Intrarenal drug infusion plays an important role in renal experimental research. Laminar flow of the blood can cause streaming and inhomogeneous intrarenal distribution of infused drugs. We suggest a simple method to achieve a homogeneous intravascular distribution of drugs infused into the renal artery of anesthetized rats. The method employs a multiple sidehole catheter inserted into the renal artery, which enables an efficient drug mixing with the arterial blood. To verify the efficiency of this method, we use laser speckle imaging and renal artery flowmetry. The results show that, compared with the conventional single-hole catheter, the multiple sidehole catheter provides a more uniform drug distribution and a homogenous vascular response on the surface of the kidney.
- blood flow imagery
- drug delivery
- laser speckle
- renal hemodynamics
intraorgan drug infusion plays a significant role in physiological research and, in particular, in renal physiology (12, 16, 20). A common approach to minimize systemic effects of the infused drugs and to activate specific renal mechanisms is to infuse drugs into the renal artery in vivo (6). An infused drug, however, is transported by laminar blood flow in the renal artery to divergent parts of the kidney depending on the position of the cannula tip of the catheter as well as the branching topology of the renal vascular tree. Consequently, some parts of the kidney are either not affected or receive lower drug concentration (18) whereas other parts receive a higher drug concentration. This issue has been addressed in clinical research with regard to drug delivery to target organs (2, 4, 5, 13), e.g., intra-arterial delivery of chemotherapy. Uneven drug distribution may also lead to a high variability of experimental results and to misinterpretation of the data. Furthermore, the uncertainty regarding local renal drug concentrations complicates experiments where concentration-specific inhibitors or activators are used.
The need to ensure adequate distribution of infused drugs to the kidney has been recognized and addressed before (3, 7, 17). Parekh (17) used a magnetic pump along with heparin infusion to disturb laminar flow and to improve drug distribution while Hamza and Kaufman (7) employed mechanical vibration of a syringe. Both of these approaches may interfere with physiological conditions by infusing heparin together with the drug into the blood (17) and by possible arterial tissue disturbance/irritation and uncontrolled vibrations (7). In addition, they require additional experimental equipment such as a pump and a vibrator.
In the present paper, we proposed the use of a catheter with multiple sideholes instead of the conventional catheter with one hole at the tip of the catheter. We designed a series of experiments to monitor the efficiency of drug distribution and concomitant vascular responses. For this purpose we applied laser speckle flowmetry in addition to traditional renal blood flow (RBF) measurements utilizing an ultrasonic flow probe. The laser speckle method has been widely used in renal research in the last few years (8, 22) and has been validated for quantification of inhomogeneous drug distribution (18).
Experimental protocols were approved by the Danish National Animal Experiments Inspectorate and conducted in accordance with guidelines of the American Physiological Society. Experiments were performed in 10 male Sprague-Dawley rats with body weight 250–350 g, purchased from Taconic (Lille Skensved, Denmark).
Before the surgery the rats were anesthetized with sevourane (8%). During the surgery sevourane concentration was reduced to a final concentration of ≈2%. The mean arterial pressure was measured with a Statham P23-dB pressure transducer (Gould, Oxnard, CA) connected to a carotid artery catheter. Two PE-10 catheters were inserted in the right jugular vein allowing continuous systemic infusion. After tracheotomy, the rat was placed on a servo-controlled heating table maintaining body temperature at 37°C. The rat was connected to a mechanical animal ventilator (60 breaths/min; 8 ml/kg body wt). Nimbex (muscle relaxant) with a concentration of 0.85 mg/ml was administered, first as bolus injection of 0.5 ml, followed by continuous infusion into the jugular vein at a rate of 20 μl/min. The left kidney was exposed and an ultrasonic blood flow probe (1PRB; Transonic T 420) was used to measure blood flow in the renal artery. The left ureter was catheterized (PE-10 connected to PE-50) to ensure free urine flow. The MoorFLPI imaging system (Moor Instruments, Millwey, Axminster, UK) was used to record changes in blood perfusion of the renal surface. The MoorFLPI camera acquired laser speckle images of 768 × 576 pixels (spatial resolution) at the rate of 1 frame/s (temporal resolution). To perform intrarenal infusions, a PE-10 catheter was inserted into the left renal artery, led there through the left iliac artery and the abdominal aorta. A schematic representation of the experimental setup is shown in Fig. 1A. Surgery was considered successful if the mean arterial pressure was stabilized at 90–130 mmHg and the renal blood flow was above 5 ml/min.
To induce a drug response we infused the vasoconstrictor angiotensin II (ANG II) for 15 s at a concentration of 4 ng/ml and an infusion rate of 144 μl/min. To establish a flow baseline, saline was infused with the same rate for 5 min before the ANG II infusion. We performed three series of experiments: 1) control experiments with the conventional PE-10 catheter inserted into the left renal artery (control group); 2) experiments where the conventional catheter was replaced by the multiple sidehole PE-10 catheter inserted into the left renal artery of the same animal (multihole group I); and 3) experiments with the multiple hole PE-10 catheter inserted into the left renal artery at the beginning of experiment (multihole group II). Ten animals were used in the experiments: five rats for the control experiments combined with series I experiments (multihole group I) and five additional rats for series II experiments (multihole group II). To prepare the multiple hole catheters we sealed the endhole in the conventional PE-10 catheter tip and then made four symmetrically positioned side holes at the distance of ~500 μm from the tip. The endhole was sealed by heating the catheter, and side holes were made with a needle. To ensure side holes position, preparations were made under the microscope with 1-mm graph paper on the background. Catheters and possible flow streaming are presented in Fig. 1, B and C.
MoorFLPI imaging software performed laser speckle contrast analysis calculating laser speckle values (SV) as the output (19). Pixels of the image located outside the kidney or/and fat spots on the renal surface were not used for further processing. To describe superficial renal perfusion in response to drug infusion we used response mapping by calculating pixel-by-pixel relative difference:(1)where and represent mean SV values over control and response periods, respectively, at each pixel (x, y) of the image. Frames were acquired over 10 s before the start of the infusion for control measurements and over 10 s around the response peak in the blood flow probe data for response measurements. M takes values from 0 to 1.
To visualize a relative difference for the control experiments and multihole group I experiments, we calculated the density distribution, i.e., the number of pixels of the renal surface image that showed a certain response strength:(2)This characteristic is equivalent to a binned response strength normalized by the total number of events for each experiment. All M values (Eq. 1) were divided into 100 intervals ranging from −1 to 1. The interval with index i = 1, for example, corresponds to M values from −1 to −0.98. Ni represents a number of pixels with M values falling into the interval with index i. S(N) is the area under N curve. The density Di is plotted as the function of M center value in the corresponding interval with index i.
To demonstrate changes in the renal artery flow, the relative maximal flow drop was calculated as:(3)where Fc is the flow averaged over 1 s before the start of the infusion and Fr is the flow at the moment of the strongest flow response. Before analysis the flow probe data were smoothed using moving average filter with the window size of one second.
Based on the normality of data distributions, we used paired and unpaired t-test to evaluate statistically significant difference between the groups. P < 0.05 were considered statistically significant.
An example of the relative difference mapping for the control group (conventional catheter) and the multihole group I for the same animal is presented in Fig. 2, A and B. Laser speckle imaging of the renal surface shows an increase in the size of the affected area. Moreover, the strength of the vascular response increases when the multihole catheter is used. Measurements of blood flow in the renal artery by means of the ultrasonic probe also detect a larger maximal decrease in RBF for the multihole catheter (9.4 ml/min) compared with the conventional catheter (3.3 ml/min) (Fig. 2, C and D). While this example shows the best improvement, the values vary between animals (see Table 1).
To highlight the difference between the effect of intrarenal infusion through the conventional catheter and the multihole catheter we calculated density distribution (Eq. 2) that defines a fraction of the renal surface that responds differently to the drug administration. Figure 3 displays such distributions for each animal. The distribution with a single peak of high magnitude indicates that the response is close to homogeneous. Multiple peaks of lower magnitude or at plateau indicate that response strength differs across the renal surface, i.e., heterogeneous drug response. A distribution shift to the left (M = 0) or to the right (M = 1) corresponds to weak or strong response, respectively. One can see that responses to the infusion via the multihole catheter (gray filled bins) are stronger and more evenly distributed. It is important to note that the shape of density distributions for the multihole catheter is similar in most cases (4 out of 5 in both multihole groups) while for the conventional catheter (white filled bins) there is a greater variation. Such variations point to high variability in the strength of response and in the size of the affected area.
To ensure proper physiological conditions, to avoid irritation of the arterial tissue, and to exclude other effects of the replacement procedure, we performed a set of experiments using only the multiple sidehole catheter inserted into the renal artery (multihole group II). Results obtained in this series of experiments are similar to the results obtained in the multihole I group (Table 1). A catheter with multiple sideholes ensures a significantly larger area of the strong response and significantly stronger relative decrease of flow in the renal artery (P < 0.05). As a condition to determine the size of the area of strong response we used a relative drop of SV values of 50% (M > 0.5). It is also important to note that the deviation in the form of density distribution of response strength for the multihole groups is less compared with the experiments with the conventional catheter. This indicates better repeatability of experiments. Details of the individual experiments are summarized in Fig. 4. Each point corresponds to a relative blood flow drop (Fig. 4A) and an area of significant response (Fig. 4B) measured in a single experiment. For both characteristics, the control group demonstrates the weakest response (diamond symbol). Use of the conventional catheter results in the maximal decrease of flow of ~3.26 times. The multihole groups I (circle) and II (square) demonstrate decrease of flow by a factor of 7.19 and 5.67, respectively. For the two experimental groups with the multiple sidehole catheter, the size of the area with strong response is significantly larger compared with the control group.
Streaming in the arterial blood flow is an important issue in the context of delivering drugs to target organs, since uneven drug distribution may lead to toxicity and reduced treatment efficiency (1, 2, 4, 5, 13). Streaming often arises at slow infusion rates where laminar bands of the infused drug primarily enter into downstream branches of the arterial vascular tree causing uneven distribution. The effect of streaming and the variability of drug delivery have been demonstrated for different arteries such as hepatic and cranial arteries (9, 11, 14, 21). An important clinical consequence is uneven distribution of, e.g., chemotherapy compounds. One possibility to resolve this problem is to increase the infusion rate (10). Infusion rates corresponding to 20% of the initial organ flow have been shown to reduce streaming significantly (2). However, in the renal circulation this poses a particular problem. Many physiological studies examining renal effects of drugs also focus on urinary excretion. Increasing renal perfusion 20% will in itself elevate the renal excretion rate. Earlier Parekh (17) and Hamza and Kaufman (7) discussed the streaming problem and possible improvements for intrarenal drug infusion. Importance of even drug distribution in renal research was also emphasized by Cupples and colleagues (3, 15). Uneven drug distribution results in increased variability of experimental results and possibly in misinterpretation of the obtained results. Using laser speckle imaging we showed that uneven drug distribution during continuous infusion of ANG II can affect blood flow dynamics (18). Furthermore, when the average RBF was measured, it is always uncertain how much of the renal parenchyma is affected by the infused drug. As shown in the present control experiments utilizing the conventional catheter, less than half of the surface area of the kidney is affected by the infused ANG II. In experiments using, e.g., inhibitors of ion channels, where the blocking effect is dependent on concentrations, interpretation of results gets even more confusing as there is no way of knowing what the local renal concentration actually is.
Parekh (17) proposed the use of a magnetic pump together with multiple catheters to draw blood back to mix with the drug before being reinfused into the animal. Such modification led to a stronger reaction of renal blood flow to vasoconstrictors and dilators and to more homogeneous distribution of the infused dye. However, there are several disadvantages in such system: 1) pumping blood back and forth can result in the catheter getting clogged. To avoid this Parekh suggested infusing heparin together with the drug. During long-term perfusion experiments, this method involves unfavorable effects of heparin; 2) an experimental setup incorporating a multicatheter system and a magnetic pump is, in addition to its higher cost, more complicated to handle.
Hamza and Kaufman (7) suggested a somewhat simpler solution to the problem. By connecting the syringe to an element providing mechanical vibrations they achieved a stronger blood flow decrease in response to phenylephrine injection and better dye distribution recorded along the renal surface. However, this method also has a few significant drawbacks: 1) it requires a specific approach to cannulate the renal artery by puncturing it with a syringe instead of inserting a polyethylene catheter through the aorta; 2) vibrations spread to the inserted syringe are hard to control; 3) according to the protocol, the system was tested with vibrations only for 60 s, and it is unclear whether prolonged vibration would lead to catheter-syringe disconnection from the renal artery; and 4) puncturing and vibration of the syringe may affect the renal artery, thus, influencing the flow and results of long-term perfusion experiments.
In the present work we demonstrated that the use of a multiple sidehole catheters greatly improves the strength of response and homogeneity of drug distribution in the renal vasculature. We monitored the effect of infused ANG II on the cortical renal blood flow by means of laser speckle flowmetry. Also, the effect on total renal blood flow measured in the renal artery by means of ultrasonic flowmetry was assessed. Compared with other methods, our method 1) does not require any special equipment except a multiple sidehole catheter; 2) allows for conventional cannulation technique through the aorta; and 3) maintains a stable drug flow. Improved mixing of blood with the drug is achieved by fewer holes and fewer diverse directions of drug streaming into the artery. We showed that the relative decrease of flow in response to ANG II infusion increased from 3.26 for the conventional catheter to 7.19 or 5.67 for the mutlihole catheter experiment groups, i.e., at least 73% stronger response. This is in agreement with the results reported by Parekh (17), although we observed a stronger blood flow drop because of larger amount of the infused drug.
Laser speckle imaging allowed additional important quantification of the drug response. Instead of photographing a dye distribution across the renal surface, we assessed direct effect of the infused drug as changes of cortical blood flow. The relative difference mapping demonstrated that the portion the kidney with strong response (M > 0:5) increased up to ≈63% with smaller deviations when the multiple sidehole catheters were used compared with ≈23% for the conventional catheters. Our experiments were performed with the vasoconstrictor ANG II but similar experiments can be conducted for any drug affecting renal blood flow.
In conclusion, using laser speckle imaging and ultrasonic flowmetry we demonstrated that the use of a simple multiple sidehole catheter significantly improves the strength of the renal vascular responses obtained after intrarenal infusion of vasoactive compounds. Also, homogeneity of the drug distribution throughout the renal vasculature is greatly increased. Thus, we propose that the properties of this simple and easy to use method is comparable or superior to more complicated and costly methods previously suggested to obtain more homogenous intrarenal infusion.
No conflicts of interest, financial or otherwise, are declared by the author(s).
D.D.P. conceived and designed research; D.D.P. performed experiments; D.D.P. analyzed data; D.D.P., M.S., C.M.S., and O.V.S. interpreted results of experiments; D.D.P. prepared figures; D.D.P. drafted manuscript; D.D.P., M.S., C.M.S., and O.V.S. edited and revised manuscript; D.D.P., M.S., C.M.S., and O.V.S. approved final version of manuscript.
This work is part of the Dynamical Systems Interdisciplinary Network, University of Copenhagen.
- Copyright © 2017 the American Physiological Society
Please sign in below with your personal user name and password.