Discovery of Potent and Selective Periphery-Restricted Quinazoline Inhibitors of the Cyclic Nucleotide Phosphodiesterase PDE1.

We disclose the discovery and X-ray cocrystal data of potent, selective quinazoline inhibitors of PDE1. Inhibitor ( S)-3 readily attains free plasma concentrations above PDE1 IC50 values and has restricted brain access. The racemic compound 3 inhibits >75% of PDE hydrolytic activity in soluble samples of human myocardium, consistent with heightened PDE1 activity in this tissue. These compounds represent promising new tools to probe the value of PDE1 inhibition in the treatment of cardiovascular disease.

Cerebrospinal fluid ß-Amyloid turnover in the mouse, dog, monkey and human evaluated by systematic quantitative analyses.

BACKGROUND: Reducing brain ß-amyloid (Aß) via inhibition of ß-secretase, or inhibition/modulation of γ-secretase, has been widely pursued as a potential disease-modifying treatment for Alzheimer's disease. Compounds that act through these mechanisms have been screened and characterized with Aß lowering in the brain and/or cerebrospinal fluid (CSF) as the primary pharmacological end point. Interpretation and translation of the pharmacokinetic (PK)/pharmacodynamic (PD) relationship for these compounds is complicated by the relatively slow Aß turnover process in these compartments. OBJECTIVE: To understand Aß turnover kinetics in preclinical species and humans. METHODS: We collected CSF Aß dynamic data after ß- or γ-secretase inhibitor treatment from in-house experiments and the public domain, and analyzed the data using PK/PD modeling to obtain CSF Aß turnover rates (kout) in the mouse, dog, monkey and human. RESULTS: The kout for CSF Aß40 follows allometry (kout = 0.395 × body weight(-0.351)). The kout for CSF Aß40 is approximately 2-fold higher than the turnover of CSF in rodents, but in higher species, the two are comparable. CONCLUSION: The turnover of CSF Aß40 was systematically examined, for the first time, in multiple species through quantitative modeling of multiple data sets. Our result suggests that the clearance mechanisms for CSF Aß in rodents may be different from those in the higher species. The understanding of Aß turnover has considerable implications for the discovery and development of Aß-lowering therapeutics, as illustrated from the perspectives of preclinical PK/PD characterization and preclinical-to-clinical translation.

Application of structure-based drug design and parallel chemistry to identify selective, brain penetrant, in vivo active phosphodiesterase 9A inhibitors.

Phosphodiesterase 9A inhibitors have shown activity in preclinical models of cognition with potential application as novel therapies for treating Alzheimer's disease. Our clinical candidate, PF-04447943 (2), demonstrated acceptable CNS permeability in rats with modest asymmetry between central and peripheral compartments (free brain/free plasma = 0.32; CSF/free plasma = 0.19) yet had physicochemical properties outside the range associated with traditional CNS drugs. To address the potential risk of restricted CNS penetration with 2 in human clinical trials, we sought to identify a preclinical candidate with no asymmetry in rat brain penetration and that could advance into development. Merging the medicinal chemistry strategies of structure-based design with parallel chemistry, a novel series of PDE9A inhibitors was identified that showed improved selectivity over PDE1C. Optimization afforded preclinical candidate 19 that demonstrated free brain/free plasma ≥ 1 in rat and reduced microsomal clearance along with the ability to increase cyclic guanosine monophosphosphate levels in rat CSF.

Cerebrospinal fluid amyloid-ß (Aß) as an effect biomarker for brain Aß lowering verified by quantitative preclinical analyses.

Reducing the generation of amyloid-ß (Aß) in the brain via inhibition of ß-secretase or inhibition/modulation of γ-secretase has been pursued as a potential disease-modifying treatment for Alzheimer's disease. For the discovery and development of ß-secretase inhibitors (BACEi), γ-secretase inhibitors (GSI), and γ-secretase modulators (GSM), Aß in cerebrospinal fluid (CSF) has been presumed to be an effect biomarker for Aß lowering in the brain. However, this presumption is challenged by the lack of quantitative understanding of the relationship between brain and CSF Aß lowering. In this study, we strived to elucidate how the intrinsic pharmacokinetic (PK)/pharmacodynamic (PD) relationship for CSF Aß lowering is related to that for brain Aß through quantitative modeling of preclinical data for numerous BACEi, GSI, and GSM across multiple species. Our results indicate that the intrinsic PK/PD relationship in CSF is predictive of that in brain, at least in the postulated pharmacologically relevant range, with excellent consistency across mechanisms and species. As such, the validity of CSF Aß as an effect biomarker for brain Aß lowering is confirmed preclinically. Meanwhile, we have been able to reproduce the dose-dependent separation between brain and CSF effect profiles using simulations. We further discuss the implications of our findings to drug discovery and development with regard to preclinical PK/PD characterization and clinical prediction of Aß lowering in the brain.

Application of the bicyclo[1.1.1]pentane motif as a nonclassical phenyl ring bioisostere in the design of a potent and orally active γ-secretase inhibitor.

Replacement of the central, para-substituted fluorophenyl ring in the γ-secretase inhibitor 1 (BMS-708,163) with the bicyclo[1.1.1]pentane motif led to the discovery of compound 3, an equipotent enzyme inhibitor with significant improvements in passive permeability and aqueous solubility. The modified biopharmaceutical properties of 3 translated into excellent oral absorption characteristics (~4-fold ↑ C(max) and AUC values relative to 1) in a mouse model of γ-secretase inhibition. In addition, SAR studies into other fluorophenyl replacements indicate the intrinsic advantages of the bicyclo[1.1.1]pentane moiety over conventional phenyl ring replacements with respect to achieving an optimal balance of properties (e.g., γ-secretase inhibition, aqueous solubility/permeability, in vitro metabolic stability). Overall, this work enhances the scope of the [1.1.1]-bicycle beyond that of a mere "spacer" unit and presents a compelling case for its broader application as a phenyl group replacement in scenarios where the aromatic ring count impacts physicochemical parameters and overall drug-likeness.

Metabolism-directed design of oxetane-containing arylsulfonamide derivatives as γ-secretase inhibitors.

A metabolism-based approach toward the optimization of a series of N-arylsulfonamide-based γ-secretase inhibitors is reported. The lead cyclohexyl analogue 6 suffered from extensive oxidation on the cycloalkyl motif by cytochrome P450 3A4, translating into poor human liver microsomal stability. Knowledge of the metabolic pathways of 6 triggered a structure-activity relationship study aimed at lowering lipophilicity through the introduction of polarity. This effort led to several tetrahydropyran and tetrahydrofuran analogues, wherein the 3- and 4-substituted variants exhibited greater microsomal stability relative to their 2-substituted counterparts. Further reduction in lipophilicity led to the potent γ-secretase inhibitor and 3-substituted oxetane 1 with a reduced propensity toward oxidative metabolism, relative to its 2-substituted isomer. The slower rates of metabolism with 3-substituted cyclic ethers most likely originate from reductions in lipophilicity and/or unfavorable CYP active site interactions with the heteroatom. Preliminary animal pharmacology studies with a representative oxetane indicate that the series is generally capable of lowering Aß in vivo. As such, the study also illustrates the improvement in druglikeness of molecules through the use of the oxetane motif.

Quantitative pharmacokinetic/pharmacodynamic analyses suggest that the 129/SVE mouse is a suitable preclinical pharmacology model for identifying small-molecule γ-secretase inhibitors.

Alzheimer's disease (AD) poses a serious public health threat to the United States. Disease-modifying drugs slowing AD progression are in urgent need, but they are still unavailable. According to the amyloid cascade hypothesis, inhibition of ß- or γ-secretase, key enzymes for the production of amyloid ß (Aß), may be viable mechanisms for the treatment of AD. For the discovery of γ-secretase inhibitors (GSIs), the APP-overexpressing Tg2576 mouse has been the preclinical model of choice, in part because of the ease of detection of Aß species in its brain, plasma, and cerebrospinal fluid (CSF). Some biological observations and practical considerations, however, argue against the use of the Tg2576 mouse. We reasoned that an animal model would be suitable for GSI discovery if the pharmacokinetic (PK)/pharmacodynamic (PD) relationship of a compound for Aß lowering in this model is predictive of that in human. In this study, we assessed whether the background 129/SVE strain is a suitable preclinical pharmacology model for identifying new GSIs by evaluating the translatability of the intrinsic PK/PD relationships for brain and CSF Aß across the Tg2576 and 129/SVE mouse and human. Using semimechanistically based PK/PD modeling, our analyses indicated that the intrinsic PK/PD relationship for brain Aßx-42 and CSF Aßx-40 in the 129/SVE mouse is indicative of that for human CSF Aß. This result, in conjunction with practical considerations, strongly suggests that the 129/SVE mouse is a suitable model for GSI discovery. Concurrently, the necessity and utilities of PK/PD modeling for rational interpretation of Aß data are established.

Modulation of NMDA receptor function by inhibition of D-amino acid oxidase in rodent brain.

Observations that N-Methyl-D-Aspartate (NMDA) antagonists produce symptoms in humans that are similar to those seen in schizophrenia have led to the current hypothesis that schizophrenia might result from NMDA receptor hypofunction. Inhibition of D-amino acid oxidase (DAAO), the enzyme responsible for degradation of D-serine, should lead to increased levels of this co-agonist at the NMDA receptor, and thereby provide a therapeutic approach to schizophrenia. We have profiled some of the preclinical biochemical, electrophysiological, and behavioral consequences of administering potent and selective inhibitors of DAAO to rodents to begin to test this hypothesis. Inhibition of DAAO activity resulted in a significant dose and time dependent increase in D-serine only in the cerebellum, although a time delay was observed between peak plasma or brain drug concentration and cerebellum D-serine response. Pharmacokinetic/pharmacodynamic (PK/PD) modeling employing a mechanism-based indirect response model was used to characterize the correlation between free brain drug concentration and D-serine accumulation. DAAO inhibitors had little or no activity in rodent models considered predictive for antipsychotic activity. The inhibitors did, however, affect cortical activity in the Mescaline-Induced Scratching model, produced a modest but significant increase in NMDA receptor-mediated synaptic currents in primary neuronal cultures from rat hippocampus, and resulted in a significant increase in evoked hippocampal theta rhythm, an in vivo electrophysiological model of hippocampal activity. These findings demonstrate that although DAAO inhibition did not cause a measurable increase in D-serine in forebrain, it did affect hippocampal and cortical activity, possibly through augmentation of NMDA receptor-mediated currents.

Diamide amino-imidazoles: a novel series of γ-secretase inhibitors for the treatment of Alzheimer's disease.

The synthesis and structure-activity relationship (SAR) of a novel series of di-substituted imidazoles, derived from modification of DAPT, are described. Subsequent optimization led to identification of a highly potent series of inhibitors that contain a ß-amine in the imidazole side-chain resulting in a robust in vivo reduction of plasma and brain Aß in guinea pigs. The therapeutic index between Aß reductions and changes in B-cell populations were studied for compound 10 h.

Design, synthesis, and in vivo characterization of a novel series of tetralin amino imidazoles as γ-secretase inhibitors: discovery of PF-3084014.

A novel series of tetralin containing amino imidazoles, derived from modification of the corresponding phenyl acetic acid derivatives is described. Replacement of the amide led to identification of a potent series of tetralin-amino imidazoles with robust central efficacy. The reduction of brain Aß in guinea pigs in the absence of changes in B-cells suggested a potential therapeutic index with respect to APP processing compared with biomarkers of notch related toxicity. Optimization of the FTOC to plasma concentrations at the brain Aß EC(50) lead to the identification of compound 14f (PF-3084014) which was selected for clinical development.

Pharmacodynamics and pharmacokinetics of the gamma-secretase inhibitor PF-3084014.

PF-3084014 [(S)-2-((S)-5,7-difluoro-1,2,3,4-tetrahydronaphthalen-3-ylamino)-N-(1-(2-methyl-1-(neopentylamino)propan-2-yl)-1H-imidazol-4-yl)pentanamide] is a novel gamma-secretase inhibitor that reduces amyloid-beta (Abeta) production with an in vitro IC(50) of 1.2 nM (whole-cell assay) to 6.2 nM (cell-free assay). This compound inhibits Notch-related T- and B-cell maturation in an in vitro thymocyte assay with an EC(50) of 2.1 microM. A single acute dose showed dose-dependent reduction in brain, cerebrospinal fluid (CSF), and plasma Abeta in Tg2576 mice as measured by enzyme-linked immunosorbent assay and immunoprecipitation (IP)/mass spectrometry (MS). Guinea pigs were dosed with PF-3084014 for 5 days via osmotic minipump at 0.03 to 3 mg/kg/day and exhibited dose-dependent reduction in brain, CSF, and plasma Abeta. To further characterize Abeta dynamics in brain, CSF, and plasma in relation to drug exposure and Notch-related toxicities, guinea pigs were dosed with 0.03 to 10 mg/kg PF-3084014, and tissues were collected at regular intervals from 0.75 to 30 h after dose. Brain, CSF, and plasma all exhibited dose-dependent reductions in Abeta, and the magnitude and duration of Abeta lowering exceeded those of the reductions in B-cell endpoints. Other gamma-secretase inhibitors have shown high potency at elevating Abeta in the conditioned media of whole cells and the plasma of multiple animal models and humans. Such potentiation was not observed with PF-3084014. IP/MS analysis, however, revealed dose-dependent increases in Abeta11-40 and Abeta1-43 at doses that potently inhibited Abeta1-40 and Abeta1-42. PF-3084014, like previously described gamma-secretase inhibitors, preferentially reduced Abeta1-40 relative to Abeta1-42. Potency at Abeta relative to Notch-related endpoints in vitro and in vivo suggests that a therapeutic index can be achieved with this compound.

Discovery, SAR, and pharmacokinetics of a novel 3-hydroxyquinolin-2(1H)-one series of potent D-amino acid oxidase (DAAO) inhibitors.

3-Hydroxyquinolin-2(1H)-one (2) was discovered by high throughput screening in a functional assay to be a potent inhibitor of human DAAO, and its binding affinity was confirmed in a Biacore assay. Cocrystallization of 2 with the human DAAO enzyme defined the binding site and guided the design of new analogues. The SAR, pharmacokinetics, brain exposure, and effects on cerebellum D-serine are described. Subsequent evaluation against the rat DAAO enzyme revealed a divergent SAR versus the human enzyme and may explain the high exposures of drug necessary to achieve significant changes in rat or mouse cerebellum D-serine.

Evaluation of cerebrospinal fluid concentration and plasma free concentration as a surrogate measurement for brain free concentration.

This study was designed to evaluate the use of cerebrospinal fluid (CSF) drug concentration and plasma unbound concentration (C(u,plasma)) to predict brain unbound concentration (C(u,brain)). The concentration-time profiles in CSF, plasma, and brain of seven model compounds were determined after subcutaneous administration in rats. The C(u,brain) was estimated from the product of total brain concentrations and unbound fractions, which were determined using brain tissue slice and brain homogenate methods. For theobromine, theophylline, caffeine, fluoxetine, and propranolol, which represent rapid brain penetration compounds with a simple diffusion mechanism, the ratios of the area under the curve of C(u,brain)/C(CSF) and C(u,brain)/C(u,plasma) were 0.27 to 1.5 and 0.29 to 2.1, respectively, using the brain slice method, and were 0.27 to 2.9 and 0.36 to 3.9, respectively, using the brain homogenate method. A P-glycoprotein substrate, CP-141938 (methoxy-3-[(2-phenyl-piperadinyl-3-amino)-methyl]-phenyl-N-methyl-methane-sulfonamide), had C(u,brain)/C(CSF) and C(u,brain)/C(u,plasma) ratios of 0.57 and 0.066, using the brain slice method, and 1.1 and 0.13, using the brain homogenate method, respectively. The slow brain-penetrating compound, N[3-(4'-fluorophenyl)-3-(4'-phenylphenoxy)propyl-]sarcosine, had C(u,brain)/C(CSF) and C(u,brain)/C(u,plasma) ratios of 0.94 and 0.12 using the brain slice method and 0.15 and 0.018 using the brain homogenate method, respectively. Therefore, for quick brain penetration with simple diffusion mechanism compounds, C(CSF) and C(u,plasma) represent C(u,brain) equally well; for efflux substrates or slow brain penetration compounds, C(CSF) appears to be equivalent to or more accurate than C(u,plasma) to represent C(u,brain). Thus, we hypothesize that C(CSF) is equivalent to or better than C(u,plasma) to predict C(u,brain). This hypothesis is supported by the literature data.

Evaluation of the utility of brain slice methods to study brain penetration.

The objective of this study was to evaluate the utility of brain tissue slices to determine the effect of plasma and brain tissue nonspecific binding on the brain-to-plasma ratio (K(p)). Mouse or rat brain slices (400 microm) were prepared using a McIlwain tissue chopper (Surrey, UK) and incubated with 1 microg/ml of compound at 37 degrees C either in a physiological buffer to determine the buffer-to-slice concentration ratio, i.e., unbound fraction in brain tissue (f(u,slice)), or in plasma to determine the slice-to-plasma concentration ratio (C(slice)/C(plasma)). The unbound fraction in plasma, f(u,plasma), was determined using equilibrium dialysis. In vitro-in vivo correlation of the brain-to-plasma ratio was examined for 13 and eight model compounds in mice and rats, respectively. C(slice)/C(plasma) and f(u,plasma)/f(u,slice) predicted the K(p) in rats, and C(slice)/C(plasma) predicted the K(p) in FVB mice for non-P-glycoprotein substrates within 3-fold but overpredicted K(p) for P-glycoprotein substrates by more than 3-fold. However, C(slice)/C(plasma) predicted the K(p) in mdr1a/1b knockout mice for both non-P-glycoprotein and P-glycoprotein substrates. Our present study demonstrates that a brain slice method can be used to differentiate whether a compound having a low K(p) is due to the effect of low nonspecific binding to brain tissue relative to plasma proteins or because of efflux transport at the blood-brain barrier.

Use of a physiologically based pharmacokinetic model to study the time to reach brain equilibrium: an experimental analysis of the role of blood-brain barrier permeability, plasma protein binding, and brain tissue binding.

This study was designed 1) to examine the effects of blood-brain barrier (BBB) permeability [quantified as permeability-surface area product (PS)], unbound fraction in plasma (f(u,plasma)), and brain tissue (f(u,brain)) on the time to reach equilibrium between brain and plasma and 2) to investigate the drug discovery strategies to design and select compounds that can rapidly penetrate the BBB and distribute to the site of action. The pharmacokinetics of seven model compounds: caffeine, CP-141938 [methoxy-3-[(2-phenyl-piperadinyl-3-amino)-methyl]-phenyl-N-methyl-methane-sulfonamide], fluoxetine, NFPS [N[3-(4'-fluorophenyl)-3-(4'-phenylphenoxy)propyl]sarcosine], propranolol, theobromine, and theophylline in rat brain and plasma after subcutaneous administration were studied. The in vivo log PS and log f(u,brain) calculated using a physiologically based pharmacokinetic model correlates with in situ log PS (R(2) = 0.83) and in vitro log f(u,brain) (R(2) = 0.69), where the in situ PS and in vitro f(u,brain) was determined using in situ brain perfusion and equilibrium dialysis using brain homogenate, respectively. The time to achieve brain equilibrium can be quantitated with a proposed parameter, intrinsic brain equilibrium half-life [t(1/2eq,in) = V(b)ln2/(PS . f(u,brain))], where V(b) is the physiological volume of brain. The in vivo log t(1/2eq,in) does not correlate with in situ log PS (R(2) < 0.01) but correlates inversely with log(PS . f(u,brain)) (R(2) = 0.85). The present study demonstrates that rapid brain equilibration requires a combination of high BBB permeability and low brain tissue binding. A high BBB permeability alone cannot guarantee a rapid equilibration. The strategy to select compounds with rapid brain equilibration in drug discovery should identify compounds with high BBB permeability and low nonspecific binding in brain tissue.

The impact of P-glycoprotein on the disposition of drugs targeted for indications of the central nervous system: evaluation using the MDR1A/1B knockout mouse model.

Thirty-two structurally diverse drugs used for the treatment of various conditions of the central nervous system (CNS), along with two active metabolites, and eight non-CNS drugs were measured in brain, plasma, and cerebrospinal fluid in the P-glycoprotein (P-gp) knockout mouse model after subcutaneous administration, and the data were compared with corresponding data obtained in wild-type mice. Total brain-to-plasma (B/P) ratios for the CNS agents ranged from 0.060 to 24. Of the 34 CNS-active agents, only 7 demonstrated B/P area under the plasma concentration curve ratios between P-gp knockout and wild-type mice that did not differ significantly from unity. Most of the remaining drugs demonstrated 1.1- to 2.6-fold greater B/P ratios in P-gp knockout mice versus wild-type mice. Three, risperidone, its active metabolite 9-hydroxyrisperidone, and metoclopramide, showed marked differences in B/P ratios between knockout and wild-type mice (6.6- to 17-fold). Differences in B/P ratios and cerebrospinal fluid/plasma ratios between wild-type and knockout animals were correlated. Through the use of this model, it appears that most CNS-active agents demonstrate at least some P-gp-mediated transport that can affect brain concentrations. However, the impact for the majority of agents is probably minor. The example of risperidone illustrates that even good P-gp substrates can still be clinically useful CNS-active agents. However, for such agents, unbound plasma concentrations may need to be greater than values projected using receptor affinity data to achieve adequate receptor occupancy for effect.

Endoleak following endovascular abdominal aortic aneurysm repair: implications for duration of screening.

OBJECTIVE: Endovascular abdominal aortic aneurysm repair (EAR) requires long-term surveillance for endoleak or increase in aneurysm diameter. We analyzed the natural history of and risk factors for endoleak development. SUMMARY BACKGROUND DATA: Endoleak is a common complication of EAR that can lead to aneurysm enlargement and even rupture. Following EAR, imaging studies are used to identify leaks since patients with endoleak may require additional endovascular interventions or conversion to open repair. No criteria currently exist for cessation or reduction in frequency of screening imaging studies. METHODS: Data on 220 patients undergoing EAR were retrospectively reviewed. Kaplan-Meier survival analysis and Cox proportional hazards regression were used with the end point being new endoleak development. Potential risk factors included preoperative aneurysm diameter, number of negative surveillance studies, and postoperative increase in diameter. RESULTS: A total of 52 patients (24%) who underwent EAR had endoleak detected during postoperative follow-up, which averaged 19 months (range, 0.4-101 months). One, 6-, 12-, and 24- month endoleak-free survival was 90%, 80%, 77%, and 73%, respectively. Three leaks occurred after year 2, at postoperative months 24, 48, and 85. Increasing number of negative screening studies was negatively associated with risk for endoleak development (B = -3.122, P < 0.001), while increase in aneurysm diameter was positively associated with risk for endoleak (B = 0.072, P = 0.04). CONCLUSION: Risk for endoleak declines as the number of negative postoperative scans increases, but new endoleaks are identified as late as 7 years following EAR. Reduction in screening frequency cannot be uniformly recommended at this time. Patients with documented aneurysm expansion should be monitored carefully and endoleak should be suspected.

Technical success from endovascular aneurysm repair in the post-marketing era: a multicenter prospective trial.

Evaluation of post-marketing success with the Ancure Endovascular Graft (AEG) was accomplished by review of a multicenter, prospective trial involving 46 centers and 163 patients. A second cohort of patients (n = 350) treated with the AEG under a controlled-use interval prior to the prospective trial was simultaneously evaluated. Technical success in both groups of patients (96.9% and 97.4%, respectively) was similar to what was reported in pre-market clinical trials. Operative implantation complications unique to the AEG included graft limb stenosis/occlusion in 35.6 and 31.4%, contralateral pull wire being caught on hooks in 33.7 and 28%, failure to seal (type I endoleak) in 17.2 and 18.3%, jacket guard being stuck in 12.9 and 11%, contralateral wire being stuck in 6.8 and 7.1%, high jacket retraction force in 16 and 8.5%, and inability to retract jacket in 1.8 and 0.5% of patients involved in the multicenter trial and controlled-use interval, respectively. One of four patients undergoing conversion in the prospective trial had graft misdeployment as a mode of failure. Three were converted for access failure. The 30-day mortality rate in the prospective trial was 3.7%. Interventions to resolve implantation-related events included stenting, guide catheter manipulations, wire exchanges, and delivery catheter disassembly. These interventions were successful in virtually every case. Open surgical procedures were not needed to correct these operative problems. Results from this study demonstrate excellent technical success with the AEG in the post-market era. Interventions to resolve implantation complications, when utilized, are highly successful in facilitating AEG implantation and providing technical success.

Duplex scan-directed placement of inferior vena cava filters: a five-year institutional experience.

PURPOSE: The purpose of this study was the assessment of the safety, efficacy, and hospital charges of bedside duplex ultrasound-directed inferior vena cava (IVC) filter placement. METHODS: All duplex ultrasound-directed IVC filters that were placed from August 8, 1995, to December 31, 2000, are reviewed. Chart review combined with mailed questionnaires and telephone follow-up examinations were used to collect demographic and outcome data. RESULTS: Three hundred twenty-five patients underwent evaluation, and 284 underwent duplex ultrasound-directed IVC filter placement. Two hundred three (71%) were male patients, and 81 (29%) were female patients. Poor IVC visualization, IVC thrombosis, and unsuitable anatomy prevented duplex-directed filter placement in 41 patients (12%). Indication for filter placement included venous prophylaxis in the absence of thromboembolism in 235 patients (83%), contraindication to anticoagulation therapy in 34 patients (12%), prophylaxis with therapeutic anticoagulation therapy in the presence of thromboembolism in 7 patients (2%), and complication of anticoagulation therapy in 8 patients (3%). There were no procedure-related deaths or septic complications. Technical complications occurred in 12 patients (4%). Filter misplacement occurred in 6 patients (2%), access thrombosis in 1 (<1%), migration in 1 (<1%), bleeding in 1 (<1%), and IVC occlusion in 3 (1%). Pulmonary emboli after IVC filter placement occurred in one patient with a misplaced filter. Average hospital charges related to duplex ultrasound-directed filter placement were $2388 less than fluoroscopic placement charges in the year 2000. CONCLUSION: Our experience indicates that duplex ultrasound-directed IVC filter placement is safe, cost-effective, and convenient for patients who need IVC filter placement.