Investigating Stability in Subgroup Identification for Stratified Medicine.

Subgroup analysis may be used to investigate treatment effect heterogeneity among subsets of the study population defined by baseline characteristics. Several methodologies have been proposed in recent years and with these, statistical issues such as multiplicity, complexity, and selection bias have been widely discussed. Some methods adjust for one or more of these issues; however, few of them discuss or consider the stability of the subgroup assignments. We propose exploring the stability of subgroups as a sensitivity analysis step for stratified medicine to assess the robustness of the identified subgroups besides identifying possible factors that may drive this instability. After applying Bayesian credible subgroups, a nonparametric bootstrap can be used to assess stability at subgroup-level and patient-level. Our findings illustrate that when the treatment effect is small or not so evident, patients are more likely to switch to different subgroups (jumpers) across bootstrap resamples. In contrast, when the treatment effect is large or extremely convincing, patients generally remain in the same subgroup. While the proposed subgroup stability method is illustrated through Bayesian credible subgroups method on time-to-event data, this general approach can be used with other subgroup identification methods and endpoints.

Adenoviral delivery of LIM mineralization protein-1 induces new-bone formation in vitro and in vivo.

BACKGROUND: The LIM mineralization protein-1 (LMP-1) gene encodes for an intracellular protein that induces the expression of several bone growth factors. The purpose of the present study was to determine the feasibility and the optimal dose of adenoviral delivery of the LMP-1 cDNA to promote spinal fusion. METHODS: A replication-deficient human recombinant adenovirus was constructed with the LMP-1 cDNA driven by a cytomegalovirus promoter. In phase 1, an in vitro dose-response experiment was performed to determine the optimal adenovirus-LMP-1 (AdLMP-1) concentration and infection time. In phase 2, nine rabbits had a single-level posterolateral arthrodesis of the lumbar spine with implantation of a carrier matrix loaded with bone-marrow-derived buffy-coat cells that had been infected for ten minutes with adenovirus containing the cDNA for LMP-1 (AdLMP-1) or beta-galactosidase (AdBgal). In phase 3, posterolateral arthrodesis of the spine was performed with implantation of cells infected with AdLMP-1 (ten rabbits) or cells infected with an empty adenovirus that did not contain LMP-1 cDNA (ten rabbits) and the results were compared. In this phase, peripheral-blood-derived buffy-coat cells were used instead of bone-marrow-derived cells and a collagen-ceramic-composite sponge was used as the carrier. RESULTS: In phase 1, the in vitro dose-response experiment showed that a multiplicity of infection of 0.25 plaque-forming units per cell was the most efficient dose. In phase 2, the implants that had received cells infected with AdLMP-1 induced a solid, continuous spinal fusion mass at five weeks. In contrast, the implants that had received cells infected with AdBgal or a lower dose of AdLMP-1 induced little or no bone formation. In phase 3, a solid spinal fusion was observed at four weeks in all ten rabbits that had received cells infected with AdLMP-1 and in none of the ten rabbits that had received cells infected with the empty adenovirus. Biomechanical and histological testing of the AdLMP-1-treated specimens revealed findings that were consistent with a high-quality spinal fusion. CONCLUSIONS: Adenoviral delivery of LMP-1 cDNA promotes spinal fusion in immune-competent rabbits.

The effect of nicotine on gene expression during spine fusion.

STUDY DESIGN: A rabbit model of posterolateral spine fusion was used to investigate the effect of nicotine on cytokine expression during spine fusion. OBJECTIVES: To determine the effects of nicotine on the known gene expression pattern of bone morphogens and related proteins expressed during spine fusion. SUMMARY OF BACKGROUND DATA: The mechanism by which nicotine increases the pseudarthrosis rate of spine fusion is unknown. Recently, a distinct temporal and spatial pattern of cytokine expression during bone formation has been described. The authors hypothesized that nicotine would alter this known pattern, thereby revealing the mechanism by which nicotine exerts its effect. METHODS: Twenty-eight New Zealand White rabbits underwent posterolateral spine fusion with autogenous bone graft. Fourteen rabbits received systemic nicotine by a miniosmotic pump. Fusions were harvested at 0, 2, 5, and 7 days and 2, 3, and 4 weeks after arthrodesis. Specimens were divided into the outer zones adjacent to the transverse processes and the central zones between the transverse processes. Gene expression of type I and II collagen, bone morphogenic protein-2, -4, and -6 and basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) was then measured at each time point in each of the two zones. RESULTS: Nicotine inhibited expression of all cytokines measured, mainly in the central zone. However, the previously described temporal and spatial patterns of expression were preserved. CONCLUSIONS: Nicotine inhibits expression of a wide range of cytokines, including those associated with neovascularization and osteoblast differentiation. Therefore, the effects of nicotine appear to involve more than just local vasoconstriction.

Gene therapy for spine fusion.

Spine fusion is a commonly performed yet often unsuccessful surgical procedure. As many as 40% of patients undergoing spine fusion may have a nonunion or failure to form a continuous bone bridge. This clinical challenge has focused much of the attention of osteoinductive bone growth factors toward spine applications. Clinical pilot and pivotal trials will show the feasibility of recombinant and purified bone growth factors to promote spine fusion in humans. Despite this, strategies of gene therapy for spine fusion and other bone healing applications are being pursued. This article reviews the state of the art of local gene therapy and highlights specific issues that must be addressed when pursuing a gene therapy program. Perhaps the most critical step in gene therapy for bone formation is choosing an appropriate osteoinductive gene. Such choices may be limited by differences in efficacy of the chosen gene and availability because of proprietary constraints. The choice of delivery vector is crucial and depends on the potency of the gene and the specific application intended. Establishing the effective dose, transduction time, and gene transfer method are important decisions. The choice of carrier material to form the scaffold for the new bone formation is paramount to successful bone formation. Finally, a strategy for in vitro and in vivo testing must be developed to maximize the chances of success in human trials.

Synthesis and characterization of 8-(dimethylamino)-1-naphthyl derivatives of aluminum, gallium, and indium.

The group 13 dichlorides of formula Ar'MCl2 [Ar' = 8-(dimethylamino)-1-naphthyl (8-(Me2N)C10H6)], M = Al (1), Ga (2), and In (3), have been prepared via the salt elimination reaction of 1 equiv of Ar'Li with MCl3 in toluene solution at -78 degrees C. The reaction of 1 with LiAlH4 in diethyl ether solution at -78 degrees C produced the dihydride [Ar'AlH2]2 (4). The X-ray crystal structures of 1-4 have been determined and show that 1 and 2 are monomeric while 3 and 4 are dimeric in the solid state. The reaction of 1 with RLi in toluene solution at -78 degrees C results in ligand redistribution and formation of Ar'2AlR (R = Me (5), t-Bu (6)). The chloride analogue of 5 and 6, Ar'2AlCl (7), can be prepared directly from the reaction of 2 equiv of Ar'Li with AlCl3 in toluene solution at -78 degrees C. The homoleptic derivative Ar'3Al (8) was obtained when 3 equiv of Ar'Li was employed. Crystal data for 1: monoclinic, space group P2(1), a = 6.534(1) A, b = 10.801(1) A, c = 9.631(2) A, beta = 105.57(2) degrees, V = 654.8(2) A3, Z = 2, R = 0.0453. Crystal data for 2: monoclinic, space group P2(1), a = 6.552(2) A, b = 10.833(2) A, c = 9.601(2) A, beta = 106.05(2) degrees, V = 654.9(3) A3, Z = 2, R = 0.0609. Crystal data for 3: monoclinic, space group P2(1)/c, a = 7.401(2) A, b = 15.746 A, c = 10.801(4) A, beta = 92.37(3) degrees, V = 1257.6(7) A3, Z = 2, R = 0.0712. Crystal data for 4: monoclinic, space group P2(1)/c, a = 13.343(2) A, b = 11.228(2) A, c = 7.505(1) A, beta = 100.64(1) degrees, V = 1105.0(4) A3, Z = 4, R = 0.0560.

Establishing an immortalized human osteoprecursor cell line: OPC1.

The present studies evaluated the feasibility of establishing a conditionally immortalized osteoprecursor cell line derived from human fetal bone tissue. Primary cultures were transfected with a plasmid in which the Mx-1 promoter drives the expression of SV40 T-antigen when activated by human A/D interferon. Several neomycin (G418)-resistant colonies were characterized for cell growth and alkaline phosphatase (ALP) enzyme activity. The clone, designated OPC1 (osteoblastic precursor cell line 1), which exhibited the highest ALP enzyme activity at passage 10 (P10), was selected for additional osteogenic phenotypic characterization. Reverse transcription-polymerase chain reaction (RT-PCR) phenotyping revealed abundant mRNA for osteocalcin (OC), osteonectin (ON), osteopontin (OP), parathyroid hormone receptor (PTHr), ALP, and procollagen type I (ProI). In addition, the levels of quantitative RT-PCR product of ON, OP, PTHr, and ProI mRNAs exhibited a marked up-regulation when maintained in medium containing an osteogenic supplement (OS). The ability to stimulate osteogenic differentiation was characterized in postconfluent OPC1 cells maintained in tissue culture medium supplemented with recombinant human bone morphogenetic protein-2 (rhBMP-2) either with or without an OS. All treatment groups exhibited a striking up-regulation of ALP enzyme activity that coincided with ALP histochemical observations. Postconfluent cells also exhibited the ability to form mineralized nodules under all treatments (confirmed by von Kossa histochemical staining and calcium deposition). An enzyme immunosorbent assay (EIA) was utilized to measure intact human OC from the OPC1 line under the various treatments. Abundant OC was evident in the tissue culture medium indicating de novo sythesis and release from the OPC1 line under appropriate conditions. The clonal human-derived OPC1 line represents a homogeneous osteogenic cell line that not only has maintained a consistent bone phenotype from P10 to at least P30, but has also exhibited the capacity to generate programmed differentiation in the presence of low dose rhBMP-2 (10 ng/ml). Thus, the OPC1 line is a human-derived osteoprecursor that provides a sensitive in vitro cell culture system to evaluate bone development, cell/biomaterial interactions, and may be a useful screen for putative bone differentiating factors.

The effect of hydrostatic pressure on intervertebral disc metabolism.

STUDY DESIGN: By the use of pressure vessels, hydrostatic pressure was applied to intervertebral disc cells cultured in an alginate. OBJECTIVE: To test the hypothesis that hydrostatic pressure directly affects the synthesis of collagen and proteoglycan by the intervertebral disc cells. SUMMARY OF BACKGROUND DATA: The influence of compression (both hydrostatic and mechanical) on chondrocyte metabolism was examined in a number of earlier studies. However, in most of these studies, articular cartilage, not intervertebral disc, was used, and in none of these was hydrostatic pressure applied to intervertebral disc cells cultured in alginate. METHODS: Fresh cells were harvested from the lumbar intervertebral discs of dogs. Before their suspension in an alginate gel system, the cells were plated and expanded until they reached confluence. Then, by use of the alginate gel system, the cells were exposed (for up to 9 days) to specific values of hydrostatic pressure inside two stainless steel pressure vessels. One vessel was kept at 1 MPa and the other at atmospheric pressure. The effects of 1 MPa were compared against atmospheric pressure by measuring the incorporation of [3H]-proline and [35S]-sulfate into collagen and proteoglycans, respectively, for the anulus cells and nucleus cells separately, and by determining whether this incorporation was reflected by changes in the levels of mRNA for aggrecan and Types I and II collagen. RESULTS: Comparisons with atmospheric pressure yielded the following findings: 1) In the incorporation studies, the nucleus and anulus cells exhibited a differential response to a hydrostatic pressure of 1 MPa. Collagen and proteoglycan syntheses were stimulated in the nucleus cells and inhibited in the anulus cells. 2) There was no significant increase in cell proliferation, as measured by DNA content, at 1 MPa for either the anulus or nucleus cells. 3) The mRNA levels of collagen (Col 1A1 and Col 2A1) and aggrecan increased at 1 MPa in both the nucleus and anulus cells. CONCLUSIONS: Hydrostatic pressure directly affects the synthesis of collagen and proteoglycan by the intervertebral disc cells.

Molecular biology and spinal disorders. A survey for the clinician.

Over the past 10 years, advances in molecular biology techniques have extended the potential for understanding spinal disorders from the microscopic (histologic) level down to the molecular level of gene expression within individual cells. These advances are initiating new avenues of research and, ultimately, novel clinical treatments. The intent of this update is to provide the spine clinician with a basic understanding of molecular biology, the type of information that may be learned from its application, and the potential for gene therapy in spine disorders.

LMP-1, a LIM-domain protein, mediates BMP-6 effects on bone formation.

Glucocorticoids can promote osteoblast differentiation from fetal calvarial cells and bone marrow stromal cells. We recently reported that glucocorticoid specifically induced bone morphogenetic protein-6 (BMP-6), a glycoprotein signaling molecule that is a multifunctional regulator of vertebrate development. In the present study, we used fetal rat secondary calvarial cultures to determine genes induced during early osteoblast differentiation as initiated by glucocorticoid treatment. Glucocorticoid, and subsequently BMP-6, was found to induce a novel rat intracellular protein, LIM mineralization protein-1 (LMP-1), that in turn resulted in synthesis of one or more soluble factors that could induce de novo bone formation. Blocking expression of LMP-1 using antisense oligonucleotide prevented osteoblast differentiation in vitro. Overexpression of LMP-1 using a mammalian expression vector was sufficient to initiate de novo bone nodule formation in vitro and in sc implants in vivo. These data demonstrate that LMP-1 is an essential positive regulator of the osteoblast differentiation program as well as an important intermediate step in the BMP-6 signaling pathway.

Lumbar spine fusion by local gene therapy with a cDNA encoding a novel osteoinductive protein (LMP-1).

STUDY DESIGN: A posterior arthrodesis animal model using local expression of a newly discovered osteoinductive protein delivered in bone marrow cells. OBJECTIVE: To introduce the concept of local gene therapy and determine its feasibility for achieving lumbar spine fusion using a gene for a novel osteoinductive protein: LIM Mineralization Protein-1 (LMP-1). SUMMARY OF BACKGROUND DATA: Extensive work is currently underway to improve the healing success and morbidity associated with the gold standard bone-grafting material of autogenous iliac crest. As a result, alternative osteoinductive proteins and new delivery methods warrant investigation. The authors' laboratory recently identified a novel gene that had osteoinductive capacity in vitro and is therefore a candidate for a new in vivo osteoinductive agent. METHODS: Single-level posterior lumbar and thoracic arthrodesis was attempted in 14 athymic rats. The graft material, which consisted of a devitalized bone matrix (no osteoinductive activity) soaked with 0.75 to 1.5 million bone marrow cells, was inserted with the dorsal spine exposed. In each rat, one site received marrow cells transfected with the cDNA encoding a novel osteoinductive protein. At the other site for a control, the marrow cells were transfected with the reverse copy of the cDNA that did not express any protein. Transfection of marrow cells for 2 hours was accomplished using the mammalian expression vector pCMV2. Rats were killed after 4 weeks, and the spines were evaluated by manual palpation, radiographs, and nondecalcified histology. RESULTS: In the pivotal experiment, successful spine fusion was obtained in 9/9 (100%) of the sites receiving marrow cells transfected with the active LMP-1 cDNA and in 0/9 (0%) of the sites receiving marrow cells transfected with the reverse (inactive) LMP-1 cDNA. Radiographs and histology confirmed the manual palpation results, demonstrating controlled new bone formation in the carrier and marrow transfected with the active LMP-1 cDNA and essentially no bone induction in the sites treated with marrow cells that did not express the protein. CONCLUSIONS: These data confirm that local delivery of the novel LMP-1 cDNA using bone marrow cells is feasible in vivo. Furthermore, these results demonstrate that posterior thoracic or lumbar spine fusion can be achieved in rats by local delivery of the LMP-1 cDNA.

The Marshall R. Urist Young Investigator Award. Gene expression during autograft lumbar spine fusion and the effect of bone morphogenetic protein 2.

A prospective animal study of posterolateral lumbar spine arthrodesis was performed to determine the temporal and spatial pattern of gene expression and to determine the effect of recombinant human bone morphogenetic protein 2 on the gene expression pattern of a healing spine fusion mass. In Group 1, 20 adult New Zealand rabbits underwent L4-L5 posterolateral intertransverse process arthrodesis using autograft alone. Two rabbits were euthanized at each of the following points: 0, 2, and 4 days, and 1, 2, 3, 4, 5, 6, and 10 weeks after surgery. The same surgical technique was used for 16 rabbits in Group II, except that the autograft first was soaked in a solution of recombinant human bone morphogenetic protein 2 before implantation. Ribonucleic acid was extracted from different regions of the fusion mass at each point and analyzed for expression of bone and cartilage related genes using reverse transcription polymerase chain reaction. A reproducible temporal sequence and spatial pattern of gene expression was found in healing spine fusions. In the central portion of the fusion mass a temporal lag in gene expression was observed that parallels the lag in healing within the central zone previously observed in histologic studies. Treatment of bone graft with recombinant human bone morphogenetic protein 2 resulted in an increase in the early expression of bone morphogenetic protein 6 which was associated with expression of higher levels of Type I collagen, osteocalcin, and other bone related genes. These findings suggest that central nonunion may be associated with delayed expression of osteoblast related genes in the central region of the forming fusion mass. The growth factor, recombinant human bone morphogenetic protein 2, increased the level of bone related gene expression throughout the fusion mass, eliminated the delay in healing within the central zone, and may decrease the likelihood of a nonunion.

Fibrin induction of tissue factor expression in human vascular endothelial cells.

BACKGROUND: For the present study, we hypothesized that fibrin is an inducer of tissue factor (TF) expression in vascular endothelial cells in vitro and in vivo. METHODS AND RESULTS: To test the in vitro aspect of this hypothesis, human umbilical vein endothelial cells (HUVECs) were cocultured with physiologically relevant concentrations of fibrin (0.03 to 1.0 mg fibrin/mL) for various times (0.5 to 24 hours), and TF expression was compared with that in unstimulated HUVECs (media control). Results demonstrated that fibrin induced a time- and dose-dependent increase in TF antigen expression, functional TF procoagulant activity, and TF mRNA in HUVECs. CONCLUSIONS: These studies demonstrate that fibrin can directly regulate TF expression in HUVECs in vitro.

Glucocorticoid-induced differentiation of fetal rat calvarial osteoblasts is mediated by bone morphogenetic protein-6.

Glucocorticoids (GCs) at physiological concentrations promote osteoblast differentiation from fetal calvarial cells, calvarial organ cultures, and bone marrow stromal cells; however, the cellular pathways involved are not known. Bone morphogenetic proteins (BMPs) are recognized as important mediators of osteoblast differentiation. Specific roles for individual BMPs during postembryonic membranous bone formation have yet to be determined. We recently reported that GC potentiated the osteoblast differentiation effects of BMP-2 and BMP-4, but not of BMP-6, which, by itself, was the most potent of the three. In the present study, we used fetal rat secondary calvarial cultures to study the role of BMP-6 during early osteoblast differentiation. Treatment with the GC triamcinolone (10(-9) M) resulted in a 5- to 8-fold increase in BMP-6 steady-state messenger RNA levels, peaking at 12 h. In contrast, BMPs -2, -4, -5, -7, and transforming growth factor (TGF)-beta1 messenger RNA levels increased by less than 2-fold, after GC treatment, compared with untreated control cultures at 24 h. BMP-6 protein secretion increased 6- to 7-fold by 12 h and 12-fold (from 7.5 to 90 ng/ml) by 24 h, as measured by quantitative Western analysis. Treatment of cells with oligodeoxynucleotides antisense to BMP-6 diminished secretion of BMP-6 protein and significantly inhibited the GC-induced differentiation, as determined by a 10-fold decrease in the number of mineralized bone nodules, compared with controls that were treated with sense oligonucleotides or no oligonucleotides (ANOVA, P < 0.05). The antisense oligonucleotide inhibition of differentiation was rescued by treatment with exogenous recombinant human BMP-6. We conclude that GC-induced differentiation of osteoblasts from the pluripotent precursors is mediated, in part, by BMP-6. These results suggest that BMP-6 has an important and unique role during early osteoblast differentiation.

Differential effects and glucocorticoid potentiation of bone morphogenetic protein action during rat osteoblast differentiation in vitro.

Bone morphogenetic proteins (BMPs) induce cartilage and bone differentiation in vivo and promote osteoblast differentiation from calvarial and marrow stromal cell preparations. Functional differences between BMP-2, -4, and -6 are not well understood. Recent investigations find that these three closely related osteoinductive proteins may exert different effects in primary rat calvarial cell cultures, suggesting the possibility of unique functions in vivo. In this study, we use a fetal rat secondary calvarial cell culture system to examine the differential effects of BMP-2, -4, and -6 on early osteoblast differentiation. These cells do not spontaneously differentiate into osteoblasts, as do cells in primary calvarial cultures, but rather require exposure to a differentiation initiator such as glucocorticoid or BMP. We determined that BMP-6 is a 2- to 2.5-fold more potent inducer of osteoblast differentiation than BMP-2 or -4. BMP-6 induced the formation of more and larger bone nodules as well as increased osteocalcin secretion. The effects of all three of these BMPs were potentiated up to 10-fold by cotreatment or pretreatment with the glucocorticoid triamcinolone (Trm). The Trm effects were synergistic with those of BMP-2 or -4, suggesting that this glucocorticoid may increase the cell responsiveness to these BMPs. Finally, BMP-6 did not require either cotreatment or pretreatment with Trm to achieve greater amounts of osteoblast differentiation than seen with BMP-2 or BMP-4 treatment, suggesting that BMP-6 may act at an earlier stage of cell differentiation.

In situ detection of tissue factor in vascular endothelial cells: correlation with the malignant phenotype of human breast disease.

Expression of tissue factor (TF) in the endothelium has been observed only rarely in human disease and has been thought to be elaborated on the surface of vascular endothelial cells (VECs) in vitro as an artifact of tissue culture. Using monoclonal antibodies and a novel probe for functional TF, we have localized TF to the VECs (and tumor cells) within the tumors of seven patients with invasive breast cancer but not in the VECs (or tumor cells) of benign tumors from ten patients with fibrocystic disease of the breast. The potent procoagulant TF was shown to be a marker of the initiation of angiogenesis in human breast cancer. Further evidence that the TF was the demonstration of a similar distribution of cross-linked fibrin only in the VECs of the malignant tumors. We interpret these data as further support for the concept that tumor cells can activate nearby VECs and regulate blood vessel growth in vivo. Large clinical pathologic studies will be necessary to determine whether TF is a useful marker for the "switch" to the angiogenic phenotype" in human breast disease and/or correlates with the thromboembolic complications of breast cancer.

Tissue factor expression in human leukemic cells.

Patients with acute leukemia are at increased risk for thrombotic and hemorrhagic complications, particularly those patients with acute promyelocytic leukemia (APL) undergoing induction chemotherapy. These serious complications have been attributed by some authors to the release of tissue factor (TF) procoagulant activity (PCA), particularly during cytotoxic chemotherapy. In previous studies of normal peripheral blood cells, only cells of the monocyte lineage have been found to express TF PCA. Therefore, several questions remain regarding the origin and characterization of the PCA in malignant leukemic cells, particularly those thought to be derived from granulocyte progenitor cells. We utilized a full-length cDNA probe, several monoclonal antibodies (MAbs) and a sensitive one-stage PCA assay to study the expression of TF in the human cell line, HL-60, in human peripheral blood monocytes/macrophages (Mo/Mø) and in highly purified populations of human polymorphonuclear leukocytes (PMN). In the HL-60 cells we detected low but significant levels of TF mRNA and TF antigen (TF:Ag). In unstimulated cells, coordinate increased levels of TF mRNA, TF:Ag and TF PCA expression were noted following phorbol-ester-induced macrophage differentiation of the cells, but a decreased level of TF mRNA with no change in the basal level of TF:Ag expression occurred following retinoic acid-induced granulocyte differentiation of this cell line. Long-term cultures of stimulated mature Mo/Mø demonstrated initial coordinate expression of TF mRNA, TF:Ag and TF PCA, but TF:Ag expression persisted even after 7 days (when TF PCA was undetectable). No TF PCA, TF:Ag or TF mRNA was demonstrated in highly purified populations of human PMN, regardless of culture conditions. Discordant expression of TF mRNA, TF:Ag and TF PCA in HL-60 cells suggests the possibility of novel, post-synthetic mechanisms for the regulation of TF PCA expression, which might be dependent on the phenotypic differentiation level of the cell. Such mechanisms (yet to be defined) might account for the ability of some leukemic cells, which frequently express characteristics of more than one cell line (e.g. monocytes and granulocytes), to express a TF gene product capable of activating blood coagulation.

Tissue factor expression in human leukocytes and tumor cells.

Tissue factor (TF) exists in a cryptic form [i.e. without procoagulant activity (PCA)] in peripheral blood monocytes and quiescent tissue macrophages but is expressed constitutively in most human tumor cells. Induction and cell surface expression of TF in these cells in vivo is associated with activation of intravascular and extravascular coagulation in patients with a variety of inflammatory or malignant diseases. The regulation of TF synthesis in cells is complex and new information from transfection studies suggests that changes in cellular glycosylation pathways impair cell surface expression of functional TF. Such dysregulation may also characterize the lineage-unfaithful expression of TF in leukemic cells and perhaps explain some of the thrombohemorrhagic complications in patients with acute progranulocytic leukemia. The importance of carbohydrate modification of TF is reviewed.

Recurrent venous thrombosis as the presenting manifestation of acute lymphocytic leukemia: leukemic cell procoagulant activity is not responsible for the hypercoagulable state.

The association of cancer with clinical abnormalities of blood coagulation, including superficial thrombophlebitis, deep vein thrombosis (DVT), and disseminated intravascular coagulation (DIC) is well-known, particularly in patients with solid tumors and acute promyelocytic leukemia (APL). Less commonly appreciated is the potential for the development of venous thromboembolic disease (TED) in patients with acute lymphocytic leukemia (ALL). Multiple mechanisms have been implicated for the activation of coagulation in these patients, with an emphasis on the contribution made by the procoagulant properties of the tumor cells themselves. We present two cases of patients with pre-B cell ALL, both of whom developed recurrent TED as the presenting manifestation of their leukemia and/or heralding relapse. The blast cells from one of the patients were studied for the presence of procoagulant activity (PCA) and by Northern blot analysis for tissue factor (TF) messenger RNA (mRNA). Neither PCA nor TF mRNA could be identified in highly purified populations of the lymphoblast cells. We conclude that recurrent TED can be a manifestation of ALL and that mechanisms other than the release of tumor cell procoagulants should be sought to explain the pathogenesis of thrombosis in some patients.

In situ characterization of antigenic and functional tissue factor expression in human tumors utilizing monoclonal antibodies and recombinant factor VIIa as probes.

Tissue factor (TF), the primary initiator of blood coagulation in vivo, is expressed in vitro by a variety of cells. Previous efforts to localize TF in tissue and cells have been limited principally to the use of immunological techniques. In the present study, we describe a novel functional probe for TF expression, which can be utilized to localize functional TF in situ in human cells and tissues. This probe, a biotinylated phe-pro-arg-chloro-methyl-ketone-labeled rVIIa (FPR-ck-VIIa), interacts with TF via high-affinity binding sites. The binding of FPR-ck-VIIa, therefore, can be correlated with the ability of TF to activate clotting. In the described studies, TF antigen (TF:Ag) expression was examined immunohistochemically with various TF-specific monoclonal antibodies (MAbs) and was correlated with functional TF expression using our novel TF-binding probe (eg, FPR-ck-VIIa). Initial results indicate that TF:Ag expression correlates with the expression of functional TF (TF:VIIa), and the specificity of both types of probes was confirmed. Parallel antigenic and functional TF expression in situ was demonstrated in various human tumors. We believe this to be the first demonstration of functional TF in situ in human cells and tissues. We suggest that FPR-ck-VIIa should prove a useful reagent for studying the role of TF in the pathogenesis of clotting complications of human disease.

Characterization of human osteoblast and megakaryocyte-derived osteonectin (SPARC).

Osteonectin is an adhesive, cell, and extracellular matrix-binding glycoprotein found primarily in the matrix of bone and in blood platelets in vivo. Osteonectins isolated from these two sources differ with respect to the complexity of their constituent N-linked oligosaccharide. In this study, osteonectin synthesized by bone-forming cells (osteoblasts) and platelet-producing cells (megakaryocytes) in vitro was analyzed to determine if the proteins produced were analogous in terms of glycosylation to those isolated from bone and platelets, respectively. Immunoblot analyses of osteonectin produced by the osteoblast-like cell lines, SaOS-2 and MG-63, indicated that secreted and intracellular forms of the molecule are structurally distinct. Endoglycosidase treatment and immunoblotting of osteonectin secreted from SaOS-2 and MG-63 cells, under serum-deprived conditions, suggested that the molecule possessed a complex type oligosaccharide unlike the high-mannose moiety found on bone matrix-derived osteonectin. Biosynthetic labeling of SaOS-2 cells and human megakaryocytes indicated that both cell types synthesize osteonectin de novo. Electrophoretic and glycosidase sensitivity analyses of [35S]-osteonectin isolated from lysates of metabolically labeled SaOS-2 cells and megakaryocytes indicated that these two cell types synthesize osteonectin molecules that are identical in oligosaccharide structure to the isolated bone and platelet proteins. These data suggest that the intracellular form of the osteonectin molecule is glycosylated differently in SaOS-2 cells and megakaryocytes but that the extracellular form which is secreted from platelets in vivo and osteoblasts in vitro is characterized by the presence of a complex type N-linked oligosaccharide.