Intramural Delivery of Agent via a Novel Drug-Delivery Sleeve
Histological and Functional Evaluation
Background The infusion sleeve is a novel drug-delivery catheter system designed to deliver an agent under controlled conditions into the arterial wall at the site of angioplasty. The purpose of the present study was to characterize the delivery agent via the infusion sleeve in ex vivo and in vivo models.
Methods and Results The delivery of horseradish peroxidase via the infusion sleeve was studied in a porcine explanted heart model. Under physiological conditions, arteries underwent balloon injury (≈10% overstretch), after which horseradish peroxidase (2.5 mL) was delivered at specific pressures. Cross-sectional analysis demonstrated greater staining when the agent was delivered at increasing pressures. The infusion sleeve was evaluated in an in vivo canine coronary model. With an infusion sleeve loaded over a standard dilatation catheter through a 9F guide, overstretch balloon injury was performed, after which fluoresceinated heparin was delivered. Animals were killed 2 hours after delivery. Fluoresceinated heparin–treated segments demonstrated high fluorescence signals, localizing with smooth muscle cell nuclei with less activity in the interstitium. The functional significance of intramural heparin delivery was studied in a porcine carotid model. In the presence of 111In-labeled platelets, arteries underwent overstretch injury followed by delivery of heparin (50 or 100 units/kg) or vehicle. Platelet deposition was reduced at 30 minutes (57%, P<.01) and 12 hours (39%, P=.06) compared with saline controls.
Conclusions Agent delivery via the infusion sleeve is pressure dependent; transmural delivery is possible with minimal disruption of arterial wall architecture; the infusion sleeve is compatible with standard angioplasty equipment; and heparin delivery at the site of balloon injury significantly reduces platelet deposition in a porcine model for a minimum of 12 hours.
PTCA continues to be limited by abrupt closure and restenosis.1 Multiple promising antirestenosis drugs have had disappointing results when tested in clinical trials. Pharmacological therapy is often limited by the inability to obtain optimal concentrations of an agent within the arterial wall. Some agents are limited by unwanted side effects distant to the angioplasty site, eg, bleeding complications with direct thrombin inhibitors and heparin.2 3 Other agents do not have favorable pharmacokinetics for systemic delivery. The ability to deliver agents locally, directly into the arterial wall, would overcome many of these problems. This strategy would provide a method of obtaining markedly higher concentrations at the site of angioplasty without being limited by extracardiac toxicities.
A successful local delivery system must deliver sufficient agent to the site of interest with an acceptable safety profile without inducing injury that might lead to restenosis. Furthermore, the delivery system must be easy to use and cost effective.
First-generation drug-delivery systems have been hampered by ineffective delivery and/or unacceptable injury.4 The structure of the first-generation drug-delivery systems consists primarily of modified angioplasty balloon catheters. The simplest of the modified PTCA balloon systems is the perforated balloon, first described by Wolinsky and Thung.4 The conditions under which an agent exits the balloon are determined by proximal inflation pressure, perforation size and number, and the fit of balloon artery. Acceptance of the perforated balloons has been hindered by the “jet” injury associated with the use of these catheters.5 The injury associated with the perforated balloon system may be due in part to the inability to precisely control the delivery conditions. With a perforated balloon catheter, both the balloon artery fit and drug-delivery conditions are determined by inflation pressure. For example, if 8 atm is required to ensure optimal fit, drug delivery must also be performed at 8 atm, even if delivery at this pressure inflicts unacceptable arterial damage.5 To overcome these limitations, balloon systems with porous membranes have been developed.6 The outer porous membrane limits the flow of the agent out of the balloon, reducing it to a low-pressure “weeping” balloon delivery method. It is uncertain whether low-pressure delivery can supply sufficient agent deep into the vessel wall.
To resolve these problems, the parameters associated with agent transfer to the arterial wall require further definition. Delivery of an agent in an aqueous vehicle requires two distinct steps: the apposition of drug-delivery elements against the arterial wall and the delivery of the agent into the wall. To better define the parameters associated with local delivery, we designed a catheter system that allows independent control of both apposition and drug delivery.
LocalMed Infusion Sleeve
The LocalMed infusion sleeve is a multilumen catheter consisting of a proximal infusion port, proximal hub, main catheter shaft, and distal infusion region with multiple side holes (Fig 1⇓). The catheter has a central lumen for dilatation catheter and guide wire access as well as four separate outer lumens for drug delivery. Side holes are located within the infusion region near the distal tip of the infusion sleeve. Radiopaque markers are also located in each drug-delivery lumen within the infusion region. The agent travels through the proximal infusion port and the outer infusion lumens and exits via side holes (nine 40-μm-diameter holes per drug-delivery lumen). The infusion sleeve is designed to track over standard dilatation balloon catheters and can be positioned relative to the balloon in one of three configurations, as outlined in Fig 2⇓. The infusion sleeve has been designed to provide independent control of the apposition of the drug-delivery elements against the arterial wall (determined by the inflation pressure of the underlying PTCA balloon) and delivery of the agent into the arterial wall (determined by the infusion pressure of the drug-delivery elements).
When used in vivo, the infusion sleeve is loaded onto a standard balloon dilatation catheter of the operator’s choice. PTCA is performed in the usual fashion, with the infusion sleeve retracted within the guide (retracted configuration). After angioplasty, the infusion sleeve is advanced over the dilatation catheter, aligning the infusion region of the infusion sleeve with the underlying PTCA balloon (aligned configuration). Drug delivery is then performed under specific apposition and infusion conditions. After drug delivery, the infusion sleeve can be retracted, allowing the unsheathed PTCA balloon to perform additional angioplasty; alternatively, the infusion sleeve/PTCA balloon catheter assembly can simply be removed.
For selected parts of the study, the infusion sleeve was specially manipulated to inactivate three lumens, leaving one active lumen. This was done to have actively treated and control sectors within each arterial segment. In the explanted heart model studies, infusion sleeves with either one or four active lumens were used as indicated. Infusion sleeves with one active lumen were used in the canine studies. Due to the larger diameter of the porcine carotid arteries (compared with porcine or canine coronary arteries), infusion sleeves with six lumens (all active) were used. All studies were performed in accordance with institutional guidelines.
Computer-Controlled Pump System
A computer-controlled pump system was used to ensure precise delivery and for monitoring pressure during delivery. The agent was loaded into a syringe mounted in a Harvard Apparatus pump, and connected to the proximal end of the infusion sleeve. Proximal pressure was monitored with a Wheatstone Bridge pressure transducer (XTC-190-200, Kulite Semiconductor Products).
Explanted Porcine Heart Model
Freshly harvested porcine hearts were obtained from a slaughterhouse and transported to our facility in iced physiological saline solution. The left coronary ostium was isolated, after which a standard 9F short side arm sheath was introduced subselectively into the left anterior descending coronary artery. The artery was perfused with physiological saline at 37°C at systemic pressures for a minimum of 5 minutes. Arterial segments (3 mm diameter) were identified by intravascular ultrasound with standard techniques (CVIS, Inc). The infusion sleeve was loaded onto a standard PTCA balloon catheter (3.0 mm). With the infusion sleeve in the retracted configuration, overstretch injury was performed by inflating the balloon to its nominal pressure for 60 seconds. After balloon deflation, the infusion sleeve was advanced over the PTCA catheter to the aligned configuration. The balloon was then reinflated to the same pressure, and the marker agent (HRP, 1.0 mg/mL; Sigma Chemical Co) was delivered at a specified proximal pressure.
The artery was then perfused with physiological saline for 5 minutes and fixed with isotonic glutaraldehyde. Frozen sections were prepared by a standard technique. HRP was detected with 3′,3-diaminobenzidine–H2O2 reaction.7 Arterial cross sections were evaluated by technicians blinded to the specific drug-delivery protocol used. Each arterial section was divided visually into 18 sectors. The presence of discoloration was noted in each sector with the use of a binary scale. The results are presented as the percentage of sectors with HRP discoloration present ±SEM. Injury to the arterial wall was graded in a similar fashion. The two-sided paired Student’s t test was used when comparing drug-delivery groups.
Injury associated with drug delivery was evaluated by delivering physiological saline. Arteries were fixed in formaldehyde and stained with hematoxylin and eosin by a standard technique.
Acute Canine Coronary Model
Under general anesthesia, a cutdown was performed on the right carotid artery. A 9F short side arm sheath was placed and a standard 9F hockey stick guide catheter was introduced into the ascending aorta to the ostium of the left main coronary artery. Intracoronary nitroglycerin (100 μg) was delivered, after which coronary arteriography was performed. The internal diameter of the arterial lumen was determined by intravascular ultrasound with a standard technique.
An infusion sleeve was then loaded onto an appropriate-size PTCA balloon catheter. With the infusion sleeve in the retracted configuration, angioplasty injury was performed at ≈110% stretch for 60 seconds. With the PTCA balloon deflated, the infusion sleeve was tracked over the PTCA catheter to the aligned configuration. The balloon was then reinflated to 4 atm, and 2.5 mL of FITC-heparin (Polysciences, Inc) (150 units/mL) was delivered over approximately 20 seconds with an estimated proximal pressure of 100 psi.
The animal was maintained under general anesthesia for approximately 2 hours after delivery, when it was killed. Frozen arterial sections were prepared with a standard technique and evaluated with fluorescence microscopy.
Porcine Carotid Overstretch Model
The functional significance of heparin delivery was evaluated in a well-described model of platelet thrombus formation at the site of angioplasty injury.8 9 Cross-breed Yorkshire swine received autologous 111In-labeled platelets that were allowed to circulate for a minimum of 12 hours. With the animal under general anesthesia, femoral arterial access was established, after which a bolus of heparin (50 units/kg) was administered via the arterial sheath. Carotid artery overstretch was performed with a standard balloon dilatation catheter (8.0 mm × 2.0 cm, Meditech, Inc). Five inflations of 30 seconds at 6 atm each were performed at 1-minute intervals. Drug delivery was then performed on the index artery. A six-lumen infusion sleeve was loaded onto a standard balloon dilatation catheter (6.0 mm×2.0 cm, Meditech, Inc) and placed at the site of previous arterial injury. Heparin was delivered (8 mL, 50 units/kg) with a balloon support pressure of 2 atm at a proximal pressure of approximately 200 psi. Saline (8 mL) was then delivered to the contralateral carotid artery at the site of previous overstretch injury after a similar protocol. Animals were killed at 30 minutes or 12 hours after the angioplasty procedure. For animals killed at 12 hours, 100 units/kg heparin was infused locally. The carotid arterial segments were then fixed in situ with antegrade perfusion of glutaraldehyde and harvested, and the radioactivity associated with labeled platelets deposited at the site of balloon injury was measured with a gamma counter. The extent of platelet deposition at the site of angioplasty injury was calculated as previously described.8 9
Explanted Porcine Heart Model
The effect of proximal pressure on delivery of HRP into the media after balloon injury was evaluated in the explanted porcine heart model. The agent (HRP, 1 mg/mL, 2.5 mL) was delivered via the infusion sleeve with one of four active drug delivery lumens at 50 (n=10), 100 (n=24), 150 (n=8), or 200 (n=5) psi proximal pressure over 39, 21, 12, and 8 seconds, respectively, resulting in 0.7±0.5%, 4.2±0.5%, 10.9±4.0%, and 36.7±8.7% coverage of the cross-sectional area evaluated (Fig 3⇓). Histological evaluation of sections treated with a proximal pressure of 100 psi demonstrated transmural discoloration with preservation of the arterial architecture (Fig 4⇓⇓). Efficacy of HRP delivery was a function of proximal delivery pressure.
Qualitative review of arterial sections treated at proximal pressures of 150 and 200 psi demonstrated frequent areas of medial separation that were not found in the control areas. In sections treated at proximal pressures of 50 and 100 psi, areas of medial separation were rarely observed. Hematoxylin and eosin slides from representative arterial sections treated with saline (8 mL, proximal pressure of 100 psi, balloon support pressure of 6 atm) via an infusion sleeve with four active lumens and control sections are shown in Fig 5⇓. The treated segments demonstrated subtle changes in the nuclear architecture with karyorrhexis and pyknosis. No significant differences were noted in terms of medial tears or dissections.
Acute Canine Coronary Model
The ability to deliver an agent in vivo with the infusion sleeve was evaluated in an acute canine coronary model. The infusion sleeve was loaded onto a 3.0-mm standard PTCA balloon catheter and introduced into a standard 9F guide catheter. With the infusion sleeve in the retracted configuration, overstretch arterial injury of the midportion of the left anterior descending coronary artery was performed. After balloon deflation, the infusion sleeve was tracked over the balloon while maintaining the position of the PTCA balloon. The balloon was reinflated, and the agent (FITC-heparin) was delivered (proximal pressure of 100 psi). The animal was killed 2 hours after local drug delivery. Gross examination revealed no abnormalities. Fluorescence microscopy revealed transmural deposition of heparin over an arc of approximately 120° in the maximal stained section. Intense fluorescence was noted to localize with the smooth muscle cell nuclei. Less intense fluorescence was noted within the interstitial spaces (Fig 6⇓).
Porcine Carotid Overstretch Model
The functional significance of heparin delivery into the angioplasty-injured arterial segment was then evaluated in the porcine overstretch injury model. At 30 minutes, there was a 57% reduction in platelet deposition at the sites of deep arterial injury compared with control segments (10.7±2.7 versus 24.8±6.5×106 platelets/cm2, n=10, P<.01). This reduction in platelet deposition persisted for 12 hours when a 39% reduction in platelet deposition compared with control segments (19.1±6.6 versus 31.4±13.6×106 platelets/cm2, n=10, P=.06) was observed (Fig 7⇓).
For a drug-delivery catheter to be effective in the treatment of abrupt reclosure or restenosis, the following criteria are important. First, the system must deliver an effective agent. Second, therapeutic concentrations must be sustained for an appropriate duration. Third, the complications associated with the drug-delivery procedure must have an acceptable rate. Fourth, the drug-delivery system must be easy to use and cost effective.
Before these issues are addressed, the ability of the system to deliver the marker agent must be evaluated. Data from the explanted heart model studies demonstrate the ability of the infusion sleeve to decouple apposition (balloon inflation) from drug infusion (proximal pressure). This feature was used to identify a protocol capable of transmural delivery of the agent with minimal changes within the medial architecture. Histological evaluation of the arteries from the explanted heart studies demonstrated subtle injury when the agent was delivered with a proximal pressure of 100 psi. It seems unlikely that injury of this nature would have an impact on the structural integrity of the artery leading to luminal compromise or perforation.
It is possible that delivery of an agent might add to the injury associated with lesion dilation leading to an increased restenosis rate. Santoian and colleagues5 reported “jet” injury associated with high-pressure delivery in connection with the Wolinsky balloon catheter. The impact of the type of injury induced by the Wolinsky balloon on restenosis is unclear.5 7 In contrast to the perforated balloon systems, the histological injury associated with delivery via the infusion sleeve seems minimal. Further studies are ongoing to evaluate the effect of the injury (if any) on myointimal hyperplasia in appropriate animal models.
The acute canine studies demonstrated the ability of the infusion sleeve to deliver a biologically appropriate agent in a setting closely resembling the clinical situation. Two hours after delivery the agent is detectable. The observation of nuclear localization suggests that the FITC-heparin is qualitatively similar to what has been observed in similar models with a passive delivery device.10 Furthermore, the ability to perform overstretch injury and drug delivery in vivo in the canine coronary circulation with the infusion sleeve in conjunction with a 9F guide demonstrates the feasibility of this approach.
Radiolabeled platelet studies in the porcine carotid overstretch model were designed to demonstrate the functional utility of heparin when delivered with a clinically relevant protocol. An important feature of this protocol is that both heparin and saline control deliveries were performed in each animal. The effect of heparin entering the systemic circulation as a result of local delivery was evaluated by comparing platelet deposition in the heparin-treated segments with that in the contralateral control segments. Local delivery of heparin resulted in a significant reduction in platelet deposition and was observed at both 30 minutes and 12 hours. It is important to note that the dosing regimen with a total of 100 to 150 units/kg heparin is clinically feasible. Recent studies have demonstrated the clinical utility of agents that have been shown to reduce platelet thrombus formation at the site of balloon injury in a similar model.11 When a GP IIb/IIIa inhibitor, 7E3, was administered systemically to patients at high risk for abrupt closure after angioplasty, a 35% reduction in ischemic events during the initial 30-day period was observed, although there was an increased risk of bleeding. This clinical benefit regarding ischemic events was sustained for at least 6 months, implying a reduction in clinical restenosis.12 These results provide indirect evidence that reduction of platelet deposition at the site of PTCA-induced arterial injury may result in a reduction in both abrupt closure and restenosis. Further studies are under way to characterize the effects of locally delivered heparin and other antithrombotics and to document whether the same benefits may be observed with less significant risks of bleeding when these agents are delivered locally.
The concept of catheter-based local drug delivery was first demonstrated by Wolinsky and Thung4 with a perforated balloon catheter. The acceptance of this catheter has been hindered by the injury associated with this type of local delivery. Other groups have sought to limit the injury associated with this delivery by limiting the energy of the agent as it exits the catheter. The Dispatch Catheter (Boston Scientific) has a novel design that permits prolonged bathing of the arterial wall with the agent. This system is an over-the-wire nondilatory system with a helical balloon that forms a spiral around a urethane sheath. Inflation brings the helical balloon against the arterial wall while expansion of the central sheath permits simultaneous distal perfusion. The agent is infused via the spaces between the coils. Heparin delivery via the Dispatch Catheter has been studied in a porcine coronary angioplasty model that is used to measure radiolabeled platelet deposition. At 1 hour, no reduction in platelet deposition was observed compared with control segments. Another low-pressure delivery system is the Microporous Infusion Catheter, which is under development by Cordis Corporation. This is a perforated balloon covered by a porous membrane. When the system is pressurized the balloon is inflated, and the agent “weeps” from the outer membrane. With this catheter in an in vivo porcine coronary model, HRP has been delivered into the media after overstretch injury. The delivery pattern is circumferential within the media.6 Other balloon-based systems have been designed to allow careful titration of drug-delivery conditions. These systems include the Transport Catheter (Cardiovascular Dynamics/Boston Scientific) and Channel Balloon Catheter (Boston Scientific). These systems have been recently reviewed.13
When comparing different drug-delivery catheter systems, the relation between proximal pressure and distal drug-delivery conditions, ie, the hydraulic head of the agent exiting the catheter, may not be equal in different systems. The distal delivery conditions may be quite different when an agent is delivered with a proximal pressure of 100 psi in the infusion sleeve compared with the delivery of the same agent at 100 psi via another catheter with different characteristics. In a closed hydraulic system, such as a standard PTCA balloon catheter, the proximal inflation pressure when stabilized will be equal to the distal pressure in the balloon. In an open hydraulic system in which there is fluid flow, there can be a substantial pressure drop along the catheter. The pressure drop between the proximal and distal ends is a function of many factors, including fluid viscosity, catheter length, cross-sectional area, and geometry of the drug-delivery conduit. In the infusion sleeve, there is a gradual pressure drop of approximately 90% between the proximal and distal ends. This has been verified with specially instrumented catheters to allow measurement of distal tip pressure during agent infusion. These studies demonstrated a pressure drop to approximately one-tenth of proximal infusion pressure. It is important to note that there is no significant change in hydrodynamics between the proximal and distal ends of the perfusion array. In the protocol used to deliver the agent in the acute canine studies, the distal pressure was ≈10 psi (data not shown). When comparing the infusion sleeve with other delivery systems, it is important not to assume that equal proximal pressures will result in equal distal drug-delivery conditions.
The studies performed in the explanted heart model or in vivo in the coronary or carotid artery were designed to evaluate drug delivery under different conditions. The drug was delivered adequately with this catheter system in these models. It is important to note, however, that all of the models described involved normal, nondiseased arteries. This substrate for drug delivery may be different for the human atherosclerotic arterial segment after angioplasty.
The results of the present study demonstrate that drug delivery via the infusion sleeve is pressure dependent and feasible with the use of standard PTCA equipment within a 9F guide. Transmural delivery is possible with minimal disruption of arterial wall architecture. Heparin delivery at the site of balloon injury significantly reduces platelet deposition in a porcine model for a minimum of 12 hours.
Local delivery of heparin with the infusion sleeve is a practical and promising approach for the prevention of abrupt closure and restenosis after PTCA. Clinical studies evaluating the use of heparin delivery via the infusion sleeve are under way.14
Selected Abbreviations and Acronyms
|PTCA||=||percutaneous transluminal coronary angioplasty|
Dr Lam was supported in part by the Medical Research Council of Canada, the Heart and Stroke Foundation of Canada–Quebec Affiliate, and the Fonds de la Recherche en Sante du Quebec. We wish to acknowledge Amy Hankley-Brongiel and Cynthia Simons (Experimental Biology Laboratory, LocalMed, Inc) for performing numerous studies in the explanted model, James Stoughton (Division of Cardiology, University of California at San Francisco) for providing technical expertise during the acute canine studies, Johanne Doucet for providing technical assistance with the porcine carotid overstretch studies, and Sherry Burr and Pam Gallant, RN, for their help in the preparation of the manuscript.
First version received March 9, 1995; accept with revision received April 27, 1995; final revision accepted May 3, 1995.
This work was supported in part by LocalMed, Inc, Palo Alto, Calif. Some of the authors are employees (A.V.K., J.R.K., G.W.G., E.J.K.) or board members (A.V.K., J.B.S.) of LocalMed, Inc, the manufacturer of the drug-delivery device discussed in the article.
- Copyright © 1995 by American Heart Association
The Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO) IIa Investigators. Randomized trial of intravenous heparin versus recombinant hirudin for acute coronary syndromes. Circulation. 1994;90:1631-1637.
Antman EM, for the TIMI 9A Investigators. Hirudin in acute myocardial infarction: safety report from the Thrombolysis and Thrombin Inhibition in Myocardial Infarction (TIMI) 9A trial. Circulation. 1994;90:1624-1630.
Santoian EC, Gravanis MB, Anderberg K, Scott NA, Karas SP, Schneider JE, King SBJ III. Use of a porous infusion balloon in swine coronary arteries: low pressure minimizes arterial damage. Circulation. 1991;84:591. Abstract.
Lincoff AM, Furst JG, Penn MS, Lee P, MacIsaac AI, Chisolm GM, Topol EJ, Ellis SG. Efficiency of solute transfer by a microporous balloon catheter in the porcine coronary model of arterial injury. J Am Coll Cardiol. 1994;23:18A. Abstract.
Lam JYT, Chesebro JM, Steele PM, Badimon L, Fuster V. Is vasospasm related to platelet deposition? Relationship in a porcine preparation of arterial injury in vivo. Circulation. 1987;75:243-248.
Lam JYT, Badimon JJ, Ellefson RD, Fuster V, Chesebro JH. Cod liver oil alters platelet-arterial wall response to injury in pigs. Circ Res. 1992;71:769-775.
Kaplan AV, Leung LLK, Leung W-H, Grant GW, McDougall IR, Fischell TA. Roles of thrombin and platelet membrane glycoprotein IIb/IIIa in platelet-subendothelial deposition after angioplasty in an ex vivo whole artery model. Circulation. 1991;84:1279-1288.
Kaplan AV, Vandormael M, Bartorelli A, Hofmann M, Stoerger H, Simpson JB, Reifart N. Heparin delivery at the site of angioplasty with a novel drug delivery sleeve: initial clinical series. J Am Coll Cardiol. 1995;25:286A. Abstract.