Introduction

Top Abstract Introduction Materials and Methods Results Discussion Conclusion References

In the pharmacological treatment of heart diseases, one of the problems to overcome is to achieve satisfactory concentrations of therapeutic agents at the target site. When agents are applied systemically, the potential risk of side effects exists and therapeutic efficacy may be low, due to metabolism and tissue binding of the agent. In addition, delivery of systemically administered substances to the heart or coronary circulation may be hampered by diminished local blood supply, e.g., in occlusive coronary artery disease. Higher therapeutic efficiencies and smaller risks of peripheral side effects are anticipated if agents are administered locally. This issue has become a matter of intense debate with recent attempts to induce myocardial angiogenesis by targeting genes or gene-derived products to the heart (Kornowski et al., 2000).

Various approaches exist to deliver agents to the heart, such as intracoronary, intramyocardial, and transvascular application or local installation of slow releasing polymers (Sellke and Simons, 1999; Kornowski et al., 2000). An alternative approach is administration of agents into the pericardial space. This approach has been successfully applied for a number of agents (see Spodick, 2000 for a brief overview). For instance, in animal models of cardiac ischemia, intrapericardially applied FGF-2 was shown to induce myocardial angiogenesis (Landau et al., 1995; Uchida et al., 1995), to increase collaterals and to improve coronary blood flow and perfusion of the ischemic heart region (Laham et al., 2000). Intrapericardially applied nitric oxide-donors (Willerson et al., 1996; Waxman et al., 1999) and antiarrhythmic agents (Ayers et al., 1996) have been shown to elicit clear effects on heart function or heart circulation in animals. In humans, pericardial effusions have been successfully treated by intrapericardially applied chemotherapeutics (Kohnoe et al., 1994) and glucocorticoids (Buselmeier et al., 1978). Finally, genes have been transferred to heart tissue with high efficiency, by applying gene-containing vectors into the pericardial space of animals (Lazarous et al., 1999; March et al., 1999; Zhang et al., 1999).

These pharmacodynamic studies suggest that intrapericardial drug application may represent a promising strategy for site-specific drug targeting to the heart. Also, from a pharmacokinetic point of view, there is reason to assume that this is the case. Since the pericardial space is a small fluid-filled closed compartment, facing the heart at one side (Spodick, 1992), it may comprise an ideal compartment for local drug delivery (Smits and Thijssen, 1986; Daemen et al., 1988). This assumption is supported by Lazarous et al. (1997), demonstrating that intrapericardial application was by far the most efficient way to deliver FGF-2 to the heart. They showed that 19% of the intrapericardially applied FGF-2 was recovered in the heart versus only 0.5% after intravenous, 1.3% after left atrial, and 3 to 5% after intracoronary administration. Nevertheless, knowledge of the pharmacokinetics of intrapericardially applied compounds is still rather limited. In particular, it is not yet known whether a potential advantage of local intrapericardial delivery can be maintained over a prolonged period of time. Therefore, in the present study, various substances were applied by intrapericardial or systemic bolus injections (fluorescent macromolecules) as well as by 7 days of infusions (fluorescent macromolecules and steroids). Pharmacokinetic profiles were determined in pericardial fluid and plasma of conscious rats instrumented with intravascular and intrapericardial catheters. Test substances were chosen because of their variable sizes and their potential therapeutic applicability (in the case of heparin, FGF-2, and cortisol) and included the fluorescent macromolecules FITC-rat IgG (155 kDa; assessed by SDS-polyacrylamide gel electrophoresis), Texas Red-RSA (67 kDa), Texas Red-FGF-2 (18 kDa), and (unfractionated) FITC-heparin (mean mol. wt., 18 kDa), the small steroid cortisol (mol. wt. = 362.5), and a carbonic acid derivative thereof (mol. wt. = 348). Finally, rats were intrapericardially or systemically infused for 7 days with ¹²⁵I-labeled FGF-2 to assess cardiac tissue penetration by autoradiography.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Conclusion References

Animals. For the experiments, male Wistar rats (Iffa Credo, Someren, the Netherlands) were used, ranging in weight between 280 and 350 g. Animals were housed at the animal facilities of the University of Maastricht with a 12-h light/dark cycle and had free access to standard rat chow and tap water. Experiments were performed according to institutional guidelines and have been approved by the local ethical committee for the use of experimental animals.

Surgical Procedure for Installing a Pericardial Catheter. Construction of the pericardial catheter and the procedure to install the catheter into the pericardial space was conducted using methods adapted from Veelken et al. (1990) and McDermott et al. (1995). The pericardial catheter consisted of silicone tubing (internal diameter: 0.51 mm; external diameter: 0.94 mm; Degania Silicone, Degania Bet, Israel), which, by assembling its endings with a polyolefin shrinking sleeve (Farnell Compounds, Maarssen, the Netherlands), was shaped as a loop (length ~2 cm, width ~1 cm). Silicone glue was applied in the middle of the loop, to create two separate chambers (for fluid injection and pericardial fluid withdrawal) that were provided with two and six holes, respectively, by use of a 25-gauge perforator. The endings of the silicone tubing were connected to polyethylene (PE-10) tubing (i.d. 0.28 mm, o.d. 0.61 mm; Portex Limited, Kent, UK). Before installment, the catheter and the polyethylene extensions were gas sterilized and filled with sterile 0.9% (w/v) NaCl. The extensions were plugged with stainless steel pins.

Preparation and Analyses of Compounds. Fluorescein-labeled heparin (unfractionated with a mean molecular weight of 18,000) was obtained from Molecular Probes (Eugene, OR). Rat serum albumin (Sigma-Aldrich, St Louis, MO) and human recombinant FGF-2 (Research Diagnostics, Flanders NJ) were labeled by reaction with the succinimidyl ester of Texas Red (Molecular Probes). Rat IgG (Sigma-Aldrich) was labeled with fluorescein by reaction of fluorescein isothiocyanate (Sigma-Aldrich). Labeling, reagent inactivation, and removal of noncovalently bound fluorescence (the latter resulting in <0.5% unbound fractions of the total fluorescence as assessed by dialysis and ultrafiltration), were conducted by standard procedures (see, e.g., Haugland, 1996).

Pharmacokinetic Studies after Bolus Injection. Immediately after implantation of the pericardial catheter, rats (still under anesthesia) were provided with a catheter in the right femoral artery essentially as described (Smits et al., 1982). Rats were allowed to recover at least 2 days before experimentation. One hour before the start of the experiment, 20 µl of pericardial fluid were withdrawn using a Hamilton 1705 (Hamilton Bonaduz AG, Bonaduz, Switzerland) syringe, and 50 µl of saline were injected into the pericardial space to check the integrity of the pericardial catheter. Injections of volumes up to 0.2 ml were previously shown to be without hemodynamic effects (Veelken et al., 1990). Blood (0.15-0.25 ml) was collected in a syringe containing a minimal volume of heparin (Organon Teknika, Boxtel, the Netherlands). These samples served as blanks for later analyses. Experiments in which substances were applied intrapericardially were started by a 50-µl bolus injection of the test substances into the pericardial space, followed by 20 µl of saline to flush the catheter. If substances were applied systemically, experiments were started by a 100-µl bolus injection of the substances and subsequent injection of 300 µl of saline into the femoral artery catheter. FITC-rat IgG, (10 mg/ml), Texas Red-RSA (10 mg/ml), and FITC-heparin (1 mg/ml) were dissolved in PBS. Texas Red-FGF-2 (20 µg/ml) was dissolved in a 10 mg/ml solution of RSA in PBS.

t

t

Infusion Studies. Directly after installment of the pericardial catheter, still-anesthetized rats were provided with a catheter in the left jugular vein (Kleinjans et al., 1984). Rats were allowed to recover for 2 days before subcutaneous implantation (under ketamine/xylazine anesthesia) of osmotic minipumps (Alzet 2001; Alza, Palo Alto, CA). Minipumps filled with solutions of the substances to be tested were primed in saline at 37°C at least 4 h before connection to the catheter. Before installing pumps, pericardial fluid and orbital sinus blood were sampled, to serve as blanks. Seven days after pump installment, rats were sacrificed by exsanguination under pentobarbitone and pericardial fluid and blood collected. To check for possible loss of substances during infusion, the remaining pump contents were analyzed. No significant changes in the concentration of the substances in the infusion fluid were found after 1 week of pumping. Infusion rates of the substances were 10 µg/h for FITC-rat IgG and Texas Red-RSA, 20 ng/h for Texas Red-FGF-2, 100 ng/h for FITC-heparin, 684 ng/h for cortisol, and 984 ng/h for the side chain-modified acid analog of cortisol. Doses were chosen to achieve concentrations that were readily measurable but without pharmacological effects (risk of bleeding in the case of heparin); similar doses were applied systemically and intrapericardially to be able to make a good comparison between the two routes of administration. The solvent was PBS, except for Texas Red-FGF-2 and cortisol, which were dissolved in a 10 mg/ml solution of RSA in PBS.

Cardiac Tissue Penetration of Intrapericardially or Systemically Infused ¹²⁵I-FGF-2. Rats were infused with ¹²⁵I-labeled FGF-2 (PerkinElmer Life Sciences, Boston, MA) at a dose rate of 25 nCi/h, similar to that described above. After 7 days of infusion, rats (n = 2 for intravenous and n = 3 for intrapericardial infusion) were sacrificed to remove the hearts. Hearts were weighed and rinsed in saline for 5 min. At the center of the hearts, transverse 2-mm slices were cut, using a razor blade. Slices were rinsed in saline after fixation by 5 min submersion in a 10% (v/v) formaldehyde solution in PBS and 5 min in 70% (v/v) ethanol in water. Slices were used for autoradiography by phosphorimaging. Cardiac tissue concentrations of ¹²⁵I-FGF-2 were determined by use of a gamma counter.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Conclusion References

Bolus Injection Studies. Pericardial fluid concentration-time profiles of intrapericardially applied and plasma concentration-time profiles of systemically applied FITC-rat IgG, Texas Red-RSA, Texas Red-FGF-2, and FITC-heparin are shown in Fig. 1.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Concentration-time profiles of fluorescent macromolecules in rat pericardial fluid after intrapericardial bolus injection (top) or in plasma after intra-arterial bolus injection (bottom). Rats were given bolus injections of 50 µl intrapericardially or 100 µl intra-arterially, and pericardial fluid and blood samples were collected at various time points for analysis as described. Data in each graph (three to seven rats) represent concentrations, standardized for body weights [i.e., body weight (kg) × concentration].

View this table: [in this window] [in a new window]   TABLE 1 Pharmacokinetic parameters of fluorescent macromolecules Parameters were derived by fitting standardized data of every individual animal to the equation Ct = A · et + B · et of one (i.e., A = 0)- and two-compartment models and are expressed as mean ± S.E.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Ratios of fluorescence measured in pericardial fluid and plasma after intrapericardial () or intra-arterial bolus injections of fluorescent macromolecules. Rats were given bolus injections of 50 µl intrapericardially or 100 µl intra-arterially, and pericardial fluid and blood samples were collected at various time points and analyzed as described. Data in each graph represent three to six rats. For FITC-heparin, data are lacking for systemic administration because fluorescence of pericardial fluid could not be detected.

Infusion Studies. Pericardial fluid and plasma substance concentrations after 7 days of infusion into pericardial space or blood are given in Table 2. Based on pilot experiments in which concentrations were determined on a daily basis, as well as on terminal half-lives (Table 1), it seems reasonable to assume that after 7 days of infusion, steady state has been reached. Following continuous infusion of fluorescent macromolecules into pericardial space, concentrations in plasma are at least 7-fold lower than in pericardial fluid (Table 2). This is also the case for the small compounds cortisol and its carbonic acid analog (Table 2). In contrast, following continuous infusion of macromolecules into blood, pericardial fluid and plasma concentrations are approximately similar.

Studies on the Cardiac Tissue Penetration of ¹²⁵I-FGF-2. To obtain information about the penetration of intrapericardially versus systemically applied substances into cardiac tissue, rats were infused with ¹²⁵I-FGF-2 for 1 week into pericardial space or blood.

View larger version (81K): [in this window] [in a new window]   Fig. 3.   Autoradiograms of cardiac tissue slices of rats that received a 1-week infusion of 125-I-FGF-2 into pericardial space or blood.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Conclusion References

Various delivery strategies have been applied to target therapeutic agents to the diseased heart, such as local installation of slow releasing polymers and intramyocardial, intracoronary, transvascular, and intrapericardial applications (Sellke and Simons, 1999; Kornowski et al., 2000). The purpose of these strategies is to improve therapeutic efficiencies and to reduce the risks of side effects inherent to systemic application. The issue, which of the application methods is superior regarding safety, feasibility, and efficiency, is still not resolved (Sellke and Simons, 1999; Kornowski et al., 2000). In this study, we investigated whether intrapericardial application offers pharmacokinetic advantages, by comparing pharmacokinetic profiles of compounds that were intrapericardially and systemically applied by bolus injection or by continuous infusion.

The results show that after intrapericardial injection of macromolecules, pericardial fluid concentrations are substantially higher than plasma concentrations over a prolonged time (Fig. 2). Similarly, if macromolecules or steroids are infused into pericardial space for 7 days, this results in relatively high pericardial fluid concentrations and low plasma concentrations (Table 2). The ratio of local and systemic steady-state concentrations after local application is termed the local advantage. Thus, for the intrapericardially infused substances, local advantages range between 7 and 420 (Table 2). The magnitude of the local advantage depends on substance exchange between the local and the blood compartments and on substance elimination from these compartments. Assuming that the clearance of a substance from the pericardial space is solely due to diffusion into plasma (without differences between back- and forward diffusion rate constants), local advantage of intrapericardial drug application would be determined by the ratio of the systemic and local clearances (Smits and Thijssen, 1986). Although this assumption is oversimplified, data derived from steady-state concentrations (Results) or bolus injection data (Table 1) indicate that the local advantages of intrapericardially applied agents can be mainly explained by their low clearances in pericardial space compared with the relatively high plasma clearances. Whether particular physicochemical properties such as size and charge of substances play a role in the local advantage of intrapericardial application is at present unclear. Although the number of agents that were tested is too small to draw any firm conclusions, the data in Table 2 do not reveal a clear relationship between molecular size and the obtained local advantage. The relatively high local advantages of the negatively charged heparin and cortisol carbonic acid may indicate that charge does play a role. On the other hand, recent studies in our laboratory demonstrated a high local advantage (of 97) for the positively charged (small) molecule sotalol as well (data not shown). That there is no clear structure-advantage relationship is not surprising, if it is taken into account that one major determinant of the local advantage of a compound is its systemic clearance, for which no simple general relationship with the structure of the compound is known.

The high local advantages that are found imply that high local drug concentrations can be obtained, whereas drug concentrations in the plasma remain low if substances are applied intrapericardially. For that reason, intrapericardial application of drugs may be expected to lead to higher therapeutic efficiencies and lower risks of side effects. However, if agents do not exert their beneficial effects by interaction with epicardial surface receptors, but act deeper in myocardial tissue or in coronary vasculature, therapeutic efficacy will depend on local tissue concentrations rather than pericardial fluid concentrations. Therefore, we assessed the cardiac tissue penetration of ¹²⁵I-labeled FGF-2 by autoradiography, since determination of cardiac tissue levels of fluorescent macromolecules was hampered by autofluorescence of the rat hearts. ¹²⁵I-labeled FGF-2 was selected because of the potential use of (intrapericardially applied) FGF-2 to achieve therapeutic angiogenesis in the heart. This experiment not only showed that total cardiac tissue concentrations of ¹²⁵I-FGF-2 are 8-fold higher in intrapericardially versus systemically infused rats, but also that if the substance was infused intrapericardially, it appeared to be present at high concentrations in most parts of the rat heart (Fig. 3). The resolution of the autoradiograms that we obtained for ¹²⁵I-FGF-2 is too low to exactly determine regional differences (e.g., epicardial versus endocardial) in ¹²⁵I-FGF-2 levels. Nevertheless, the current experiments indicate that the intrapericardial infusion of FGF-2 really seems to be beneficial in that high cardiac tissue levels of the agent can be reached. In line with these observations, ongoing studies in our laboratory indicate that in a rat model, a dose of FGF-2 improves cardiac perfusion only when infused intrapericardially, not when infused systemically (our observations), pointing to high efficiency of the intrapericardial drug infusion. The above observations also correspond with findings by others (Lazarous et al., 1997; Stoll et al., 1998), comparing various administration routes to deliver FGF-2 by bolus injection. These authors found that the intrapericardial route was the most effective to obtain high cardiac FGF-2 levels.

That the concept of the high local advantage by intrapericardial drug application can indeed be translated into high therapeutic efficiencies (and low risk of peripheral side effects) has been demonstrated not only for FGF-2 but for other substances as well (e.g., Labhasetwar et al., 1994; Landau et al., 1995; Uchida et al., 1995; Ayers et al., 1996; Willerson et al., 1996; Waxman et al., 1999; Laham et al., 2000; Spodick, 2000). For instance, nitric oxide-donors are more effective in exerting regional effects, i.e., reduction of cyclic flow variations of mechanically injured coronary arteries (Willerson et al., 1996) or induction of coronary vasodilatation (Waxman et al., 1999), but produce smaller blood pressure decreases (systemic effects), when applied intrapericardially.

As already discussed, we found that if substances are infused into the pericardial space, their concentration in pericardial fluid will be relatively high, due to low clearances of the substances in the pericardial fluid compared with the higher clearances in the blood. In pericardial fluid, concentrations of various endogenous agents have been shown to exceed those in plasma. This is observed, for instance, for endothelin-1 (Horkay et al., 1995), FGF-2 (Corda et al., 1997), and atrial natriuretic peptide (Amano et al., 1993; Corda et al., 1997) in heart surgery patients. Also, in rats, concentrations of atrial natriuretic peptide in pericardial fluid are higher than in plasma (Klemola et al., 1995). The present study shows that these high pericardial fluid/plasma concentration ratios not only are a matter of local syntheses of the agents in the heart or in the mesothelial cells of the pericardium (Eid et al., 1994; Mebazaa et al., 1998), but can also be attributed to the relatively slow clearances of these agents, once "trapped" into pericardial space.

From the pharmacokinetic data, plasma volumes as well as pericardial volumes of the rats can be derived. Upon systemic bolus injections, central compartment volumes (V_c) (Table 1) range between 33 and 46 ml/kg. These volumes probably represent plasma volumes and are well within the range of previously determined plasma volumes of male Wistar rats (Lee and Blaufox, 1985; Wauquier and Devynck, 1989). Given the almost instantaneous mixing of injected compounds in pericardial space (Santamore et al., 1990), it seems reasonable to assume that calculated V_c values indicate pericardial fluid volume. Hence, pericardial fluid volumes of catheter-instrumented rats would be 0.5 to 0.9 ml/kg. Since the catheter will create additional space in the pericardial cavity (estimated to be about 0.1 ml), which presumably is fully filled with fluid, "real" rat pericardial fluid volumes will probably be 0.25 to 0.35 ml/kg lower. Previous studies showed that mean pericardial fluid volumes are 0.23 ml/kg in greyhounds (Gibson and Segal, 1978) and 0.29 ml/kg in mongrel dogs (Santamore et al., 1990). In non-heart-diseased humans, pericardial fluid volume ranges between 15 and 50 ml (Spodick, 1992), which at a body weight of 75 kg would correspond to 0.2 to 0.67 ml/kg.

	Conclusion

Top Abstract Introduction Materials and Methods Results Discussion Conclusion References

After intrapericardial application, high local drug concentrations can be obtained in the heart, whereas plasma drug concentrations remain low. This can be explained by the fact that the clearances of substances in pericardial fluid are low, relative to substance clearances in plasma. Because of this pharmacokinetic advantage, a desirable local drug concentration may be achieved at lower doses, while the potential risk of peripheral side effects is reduced by intrapericardial drug application. Therefore, intrapericardial application of therapeutic agents provides a promising tool to obtain site-specific treatment of heart or coronary diseases.