Long-term blood product transfusion support for patients with myelodysplastic syndromes (MDS): cost analysis and complications.

Patients with myelodysplastic syndromes (MDS) frequently become dependent on blood transfusions. We analyzed the total transfusion support required, and its complications and cost, following the diagnosis of MDS (total period = 79.7 patient-years) in 50 patients followed at the Minneapolis VA Medical Center. From diagnosis of MDS to transformation to AML or death (the MDS phase), 41 patients (82%) required transfusions. The median numbers of transfused blood products per patient per year of follow-up in the MDS phase were: packed red blood cells (pRBC), 11.1 (range, 0-91.3) units, random donor platelets (RDP), 6.8 (range, 0-581) units, and single donor apheresis platelet packs (SDP): 0 (range, 0-40) collections. In the AML phase (time from diagnosis of secondary AML to death or last follow-up), median transfusion requirements per patient (n = 5) were 24 (range, 8-88) units pRBC, 94 (range, 24-480) units RDP and 3 (range, 0-19) collections of SDP. Overall, 80% of patients required either special processing or selection of blood products, had reactions to blood products or required premedications (specified/complicated transfusions); 94% of all pRBC and 97% of all platelet transfusions were specified/complicated. The median cost of transfusions per patient was $4048 (range, $0-73210) during the MDS phase and $13210 (range, $5288-59010) during the AML phase. During the MDS phase, the median cost was $4877 (range, $0-67050) per patient-year of follow-up; the major proportion of this cost was for pRBC transfusions. Long-term support with frequent transfusions for MDS usually requires specially selected or processed blood products, and is associated with a high incidence of transfusion reactions. This study provides baseline data on the costs of transfusion support for MDS, and can be used for comparing resource utilization and costs of long-term transfusion support (supportive care) with growth factor therapy or disease-modifying modalities such as allogeneic transplantation.

HL-60 cells degrade alpha-actinin to produce a fragment that promotes monocyte/macrophage maturation.

An amino-terminal fragment of alpha-actinin can promote monocyte/macrophage maturation. This fragment was initially isolated from media of HL-60 myeloid leukemia cells cultured on extracellular bone marrow matrix. To determine the source of this fragment in this culture system, we investigated whether HL-60 cells grown on bone marrow stroma have increased intracellular levels of alpha-actinin that may be released into the media during cell apoptosis. HL-60 cells grown on matrix showed no evidence of increased cellular alpha-actinin compared to cells grown on plastic substrata as measured by flow cytometry. In addition, there was no evidence of increased apoptosis as determined by DNA fragmentation assays or flow cytometry. However, 100 kD alpha-actinin was found in the extracellular matrix of bone marrow stroma by Western blot analysis and immunofluorescence microscopy. The alpha-actinin content in the stroma was markedly decreased after exposure to HL-60 cells. Furthermore, lysates of HL-60 cells or of peripheral blood monocytes can degrade exogenous alpha-actinin to produce a 31 kD fragment, which promotes monocyte/macrophage maturation. We conclude that when alpha-actinin is present in the extracellular matrix, it can be modified by HL-60 cells to produce a maturation promoting 31 kD fragment.

A fragment of alpha-actinin promotes monocyte/macrophage maturation in vitro.

Conditioned media (CM) from cultures of HL-60 myeloid leukemia cells grown on extracellular bone marrow matrix contains a factor that induces macrophage-like maturation of HL-60 cells. This factor was purified from the CM of HL-60 cells grown on bone marrow stroma by ammonium sulfate precipitation, then sequential chromatography on DEAE, affi-gel blue affinity, gel exclusion, and wheat germ affinity columns, followed by C-4 reverse phase HPLC, and SDS-PAGE. The maturation promoting activity of the CM was identified in a single 31 kD protein. Amino acid sequence analysis of four internal tryptic peptides of this protein confirmed significant homology with amino acid residues 48-60, 138-147, 215-220, and 221-236 of human cytoskeletal alpha-actinin. An immunoaffinity purified rabbit polyclonal anti-chicken alpha-actinin inhibited the activity of HL-60 conditioned media. A 27 kD amino-terminal fragment of alpha-actinin produced by thermolysin digestion of chicken gizzard alpha-actinin, but not intact alpha-actinin, had maturation promoting activity on several cell types, including blood monocytes, as measured by lysozyme secretion and tartrate-resistant acid phosphatase staining. We conclude that an extracellular alpha-actinin fragment can promote monocyte/macrophage maturation. This represents the first example of a fragment of a cytoskeletal component, which may be released during tissue remodeling and repair, playing a role in phagocyte maturation.

Phase I trial of etoposide, carboplatin, and GM-CSF in extensive small-cell lung cancer: a Cancer and Leukemia Group B study (CALGB 8832).

The maximum tolerated dose (MTD) of etoposide and carboplatin without growth factor support was previously defined by Cancer and Leukemia Group B (CALGB) as 200 and 125 mg/m2/day x 3, respectively, given every 28 days to previously untreated patients who have extensive, small-cell lung cancer (SCLC). Myelosuppression was dose-limiting. The purpose of this phase I trial was to determine if granulocyte macrophage colony-stimulating factor (GM-CSF) support allows the dosage of the combination of etoposide and carboplatin to be increased above the previously determined MTD. In this CALGB study of 44 evaluable patients with performance status 0-2, cohorts were treated with etoposide and carboplatin given intravenously on days 1-3 followed by GM-CSF (molgramostim) given subcutaneously on days 4-18. Four dose levels of bacteria-derived recombinant GM-CSF (5, 10, 20 microg/kg/day and 5 microg/kg every 12 h), three dose levels of etoposide (200, 250, and 300 mg/m2/day x 3), and two dose levels of carboplatin (125 and 150 mg/m2/day x 3) were evaluated. There was no chemotherapy dose escalation in individual patients. With 5 microg/kg/d GM-CSF, the first etoposide and carboplatin cycle of 300 and 150 mg/m2/day x 3, respectively, could be administered with acceptable toxicity. However, GM-CSF did not allow repeated administration of this dose-escalated regimen every 21 days, since delayed platelet and/or neutrophil recovery was dose limiting in later cycles. These results demonstrate that GM-CSF alone has limited capability to support the repeated administration of high doses of etoposide and carboplatin. CALGB currently is testing the ability of interleukin (IL)-6 given with GM-CSF to ameliorate the cumulative myelosuppression of this intense regimen.

A phase I/II trial of etoposide and cisplatin in extensive small cell lung cancer: a cancer and leukemia group B study.

Patients with untreated extensive small cell lung cancer (SCLC) with CALGB performance scores 0-2 were treated with etoposide 200 mg/m2/day on days 1-3 and cisplatin doses of 20, 30, or 35 mg/m2/day days 1-3 in a Phase I/II format. Of the nine patients treated at the 35 mg/m2/day cisplatin dose in the Phase I portion of the study, Grade 4 leukopenia occurred in five patients and Grade 4 thrombocytopenia in four. There were two deaths due to myelosuppression and sepsis. This dose was thus considered the maximum tolerated dose (MTD), and a Phase II trial was then conducted using this treatment program. In the Phase II trial of 39 patients, the objective response rate was 67% (95% confidence interval, 50-81%) with 21% complete responses (CI 9-36%). Median survival was 10.5 months. Grade 4-5 leukopenia was seen in 57% and Grade 4-5 thrombocytopenia in 56%. The MTD defined by this Phase I trial represents a 67-100% increase in etoposide and a 32-42% increase in cisplatin dosage compared to prior studies. The observed objective response rates with this regimen are comparable to studies using conventional doses, but hematological toxicity was higher.

Marrow-derived heparan sulfate proteoglycan mediates the adhesion of hematopoietic progenitor cells to cytokines.

Heparan sulfate proteoglycan (HS-PG), an important component of the human bone marrow extracellular matrix (ECM), is believed to influence hematopoietic progenitor cell development by binding and localizing growth factors to specific niches within the hematopoietic microenvironment. We utilized a model ECM system, which uses immobilized ECM proteins and/or cytokines and bone marrow populations enriched for human hematopoietic stem cell (HSC), to assess the effects of HS-PG on the development of primitive hematopoietic progenitor cells. HS-PG alone failed to bind hematopoietic progenitor cells cloned from bone marrow CD34+CD15-HLA-DR- cells, which are enriched for HSC. HS-PG alone failed to function as a mitogen. In sharp contrast, the interaction of HS-PG with either growth factors (interleukin-3 [IL-3] or stem cell factor/Kit ligand [KL] or an ECM protein (thrombospondin [TSP]) markedly influenced progenitor cell adherence. The binding of either IL-3 or KL to HS-PG resulted in a two-fold increase in attachment of the colony-forming unit-granulocyte/macrophage (CFU-GM), a 1.5-fold increase in attachment of the burst-forming unit-erythroid (BFU-E) and the high-proliferative-potential colony-forming cell (HPP-CFC), and a two- to three-fold increase in attachment of the colony-forming unit-granulocyte/erythroid/macrophage/megakaryocyte (CFU-GEMM) compared to localized growth factor alone. Attachment of the BFU-megakaryocyte (BFU-MK), however, was slightly reduced by the interaction of either IL-3 or KL with HS-PG. The interaction of HS-PG with TSP resulted in a two-fold increase in CFU-GM and CFU-GEMM attachment, while the attachment of BFU-E, HPP-CFC, and BFU-MK was unaltered. We conclude that HS-PG cooperatively interacts with both growth factors and ECM proteins to augment progenitor cell localization within the hematopoietic microenvironment.

Human bone marrow heparan sulfate induces leukemia cell differentiation.

The proliferation and development of hematopoietic cells occurs in close association with bone marrow stroma. Heparan sulfate is a major component of the stroma. We have isolated a form of heparan sulfate proteoglycan from a human stromal cell line grown in vitro in the presence of [35S]sulfate. This proteoglycan contains a phosphatidylinositol component which likely anchors it to the stromal cell membrane. The glycosaminoglycan chains of this proteoglycan could induce maturation of the HL-60 myeloid leukemia cell line. A less hydrophobic heparan sulfate proteoglycan was also present in the stroma, but could not induce HL-60 maturation. The two heparan sulfates had glycosaminoglycan chains that were similar in size (36 Kd) and charge density. Structural studies suggested only minor but perhaps significant differences in the carbohydrate sequences of the two heparan sulfates. The relationship of these subtle structural differences to the difference observed in differentiation-inducing activity remains to be elucidated.

Phase I/II trial of etoposide and carboplatin in extensive small-cell lung cancer. A report from the Cancer and Leukemia Group B.

Previously untreated extensive small-cell lung cancer (SCLC) patients with performance status 0-2 were treated with etoposide 200 mg/m2/day on days 1-3 and carboplatin doses of 50, 100, or 125 mg/m2/day on days 1-3 in a Phase I format. Among the ten eligible patients treated with 125 mg/m2/day of carboplatin, grade 3 or 4 infection occurred in six patients, grade 4 thrombocytopenia in four patients, and there was one death with myelosuppression. Thus, this dose was considered the maximum tolerated dose (MTD), and a Phase II trial was then conducted utilizing this treatment program. In the Phase II trials, 81% of the 48 eligible patients had grade 3 or 4 leukopenia, 76% had grade 3 or 4 thrombocytopenia, and 55% had grade 3 or 4 anemia. There were three (6%) toxic deaths from myelosuppression. The objective response rate was 63% (17% complete responders) with a median response duration of 6.2 months for complete responders and 6.4 months for partial responders. Median survival was 12 months. The MTD defined by this Phase I trial represents a 67-100% increase in etoposide and a 25% increase in carboplatin compared to prior studies. Cancer and Leukemia Group B (CALGB) plans to study further dose intensification of this regimen with colony-stimulating factors.

Epidermolysis bullosa acquisita induced by GM-CSF: a role for eosinophils in treatment-related toxicity.

A patient treated with granulocyte-macrophage colony-stimulating factor (GM-CSF) developed eosinophilia and epidermolysis bullosa acquisita. The bullae were subepidermal, and filled with an inflammatory infiltrate composed predominantly of eosinophils. Immunofluorescence studies disclosed linear deposition of IgG, IgA and C3 at the basement membrane zone and immunoelectron microscopy demonstrated antibody deposition in the lamina densa and sublamina densa region; however, the patient's serum did not contain circulating antibody to basement membrane zone antigens. Staining with monoclonal antibodies revealed dense deposits of both eosinophil peroxidase and eosinophil major basic protein at the dermal-epidermal junction. The eosinophilia and skin lesions resolved upon discontinuation of GM-CSF. This case provides evidence for two hypotheses: (1) GM-CSF induced proliferation and activation of eosinophils may contribute to some of the toxicities of GM-CSF treatment, and (2) activated granulocytes, including eosinophils, may mediate blister formation in epidermolysis bullosa acquisita.

Bone marrow matrix promotes differentiation and prolongs the cell cycle of U-937 cells.

The extracellular matrix influences the growth and differentiation of a variety of cell types. In this study, the effects of bone marrow extracellular matrix on U-937 cells, a human histiocytic lymphoma cell line, were assessed. Sixty percent of U-937 cells adhered to extracellular matrix, whereas only 1% adhered to uncoated plastic. U-937 cells grown on extracellular matrix released significantly more lysozyme into the medium (8.3 +/- 0.3 micrograms/10(6) cells) compared to those grown on plastic (4.2 +/- 0.5 micrograms/10(6) cells). FMLP (f-met-leu-phe) receptor expression was also enhanced suggesting a more mature phenotype in cells grown on matrix (2980 cpm/10(6) cells vs 230 cpm/10(6) cells on plastic). Furthermore, bone marrow extracellular matrix inhibited proliferation of U-937 cells. After four days in culture, there was a 65% inhibition of cell growth in matrix-coated flasks compared to uncoated flasks. Since an arrest in G0/G1 usually precedes mammalian cell differentiation, DNA histograms were performed on U-937 cells grown on matrix to detect such an arrest. However, the cell cycle distribution of U-937 cells grown on extracellular matrix or uncoated plastic for various time periods was similar. In contrast, bromodeoxyuridine pulse labeling revealed approximately a 5 hr prolongation in cycle length in cells grown on extracellular matrix. We conclude that bone marrow extracellular matrix induced macrophage-like differentiation and inhibited proliferation of U-937 cells with a prolongation of the cell cycle that was not G0/G1 phase specific.

Bone marrow extracellular matrix induces HL-60 cells to produce an autonomous differentiation factor.

Conditioned medium from cultures of HL-60 myeloid leukemia cells grown on extracellular bone marrow matrix induces macrophage-like differentiation of fresh HL-60 cells. The active medium component is sensitive to protease treatment, indicating that it is a protein, but it is heat stable. Conditioned medium from HL-60 cells grown on protease-treated bone marrow matrix still contains the active component. Thus, it appears that the differentiation-inducing protein is produced by HL-60 cells and is not released from the bone marrow matrix. To identify this differentiation factor, RNA was isolated from HL-60 cells grown on bone marrow matrix and assayed by Northern analysis for expression of mRNA for human differentiation factor, tumor necrosis factor, and macrophage colony-stimulating factor, all inducers of monocyte/macrophage differentiation. Expression of differentiation factor, tumor necrosis factor, or macrophage colony-stimulating factor mRNA was not enhanced in HL-60 cells grown on matrix compared to cells grown on uncoated plastic flasks. Thus, the maturation factor does not appear to be differentiation factor, tumor necrosis factor, or macrophage colony-stimulating factor within the limits of detection of Northern analysis. Elution of the active conditioned medium fraction on a Sephacryl S-200 column revealed a molecular weight of approximately 40,000. The active protein eluted on a DEAE-cellulose ion-exchange column at an ionic strength of 0.3 M NaCl, indicating that it is fairly anionic. Thus, bone marrow matrix is able to induce HL-60 cells to produce a maturation-inducing 40 kilodalton protein.(ABSTRACT TRUNCATED AT 250 WORDS)

Corynebacterium JK: a new pathogen in ventriculostomy infections.

In the past decade, Corynebacterium JK has emerged as a pathogen in several distinct clinical settings, including sepsis in immunocompromised patients and prosthetic valve endocarditis. It is also recognized as a nosocomial pathogen in infections of prosthetic devices. We present a case of a patient with carcinomatous meningitis who developed a Corynebacterium JK infection of an internal ventriculostomy which was used for intraventricular chemotherapy. Treatment with systemic and intraventriculostomy vancomycin for three weeks resulted in bacteriologic resolution of the infection. Removal of the prosthetic device was not essential for cure in this patient. The clinical spectrum of infection with this organism and aspects of therapy are reviewed. As a greater awareness of the pathogenic nature of this organism develops, it is likely to be implicated as a causative agent in a variety of infections.

BCNU treatment of marrow stromal monolayers reversibly alters haematopoiesis.

The marrow stromal microenvironment is essential for maintaining normal haematopoiesis. Chemotherapy drugs, such as the nitrosoureas, may impair the ability of the stroma to support haematopoiesis. To assess the effects of 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU) on in vitro haematopoiesis, stromal monolayers were treated with BCNU, 5 micrograms/ml weekly for 3 weeks, then seeded 24 h after the third treatment with haematopoietic progenitors. Three weeks after seeding, total adherent and non-adherent cell numbers were similar in treated and control flasks as were adherent granulocyte/macrophage colony forming cell (GM-CFC) numbers. In contrast, non-adherent GM-CFC were significantly reduced in treated flasks, to 40-60% of controls. However, no reduction in non-adherent GM-CFC number was seen if seeding was delayed for 7 d following BCNU treatment, suggesting the effects are reversible. Conditioned media from treated and control monolayers, harvested at a time corresponding to the time of seeding, showed no difference in colony stimulating activity. In addition, extracts of bound growth factors from treated and control monolayers also showed no differences in colony stimulating activity. Thus, BCNU can alter haematopoiesis through a reversible effect on the marrow stroma and this effect does not appear to be mediated by a change in stromal growth factor production.

The effect of transforming growth factor-beta 1 on glycosaminoglycan production by human marrow cultures.

We have assayed the effect of transforming growth factor-beta 1 (TGF-beta 1), a potent modulator of hematopoiesis, on glycosaminoglycan production in human marrow cultures. Glycosaminoglycans are a component of the extracellular matrix known to affect cell growth and differentiation. TGF-beta 1 and [35S]sulfate were added simultaneously to hematopoietically active human marrow cultures, and radiolabeled glycosaminoglycan production was determined by cetylpyridinium chloride precipitation. TGF-beta 1 at 15 ng/ml for 72 h increased [35S]sulfate incorporation into media glycosaminoglycans to 190% of control levels but did not affect the [35S]sulfate incorporation into cell-associated glycosaminoglycans. Approximately 90% of the glycosaminoglycans in the media fraction and 85% of the glycosaminoglycans in the cell-associated fraction were susceptible to degradation by chondroitin ABCase in both treated and control cultures. Pulse-chase experiments suggested that the increase in glycosaminoglycan [35S]sulfate incorporation was not due to decreased glycosaminoglycan degradation. This concentration of TGF-beta 1 did not alter nonadherent granulocyte-macrophage colony-forming unit (CFU-GM) number per flask but significantly decreased the more primitive adherent CFU-GM number per flask (by 50%-70%). These data suggest that the ability of TGF-beta 1 to modulate hematopoiesis may be due, in part, to its effects on glycosaminoglycan production.

A heparan sulfate-containing fraction of bone marrow stroma induces maturation of HL-60 cells in vitro.

Constituents of the bone marrow microenvironment have the capacity to influence both normal and malignant hematopoietic cell behavior. For example, HL-60 human promyelocytic leukemia cells in vitro display a more mature phenotype when grown on a bone marrow stroma-derived matrix. To elucidate which component(s) of the stromal matrix is capable of modulating HL-60 cell phenotype, matrices were treated with a variety of chemicals and enzymes prior to being used in the differentiation assay. Treatment of matrices with collagenase, pronase, chondroitinase, or chloroform:methanol:ether could not abolish the differentiation-promoting activity of bone marrow stroma. In contrast, the activity was destroyed by alkali treatment (0.5 M NaOH for 18 h) or heparinase/heparitinase enzymes. Heparin added to cultures increased maturation of HL-60 cells as determined by esterase production, Fc rosette formation, and morphological appearance. Other stromal components such as laminin, fibronectin, collagen I, collagen IV, or chondroitin sulfate did not alter the HL-60 leukemia cell phenotype. Stroma-derived matrix material which labeled with [35S]sulfate and eluted on a DEAE ion-exchange column as a high ionic fraction in 1.5 M LiCl and 7.5% sodium dodecyl sulfate contained the active fraction. A heparan sulfate proteoglycan component isolated by polyacrylamide-agarose gel electrophoresis induced a more mature HL-60 phenotype, and digestion with heparinase/heparitinase in the presence of protease inhibitors abrogated the effects on HL-60 phenotype. We conclude that a heparan sulfate-associated fraction of the bone marrow matrix plays a key role in the regulation of leukemic cell maturation.

Serum and urine glycosaminoglycans in myeloid leukemia and myelodysplasia.

Urinary glycosaminoglycan excretion is increased in a variety of human diseases, including malignancy. We have measured serum and urine glycosaminoglycan levels by the carbazole method of uronic acid determination in patients with myeloid leukemia or myelodysplasia. Eleven patients were studied during active disease as well as eight in complete remission. Serum levels in patients with active disease did not differ significantly from 11 healthy volunteers with no hematological disease. In contrast, the median urine level for the patients with active disease was 7.6 mg uronate/g Creatinine (Creat) compared to 2.6 for controls (p less than 0.002). Interestingly, the eight patients in complete remission also had a significant increase in uronate excretion with a median of 7.3 (p less than 0.002). These results suggest that elevated urinary glycosaminoglycan levels in leukemia are not due to impaired ability of the liver to clear circulating glycosaminoglycans or overproduction by leukemic cells. The observed increase in glycosaminoglycan excretion may be due to altered bone marrow matrix metabolism that is often not reversed by the achievement of hematologic remission.

BCNU-induced increase in sulfated glycosaminoglycan production by human bone marrow stromal cells.

The etiology of alkylator-induced leukemia is obscure, but may be due in part to alternations in the bone marrow stromal microenvironment. Marrow extracellular matrix, including collagen, glycosaminoglycans/proteoglycans, and glycoproteins, may play a crucial role in the control of normal and abnormal hematopoiesis. Twenty-four hours after seeding, confluent human bone marrow stromal cell cultures were exposed for 3 h to 15 micrograms/ml of 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU), an alkylating agent with leukemogenic potential. On the eighth day of culture, [35S]sulfate was added and radiolabeled glycosaminoglycan(s) (GAGs) was harvested 24 h later. BCNU treatment resulted in a 104% increase of the radiolabel incorporation into cetylpyridinium chloride precipitable GAG. In addition, spectrophotometric measurement of total GAG in treated cells revealed a similar GAG increase. However, BCNU treatment did not alter compartmental GAG distribution or GAG species. Our results demonstrate a profound quantitative change in the production of important extracellular matrix components by bone marrow stromal cells after exposure to a nitrosourea. This increase may be a factor in microenvironmental alterations leading to bone marrow toxicity following alkylator exposure.

Degradation of cartilage proteoglycans by myeloid leukemia cells.

Polymorphonuclear leukocytes contain proteases that are capable of degrading articular cartilage matrix in disease states such as rheumatoid arthritis and osteoarthritis. In this study, the HL-60 human promyelocytic leukemia cell line was examined for ability to degrade cartilage proteoglycans. The HL-60 cells contained proteoglycan-degrading enzymes, which may contribute to the joint inflammation sometimes seen in acute leukemia. However, the protease activity was much less than in mature neutrophils and was not enhanced by the induction of myeloid maturation with dimethyl sulfoxide or retinoic acid. The diminished enzyme activity of induced HL-60 cells compared to normal neutrophils is another functional deficiency of these cells.

Effects of extracellular matrix on the malignant phenotype.

Extracellular matrix molecules, including collagen, glycosaminoglycans (usually linked to a protein core as proteoglycan), elastin, and glycoproteins, influence the initiation and maintenance of differentiation of a variety of cell types. These molecules bind to the cell surface at specific sites and nonspecifically by electrostatic forces. Such interactions may alter the cell's response to growth and differentiation factors. After neoplastic transformation, most cells retain some dependence on these factors. This paper reviews the influence of matrix components on the phenotype of a variety of malignant cells and concludes that in vitro studies of malignant cell behavior require the utilization of an appropriate microenvironment.

Bone marrow matrix modulation of HL-60 phenotype.

The initiation and maintenance of cellular differentiation for a variety of cell types has been shown to be influenced by the microenvironment. To investigate the influence of bone marrow stroma on leukemic cell differentiation, HL-60 human promyelocytic leukemia cells were grown in the presence of Triton-treated extracellular matrix derived from normal human bone marrow stromal cells. This bone marrow matrix microenvironment had a dramatic impact on the phenotypic expression of this malignant line. HL-60 cellular proliferation, morphology, nonspecific esterase activity, formation of Fc rosettes, and sensitivity to induction by 12-O-tetradecanoyl-phorbol-13-acetate (TPA) were all influenced by the presence of matrix molecules. In contrast, stromal cell-conditioned media did not alter HL-60 phenotype. Thus, HL-60 cells appear to retain responsiveness to a human bone marrow stromal cell-derived matrix despite their ability to grow autonomously. Studies of the interaction of leukemic cells and marrow stroma in vitro may provide important information concerning the regulation of leukemic cell behavior.