CA2147010A1 - Immunological purging of tumor cells from bone marrow using microspheres and monoclonal antibodies - Google Patents

Abstract

A method is described for the immunological purging of tumor cells from the bone marrow of a patient having B cell lym-phoma. The purged bone marrow may be used for therapeutic autologous bone marrow transplantation. Microspheres and a plu-rality of anti-B cell monoclonal antibodies are used to remove tumor cells from bone marrow, without removal or lysis of non-tu-mor cells, to a level not detectable by polymerase chain reaction assay.

Description

WO94/08621 21 4 7 ~ 1 0 PC~/U593/0989l IMMUNOLOGICAL PURGING OF TUMOR CELLS FROM BONE MARROW
USING MICROSPHERES AND MONOCLONAL ANTIBODIES

GOVERMMENT SUPPORT

This invention was made with Government support and the Government has certain rights to this invention under National Institutes of Health Grants CA40216 and CA34183.

TECHNICAL FIELD
This invention relates to the immunologic purging of tumor cells from bone marrow using a unique combination of monoclonal antibodies and microspheres, and to a method of treating persons having B cell lymphoma by the autologous treatment and transplantation of bone marrow which has been purged of tumor cells by use of said unique combination.

BACKGROUND ART
Therapy by high dose radio/chemotherapy followed by autologous bone marrow transplantation (ABMT) has become a major treatment option for an increasing number of patients with hematologic and solid tumors [1-10]. While the infusion of autologous bone marrow provides suffi-cient hematopoietic stem cells to avert the attendant severe myelosuppresion, there is concern that occult clonogenic tumor cells harbored within the autologous marrow might contribute to relapse through multiplication after the marrow is transplanted back into the patient.
In non-Hodgkin's lymphoma (NHL), bone marrow infiltration by the disease is common at the time of diagnosis and relapse [11-12].
Attempts to purge bone marrow of the lymphoma cells using immunologic and pharmacologic agents are well documented [2,4,13-14]. These studies have shown that a patient's bone marrow can be purged in vitro without significantly impairing hematologic engraftment. After SU~S~I~UTE ~FT ~U~ 26) WO94/08621 21~ 7 ~ ~ O PCT/US93/09891 .

nearly a decade of scientific investigation, however, controversy persists as to whether it is necessary to remove small numbers of residual or even histologically evident tumor cells from the harvested bone marrow.
Moreover, such investigations have not fully answered the question of whether or not bone marrow purging contri-butes to improving a patient's chance of disease-free survival (DFS).
An important obstacle in the determination of whether or not tumor cell purging increases a patient's chances of disease free survival has been the inability accurately identify occult or hidden lymphoma cells in bone marrow before and after in vitro purging. Using traditional morphological methods, a bone marrow specimen ~udged histologically normal may still be infiltrated with up to 5% lymphoma cells. This 5~ figure is the lower limit of morphologic tumor cell detection. Recent-ly, however, more sensitive techniques of assaying bone marrow infiltration by tumor cells have boen developed.
The new techniques have confirmed that a given "histologically normal" bone marrow specimen assayed using the traditional morphological method may indeed contain a substantial number of ly~phom~ cells [15-17].
A technique of particular utility in assaying for occult cells in bone marrow is the polymerase chain reaction (PCR). Several publications [18-28] have shown that PCR amplification can be used to detect the inter-~ chromosomal translocation involving the bc1-2 proto-- oncogene on chromosome-18 and the immllnoglobulin heavy chain locus on chromosome-14. The translocation occurs in approximately 85% of patients with follicular NHL and 30% of patients with diffuse NHL [21-26]. The extreme sensitivity of the PCR t6echnique permits the detection of one ly~pho~ cell in 10 normal cells [18,20]. As wil ~ 35 be shown herein, determining whether marrow is PCR
- positive or PCR negative allows a more accurate assess-_ ment of the efficacy of the materials used to purge bone _ marrow. This accuracy, in turn, permits a more accurate WO94/08621 ~1~ 7 ~ 1 0 PCT/US93/09891 assessment of a patient's disease-free survival prospects.
The use of monoclonal antibodies in the presence of complement to purge bone marrow has been reported by a number of investigators over the past decade. L.M.
Nadler et al.,Lancet, ii:427-431 (1984) reported the use of an anti-B5 (anti-CD20) monoclonal antibody and complement for autologous bone marrow transplantation in patients with relapsed B-cell NHL. Baumgartner et al., Proceedings of the 1st International Symposium on Autologous Bone Marrow Transplantation, eds. K. Dicke, G.
Spitzer and A.R. Zander (1985), pp-377-381. reported treating and transplanting bone marrow treated with complement and the monoclonal antibody anti-Y 29/55, an antibody which recognizes, among other cells, malignant lymphoid cells derived from human plasma cell precursors, but does not react with acute lymphocytic leukemic cells of the null, common, pre-B, and T types, nor with normal hematopoietic elements including granulocyte-macrophage precursors. Feeney et al., Cancer Res., 41: 3331-3335 (1981) reported the elimination of leukemia cells from rat bone marrow using complement and antibody.
More recently, Armitage et al. reported on "Bone marrow transplantation in the treatment of patients with lympho~. " [1]; Freedman et al. studied at "Autologous bone marrow transplantation in B-cell non-Hodgkin's lymphoma...." [2]; Hurd et al. discussed "Autologous bone marrow transplantation in non-Hodgkin lymphom~: mono-clonal antibodies plus complement for ex vivo marrow treatment." [4]; Ball et al. reported "Autologous bone marrow transplantation for acute myeloid leukemia using monoclonal-purged bone marrow." [6]; and Frei et al.
studied the prospects of "Bone marrow transplantation for solid tumors...." [9]. None of these publications, however, has demonstrated a reagent or an antibody or a combination thereof and/or a method of treating bone marrow so as to enable long term patient survival after in vitro autologous bone marrow treatment.

WO94/08621 2l ~7 a l~ PCT/US93/09891 The present invention overcomes the difficulties encountered in the known art and achieves important advantages in the purging of tumor cells from bone marrow by using a combination of microspheres and a plurality of monoclonal antibodies (hereafter abbreviated MmAb). The data compiled and disclosed herein using the claimed invention of mlcrospheres and selected anti-B cell monoclonal antibodies clearly indicates that there is an unexpected synergistic interaction which results in a significantly higher percentage of samples being depleted of tumor cells relative to the known art method of using complement mediated lysis and a plurality of monoclonal antibodies (hereafter abbreviated CmAb). This synergis-tic interaction is evidenced by the fact that of the tumor cell containing bone marrow samples treated herein, all samples can be purged to be PCR negative in tumor cells after three rounds of treatment with a plurality of selective monoclonal antibodies and microspheres. In contrast, using complement mediated lysis and the same plurality of monoclonal antibodies, approximately fifty percent (50%) of CmAb purged samples contained tumor cells as indicated by their failure to be shown PCR
negative after three rounds of CmAb treatment. Tests using microspheres and selected T-cell monoclonal ~ 25 antibodies indicate that the use of microspheres, by = themselves, does not facilitate the removal of tumor cells from bone marrow. Consequently, the synergistic ~ interaction which is observed by using MmAb to purge _ tumor cells from bone marrow is not merely an additive _ 30 effect summed along the separate effects of the micros-pheres and the antibodies. Lastly, the CmAb data included herein indicates that there is a highly significant correlation between the ability to purge bone marrow to PCR negativity for the bc1-2 translocation and ~ 35 disease-free survival of a patient after ABMT. However, = only about 50% of marrow samples have been found purgable _ to PCR negativity using CmAb. Since the data presented ~ herein indicates that the use of MmAb significantly W094/08621 ~ ~ 7~I~ PCT/US93/09891 increases the percentage of samples wh$ch can be purged to PCR negativity (to 90-100%), it is believed that the use of MmAb may significantly i,.l~L~ve the disease-free survival statistics.
DISCLOSURE OF THE INVENTION
The invention discloses a unique method for ~t1nologically purging tumor cells from the bone marrow of a patient having B cell lymphoma for the purpose of therapeutic autologous bone marrow transplantation. The method comprises the steps of:
(a) collecting the marrow;
(b) treating the marrow with a plurality of selective monoclonal antibodies and microspheres in a specified sequence to deplete the marrow of tumor cells, without depletion of non-tumor cells, to a level where the tumor cells are not detectable in the marrow by polymerase chain reaction assay; and ( C ) A~mi ni stering the treated bone marrow to the patient from whom it was obtained.
The method entails the use of antibodies which conjugate to the tumor cells and microspheres, preferably magnetic microspheres of size in the range about 0.3 to about 5.0 microns, coated with goat anti-mouse immunoglobulin(s) (Ig) when then conjugate to the antibodies. Goat anti-rabbit and rabbit anti-mouse Igs may also be used according to the invention. Goat anti-mouse Ig is preferred. The microspheres, with antibodies and tumor cells conjugated thereto, were then separated from the marrow sample. This unique method of using a plurality of monoclonal antibodies and microspheres has been found to remove tumor cells from bone marrow to PCR
negativity without the use of complement. Furthermore, whereas the use of a plurality of selective monoclonal antibodies and complement has been found to purge about 50% of tumor containing bone marrow samples to PCR
negativity, 90-100% of tumor containing samples have been purged to PCR negativity using a plurality of monoclonal antibodies and microspheres.

WO94/08621 214~ Ql~ - PCT/US93/09891 -In further embodiments using microspheres, instead of sequencing the antibody and microsphere treatments, the monoclonal antibodies may be bound to the micro-spheres, or to microspheres conjugated or coated with immunoglobulins or other substances, before contact with tumor cell containing bone marrow. These embodiments, stereochemical and other physical considerations known to those skilled in the art will have to be considered. For example, if the antibodies are to be conjugated to the microspheres before contact with tumor cell containing samples, it may be necessary to insert a bridging group of about 1-20 atoms long between the surface of the microsphere and the antibody. This in turn may re~uire washing the microspheres after a purging cycle and combining the washing with a bulk sample in order to m~ ni ~ ze the loss of non-tumor cells.
BRIEF DESCRIPTION OF THE DRAWINGS
~ Fig.1 is a graph showing the efficacy of immunologic = purging of Ra~i cells with various complement and/or antibodies as assessed by clonogenic cell growth;
Figs. 2A and 2B is the Southern Blot analyses of bc1-2 translocation sequences amplified by polymerase chain reactlon (PCR) before and after treatment according to the invention;
Fig.3 is a graph of the actuarial probabillty of disease free survival after ABMT in 114 persons with B-~ cell NHL;
_ Figs. 4A-4C is a graph of the actuarial probability ~ of disease free survival after ABMT subclassified by =~ 30 disease status at ABMT;
Figs. 5A-5D is graph of the actuarial probability of disease-free survival after ABMT subclassified by bone ~ marrow involvement at ABMT; and - Fig. 6 is the Southern Blot analyses of bone marrow (A) and multiple peripheral blood samples (B-G) taken from two patients for bc1-2 translocation sequences as amplified by polymerase chain reaction.
Figs. 7A-7C, illustrate detection by Southern Blot ~- W094/08621 2l470la PCI'/U593/~9891 analysis of the bc1-2 translocation sequences amplified by polymerase chain reaction.
Fig. 8 illustrates that PCR analysis is capable of detecting one lymphoma cells on 10 normal mononuclear cells following MmAb-4 treatment.
BEST MODE FOR CARRYING OUT THE IN~ENTION
References 1. J.O. Armitage et al., Blood, 73: 1749-1758 (1989).

2. A.S. Freedman et al., J. Clin. Oncol., 8; 1-8 ( 1990 ) .

3. J.G. Gribben et al., J. Clin. Oncol., 7: 1621-1629 (1989).

4. D.D. Hurd et al., Am. J. Med., 85: 829-834 (1988).

5. I. Philip et al., N. Eng. J. Med., 316: 1593-1498 (1987).

6. E.D. Ball et al., Blood, 75: 1199-1206 (1990).

7. J.G. Gribben et al., Blood, 73: 340-34 (1989).

8. R. Wallerstein et al., J. Clin. Oncol., 8:
1782-1788 (1990).

9. E. Frei et al., J. Clin. Oncol., 7: 515-526 (1989).

10. W.P. Peters et al., J. Clin. Oncol., 6: 1501-1515 (1988).

11. F. Dick et al., Cancer, 33: 1382-1398 (1874).

12. R.S. Stein et al., Cancer, 37: 629-636 (12976).

13. F.M. Unkun et al., Blood, 69; 361-366 (1987).

14. T. Takvorian et al., N. Eng. J. Med., 316:
1499-1505 (1987).

15. N. Berliner et al., Blood, 67: 80-855 (1986).

16. D. Benjamin et al., Blood, 61: 1017-1019 (1983).

17. M.L. Cleary et al., Proc. Natl. Acad. Sci. USA, 81: 593-597 (1984).

18. M.S. Lee et al., Science, 237: 175-178 (1987).

19. M. Crescenzi et al., Proc. Natl. Acad. Sci.
USA, 85: 4869-4873 (1988).

æ~Q~

20. B.Y. Ngan Qt al., Blood, 73: 1759-1762 (1989).

21. J.J. Yunis et al., N. Eng. J. Med., 307: 1231-1236 (1987).

22. L.M. Weiss et al., N. Eng. J. Med., 317: 1185-1189 (1987).

23. M.S. Lee et al., Blood, 70: 90-94 (1989).

24. J.J. Yunls et al., N. Eng. J. Med., 320: 1047-1054 (1989).

25. A.C. Aisenberg et al., Blood, 71:969-972 (1988).

26. W.B. Graninger et al., J. Clin. Invest., 80:
1512-1515 (1987).

27. R. Higuchi, 'Simple and rapid preparation of samples in PCR:, PCR Technology, H.A. Erlich Ed. (New York: Stockton Press, 1989), pp. 31-38.

28. L.M. Nadler et al., J. Clin. Invest., 67: 134-140 (1981).

29. A.S. Fr~P~m~n et al., J. T~mtl~ol., 134: 2228-2235 (1985).

30. J. Ritz et al., Nature, 283: 5383-583 (1980).

31. J. Ritz et al., Blood, 58: 648-652 (1981).

32. L.M. Nadler et al., Lancet, ii: 427-431 (1981).

33. D.C. Roy et al., Leukemia Res., 14: 407-416 (1990) .

34. C. Taswell, J. Immunol., 126: 1614-1618 (1981).

35. M.L. Cleary et al., Proc. Natl. Acad. Sci. USA, 81: 593-597 (1985).

36. M.L. Cleary et al., J. Exp. Med., 164: 315-325 (1986).

37. S. Kwok et al., Nature, 339: 237-238 (1989).

2~i 47~10 g 38. R. Peto et al., J. Royal Stat. Soc. A, 135:
185-206 (1972).

39. D.R. Cox, Analysis of Binary Data (London:
Methuen & Co., 1970).

40. C.R. Mehta et al., Biometrics, 40: 819-825 (1984).

41. D.R. Cox, J. Royal Stat. Soc. B, 34: 181-220 (1972).

42. E.L. Kaplan et al., J. Amer. Stat. Soc., 53:
457-481 (1958).

43. R.S. Negrin et al., Blood, 77: 654-660 (1990).

44. A.S. Freedman et al., Leukemia, 1: 9-15 (1987).

45. J. Cossman et al., Am. J. Pathol., 115: 117-124 (1984).

46. S. Hellman et al., N. Eng. J. Med., 234: 1585-1590 (1991).

47. E. Passamani, N. Eng. J. Med., 234: 1589-1592 (1991) .

48. F. Herrmann et al., Blood, 246-254 (1987).

49. S.I. Schlager et al., Cancer Res., 37: 1432-1537 (1977) 50. S.I.Schlager et al., J. Immunol., 120: 472-480 (1978).

51. A.P. Gee et al., J. Natl. Cancer Inst., 75:
441-445 (1985).

52. A. Circolo et al., J. Immunol., 128: 1118-1121 (1982).

53. T. Borsos et al., Molec. Immunol., 20: 433-438 (1983).
I. DEPLETION OF TUMOR CELLS USING COMPLEMENT MEDIATED
LYSIS
All bone marrow samples taken and described herein were obtained with the informed consent of patients or volunteers and after validation by the Human Protection Committee of the Dana-Farber Cancer Institute in Boston, Massachusetts. All treatments given and described herein were given with the informed consent of the patients and after validation by the Human Protection Committee of the ~ WO94/08621 2~ PCT/US93/09891 -Dana-Farber Cancer Institute. The data and results presented throughout this application are to demonstrate the utility of the invention and are not to be construed as limiting the invention. The complement mediated lysis data presented herein is comparative to the invention = claimed herein.
A major concern in using ABMT has been that clonogenic tumor cells might be re-infused with autologous bone marrow and contribute to patient relapse.
This invention enables effective immunologic purging using a unique combination of microspheres and selective monoclonal an3tibodies6. This combination is capable of inducing a 10 to 10 tumor cell kill in clonogenic = assays. As a consequence of this reduction in tumor ~ 15 cells after treatment according to the invention, the = bone marrow was PCR negative in 57 of 114 patients who had a PCR amplifiable breakpoint involving the bc1-2 translocation at the time of bone marrow harvest. More importantly, those patients who were re-infused with ~ 20 autologous bone marrow depleted of residual lymphoma - cells, i.e., PCR negative, had a highly statistically significant increase in disease-free survival compared to those patients whose bone marrow could not be purged to a PCR negative value and were, consequently, re-infused with autologous marrow containing highly reduced, but = residual lymphoma cells.
One hundred and fourteen patlents wlth B-cell non-Hodgkin's lymphoma were studied. Al l patients had ~ radio/chemosensitive disease at the time of ABMT. This _ 30 was assessed as the achievement of protocol eligible _ m;n;m~l disease state following aggressive induction or salvage radio/chemotherapy. Protocol eligible minimal disease criteria included (1) whether there was complete ~ remission (CR) or partial remission (PR) of tumor masses = 35 two centimeters or less and (2) whether there was bone marrow infiltration of 5% or less by tumor cells in the initial protocol, but less than 20% of the intertra-becular space in subsequent protocols. All cases were W094/08621 2 1 ~ 7 ~ 1 0 PCT/US93/09891 documented by immunotyping to express Bl (CD20).
Bone marrow was re-infused following immunologic purging using complement and a tumor removing sufficiency of a plurality of anti-B cell and anti-CALLA (antl-Common Acute Lymphoblastic Leukemla Antigen) monoclonal antibodies. The presence of a sufficiency of monoclonal antibodies is determined by pretreatment tests whlch determine the approximate amount of tumor cells present in the bone marrow. Rabbit complement was preferred.
PCR negativity was typically achieved after three treatment or tumor cell purging cycles. Consequently, in all cases, purged bone marrow was transplanted only after three marrow treatment or purging cycles. Vials of marrow from all patients were cryopreserved before and after purging. DNA was extracted from 38 patients' non-clyo~Leserved samples before and after purging. For 10 of these patients, DNA was also extract2d from aliquots of bone marrow obtained after each of the three complement/antibodies purgings. Cells isolated from diagnostic lymph nodes were cryopreserved in the ma~ority of patient~. Normal bone marrow was obtained from healthy volunteer donors. Genomic DNA was lsolated in all cases by cell lysis treatment using non-ionic detergents and proteinase K (Sigma, St. Louis M0). An additional 50 DNA extractions were also conducted using known established procedures.
Bone marrow was harvested from the iliac crest under general anesthesia and collected in RPMI media (Whitakker, Piscataway, NJ) containing preservative free heparin. The harvested bone marrow was concentrated to yield a buffy coat and washed on a Cobe 2991 cell washer.
The mononuclear cell fraction was isolated by centrifu-gation over Ficoll-Hypaque (Pharmac7a, Piscataway, NJ) and the cells resuspended at 2 x 10 cells/ml in RPMI
1640 containing 0.5% fetal bovine serum (FBS, Hycone Lab, Logan UT). The cells were incubated with a tumor removing sufficiency, preferably a saturating concentra-tion, of the monoclonal antibodies for a time in the WO94/08621 ~ ~ PCT/US93/09891 -range of 15 minutes to about 1 hour, preferably for about = 15 minutes, at 4C. A saturating concentration for each antibody is determined from the number of tumor cells present and the antigen density on the cells. Rabbit complement (3-4 week old rabbit serum, Pel Freeze Inc., Brown Deer, WI) was added at predetermined dilutions for each lot and incubated with target cells for 30 minutes at 37C in the presence of deoxyribonuclease (Sigma, St.
Louis, M0) at 2.5 mg/ml to prevent cell clumping. The amount of complement added is dependent on the volume of the sample is in the range 1/1 to 1/10 v/v. The cells were pelleted by centrifugation and the procedure repeated twice for a total of three treatments in vitro with the monoclonal antibodies and rabbit complement.
= 15 The cells were washed three times, resuspended in autologous serum and 10% dimethylsulfoxide (Sigma, St.
- Louis, M0) and ~L yO~ eserved according to the method of - Takovarian et al., N. Engl. J. Med, 316: 1499-1505 (1987). Before reinfusion, the cLy~Leserved cells were rapidly thawed, diluted in medium containing DNAase as described by Takovarian et al., NEMJ, 316: 1499-1505 - (1987). Vlability was measured by trypan dye exclusion.
= Generally, bone marrow was reinfused withln 4 weeks after = it was obtained.
= 25 The ly-m~hom-~ cell lines Ra~i and DHL-6 were used ln evaluating the invention. Ra;i is a human Burkitt's lymphoma B cell line expressing CD20, CD10 and B5 anti-- gens. Raji cells were grown in RPMI 1640 medium contain-_ ing 10% heat inactivated FBS, 2% l-glutamine, 1~ sodium - 30 pyruvate and 1% penicillin and streptomycin. DHL-6 is a human B cell line containing a bc1-2 translocation and ~ was received from Dr. A. Epstein (Univ. So. Calif., Los - Angeles, CA).
The monoclonal antibodies used according to the invention are an anti-CD (anti-B1) monoclonal antibody, a monoclonal antibody specific for actlvated B cells and B
cell ly~phom~ (anti-B5), an anti-CD10 (anti-J5) mono-clonal antibody and an anti-CD19 (anti-B4) monoclonal antibody. The antibodies usQd in accordance with the invention must induce lysis of tumor cells in the bone marrow in the preRence of complement. The antibodies used herein all induce lysis in the presence of rabbit complement.
Anti-Bl is an IgG2a murine monoclonal antibody specific for a 35,000 dalton cell surface glycoprotein present on normal and malignant cells. The Bl antigen is found on all B cells isolated from peripheral blood, lymph node, spleen, tonsil and bone marrow. This antigen is also found on 50~ of CALLA positive acute lymphoblas-tic leukemias (ALL); but not normal T cells, monocytes, granulocytes or tumors of these lineages. The Bl antigen is an integral component of the B cell membrane and is also found on some null cell lymphomas and non-T cell ALLs. The anti-Bl monoclonal antibody was prepared by P.
Stashenko et al., "Characterization of a human B lympho-cyte-specific antigen", J. Immunol., 125: 1678 et seq.
(1980) and L. M. Nadler et al., "A unique cell ~urface antigen identifying lymphoid malignancies of B cell origin", J. Clin. Invest., 67: 134 et seq. (1981). Any monoclonal antlbody which reacts with the Bl antigen in the same manner as anti-Bl may be used in practicing the invention. Complement mediated lysis is dependent upon the complement mediated cytolytic activity of the mono-clonal antibody.
Anti-B5 is a murine IgM monoclonal antibody described in U.S. Patent No. 4,692,405. The B5 antigen has a molecular weight of 75,000 daltons and is expressed on a small number of activated B cells in unstimulated lymph node, spleen and tonsil. It is not expressed on resting B cells isolated from peripheral blood, lymph node, spleen or tonsil. Resting spleenic B cells activated in vitro with protein A, anti-Ig, Epstein-Barr virus or pokeweed mitogens demonstrate the appearance of B5 antigen. B5 is found on a wide variety of B cell neoplasms and the majority of B-Cll, Burkett's lymphomas, nodular poorly differentiated lymphocytic lymphomas, WO94/08621 PCT/US93/09891 ~
2 ~

diffuse poorly differentiated lymphocytic lymphomas, diffuse large cell ly~pho~, hairy cell leukemias, and diffuse well differentiated lymphocytic lymphomas. B5 is not expressed on non-T cell Alls, T cell or myeloid leukemic cells, normal T cells, monocytes, granulocytes t red blood cells or platelets. It has, however, been noted on some non-hemopoietic malignancies, especially small cell lung carcino~s. Any monoclonal antibody which reacts with the B5 antigen in the same manner as 10 anti-B5 may be used in practicing the invention.
Anti-J5 is an IgG2a murlne monoclonal antibody directed agalnst the human Common Acute Lymphoblastlc Leukemla Antlgen (CALLA) which has a molecular weight of about 100,000 daltons. The antigen is found on tumor 15 cells from 80% of patlents wlth non-T cell ALLs and 40-50% of patients with chronic myelocytic leukemia in blast crisls. This antigen is present on a small number of cells ln normal bone marrow and in fetal liver. The J5 antigen is also found on the tumor cells from some B cell and ~ cell ly~phc-~, including poorly differentiated nodular lymphocytic lymphoma, Burkett's lymphoma and T
cell lymphoblastic ly~rho-~, as well as normal renal tubular and glomerular epithelial and breast myoepihtelial cells. The anti-J5 monoclonal antibody was derived from the hybridization of mouse NS/l-AG cells with spleen cells from BALB/cJ mice immunized with tumor cells from a patient with CALLA positive non-T cell ALL
by J. Ritz et al., "A monoclonal antibody to human acute lymphoblastic leukemias antigen.", Nature, 283: 583 et seq. (1980). Any monoclonal antibody which reacts with the J5 antigen in the same manner as the anti-J may be used in practicing the invention.
Anti-B4 (anti-CDl9) is a murine IgG2 monoclonal antibody for CDl9. The antibody recognizes the B4 antigen which is present on all B cells isolated from lymphoid organs and on approximately 5% of normal adult = bone marrow cells. In addition, the B4 antigen is expressed on greater than 95% of non-T cell acute lympho-WO94/08621 2 1 ~ 7 0 1 0 PCT/US93/09891 blastic leukemias, on 90% of B cell lymphom~ and B cell chronic lymphocytic leukemias.
The anti-Bl, anti-B5, anti-B4 and anti-J5 monoclonal antibodies are co~rcially available from Coulter Immunology Division, Coulter Corporation, Hialeah, Florida.

Clonogenic Assay Bone marrow was obtained from healthy volunteer donors and anticoagluted with preservative free heparin.
The mononuclear cell fraction was lsolated by Ficoll-Hypaque density gradient centrifugation and cells irradiated to 40 Gy at ll.l Gy/min ( Cs, Gamma cell, Atomic Energy of Canada, Ottawa, Canada) before they were mixed with lymphoma cells in purging experiments. The lymphoma cell line Ra~i was added to irradiated normal marrow ~o~on1~clear cells in a l:20 ratio and suspended in media at 2 x l0 cells/ml. The cell suspensions were treated with the monoclonal antibodies and complement combination for a total of three treatments using the same protocol used for the marrow harvest samples. The cells were then washed three times and plated in a limiting dilution assay [33]. Each sample was serially diluted from 5 x l0 to 0.5 cells per l00 ml in RPMI
media supplemented with 10% FBS. From 48 to 96 aliquots of each dilution were plated in flat bottomed micro-culture wells (Nunclon, Nunc, Denmark). Fresh media was added every four days and incubated at 37C in a 5% CO
atmosphere for 14-18 days. Growth at each serial dilution was assessed in "all or nothing", positive or negative fashion under an inverted phase microscope.
Under these conditions, the only tumor cells which will grow are those with high cloning efficiency. The plot of the number of negative wells at each dilution against the total number of cells for each point follows a Poisson probability distribution. Frequency of clonogenic cells within the test population was estimated using Chi-square m; ni i zation.

W094/08621 21~7 ~ I ~ PCT/~lS93/Og891 ~

Polymerase Chain Reaction DNA was heated to 96C for lO minutes to destroy proteinase K activity before amplification. Nested oligonucletide amplification was performed at both the ma~or breakpoint (MBR) and minor cluster (mcr) region of the bcl-2/Ig hybrid gene for each sample. Conditions for the PCR amplif~cation at MBR were optimized by using serial dilutions of the cell line DHL-6 in normal bone marrow mononuclear cells So that dilutions of one tumor = lO cell in lO normal cells were readily detectable. No cell line was available to determine accurately the sensitivity of the PCR by using the primers for the mcr.
However, dilution studies using patient tumor cells 6 suggest a sensitivity of at least one tumor cell in lO
normal marrow cells. Standard precautions against cross-cont~in~tion of amplified material were taken and amplified material was never taken to the areas where DNA
extraction was performed.
PCR amplification was performed for 25 cycles in a Perkin Elmer Cetus thermal cycler (Cetus, Emeryville, CA) in 50 ml of buffer containing 50 mM KCl, lOmM Tris-Cl, 2.25 mM MgCl2, 0.01% gelatin, l.5 mg of DNA, 20nM of oligonucleotide primers, 200 mM each of dATP, dCTP, dGTP
and dTTP, 1.5 U Taq polymerase (Cetus, Emeryville, CA).
The $nitial amplification was performed using oligonucleotides 5'CAGCCTTAAACATTGATGG(Seg l) for the MBR, 5'CGTGTGGTACCACTCCTG3'(Seq 2) for the mcr and 5'ACCTGAGGAGACGGTGACC3'(Seq 2) for the J consensus region. Each cycle was performed with denaturation at 94C for one minute, annealing at 55C (MBR) or 58C
(mcr) for one minute and extension at 72C for one minute. The final extension period was extended to ten minutes.
A 5 ml aliquot of the amplified mixture was made to 50 ml and re-amplified for 30 cycles using oligonucleotides internal to the original primers, 5'TATGGTGGTTTGACCTTTAG5'(Seq 4) for the MBR, 5'GGACCTTCCTTGGTGTGTTG3'(Seq 5) for the mcr and WO94/08621 214 7 ~ 10 PCT/US93/09891 5'ACAGGGTCCTTGGCCCCA3'(Seq 6) for the J consensus region, with one minute denaturation at 94C, one minute annealing at 58C and one minute extension at 72C. The final extension period was again extended to ten minutes.
Aliquots of the final reaction product were analyzed by electrophoresis in 4~ agarose gel (NuSieve, FMC, Rockland ME) containing ethidium bromide and visualized under UV
light. DNA was blotted onto Zeta-probe blotting mem-branes (Bio Rad, Richmond, CA) and bc1-2 specific DNA
detected by hybridization overnight with P-labelled oligonucleotide probes, 5'CCCTCCTGCCCTCCTTCCG3'(Seq 7) for the MBR, and 5'GGACCTTC~ll~lGTGTTG3'(Seq 5)3for the mcr. Oligonucleotides were radiolabelled with ( P)ATP
using T4 polycucleotide kinase (New England Biolabs, Beverly, MA) according to the manufacturer's instructions.
Control tests were performed with each amplification usi5ng a weak positive control consisting of DNA from a dilution of the cell line DHL-6 in normal bone marrow cells and a negative control conslstlng of PCR
buffer containing heat inactivated proteinase K. Each sample was analyzed at least three times at each breakpoint site. In addition, in all samples having no detectable PCR product, PCR reactions were repeated using the original oligonucleotides and primers specific for the human B7 gene to ensure that DNA could be amplified in all samples. Serial dilutions of each DNA sample were then made in the PCR sample buffer. PCR was performed to determine the titer at which PCR became negative in order to estimate the quantity of DNA with the translocation in the starting material.

Evaluation and Statistical Methods.
Before treatment, patients were evaluated by com-plete blood chemistry, bone marrow aspirate and biopsy, physical examination, cell-surface phenotype studies of bone marrow mononuclear cells and peripheral blood.
Other evaluations included computer tomography, X-ray and gallium scanning as needed to determine the extent of W O 94/08621 PC~r/US93/09891 2~7~ 18-disease. Following reinfusion, retesting was done every six months or as clinically necessary.
The relationships between patient characteristics, the results of PCR post-lysis and time to relapse were S assessed using the logrank test t38]. Relationships between patient characteristics and PCR were assessed with the Fisher exact test [39]. The Wilcoxon test for ordered contingency variables was used for ordered categorical data. Stratified logrank tests were also considered using patient variables thought to be prognostic of a longer time to relapse. Cox proportional hazards regression [41] was used to attempt to build a model to estimate the effect of co-variates on the risk of relapse. Disease-free survival curves were estimated using the method of Kaplan and Meier [42] and compared by the logrank test. Survival was calculated from the day of marrow transplantation. Table 1 summarizes relevant characteristics of the patients followed during the course of testing the invention.
= 20 Table 1. Characteristics of B-cell NHL patie*nts with bc1-2 translocation undergoing ABMT
PCR- PCR+ Total Total 57 57 114 Male 37 35 72 Female 20 22 42 Histology Low 36 38 74 Intermediate 20 18 38 High 1 1 2 BM never involved 20 9 29 BM previously involved 37 48 85 Previous EN in*volvement16 25 41 No previous EN involvement 41 32 73 WO94/08621 ~7 Q 1 0 PCT/US93/09891 PCR- PCR+ Total At ABMT

BM negative 35 30 65 BM <5% 20 18 38 BM >5% 2 9 ll * Extra_odal ** Complement and monoclonal antibodies anti-Bl, anti-B5 and anti-J5 were used.
The ability to purge tumor cells from normal bone marrow by using monoclonal antibodies was demonstrated with an in vitro model. The hybridoma cells line Raji was added to normal donor bone marrow mononuclear cells in a l:20 ratio. This suspension was then treated with the anti-B cell monoclonal antibodies anti-Bl and anti-B5, with anti-J5, either singly or in combination, in the presence of complement. The cell suspension was treated a total of three times in order to deplete Ra~i clonogenic cells.
Fig. l indicates that treatment of the RaJl contA~ n~ ng suspension with complement alone or wlth the comblnation antl-(B5+Bl+J5) without complement did not slgnificantly reduce the fraction of clonogenic cells in the suspension sample. However, when complement and one of anti-B, anti-Bl, or anti-J5 was used, there was a significant reduction in clonogenic cell growth following in vitro purging as compared to purging with an antibody or complement alone (p=0.014).
In these evaluations, antibody and component ly3is was capable of inducing approximately three logs (lO ) of cell kill. The combination two or three antibodies with complement produced an even greater reduction in clonogenic tumor cell growth compared to complement and a single antibody (p=0.0066). According to Fig. l, the combinations anti-Bl + anti-J5 I complement and anti-Bl -W094/08621 ~ PCT/US93/0~891 -anti-B5 + anti-J5 + complement both reduced the fraction of clonogenic cells to about 10 . The latter combination is preferred.
To determine whether immunogenic purging could eradicate lymrho~ cells from patients' bone marrow at the time of ABMT, samples from 10 patients whose harvested marrow contains cells with translocation of the bc1-2 oncogene were analyzed to determine whether these cells could be depleted to PCR negatlvity by immunogenlc purging. Each of these bone marrows was treated with a cocktail comprising anti-B1 plus anti-B5 plus anti-J5 plus complement. Samples were analyzed for the bc1-2 translocation by PCR before and after each of the three rounds of monoclonal antibodies plus complement treatment. Six of the 10 patients' bone marrow purged to negativity. In each case, three cycles of monoclonal antibodies plus complement treatment was required to achieve PCR negativity. The results obtained with three representative patients from this group are shown in Fig.
2a. Patients 1 and 2 reached PCR negativity following the third round of treatment while patient 3 remained PCR
positive for the bc1-2 translocation. Negrin et al., Blood, 77:654-660 (1991) [43] have noted that lmmunologic purging of lymphoma cell line could result ln PCR
negativity, although these authors also notod that there was a difference among ly ,hom~ cell lines with regard to susceptibility to purging. Overall, the bone marrow of 50% of treated patients became PCR negative for bc1-2 after three complement/antibodies treatment cycles.
The results described herein were obtained from analysis of the treatment of patients who had a docu-mented PCR amplifiable breakpoint for the bc1-2 trans-location and for whom both pre- and post-lysis samples were available for analysis. The bc1-2 translocation was identified by PCR in the diagnostic tissue of 125 patients and the requisite samples were available for 114 of these cases. In addition, pre- and post-lysis samples wo g4/n862~ 7 D~ D rc~/us93/o989l were also available from 47 patients who had B-cell NHL
with no PCR amplifiable translocation involving the bcl translocation. These 47 samples were used as controls.
DNA was extracted from the pre- and post-lysis samples and PCR was performed at the MBR and mcr of bc1-2 to assess whether residual cells with the bc1-2 trans-location were present at harvest and following compl-ment/antibodies purging. Each of the pre- and post-lysis samples was analyzed three times and was blinded with regard to clinical outcome or to whether a bcl-trans-location had previously been identified for that patient. At the time of bone marrow harvest, PCR
detected bone marrow infiltration in the pre-lysis marrow in all of the 114 cases for whom a bc1-2 translocation was detected in the diagnostic tissue.
Immunologic purging resulted in the loss of detectable PCR product in the post-lysis samples of 57 of the 114 patients involved herein. In the other 57 post-lysis samples, PCR of the bc1-2 translocation was positive after purging. The results from nine represen-tative pre- and post-lysis samples is shown in Fig. 2b.
Lanes 1-4 represent cases where ly~rhom~ cells with the bc1-2 translocation could not be detected by PCR in the post-lysis sample. Lanes 5-9 are representative of patients who rPm~;ned PCR positive after lysis. Although equal amounts of DNA were added to each PCR reaction, the ability to purge to PCR negativity did not appear to correlate with the intensity of the PCR signal in the pre-lysis sample. In fact, in the samples illustrated in Lane 8, there was a slight increase in the intensity of the PCR product detected after purging. This may be ~ust the normal variation in the technique or it may be that in this patient the loss of normal cells was relatively greater than the loss of neoplastic cells. Since PCR is not a quantitative assay, PCR was performed using serial dilutions of the pre- and post-lysis samples to determine the titer at which the PCR product could no longer be detected. ThiS provided a semi-quantitative estimate of ~ WO94/08621 ~1~70~0 PCT/US93/09891 -the amount of DNA in each sample that has the bc1-2 translocation. There is no correlation between the amount of DNA in a pre-lysis sample and the ability to purge to PCR negativity (p=0.138, Wilcoxon test for ordered contingency variables). The lack of correlation between the amount of DNA in the pre-lysis sample and the ability to purge to PCR negativity suggests that there may be intrinsic difference in the susceptibility of lymrhor~ cells from different patients to the purging regimen.
As part of the test analysis, an attempt was made to correlate patients' clinical characteristics with the ability to purge to PCR negativity. As Table 1 indicates, PCR negative and posltlve subgroups were equally weighed with regard to gender, histology and previous extranodal (EN) disease. While a number of - different correlations were attempted, the best results were achieved using the history of histological bone marrow involvement. In the 29 patients with no history of bone marrow involvement, 69~ became PCR negative. In the 85 patients with a history of bone marrow involve-ment, 44% were PCR negative after purging (p=0.0169). No other available on-study variable significantly improved this model.
The effect of the ability of the complement/anti-bodies combination to purge all residual lymphoma from the marrow was correlated with disease free survival (DFS) after ABMT. Fig. 3 indicates that DFS after ABMT
correlates strongly with whether the patient purged PCR
negative or positive after lysis (p,0.00001). In the group of patients who purged PCR negative, only four of 57 patients relapsed after as long as 8 years. Two additional patients in this group were removed from analysis at 24 and 28 months after ABMT after they died from unrelated causes. Neither of these two patients was found to have lymphoma after gross and microscopic e~ n~tion during autopsy. In contrast, of the 57 patients who r~m~;ned PCR positive for the bc1-2 trsns-~ W O 94/08621 2 1 4 7 0 1 ~ PC~r/US93/09891 location, 26 patients have relapsed and the median DFS ofthis group was reached at 19.7 months.
Figs. 4A-4C compare patients in complete recovery and partial recovery, and indicates that those in complete remission had increased DFS (p=0.0021).
However, Fig. 4 also indicates that of the 49 patients in complete remission at ABMT, the 29 patients who purged PCR negative for bc1-2 had increased DFS after ABMT
compared to the 20 patients who r~-A~ned PCR positive after purging (p=0.0012). Similarly, of the 65 patients in partial remission at ABMT, the 28 patients who purged PCR negative had increased DFS compared to the 37 patients whose post-lysis sample remained PCR positive after purging (p=0.0011).
Figs. 5A-5D show that histologic bone marrow involvement at ABMT also correlates with DFS after ABMT
(p=0,0054). Of the 65 patients with no morphological evidence of bone marrow infiltration at the time of bone marrow harvest, the 35 patients who purged PCR negative after lysis had increased DFS compared to the 30 patients who r~m~in~d PCR positive after purging (p=O.OOOl).
Similarly, in the 38 patients assessed by morphology to have ~; n; r~l bone marrow involvement at ABMT (<5% of the intertrabecular space), the 20 patients who purged PCR
negative also had increased DFS compared to the 18 patients who remained PCR positive (p=0.005). Eleven patients had overt histologic bone marrow involvement (10-20% of the intertrabecular space) at the time of bone marrow harvest. Only two of these patients became PCR
negative after purging. Histologic subtype of the NHL
did not correlate with DFS after ABMT (p=0.249).
However, within each subtype, those who purged negative after lysis had increased DFS. After examining the data presented in Fig. 4, the conclusion is that purging to PCR negativity increases DFS, and that the combination of complement and antibodies will le...ove tumor cells from bone marrow to the point of PCR negativity, but not for all different bone marrow samples.

'~47~`10 The relationship between PCR status after purging and tlme to relapse was assessed uslng the logrank test.
Stratified logrank tests were also performed using as stratification factors those clinical variables poten-tially prognostic of longer time to relapse. Univariatelogrank analysis of the data identified several variables that were associated with suggestive differences in the time to relapse a~ter ABMT. This analysis identified complete remission (p=0.0021), bone marrow infiltration at ABMT (p=0.0054) and history of bone marrow infiltra-tion (p=0.05) as potentially explanatory variables in addition to PCR. In all cases, the effect of PCR
persisted when other variables were used for stratifica-tion. The effect of PCR remained significant within each strata. The results indicate the utility of the comple-ment/antibodies combination for providing a readily evaluated means for purging bone marrow of hidden tumor cells. The combination can reduce tumor cells in bone marrow of at least some patients to PCR negativity. PCR
negativity is shown to correlate with increased DFS. It is unclear as to why the bone marrow of all patients cannot be purged to PCR negativity. It is within the scope of the claimed invention that other anti-B1 and anti-J5 monoclonal antibodies, and other monoclonal antibodies specific to activated B cells and B cell lymrho~ will increase the percentage of patients whose bone marrow may be purged of tumor cells.
The detection of peripheral blood cells with the bc1-2 translocation after the ABMT procedure has been carried out was found to correlate with the ability to purge the bone marrow of tumor cells. Durlng the course of perfecting the invention, analyses were carried out to determine whether there were detectable lymphoma cells in the peripheral blood of patients at the time of ABMT.
Twenty-five patients were involved in the study. Blood samples were also tested six months after ABMT. Multiple blood samples were obtained from each patient and analyzed by PCR. The results from the analysis of bone ~ W094/~862l 21 ~ 7 0 I 0 PCT/US93/09891 marrow and multiple blood samples from two representative patients, patients 1 and 8 from Fig. 2b, are shown in Fig. 6. There was no correlation as to which patients had peripheral blood cells positive for bc1-2 translocation before and after ABMT (data not shown).
However, cells containing the bc1-2 translocation were detectable in the peripheral blood of 13 of the 14 patients in this group who rPm~;~ed PCR positive after purging. In contrast, no cells having the bc1-2 translocation were found in the peripheral blood of the 11 patients who became PCR negative after CmAb marrow purging.
The use of three and four monoclonal antibodies in conjunction with complement was compared. Tumor cell contAi~ng bone marrow samples from 19 patients were treated through three cycles as described above using complement and either anti-B1, -B5 and -J5 or anti-B1, -B4, -B5 and J5. The results are shown below in Table 2.
Using three monoclonal antibodies, 10/19 samples remained PCR positive. Using four monoclonal antibodies, 5/19 samples rem~lned PCR positive.

II. Depletion Of Tumor Cells Using Microspheres Example 1. Comparison of Complement Induced Lysis And Magnetic Sphere Depletion of Tumor Cells From Bone Marrow Samples.
Normal bone marrow samples from healthy volunteers were contaminated with serial dilutions of lymphoma cell lines DHL and RL, both of which contain the bc1-2 trans-location. Although both cell lines equally expressed the identical targeted antigens, there were marked differen-ces in the susceptibility of the cell lines to complement mediated lysis. These differences were maintained whether or not the anti-B4 (anti-CD19) monoclonal anti-body was added to the anti-B1, -B5 and -J5 combination of antibodies. However, when anti-B1, -B5 and -J5 were used in conjunction with magnetic microspheres, the bone marrow samples were purged to PCR negativity in a signif-W094/n8621 21 ~7 0 ~ PCT/US93/09891 icantly more efficient ~n~er regardless of whether themarrow was contaminated with the DHL or RL cell lines.
The positive results obtained using normal bone marrow deliberately cont~minated with lymphoma cells encouraged further laboratory testing using aliquots of bone marrow contA; ni ng tumor cells. The tumor cell cont~;n;ng marrow was harvested from five non-Hodgkin's ly~pho~ patients. Aliquots from all patients were first treated with complement and the monoclonal antibodies anti-Bl, -B5 and -J5. After three rounds of treatment as described above, aliquots from two of the five patients reached PCR negativity. Additional aliquots were then treated in the same ~nner~ but with the anti-B4 mono-clonal antibody added to the combination. The same results, two of five aliquots reaching PCR negativity, were obtained. The aliquots reaching PCR negativity were from the same source in both tests.
Following this series of tests with compliment and a combination of either three or four monoclonal antibod-ies, further laboratory tests were conducted using tumorcell containing bone marrow aliquots from five patients and magnetic microspheres with a combination of three (anti-Bl, -B5 and -J5) and four monoclonal antibodies (anti-Bl, -B5, -J5 and -B4). In contrast to the results obtained above using complement induced lysis, marrow - samples from all five patients were purged to PCR nega-tivity after three rounds of treatment. PCR negativity was attained using both three and four monoclonal antibodies. These results suggest that there are differences in the susceptibility of lymphoma cells to complement mediated lysis and that these differences may be overcome or levelled out by the use of a combination monoclonal antibodies and magnetic microspheres instead of complement plus a combination of monoclonal anti-bodies.

Example 2. Further Comparison of Complement Induced Lysis And Magnetic Sphere Depletion W094/0861l 2 ~ 4 7 0 ~ O pa~us~3~o989l Of Tumor Cells From The Bone Marrow Samples Of Lymphoma Patients.
Combinations of three (anti-Bl, -B5 -and -J5) and four (anti-Bl, -B4, -B5 and -J5) monoclonal antibodies were prepared. Harvested bone marrow from ly~phl~ ~
patients was obtained and divided. One portion of the marrow was purged of tumor cells using complement and the three monoclonal antibody combination as previously described. The mononuclear cell fraction from the rem~n~er of the harvested marrow was isolated by Ficoll-Hypaque density centrifugation and divided into four fractions. The four fractions were treated separately.
Two fractions of the centrifuged marrow were lncubated at about 4C for about 30 minutes, with a saturating concen-tration of three monoclonal antibodies and two were incu-bated with four antibodies. The antibody incubation time may be in the range of about 15 minutes to about 1 hour.
Following incubation with the monoclonal antibodies, a three antibody and a four antibody incubate were depleted of tumor cells by complement lysis using a sufficient concentration of rabbit complement from the serum of 3-4 week old rabbits (Pel-freeze, Brown Deer, WI). The cells were incubated with complement at about 37C for a time in the range of about 15 minutes to about 1 hour, preferably about 30 minutes. After complement treatment, the cells were washed in media containing deoxyribonuclease (Sigma, St. Louis, MO) at a final concentration of 2.5 mg/ml. This prevented cell clumping and ensured that DNA from the lysed cells was not co-purified during DNA extractions.
The remaining three and four antibody incubates weredepleted by magnetic separation using goat anti-mouse Ig conjugated magnetic beads. Goat anti-mouse IgG and IgM
conjugated magnetic beads (Advanced Magnetics, Cambridge, MA) were mixed together, washed twice according to direc-tions and added at6a final concentration of 100 micro-liters (~L) per lO antibody coated cells. The cells were incubated with the magnetic beads at 4C for a time WO94/08621 PCT/US93/09891 ~
~4~ ~ -28-in the range of about 15 minutes to about 1 hour, prefer-ably about 30 minutes. After incubation, the magnetic beads were collected in two successive procedures for about 15 minutes using a magnetic particle concentrator (Dynal, Great Neck, NY).
After complement mediated lysis or magnetic bead depletion, the cells were pelleted by centrifugation and the procedures were twice repeated for a total of three treatment cycles. Before and after each of the three treatment cycles, an aliquot of cells was collected from each fraction and genomic DNA was isolated by cell ly818 with non-ionic detergents and proteinase K (Sigma, St.
Louis, MO). Nested oligonucleotide amplification was performed at the MBR or mcl region of the bcl-2/Ig hybrid gene using the method described above. The PCR
results for 19 samples are shown in Table 2.
Table 2. Comparative PCR Results Sample # 3 mAb + C 4 mAb + C 3 mAb + MB 4 mAb + MB
Cycle No- 1 2 3 1 2 3 1 2 3 1 2 3 1 + + + + + +
2 + + - + + _ + _ _ +
3 + + + + + + + + _ +
4 + + + + + - + + _ +
+ + - + + - + _ _ +
6 + + + + + +
7 + + - + + - + _ _ +
8 + + - + + - + + _ +
g + + + + + _ + + _ + +
+ + + + + + + + -- + +
11 + + + + + - + _ _ +
12 + + + + + + + + - + +
13 + + - + + - + + - + NT -14 + + - + + - + + _ +
+ + + + + - + _ _ +
16 + + - + + _ + _ _ +
17 + + + + + + + _ _ +

19 + + -- + + -- + -- _ +

W O 94/08621 2 1 ~ ~ O 1 0 PC~r/US93/09891 (+) = PCR positive (-) = PCR negative NT = Not Tested The results shown in Table 2 indicate that using magnetic microspheres to deplete antibody coated tumor cells ~h~nces the purging of tumor cells from bone marrow samples and results in more samples being purged to PCR negativity. The PCR assay indicates that 50% of the samples treated with complement plus three monoclonal antibodies can be purged to PCR negativity after three rounds of treatment. These results suggested that there must exist a subpopulation of tumor cells that is resistant to complement mediated lysis and that an alternative method might be more efficient in removing tumor cells. The addition of an anti-B4 monoclonal antibody to complement plus three monoclonal antibodies increased the number of samples which could be purged to PCR negativity; but neither of complement mediated lysis methods achieved results equaling those achieved using the magnetic microspheres. Incubation of tumor containing marrow samples with three or four monoclonal antibodies followed by depletion using magnetic microspheres resulted in all samples purging negative.
The magnetic microspheres used according to the invention may be any magnetic microspheres such as magnetic polystyrene latex spheres, magnetic dextran or gelatin microspheres and similar microspheres. As used herein, the term microspheres or immunologically accept-able substrate includes magnetic or non-magnetic parti-cles of suitable size and of any shape; for example,round, cubic, rectangular, or other shapes. The magnetic particles may be enclosed by a coating material such as poly~ylene~ or gelatin or the particles may be embedded in the coating. The size of the magnetic microspheres used in the invention may range from about 0.3 microns to about 5 microns. Non-magnetic equivalents to the mag-netic species described herein may also be used according to the invention. Appropriate separation techniques are ~ 30-selected when such non-magnetic particles are used.
In addition to using immmunoglobulin coated magnetic microspheres to deplete tumor cells saturated with mono-clonal bodies as described in Example 2, the monoclonal antibodies used in the invention may be attached to the microspheres by techni~ues known in the art before the microspheres are mixed with the bone marrow samples. For example, the monoclonal antibodies may be attached to immunoglobulin coated microspheres prior to mixing the microspheres with the tumor cell containing marrow.
Another example is the use of a carboxylate-diamine couple as a bridge between the antibody and the micro-sphere. If the microsphere has pendent groups such as an amine, a carb~xy group or a thiol group, the antibody may be attachable to the microsphere without further functionalization of either species.
The following examples are given to illustrate the further embodiments of the invention and are not intended to be limiting. Variations in the composition of the magnetic microspheres and the methods of attaching the combination of monoclonal antibodies to the microspheres are deemed within the scope of the invention. These include the methods described in the following U.S.
Patents whose teaching are incorporated herein by reference: Patent Nos. 4,152,563; 4,253,844; 3,639,558;
4,~52,773; 4,452,773; 3,639,558; 4,419,444; 4,738,932;
4,360,358 and 4,414,324. Additional teachings incorpor-ated by reference may be found in PCT International Publication W0 90/04178.
Example 3. Preparation Of T~u~oglobulin Coated Magnetic Microspheres Having A Combination Of Selective Monoclonal Antibodies.
Non-porous magnetic mlcrospheres (sometimes called beads) of generally monodispersed or uniform size in the range of about 0.3 to about 5.0 mlcrons are conditioned for bi n~i ng of a combination of monoclonal antibodies thereon by pre-coating the microspheres with rabbit or goat anti-mouse immunoglobulin (GAM or RAM) as follows.

W094/08621 2 1 ~ 7 0 1 0 PCT/US93/09891 First, 250 mg of beads are dispersed in 3 ml of distilled water and sonicated for 2-3 minutes. The mixture of beads in water is then cooled for several hours at 4C.
The beads are then magnetically separated and the water discarded. The beads are resuspended in 5mg of RAM or GAM and diluted with 500 microliters of phosphate-buffered saline (PBS). This mixture is incubated at room temperature and is mixed for 4-5 hours. Thereafter the beads are washed six times with 4 ml portions of PBS
contA;n;ng l~ bovine serum albumin (BSA). The beads are then resuspended in 4 m8 f PBS-1% BSA and mixed.
A quantity of 5xlO of the resuspended beads is pipetted into a siliconized test tube. The beads are magnetically separated from the liquid which is dis-carded. The beads are then resuspended in PBS and thecombination of three or four selective monoclonal anti-bodies as described above is added to the suspension such that the concentration of antibody combination totals approximately 0.5 mg/ml of solution. The solution cont~n~ng beads and antibodies is then incubated at a temperature in the range of 10-30C, for about one hour.
The beads are then washed a plurality of times with PBS-l~ BSA and resuspended in the same.

Example 4. The Removal Of Tumor Cells From Bone Marrow Using A Combination Of Selective Monoclonal Antibodies On Immunoglobulin Coated Magnetic Microspheres Bone marrow contA1n~ng tumor cells are harvested, concentrated, washed, fractionated and suspended in RPMI
medium as described above for complement mediated lysis.
The antibody containing magnetic beads prepared according to Example 3 are then added to the marrow suspension and the resulting mixture is incubated for a time in the range of about lO minutes to about l hour, preferably for about 15 minutes, at a temperature of about 4C. The magnetic spheres are then removed, and the cells are WO94/08621 PCT/US93/0~891 -~ 32-washed and pelleted. The procedure is twice repeated for a total of three treatments. The bone marrow cells are then washed and tested for tumor cells by PCR.
DEPLETION OF TUMOR CELLS USING MICROSPHER~S
Using the extremely sensitive technique of polymerase chain reaction to detect the bc1-2 translocation, it was found that the use of complement mediated lysis and three monoclonal antibodies (CmAb-3) would purge only 50~ of different bone marrow samples of PCR detectable lymph~m~ cells. This observation is of clinical importance because those patients whose re-infused bone marrows harbored residual lymphoma cells ~ ctrated a significantly increased incidcnce of relapse. In view of the fact that only about 50~ of patients' marrow could be purged to PCR negativity using CmAb, microsphere depletion of tumor cells was investi-gated further. Microsphere use produces a significant i...~ovement of unexpected magnitude in tumor cell deple-tion. Using microspheres and either three or four mono-clonal antibodies (MmAb-3 or MmAb-4), all PCR detectable tumor cells were purged from aliquots of twenty-five (25) different tumor cells cont~ ng bone marrow samples.
When different aliquots of these same 25 samples were treated wlth complement and three monoclonal antibodies (CmAb-3), only ll out of 25 (44~) purged to PCR negativ-ity. The addition of a fourth monoclonal antibody followed by complement mediated lysis (CmAb-4) purged the marrow of an additional five (5) patients, bringing the total to 16 out of 25. The use of MmAb also was found to be specific because there was no loss of committed mye-loid progenitor cells. The microsphere results suggests that the use of microspheres will be superior to comple-ment mediated lysis, and the lack of non-specific toxicity to myeloid progenitor cells predicts that microsphere use will not impair engraftment. The follow-ing describes further testing using microspheres and monoclonal antibodies to remove tumor cells from bone marrow samples.

WO94/08621 21~ 7 ~1 ~ PCT/US93/09891 Materials and Methods.
Bone marrow was obtained from twenty-five patients with B cell NHL after Human Protection Committee validation and informed consent. Immunotyping documented that all lymphomas expressed CD20. All patients had achieved a protocol eligible minimal disease level following induction or salvage chemotherapy at the time of bone marrow harvest. M1 n~ rAl disease criteria includes either a complete remission (CR) or a partial remission (PR) to tumor masses of 2 cm or less and marrow infiltration of less than 20% of the intertrabecular space. Bone marrow aspirates and biopsies were obtained one month before marrow harvest to assess lymphomatous infiltration and to confirm that the bc1-2 translocation could be detected by PCR amplification. All samples contained a PCR detectable bc1-2 translocation.
Patient characteristics are summarized as follows.
Twenty patients had follicular small cleaved cells, one patient had follicular mixed small and large cells and four patients had diffuse small cell histology. Nlne patients achieved CR following induction or salvage chemotherapy and the re~Ai n1 ng patients achieved a protocol eligible PR as defined herein. Fifteen patients had a prior history of morphologic bone marrow infiltra-tion ranging from local infiltrates to 90% of the inter-trabecular space. Eleven patients had morphological evidence of bone marrow infiltration at the time of bone marrow harvest. All twenty-five patients had residual detectable lympho~ cells in their harvested bone marrow when it was assessed by PCR. Nineteen patients had a a translocation involving the major breakpoint region (MBR) of bc1-2 and six patients had a translocation at the minor cluster region (mcr).

Microspheres The term microspheres or beads as used herein is descriptive of a particle of 0.1 to 5 microns in its W094/08621 2~ 4~ ~ a ~ PCT/US93/09891 -ma~or dimension and is inclusive of spherical, cubic, and rectangular particles as well as particles having other shapes. Preferred particles are about 0.3 to 5.0 microns. The particles may be made of a polymeric mater-ial such as polys~yLene, polyacrylate, polymethacrylate,polyester, a styrene-divinylbenzene copolymer. poly-phenylene oxide and other polymers, copolymers or sub-qtances such as gelatin, dextran or an aminodextran, or may be made of a substance which is coated with these polymers, copolymers or substances.
The microspheres may be magnetic or non-magnetic.
Magnetic microspheres are preferred for their ease of separation. Magnetic mlcrospheres may be polymeric particles which have a magnetic material embedded ln a polymer matrlx or they may comprise a magnetic nucleus or core having a polymeric coating. An examples of gelatln coated particles may be found ln U.S. Patent No.
5,062,99l.
The magnetic beads used to treat the twenty-five patients' bone marrow were purchased from Advanced Magnetics, Cambridge, Massachusetts and consisted of goat anti-mouse (GAM) IgG and IgM conjugated magnetic beads.
Their use herein is not to be understood as limiting the invention. Any beads of type described above may be used according to the invention. IgG and IgM coating of uncoated beads is well known in the art.

Monoclonal Antibodies The anti-B cells monoclonal antibodies have been previously described herein. The three antibody combin-ation (m Ab-3) was a mixture of murine anti-Bl, anti-B5 and anti-J5 monoclonal antibodies. The four antibody combination (mAb-4) was formed by a~d;n~ an anti-B4 mono-clonal antibody to the three antibody combination.
Excess saturating concentrations of the monoclonal anti-bodies were used.
The anti-T cell monoclonal antibodies used as controls were specific to CD2, CD3, CD4 and CD28, and W O 94/08621 21 4 7 0 1 0 PC~r/US93/09891 were obtained from Dr. C. Morimoto, Dana-Farber Cancer Institute, Boston, Massachusetts. Other anti-T cell monoclonal antibodies having the indicated CD
specificities may be used in their place. The anti-T
cell monoclonal antibodies were isotype matched to the anti-B cell monoclonal antibodies. Excess saturating concentrations of the antibodies were used.

Bone Marrow Purging Aliquots of the harvested bone marrow were purged using three monoclonal antibodies and complement mediated lysis (CmAb-3) as described herein. These CmAb-3 purged samples were used as reference samples for determining the utility of the microsphere procedure. Unless other-wise specified, small aliguots of the marrow samples were used in all tests. The bulk of the marrow samples were designated as clinical samples and were reserved for use as indicated herein.
Additional aliquots of bone marrow were purged using the MmAB technique which follows a specific sequence. In a single purging cycle, the aliquots were first incubated with excess saturating concentrations of the three anti-bodies for 30 minutes at 4C. The cells coated with the monoclonal antibodies were then depleted using a twice washed mixture of GAM IgG and IgM con;ugated magnetic beads to which the antibodies also conjugated. The 6eads were added to the cells at the rate of 100 ~L per 10 cells. This provides a bead-to-cell ratio of about 50.
Since the monoclonal antibodies used herein target normal B cells as well as lymphom~ cells, the estimated bead-to-tumor cell ratio was in the range of about 250-500. The samples were incubated with the beads for 30 minutes at 4C. After incubation, the magnetic beads were collected for 15 minutes using a magnetic particle concentrator (Dynal, Great Neck, New York) or other means of magnetic separation. A depleted bone marrow cell sample was transferred to a clean vessel and any residual magnetic beads contained therein were collected during a second 15 WO 94/08621 PCr/US93/09891 . ~
2I470~ -36-minute magnetic separation.
After the first CmAb or MmAb depletion, the treated bone marrow cells were pelleted by centrlfugation, washed three times and the tumor cell depletlon or purging repeated for a total of three cycles. Before the first depletion cycle and after each depletlon cycle, a sample of cells was collected and genomic DNA was isolated by cell lysis with non-ionic detergents followed by treatment with proteinase K (Sigma, St. Louis, Missouri).

10 PCR Amplification.
PCR amplification of any tumor cells present in a sample was conducted as described elsewhere herein. In the present analysis, 50 ~L of samples were used and the cell line RL was used in place of DHL-6. The cell line RL was a gift from Dr. W. Urba, National Cancer Institute, Bilogic Response Modifiers Branch, Frederick, Maryland. RL has a PCR amplifiable translocation at the major breakpoint. Other cell lines which have a similar translocation may be used in place of RL. No cell line was available with a translocation at the minor break-point region. PCR amplification was also performed using oligonucleotides for the human B cell activation antigen B7 in order to confirm that extracted DNA could be amplified by PCR in all samples.

Colony Assays Bone marrow samples from eight of the twenty-five patients were assayed for hematopoietic progenitor cell growth both before purging and after the third purging cycle uslng the CmAb-4 and MmAB-4 procedures. The samples were blinded from the person who was conducting the colony assays. Granulocyte-macrophage colony-formlng units (CFU ) were assayed by a standard procedure using a two layer agar assay [48]. Bladder carcinoma cell line 5637 (ATCC, Rockville, Maryland) conditioned medlum wa~
used at a final concentration of 10% for the s4ource of colony stimulating factors. A total of 5 x 10 bone WO94/08621 2 1 4 ~ O 1 0 PCT/US93/09891 marrow ~onon~clear cells from each of the marrow frac-tions were plated into each well in the overlayer of the double layer agar system. Quadruplicate CFU cultures were harvested on days 7 and 14, and 0.3~ agar overlayers were dried onto glass slides and stained with Gill's hematoxylin (Sigma, St. Louis, Missouri). CFU were counted on4days 7 and 14, and expressed as total colonies pe4r 5 x lO cells plated and as total colonies per 5 x lO original "pre-depletion bone marrow mononuclear cells.
Evaluation In the present investigation, bone marrow having PCR
detectable tumor cells was obtained from twenty-five patients and was treated using the MmAB procedure to determine if an i~ ovement could made in the number of samples which could be depleted below the level of PCR
tumor cell detection as compared to the CmAB procedure.
The results of the PCR analysis of samples from two representative patients are shown in Figs. 7A-7B. As shown in Fig. 7A, patient 6 had residual PCR detectable tumor cells following three cycles of CmAb-3 treatment.
However, all PCR detectable tumor cells were removed after three rounds of CmAb-4 treatment. Depletion using MmAB-3 and MmAb-4 resulted in the removal of all PCR
detectable tumor cells after two rounds of treatment.
Fig. 7C shows the results obtained by PCR amplification using oligonucleotides for the B7 gene of this patient.
The results confirm that PCR amplifiable DNA was extracted from each of the samples.
As shown in Fig. 7B, patient 7 had no PCR detectable lymphoma cells following the third round of treatment using CmAB-3 or CmAb-4. When analogous MaAb treatment was used, no PCR detectable tumor cells were found after the second treatment cycle.
The purging of bone marrow samples using microspheres appears to be specific and does not result in the loss of antibody-free cells through conjugation such cells to the microspheres. In repetitive experi-WO94/08621 214~010 PCT/US93/09891 ments, when bone marrow samples were treated through three cycles using anti-T cell monoclonal antibodies and GAM coated microspheres, there was never a loss of PCR
detectable tumor cells through non-specific binding of the tumor cells to the microspheres. Furthermore, the use of microspheres did not effect the sensitivity of PCR
amplification. As the results shown in Fig. 8 indicate, PC6 analysis is capable of detecting one lymphoma cell in normal bone marrow cells.
Table 3 gives the results of the PCR amplification at the ma~or and minor breakpoint regions of the bc1-2 translocation after each of the three CmAb-3, CmAb-4, MmAb-3 and MmAb-4 treatment cycles for the bone marrow samples from all twenty-five patients. After the third cycle of CmAb-3 treatment, only 11 out of 25 samples (44%) had no PCR detectable lymphoma cells. In each of these 11, three cycles was required to reach this state.
Identical results were obtained using the bulk (clinical) marrow samples and aliquots for 24 of the 25 patients.
For patient 11, the clinical sample contained residual detectable ly~ph~.~ cells while the aliquoted fraction was purged of all PCR detectable cells. This suggests that the results obtained using small bone marrow aliquots are representative of the whole harvested bone marrow sample.

WO94/08621 2 1 ~ 7 ~ 1 0 PCT/US93/09891 Table~3. Comparative PCR Results Patlent # 3 mAb + C 4 mAb + C 3 mAb + MB 4 mAb + MB
Cycle No- 1 2 3 1 2 3 1 2 3 1 2 3 + + + + + + + + _ + +
2 + + - + + _ + _ _ +
3 + + + + + + + + _ +
4 + + + + + - + + _ +
+ + - + + - +
6 + + + + + - + _ _ +
7 + + - + + - + - - +
8 + + - + + - + + _ +
g* + + + + + _ + + _ + +
+ + + + + + + -- -- +
11 + + -- + + -- + -- -- +
12 + + + + +
13 + + - + + - + + _ +
14 + + - + + _ + + _ +
+ + - + + _ + + _ +
16 + + - + + - + - - +
17* + +
18 + + + + + + + + - + +
19 + + -- + + -- +
+ + - + + - + + - +
21 + + + + + +
22 + + + + + +
23 + + + + + + + + _ +
24* + + + + + - + _ _ +
+ + + + + + + + _ + +

(+) = PCR posit$ve (_ = PCR negative CmAb-4 treatment of marrow samples increased the number samples havlng no PCR detectable tumor cells after three cycles to 16 out of 25 (64%). The five additional patients whose samples could be purged of residual lymph-oma cells by four, but not three, monoclonal antibodiesare indicated by an asterisk after the patient number.
However, regardless of whether three or four monoclonal ~7~ 40-antibodies were use, three CmAb treatment cycles were required before any of the samples were purged of PCR
detectable tumor cells. In view of the number of patients studied, there does not seem to be a statis-tically significant advantage in adding a fourth mono-clonal antibody to the CmAb treatment procedure (p =
0.256, Fisher exact test).
An attempt was made to correlate patient character-istics with purging success. No associations were found between the patients' clinical characteristics (his-tology, status at ABMT [Cr versus PR], history of pre-vious bone marrow involvement or bone marrow involvement at the time of harvest) and the ability to purge the bone marrow of all PCR detectable cells using CmAb-3 or CmAb-4.
The results obtained using microspheres were significantly different and better than those obtained using complement mediated lysis. Following three treatment cycles using microspheres. all 25 patient samples could be depleted of PCR detectable lymphoma cells using either three or four monoclonal antibodies.
Statistically, the improvement was p = O.OOOl using three antibodies and p = 0.0016 using four antibodies. After only two MmAb-3 treatment cycles, ll patient samples (44%) were purged of PCR detectable cells. After only two treatment cycles using MmAb-4, 20 patient samples (80%) were PCR negat~ve for tumor cells. Tha use of four monoclonal antibodies was significantly more efficient in removing residual lymphoma cells after the second treatment cycle (p = 0.0186) that was the use of three monoclonal antibodies at the same point in CmAb. Those patients whose marrow contained residual PCR detectable lymrhomA cells after two MmAb-3 treatment cycles, but not after two MmAb-4 cycles are indicated by the symbol ^^ in Table 3. Significant in the results shown in Table 3 is that four patients' marrow, numbers 5, 14, l9 and 21, showed no PCR detectable tumor cells after a single MmAb-4 treatment. Patient 21's sample is especially signifi-W O 94/08621 ' 2 1 ~ 7 Q 1 ~ PC~r/US93/09891 cant because it could not be depleted of PCR detectable cells using either CmAB-3 or CmAb-4.
A concern throughout the tests was that the micro-sphere purging procedure might lead to an unacceptable loss of bone marrow mononuclear cells. The number of ~o~onl~clear cells from the patient9s which were treated was in the range of 6.44 ~ 1.23 10 . Following CgmAb-3 treatment, the clinical ~ample had 4.4 1 1.2 x lO cells, a recovery of 68%. Cell recovery from the evaluation aliquots using CmAb-3 was 72%. The two values closely agree. The use of microspheres seems to reduce non-specific cell loss relative to the use of complement.
Cell recovery was 78% following MmAb-3 and 82% following MmAb-4. In all cases, CmAb or MmAb, about 5% of total cells were removed for analyses during the overall purg-ing cycles. Therefore, maximum recovery is 95%. Flow cytometry analyses indicate that there is no loss of B
lineage cells after either CmAb or MmAb (data not shown).
An additional concern was that the use of micro-spheres or complement mediated lysis might result in the loss of hematopoietic progenitors from the bone marrow.
In parallel experiments using CmAb-4 or MmAb-4 through three treatment cycles, colony assays were established as described for patients 18-25. The results for three patient samples, corrected for cell loss, are shown in Table 4. The results indicated that4there was no significant loss of CFUGM per 5 x lO cells at either day 7 or 14 following microsphere depletion compared to an undepleted marrow sample (Pre) or CmAb. When the nu4mbers of colonies were corrected and expressed per 5 x 10 "pre-depletion" bone marrow mononuclear cells, there was no significant loss (NS) in CFU at day 7 or 14 for MmAb depleted samples relative to the "Pre" samples. However, corrected CmAb samples did show significant loss in CFU
at days 7 and 14 for 7 of the eight patients studied. It was noted, however, that though the was a significant loss in CFU at days 7 and 14 following CmAb for 7 of GM
the 8 patients, all these patients engrafted rapidly W 094/08621 21~7 01~ PC~r/US93/09891 -following re-infusion of their marrow after high dose therapy.

Table 4. Granulocyte-Marcophage Colony ~orming Unlts (CFU
5 Patient# Day# Pre Complement Microspheres 21 7 247 1 50 212 1 27 NS 179 ~ 9 NS
14 69 ~ 11 50 1 7 p < 0.05 58 l 8 NS

23 7 284 1 28 198 1 13 p < 0.025 236 119 NS
14 29 ~ 8 23 1 3 NS 23 12 NS

7 234 1 26 171 1 14 p < 0.01219 ~ 34 NS
14 11 ~ 3 7 1 1 p < 0.05 9 1 1 NS
The results shown herein indicate that tumor cell depletion using a plurality of monoclonal antibodies and microspheres is significantly more efficient than the procedure using a plurality of monoclonal antibodies and complement mediated lysis. The results also indicate that three rounds of MmAb depletion is required, that re-infusion of the MmAb depleted marrow may produce superior results compared to CmAb depleted marrow and that the lack of non-specific toxicity to myeloid progenitor cells in an indication that microsphere use will not impair engraftment.
It is belleved that the failure of CmAb purglng to remove all tumor cells can be explained and attributed to three possible mechanisms. In the first, the clonogenic tumor cells might not express the surface antigens expressed by a ma~ority of tumor cells. In the second, modulation of one or more of the surface antigens following monoclonal antibody attachment to its ligand W O 94/08621 2 1 4 7 0 1 0 PC~r/US93/09891 might limit complement mediated lysis. In the third, a subgroup of patients may have lymphomas which are intrinsically more resistant to complement mediated lysis. Since, when the same number of monoclonal anti-bodies was used in both procedures, the microsphereprocedure was capable of depleting residual lymphoma cells in all cases whereas complement mediated lysis procedure failed to completely remove tumor cells, the third mechanism is believed the most likely. A number of possible mech~n~mc for cell resistance to the lytic effects of complement have been described in the tech-nical literature. Associations have been shown between biochemical events in the cell and sensltivlty to comple-ment medlated lysls ~49, 50]. The presence of an anti-complement factor ln human bone marrow has been demon-strated [51]. Lastly, lt ls also posslble that some lymphoma cell~ may express such a low density of the target antigens that elimination is feasible by use of microspheres, but complement mediated lysis is not possible ~52, 53].
While there is controversy as to whether re-infusion of lympho~ cells in autologous marrow contributes to subsequent relapse after ABMT, the CmAb results presented herein indicate that there is an increased incidence of relapse in those patients whose autologous marrow contained PCR detectable lymphoma cells after purging.
However, this observation does not prove that re-infused lym--phnm~ cells contribute to relapse. Endogenous disease is likely to contribute to relapse in some patients, even after re-infusion of syngeneic marrow. While some have argued that either a r~n~om;zed trial or gene marker study is necessary to delineate the relative contribution to relapse made by endogenous disease versus re-infused tumor cells, the approach of the claimed invention is that if the relapse rate significantly decreases when tumor-free marrow is re-infused, this is a strong indi-cator that marrow derived cells contribute to relapse.
The long-term CmAb results shown in Fig. 3 indicates that 21~7010 ~44~

disease-free survival after ABMT correlates strongly with whether the patient"s marrow purged PCR n~gative or positive. In the group that purged negative, only 4 out of 57 patients relapsed after as long as 8 years. In contrast, out of 57 patients whose marrow was PCR
positive after CmAb, 26 relapsed and the median disease-free survival time was l9.7 months. As a consequence of the CmAb results, it is believed that any ~ ~lovement in removing tumor cells, such as that described herein using microspheres, may increase the number of patients experiencing disease free survival and/or tha length of survival time.

Example 5. The Removal Of Tumor Cells From Bone Marrow Using A Combination Of Selective Monoclonal Antibodies On Gelatin - Coated Magnetic Microspheres Magnetic microspheres are prepared according to U.S.
Patent No. 5,062,991 and monoclonal antibodies are conjugated there to using bridging groups 1-20 atoms long. Bone marrow cont~ining tumor cells are harvested, concentrated, washed, fractionated and suspended in RPMI
medium as described above for complement mediated lysis.
The antibody containing magnetic beads are then added to the marrow suspension and the resultiny mixture is incubated for an experimentally determined time, possibly lO minutes to about 2 hours, at a temperature of about 4C. The magnetic spheres are then 1~ ~ved, and the cells are washed and pelleted. The procedure is twice repeated for a total of three treatments. The bone marrow cells are then washed and tested for tumor cells by PCR. Alternatively, the magnetic microspheres prepared according to Patent No. 5,062,99l may be coated wlth GAM and used as described in Example 2.

We claim:

Claims (25)

IN THE CLAIMS

1. A method for immunologically purging tumor cells from the bone marrow of a patient having B cell non-Hodgkin's lymphoma, said method comprising treating said marrow with a plurality of selective anti-B cell monoclonal antibodies and microspheres in a specified sequence to deplete said marrow of tumor cells, without depletion of non-tumor cells, to a level where tumor cells are not detectable in said marrow by polymerase chain reaction assay; said plurality of monoclonal antibodies including:
(1) an anti-B5 monoclonal antibody specific to a 75,000 dalton molecular weight antigen on activated B
cells and B cell lymphomas, (ii) an anti-CD20 monoclonal antibody, and (iii) an anti-CD10 monoclonal antibody specific against the common acute lymphoblastic leukemia antigen.

2. The method of claim 1 wherein said microspheres are immunoglobulin coated microspheres of a size in the range of 0.1 microns to 5 microns.

3. The method of claim 2 wherein said immunoglobulin is goat anti-mouse immunoglobulin

4. The method of claim 1 wherein an anti-CD19 monoclonal antibody is added to the monoclonal antibodies described therein

5 (Amended). A method for immunologically purging tumor cells from the bone marrow of a patient having B
cell non-Hodgkin's lymphoma, said method comprising:
(a) incubating said marrow with a plurality of monoclonal antibodies in tumor treating sufficiency quantities at a selected temperature for a time in the range of 10-30 minutes, said plurality of antibodies consisting of:
(i) an anti-CD20 monoclonal antibody, (ii) an anti-CD10 monoclonal antibody and (iii) an anti-B5 monoclonal antibody specific to a 75,000 dalton molecular weight antigen on activated B
cells and B cell lymphomas;
(b) incubating the product of step (a) with immunoglobulin coated microspheres at a temperature of about 4°C for a time in the range of 15 minutes to one hour;
(c) separating the magnetic microspheres from the product of step (b); and (d) repeating the steps (a), (b) and (c) a plurality of times to deplete the tumor cells in said marrow, without depletion of non-tumor cells, to a level where the tumor cells are not detectable by polymerase chain reaction assay.

6. The method of claim 5 wherein an anti-CD19 monoclonal antibody is added to the monoclonal antibodies used in step (a).

7. The method of claim 5 wherein said microspheres are of a size in the range of 0.1 to 5.0 microns.

8. The method of claim 5 wherein said immunoglobulin is goat anti-mouse immunoglobulin.

9. A method for treating a patient having B cell non-Hodgkin's lymphoma by autologous bone marrow transplantation, said method comprising:
(a) collecting bone marrow from said patient;
(b) treating said marrow with a plurality of selective anti-B cell monoclonal antibodies and microspheres in a specified sequence to deplete said marrow of tumor cells, without depletion of non-tumor cells, to a level where tumor cells are not detectable by polymerase chain reaction assay; and said plurality of monoclonal antibodies including:
(i) an anti-B5 monoclonal antibody specific to a 75,000 dalton molecular weight antigen on activated B
cells and B cell lymphomas, (ii) an anti-CD20 monoclonal antibody, and (iii) an anti-CD10 monoclonal antibody specific against the common acute lymphoblastic leukemia antigen;
and (c) transplanting the treated marrow back into the patient from whom it was obtained.

10. The method of claim 9 wherein said microspheres are immunoglobulin coated microspheres of a size in the range of 0.1 microns to 5 microns.

11. The method according to claim 10 wherein said immunoglobulin is goat anti-mouse immunoglobulin.

12. The method of claim 9 wherein an anti-CD19 monoclonal antibody is added to the monoclonal antibodies described therein.

13. A method for immunologically purging tumor cells from the bone marrow of a patient having B cell non-Hodgkin's lymphoma, said method comprising treating said marrow with a plurality of selective anti-B cell monoclonal antibodies con;ugated to microspheres to deplete said marrow of tumor cells, without depletion of non-tumor cells, to a level where tumor cells are not detectable by polymerase chain reaction assay; said plurality of monoclonal antibodies including:
(i) an anti-B5 monoclonal antibody specific to a 75,000 dalton molecular weight antigen on activated B
cells and B cell lymphomas, (ii) an anti-CD20 monoclonal antibody, and (iii) an anti-CD120 monoclonal antibody specific against the common acute lymphoblastic leukemia antigen.

14. The method of claim 13 wherein said microspheres are magnetic microspheres of a size in the range of 0.1 microns to 5 microns and are coated with a substance selected from the group consisting of immunoglobulins, gelatins, dextrans and aminodextrans.

15. The method according to claim 14 wherein said immunoglobulin is goat anti-mouse immunoglobulin.

16. The method according to claim 13 wherein an anti-CD19 monoclonal antibody is added to the monoclonal antibodies described therein.

17. A method of treating by autologous bone marrow transplantation a patient having B cell non-Hodgkin's lymphoma, said method comprising:
(a) obtaining bone marrow from said patient;
(b) incubating said marrow with magnetic microspheres to which are conjugated a plurality of anti-B cell monoclonal antibodies, said incubation being at a temperature of about 4°C for an experimentally determined time and said plurality of monoclonal antibodies consisting of:
(i) and anti-CD20 monoclonal antibody;
(ii) an anti-CD10 monoclonal antibody and (iii) an anti-B5 monoclonal antibody specific to a 75,000 dalton molecular weight antigen on activated B
cells and B cell lymphomas;
(c) separating the magnetic microspheres from the product of step (b);
(d) repeating steps (a), (b) and (c) a plurality of times to deplete the tumor cells in said marrow, without depletion of non-tumor cells, to a level where the tumor cells are not detectable by polymerase chain reaction assay; and (e) transplanting the treated marrow back into the patient from whom it was obtained.

18. The method according to claim 17 wherein an anti-CD19 monoclonal antibody is added to the monoclonal antibodies used in step (a).

19. The method of claim 17 wherein said microspheres are of a size in the range of 0.1 to 5.0 microns

20. The method of claim 17 wherein said microspheres are coated with a substance selected from the group consisting of immunoglobulins, gelatins, dextrans and aminodextrans, and said monoclonal antibodies are conjugated to said coating.

21. The method of claim 20 wherein said immunoglobulin is goat anti-mouse immunoglobulin.

22. Particles for removing tumor cells from the bone marrow of a patient having B cell non-Hodgkin's lymphoma, said particles comprising microspheres having a plurality of anti-B cell monoclonal,antibodies conjugated thereto, and said particles being capable of depleting tumor cells without depletion of non-tumor cells from said marrow and said plurality of monoclonal antibodies including:
(i) an anti-B5 monoclonal antibody specific to a 75,000 dalton molecular weight antigen on activated B
cells and B cell lymphomas;
(ii) an anti-CD20 monoclonal antibody, and (iii) an anti-CD10 monoclonal antibody specific against the common acute lymphoblastic leukemla antigen.

23. The particles of claim 22 wherein said microspheres are magnetic microspheres of a size in the range of o.3 microns to 5 microns and are selected from the group consisting of immunoglobulin coated microspheres, immunoglobulin coated microspheres having said biological substances bound thereto and non-immunoglobulin coated microspheres having said biological substances attached thereto.

24. The particles according to claim 23 wherein said immunoglobulin is goat anti-mouse immunoglobulin.

25. The particles according to claim 22 wherein an anti-CD19 monoclonal antibody is added to said plurality of monoclonal antibodies.