Purging in BCR-ABL-positive acute lymphoblastic leukemia ... - Nature

Bone Marrow Transplantation (2000) 25, 97–104  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt

Purging in BCR-ABL-positive acute lymphoblastic leukemia using immunomagnetic beads: comparison of residual leukemia and purging efficiency in bone marrow vs peripheral blood stem cells by semiquantitative polymerase chain reaction J Atta1, F Fauth1, M Keyser2, E Petershofen2, C Weber1, G Lippok1, D Hoelzer1 and H Martin1 1

Department of Hematology and Oncology and 2Institute of Transfusion Medicine, Johann Wolfgang Goethe-University, Frankfurt/Main, Germany

Summary: Twenty autologous bone marrow (BM) and 25 peripheral blood stem cell (PBSC) grafts were collected from a total of 40 consecutive patients with BCR-ABL+ acute lymphoblastic leukemia (ALL) in first (n = 37) or second (n = 3) complete morphological remission and subsequently purged with a cocktail of anti-CD19, -CD10, AB4 MoAbs and immunomagnetic beads (IMB). Residual BCR-ABL-positive cells before purging were detected in 19 of 20 BM grafts at a median of 4 (range 0–6) logs and in 17 of 25 evaluable PBSC grafts at a median of 1 (range 0–3) log above the limit of detection assessed by a semiquantitative limiting log10-dilution RT-PCR (P ⬍ 0.0001). IMB purging depleted a median of 2.5 (range 1–4) log of residual BCR-ABL+ cells from BM and a median of 1 (range 0–2) log from PBSC grafts, achieving RT-PCR negativity in 1/20 BM and 12/25 PBSC grafts after purging. Cell recoveries were 62% and 86% (P ⬍ 0.0001) of MNC and 74% and 97% (P = 0.065) of CD34+ cells after BM and PBSC purging, respectively. BM purging was superior using the triple MoAb cocktail which depleted 2.64 ⴞ 0.4 log (n = 14) compared to 1.6 ⴞ 0.4 log (n = 5) using the MoAb cocktail not including AB4 (P = 0.02). We conclude that unpurged BM grafts contain 2–3 log more residual BCR-ABL+ cells than unpurged PBSC grafts and that purging efficacy is superior in BM compared to PBSC grafts, but median titers in purged BM grafts still exceed those in purged PBSC grafts. Bone Marrow Transplantation (2000) 25, 97–104. Keywords: acute lymphoblastic leukemia; BCR-ABL rearrangement; RT-PCR; immunomagnetic beads; minimal residual disease; purging

The t(9;22) Philadelphia translocation and/or the corresponding BCR-ABL rearrangement are found in about 20–25% of adult patients with acute lymphoblastic leukemia.1–5 Correspondence: Dr H Martin, Department of Hematology and Oncology, Johann Wolfgang Goethe University, Theodor-Stern-Kai 7, D-60590 Frankfurt/Main, Germany Received 8 April 1999; accepted 5 August 1999

Conventional chemotherapy induces complete morphological remissions in 60–70% of patients, but these are rarely, if ever, durable.5 Thus, bone marrow or stem cell transplantation is usually essential to prevent fatal relapse. Allogeneic bone marrow transplantation from an HLA-matched family donor cures about 30–40% of patients.6–9 After transplantation from matched unrelated donors (MUD) long-term disease-free survival was 49% in a series of 18 patients with a median age of 25 years.10 However, the median age of patients with BCR-ABL-positive ALL in the German multicenter adult ALL (GMALL)5 and other large multicenter studies is 45 years. Thus, for older patients without a matched sibling donor, autologous transplantation remains an option for cure with a disease-free survival at 3 years of 20–30%.11 Relapse rates of 65–70% account for the majority of treatment failures after autologous transplants. The higher incidence of post-transplant relapse after autologous compared to allogeneic transplantation is likely due to the lack of a graft-versus-leukemia effect but possibly also to reinfusion of leukemic cells with the graft12 or a combination of the two. It is difficult to distinguish lack of GVL from reinfusion of residual leukemia in contributing to post-transplant relapse. As a first approach we characterized residual leukemia in the grafts and studied the feasibility and the depletion efficiency of available purging methods.13,14 We previously demonstrated that marrow harvested in morphological remission from 14 patients with BCR-ABLpositive ALL was consistently contaminated with BCRABL-positive cells,14 providing a rationale for subsequent in vitro purging using anti-CD10, -CD19 and AB-4 antibodies and immunomagnetic beads. Purging depleted a median of 3 log of BCR-ABL-positive cells without achieving complete PCR negativity. In contrast, leukapheresisderived peripheral blood stem cell grafts from four patients with BCR-ABL-positive ALL revealed 2 to 3 log less contamination with residual BCR-ABL-positive cells and became PCR negative after immunomagnetic bead purging.13 The objectives of the present study were first to investigate the difference in leukemic contamination between bone marrow and peripheral blood autografts in a larger prospective series comprising 40 consecutive BCR-ABLpositive ALL patients, second to compare the efficiency of immunomagnetic bead purging of bone marrow vs periph-

MRD and purging in BCR-ABL-positive ALL J Atta et al

98

eral blood grafts, and third, to define factors with potential impact on purging efficiency.

Patients and methods Patients A total of 40 patients with BCR-ABL-positive ALL without an HLA-identical sibling donor were included in this study. They were referred to our institution for autologous bone marrow (n = 20) or peripheral blood stem cell harvests (n = 20) with subsequent IMB purging from 25 different hospitals from June 1992 to February 1998. The median age at the time of diagnosis was 40 (range 6–63) years. The patients’ characteristics are detailed in Table 1. The BCR-ABL rearrangement at diagnosis was routinely assessed in a central laboratory of the German multicenter adult ALL (GMALL) study.15,16 Thirty-seven of 40 patients were recruited in first CR and three patients in second CR. At the time of harvest all patients were in complete remission by cytological criteria. During the initial phase of the study between June 1992 and June 1994, only autologous BM harvests were performed (n = 14).14 The subsequent protocol including PBSC harvest was initiated in 1994.13 Since then, patients mobilizing sufficient numbers of CD34+ cells were leukapheresed (n = 20) while patients with unsatisfactory CD34 mobilization (⬍10 CD34+ cells/␮l PB) remained candidates for BM harvest (n = 6). Written informed consent was obtained from each patient. Chemotherapy prior to stem cell harvest Six of 40 patients (patients 1–5, 22), diagnosed prior to 1994, received induction and consolidation therapies according to the German multicenter adult ALL (GMALL) 04/89 protocol,4 another 32 patients were treated according to the subsequent GMALL 05/93 protocol.5 The contents of these protocols5 and the relevant cycles for this study13,14 have been described in detail elsewhere. One single infant patient received the pediatric COALL-92 protocol,17 one patient was treated according to an Austrian adult ALL protocol.18 Table 1 details the chemotherapy protocol, the time from diagnosis to harvest and the interval from completing the preceding cycle of chemotherapy for each individual patient. Harvest and processing of bone marrow (BM) Seven patients (patients 1–4, 15, 18, 19) received prestimulation with 300 ␮g G-CSF s.c. for 5 consecutive days prior to collection of bone marrow. Bone marrow cells were harvested under general anesthesia by multiple aspirations from the posterior iliac crest on both sides. Mononuclear cells (MNC) were separated over a Ficoll density gradient using a COBE 2991 (COBE, Lakewood, CA, USA) blood cell washer. The MNC fraction was washed twice in normal saline supplemented with 2% human serum albumin and resuspended for purging procedures in 100 ml S-MEM containing 50% AB0-compatible fresh frozen plasma. Bone Marrow Transplantation

Harvest and processing of peripheral blood stem cells (PBSC) Patients scheduled for peripheral blood stem cell harvest were mobilized with 5–10 ␮g/kg rhG-CSF (Neupogen, Amgen, Thousand Oaks, CA, USA), procedures beginning 1–5 days after the last day of chemotherapy prior to leukapheresis. Blood counts and monitoring of CD34+ cells were performed daily as described previously.13 Leukaphereses were performed using a Fenwal CS 3000 plus (Baxter, Munich, Germany) with the Baxter stem cell collection program. Ten to 16 l of blood were processed at a flow rate of 50–70 ml/min. The apheresis product was collected in a volume of 50 ml. Immunomagnetic bead purging Dynabeads (Dynal, Hamburg, Germany) and the Baxter MaxSep Magnetic cell separator (Baxter, Deerfield, IL, USA) were used for IMB purging in all patients. From 1992 to 1996, a triple cocktail of directly coupled Dynabeads as detailed below was used in the first 15 of a total of 20 BM patients and the first five of a total of 20 PBSC patients. From 1996 directly coupled CD10- and AB4-Dynabeads were no longer available and the purging protocol had to be modified. Thus, the purging methodology was not homogeneous throughout this study. According to the available stock of beads, a double cocktail of direct CD19 and CD10 Dynabeads was used for five purging procedures in another three PBSC patients as detailed in Table 1. Only direct CD19 beads were used in another two BM and 10 PBSC purging procedures. Another three BM and two PBSC grafts were purged indirectly using a cocktail of CD19 and CD10 MoAbs. Purging was always completed on the day of the stem cell harvest. Direct immunomagnetic bead purging For direct IMB purging, mouse IgM monoclonal antibodies (MoAb) were directly coupled to Dynabeads, and were supplied in 10 ml vials each containing approximately 4 × 109 MoAb beads (Dynal). Three different MoAbs were used: an IgM-anti-CD19 (clone AB119), an IgM-anti-CD1020 and AB4, an anti-MHC class II IgM-MoAb.21,22 The purging procedure was described in detail previously.14 Briefly, the absolute numbers of B cells in the leukapheresis product or in the MNC fraction of the bone marrow, respectively, were estimated by FACS analysis before purging. A ratio of 40 beads per target cell (=B cell) was aimed at for the first purging cycle. Incubation time was 30 min. After incubation, the bound cell/bead rosettes as well as the excess beads were removed using the Baxter MaxSep with the primary magnet being precooled to 4°C. The bead purging procedure was repeated once (second cycle). Indirect immunomagnetic bead purging Primary mouse IgG MoAbs and sheep anti-mouse (SAM) Dynabeads for indirect immunomagnetic bead purging were purchased at GMP quality from Baxter (Munich, Germany). Cells were first incubated for 30 min with a

F M F M M M M F M F M M M M M M M F F M

49 44 45 22 56 6 18 34 58 33 37 37 25 49 52 48 38 60 43 55

31 55 55 36 34 32 55 30 21 57 43 37 53 39 34 63 25 40 57 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

3400 20 000 285 500 122 700 2000 1300 57 900 22 000 6600 23000 21 900 53 800 10 300 NA 6400 7100 NA 30 000 94 800 305 000

37 200 9 890 18 400 33 000 14 300 NA 34 800 22 000 67 700 2900 6000 128 000 26 100 184 000 29 770 NA 189 000 28 000 51 000 15 500

WBC/␮l

p190 p210 p190 p190 p210 p190 p210 p190 p190 p210 p190 p190 p190 p190 p190 p190 p190 p190 p210 p210

p190 p190 p190 p210 p210 p210 p210/p190 p190 p210 p190 p190 p190 p190 p190 p190 p190 p190 p190 p190 p190

BCR-ABL

At diagnosis

05/93 04/89 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93 05/93

04/89 04/89 04/89 04/89 04/89 COALL-92 05/93 05/93 05/93 05/93 05/93 05/93 05/93 Austrian ALL 05/93 05/93 05/93 05/93 05/93 05/93

GMALL protocol

CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR2 CR1 CR1 CR1

CR1 CR1 CR1 CR1 CR2 CR1 CR1 CR1 CR1 CR1 CR1 CR2 CR1 CR1 CR1 CR1 CR1 CR1 CR1 CR1

Status at harvest

157 169 69 73 70 84 79 76 142 81 78 85 84 69 67 76 1282 84 153 67

169 141 189 215 460 210 211 97 71 105 463 135 78 155 87 117 117 205 149 132

Diagnosis

NA 21 15 19 22 24 20 18 12 17 17 21 18 16 17 22 11 17 13 24

38 39 45 45 69 NA 35 39 35 39 NA 52 33 40 26 19 21 20 29 22

Last chemo

Days to harvest from

PB PB PB PB PB PB PB PB PB PB PB PB PB PB PB PB PB PB PB PB

BM BM BM BM BM BM BM BM BM BM BM BM BM BM BM BM BM BM BM BM

Source

D D D D D D D D D D D D D D D D D D IND IND

D D D D D D D D D D D D D D D D D IND IND IND

Purging

+ + + + + + + + + + + + + + + − − + + +

+ + + + + + + + − − − − − − − − − − + +

+ + + + + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + + + + +

+ + + + + − − − − − − − − − − − − − − −

+ + + + + + + + + + + + + + + − − − − −

CD19 CD10 AB4

MoAbs

1 0 0 1 1 2 neg 1 2 neg neg 1 1 neg 2 1 2 3 2 2

5 6 3 4 4 4 3 5 4 4 3 3 3 neg 5 2 6 3 5 4

Pre

Dx = date of diagnosis; D = direct purging; IND = indirect purging; NA = not available; NE = not evaluable; GMALL = German multicenter ALL.

M F F M M M F F M M F M F M M M F F M M

Sex

Patient characteristics, purging modalities and RT-PCR titers

UPN Age at Dx

Table 1

4 4 2 2 3 2 3 2 3 3 2 3 2 NE 2 1 3 2 1 1

Depletion

neg neg neg neg 1 1 neg 0 0 neg neg 0 0 neg 2 1 1 1 1 1

⭓2 ⭓1 ⭓1 ⭓2 0 1 NE 1 2 NE NE 1 1 NE 0 0 1 2 1 1

1st leukapheresis

1 2 1 2 1 2 0 3 1 1 1 0 1 neg 3 1 3 1 4 3

Post

NE NE 0 NE NE

neg neg neg neg 1 1 neg neg neg neg

2nd leukapheresis

Pre Post Depletion

Log RT-PCR titer


99

Bone Marrow Transplantation


100

cocktail of 1 ml of anti-CD10 (Baxter code R9901) as well as anti-CD19 (clone HD237; Baxter code R9909) mouse IgG MoAbs as primary antibodies. After washing out unbound primary antibodies, the cell–antibody complexes were incubated with SAM Dynabeads (Baxter code R9950). The estimation of beads per target cell as well as the conditions for the purging procedure were as described above. Purged stem cells were concentrated and cryopreserved in S-MEM + 50% AB0-compatible fresh frozen plasma + 10% (final concentration) DMSO using a computer-controlled cryopreservation device (Cryoson BV-6; Cryoson, Schoellkrippen, Germany).

ditions for PCR performance were 95°–56°–72°C for MBCR-ABL-primers, and 95°–60°–72°C for m-BCR-ABLprimers (denaturation/annealing/extension). To further enhance the sensitivity of detection amplified DNA samples were reamplified using BCR-ABL-specific nested primers with cycling conditions as before. Simultaneous amplification of normal c-abl sequences served as a standard for the quality of transcripts amplified and to demonstrate the presence of amplifiable DNA in those samples yielding no signal specific for the BCR-ABL fusion. Amplification products were run on an ethidium bromide stained 2% agarose gel. The results were expressed as log10 of the highest dilution with a positive RT-PCR signal.

Flow cytometric analysis CD34+ cells and CD19+ cells were enumerated by flow cytometry as described in detail previously.13 CFU-GM assay CFU-GM colony numbers were determined using a semisolid methylcellulose assay as described previously.14 Briefly, MNC were plated at 0.5–1 × 105 MNC/ml in quadruplicate 1 ml aliquots containing 1.2% methylcellulose in IMDM, 30% FCS (HyClone, Logan, UT, USA), 5 × 10−5 m 2-mercaptoethanol (2-ME), 10 ng/ml recombinant human GM-CSF (Behring, Marburg, Germany) 10 ng/ml recombinant human G-CSF (Amgen), and incubated at 37°C, in a fully humidified 5% CO2 atmosphere. Colonies of more than 50 cells were enumerated after 14 days and scored as CFU-GM. Semiquantitative RT-PCR Limiting log10 dilution: Mononuclear cells from unpurged as well as purged samples were washed twice in PBS, and subsequently diluted in logarithmic steps in a series of Eppendorf tubes (2 ml) each containing 1.8 × 107 MNC from healthy volunteer blood donors down to a 1:108 dilution.14 Two BCR-ABL-positive cell lines, BV-173 (MBCR-ABL-positive)23 and SD-1 (m-BCR-ABL-positive)24 were used as positive controls and for standard logarithmic dilutions. Reverse transcriptase polymerase chain reaction (RTPCR): Total RNA was prepared from each tube using the acid guanidinium/phenol/chloroform method.25 Five ␮g RNA were subjected to random-primed reverse transcription using 200 units of Moloney murine leukemia virus reverse transcriptase (GIBCO/BRL, Grand Island, NY, USA) in 20 ␮l 10 × PCR buffer (Perkin Elmer, Emeryville, CA, USA), and 0.4 mmol/l of each desoxyribonucleotide, primed by 1500 pmol/␮l random hexamers. Ubiquitous RNAses were inhibited by 1.8 U/␮l RNA-Guard (Pharmacia, Uppsala, Sweden). PCR was performed using a Perkin Elmer Cetus Thermal Cycler in 50 ␮l 10 × PCR buffer containing 500 mm KCl, 100 mm Tris-HCl, and 15 mm MgCl2, pH 8.3, and 1 unit of Taq DNA Polymerase (Perkin Elmer) primed by 2.5 mmol/ml oligomers specific for detecting 5⬘ and 3⬘ sequences of the M-BCR-ABL and m-BCR-ABL-rearranged RNA transcripts.3 Cycling conBone Marrow Transplantation

Estimation of total BCR-ABL+ cell counts in BM and PBSC grafts It was not possible to determine the sensitivity of the nested RT-PCR with leukemic cells from each individual patient. Thus, percentages and absolute counts of BCR-ABL+ cells were estimated indirectly on the basis of the median RTPCR sensitivity as determined in dilutions of BM samples from patients at relapse and of the cells lines BV-173 and SD-1. The median sensitivity was 1 BCR-ABL+ cell in 106 normal cells.13,14 The total counts of BCR-ABL+ cells per graft was estimated as follows: BCR-ABL+ cellsestimated = TNC × highest dilution with positive PCR signal/106. Statistical analysis Descriptive statistics, data comparison with the two-tailed Student’s t-test and linear regression analysis were performed and plotted using the GraphPad Prism software package.

Results Yield of nucleated cells, CFU-GM and CD34+ cells in autologous BM and PBSC grafts The yields of total nucleated cells (TNC) were a median of 16 × 109 (range 0.5–48 × 109) in 25 PBSC grafts from 20 patients and 24 × 109 (range 1.4–63 × 109) TNC in 20 BM harvests, respectively. Total cell numbers in the MNC fraction of BM grafts after Ficoll separation were a median of 4 × 109 (range 0.7–25.9 × 109). The median total numbers of CD34+ cells and CFU-GM were 1.9 × 107 and 1.2 × 107 in BM, and 3.8 × 108 and 4.9 × 107 in PBSC, respectively (Figure 1). Cell recoveries after purging of BM vs PBSC grafts After bone marrow purging, medians of 62% (range 46– 97%) of MNC, 74% (range 14–128%) of CD34+ cells and 61% (range 11–106%) of CFU-GM were recovered. Median cell recoveries after PBSC purging were 86% (range 52–114%) for TNC, 97% (range 22–130%) for CD34+ cells, and 118% (range 26–193%) for CFU-GM, significantly exceeding recoveries after BM purging (P ⬍


CD34+

CFU-GM

Cells per graft

109

108

107

106

105

PB BM P = 0.0005

PB BM P = 0.0006

1010

CD19+

PB BM P = 0.0007

+

+

Figure 1 Absolute numbers of CD34 cells, CFU-GM and CD19 cells in peripheral blood and bone marrow grafts.

0.0001 for NC, P = 0.065 for CD34+ cells, P ⬍ 0.0001 for CFU-GM). Quantification of CD19+ cells in BM and PBSC grafts by flow cytometry The percentage of CD19+ cells in Ficolled, but unpurged BM grafts was a median of 1.9% (range 0.04–12.6%) of MNC, corresponding to absolute counts of median 4.1 × 107 (range 2.4 × 106 to 4.8 × 108) CD19+ cells per BM graft, whereas the percentage of CD19+ cells in PBSC grafts was a median of 0.04% (range 0.0–0.9%) of MNC, yielding significantly lower absolute numbers of median 5.4 × 106 (range 0–1.4 × 108) B cells per PBSC graft (P ⬍ 0.001). Interval between chemotherapy and BM vs PBSC harvest Autologous bone marrow was harvested in 20 patients at a median of 36 (range 19–69) days after completing the previous cycle of chemotherapy. In contrast, the intervals between completing the corresponding previous cycle of chemotherapy and PBSC leukapheresis were significantly shorter: 25 PBSC were harvested in 20 patients after a median of 18 (range 11–24) days. Residual BCR-ABL-positive cells in unpurged BM grafts Residual BCR-ABL-positive cells were consistently detected in all except one of 20 autologous BM grafts tested. By semiquantitative RT-PCR, the median prepurge titer in the Ficolled MNC fraction was 4 (range 2–6) log above the limit of detection. The median estimated proportions of BCR-ABL+ cells were 1% of Ficolled MNC and 0.2% of unseparated TNC, respectively. Total BCRABL+ cells in unpurged BM grafts were estimated as median 2.9 × 107 (range ⬍103 to 2.5 × 109) (Figure 2). Residual BCR-ABL-positive cells in unpurged PBSC grafts In contrast to BM grafts, eight of 25 (32%) unpurged PBSC grafts were already BCR-ABL negative. In 17 of 25 PBSC grafts, residual BCR-ABL-positive cells were detectable,

Estimated total BCR-ABL+ cells

1010

101

P < 0.0001

109

P < 0.0086

108 107 106 105 104

neg PB BM Before purging

PB BM After purging

Figure 2 Estimated total numbers of BCR-ABL+ cells in bone marrow (BM) and peripheral blood (PB) autografts before and after purging with two cycles of immunomagnetic beads. The dots below the dotted line represent grafts with negative RT-PCR results.

but at a significantly lower level as compared to autologous BM grafts. The median RT-PCR-titer was 1 (range 0–3) log above the limit of detection. The estimated proportion of BCR-ABL+ cells was a median of 0.001% (range ⬍0.0001–0.01%) of total nucleated cells. The estimated total numbers of BCR-ABL+ cells were a median of 1 × 105 (range ⬍5 × 103 to 2 × 107) in unpurged PBSC grafts (Figure 2). The difference between RT-PCR titers in BM and PBSC is highly significant (P ⬍ 0.0001). Depletion of residual BCR-ABL-positive cells assessed by semiquantitative RT-PCR Immunomagnetic bead purging depleted median 2.3 log (range 1–4 log, n = 19) of BCR-ABL-positive cells from autologous BM grafts without achieving PCR negativity in any of the purged BM grafts with detectable signal prior to purging. The median PCR titer after BM purging was 1 log (range 0–4 log) above the limit of detection. The estimated total numbers of BCR-ABL+ cells in purged BM grafts were a median of 6 × 104 (range ⬍1 × 103 to 2 × 107) (Figure 2). Four of 17 PBSC grafts with a detectable prepurge signal became PCR negative by purging. Including those grafts which were already PCR negative prior to purging (n = 8), a total of 12 of 25 purged PBSC grafts (48%) were PCR negative. Thirteen of purged PBSC grafts (52%) remained PCR positive at a median level of 1 log above the limit of detection. In these grafts, the factor of depletion was a median of 1 log (range 0–⭓2 log). The estimated total numbers of BCR-ABL+ cells in purged PBSC grafts were a median of 6 × 103 (range ⬍1 × 103 to 7 × 105) (Figure 2). Purging efficiency in bone marrow compared to PBSC grafts The depletion achieved in BM grafts significantly exceeded the depletion achieved in PBSC grafts (median 2.3 log vs median 1 log, P ⬍ 0.001). Nevertheless, due to the median 3 log higher prepurge titers in BM, the median postpurge Bone Marrow Transplantation


102

RT-PCR titer in purged BM grafts still exceeded the median titer in postpurge PBSC grafts (median 1 log vs median 0 log above the limit of detection, P ⬍ 0.0001). Variables influencing efficiency of immunomagnetic bead purging Since the purging methodology was not homogeneous throughout this study, the following parameters in BMpurging were analyzed. Trends over time: A linear regression of RT-PCR titers over consecutive patient numbers did not show a significant drift with time in median prepurge PCR titers (P = 0.47) (see Figure 3). However, depletion efficiency deteriorated from 3 to 1.5 log (P = 0.006) resulting in an increase of postpurge RT-PCR titers (P = 0.2). MoAb cocktail: For 15 patients, bone marrow was purged with a triple MoAb cocktail including anti-CD19, -CD10 and -AB4 antibodies, whereas in a heterogenous group of five BM patients AB4 was not included. The purging efficiency in the AB4+ group was significantly superior compared to the AB4− group (2.64 ⫾ 0.4 log vs 1.6 ⫾ 0.4 log; P = 0.02), explaining the deterioration over time. Other variables such as total bead numbers used per purging cycle or bead numbers in relation to MNC, CD19+ target cells or incubation volume did not significantly influence postpurge PCR titers (data not shown). Discussion In this prospective study we compared the contamination with residual BCR-ABL-positive leukemic cells in 25 peripheral blood and 20 bone marrow-derived autologous grafts collected from a total of 40 patients with BCR-ABLpositive ALL in morphological remission. 6 5

Log RT-PCR titer

BM before purging (P = 0.47) 4

Depletion (P = 0.006)

3 2 1

BM after purging (P = 0.2) 0

neg 92

93

94

95

96

97

Year Figure 3 Linear regression lines of log RT-PCR titers over time in BMharvested patients. Linear regression parameters in BM titers before purging are: slope = −0.11 ± 0.15 per year, r2 = 0.03, P = 0.47, indicating consistency of BM titers before purging over time; in BM titers after purging: slope = 0.20 ± 0.15 per year, r2 = 0.1, P = 0.2; in log-depletion; slope = −0.32 ± 0.1 per year, r2 = 0.36, P = 0.006, indicating a significant decline in purging efficiency over time. Bone Marrow Transplantation

We previously found median 1.5 log lower RT-PCR titers in unmobilized peripheral blood as compared to bone marrow in nine paired samples from remission patients26 and hypothesized that PBSC grafts might be superior to BM grafts due to a significantly lower level of residual leukemia. In another small series of four G-CSF-mobilized leukapheresis samples there was a trend for even lower RTPCR titers.13 These findings are substantiated to high significance by the larger series of patients presented here. The median 3 log lower RT-PCR titers in PBSC grafts were most likely not only due to the different sources of stem cells, but also to the shorter interval from previous chemotherapy to leukapheresis of a median of 18 days compared to a median of 36 days to marrow collection. Referring to a difference of about 1.5 log RT-PCR titer in paired unmobilized blood and marrow samples,26 we estimate that earlier collection of PBSC may account for another 1.5 log. Thus we extrapolate a 1 log increase in peripheral RT-PCR titers within a 2 week period, which is in accordance with the clinical experience of early relapses seen in this disease. A point of theoretical concern was that leukemic cells were also mobilized by G-CSF27,28 in a condition highly efficient in mobilizing normal CD34+ cells. No data are yet available to compare peripheral MRD in a cohort of patients not receiving G-CSF. We conclude from the consistently low RT-PCR titers in leukapheresis samples that, if at all, mobilization of total numbers of BCR-ABL-positive cells does not appear to be a major problem in the patients studied. To our knowledge, no comparable data have been published by others except those by van Rhee et al,29 who pursued a different approach to quantitatively assess MRD in marrow and blood samples from patients with BCR-ABL-positive ALL. While we use a limiting log-dilution RT-PCR assay in an attempt to focus on the quantitation of the number of residual BCR-ABL-positive cells, van Rhee quantitated the ratio of BCR-ABL vs ABL mRNA transcripts. He found no significant difference in BCR-ABL mRNA expression between BM and PB. Another methodological difference to our study was the interval between sampling and RNA extraction. We always extracted RNA within a maximum of 24 h after cell sampling. Van Rhee attributed negative findings in PB with a corresponding positive BM sample in a few discordant cases to poorer quality of PB specimens, a consequence in part of an interval of up to 4 days between collection of samples and processing. In vitro purging in BCR-ABL-positive ALL reported by other investigators included either anti-CD10 MoAb and complement,30 mafosfamide,8,18 combination of anti-CD10 immunotoxins and 4-HC,31 or in two cases long-term marrow cultures.32,33 A total of 20 out of 55 patients autografted for BCR-ABL+ ALL8,9,18,30–38 (not including our series11) received purged marrow.8,18,30–33 None of these reports included monitoring of residual disease before and after purging by semiquantitative PCR. In our study the regression analysis of prepurge RT-PCR titers in bone marrow over a 6 year period did not show a significant drift in median PCR titers. Thus, we assume that the sensitivity of our semiquantitative RT-PCR assay did not change significantly over time. The purging efficiency in marrow grafts, however,


decreased significantly by 1–1.5 log indicating an emerging methodological problem in the purging procedure. The most obvious change over time was that we had to modify our cocktail of monoclonal antibodies, since the AB4 antibody21,39 was only available for us until 1995. At present we assume that AB4 contributed to higher purging efficiencies since we were not able to identify any other variable with significant impact on purging efficiency. In PBSC purging, the AB4 antibody was included in five patients. There was a trend towards a higher proportion of postpurge PCR-negative grafts in this group. However, possibly due to the small number of patients the impact on purging efficiency did not reach statistical significance. We cannot fully explain the remarkable difference of 2.5 log in purging efficiency between BM and PBSC, but we assume that there are both methodological as well as potential biological factors. Methodologically, this difference cannot solely be attributed to the use of the AB4 antibody. Another factor which may account for differences in purging efficiency in PBSC vs BM is the different median number of beads per incubation volume. With significantly higher median ratios of beads per target B cell for PBSC, the concentration of beads per volume, however, was a median of four times higher for BM as it was for PBSC purging. Finally, we cannot exclude that there was a variation in the quality of the CD19 antibodies provided over the 6 year period of this study. A possible biological explanation could be the presence of low numbers of BCR-ABL-positive leukemic progenitors with low or absent expression of the CD19 antigen40 thus escaping the purging procedure. Further studies to define the impact of each of these as yet hypothetical factors are ongoing. In addition, we are currently evaluating a double-purging strategy comprising a CD34 preselection followed by depletion of CD19-positive cells using the MACS (Miltenyi, Bergisch-Gladbach, Germany) cell separation device. Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft (Ho 684/2-1). We appreciate the contribution of the hematologists participating in the German multicenter adult ALL studies.

References 1 Bloomfield CD, Goldman AI, Alimena G et al. Chromosomal abnormalities identify high-risk and low-risk patients with acute lymphoblastic leukemia. Blood 1986; 67: 415–420. 2 Secker-Walker LM, Craig JM, Hawkins JM, Hoffbrand AV. Philadelphia positive acute lymphoblastic leukemia in adults: age distribution, bcr breakpoint and prognostic significance. Leukemia 1990; 5: 196–199. 3 Maurer J, Jannsen JWG, Thiel E et al. Detection of chimeric BCR-ABL genes in acute lymphoblastic leukemia by the polymerase chain reaction. Lancet 1991; 337: 1055–1058. 4 Hoelzer D. Therapy of the newly diagnosed adult with acute lymphoblastic leukemia. Hematol Oncol Clin North Am 1993; 7: 139–160.

5 Hoelzer D. Treatment of acute lymphoblastic leukemia. Semin Hematol 1994; 31: 1–15. 6 Barrett AJ, Horowitz MM, Ash RC et al. Bone marrow transplantation for Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood 1992; 79: 3067–3070. 7 Chao NJ, Blume KG, Forman SJ, Snyder DS. Long-term follow-up of allogeneic bone marrow recipients for Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood 1995; 85: 3353–3354. 8 Stockschla¨der M, Hegewisch S, Kru¨ger W et al. Outcome of 14 patients with Philadelphia chromosome-positive ALL receiving autologous or allogeneic bone marrow transplantation. Bone Marrow Transplant 1995; 15: S71 (Abstr. 306). 9 Radich J, Gehly G, Lee A et al. Detection of bcr-abl transcripts in Philadelphia chromosome-positive acute lymphoblastic leukemia after marrow transplantation. Blood 1997; 89: 2602–2609. 10 Sierra J, Radich J, Hansen JA et al. Marrow transplants from unrelated donors for treatment of Philadelphia chromosomepositive acute lymphoblastic leukemia. Blood 1997; 90: 1410–1414. 11 Martin H, Atta J, Fauth F et al. Autologous BMT/PBSCT in patients with BCR-ABL-positive acute lymphoblastic leukemia. Bone Marrow Transplant 1998; 21: S2 (Abstr. 5). 12 Brenner MK, Rill DR, Moen RC et al. Gene-marking to trace origin of relapse after autologous bone marrow transplantation. Lancet 1993; 341: 85–86. 13 Martin H, Atta J, Zumpe P et al. Purging of peripheral blood stem cells yields BCR-ABL-negative autografts in patients with BCR-ABL-positive acute lymphoblastic leukemia. Exp Hematol 1995; 23: 1612–1618. 14 Atta J, Martin H, Bruecher J et al. Residual disease and immunomagnetic bead purging in patients with BCR-ABL-positive acute lymphoblastic leukemia. Bone Marrow Transplant 1996; 18: 541–548. 15 Maurer J, Bartram CR, Ludwig WD et al. Frequency of BCR/ABL m-RNA in adult lymphoblastic leukemia (ALL) as determined in a prospective study. Ann Hematol 1993; 67 (Suppl.) A80 (Abstr. 312). 16 Yokota S, Hansen-Hagge TE, Ludwig WD et al. Use of polymerase chain reactions to monitor minimal residual disease in acute lymphoblastic leukemia patients. Blood 1991; 77: 331–339. 17 Gobrecht O, Gobel U, Graubner U et al. Effect of dose intensity and therapy-induced leukocytopenia in intensive therapy on the prognosis of acute lymphatic leukemia in childhood. Results in 213 patients of the COALL-85 study. Klin Padiatr 1992; 204: 230–235. 18 Mitterbauer G, Fodinger M, Scherrer R et al. PCR-monitoring of minimal residual leukaemia after conventional chemotherapy and bone marrow transplantation in BCR-ABL-positive acute lymphoblastic leukaemia (published erratum appears in Br J Haematol 1995; 90: 492). Br J Haematol 1995; 89: 937–941. 19 Funderud S, Erikstein B, Asheim HC et al. Functional properties of CD19+ B lymphocytes positively selected from buffy coats by immunomagnetic separation. Eur J Immunol 1990; 20: 201–206. 20 Campana DE, Janossy G. Leukemia diagnosis and testing of complement fixing antibodies for bone marrow purging in ALL. Blood 1986; 68: 1264–1271. 21 Kvalheim G, Sorensen O, Fodstad O et al. Immunomagnetic removal of B-lymphoma cells from human bone marrow: a procedure for clinical use. Bone Marrow Transplant 1988; 3: 31–41. 22 Wang MY, Fodstad O, Ludwig WD et al. An effective, direct immunomagnetic procedure for purging acute lymphoblastic

103



104

23 24 25 26

27

28

29

30

31 32

leukemia cells from human bone marrow. Pathol Oncol Res 1995; 1: 32–37. Pegoraro L, Matera L, Ritz J et al. Establishment of a Ph1positive human cell line (BV173). J Natl Cancer Inst 1983; 70: 447–451. Dhut S, Gibbons B, Chaplin T, Young BD. Establishment of a lymphoblastoid cell line, SD-1, expressing the p190 bcr-abl chimaeric protein. Leukemia 1991; 5: 49–55. Chomczynski P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162: 156–159. Martin H, Atta J, Bruecher J et al. In patients with BCR-ABLpositive ALL in CR peripheral blood contains less residual disease than bone marrow: implications for autologous BMT. Ann Hematol 1994; 68: 85–87. Tsuchiya H, Adachi N, Asou N et al. Responses to granulocyte colony-stimulating factor (G-CSF) and granulocyte– macrophage CSF in Ph1-positive acute lymphoblastic leukemia with myeloid surface markers (letter) (see comments). Blood 1991; 77: 411–413. Komada Y, Zhou YW, Zhang SL et al. Role in growth regulation of cytokines and cytokine receptors in acute lymphoblastic leukaemia expressing myeloid markers. Br J Haematol 1993; 84: 408–415. van Rhee F, Marks DI, Lin F et al. Quantification of residual disease in Philadelphia-positive acute lymphoblastic leukemia: comparison of blood and bone marrow. Leukemia 1995; 9: 329–335. Miyamura K, Tanimoto M, Morishima Y et al. Detection of Philadelphia chromosome positive acute lymphoblastic leukemia by polymerase chain reaction: possible eradication of minimal residual disease by marrow transplantation. Blood 1992; 79: 1366–1370. Brennan C, Weisdorf D, Kersey J et al. Bone marrow transplantation for Philadelphia chromosome-positive acute lymphoblastic leukemia. Proc ASCO 1991; 10: 222. Fabrega S, Laporte JP, Giarratana MC et al. Polymerase chain


33

34

35

36

37

38

39

40

reaction: a method for monitoring tumor cell purge by longterm culture in BCR/ABL positive acute lymphoblastic leukemia. Bone Marrow Transplant 1993; 11: 169–173. Grande M, Barbu V, van den Akker J et al. Autologous bone marrow transplantation in ALL: relapse linked to infusion of tumor cells with the back-up marrow. Bone Marrow Transplant 1994; 14: 477–480. Gehly GB, Bryant EM, Lee AM et al. Chimeric BCR-abl messenger RNA as a marker for minimal residual disease in patients transplanted for Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood 1991; 78: 458–465. Annino L, Ferrari A, Cedrone M et al. Adult Philadelphiachromosome-positive acute lymphoblastic leukemia: experience of treatments during a ten-year period. Leukemia 1994; 8: 664–667. Preti HA, O’Brien S, Giralt S et al. Philadelphia-chromosomepositive adult acute lymphocytic leukemia: characteristics, treatment results, and prognosis in 41 patients. Am J Med 1994; 97: 60–65. Carella AM, Frassoni F, Pollicardo N et al. Philadelphia-chromosome-negative peripheral blood stem cells can be mobilized in the early phase of recovery after a myelosuppressive chemotherapy in Philadelphia-chromosome-positive acute lymphoblastic leukaemia. Br J Haematol 1995; 89: 535–538. Dunlop LC, Powles R, Singhal S et al. Bone marrow transplantation for Philadelphia chromosome-positive acute lymphoblastic leukemia. Bone Marrow Transplant 1996; 17: 365–369. Wang MY, Fodstad O, Funderud S et al. A direct and effective immunomagnetic method for bone marrow purging in non-T ALL malignancies. The use of an anti-HLA-DR antibody gives sustained engraftment. Prog Clin Biol Res 1994; 389: 119–124. Vormoor J, Baersch G, Baumann M et al. Flow cytometric identification of candidate normal stem cell populations in CD45-negative B-cell precursor acute lymphoblastic leukaemia. Br J Haematol 1998; 100: 501–508.