Associations between KIR/KIR-ligand genotypes and clinical outcome for patients with advanced solid tumors receiving BEMPEG plus nivolumab combination therapy in the PIVOT-02 trial.

Bempegaldesleukin (BEMPEG), a CD122-preferential IL2 pathway agonist, has been shown to induce proliferation and activation of NK cells. NK activation is dependent on the balance of inhibitory and excitatory signals transmitted by NK receptors, including Fc-gamma receptors (FCγRs) and killer immunoglobulin-like receptors (KIRs) along with their KIR-ligands. The repertoire of KIRs/KIR-ligands an individual inherits and the single-nucleotide polymorphisms (SNPs) of FCγRs can influence NK function and affect responses to immunotherapies. In this retrospective analysis of the single-arm PIVOT-02 trial, 200 patients with advanced solid tumors were genotyped for KIR/KIR-ligand gene status and FCγR SNP status and evaluated for associations with clinical outcome. Patients with inhibitory KIR2DL2 and its ligand (HLA-C1) observed significantly greater tumor shrinkage (TS, median change -13.0 vs. 0%) and increased PFS (5.5 vs. 3.3 months) and a trend toward improved OR (31.2 vs. 19.5%) compared to patients with the complementary genotype. Furthermore, patients with KIR2DL2 and its ligand together with inhibitory KIR3DL1 and its ligand (HLA-Bw4) had improved OR (36.5 vs. 19.6%), greater TS (median change -16.1 vs. 0%), and a trend toward prolonged PFS (8.4 vs. 3.6 months) as compared to patients with the complementary genotype. FCγR polymorphisms did not influence OR/PFS/TS.These data show that clinical response to BEMPEG plus nivolumab treatment in the PIVOT-02 trial may be associated with the repertoire of KIR/KIR-ligands an individual inherits. Further investigation and validation of these results may enable KIR/KIR-ligand genotyping to be utilized prospectively for identifying patients likely to benefit from certain cancer immunotherapy regimens.

Antibody landscape of C57BL/6 mice cured of B78 melanoma via immunotherapy.

Hoefges et al. utilized a whole-proteome peptide array approach to show that C57BL/6 mice develop a large repertoire of antibodies against linear peptide sequences of their melanoma after receiving a curative immunotherapy regimen consisting of radiation and an immunocytokine. Antibodies can play an important role in innate and adaptive immune responses against cancer, and in preventing infectious disease. Flow cytometry analysis of sera of immune mice that were previously cured of their melanoma through a combined immunotherapy regimen with long-term memory showed strong antibody-binding against melanoma tumor cell lines. Using a high-density whole-proteome peptide array, we assessed potential protein-targets for antibodies found in immune sera. Sera from 6 of these cured mice were analyzed with this high-density, whole-proteome peptide array to determine specific antibody-binding sites and their linear peptide sequence. We identified thousands of peptides that were targeted by 2 or more of these 6 mice and exhibited strong antibody binding only by immune, not naive sera. Confirmatory studies were done to validate these results using 2 separate ELISA-based systems. To the best of our knowledge, this is the first study of the "immunome" of protein-based epitopes that are recognized by immune sera from mice cured of cancer via immunotherapy.

Recognitive specificity of human cytotoxic T lymphocytes. I. Antigen-specific inhibition of human cell-mediated lympholysis.

The specificity of antigen recognition by in vitro sensitized human cytotoxic T lymphocytes (CTLs) has been studied using a sensitive cell-mediated lympholysis (CML) assay. Frequently, high levels of cytotoxicity are observed on third-party targets unrelated to sensitizing or responding cells; however, no cytotoxicity differing significantly from zero has been observed on targets autologous to the responding CTLs. This "cross-killing" of third-party target cells has been observed when stimulating and third-party cells bear no cross-reacting serologically defined (SD) antigens, thought to be the target antigens recognized by CTLs. CML-blocking studies, using unlabeled normal human lymphocytes to inhibit 51Cr release from radiolabeled target cells, have shown that cross-killing, even in the absence of shared SD determinants, results from CTLs recognizing antigens shared by the third-party targets and the initial stimulating population. Furthermore, these antigens have been mapped to the major histocompatibility complex (MHC). The ability of human CTLs to specifically recognize MHC-controlled antigens not detected serologically suggests that SD antigens may be recognized differently by alloantisera and CTLs, or that MHC antigens other than SD may be the targets of CTLs in CML.

Characterization of the interleukin 2 receptors (IL-2R) expressed on human natural killer cells activated in vivo by IL-2: association of the p64 IL-2R gamma chain with the IL-2R beta chain in functional intermediate-affinity IL-2R.

Interleukin 2 (IL-2) receptors expressed on the surface of activated T cells and natural killer (NK) cells exhibit a variety of affinity states depending on their subunit composition. Low-affinity binding is associated with a 55-kDa alpha chain, intermediate-affinity binding with a 70-75-kD beta chain, and high-affinity binding with a bimolecular complex of the alpha and beta subunits. In a previous study of the IL-2 receptors expressed on NK cells obtained from cancer patients after in vivo IL-2 therapy, we documented a discrepancy between the level of beta chain and the level of intermediate-affinity IL-2 binding sites expressed on the cell surface. Based on this result, we postulated that formation of intermediate-affinity receptor sites required a component in addition to the beta chain, and that this component was present at limiting levels in the patient NK cells. In the present study we have examined the structure of the intermediate-affinity receptor complex using monoclonal antibodies that recognize the beta chain, but that do not interfere with its ability to bind IL-2. Evidence is presented establishing the physical association of a novel protein of 64 kD with the beta chain in intermediate-affinity IL-2 binding sites. This molecule, termed IL-2R gamma chain, coprecipitated with beta chains prepared from cells that had been incubated with IL-2, but was undetectable in immunoprecipitates prepared in the absence of IL-2. Examination of gamma chain expression in post-IL-2 therapy NK cells, where only low levels of intermediate-affinity IL-2 binding were detectable, revealed that the gamma chain was associated with, on average, only 10-12% of the beta chains expressed on such cells. This contrasted with approximately equal levels of beta and gamma chain expression on YT cells, a cell line that has both high levels of cell surface beta chain expression and high levels of IL-2 binding. Thus, the ratio of gamma chain to beta chain present in the immunoprecipitates roughly correlated with the proportion of beta chain involved in intermediate-affinity receptor sites. This result suggests that the 64-kD gamma chain is the component responsible for regulating the affinity of IL-2 association with the beta subunit. By further defining the structural components necessary for IL-2 receptor formation, these studies provide additional insight into mechanisms whereby lymphocytes might regulate their responsiveness to IL-2.

Pre-emption of human cell-mediated lympholysis by a suppressive mechanism activated in mixed lymphocyte cultures.

The regulation of B-cell and T-cell immune responses has been extensively examined and in the experimental animal appears to involve regulatory or "suppressor" T cells (1-4). The limitations of in vitro experimentation have made comparable study of nonpathological human suppression quite difficult (5). We report here an in vitro method that generates and quantitates suppressor activity in man after antigen-specific activation in mixed leukocyte culture (MLC). The one-way MLC induces both a proliferative response (6) and the generation of cytotoxic T lymphocytes (CTLs) (7). Both of these responses are mediated by antigen-specific T-cell subpopulations (8,9) and have been correlated with recognitive and destructive phases of allograft rejection. Recent reports have examined the antigen reactivity of mouse (10,11), rat (12), or human (13,14) lymphocytes obtained after proliferation in MLC. In all cases, after the primary MLC proliferative peak, the recovered lymphocytes rapidly differentiate upon re-exposure to the initial stimulating population, but do so only weakly when exposed to a presumably noncross-reactive third-party stimulating population. Velocity sedimentation separation studies have shown that the blast cells produced in a primary MLC revert to small lymphocytes that rapidly differentiate into proliferating and/or cytotoxic T lymphocytes upon restimulation with the initial antigen (15). These findings demonstrate that positive selection for the responding population in primary MLC does exist and may account for at least part of the specificity of the secondary response. However, this positive selection does not preclude possible involvement of a suppressor mechanism. In fact we have detected suppressor activity in primary MLC sensitization cultures at a time when the proliferation responsible for positive selection does not preclude possible involvement of a suppressor mechanism. In fact we have detected suppressor activity in primary MLC sensitization cultures at a time when the proliferation responsible for positive selection in not yet significant, suggesting that suppression may be overriding importance in the specificity of MLC-activated secondary responses.

Responsiveness of germfree mice in mixed leukocyte culture. Reexamining the nature of the responding unit.

Conventional (cv) and germfree (gf) mice are able to give a good proliferative response to allogeneic cells in the mixed leukocyte culture (MLC) test, while the response to xenogeneic stimulating cells has been in question. Previous studies by others have suggested only a low MLC response in cv animals and none in gf ones. We have found that both cv and gf animals can give a good MLC response to xenogeneic as well as allogeneic cells. These findings are of importance for our understanding of both MLC stimulation and response.

Three HLA-D region antigens defined by primed LD typing.

We have recently described a new method, primed LD typing or PLT, for specific identification of HLA-D antigens. Highly discriminatory PLT cells have been developed which clearly differentiate between cells of individuals that restimulate strongly and those that restimulate weakly. Seven such discriminatory PLT cells have been used to define three antigens called PL1, PL2, and PL3; two more PLT cells may define antigen(s) PL4.

Secondary cell-mediated lympholysis: importance of H-2 LD and SD factors.

Lymphocytes stimulated in mixed leukocyte cultures and left for 13-17 days, i.e. beyond their peak proliferative and cytotoxic reactivities, can be restimulated to give a secondary-type rapid and strong proliferative and cytotoxic response when confronted with cells of the original sensitizing cell donor. We have concerned ourselves primarily with the requirements of restimulation for the presence of LD and/or SD stimuli on the restimulating cells. (a) The low level cell-mediated lympholysis (CML) associated with LD differences in a primary CML can be restimulated to give a secondary-type response by those same LD antigens. (b) If the original sensitizing cells differ from the responding cells by both LD and SD antigens, restimulation with only the LD antigens, or third-party cells presumably carrying cross-reactive LD antigens, can restimulate the secondary CML responses directed against the SD antigens on the original sensitizing cells. (c) The presence of SD antigens on the restimulating cells that are cross-reactive with the primary sensitizing SD antigens (as determined in a primary CML) leads to the preferential activation of cytotoxic T lymphocytes reactive to those antigens although maximum cytotoxicity is still directed at cells carrying the original sensitizing SD antigens. A model to explain these results is presented.

Increased expression of the interleukin 2 (IL-2) receptor beta chain (p70) on CD56+ natural killer cells after in vivo IL-2 therapy: p70 expression does not alone predict the level of intermediate affinity IL-2 binding.

The expression of the 70-kD beta subunit of the interleukin 2 receptor (IL-2R) has been examined on peripheral blood lymphocytes (PBL) obtained from patients receiving systemic infusions of IL-2. Using monoclonal antibodies directed against p70, flow cytometric analyses revealed a greater than threefold increase in expression of the IL-2R beta chain on CD56+ natural killer (NK) cells from post-IL-2 therapy PBL relative to pre-therapy cells. The level of p70 expression on the post-therapy cells was three- to fourfold greater (based on fluorescence intensity) than the level of p70 expression on YT cells, an NK-like cell line that expresses approximately 12,000 intermediate affinity IL-2 binding sites/cell. Despite the high level of p70 expression, in 125I-IL-2 binding assays only 790-1,290 intermediate affinity IL-2 binding sites/cell were detected on post-therapy cells from six patients. These data represent the first report of increased p70 expression after in vivo IL-2 administration and suggest a requirement for at least one additional subunit for the formation of functional intermediate affinity IL-2Rs. Furthermore, the presence on the surface of post-therapy NK cells of excess p70 that does not bind IL-2 with intermediate affinity implies that the formation of intermediate affinity IL-2Rs is not solely determined by the level of p70 expression, and that the response of NK cells to IL-2 might be regulated by altering the expression of p70 or some other IL-2R subunit.

Gamma/delta T cell clones and natural killer cell clones mediate distinct patterns of non-major histocompatibility complex-restricted cytolysis.

Non-MHC-restricted killer cells are cytotoxic lymphocytes that can mediate cytolysis of most tumor targets without apparent selectivity and restriction by the MHC, particularly when activated with IL-2. These effector cells include predominantly NK cells and T cells expressing the TCR-gamma/delta. We found that TCR-gamma/delta-1+, delta TSC1-, BB3+, Ti gamma A+ T cell clones mediate a characteristic cytolytic pattern of non-MHC-restricted cytolysis that is markedly different from NK clones and alpha/beta T cell clones derived from the peripheral blood of the same normal individuals. The characteristic finding is that all BB3/Ti gamma A+ gamma/delta clones mediate strong cytolysis of Daudi cells but they do not lyse Raji cells. In contrast, NK clones from the same donors mediate strong cytolysis of both Daudi and Raji targets. Cytotoxicity by the gamma/delta clones on certain target cells such as Daudi and Molt 4 can be specifically inhibited by mAbs reactive against the TCR-gamma/delta. Therefore, the TCR-gamma/delta on these clones either directly recognizes target epitopes on some tumor targets or it is involved in the regulation of their cytotoxic function. The expression of TCR-gamma/delta products reacting with the BB3 and Ti gamma A mAbs reflects the usage of identical TCR-gamma/delta V region genes that appear to be associated with the characteristic pattern of non-MHC-restricted cytotoxicity displayed by this major subset of human peripheral blood gamma/delta cells.

HL-A LD (lymphocyte defined) typing: a rapid assay with primed lymphocytes.

When human lymphocytes are cultured for 9 to 14 days with stimulating cells of a family member differing by a single HL-A haplotype they become "primed" to recognize specific HL-A LD (mixed lymphocyte culture) antigens. These primed lymphocytes respond specifically and rapidly when "restimulated" with cells of a person that contain the same LD antigens as those of the priming haplotype. Specific HL-A LD antigens can be detected within 24 hours by this primed LD typing.

Cell mediated immunity: separation of cells involved in recognitive and destructive phases.

The mixed leukocyte culture (MLC) and the cell mediated lympholysis (CML) assays are used as in vitro models of the afferent, or recognitive, and efferent, or destructive, phases of the homograft reaction. Activity in both of these tests has been related to differences at the major histocompatibility complex, HL-A in man and H-2 in mouse. Recent evidence suggests that the presumed cell surface differences which lead to cell proliferation in MLC are different from those which act as a target for CML. Data are presented providing further support for this hypothesis; in addition separate cell populations may respond to the differences which activate cells in MLC and to the differences which serve as targets for CML. There thus appears to be a dichotomy both for genetic control of, and cell populations involved in, the recognitive and destructive phases of cell mediated immunity.

HLA identical leukemia cells and T cell growth factor activate cytotoxic T cell recognition of minor locus histocompatibility antigens in vitro.

Lymphocytes from a healthy HLA-identical bone marrow transplant donor were tested for their ability to destroy her brother's acute myelogenous leukemia blasts in vitro. Primary mixed lymphocyte culture (MLC) and cell-mediated lysis (CML) responses between the patient's remission (pretransplant) and donor's lymphocytes were negative. Stimulation of donor lymphocytes for 7 d in vitro with irradiated leukemia cells, leukemia cells plus allogeneic irradiated lymphocytes, or a pool of irradiated lymphocytes from 10 donors, did not activate any cytotoxic cells able to destroy the HLA identical leukemic blasts. Further culturing for 7 additional d in T cell growth factor (TCGF) generated lymphocytes that induced effective cytotoxicity against the leukemic blasts, but not against autologous lymphocytes. Effective killing against the leukemia was observed only in cultures initially stimulated with the irradiated leukemia cells. These cytotoxic cells were maintained in TCGF and mediated persistent killing against the leukemic target cells. They were also able to destroy lymphocytes from the patient's mother and father, but not from an unrelated cell donor. This suggested specific recognition of non-HLA antigens inherited by the patient, that were foreign to the HLA identical bone marrow donor. These lymphocytes were cloned by a limiting dilution technique and one clone maintained cytotoxicity to the AML blasts and the father's lymphocytes, but not lymphocytes from the mother or an HLA-identical donor. This cytotoxicity was inhibited by a monoclonal anti-HLA antibody. Thus, in vitro sensitization of this sibling's lymphocytes with AML blasts followed by TCGF expansion, and cloning, enabled the detection of HLA-restricted cytotoxic cells that recognize minor locus histocompatibility antigens. This immune recognition may be relevant to the "graft vs. leukemia" effect that has been observed in leukemic animals and patients following histocompatible hematopoietic transplants.

A human T cell clone that mediates the monocyte procoagulant response to specific sensitizing antigen.

A panel of human purified protein derivative of the tubercle bacillus (PPD)-reactive T cell clones was derived by cloning out of soft agar followed by cultivation on inactivated feeder cells in the presence of interleukin-2. 1 of 4 clones tested was able to mediate an increase in monocyte procoagulant activity (PCA) in response to PPD. All four clones had identical surface marker phenotypes (T4+, T8-) and proliferated in response to antigen. The reactive T cell clone possessed no PCA of its own, but upon being presented with PPD was able to instruct monocytes to increase their expression of PCA. Antigen presentation could be performed only by autologous monocytes; allogeneic monocytes from donors unrelated to the donor of the reactive clone could not present antigen to cells of the clone in a way that would initiate the procoagulant response. Cells of the reactive clone did not mediate increased monocyte PCA in response to Candida, even though peripheral blood mononuclear cells from the donor demonstrated increased PCA to both Candida and PPD. Thus, the PCA response to specific antigen can be mediated by a single clone of cells that shows specificity in the recognition of both antigen and antigen presenting cell.

Repetitive weekly cycles of interleukin-2. II. Clinical and immunologic effects of dose, schedule, and addition of indomethacin.

Clinical trials with high doses of interleukin 2 (IL-2) have shown antitumor responses, but many of the patients have experienced severe and occasionally life-threatening toxic effects. Preclinical studies indicate that modifications in IL-2 dose, route, and schedule can influence both immune activation and antitumor effects. This study evaluated the clinical tolerance to and immunologic modifications induced by four repetitive weekly cycles of IL-2, with two dose levels (1 X 10(6) and 3 X 10(6) U/m2 per day) of IL-2 and three different daily administration schedules [bolus, continuous, or combined (bolus and continuous)], with and without indomethacin treatment. Patients in all treatment groups experienced acceptable, non-life-threatening toxic effects and immune system stimulation characterized by rebound lymphocytosis with increased numbers of natural killer and lymphokine-activated killer cells and enhanced direct cytolytic function. These immune changes were significantly enhanced by the repetition of IL-2 cycles beyond the first week of therapy. At an IL-2 dose of 3 X 10(6) U/m2 per day, bolus IL-2 was less immunostimulatory than continuous-infusion IL-2. The combined regimen (with half of each daily dose given as a bolus and half as a 24-hr infusion) was as stimulatory as continuous-infusion IL-2 and also induced antitumor effects. Finally, the addition of indomethacin to this regimen did not significantly modify in vitro or in vivo immune response parameters but appeared to worsen the systemic toxic effects of renal dysfunction and capillary leakage. These results suggest that continuous or combined infusion of IL-2 at 3 X 10(6) U/m2 per day on this schedule should be considered for further testing in phase II trials or in combination with other therapeutic modalities.

Repetitive weekly cycles of recombinant human interleukin-2: responses of renal carcinoma with acceptable toxicity.

A phase I trial of repetitive weekly cycles of human recombinant interleukin-2 (IL-2) was performed in 23 patients with metastatic carcinoma. Patients received 4 days of IL-2 each week, followed by 3 days of observation, for 4 consecutive weeks. IL-2 was administered iv at 1.0 or 3.0 X 10(6) U/m2/day by one of three schedules involving continuous or bolus infusions. All treatment was carried out in a general hospital ward without intensive care unit monitoring or support. Seventeen patients had metastatic renal cell carcinoma; three of these demonstrated measurable (greater than 50% shrinkage) partial responses. This study demonstrates that IL-2 given alone without lymphokine-activated killer cells in this manner can induce antitumor effects with acceptable toxicity.

Lymphokine-activated killer activity induced by in vivo interleukin 2 therapy: predominant role for lymphocytes with increased expression of CD2 and leu19 antigens but negative expression of CD16 antigens.

The phenotype and function of lymphocytes from cancer patients treated with repetitive weekly cycles of continuous i.v. infusions of recombinant interleukin 2 (IL-2) were examined. Peripheral blood lymphocytes (PBL) obtained after IL-2 therapy showed an increased percentage of cells bearing the CD16 and leu19 markers which are associated with natural killer cells. These PBL mediated significantly increased levels of IL-2-dependent lymphokine-activated killer (LAK) activity against the Daudi cell line. Depletion of CD16+ cells from PBL obtained after in vivo IL-2 caused only slight inhibition of their LAK activity or their proliferative response to IL-2 in vitro. This indicates that CD16+ cells are involved but play only a minor role in these responses. In contrast, depletion of leu19+ cells, from PBL activated in vivo with IL-2, virtually abrogated their LAK activity and their proliferative response to IL-2. Two-color flow cytometry studies showed that a leu19+/CD16- population was expanded by in vivo IL-2 therapy and was responsible for the majority of LAK activity by in vivo-activated PBL. Moreover, this CD16- population showed an increased density of leu19 and CD2 (E rosette receptor) antigens when compared to the resting PBL obtained prior to IL-2 treatment. These data show that the predominant population mediating in vitro LAK activity, induced by in vivo IL-2 therapy, consists of activated natural killer cells with a high density of leu19 and CD2 antigens but negative for the CD16 antigen.

In vivo induction of the lymphokine-activated killer phenomenon: interleukin 2-dependent human non-major histocompatibility complex-restricted cytotoxicity generated in vivo during administration of human recombinant interleukin 2.

The availability of purified human recombinant interleukin 2 (IL-2) has enabled clinical trials to test its in vivo effects. We report here the immunological effects of 7 consecutive days of IL-2 treatment given to 25 patients with cancer in a clinical Phase I study. Peripheral blood lymphocytes obtained from patients following therapy with IL-2 had enhanced proliferative responses to IL-2 and enhanced direct cytotoxic activity on K562 target cells. This lytic activity was further augmented by the addition of IL-2 during the 51Cr release assay. Fresh peripheral blood lymphocytes from some patients who had just completed treatment at the higher IL-2 dose levels were able to kill both the natural killer-resistant Daudi cell line and fresh tumor cells while pretreatment samples and peripheral blood lymphocytes from healthy controls were not. This lytic activity was best detected when IL-2 was present in the in vitro effector assay. These results demonstrate that the administration of IL-2 to patients with cancer induces a population of effector cells able to directly destroy natural killer-resistant target cells, when assayed in the presence of IL-2.

Clinical and immunological effects of recombinant interleukin 2 given by repetitive weekly cycles to patients with cancer.

Eleven patients received four consecutive weekly cycles of human recombinant interleukin 2 (IL-2) by continuous infusion for 4 days/week. Two dose levels were tested, 1 and 3 X 10(6) units/m2/day. Toxicities experienced by most patients included fever, rigors, fatigue, anemia, eosinophilia, and liver function abnormalities. All side effects from treatment reversed and no severe or life-threatening problems occurred. A marked lymphocytosis was seen following the 4 weeks of therapy. Fresh lymphocytes obtained during this lymphocytosis mediated enhanced destruction in vitro of a natural killer cell-resistant tumor cell line (Daudi). The increase in the absolute number of circulating lymphocytes and their enhanced ability to mediate direct lysis of Daudi targets resulted in a greater than 100-fold mean increase in cytotoxic potential by the end of IL-2 treatment. One patient, with renal carcinoma, who was treated at 3 X 10(6) units/m2/day experienced a sustained measurable response with greater than 50% regression of pulmonary and hepatic metastases. Five patients were retreated with a second course of IL-2, lasting 4 weeks. This therapy was well tolerated in four of these five patients, with similar immunological changes occurring. No further antitumor responses were seen in these patients. Thus, a relatively well tolerated immunotherapy regimen using IL-2 can induce dramatic increases in lymphocyte number and augment their in vitro antitumor reactivity.

In vitro analysis of donor bone marrow following monoclonal antibody treatment for the prevention of acute graft versus host disease.

Graft-versus-host disease (GVHD) remains a major cause of morbidity and mortality following bone marrow transplantation. The in vitro removal of the GVHD-causing T-lymphocytes from donor marrow is one approach which could control this complication. Treatment of the donor bone marrow with lectins and erythrocyte-forming rosette depletion, anti-T-cell antisera or monoclonal antibodies are methods currently being tested to accomplish this. CT-2 is an immunoglobulin monoclonal antibody specific for the T-cell erythrocyte-forming rosette receptor. Bone marrow from 23 consecutive donors was treated in vitro with CT-2 and complement, prior to infusion, as a potential means of controlling GVHD. Surface marker analysis using erythrocyte-forming rosetting, and OKT-3 and OKT-11 monoclonal antibodies on paired samples of treated and untreated marrow demonstrated a mean depletion to 1% of the original number of T-cells. Proliferative responses to alloantigens and mitogens as well as cytotoxic and natural killer cell function were tested and found to be markedly reduced. Despite these effects on T-lymphocytes, viable hematopoietic stem cell colonies were retained. Clinical results following the in vitro T-lymphocyte depletion of donor bone marrow for the 8 histocompatible and 15 nonhistocompatible bone marrow transplantation are reported. Prompt engraftment with minimal GVHD, despite no posttransplant GVHD prophylaxis, was seen in seven of the matched patients. In the nonhistocompatible bone marrow transplantation, failure of engraftment occurred in 11 patients. Grades III-IV GVHD were seen in two of the four patients that engrafted despite good T-lymphocyte depletion. No predictive correlation could be found between the in vitro analysis of marrow following CT-2 treatment and clinical outcome.