A phase III clinical trial of a mixture agent of plasma-derived factor VIIa and factor X (MC710) in haemophilia patients with inhibitors.

INTRODUCTION: MC710, a 1:10 protein weight ratio mixture of plasma-derived activated factor VII (FVIIa) and factor X (FX), is a novel bypassing agent for haemostasis in haemophilia patients with inhibitors. We evaluated the haemostatic efficacy and safety of one to two administrations of MC710 in 21 joint, muscle, and subcutaneous bleeding episodes in 14 male patients, in a multi-centre, open-label, non-randomized clinical trial. METHODS: Subjects were intravenously administered one or two doses of 60 or 120 µg kg-1 MC710 (as FVIIa) once or twice (to a maximum of 180 µg kg-1 ) over up to five bleeding episodes per subject. The haemostatic efficacy of MC710 was determined for each episode by investigator evaluation, using changes in visual analogue scale (VAS) for pain relief, and/or knee joint or muscle circumference for swelling reduction, and range of motion (ROM) for improvement of joint mobility. RESULTS: In 21 treatments for bleeding episodes, 19 were rated "excellent" or "effective" 8 h after the last treatment. VAS significantly decreased over time, and ROM significantly improved over time compared with the values before treatment. One mild adverse reaction, decreased blood potassium, and two serious adverse events, both knee joint bleeding, were observed within 1 week after first administration, with no significant effect on safety. Furthermore, diagnostic markers did not show any signs of disseminated intravascular coagulation (DIC). CONCLUSION: These results show that MC710 has sufficient haemostatic efficacy and safety, and can be used as a potential bypassing agent to control bleeding in haemophilia patients with inhibitors.

Activated protein C prevents endotoxin-induced hypotension in rats by inhibiting excessive production of nitric oxide.

BACKGROUND: Excessive production of nitric oxide (NO) by the inducible isoform of NO synthase (iNOS) is critically involved in endotoxin (ET)-induced hypotension. Tumor necrosis factor-alpha (TNF-alpha) plays an important role in induction of iNOS. Because activated protein C (APC), a physiological anticoagulant, inhibits TNF-alpha production, it might prevent hypotension by inhibiting excessive production of NO. In this study, we examined this possibility using a rat model of septic shock. METHODS AND RESULTS: Intravenous administration of APC prevented both ET-induced hypotension and the increases in plasma levels of NO(2)(-)/NO(3)(-). The hypotension was also inhibited when APC was administered 30 minutes after ET administration. APC inhibited the increases in lung levels of iNOS activity by inhibiting expression of iNOS mRNA in animals given ET. APC significantly inhibited the increases in lung tissue levels of TNF-alpha and expression of TNF-alpha mRNA in animals given ET. Neither DEGR-F.Xa, a selective inhibitor of thrombin generation, nor DIP-APC, an active site-blocked APC, showed any effect on these ET-induced changes. Both inhibition of TNF-alpha production by leukocytopenia and treatment with anti-rat TNF-alpha antibody produced effects similar to those induced by APC. Aminoguanidine, a selective inhibitor of iNOS, inhibited both the hypotension and the increases in plasma levels of NO(2)(-)/NO(3)(-) in this animal model. CONCLUSIONS: These observations strongly suggest that APC inhibits iNOS induction by decreasing TNF-alpha production, leading to the prevention of ET-induced hypotension. Furthermore, such effects of APC were not dependent on its anticoagulant effects but rather on its serine protease activity.

Role of neutrophils in spinal cord injury in the rat.

Activated neutrophils are thought to be involved in tissue injury through the release of various inflammatory mediators. To understand the role of neutrophils in spinal cord injury, the effects of nitrogen mustard-induced leukocyte depletion and the administration of an anti-P-selectin monoclonal antibody on motor disturbances observed following spinal cord compression were examined in rats. Spinal cord injury was induced by applying a 20-g weight for 20 min at the level of the 12th thoracic vertebra, resulting in motor disturbances of the hindlimbs 24 h postcompression. Motor disturbances, evaluated using Tarlov's index, an inclined-plane test and climbing ability, were markedly attenuated in rats with nitrogen mustard-induced leukocytopenia. Administration of the anti-P-selectin monoclonal antibody, by which adhesion of activated neutrophils to endothelial cells may be inhibited, also attenuated motor disturbances. Histological examination revealed that intramedullary hemorrhages observed 24 h after compression at the 12th thoracic vertebra of the spinal cord were significantly attenuated in leukocytopenic animals and those which received the anti-P-selectin monoclonal antibody. The accumulation of neutrophils at the site of compression, as evaluated by measuring the tissue myeloperoxidase activity, significantly increased with time following the compression, peaking at 3 h postcompression. Spinal cord myeloperoxidase activity did not increase in sham-operated animals. Leukocyte depletion and administration of the anti-P-selectin monoclonal antibody both reduced the accumulation of neutrophils in the damaged spinal cord segment 3 h postcompression. These observations strongly suggest that activated neutrophils play an important role in compression-induced thoracic spinal cord injury and that a P-selectin-mediated interaction between activated neutrophils and endothelial cells may be a critical step in endothelial cell injury leading to spinal cord injury.

Neuroprotection by recombinant thrombomodulin.

We examined whether recombinant human soluble thrombomodulin (rhs-TM) reduces compression trauma-induced spinal cord injury through protein C activation in rats. Administration of rhs-TM, either before or after the induction of spinal cord injury (SCI), markedly reduced the resulting motor disturbances. However, neither rhs-TM pretreated with an anti-rhs-TM monoclonal antibody (MAb) F2H5, which inhibits thrombin binding to rhs-TM, nor those pretreated with MAb R5G12, which selectively inhibits protein C activation by rhs-TM, prevented the motor disturbances. Intramedullary hemorrhages, observed 24 h after trauma, were significantly reduced in animals given rhs-TM. The increase in the tissue levels of tumor necrosis factor-alpha (TNF-alpha), TNF-alpha mRNA expression, and the accumulation of leukocytes in the damaged segment of the spinal cord were significantly inhibited in animals receiving rhs-TM, but these effects were not observed following administration of rhs-TM pretreated with MAb R5G12 or MAb F2H5. Leukocytopenia and activated protein C all produced effects similar to those of rhs-TM. These findings suggest that rhs-TM prevents compression trauma-induced SCI by inhibiting leukocyte accumulation by reducing the expression of TNF-alpha mRNA and such therapeutic effects of rhs-TM could be dependent on its protein C activation capacity. Findings further suggest that thrombomodulin can be implicated not only in the coagulation system but in regulation of the inflammatory response.

Recombinant tissue factor pathway inhibitor prevents lipopolysaccharide-induced systemic hypotension in rats by inhibiting excessive production of nitric oxide.

Excessive production of nitric oxide (NO) by the inducible form of NO synthase (iNOS) plays a key role in the development of endotoxin shock. Tumor necrosis factor-alpha (TNF-alpha) induces iNOS, thereby contributing to the development of shock. We recently reported that recombinant tissue factor pathway inhibitor (r-TFPI), an important inhibitor of the extrinsic pathway of the coagulation system, inhibits TNF-alpha production by monocytes. In this study, we investigated whether r-TFPI could ameliorate hypotension by inhibiting excessive production of NO in rats given lipopolysaccharide (LPS). Pretreatment of animals with r-TFPI prevented LPS-induced hypotension. Recombinant TFPI significantly inhibited the increases in both the plasma levels of NO2-/NO3- and lung iNOS activity 3 h after LPS administration. Expression of iNOS mRNA in the lung was also inhibited by intravenous administration of r-TFPI. However, neither DX-9065a, a selective inhibitor of factor Xa, nor an inactive derivative of factor VIIa (DEGR-F.Vlla) that selectively inhibits factor VIIa activity, had any effect on LPS-induced hypotension despite their potent anticoagulant effects. Moreover, neither the plasma levels of NO2-/NO3- nor lung iNOS activity were affected by administration of DX-9065a and DEGR-F.VIIa. These results suggested that r-TFPI ameliorates LPS-induced hypotension by reducing excessive production of NO in rats given LPS and this effect was not attributable to its anticoagulant effects, but to the inhibition of TNF-alpha production.

Recombinant human soluble thrombomodulin reduces endotoxin-induced pulmonary vascular injury via protein C activation in rats.

Adult respiratory distress syndrome (ARDS) is a serious complication of disseminated intravascular coagulation (DIC) or multiple organ failure. To determine whether recombinant soluble human thrombomodulin (rsTM) may be useful in treating ARDS due to sepsis, we investigated the effect of rsTM on lipopolysaccharide (LPS)-induced pulmonary vascular injury in rats. The intravenous administration of rsTM prevented the increase in pulmonary vascular permeability induced by LPS. Neither heparin plus antithrombin III (AT III) nor dansyl Glu Gly Arg chloromethyl ketone-treated factor Xa (DEGR-Xa), a selective inhibitor of thrombin generation, prevented LPS-induced vascular injury. The agents rsTM, heparin plus AT III, and DEGR-Xa all significantly inhibited the LPS-induced intravascular coagulation. Recombinant soluble TM pretreated with a monoclonal antibody (moAb) that inhibits protein C activation by rsTM did not prevent the LPS-induced vascular injury; in contrast, rsTM pretreated with a moAb that does not affect thrombin binding or protein C activation by rsTM prevented vascular injury. Administration of activated protein C (APC) also prevented vascular injury. LPS-induced pulmonary vascular injury was significantly reduced in rats with leukopenia induced by nitrogen mustard and by ONO-5046, a potent inhibitor of granulocyte elastase. Results suggest that rsTM prevents LPS-induced pulmonary vascular injury via protein C activation and that the APC-induced prevention of vascular injury is independent of its anticoagulant activity, but dependent on its ability to inhibit leukocyte activation.

A novel platelet activating factor antagonist, SM-12502, attenuates endotoxin-induced disseminated intravascular coagulation and acute pulmonary vascular injury by inhibiting TNF production in rats.

Adult respiratory distress syndrome and disseminated intravascular coagulation are important pathologic conditions affecting the outcome of patients with sepsis. To elucidate the possible therapeutic efficacy of SM-12502, a novel platelet activating factor antagonist, on acute lung injury and disseminated intravascular coagulation in sepsis, we investigated the effect of SM-12502 on an endotoxin (ET)-induced septic model in rats. SM-12502 prevented ET-induced increases in pulmonary vascular permeability and ET-induced histologic changes, such as leukocyte infiltration and pulmonary interstitial edema, 6 h following the administration of ET (5 mg/kg). SM-12502 also inhibited the decrease in fibrinogen and the increase in fibrin and fibrinogen degradation products observed following ET administration. SM-12502 prevented increases in the serum concentration of tumor necrosis factor (TNF) 90 min following ET administration in vivo, and significantly inhibited the production of TNF-alpha by ET-stimulated monocytes in vitro. These findings suggest that SM-12502 attenuates the actions of endotoxin by the inhibition of TNF production.

Effects of plasma kallikrein specific inhibitor and active-site blocked factor VIIa on the pulmonary vascular injury induced by endotoxin in rats.

The acute respiratory distress syndrome (ARDS) is a serious complication of sepsis. To evaluate the role of the coagulation system in the pathogenesis of ARDS in sepsis, we examined the effects of the administration of a synthetic plasma kallikrein specific inhibitor (PKSI) and of active-site blocked factor VIIa (DEGR-VIIa) on the pulmonary vascular injury induced by E. coli endotoxin (ET) in rats. Administration of PKSI prevented the pulmonary vascular injury induced by ET as well as pulmonary histological changes in animals administered ET, but it did not affect the intravascular coagulation. The opposite effect was seen with DEGR-VIIa, which prevented the intravascular coagulation but not the pulmonary vascular injury. PKSI did not inhibit the activation of the complement system induced by ET leading to the activation of neutrophils. Findings suggest that PKSI may prevent the pulmonary vascular injury induced by ET by inhibiting kallikrein, which activates the neutrophils. The intrinsic pathway of coagulation may be more important than the extrinsic pathway in the pulmonary vascular injury produced by ET.

Methylprednisolone reduces spinal cord injury in rats without affecting tumor necrosis factor-alpha production.

Methylprednisolone (MPS) is the only therapeutic agent currently available for traumatic spinal cord injury (SCI). However, little is known about its therapeutic mechanisms. We have demonstrated that tumor necrosis factor-alpha (TNF-alpha) plays a critical role in posttraumatic SCI in rats. Since MPS has been shown to inhibit TNF-alpha production in vitro, it is possible that MPS can reduce SCI by inhibiting TNF-alpha production. To examine this possibility, we investigated the effect of MPS on TNF-alpha production in injured segments of rat spinal cord. Leukocytopenia and high-dose intravenous administration of MPS markedly reduced the motor disturbances observed following spinal cord trauma. Both treatments also reduced the intramedullary hemorrhages observed histologically 24 hr posttrauma. Leukocytopenia significantly reduced tissue levels of both TNF-alpha mRNA and TNF-alpha, 1 and 4 hr posttrauma, respectively, and it also inhibited the accumulation of leukocytes in the injured segments 3 hr posttrauma, while MPS had no effects. Lipid peroxidation and vascular permeability at the site of spinal cord lesion were both significantly increased over time after the induction of SCI, peaking 3 hr posttrauma. These events were significantly reduced in animals with leukocytopenia and in those given anti-P-selectin monoclonal antibody compared to sham-operated animals. Administration of MPS significantly inhibited both the increase in lipid peroxidation and the vascular permeability. These findings suggested that MPS reduces the severity of SCI, not by inhibiting the production of TNF-alpha at the site of spinal cord trauma, but by inhibiting activated leukocyte induced lipid peroxidation of the endothelial cell membrane. This suggests that MPS may attenuate spinal cord ischemia by inhibiting the increase in endothelial permeability at the site of spinal cord injury.

Effects of various doses of antithrombin III on endotoxin-induced endothelial cell injury and coagulation abnormalities in rats.

We previously demonstrated that antithrombin III reduced the injury to endothelial cells caused by activated leukocytes in rats administered endotoxin. This occurred via the increase of the endothelial release of prostaglandin I2, which is a potent inhibitor of leukocyte activation. We evaluated the dose of antithrombin III required to prevent such endothelial cell injury in rats administered endotoxin, by comparing the effects of various antithrombin II doses on the pulmonary vascular injury. The intravenous administration of endotoxin, 5 mg/kg, produced a transient accumulation of leukocytes in the lung, followed by pulmonary vascular injury, as indicated by an increase in the pulmonary vascular permeability, and coagulation abnormalities. The dose of 250 U/kg significantly inhibited all such effects of endotoxin. While lower doses of antithrombin III (50 and 100 U/kg) significantly inhibited such coagulation abnormalities, they failed to prevent either the pulmonary accumulation of leukocytes or the subsequent pulmonary vascular injury. Rats administered endotoxin exhibited an accumulation of neutrophils and edematous changes in the pulmonary interstitial space. Although such changes were reduced after 250 U/kg of antithrombin III, they were unaffected by lower doses of 50 and 100 U/kg. Plasma levels of 6-keto-PGF1alpha were markedly increased in rats 90 min after the administration of endotoxin, and were significantly decreased in the endotoxin-treated rats administered the lower doses of antithrombin III (50 and 100 U/kg), but not altered in those endotoxin-treated rats receiving 250 U/kg of antithrombin III. These findings suggest that a higher antithrombin III dose is necessary to prevent endothelial cell injury than is required to inhibit coagulation abnormalities in an animal model of sepsis. These observations support the notion that antithrombin III may prevent endotoxin-induced endothelial cell injury by promoting endothelial release of prostaglandin I2 and thus inhibiting leukocyte activation.

Effects of antithrombin III (AT III) and Trp49-modified AT III on plasma level of 6-keto-PGF1 alpha in rats.

We evaluated the effect of antithrombin III (AT III) on the plasma level of 6-keto-PGF1 alpha in rats to determine whether AT III may promote the release of prostacyclin (PGl2) from endothelial cells in vivo. The intravenous administration of AT III (250 U/kg) significantly increased the plasma levels of 6-keto-PGF1 alpha, with a peak seen 90 min post-administration. Neither Trp49-modified AT III, which lacks affinity for heparin but retains an inhibitory capacity for thrombin, nor heparin plus AT III, increased the plasma level of 6-keto-PGF1 alpha 90 min after administration. Indomethacin pretreatment inhibited the increase in plasma levels of 6-keto-PGF1 alpha produced by AT III. Observations suggest that AT III may promote the release of PGl2 from endothelial cells by interacting with heparin-like glycosaminoglycans in vivo, consistent with previous observations in cultured endothelial cells.

Endotoxin-induced pulmonary vascular injury is mainly mediated by activated neutrophils in rats.

Acute respiratory failure is a common complication in patients with disseminated intravascular coagulation associated with sepsis. To elucidate the role of coagulation abnormalities in acute lung injury in sepsis, we investigated the effect of anticoagulants on the pulmonary vascular injury in rat induced by lipopolysaccharide (LPS). When administered intravenously, LPS (5 mg/kg body weight) significantly increased the accumulation of 111indium-labeled neutrophils in lung 30 min after administration. Subsequently, the pulmonary vascular permeability and the serum level of fibrin and fibrinogen degradation products (E) [FDP (E)] increased and remained elevated for several hours. Neither heparin alone, heparin plus antithrombin III, or dansyl-Glu-Gly-Arg-chloromethyl ketone-treated factor Xa, a selective inhibitor of thrombin generation, prevented LPS-induced vascular injury 6 hours after LPS administration, whereas these substances significantly inhibited the increase in serum FDP (E) at that time. LPS-induced pulmonary vascular injury was significantly attenuated in rats with methotrexate-induced leukocytopenia or treated with ONO-5046, a potent granulocyte elastase inhibitor, although ONO-5046 did not inhibit the LPS-induced increase in serum FDP (E). Thus, activated leukocytes play a more important role than coagulation abnormalities in the pathogenesis of LPS-induced pulmonary vascular injury in an experimental rat model of endotoxemia.

Effect of nafamostat mesilate, a synthetic protease inhibitor, on tissue factor-factor VIIa complex activity.

Nafamostat mesilate (NM), a synthetic protease inhibitor, is frequently used for the treatment of disseminated intravascular coagulation (DIC) in Japan. NM inhibits several proteases which may be importantly involved in the pathophysiology of DIC. Since tissue factor (TF) plays a critical role in DIC associated with septicemia, inhibition of the extrinsic pathway of coagulation by coagulation inhibitors may be useful for the treatment of DIC. NM inhibited extrinsic pathway activity (TF-F.VIIa mediated-F.Xa generation) in a concentration dependent manner; the IC50 was 1.0 x 10(-7) M. F.Xa was not inhibited by NM at the concentrations used in the experiment, suggesting that NM might inhibit TF-F.VIIa complex activity. When incubated with TF-F.VIIa complex, NM inhibited the complex activity with an IC50 of 1.5 x 10(-7) M, the same value that found for inhibition of extrinsic pathway activity. A Lineweaver-Bulk's plot of the inhibition demonstrated that NM inhibited TF-F.VIIa complex in a competitive fashion, with an inhibition constant (Ki) of 2.0 x 10(-7) M. These findings suggested that NM may be a potent inhibitor of TF-F.VIIa complex and the therapeutic effect of NM in DIC patients could be partly explained by inhibition of the extrinsic pathway of the coagulation system.

Pulmonary vascular injury induced by hemorrhagic shock is mediated by P-selectin in rats.

To investigate whether the P-selectin-mediated leukocyte adhesion to the endothelial cells is involved in pulmonary vascular injury after hemorrhagic shock, we examined the effect of an anti-P-selectin monoclonal antibody (MAb PB1.3) on the pulmonary accumulation of leukocytes and the subsequent pulmonary vascular injury observed after hemorrhagic shock in rats. Two hours after hemorrhagic shock, pulmonary accumulation of leukocytes, as evaluated by measuring myeloperoxidase activity, began to increase and peaked after 6 hours. Pulmonary vascular injury, as evaluated by the extravascular leakage of 125I-albumin, was significantly increased 6 hours after hemorrhagic shock. MAb PB1.3 significantly prevented both the pulmonary accumulation of leukocytes and subsequent pulmonary vascular injury. MAb PNB1.6, an anti-P-selectin monoclonal antibody incapable of inhibiting P-selectin-mediated leukocyte adhesion, did not prevent either of these effects. These observations strongly suggest that the pulmonary sequestration of leukocytes and the subsequent pulmonary vascular injury after hemorrhagic shock are mediated by P-selectin.

Reduction of spinal cord injury by administration of iloprost, a stable prostacyclin analog.

To investigate whether iloprost, a stable analog of prostacyclin, is useful for the prevention of posttraumatic spinal cord injury, we examined its effects on compression trauma-induced spinal cord injury in rats. Spinal cord injury was induced by applying a 20-g weight for 20 minutes to the spinal cord at the level of T-12, resulting in motor disturbances in the hindlimbs. These motor disturbances, evaluated using Tarlov's index, were markedly attenuated in rats with nitrogen mustard-induced leukocytopenia. Administration of iloprost also attenuated the motor deficits. Histological examination revealed that intramedullary hemorrhages observed 24 hours after trauma were significantly attenuated in leukocytopenic animals and in animals that received iloprost. The accumulation of leukocytes at the site of trauma, evaluated by measuring tissue myeloperoxidase activity, significantly increased with time following the trauma, peaking at 3 hours postinjury. Spinal cord myeloperoxidase activity in sham-operated animals did not increase postoperatively. Leukocyte depletion and administration of iloprost reduced the accumulation of leukocytes in the damaged spinal cord segment 3 hours posttrauma. These findings indicate that iloprost attenuates motor disturbances induced by spinal cord trauma and that its therapeutic efficacy can be partly explained by its inhibition of leukocyte accumulation at the traumatized site.

[Endothelial cells and coagulation abnormalities].

Endothelial cells have two important anticoagulant systems, heparan sulfate-antithrombin system and thrombomodulin-protein C system. Under physiological conditions, these two systems inhibit activation of coagulation on endothelial cells. However, under inflammatory conditions, tumor necrosis factor(TNF)-alpha or other cytokines produced by monocytes reduce the anticoagulant properties of endothelial cell by downregulating expression of heparan sulfate and thrombomodulin on endothelial cells. Antithrombin stimulates prostacyclin generation from endothelial cells by interacting with heparan sulfate of endothelial cells and generated prostacyclin inhibits TNF-alpha production by monocytes. Activated protein C inhibits TNF-alpha production by monocyte dependent of its protease activity. Thus, antithrombin and activated protein C might inhibit the endothelial perturbation induced by cytokines. Antithrombin regulates TNF-alpha induced tissue factor expression on endothelial cells by an unknown mechanism. Thus, antithrombin and activated protein C might be useful agents for treating coagulation abnormalities associated with sepsis or other inflammation because these agents inhibit not only coagulation but also downregulation of anticoagulant activities of endothelial cells.

[Studies on the role of leukocytes in the activation of intravascular coagulation in septicemia].

To elucidate the role of leukocytes in intravascular clotting in patients with septicemia, plasma levels of thrombin-antithrombin III complex (TAT), soluble fibrin monomer complex (SFMC) and fibrinogen (Fbg) were determined in 33 patients with septicemia. Twenty out of 33 patients revealed a marked leukopenia (leukocyte count was less than 1,000/microliters) due to suppression of hematopoiesis by the administration of chemotherapeutic agents for the treatment of hematological malignancies. Thirteen out of 33 patients showed normal or increased leukocyte counts. Plasma levels of TAT and SFMC in septicemic patients with leukopenia were significantly lower than those in patients whose leukocyte counts were higher than 1,000/microliters. Plasma fibrinogen levels were significantly lower in patients whose leukocyte counts were higher than 1,000/microliters than those in patients with leukopenia. Plasma TAT levels were found to correlate well with the number of leukocytes in these patients with the correlation coefficient (R) of 0.67 (p less than 0.001). Significantly high positive correlation was observed between plasma TAT levels and the number of monocytes (R = 0.92, p less than 0.001). Significant correlation was also observed between plasma SFMC levels and the number of monocytes (R = 0.72, p less than 0.001). No significant correlation was found between the number of platelets and TAT levels. These findings suggest that leukocytes (especially monocytes) play a critical role in activating intravascular coagulation in septicemic patients.

[Reactive histiocytosis during initial remission induction therapy for acute myeloblastic leukemia].

Six of 14 patients with acute myeloblastic leukemia (AML) complicated reactive histiocytosis during initial remission induction therapy. All six patients had a high fever without signs of infection during initial chemotherapy, and periods of myelosuppression were prolonged. Histiocytes with a mature appearance, some of which phagocyted erythrocytes, thrombocytes or neutrophils, increased in the bone marrow. All of 3 patients tested showed high serum levels of ferritin. Two of 3 patients treated with 125 mg/day methylprednisolone achieved complete remission. In the remaining 3 patients, one patient achieved complete remission, but the others died of fungal pneumonia or sepsis. Thus, reactive histiocytosis is one of the severe complications in patients with AML undergoing chemotherapy.