Site-Specific Intracoronary Heparin Delivery in Humans After Balloon Angioplasty
A Radioisotopic Assessment of Regional Pharmacokinetics
Background Demonstration and quantification of site-specific intracoronary administration of compounds has been confined thus far to the experimental animal laboratory. The aim of this study was to describe a scintigraphic method to demonstrate site-specific intracoronary drug delivery in humans. The methods allow on-line visualization and off-line quantification of site-specifically infused γ-emitting compounds.
Methods and Results In 12 patients after balloon angioplasty, 99mTc-labeled heparin was administered at the site of dilatation by use of a coil balloon. Both the infusion period and the washout period after the end of infusion were monitored with a γ-camera. A curve of counts per pixel as a function of time was derived that showed an accumulation phase during infusion followed by a washout phase after the end of infusion. Both phases were fitted by regression analysis and showed a linear accumulation pattern and a biexponential washout pattern. After correction for background counts, 99mTc decay, and body attenuation, peak heparin amount and regional bioavailability were calculated. Peak amount was defined as the initial point of the slow washout component of the biexponential curve (elimination component), and regional bioavailability was defined as the area under the curve of accumulation and washout phase. Half-life and retention time, defined as seven half-lives, were obtained by use of the elimination component after correction for 99mTc decay. Mean peak delivered amount was 45±44 IU (236±228 μg), corresponding to an efficiency of delivery ranging from 1% to 8% of the totally infused dose. Total regionally bioavailable heparin reached 244±194 IU·h (1.28±1.01 mg·h). Retention time varied from 12 to 90 hours (mean, 50:33±22:50 hours:minutes).
Conclusions Site-specific intracoronary heparin delivery after angioplasty by means of the coil balloon was demonstrated in humans, and regional pharmacokinetics was quantified by use of a radioisotopic technique.
After PTCA, restenosis occurs in 35% to 50% of initially successful procedures.1 2 Our current understanding of restenosis suggests that it involves a sequence of pathophysiologically interrelated mechanisms that can be classified as (1) thrombus formation,3 (2) cell migration and neointimal proliferation,4 5 6 and (3) acute recoil7 and chronic remodeling.8 Attempts to reduce the incidence of restenosis by use of systemic administration of many different drug classes during and/or after PTCA9 10 have all failed to be effective. One exception has been the use of platelet glycoprotein IIb/IIIa receptor antibodies, but these findings await confirmation.11 The lack of efficacy of systemic drug administration to prevent restenosis has been attributed to insufficient compound concentration at the site of angioplasty. Site-specific delivery has been proposed to result in higher concentrations of chemical compounds at the site of angioplasty.12 13
Because previously used methods for quantification require tissue sampling, evaluation of the effectiveness of site-specific intracoronary delivery of drugs has been confined thus far to animal studies. Previously described techniques consist of either histological or autoradiographic quantification of penetration of a compound into the arterial wall13 14 15 16 or well counting of radioactivity levels of tissue samples.17 18 19 20
The aim of the present study is to describe a clinically applicable strategy for assessment of site-specific postangioplasty intracoronary delivery of 99mTc-labeled heparin by use of a particular infusion catheter. This method, which allows on-line visualization and off-line quantification of site-specifically administered γ-emitting compounds, need not be confined to the cardiovascular field but could find application in the fields of oncology and infectious diseases as well. Ultimately, this technology could be used to optimize site-specific efficacy of a compound using the information derived from regional pharmacokinetics.
The local drug-delivery device (Dispatch; Scimed Systems Inc, Maple Grove, Minn) is depicted and described in Fig 1⇓. By design, the coil balloon is a nondilatational infusion catheter meant to be used after percutaneous revascularization procedures. With inflation of the coil balloon, a nonporous polyurethane sheath forms a tube-shaped structure that allows both coronary perfusion through the central conduit and drug infusion over the external surface between the coils (drug compartments). The feasibility and safety of this device in humans have been reported previously.21 22
Study Population and Protocol
The study cohort consisted of patients with stable or unstable angina who were eligible for balloon angioplasty of a native coronary artery lesion. Angiographic exclusion criteria were as follows: target lesion at a bifurcation of a major side branch, target lesion >15 mm in length, reference portion of the vessel <2.5 mm in diameter, and subtotal stenosis with TIMI flow grade 1 or total occlusions. The protocol was accepted by the medical ethics committee of the hospital, and written informed consent was obtained from all patients before inclusion in the study.
After balloon angioplasty with an angiographically optimal end result, the coil balloon was inserted through an 8F guiding catheter and inflated at the angioplasty site. An inflation time of 30 minutes was planned unless symptomatic intolerance or ECG signs of ischemia occurred. A saline solution containing radiolabeled heparin (100 IU/mL) was infused at a rate of 36 mL/h at the site of angioplasty by use of a volume-controlled syringe pump (P 3000 syringe pump; Welmed). During intracoronary inflation, fluid-mediated pressure in the drug-delivery compartment of the device was recorded through the infusion port, allowing an estimation of the pressure in the drug compartments.
Lesion-site evaluation was performed angiographically after intracoronary injection of 2 mg isosorbide dinitrate (Cedocard; Cedona Pharmaceuticals BV). Preintervention lesion type was classified according to the modified American Heart Association classification,23 and the presence of thrombi and calcifications at the lesion site was recorded. Dissections occurring after angioplasty and drug delivery were documented according to the classification described by Huber et al.24 Dissection length and lesion diameters before PTCA, after PTCA, and after drug delivery were determined by quantitative coronary angiography (CAAS II; Pie Medical).25 26
Radiopharmaceutical and Scintigraphic Methods
The efficiency of site-specific heparin delivery was assessed by the infusion of 99mTc-labeled heparin. 99mTc-labeled heparin was prepared by use of a modification of a previously described method27 : 1 mL (5000 IU) of commercially available heparin (heparin Leo; Leo Pharmaceutical Products BV) and 103 MBq of 99mTc pertechnetate were added to a 0.35 g/L stannous chloride solution. The stannous chloride solution was prepared by the hospital pharmacy in 10-mL vials containing 0.35 mg stannous chloride dehydrate (Merck) in 1 mL of water for injection and was sterilized by membrane filtration (0.2 μm) and autoclave (15 minutes, 121°C). The labeled heparin solution was then diluted to 50 mL with sodium chloride 0.9% in a sterile, 60-mL syringe. The syringe was provided with a membrane filter (0.2 μm; Sterifix EF; B. Braun Melsungen AG). Before infusion of the solution, the filter was preconditioned with 10 mL of the heparinized solution. Quality control of 99mTc-to–heparin complex formation (paper chromatography: solid phase, Whatman No. 1; liquid phase, acetone) was performed before and after filtration and showed a labeling yield of >95% (99mTc-heparin activity/99mTc-heparin activity plus 99mTc pertechnetate). The stability of the 99mTc-labeled heparin in the injection fluid over time was tested by use of routine Sephadex G-25 gel filter with PD-10 columns (Pharmacia) immediately after preparation, at the time of administration to the patient, and 24 hours after preparation. 99mTc pertechnetate (99mTcO4−) was used as a control. During elution with 0.9% NaCl, 0.5-mL fractions were collected. After elution of the void volume, the high-molecular-weight 99mTc heparin complex elutes in the first 3.5 mL, whereas free 99mTcO4− elutes mainly between 9 and 15 mL. Thirty-seven megabecquerels (1 mCi) of 99mTc heparin (corresponding to 1800 IU) was to be administered during a 30-minute period.
Scintigraphic images were acquired in a planar 45° LAO projection with a small-field (diameter, 15 cm) mobile γ-camera (Cardiac; Siemens) equipped with a 140-keV low-energy, parallel-hole collimator. The analyzer was set to 140 keV with a 15% window. Acquired images (matrix, 64×64 pixels) were stored in a PDP 11/34 computer system (Digital). Regional activity was registered as dynamic acquisitions for 40 minutes (1 frame/40 s) in the catheterization suite followed by dynamic acquisitions of 15 minutes each hour (1 frame/40 s; except for the first two patients, in whom one static acquisition of 15 minutes was performed) for ≤8 hours after the procedure. Regional activity was quantified by counts of the site of interest (99mTc-labeled heparin delivery area). Counts were expressed as counts per pixel (absolute counts normalized by number of pixels of the area of interest) and depicted over time. The resulting curve of counts per pixel over time showed an accumulation phase corresponding to the 99mTc-labeled heparin infusion period followed by a washout phase starting at the end of infusion (Fig 2⇓). To ensure a sharp transition between accumulation and washout phase, coil-balloon deflation and discontinuation of infusion were performed simultaneously before removal of the catheter.
A polyurethane bottle filled with a well-defined aliquot of the injection fluid was used as standard. Activity was measured as static acquisition during a 180-second period before the start of the procedure.
The thyroid was scanned for ≤8 hours after administration of the radiolabeled heparin. Thyroid uptake was blocked neither by potassium iodide nor potassium perchlorate.
From Scintigraphic Acquisition to Pharmacokinetics: Concept and Definitions
The curve of the counts per pixel as a function of time shows an accumulation phase during infusion of 99mTc-labeled heparin followed by a washout phase at the end of the infusion (Fig 2⇑). This curve is used to derive regional pharmacokinetics.
The accumulation phase was expected to obey a logarithmic or linear accumulation pattern (Fig 3⇓, top) dependent on the achievement of a steady-state condition between the central compartment (extraluminal region) and the peripheral compartment (endoluminal cavity).
In accordance with the open two-compartment model, the washout phase was assumed to be biexponential, reflecting the superposition of two kinetics of washout: a steep distribution component (α-phase) and a flat elimination component (β-phase). The distribution component reflects a fast washout of readily accessible heparin in the peripheral compartment; the elimination component reflects a slow washout of not readily accessible heparin from the central compartment (Fig 3⇑, bottom).
Calculation of pharmacokinetic parameters by use of the scintigraphic acquisition is described in the “Appendix.” The method will be referred to as RIT-PK (RadioIsotopic Technique to determine PharmacoKinetics).
Peak regional heparin amount is defined as the initial value of the flat elimination component (β-phase), obtained by back-extrapolation to zero time (end of infusion or beginning of washout). It is expressed either as an absolute value (in international units or micrograms) or as a percent of the totally infused drug amount (Fig 4⇓, top).
Biological half-life is defined as the time period to decrease regional drug amount to half of its initial value according to the flat elimination component (β-phase) (Fig 4⇑, middle).
Retention time is defined as seven half-lives, which corresponds to a >99% decrease of the initial regional drug amount according to the flat elimination component (β-phase).
Regional bioavailability of heparin is defined as the area under the curve, which includes the accumulation phase as well as the washout phase. It is expressed either as drug amount as a function of time (international units times hour or milligrams times hour) or as drug concentration as a function of time (international units per gram times hour or milligrams per gram times hour) (Fig 4⇑, bottom).
Vessel permeability is defined as the physical property that permits pressure-induced centrifugal fluid flow through an arterial segment (20-mm coil-balloon length).
Continuous variables were expressed as mean±SD. Regression analysis was performed on the uncorrected counts per pixel over time of the accumulation phase. Linear and logarithmic models were fitted. The absolute values of the residuals of these models were compared by two-tailed, paired t tests. A probability value less than .05 was considered to be significant. Nonlinear regression analysis (biexponential) was performed on the uncorrected counts per pixel over time of the washout phase and for 99mTc decay–corrected counts. Goodness of fit of all models was defined as the ratio of regression sum of squares to total sum of squares (r2 value).
Effects of Conventional Balloon Angioplasty and Prolonged Coil-Balloon Inflation
MLD of the treated lesions was 0.93±0.28 mm before PTCA. Post-PTCA MLD increased to 1.95±0.52 mm with balloon nominal values of 2.5 mm to 4.0 mm, a mean inflation pressure of 10.8±3.4 atm (range, 6 to 18 atm), and a mean inflation time of 317±157 seconds (range, 120 to 600 seconds). Angioplasty balloon–to-artery ratio was 1.08±0.12 mm. After insertion of a coil balloon (ranging in size from 3.0 to 4.0 mm nominal size) at a mean inflation pressure of 7.7±1.9 atm during an average inflation time of 30±4 minutes, MLD increased to 1.98±0.52 mm (P=NS). In 9 of 12 cases (exceptions: patients 4, 10, and 11), the coil-balloon diameter size was 0.5 mm larger than the largest-used angioplasty balloon. In 2 cases (patients 3 and 5), the MLD decreased despite the use of a high inflation pressure in this nondilatational device (radial pressure <2 atm), perhaps reflecting a change of the shape of the coil section (shift from round to oval). Angiographically, this may result in a decreased luminal diameter, especially in the single working projection.
Three dissections (type A; type B, length 10.4 mm; and type D, length 28.7 mm) were observed after balloon angioplasty. The type B dissection (patient 4) was angiographically no longer visible after prolonged inflation of the coil balloon, whereas the type A and D dissections remained unchanged after site-specific heparin administration (patient 9, type A; patient 11, type D, length 29.5 mm).
Individual infusion parameters are reported in Table 3⇓.
Infused volume was 18.1±2.2 mL (range, 12.6 to 21.2 mL) and was derived from the rate of infusion and duration of infusion. The average dose of infused heparin was 1807±222 IU (range, 1260 to 2120 IU).
The mean back pressure (pressure measured at the proximal infusion hub) during infusion ranged from 57 to 136 mm Hg. The mean back pressure of the device on bench testing with 100 IU/mL heparin at an infusion rate of 36 mL/h was 23.92±0.09 mm Hg (mean value for 15 measurements). The pressure difference is indicative of the pressure in the isolated drug-infusion compartment and ranged from 33 to 112 mm Hg (mean, 83±26 mm Hg).
The counts per pixel curves of uncorrected counts for 8 patients are shown in Fig 5⇓. The washout phase was not available in 4 of this initial cohort of 12 for the following reasons: (1) time lapse between the last scintigraphic acquisition in the catheterization laboratory and that outside the catheterization laboratory >2 hours (n=2; patients 5 and 6) and (2) liver superimposed on the drug-delivery site (n=2; patients 10 and 12).
Calculated pharmacokinetic parameters for regionally delivered 99mTc-labeled heparin are tabulated on an individual basis in Table 6⇓.
The mean amount of regionally delivered heparin, expressed as percent of the infused dose, was 2.5±2.4% (range, 0.6% to 7.8%), corresponding to 45±44 IU (range, 11 to 141 IU) or 236±228 μg (range, 58 to 739 μg) (190.9 IU=1 mg).
Average half-life reached 7:13±3:16 h:min (range, 2:29 to 12:50 h:min), corresponding to a mean retention time of 50:33±22:50 h:min (range, 12:25 to 89:51 h:min), reflecting a washout of >99% of the amount of site-specifically delivered heparin.
The averaged regional amount of bioavailable heparin (area under the curve) was 244±194 IU·h (range, 57 to 611 IU·h).
The slope of the linear accumulation phase ranged from 3.4 to 38, reflecting an ≈10-fold difference in permeability among the evaluated vessels.
Site-specific administration of pharmacological compounds labeled with γ-emitting radionuclides allows for the assessment of regional intracoronary delivery without the requirement of tissue sampling (Fig 6⇓) and therefore can be used for clinical application. The derived regional pharmacokinetics of the compound reflects its chemical properties, the modality of delivery, and the characteristics of the site of delivery. In addition, quantification of regional pharmacokinetics allows interindividual comparison and assessment of the short-term regional behavior of the compound in a clinical perspective (short- and long-term outcomes). Regional pharmacokinetics determined by RIT-PK becomes a surrogate for the prediction of optimal efficacy of site-specifically delivered drugs.
The accumulation phase reflects the period during infusion of the compound (Fig 2⇑). The resulting accumulation curve expressed in counts per pixel as a function of time was fitted for a linear model as well as for a logarithmic model (Fig 3⇑, top). Using the coil balloon with a heparin solution of 100 IU/mL and an infusion rate of 36 mL/h for 30±6 minutes, the model that defined a linear relationship more appropriately described the accumulation phase in 10 of 12 patients, with a mean correlation coefficient of .95±.04 (Table 4⇑). This suggests that inflow in the central compartment (extraluminal region or regional fluid infiltration) and outflow in the peripheral compartment (endoluminal cavity or runoff) remained proportional throughout the infusion period in this hypothetical two-compartment, open model. Thus, steady state was not reached using the selected infusion device and infusion parameters. According to this accumulation pattern, one can predict that with higher drug concentrations and/or longer infusion times, an increased delivery can be achieved in the central compartment.
Infusion Pressure, Accumulation Phase, and Vessel Permeability
A hypothesis could be made that the pressure of infusion and/or the slope (b) of the linear accumulation curve (y=a+bx) describes the permeability characteristics of the delivery site. Permeability of the delivery site is expected to be influenced by two types of variables: runoff conditions and filtration conditions.14 Runoff conditions or outflow in the peripheral compartment depends on local anatomy (presence of side branches and amount of vasa vasorum) and on the degree of apposition of the device to the vessel wall. Filtration conditions or inflow in the central compartment depends on tissue characteristics (composition of the vessel wall and surrounding tissue)28 29 and postangioplasty degree of vessel disruption. A link between permeability of the site of delivery and delivery efficiency can be expected if runoff conditions are predictable and possibly contained. The expected relationship would be the steeper the slope, the more permeable the site of delivery and the higher the amount of regionally delivered heparin. Assuming optimal apposition, in our series of 8 patients with accumulation data as well as complete pharmacokinetic parameters, no consistent relationship between delivery pressure, slope, and delivery amount of heparin (Tables 3⇑, 4⇑, and 6⇑) could be found, even in the subgroup (n=5; patients 2, 3, 4, 7, and 11) without angiographically visible side branches at the drug-delivery site (Table 2⇑). The absence of a predictable dependence between runoff conditions and filtration conditions makes quantification of delivery efficiency from permeability characteristics impossible. Thus, the quantification of delivery efficiency is dependent on an accurate assessment of the washout phase.
The postulated biexponential washout was interpreted as an early fast washout component due to endoluminal distribution (or runoff) of heparin not adherent to vascular and perivascular structures and a second slow component assumed to represent tissue elimination after regional infiltration (Fig 3⇑, bottom). This two-compartment open model of washout was corroborated by a correlation coefficient >.95 (r2=.97±.01; Table 5⇑). On the basis of this assumption, delivery amount, half-life, and retention time of the site-specifically infused heparin were calculated using the flat elimination curve (β-phase).
Peak regional delivery amount ranged between 0.6% and 7.8% of the total dose infused, corresponding to 11 to 141 IU or 58 to 739 μg heparin (190.9 IU=1 mg). The regional concentrations of a compound are difficult to predict because the distribution volume can only be estimated. Assuming intramural delivery and a volume of distribution of 1 mL corresponding to 1 g of coronary artery tissue, tissue concentration achieved would be 45±44 IU/g or 236±228 μg/g (arterial model of 3-mm ID, 2-mm wall thickness, and 20-mm length [volume=1 mL] and tissue density of water [density=1 g/cm3]). Although this amount may seem small, a systemic dose of >3 000 000 IU would be required to reach the same tissue concentration in a person weighing 70 kg if the body were considered as a single homogeneous compartment.
A recent study19 using delivered microspheres and combining histological and autoradiographic approaches has shown that the microspheres penetrated beyond the vessel wall, reflecting a regional delivery pattern, and <0.17% could be located intramurally. Thus, an assumption of exclusively intramural delivery overestimates delivery efficiency by a factor of ≈15 (ratio of mean delivery efficiency: 2.54%/0.17%). By use of this correction factor, calculated “intramural” concentration in humans is expected to be 3±2.9 IU/g (16±15 μg/g).
In both situations, the regional and “intramural” concentrations obtained (236±228 μg/g or, after correction for intramural delivery, 16±15 μg/g) are of similar magnitudes to those that have been shown to be effective for the suppression of cell proliferation in vitro30 and in vivo31 (10 to 20 μg/mL).
An alternative way to estimate efficiency of delivery is to consider the regionally bioavailable amount of heparin derived from the area under the curve. This integrated approach has the advantage of accounting for delivery modality, peak delivery amount, and retention time. The resulting regional exposure to 244±194 IU·g−1·h (1.28±1.01 mg·g−1·h) of heparin per gram of tissue or, after correction for “intramural” delivery, to 16±13 IU·g−1·h (85±67 μg·g−1·h) is 5.3× higher or 2.8× lower, respectively, than the integrated dose during a 24-hour period to suppress in vitro rat smooth muscle cell and BHK fibroblast proliferation.30
Data derived from systemic heparin administration have established that heparin half-life is dose dependent: the higher the dose, the longer the expected half-life.32 The systemic half-life of heparin ranges between 30 minutes and 5 hours.33 34 One would expect from plasma kinetic–derived data that the attainment of higher regional concentrations would result in a longer regional half-life due to slower metabolism. In fact, regional half-life (mean, 7:13±3:16 h:min; range, 2:29 to 12:50 h:min) is longer than plasmatic half-life but shorter than predicted [calculated “intramural” concentration, 3 IU/g or 3000 IU/kg; expected biological half-life, (26+0.323×D)±12 minutes=16:58 h:min, where D is the dose in international units per kilogram].32 This suggests that redistribution according to the postulated two-compartment open model and, to a lesser degree, regional metabolism are rate determining for regional heparin half-life. The flat washout component therefore seems to be related more to cellular adhesion31 or extracellular binding35 of heparin.
Comparison With Literature
Previous studies in normal porcine arteries have demonstrated an efficiency of intramural delivery of 3H-labeled heparin of 0.05% and 1.3% using the coil balloon,20 36 0.6±0.2% with the microporous balloon,37 and 2.3±1.63% (0.21% to 5.48%) with the hydrogel-coated balloon.17 These data show that peak delivery efficiency is in the range of a few percent regardless of animal model, type of device, or quantification technique. The values exhibited a scatter of one order of magnitude despite the use of healthy coronary arteries and limitation of the evaluation of efficiency to intramural deposition.
Retention time was estimated to be 90 minutes for 85% washout (≅3 half-lives) after coil-balloon application36 and up to 1 hour after microporous balloon application.37 Fluoresceinated heparin delivered by the hydrogel-coated balloon was detected intramurally up to 24 hours after administration in canine arteries38 and up to 48 hours if delivered by the porous balloon.13
A recent report20 described the infusion of 3H-labeled heparin by use of the coil balloon in an in vivo healthy porcine coronary artery–injury model. Intramural pharmacokinetic data were determined by scintillation counting of the explanted artery at successive time points. This approach approximates our methods and provides an interesting comparison, especially for one of the selected infusion modalities (heparin solution, 152 IU/mL, 30 mL/h during a 30-minute period). The authors found a biexponential washout pattern and an intramural delivery efficiency of 0.04±0.01% with a half-life of 53 hours. By comparison, our human data showed an “intramural” delivery efficiency (regional delivery of 2.5±2.4% divided by 15) of 0.16±0.16% (range, 0.04% to 0.52%) and a half-life of 7:13±3:16 h:min. Efficiency of delivery was of a similar magnitude, but the human data showed a wider dispersion of the same range, as previously described in the literature.17 36 37 Half-life was substantially different, underlining the limitations of the discontinuous nature of data acquisition in the animal experimental setting. In fact, each evaluated time point in the present study represents one animal study. Quantification of dynamic processes such as kinetics is thus less reliable. RIT-PK allows the determination of pharmacokinetics (accumulation and washout) as a continuous process performed in the same subject. In our setting, the mean time period of data acquisition was 7:58±32 h:min, corresponding to an average of 185±32 data points.
Lessons From Individual Cases
A case-by-case analysis to evaluate pharmacokinetics in the context of angiographic morphology seems appropriate owing to (1) the amount of data collected with a high correlation coefficient (>.95) for the postulated pharmacokinetic model and two-compartment open model and (2) the observed interindividual dispersion of the results.
Accumulation Pattern, Regionally Delivered Amount, and Retention Time
Three patients (patients 4, 9, and 11) presented an angiographically visible dissection after angioplasty.
In 2 patients (patients 4 and 9), a logarithmic accumulation pattern was observed, whereas the remaining 10 patients presented a linear pattern. Patient 4 experienced a type B dissection after PTCA, which could be covered in its entire length (10.4 mm) by the coil balloon and which was reduced by the prolonged inflation-infusion period. Two mechanisms may be responsible for the observed logarithmic accumulation pattern in this setting: (1) a reduction of the permeability characteristics of a vessel by precluding access to side branches, thereby decreasing runoff conditions, and (2) an augmentation of access to the layers underlying the endothelium (plaque and media), which possess a higher affinity to heparin. An increased affinity of heparin to exposed, traumatized, and positively loaded tissue has been reported previously in the literature.39 However, a logarithmic accumulation pattern was not found in patient 11, who also experienced a clearly visible long dissection plane (type D dissection) on angiography. The main difference between the two patients was the relationship between dissection length and device length. In patient 11, the extremes of the dissection were not covered by the coil balloon. In this setting, runoff conditions may prevail, as illustrated by the increase in dissection length after site-specific drug infusion (length, 28.7 mm before infusion versus 29.5 mm after infusion). Furthermore, apposition conditions were not optimal (balloon coil–to-balloon ratio of 1:1). As might be predicted, this patient exhibited a low amount of deposited heparin (<1%; 15 IU or 79 μg). Patients 4 and 11 illustrate that local trauma may not influence site-specific drug delivery in the same way and that other parameters, such as the relationship between the length of the device and the extent of the trauma and/or degree of apposition, may play an additional role.
Patient 9 presented a type A dissection after PTCA that remained unchanged after prolonged site-specific heparin infusion. The trend toward a logarithmic accumulation in this second patient can be explained by altered permeability characteristics of the vessel due to the presence of thrombotic material (eg, side branches and vasa vasorum). This patient presented the lowest regional amount of deposited heparin (<1%; 11 IU or 58 μg). Local thrombus was able to alter not only the delivery pattern and delivery amount but the retention time as well. In fact, heparin retention time was prolonged, which may be explained by adherence of heparin to the thrombus surface secondary to the affinity of heparin to antithrombin III. This suggests that thrombus-rich lesions have less uptake of regional heparin even though they compensate, at least in part, with a longer local retention. Whether this means that a thrombus-rich lesion is less suitable for site-specific heparin deposition must be confirmed by further studies.
In our series of patients, there were two striking observations concerning delivery efficiency: the behavior of a restenotic lesion and of a thrombus-containing lesion (patient 9, previously described).
In patient 2 (restenotic lesion), the uptake of the lesion (14 IU, 76 μg) was the second lowest of any patient in our series and was threefold lower than the averaged de novo lesions (49±45 IU, 258±236 μg). These results may reflect changes in local vessel structure subsequent to the healing process and confirm that permeability of the site of delivery is influenced by tissue characteristics, as previously reported.28 29 If confirmed, this finding would have the practical implication that a restenotic lesion would not be an ideal candidate for local drug delivery with the coil balloon.
Patient 1 presented a shorter half-life than expected on the basis of the amount of locally delivered heparin (highest delivery amount and paradoxically shortest half-life). This was the only patient in the series who underwent a second PTCA (left circumflex artery, segment 12) distal to the drug-delivery site (performed after the first scintigraphic acquisition). This second procedure lasted 48 minutes and required stent implantation (Wallstent; 23×5 mm). The drug-delivery site was crossed by two balloon passages before stenting, by the Wallstent delivery system, and by one further balloon passage to optimize stent deployment. Contrast injections, temporary balloon occlusions of the artery followed by reactive hyperemia, and mechanical friction of plastic surfaces at the site of local drug delivery may have considerably altered the washout kinetics and may explain the shorter half-life. Practically, site-specific administration should be performed as a last intervention to avoid any artificial shortening of retention time.
Limitations of Radioisotopic Study
Nonspecific myocardial uptake. Ischemic myocardium can be a substrate for nonspecific uptake of 99mTc-labeled heparin in vivo (canine animal model) and could potentially influence the quantification of pharmacokinetics of transvascularly delivered 99mTc-labeled heparin.39 To evaluate nonspecific myocardial uptake, we administered 37 MBq of 99mTc-labeled heparin to three patients 15 minutes after angioplasty of a left anterior descending artery lesion (in our series, the time period until the start of site-specific heparin infusion was 26±3 minutes). No nonspecific myocardial uptake was observed up to 8 hours after administration of the radiolabeled drug (data not shown). Thus, transient occlusion with chest pain and documented ECG signs of myocardial ischemia in the anterior wall did not result in nonspecific myocardial uptake. However, myocardial ischemia could also be induced during local drug infusion as a result of side-branch occlusion and not primarily as a consequence of a lack of distal perfusion of the device.22 In our series, two patients (patients 4 and 8) presented signs of myocardial ischemia during site-specific drug administration (infusion time <30 minutes). Regional heparin bioavailability in these two patients was not significantly different than in the other six patients (151±31 versus 276±218 IU/h; P=NS). Thus, even though nonspecific myocardial uptake was observed in nonanticoagulated dogs after transient coronary ligation (40 to 80 minutes), these results seem unlikely to be applicable to fully anticoagulated patients in a postangioplasty and/or post–site-specific heparin-delivery situation.
Spatial extension of regional heparin delivery. According to the number of pixels of the region of interest, the area of delivery ranges from 5 to 20 cm2, confirming that “local delivery” is not confined to intramural delivery but has to be viewed as regional delivery occurring in adjacent tissues (perivascular tissue, myocardium, and pericardial space). In this regard, a recent histological report19 demonstrated transmural delivery of nanoparticles in the perivascular fat and musculature of rabbit atherosclerotic arteries.
Limitations of the scintigraphic evaluation of spatial extension of delivery can result from three aspects of the technique:
(1) Projection: Depending on the site treated, the LAO 45° planar projection may result in an orthogonal view or, in the extreme, a frontal view of the site of delivery, creating a projection-linked variability in size.
(2) Motion artifact of the area of interest: Systolic-diastolic heart movement and respiratory excursion have a tendency to increase the size of the “hot spot.” Acquisition in a gated modality could potentially eliminate interference due to heart motion. Unfortunately, the amount of energy used would not permit the acquisition of an image of sufficient quality. Respiratory interference remains a confounding factor.
(3) Three-dimensional extension of the area of interest: Planar acquisition does not permit assessment of a three-dimensional extension of the “hot spot.” Acquiring images with a three-head, rotating γ-camera makes three-dimensional reconstruction possible. However, this could be performed only outside of the catheterization suite during the follow-up phase. The focus of the study was to assess kinetics using evolution in time of regional counts, and therefore any change in acquisition modality was avoided.
Stability of 99mTc-labeled heparin. The prepared 99mTc-labeled heparin was confirmed to be stable in vitro over time up to 24 hours after preparation, indicating the use of a reliable product during the experiment. The thyroids of the patients were not imaged with the γ-camera for ≤8 hours after administration of the radiolabeled heparin, indicating no circulating free pertechnetate in vivo at the time of the measurements (free pertechnetate is a radiopharmaceutical for routine thyroid scintigraphy).
Physical half-life of the radionuclide. Calculated biological half-life can be unreliable when biological half-life is longer than physical half-life of the radionuclide. For this reason, it is of utmost importance to correct for the physical half-life of the tracer when biological half-life and retention time of the drug are calculated. An alternative would be to use radionuclides with a longer physical half-life (eg, 111In [36.44 hours]).
Imaging interference. Heparin is metabolized in both the endothelium and the reticuloendothelial system, particularly in the liver and spleen, and is eliminated via the kidney.33 40 After <30 minutes, the liver and spleen become visible scintigraphically (Fig 6⇑). Superimposition of the liver and spleen at the site of intracoronary delivery may occur, which can interfere with the definition of the region of interest. This may be a problem, particularly for the right coronary artery (segment 2 and proximal 3, depending on its relationship to the position of the heart and the diaphragm). Angulation of the γ-camera head is limited to a range of 20° to 90° due to steric hindrance of the roentgenography equipment (amplifier or tube).
Interval of scintigraphic acquisitions. To allow an appropriate mathematical fitting of the data points, the time interval between the catheterization laboratory acquisition and the next acquisition should not exceed 90 minutes. When the procedure after drug delivery is prolonged because of an unsatisfactory final result, the scintigraphic acquisition may be delayed. This delay may be of such an extent that the iterative fitting of the data points can be compromised.
In conclusion, after site-specific intracoronary delivery of a γ-emitting drug after angioplasty, regional persistence of the delivered compound was proved in humans. RIT-PK allows on-line visualization and off-line determination of regional pharmacokinetics. Pharmacokinetic parameters such as peak delivered amount, regional bioavailability, and half-life and retention time were calculated for heparin. Further studies are needed to define the relationship between regional efficiency (pharmacokinetics of site-specifically administered heparin) and clinical efficacy (clinical events such as acute thrombosis or long-term restenosis).
Selected Abbreviations and Acronyms
|LAO||=||left anterior oblique|
|MLD||=||minimal luminal diameter|
|PTCA||=||percutaneous transluminal coronary angioplasty|
|RAO||=||right anterior oblique|
|RIT-PK||=||radioisotopic technique to determine pharmacokinetics|
Correction Factors and Calculation of Pharmacokinetic Parameters
Correction factors are derived from physical laws and applied to measured counts to calculate “corrected” counts to express a closer approximation of locally deposited, radiolabeled heparin.
Regionally measured counts were corrected for (1) background activity, (2) body attenuation, and (3) 99mTc decay rate.
1. Background activity: Background counts represent nonspecific, nonregional activity. Background counts were measured in an area of the heart adjacent to the region of interest. This background activity takes into account both mixed blood-pool counts (AV blood) and counts of perfused myocardium. Background correction is performed by subtraction of background counts per pixel from regional counts per pixel for each acquisition frame.
2. Body attenuation: Attenuation corresponds to soft-tissue absorption of a radiating source. Attenuation is calculated according to the formula where I (in counts per second) is the intensity after absorption or measured intensity, I0 is the primary intensity or intensity at the site of delivery, μ (per centimeter) is the linear attenuation coefficient (0.16 cm−1 for 99mTc in water and assuming body attenuation as in water), d (in centimeters) is the thickness (angioplasty site–to-skin distance in LAO 45° projection corresponding to the angulation of the γ-camera acquisition head), and e is the base of the system of the natural logarithm (approximate numerical value=2.72). Thus, primary intensity (I0) is dependent on the parameter e−μd.
Angioplasty site–to-skin distance or soft-tissue thickness (d) is measured by fixing a radiopaque glass marble to the chest of the patient in alignment with the PTCA balloon marker positioned at the angioplasty site. The exact alignment of the marble and the balloon marker in the center of the frame is performed under fluoroscopy, in an LAO 45° projection. The balloon marker and marble are then filmed in the perpendicular projection (45° RAO) with the lowest amplification field necessary to get the balloon marker and marble on the same film frame during one cardiac cycle. The off-line projection (Tagarno 35 CX, Tagarno AS; focal distance, 45 cm) of this frame allows for determination of the angioplasty site–to-skin distance (dx) from the equation or where Dr is the actual marble diameter, Dm the marble diameter measured on the projection, dx the unknown angioplasty site–to-skin distance, and dm the balloon marker–to-marble distance measured on the projection.
On-line, the visual intercept of systolic-diastolic movement and respiratory excursions of the balloon marker was aligned with the center of the marble in the LAO 45° projection. Off-line, the balloon marker–to-marble distance (dm) was determined from the RAO 45° projection using the median point of systolic-diastolic movement and respiratory excursions and the closest point of the marble surface.
3. 99mTc decay rate: Decay of 99mTc was calculated according to the formula where At and A0 (becquerels=disintegration/s) is activity at time t and t0, respectively, and τ is the physical half-life of 99mTc (6 hours).
Calculation of Pharmacokinetic Parameters
The calculated pharmacokinetic parameters for site-specifically administered heparin were (1) peak amount, (2) half-life and retention time, and (3) regional bioavailability.
1. Peak heparin amount: Peak regional heparin amount is calculated as follows:
(A) Using the flat exponential component (elimination component) of the washout curve (y=a×e−bx), time point x=0 (beginning of the washout phase=end of the infusion phase) defines peak regional heparin amount (y0). Thus, The flat exponential curve is obtained from the biexponential fitting of the uncorrected counts per pixel over time of the washout phase.
(B) The obtained value is corrected by the three correction factors: background, attenuation, and decay.
(C) The corrected regional counts per pixel are multiplied by the number of pixels in the region of interest to obtain the absolute regional counts.
(D) The ratio of the corrected, regional counts over the infused counts yields the proportion of regionally delivered heparin to totally infused heparin.
The infused counts are obtained using the activity of the standard normalized for infused volume and for acquisition time (180 seconds’ static acquisition versus 40 s/frame dynamic acquisition) and corrected by 99mTc decay (activity at start of washout).
(E) The obtained ratio multiplied by the infused amount of heparin (in international units) results in the regionally delivered heparin amount. To convert from international units to milligrams, a conversion factor of 0.00524 was applied (1 mg=190.9 IU).
Peak regional heparin amount (in percent) was obtained by dividing regional heparin amount (in international units or milligrams) by the totally infused amount of heparin (in international units or milligrams).
2. Half-life and retention time: Half-life was calculated by means of the elimination component (y=a×e−bx) of the washout curve corrected by 99mTc decay.
The pharmacokinetic formula is Ct=C0×e−kt, where Ct and C0 are the concentrations at time points t and 0, respectively, k is the elimination constant, and t is the time period. Translated to the scintigraphic approach, Ct and C0 correspond to the counts per pixel corrected by 99mTc decay at time points t and 0, respectively, where C0=a and Ct represents the measured counts per pixel.
According to the definition of biological half-life According to the formula for the elimination component Retention time ) was obtained according to the definition 3. Regional bioavailability: Regional bioavailability (AUCtot) of heparin was calculated as the sum of the areas under the curve of the linear accumulation phase (AUClin) and biexponential washout phase (AUCexp1 and AUCexp2) using the corrected counts per pixel curve (corrected for 99mTc decay, background, and attenuation). To convert counts per pixel to heparin amount, the same procedural steps (C-E) were followed as for the calculation of peak heparin amount.
AUC is expressed in either IU·h (mg·h) or IU·mL−1·h (mg·mL−1·h).
Drug-delivery catheters were graciously provided by SCIMED Systems Inc. This project was supported in part by a grant from The Netherlands Heart Foundation (No. 95087) and from the Glaxo-Wellcome Co. Dr Edoardo Camenzind is the recipient of the Swiss Conrad Gessner grant and fellowship award from Ciba-Geigy, Basel, Switzerland. We gratefully acknowledge Jürgen Ligthart and Roel de Ruiter and all personnel of the catheterization laboratory for excellent technical assistance and Marcel van der Pluÿm for his technical expertise in preparing the radiopharmaceutical. We are indebted to Paolo Fioretti, MD, for technical advice and to Peter Paul Kint and Maud van Nierop for preparing the figures.
Presented in part at the 44th Scientific Sessions of the American College of Cardiology, New Orleans, La, March 19-23, 1995, and at the 68th Scientific Sessions of the American Heart Association, Anaheim, Calif, November 13-16, 1995, and published in abstract form (Circulation. 1995;92[suppl I]:I-785).
- Received November 14, 1996.
- Revision received December 17, 1996.
- Accepted January 9, 1997.
- Copyright © 1997 by American Heart Association
Serruys PW, Luijten HE, Beatt KJ, Geuskens R, de Feyter PJ, van den Brand M, Reiber JHC, ten Katen HJ, van Es GA, Hugenholtz PG. Incidence of restenosis after successful coronary angioplasty: a quantitative angiographic study in 342 consecutive patients at 1, 2, 3, and 4 months. Circulation. 1988;77:361-371.
Schwartz RS, Holmes DR, Topol EJ. The restenosis paradigm revisited: an alternative proposal for cellular mechanisms. J Am Coll Cardiol. 1992;84:1426-1436.
Isner JM. Vascular remodeling: honey, I think I shrunk the artery. Circulation. 1994;89:2937-2941.
Topol EJ, Califf RM, Weisman HF, Ellis SG, Tcheng JE, Worley S, Ivanhoe R, George BS, Fintel D, Weston M, Sigmon K, Anderson KM, Lee K, Willerson JT, on behalf of the EPIC Investigators. Randomised trial of coronary intervention with antibody against platelet IIb/IIIa integrin for reduction of clinical restenosis: results at six months. Lancet. 1994;343:881-886.
Wolinsky H, Thung SN. Local introduction of drugs into the arterial wall: a percutaneous catheter technique. J Intervent Cardiol. 1989;2:219-228.
Fernandez-Ortiz A, Meyer BJ, Mailhac A, Chesebro JH, Badimon L, Hassinger N, Owen WG, Fuster V, Badimon JJ. Intravascular local delivery: an iontophoretic approach. Circulation. 1994;89:1518-1522.
Azrin MA, Mitchel JF, Fram DB, Pedersen CA, Cartun RW, Barry JJ, Bow LM, Waters DD, McKay RG. Decreased platelet deposition and smooth cell proliferation after intramural heparin delivery with hydrogel-coated balloons. Circulation. 1994;90:433-441.
Thomas CN, Robinson KA, Cipolla GD, Jones M, King SB III, Scott NA. In-vivo local delivery of heparin to coronary arteries with a microporous infusion catheter. J Am Coll Cardiol. 1994;23:187A. Abstract.
Fram DB, Mitchel JF, Chow MSS, McKay RG. Pharmacokinetic modelling of intramural heparin washout following local delivery in the porcine coronary artery injury model. Circulation. 1995;92(suppl I):I-784. Abstract.
Mitchel JF, Fram DB, Palme DF II, Foster R, Hirst JA, Azrin MA, Bow LM, Eldin AM, Waters DD, McKay RG. Enhanced intracoronary thrombolysis with urokinase using a novel, local drug delivery system: in vitro, in vivo, and clinical studies. Circulation. 1995;91:785-793.
Camenzind E, Kint PP, Di Mario C, Ligthart J, van der Giessen W, Boersma E, Serruys PW. Intracoronary heparin delivery in humans: acute feasibility and long-term results. Circulation. 1995;92:2463-2472.
Ellis SG, Vandormael MG, Cowley MJ, DiSciascio G, Deligonul U, Topol E, Bulle TM, and the Multivessel Angioplasty Prognosis Study Group. Coronary morphologic and clinical determinants of procedural outcome with angioplasty for multivessel coronary disease. Circulation. 1990;82:1193-1202.
Reiber JHC, Serruys PW, Kooijman CJ, Wijns W, Slager CJ, Gerbrands JJ, Schuurbiers JCH, Den Boer A, Hugenholtz PG. Assessment of short-, medium-, and long-term variations in arterial dimensions from computer-assisted quantification of coronary cineangiograms. Circulation. 1985;71:280-288.
Chosbanian AV, Menzoian JO, Shipman J, Heath K, Haudenschild CC. Effects of endothelial denudation and cholesterol feeding on in vivo transport of albumin, glucose, and water across rabbit carotid artery. Circ Res. 1983;53:805-814.
Hoover RL, Rosenberg R, Haering W, Karnovsky MJ. Inhibition of rat arterial smooth muscle cell proliferation by heparin, II: in vitro studies. Circ Res. 1980;47:578-583.
Guyton JR, Rosenberg RD, Clowes AW, Karnovsky MJ. Inhibition of rat arterial smooth muscle cell proliferation by heparin: in vivo studies with anticoagulant and nonanticoagulant heparin. Circ Res. 1980;46:625-634.
Mayerus PW, Broze GJ, Miletich JP, Tollefsen DM. Anticoagulant, thrombolytic and antiplatelet drugs. In: Goodman Gilman A, Kall TW, Nies AS, Taylor P, eds. The Pharmacological Basis of Therapeutics. Elmsford, NY: Pergamon Press; 1990:1311-1317.
Fram DB, Mitchel JF, Eldin AM, Waters DD, Norenberg FW, McKay RG. Intramural delivery of 3H-heparin with a new site-specific drug delivery system: the D3 catheter. J Am Coll Cardiol. 1994;23:186A. Abstract.
Thomas CN, Robinson KA, Cipolla GD, Jones M, King SB III, Scott NA. In-vivo local delivery of heparin to coronary arteries with a microporous infusion catheter. J Am Coll Cardiol. 1994;23:187A. Abstract.
Fram DB, Aretz T, Azrin MA, Mitchel JF, Samady H, Gillam LD, Sahatjian R, Waters D, McKay RG. Localized intramural drug delivery during balloon angioplasty using hydrogel-coated balloons and pressure-augmented diffusion. J Am Coll Cardiol. 1994;23:1570-1577.
Kulkarni PV, Parkey RW, Buja LM, Wilson JE III, Bonte FJ, Willerson JT. Technetium-labeled heparin: preliminary report of a new radiopharmaceutical with potential for imaging damaged coronary arteries and myocardium. J Nucl Med. 1978;19:810-815.
Boneau B, Caranobe C, Saivin S, Dol F, Sié P. Pharmacokinetics of heparin and of dermatan sulfate: clinical implications. In: Lane DA, ed. Heparin and Related Polysaccharides. New York, NY: Plenum Press; 1992:237-247.