphocytic leukemia: report of disease assessment in chronic lymphocytic leukemia. Blood 1996; the NCI-sponsored Working Group. Am J Hematol 1988; 29: 88: 77a, Abstr. 297. 152–163. 25 Kaplan E, Meier P. Nonparametric estimation from incomplete 15 Gribben JG, Freedman AS, Neuberg D, Roy DC, Blake KW, Woo observations. J Am Stat Assoc 1958; 53: 457–481. SD, Grossbard ML, Rabinowe SN, Coral F, Freeman GJ, Ritz J, 26 Leporrier M, Reman O, Troussard X, Levaltier X, Vie B. Double Nadler LM. Immunologic purging of marrow assessed by PCR hemibody irradiation with GM-CSF as salvage therapy for refrac- before autologous bone marrow transplantation for B cell lym- tory chronic lymphocytic leukemia. Leuk Lymphoma 1994; 16: phoma. New Engl J Med 1991; 325: 1525–1533. 121–124. 16 To LB, Haylock DN, Simmons PJ, Juttner CA. The biology and 27 O’Brien S, Kantarjian H, Beran M, Freireich E, Kornblau S, Koller clinical uses of blood stem cells. Blood 1997; 89: 2233–2258. C, Lerner S, Gilbreath J, Keating M. Fludarabine and cyclophos- 17 Cheson BD, Bennet JM, Grever M, Kay N, Keating MJ, O’Brien S, phamide therapy in chronic lymphocytic leukemia. Blood 1996; Rai KR. National Cancer Institute-sponsored Working Group 88: 480a, Abstr. 1910. Guidelines for Chronic Lymphocytic Leukemia: revised guidelines 28 Rozman C, Montserrat E. Chronic lymphocytic leukemia. New for diagnosis and treatment. Blood 1996; 87: 4990–4997. Engl J Med 1995; 333: 1052–1057. 18 Robertson LE, Huh YO, Butler JJ, Pugh WC, Hirsch-Ginsberg C, 29 Montserrat E, Gomis F, Vallespi T, Rios A, Romero A, Soler J, Stass S, Kantarjian H, Keating MJ. Response assessment in chronic Alcala A, Morey M, Ferran C, Diaz-Mediavilla J, Flores A, lymphocytic leukemia after fludarabine plus prednisone: clinical, Woessner S, Batlle J, Gonzalez-Aza C, Monserrat R, Reverter JC, pathologic, immunophenotypic and molecular analysis. Blood Rozman C. Presenting features and prognosis of chronic lympho- 1992; 80: 29–36. cytic leukemia in younger adults. Blood 1991; 78: 1545–1551. 19 Rozman C, Montserrat E, Rodriguez-Fernandez JM, Ayats R, Valle- 30 O¨ sterborg A, Fassas AS, Anagnostopoulos A, Dyer MJS, Catovsky spi T, Parody R, Rios A, Prados D, Morey M, Gomis F, Alcala A, D, Mellstedt H. Humanized CD52 monoclonal antibody Cam- Gutierrez M, Maldonado J, Gonzalez C, Giralt M, Hernandez- path-1H as first-line treatment in chronic lymphocytic leukaemia. Nieto L, Cabrera A, Fernandez-Ranada JM. Bone marrow histol- Br J Haematol 1996; 93: 151–153. ogic pattern – the best single prognostic parameter in chronic lym- 31 Bowen AL, Zomas A, Emmett E, Matutes E, Dyer MJS, Catovsky phocytic leukemia: a multivariate survival analysis of 329 cases. D. Subcutaneous Campath-1H in fludarabine-resistant/relapsed Blood 1984; 64: 642–648. chronic lymphocytic leukaemia and B-prolymphocytic leukemia. 20 Bennet JM, Catovsky D, Daniel MT, Flandrin G, Galton DAG, Br J Haematol 1997; 96: 617–619. Gralnick HR, Sultan C. Proposals for the classification of chronic 32 Maloney DG, Grillo-Lopez AJ, White CA, Bodkin D, Schilder RJ, (mature) B and T lymphoid leukemias. J Clin Pathol 1989; 42: Neidhart JA, Janakiraman N, Foon KA, Liles TM, Dallaire BK, Wey 567–584. K, Royston I, Davis T, Levy R. IDEC-C2B8 (Rituximab) anti-CD20 21 Nadler LM. B cell leukemia panel workshop: summary and com- monoclonal antibody therapy in patients with relapsed low-grade ments. In: Reinherz EL, Haynes BF, Nadler LM, Bernstein ID (eds). non-Hodgkin’s lymphoma. Blood 1997; 90: 2188–2195. Leukocyte Typing II. Springer-Verlag: New York, NY, 1986, pp 33 Michallet M, Archimbaud E, Bandini G, Rowlings PA, Deeg HJ, 3–43. Gahrton G, Montserrat E, Rozman C, Gratwohl A, Gale RP. HLA- 22 Vuillier F, Claisse JF, Vandenvelde C, Travade P, Magnac C, Chev- identical sibling bone marrow transplantation in younger patients ret S, Desablens B, Binet JL, Dighiero G. Evaluation of residual with chronic lymphocytic leukemia. Ann Intern Med 1996; 124: disease in B cell chronic lymphocytic leukemia patients in clinical 311–315.