Impact of abciximab and coronary stenting on outcomes and costs of percutaneous coronary interventions in a community hospital.

OBJECTIVE: To assess costs and outcomes of coronary stenting and balloon angioplasty with and without adjunctive treatment with abciximab for 3758 consecutive elective percutaneous coronary interventions at a single community center over the 2.5-year period between 1 January 1995 and 30 June 1997. RESULTS: Abciximab was more common among patients who had recently suffered myocardial infarction, patients with unstable angina, and patients with more complex coronary lesions. Use of abciximab in conjunction with balloon angioplasty or stenting and stenting alone was associated with significant reductions in incidence of major adverse cardiovascular events in hospital. Multivariate analysis indicated that use of abciximab and stenting were associated with significant independent effects on risk of an event. Hospital costs were increased for patients administered abciximab, treated with stenting, or both. Total costs and costs inclusive of those incurred in catheterization laboratory and pharmacy increased significantly with increasing complexity of lesions. Multivariate regression analysis (baseline cost US$5621) identified death (US$16098), emergency revascularization (US$13678), usage of multiple stents (US$1423 for each stent), and use of abciximab (US$1269) as independent predictors of a greater cost. One-year follow-up revealed significant differences among treatment strategies in terms of risk of need for subsequent revascularization procedures. Lack of stenting but not use of abciximab was identified as a significant predictor of need for repeat revascularization procedures. CONCLUSIONS: Our findings are in general agreement with cost analyses of use of abciximab for populations in clinical trials and suggest that improvements of early clinical outcome with abciximab treatment and stenting justify the incremental cost of treatment in a community hospital setting.

New approaches to diagnosis and management of unstable angina and non-ST-segment elevation myocardial infarction.

Recently, it has been demonstrated in multiple clinical research studies that non-Q-wave myocardial infarction shares many of the features of unstable angina pectoris and that both diseases initially are managed similarly. Important new antiplatelet drugs (glycoprotein IIb-IIIa inhibitors) and antithrombin agents (low-molecular-weight heparin) are currently recommended for patients with unstable angina pectoris/non-ST-segment elevation MI who are at high or intermediate risk on the basis of symptoms, electrocardiographic findings, and the presence or absence of serum markers (eg, troponin I, troponin T, and creatine kinase-MB). This review provides important information concerning the results of clinical studies of glycoprotein IIb-IIIa inhibitors (tirofiban hydrochloride and eptifibatide) when used with unfractionated heparin in patients with this syndrome or with low-molecular weight heparin (enoxaparin sodium) in similar patients. The Thrombolysis in Myocardial Infarction IIIB, Veterans Affairs Non-Q-Wave Infarction Studies in Hospital, and Fast Revascularization During Instability in Coronary Artery Disease II studies evaluating a conservative, ischemia-guided approach vs an early aggressive approach to such patients are presented, with a practical algorithm for treating such patients.

Uneventful use of tirofiban as an adjunct to coronary stenting in a patient with a history of abciximab-associated thrombocytopenia 10 months earlier.

A 52-year-old male undergoing a percutaneous coronary intervention after an inferior infarction treated with thrombolytic therapy was given abciximab with development of early severe reversible thrombocytopenia (platelet count nadir of 50,000 per mm3). The same patient underwent another percutaneous coronary intervention 10 months later and was given tirofiban without development of thrombocytopenia. This observation suggests that the development of thrombocytopenia with abciximab is not necessarily generalizable to other unique IIb/IIIa receptor antagonists. Tirofiban may be an appropriate consideration for patients who would benefit from a IIb/IIIa receptor antagonist and who are not candidates for abciximab therapy due to an episode of abciximab-associated thrombocytopenia. Further investigation of thrombocytopenia associated with IIb/IIIa receptor antagonists is warranted to determine appropriate approaches to the management of patients with a history of abciximab-associated thrombocytopenia.

Clopidogrel as adjunctive antiplatelet therapy during coronary stenting.

OBJECTIVES: We examined the procedural and 30-day clinical outcomes among patients receiving aspirin and either ticlopidine or clopidogrel during coronary stenting. BACKGROUND: Ticlopidine-plus-aspirin has become standard antiplatelet therapy for the prevention of thrombotic complications after coronary stenting. Clopidogrel has a similar mechanism of action as ticlopidine, but both its efficacy and its safety as a pharmacologic adjunct to coronary stenting have not been well described. METHODS: This single-center, prospective analysis examined the in-hospital procedural and 30-day clinical outcomes among 875 consecutive patients undergoing coronary stenting who received adjunctive aspirin and either clopidogrel (n = 514; 58.7%) or ticlopidine (n = 361; 41.3%) therapy. RESULTS: Procedural success rates were similar among the clopidogrel- (99.6%) and ticlopidine-treated patients (99.4%). Subacute stent thrombosis (i.e., >24 h < or =30 days) occurred in one clopidogrel-treated (0.2%) and in one ticlopidine-treated (0.3%) patient (p = 0.99). By 30 days following the index procedure, the combined rates of death, nonfatal myocardial infarction and need for target vessel revascularization were similar among patients who received either clopidogrel (2.1%) or ticlopidine (1.4%; p = 0.57) therapy. CONCLUSIONS: In this analysis the antiplatelet combination therapy of aspirin-plus-clopidogrel was an effective regimen for preventing thrombotic complications and major adverse cardiovascular events among a broad spectrum of patients undergoing coronary artery stenting.

Economic implications of coronary stenting with adjunctive IIb/IIIa receptor antagonists in a community hospital.

To assess the implications of coronary stenting with several IIb/IIIa receptor antagonists, total hospital cost and adverse events were reviewed for 674 elective stent procedures from June 1998 through December 1998. The use of IIb/IIIa receptor antagonism and the agent selected were at the discretion of the interventional cardiologist. In-hospital, 30-day and 6-month adverse cardiac events were similar among the treatment strategies. Target vessel revascularization at six months was similar among the treatment strategies. Patients who received a IIb/IIIa receptor blocker with their stent procedure were less likely to be rehospitalized within 30 days. Multivariate regression analysis identified specific factors responsible for prolongation of hospital stay including adverse cardiac events, physician practice pattern and age greater than 70 years (all p < 0.002). Overall hospital cost for patients receiving tirofiban as an adjunct to coronary stenting was approximately $1,000 less than patients receiving abciximab. Total cath lab expenditures were similar for these groups and the savings in hospital cost was directly attributable to a lower pharmacy cost in the tirofiban group. Multivariate regression analysis identified adverse cardiac events, left ventricular systolic dysfunction, multiple stent placement, physician practice and abciximab as significant contributors to increased hospital cost (all p < 0.002). Tirofiban as an adjunct to coronary stenting was not identified by multivariate analysis as a significant contributor to hospital cost. Bleeding rates were similar among the treatment strategies. Thus, coronary stenting in our community hospital is associated with acceptable outcomes regardless of treatment strategy and hospital cost is significantly influenced by the use of IIb/IIIa blockade with stenting and the type of agent selected.

One balloon, one stent: an effective cost-containment strategy for elective stent deployment.

A cost-containment strategy to reduce stent procedure-related resources utilizing an integrated system comprised of a Stablizer guide wire, a Brite-Tip guiding catheter, and a single Titan balloon catheter for both lesion predilatation and poststent deployment was compared to a conventional strategy utilizing a nonintegrated guide wire, guiding catheter, and balloon components. Both groups were comparable with respect to demographics, number of lesions stented, and stents deployed per lesion. No differences in lesion length or pre- and poststent minimal luminal diameter were observed. Balloon use was significantly reduced using the integrated strategy when compared to the conventional strategy (1.3 +/- 0.5 vs. 2.1 +/- 1.1; P < 0.01); overall nonstent-related resource utilization was significantly reduced ($747 +/- $401 vs. $1,093 +/- $467; P < 0.01). Procedural success rates were identical in both groups (100%), and no patient sustained subacute stent thrombosis or required target vessel revascularization at 1 mo follow-up. We conclude that the use of a single Titan balloon catheter as part of an integrated cost-containment strategy for both lesion predilatation and poststent deployment results in considerable cost savings while maintaining high procedural and clinical success rates.

Nephrotoxicity from contrast media: attenuation with theophylline.

PURPOSE: To determine if depression of creatinine clearance after administration of contrast medium may be prevented with theophylline. MATERIALS AND METHODS: A nonionic, low-osmolality contrast medium (iopamidol) or an ionic, high-osmolality contrast medium (sodium diatrizoate) was administered to 93 patients. Before the examination, these patients were given theophylline or a placebo orally. There were also 30 patients who received an adenosine-uptake inhibitor (dipyridamole). Creatinine clearance and urinary adenosine levels were measured before and after angiography. RESULTS: Creatinine clearance decreased 18% +/- 4 in the placebo-iopamidol group but did not decrease in the theophylline group; urinary adenosine increased 67% +/- 7. Creatinine clearance decreased 42% +/- 5 in the placebo-sodium diatrizoate group and decreased 24% +/- 3 in the theophylline group; urinary adenosine increased 119% +/- 8. In the dipyridamole group in which iopamidol was given, urinary adenosine increased 96% +/- 7 and creatinine clearance decreased 37% +/- 5. CONCLUSION: Intrarenal adenosine can be implicated in the pathogenesis of hypertonic contrast medium nephrotoxicity.

Abnormal expression of plasminogen activators in aortic aneurysmal and occlusive disease.

PURPOSE AND METHODS: Aortic aneurysms are characterized by the destruction of the extracellular matrix of the media, whereas occlusive disease involves excess matrix accumulation within the intima. Plasmin degrades extracellular matrix directly and indirectly by activation of latent metalloenzymes. To determine the expression of tissue- and urokinase-type plasminogen activators, immunoassay, fibrin autography, Northern analysis, and immunohistochemistry were performed on specimens of aneurysmal (n = 12), occlusive (n = 8), and healthy (n = 6) aorta. RESULTS: Immunoassay of tissue-type plasminogen activator revealed 8.7 +/- 0.9 ng tissue-type plasminogen activator/mg extracted protein in aneurysmal aorta, 5.7 +/- 0.3 ng/mg in normal aorta, and 2.5 +/- 0.3 ng/mg in occlusive aorta (p < 0.05 for comparisons between all groups). No urokinase-type plasminogen activator antigen was detected by urokinase-type plasminogen activator immunoassay. Fibrin autography exhibited lytic activity at 64 kDa and 54 kDa attributable to tissue-type plasminogen activator and urokinase-type plasminogen activator. The vast majority of fibrinolysis was secondary to free tissue-type plasminogen activator and was greatest in aneurysmal disease and least in occlusive disease. There was only a small amount of lysis secondary to urokinase-type plasminogen activator. Expression of tissue-type plasminogen activator and urokinase-type plasminogen activators mRNA was comparable in aneurysmal and occlusive aortas. In contrast to occlusive disease, aneurysms had an inflammatory cell infiltrate characterized by the expression of urokinase-type plasminogen activator by specific mononuclear cells. Tissue-type plasminogen activator expression was evident in the intima of normal and diseased aorta and in the media of diseased aorta. CONCLUSION: Differential expression of plasminogen activators within the arterial wall may contribute to the unique pathogenesis of aneurysmal and occlusive aortic disease.

Factors responsible for impaired fibrinolysis in obese subjects and NIDDM patients.

Accelerated atherosclerosis is the leading cause of death in patients with non-insulin-dependent diabetes mellitus (NIDDM). Impaired endogenous fibrinolytic activity may accelerate atherosclerosis by exposing vascular luminal wall surfaces to persistent and recurrent thrombi and clot-associated mitogens. This study was conducted to further characterize endogenous fibrinolysis in lean and obese nondiabetic subjects and in NIDDM patients and to identify mechanisms responsible for the alterations identified. Obese and diabetic subjects had threefold elevations of plasma concentrations of plasminogen activator inhibitor type 1 (PAI-1) compared with values in lean control subjects. Despite the lack of significant differences in plasma concentrations of tissue-type plasminogen activator in the obese and diabetic subjects, both basal and stimulated endogenous fibrinolytic activities were decreased. The decreases were associated with increased activity of PAI-1 in plasma, in turn correlated with increased concentrations of immunoreactive insulin and C-peptide. These results are consistent with our previous observations demonstrating direct stimulatory effects of insulin and its precursors on cellular expression of PAI-1 in vitro and observations by others demonstrating decreased basal fibrinolytic activity in NIDDM patients. Impaired endogenous fibrinolytic activity could lead to prolonged or recurrent exposure of luminal surfaces of vessel walls to microthrombi and clot-associated mitogens that may accelerate atherosclerosis in hyperinsulinemic subjects.

Induction of endothelial cell expression of the plasminogen activator inhibitor type 1 gene by thrombosis in vivo.

BACKGROUND: We have shown previously that products from activated platelets can augment synthesis of plasminogen activator inhibitor type 1 (PAI-1) in cultured endothelial and hepatoma (Hep G2) cells in vitro and increase plasma PAI-1 activity in vivo in rabbits. Accordingly, the effects of activation of platelets associated with thrombosis and thrombolysis in vivo on plasma PAI-1 activity and expression of the PAI-1 gene in endothelium, liver, and other organs were characterized. METHODS AND RESULTS: Endothelial injury giving rise to platelet-rich thrombi was induced with electrical stimulation in carotid arteries in rabbits. Clot lysis and recanalization were induced subsequently with intravenous tissue-type plasminogen activator (t-PA) and verified with Doppler flow probes. Plasma PAI-1 activity (mean +/- SD) increased from 6 +/- 2 arbitrary units (AU)/ml to 129 +/- 48 AU/ml (n = 15) within several hours after recanalization. When t-PA had failed to induce recanalization, the increase was much less (from 7 +/- 2 to 42 +/- 23 AU/ml, n = 11). To define mechanisms responsible for these changes, PAI-1 messenger RNA (mRNA) was evaluated by Northern blot analysis and localized in tissues by in situ hybridization. Strong and consistent induction of PAI-1 mRNA was evident in aorta, heart, and liver of animals subjected to thrombosis (twofold to threefold increases compared with values in controls), particularly in those in which thrombolysis had been induced (fourfold to sixfold). After thrombolysis, an intense, PAI-1 mRNA-specific signal was detected in endothelium of aorta, liver, and heart, with less intense signals in endothelium of lung, adrenals, and kidneys. CONCLUSIONS: The increases in plasma PAI-1 activity follow a preceding increase in endothelial cell expression of the PAI-1 gene as reflected by PAI-1 mRNA levels. Thus, increased synthesis of endothelial cell PAI-1 after thrombosis and thrombolysis may attenuate endogenous fibrinolysis early after coronary thrombolysis, thereby potentiating early, thrombotic reocclusion.

Induction of synthesis of plasminogen activator inhibitor type-1 by tissue-type plasminogen activator in human hepatic and endothelial cells.

Plasminogen activator inhibitor type-1 (PAI-1) can modify fibrinolytic activity in vitro and in vivo. The present study was performed to determine whether pharmacologic concentrations of tissue-type plasminogen activator (t-PA) can initiate negative feedback by stimulating PAI-1 synthesis. In both human hepatoma cells (Hep G2) and human umbilical vein endothelial cells (HUVEC), t-PA increased the total concentrations and appearance of newly synthesized protein in conditioned media of free PAI-1 and PAI-1 complexed with t-PA in a dose and time dependent fashion judging from results after immunoprecipitation of metabolically labeled PAI-1. The t-PA effect was not attributable simply to release of stored or matrix-bound PAI-1. In HUVEC, Northern blot analyses indicated that t-PA increased steady-state levels of PAI-1 mRNA two-fold. In contrast PAI-1 mRNA expression was not increased in Hep G2 cells. Thus, mechanisms of stimulation appeared to differ in the two cell lines. The results obtained are consistent with the hypothesis that increased PAI-1 synthesis and secretion in response to t-PA may limit or attenuate fibrinolysis locally or systemically in vivo.

Dependence of fibrinolytic activity on the concentration of free rather than total tissue-type plasminogen activator in plasma after pharmacologic administration.

To identify factors responsible for the decline of plasma tissue-type plasminogen activator (t-PA)-specific activity that we have observed after infusions of the activator and to define the potential usefulness of selected variants of t-PA in obviating them in patients with infarction, serial plasma samples from patients (n = 4) and rabbits (n = 15) given t-PA were assayed for total t-PA antigen, t-PA activity, and free as opposed to type-1 plasminogen activator inhibitor (PAI-1)--complexed t-PA. In patients, attenuation of t-PA specific activity after infusions was evident with concentrations of total t-PA antigen that were as much as sevenfold greater than pretreatment values (62 compared with 9 ng/ml). Attenuation of t-PA activity corresponded with the disappearance of free t-PA from plasma and was associated with persistence of complexes of t-PA with PAI-1. In normal rabbits (n = 4) given wild-type t-PA by bolus injection, PAI-1 activity was 4 +/- 1 arbitrary units/ml. Attenuation of t-PA activity was not evident until 60 minutes after injection at a time when total plasma t-PA antigen concentration was as low as 13 +/- 8 ng/ml. Under these conditions, plasma t-PA was composed predominantly of free t-PA. In rabbits (n = 5) given lipopolysaccharide to increase plasma PAI-1 activity to 193 +/- 84 arbitrary units/ml, the specific activity of t-PA was attenuated as early as 15 minutes after injection at a time when total t-PA antigen concentration was as high as 164 +/- 79 ng/ml. As was the case with samples from patients, attenuation was associated with the disappearance of free t-PA and the persistence of complexes of t-PA with PAI-1. A genetically engineered variant of t-PA with comparable specific activity and a comparable rate constant of association with PAI-1 but designed to persist in the circulation manifested prolonged clearance from plasma of normal rabbits (n = 3) (t1/2 = 24.6 +/- 1.6 minutes compared with an alpha phase t1/2 of 1.9 minutes for wild-type t-PA). The variant lacked the epidermal growth factor and kringle one domains and contained a duplicated kringle two domain.(ABSTRACT TRUNCATED AT 400 WORDS)

Potential attenuation of fibrinolysis by growth factors released from platelets and their pharmacologic implications.

Increased concentrations of the fast-acting tissue-type plasminogen activator (t-PA) inhibitor attenuate the fibrinolytic activity of pharmacologically administered activators of the fibrinolytic system such as t-PA. Accordingly, it was hypothesized that augmentation of synthesis and elaboration of inhibitor from the liver, leading to increased concentrations of inhibitor in plasma, or from endothelial cells in the vicinity of thrombi undergoing lysis, leading to increased concentrations locally, may contribute to failure of pharmacologically induced thrombolysis or to early reocclusion. Because platelets are rich in transforming growth factor beta and epidermal growth factor-like activity, it was thought that release of growth factors from platelets activated in vivo could mediate increases of the inhibitor in plasma by stimulating its formation in the liver and its local release from endothelial cells in the vicinity of thrombi. If so, fibrinolysis might be rendered more effective by concomitant prevention of platelet growth factor release. Transforming growth factor beta, a major constituent of platelets, increased concentrations of the t-PA inhibitor messenger ribonucleic acid (mRNA) in human hepatoma cells in a specific and dose-dependent manner. A peak effect was seen with 5 ng/ml and a 10-fold increase in 6 hours. Release of inhibitor protein into conditioned media increased as well. Induction of the inhibitor mRNA increase was elicited by exposure as brief as 30 minutes. Cycloheximide, an inhibitor of protein synthesis, was not inhibitory. The mechanisms responsible differed from those seen with epidermal growth factor, shown previously in the laboratory to increase inhibitor mRNA. In addition, the 2 factors were synergistic. Platelet lysates elicited effects simulating those of the purified growth factors.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of interaction of active-site serine mutants of tissue-type plasminogen activator with plasminogen activator inhibitor-1.

To define determinants of interactions of tissue-type plasminogen activator (t-PA) with plasminogen activator inhibitor type-1 (PAI-1), we utilized site-directed mutagenesis to substitute either threonine or glycine for the active-site serine of tissue-type plasminogen activator. Assays of conditioned media of transfected cells demonstrated that the threonine substitution markedly decreased but did not entirely abolish plasminogen activating activity. In contrast, the glycine substitution yielded a mutant with absolutely no detectable plasminogen activating activity. Wild-type t-PA formed stable complexes with PAI-1. However, even when exogenous inhibitor was present in the medium or purified mutant was added to plasma that had been rendered PAI-1-rich in vivo, the mutants were present in the free form exclusively judging from results of fibrin autography and Western blot analysis. Thus, despite maintenance of some residual plasminogen-activating activity associated with preservation of the hydroxyl group at the active site, the threonine mutant did not form stable complexes with inhibitor. The glycine mutant, developed so that steric hindrance or other unfavorable interactions at the modified active site would be minimal, was similarly incapable of forming complexes with PAI-1. These results show that the presence of an active site serine residue is necessary for formation of stable complexes between t-PA and PAI-1.

Regulation of the expression of type 1 plasminogen activator inhibitor in Hep G2 cells by epidermal growth factor.

To identify factors potentially influencing expression of type 1 plasminogen activator inhibitor (PAI-1), we characterized the human tissue-specific distribution of PAI-1 mRNA and the influence of epidermal growth factor (EGF) on expression of steady state levels of PAI-1 mRNA and secretion of PAI-1 by Hep G2 cells. Two species of PAI-1 mRNA (3.2 and 2.2 kilobases) were detected, and the ratio of the two varied among tissues (3 to 5:1) in contrast to the 1:1 ratio detected in Hep G2 cells. Expression of PAI-1 mRNA was inversely related to the distribution of tissue-type plasminogen activator mRNA (2.3 kilobases). Nu-Serum, a growth media supplement, increased steady state levels of PAI-1 mRNA 5-fold within 3 h. Factors responsible were found to be trypsin-sensitive and dialysis-resistant. Antisera to EGF attenuated Nu-Serum-induced increases of PAI-1 mRNA by 57%, suggesting that EGF or EGF homologous peptides contributed to the response. EGF elicited increases of PAI-1 mRNA levels in a dose-dependent manner. Induction was rapid (7-fold at 3 h with 5 ng/ml) and complete within 10 h. The response was not attenuated by cycloheximide (25 micrograms/ml). Factor X and glyceraldehyde-3-phosphate dehydrogenase mRNA did not increase. Increased levels of PAI-1 antigen were detected in conditioned media of Hep G2 cells by 4 h and were maximal at 8 h (6-fold). We conclude that the expression of PAI-1 mRNA is tissue-specific and regulated by epidermal growth factor in Hep G2 cells.

Biochemical determinants of clearance of tissue-type plasminogen activator from the circulation.

Biochemical modification of tissue-type plasminogen activator (t-PA) designed to alter pharmacokinetics and pharmacodynamics offers promise for development of pharmaceuticals particularly suitable for treatment of specific disorders and for induction of coronary thrombolysis by intramuscular as well as intravenous administration. Accordingly, to identify biochemical determinants of clearance of t-PA from the circulation, we injected rabbits intravenously with three different preparations of t-PA synthesized from the same human gene and expressed in Chinese hamster ovary cells cultured under disparate conditions. Influences of glycosylation on clearance were defined by experiments with enzymatically treated t-PA in which clearance was assessed with concomitant administration of selected neoglycoproteins that compete with t-PA for specific glycoprotein receptors. The role of an intact active catalytic site, as reflected by differences in clearance with and without prior treatment of t-PA with the protease inhibitor PPACK, was defined also. Results indicate that clearance is altered by inhibition of the active site and that the nature and extent of glycosylation--not evident simply by analysis of peptide structure--influence clearance as well. These findings suggest that mannose/N-acetylglucosamine-specific glycoprotein receptors expressed on hepatic reticuloendothelial cells participate in clearance of t-PA from the circulation but that galactose-specific glycoprotein receptors probably do not. The observations may explain differences in clearance seen with different preparations of t-PA that have been seen in clinical pilot studies and may identify biochemical determinants of clearance amenable to modification for development of agents with potentially desirable, specific biological properties.

Interactions of tissue-type plasminogen activator with plasma inhibitors and their pharmacologic implications.

To delineate interactions of infused tissue-type plasminogen activator (t-PA) with inhibitors in plasma and their impact on fibrinolytic activity, serial plasma samples from patients with acute myocardial infarction and from normal rabbits given infusions of t-PA were assayed for t-PA antigen, activity of "fast acting" plasminogen activator inhibitor (PAI-1), and the presence and nature of t-PA-inhibitor complexes. In patients, endogenous t-PA circulated predominantly as a 100 kilodalton (kDa) complex with PAI-1, as verified by immunoprecipitation. During infusions, t-PA circulated not only as free t-PA (55 kDa) but also in complexes with PAI-1 (100 kDa), alpha 2-antiplasmin (110 kDa), and C1-esterase inhibitor (170 kDa). After termination of infusions, levels of free t-PA declined, while inhibitor complexes remained prominent. Free PAI-1 activity, assayed spectrophotometrically, was markedly elevated in the 24 hr interval after infusion of t-PA in 47% of patients with infarction. The specific activity of t-PA during infusions was 0.4 IU/ng or greater. However, during the 3 hr interval after infusions in patients, specific activity declined in association with prominence of t-PA complexes, predominantly with PAI-1. Infusions of t-PA in normal rabbits did not result in reactive increases in PAI-1 activity or in the t-PA-PAI-1 complex. After infusions, t-PA was associated predominantly with alpha 2-antiplasmin and C1-esterase inhibitor rather than PAI-1. t-PA inhibitor complexes were seen despite immediate acidification of whole blood, indicating that they were present in vivo.(ABSTRACT TRUNCATED AT 250 WORDS)

Experimental anaerobic liver abscess: observations on treatment.

Various treatment modalities for solitary anaerobic liver abscesses were evaluated in a recently-described rabbit model. In the first phase of the experiment, 35 rabbits with liver abscesses induced with Bacteroides fragilis and Fusobacterium necrophorum were randomized into four groups: surgical drainage alone, drainage plus clindamycin 150 mg IM 8-hourly, clindamycin alone, and untreated controls. Serum clindamycin concentrations in rabbits were similar to those achieved in humans. The survival of rabbits receiving antibiotic chemotherapy alone was significantly better than controls, whereas the survival of those having surgical drainage with or without chemotherapy was not. However, successful surgical drainage was followed by weight gain in surviving rabbits. In the second phase of the experiment 18 rabbits with abscesses were randomized into the same groups. Aspirates of pus from all rabbits receiving clindamycin were sterile by day 7 of treatment, but high bacterial counts were still present in the abscess cavities of control rabbits and of those undergoing drainage alone. These findings illustrate the application of a new model for pyogenic liver abscess in laboratory investigation. Their relevance to management of human pyogenic liver abscesses remains to be assessed.

An experimental model for pyogenic liver abscess.

We have developed a reproducible small-animal model for pyogenic liver abscess, suitable for investigating diagnostic and therapeutic modalities. Male New Zealand white rabbits weighing 2-3 kg were anaesthetized and the liver exposed. Gentle pressure was applied with forceps to the right hepatic lobe. A suspension of 10(5) colony forming units (cfu) Escherichia coli plus Fusobacterium necrophorum (10(6) cfu) plus Bacteroides fragilis (10(6) cfu) was immediately injected into a mesenteric vein. Two weeks later a palpable mass (mean diameter 4 cm) had developed. Thick pus could be aspirated percutaneously. Necropsy revealed a single, but often multiloculated, abscess at the site of the previous trauma. Injection of E. coli alone did not produce any abscesses and B. fragilis alone only small abscesses, with low and variable frequency. Inoculation with F. necrophorum alone produced large abscesses, and a dose-response relationship was established. This is a simple and reliable small-animal model useful in studies of imaging techniques, antibiotic regimens and invasive treatments for pyogenic liver abscess.