The Isolated Blood And Perfusion fluid Perfused Heart

Fiona J Sutherland and David J Hearse
Cardiovascular Research
The Centre for Cardiovascular Biology and Medicine
The Rayne Institute, King’s College
St Thomas’ Hospital, London SE1 7EH, United Kingdom
Telephone: 44 0207 928 9292 ext. 5711, Fax: 44 0207 922 8139 

The isolated heart as a model for the study of cardiovascular disease.

As discussed in an earlier paper1, the isolated perfused small mammalian heart probably represents the optimal compromise in the conflict between the quantity and quality of data that can be acquired from an experimental model versus its clinical relevance - especially in relation to the modelling of ischemia. In exploiting this preparation to its full potential it is important to consider a number of key questions each of which will be addressed in the following sections. In the course of this discussion, a practical guide to perfusion methodology will emerge.

What are the advantages and disadvantages of the isolated perfused heart?

At a practical level, the isolated heart, especially from small mammals, provides a highly reproducible preparation which can be studied quickly and in large numbers at relatively low cost. It allows a broad spectrum of biochemical, physiological, morphological and pharmacological indices to be measured (see later). These measurements can be made in the absence of the confounding effects of other organs, the systemic circulation and a host of peripheral complications such as circulating neurohormonal factors. This characteristic may be considered as an investigational advantage in that it allows the dissection of peripheral from cardiac responses or a disadvantage in that it makes the preparation one step further removed from the in vivo state. Whilst the fact that the isolated heart is denervated must always be taken into account, this can be turned to advantage allowing the separation of cardiac from sympathetic and vagal stimulation. Denervation and the absence of other peripheral factors can often be compensated for; thus, catecholamines or other neurotransmitters may be included in perfusates and many other peripheral factors can be added exogenously and in a controlled manner which again can represent a significant investigational strength. Certainly, the isolated perfused heart provides an excellent test-bed for undertaking carefully controlled dose-response studies of metabolic or pharmacological interventions.

It must be recognised that, as an ex vivo preparation, the isolated heart is a constantly deteriorating preparation but nonetheless it is capable of study for several hours. The preparation also readily allows the induction of whole heart or regional ischemia and this can be achieved at various levels of flow. Similarly, anoxia or hypoxia at various degrees of oxygen deprivation (in the presence of normal flow) can be easily imposed. The isolated heart preparation is amenable to reperfusion or reoxygenation at various rates and with various reperfusate compositions thus providing a powerful tool for assessing many aspects of ischemia- and reperfusion-induced injury. Arrhythmias are readily induced and studied, especially in the larger hearts where conduction pathways can be mapped and a variety of electrophysiological recordings made. Of particular importance, the isolated heart preparation allows experiments to be continued in the face of events (e.g. infarction-induced loss of pump function, cardiac arrest or arrhythmias) which would normally jeopardise the survival of an in vivo experiment.

Which species is best for perfusion?

In essence the hearts from any mammalian species (together with non-mammalian hearts such as those from frogs or birds) may be perfused. However, although isolated perfusion of large animal hearts such as pigs, monkeys, sheep, dogs and even man has been reported, these are less frequently used. This is probably on account of the high cost, greater variability, large volumes of perfusion fluids and cumbersome equipment that is required. Without doubt the most frequently studied heart is that of the rat but there are numerous reports of studies with other species such as the rabbit, guinea pig, hamster, gerbil, ferret and mouse.2 The advent of transgenic technology will undoubtedly result in increasing numbers of studies using murine preparations and this will require investigators to become adept at using this, as yet, poorly characterised model. Unfortunately, although the literature contains an increasing number of studies which employ mouse hearts, the fundamental characteristics of the preparation (e.g. pressure-volume and calcium-function relationships) have yet to be completely characterised. Another caution relates to the high heart rate of the mouse and the miniaturisation of equipment which requires investigators to take due account of the frequency-response limitations of recording equipment.

In practical terms, the rat heart is by far, the best characterised, it is also the heart most frequently used for more complex perfusion preparations such as working and blood perfused hearts. In terms of ease of handling, the rat has a great advantage over smaller hearts such as the mouse where intraventricular pressure recordings are more difficult. The rat does however suffer from one, frequently cited, limitation, namely its very short action potential duration which can limit its value (in terms of extrapolating to the human) of some studies of arrhythmogenesis and anti-arrhythmic drugs. Other species, such as the rabbit, suffer problems with anesthesia and the guinea pig heart differs from other species in that it is totally collateralised effectively preventing the study of regional ischemia in this species. Thus, as no ideal species exists, in selecting a species for study it is crucial to recognise weaknesses and exploit advantages.

What kind of perfused heart preparation can be used?

Although a number of other variants exist, isolated perfused heart preparations are largely based on adaptations of that originally described by Langendorff3 or the more complex working preparation described by Neely4.

The Langendorff heart preparation: As shown in Figure 1, this involves the cannulation of the aorta which is then attached to a reservoir containing oxygenated perfusion fluid. This fluid is then delivered in a retrograde direction down the aorta either at a constant flow rate (delivered by an infusion or roller pump) or a constant hydrostatic pressure (usually in the range of 60-100mmHg). In both instances, the aortic valves are forced shut and the perfusion fluid is directed into the coronary ostia thereby perfusing the entire ventricular mass of the heart, draining into the right atrium via the coronary sinus.

Constant flow or constant pressure perfusion?

Depending on the requirements of the experiment, both of the two modes of perfusate delivery have advantages and disadvantages. In both instances (in the absence of any imposed ischemia) and depending upon species, the resulting coronary flow with a blood-free perfusate is often in the range of 8-12 ml/minute/g wet weight of tissue (a value which is several times that of blood flow in vivo). Whilst constant flow perfusion adds an additional element of constancy to an experiment it has the disadvantage that, unlike constant pressure perfusion, autoregulatory mechanisms are overridden and it does not automatically alter the amount of perfusate delivered to the whole heart when there are changes in heart rate or work or when regional ischemia is imposed (under which circumstance the same volume of perfusate may be forced through a much smaller perfusion bed). Switching between constant flow and constant pressure modes of perfusion is not straightforward and, with simple apparatus, may not be feasible within a single experimental protocol. Similarly, blood-perfused hearts where circulating volumes are small (see later), may not easily lend themselves to constant pressure perfusion. In order to address this problem, Shattock et al5 developed an electrical feedback system designed to control an isolated heart perfused with a peristaltic pump. This system (now commercially available through ADInstruments Ltd, Australia) allows the perfuser to switch instantly between constant pressure and constant flow modes of perfusion, enabling perfusion pressures and coronary flow to be controlled over a wide range, and providing a continuous on-line measurement of both perfusion pressure and coronary flow if desired. This latter feature is of considerable advantage in studies of vascular function. This system has been used successfully with hearts from mice, rats, guinea pigs and rabbits.

Using the rat as an example, the general procedure for Langendorff perfusion is as follows:

Excision of the heart from the donor animal: Isolation of the heart requires the donor to be rendered unconscious prior to excision. Anesthesia can be induced by inhalation of agents such as ether, halothane or methoxyflurane or injection (via an intravenous or intraperitoneal route) with agents such as pentobarbitone. Intravenous administration to a conscious animal is usually via a tail vein, whereas in the anesthetised state a femoral vein would be the preferred route (the vein being accessed by a small skin incision). An alternative to anesthesia is cervical dislocation or concussion but in both instances there are major effects on catecholamines and other circulating factors. However, there is no correct or ideal procedure for rendering an animal unconscious, nor is there an ideal anesthetic, each has its advantages and disadvantages and these will vary from species to species. Ether is hazardous as it is highly flammable and an irritant to the animal and must only be used in a well ventilated area. The most widely used anesthetic is pentobarbitone, but this is a cardiovascular and respiratory depressant that can lead to a reduction of cellular high energy phosphates6. Whatever the choice of procedure (and this may be influenced by local animal welfare regulations), every effort should be made to minimise stress by keeping the animal in a quiet environment prior to anesthesia and by minimising handling. Induction of anesthesia should be as swift as possible, with the perception of pain completely suppressed (this can be assessed by determining the animal’s response to a stimulus such as the pedal withdrawal reflex). Unless studying lipid or fatty acid metabolism (where heparin has a lipolytic action) it is advisable to administer an appropriate intravenous dose of heparin or another anticoagulant, to prevent the formation of thrombi in the excised heart.

Once the animal is anesthetised the heart can be excised. Generally, the diaphragm is accessed by a transabdominal incision and cut carefully to expose the thoracic cavity. The thorax is opened by a bilateral incision along the lower margin of the last to first ribs, the thoracic cage is then reflected over the animals head, exposing the heart. Some investigators then cradle the heart between their fingers (it is essential to do this gently to avoid contusion injury) and then lift the heart slightly before incising the aorta, vena cava and pulmonary vessels. Immediately after excision, hearts are usually immersed in cold perfusion solution (4 degrees C to limit any ischemic injury during the period between excision and the restoration of vascular perfusion). Some investigators prefer to cannulate the aorta in situ prior to excision of the heart, but whichever the preferred procedure, it is important that vascular perfusion be re-established as soon as possible after excision of the heart. With practice and close proximity of the perfusion apparatus and site of heart excision the entire process can be accomplished in less than 30 seconds although some investigators report times as long as 10 minutes (sufficient for unintentional ischemic preconditioning or even stunning!)7.

Cannulation and re-establishment of vascular perfusion: The aortic perfusion cannula (Figure 2) can be constructed from a variety of materials including glass, plastic or thin walled stainless steel. The external diameter is typically similar to, or slightly larger than, that of the aorta (about 3mm for a heart from a 250g rat). Several small circumferential grooves (Figure 2A) or a single flange is usually machined into the distal end of the cannula to prevent the aorta from slipping off. Some cannulae (Figure 2B) are heated with water-jackets to prevent any unwanted fall in perfusate temperature as it is delivered to the heart (see later section on the importance of temperature regulation).

A water-jacketed reservoir, situated above the aortic cannula, contains the perfusion fluid which is oxygenated via a sintered glass gas distributor (for bicarbonate-based perfusion fluids 95%O2 + 5%CO2 is normally used). It is advisable to have the perfusion fluid gently dripping from the aortic cannula prior to cannulation since this helps minimise the chance of air emboli at the time the heart is attached to the cannula. Cannulation is aided by cutting along the aortic arch to open it, thus giving a larger area for cannulation. Hearts should be held gently between the tips of blunt-ended fine curved forceps, taking care to avoid stretching or ripping of the aortic wall. The aorta is then gently eased over the end of the cannula, taking care not to insert the cannula too far into the aorta since this would occlude the coronary ostia or damage the aortic valve. The
aorta is then clamped to the cannula with a small blunt artery clip, whilst a ligature is rapidly tied around the aorta, locking into the grooves; the artery clip can then be removed. In the case of the flanged cannula the aorta is slid down the cannula so that the tie is against the flange. Full flow of perfusate should be initiated as soon as the heart is mounted on the cannula.

Once the heart is securely attached to the cannula any surplus tissue (such as bits of thymus, fat or lungs) can be trimmed away. Drainage of coronary perfusate from the right side of the heart via the pulmonary artery should be unimpeded, however, in the course of cannulation it is possible to accidentally ligate the pulmonary artery. Thus, to facilitate adequate drainage it is advisable to make a small incision in the base of the pulmonary artery using small pointed scissors. Some investigators elect to cannulate the pulmonary artery, particularly if they are interested in measuring A-V differences in pO2 or are making a separate collection of cardiac lymph. Some investigators also insert a small drainage catheter in the apex of the left ventricle to prevent the accumulation of any Thebesian flow in the left ventricular lumen. After its passage through the coronary vasculature the coronary flow can either be discarded or collected for analysis; if recirculating perfusion is required the coronary effluent can be returned (preferably via a 5m filter) to the perfusion fluid reservoir for reoxygenation.

Once cannulation is completed and coronary perfusion initiated, contractile function and regular heart rhythm will return within a few seconds but it may be 10 minutes or more before maximum function is established. During this time, various instrumentations of the heart can be undertaken. Most studies require contractile function to be measured (not only for the provision of baseline data but also to monitor the stability of the heart and the extent of any disturbances of cardiac rhythm). Although contractile activity can be assessed via a strain gauge attached to the apex of the heart or an open tip pressure transducer inserted into the left ventricle, the preferred procedure involves the insertion of a compliant intraventricular balloon. These balloons are often made from thin silicone rubber or domestic food wrap. Not only is this ideal for the measurement of isovolumic left ventricular function but it is also a convenient means of measuring heart rate. For the balloon insertion, the left atrial appendage is removed to provide a clear field of view and a deflated balloon, attached to a short rigid catheter and pressure transducer, is introduced into the left ventricle via the mitral valve. The neck of the balloon is then either sutured into position or firmly attached to the cannula with a small piece of plasticine. Great care must be taken while inserting and securing the balloon as it is very easy to damage the heart - especially small hearts such as those from mice or neonatal rats. The balloon is inflated with water from a microsyringe until a left ventricular end diastolic pressure of between 4-8 mm Hg is obtained (an excessively high balloon volume should be avoided since it may cause tissue compression and subendocardial ischemia which in turn will lead to an unstable preparation with an increased propensity to arrhythmias). Once the balloon is in position, left ventricular systolic, diastolic, and developed pressure can be recorded. If the heart is to be paced, one silver wire recording electrode can be hooked on the ventricle with the reference electrode attached to the stainless steel cannula. A bipolar stimulator is recommended in order to avoid toxicity from electrolysis of the perfusion fluid which can occur when unipolar stimulation is used. If an ECG is to be recorded, the electrodes can be positioned as required, with again, the stainless steel cannula making a suitable indifferent electrode. Once instrumentation of the heart has been completed, a temperature regulated heart chamber should be placed around the heart and the top covered over with domestic food wrap or other thin material.

The working heart preparation: As shown in Figure 3, this is a more complex preparation with ventricular filling via the left atrium and ejection in the normal direction via the aorta. This preparation offers the advantage of an ability to measure pump function with different filling pressures and afterloads. Rat hearts are the most frequently used species for working heart preparations but all species can be used - even the dog or pig.

Excision of the heart from the donor animal: Since the first step in establishing a working heart preparation is to set up a Langendorff preparation, the procedure for anesthesia and excision is identical to that described earlier.

Cannulation and re-establishment of vascular perfusion: Again, the first steps are identical to the Langendorff procedure described earlier, except that the Langendorff perfusion line is usually attached to a side arm of the aortic cannula (Figure 2C). Following the establishment of perfusion (without the insertion of an intraventricular balloon) there is just one additional step (which proves most difficult to those learning the procedure) namely the cannulation and secure tying off, without any leaks, of the left atrium via one of the orifices of the pulmonary veins. The dimensions and relative positions of the aortic and atrial cannulae are critical to a successful preparation. Once aortic and left atrial cannulation are accomplished, the Langendorff perfusion line is clamped ("a" closed) and perfusion initiated by unclamping the perfusion line from the left atrium ("c" opened) whilst simultaneously unclamping the aortic outflow line ("b" opened). In this way, oxygenated perfusion fluid from a constant pressure head left atrial perfusion reservoir (which is continuously filled by a roller pump from a gassing reservoir which is frequently referred to as ‘the lung’) flows under gravity into the left atrial cannula. The preload of the preparation is determined by the height of the overflow from the atrial perfusion reservoir above the heart. This is usually set to 20cm for isolated rat hearts but can be varied to suit other preparations or to allow construction of "Starling curves" relating preload to cardiac function. It is important to stress that, in vivo, cardiac output is equal to the venous return from the lungs to the left atrium - in the isolated working heart the venous return is represented by the flow from the left atrial cannula. An important point, often overlooked by investigators when constructing a working heart apparatus, is that the left atrial perfusion line must be capable of delivering perfusion fluid at a rate sufficient to support the maximum cardiac output of a working heart at any particular preload. If the left atrial cannula is too small it will artificially limit the cardiac output of the preparation. The problem is compounded by the pulsatile nature of atrial filling, which means that the atria only fill during about half of the cardiac cycle. To ensure that this problem does not arise and that left ventricular filling is not limited by inadequate left atrial inflow it is essential to check that the left atrial perfusion line can deliver a flow rate of at least twice the expected maximal cardiac output. This is easily checked by running the apparatus without a heart attached and measuring the flow from the left atrial line - a rate of at least 150ml/min is recommended for a 1g heart. Having flowed from the left atrial cannula into the left atrium, the perfusion fluid is ejected via the mitral valve into the left ventricle from where it is ejected through the aortic cannula against a hydrostatic pressure via the elasticity chamber and flowmeter to the top of the lung. The afterload is determined by the height of the column of fluid above the aortic cannula. The elasticity chamber, which contains a known volume of air for the working rat heart, mimics normal vascular elasticity. It is an essential component of the perfusion circuit and without it the heart will rapidly fail. In the course of left ventricular ejection a portion of the perfusion fluid is forced into the coronary ostia and thereby perfuses the coronary vessels of the heart. The coronary effluent exits from the right heart into the heart chamber from whence it may be sampled for assay or returned (via a roller pump) to the lung for reoxygenation. Depending on the species and experimental design, filling pressures are usually in the range of 10-20 cmH2O and afterloads in the range 60-100 cmH2O. Under these conditions and using the heart from a 250g rat, coronary flows of up to 25 ml/minute and aortic flows of 50-80 ml/minute can be expected. These can be measured by timed collection into measuring cylinders or by in-line float or electromagnetic flow meters. Summation of coronary and aortic flow gives the cardiac output. Hearts may be paced or allowed to beat spontaneously under which circumstances heart rate may be derived from a pressure recording which is usually via a side arm of the aortic cannula. Because of the large volumes of perfusion fluid pumped by the heart, the working preparation usually operates in the recirculating mode and for this reason it is essential to have an in-line filter (5m porosity) in the circuit to remove any particulate contaminants which may originate from the heart, connecting tubing, glassware or perfusion solutions.

What is the best perfusion temperature?

Irrespective of whether a Langendorff or working heart preparation is used it is obviously preferable to perfuse at or near the normal body temperature of the species under study. This value varies somewhat between species but in general (unless as in some surgical studies hypothermia is deliberately employed), most investigators elect to perfuse their hearts at 37.0-37. 5C. It cannot be stressed too strongly how critical it is (and how difficult it is) to maintain good and uniform temperature control, it is strongly recommended to have permanent temperature sensing microthermisters at various parts of the circuit. There are two basic approaches to maintaining the heart, the perfusion circuit and the fluid it contains at the correct temperature. The first is a thermostatically-regulated cabinet in which moist warm air is circulated. These are rarely used, they seldom work well (in part due to the loss of heat that occurs every time the door is opened) and they can restrict access to the heart. Most investigators choose to use a thermostatically-controlled water-jacketed system in which all glass reservoirs, the heart perfusion chamber and as many of the delivery lines as possible are surrounded by rapidly flowing water at 37.0-37. 5C. To be effective, this requires a high output, well regulated, water circulator and the careful design of the circuit since some temperature drop across the apparatus is inevitable. For this reason, key compartments such as the heart chamber, the cannula assembly (if water-jacketed) and the Langendorff or left atrial inflow lines should receive flow from the circulator first. Bifurcation of the water lines should be avoided as kinking of tubes may lead to different flow rates to different parts of the apparatus. Finally, investigators should avoid falling into the trap of setting the output of the circulator at hyperthermic temperatures (e.g. 39C) in an attempt to compensate for any temperature drop between the circulator and the heart chamber as this will inevitably cause problems at some time or another. In general it is far better to suffer a small degree of hypothermia in the circuit since this will not damage the tissue and the relationship between cardiac function and temperature, which, although very steep below 35C, is relatively flat between 35 and 37C. A key point is to avoid over-heating the tissue and to maintain a constant temperature even if it is a little below that desired. Not surprisingly, due to the necessarily very long perfusate delivery lines, temperature control in hearts that are perfused in NMR spectrometers are very vulnerable to poor temperature control. Heat loss from the heart chamber can be reduced by the use of cannulae attached to silicon (not red rubber) bungs the diameter of which matches that of the heart chamber, thus creating a tight seal. While this is easy to achieve in a working heart preparation, it is harder in the Langendorff heart due to the need for balloon insertion, although this can be overcome by the creation of a small groove in the bung for the balloon catheter. As mentioned earlier, an alternative approach is to cover the area between a smaller bung and the heart chamber with waxed laboratory film.

What indices of tissue function, integrity and injury can be measured?

As already discussed, the isolated heart, whether it be a Langendorff or working preparation, provides the opportunity for the acquisition of a very wide spectrum of highly reproducible data in a rapid and cost-effective manner. A variety of physiographs can be used to record the data, computer based recording devices offer some advantages over the more traditional recorders in that they allow better data storage and analysis.

Morphology and vascular anatomy: The perfused heart can readily be taken for examination by either light or electron microscopy. In both instances it offers the very important advantage that fixation can be by coronary perfusion. For sequential studies, multiple microbiopsies can be taken (starting at the apex of the heart and working towards the base) at different times during a perfusion protocol (this necessitates immersion fixation). In addition, whole hearts can be used for gross morphology such as is required during infarct size studies. Hearts can also be perfused with a variety of gels, particles and resins that allow the specific visualisation of vascular perfusion beds.

Biochemistry: There are a multitude of measurements that can be made using the perfused heart. Arterio-venous differences in substrates, metabolites such as lactate, oxygen and a host of other markers of normal and abnormal metabolism can be made. Leakage of cellular constituents such as enzymes and proteins can readily be made for the assessment of tissue injury and whole hearts or sequential biopsies can be taken for metabolic analysis - typically of high energy phosphates and the way in which they are influenced by conditions such as ischemia and reperfusion. The ability to perfuse hearts in an NMR spectrometer allows continuous on-line measurement of metabolites and intracellular ions such as calcium, protons or sodium. Transluminal or surface reflectance fluorescence spectroscopy allows the continuous monitoring of intermediates such as NADH or the contractile transients of ions such as calcium. Suction microelectrodes allow continuous measurement of interstitial potassium, calcium, pH or monophasic action potentials. The perfused heart also provides an ideal model for the delivery of vectors in gene transfer studies where adenovirus or other vectors can be selectively delivered and trapped in the coronary vasculature.

Cardiac rhythm and electrophysiology: Electrocardiographic recordings allow the detection, identification and quantification of abnormalities of cardiac rhythm and microelectrodes allow further analysis - indeed, in larger hearts such as the rabbit and the pig, conduction pathway mapping and selective ablation is possible.

Cardiac contractile function: Whereas the Langendorff preparation provides valuable information on left ventricular systolic and diastolic pressures and their derivatives, the working heart gives valuable data on cardiac pump function. In addition, as mentioned earlier, various tension recording devices may also be attached to the heart. Ultrasonic crystals can be readily used for regional or transmural function studies and various echo techniques can also be employed for measurements of wall thickening. Pressure-volume relationships can be studied with ease as can more rarefied indices of contractile function such as Emax.8

Pharmacology: Both isolated heart preparations are extremely valuable for assessing the direct cardiovascular effects of various therapeutic agents in terms of contractile function, electrical activity or metabolic function. The option for recirculating and non-recirculating preparations allows drug dose-response studies to be carried out with great speed and reproducibility and with precise control over concentration. Another advantage is to be able to rapidly washout drugs from the circulation by replacing perfusion fluids.

Vascular biology: Whilst most research in the isolated perfused heart has tended to focus on the function and malfunction of the myocyte, using contractile and metabolic endpoints, it should be stressed that both the Langendorff and working heart can be used to study vascular reactivity, endothelial and smooth muscle function and the effect of a variety of interventions on coronary flow and its distribution. Indeed, the isolated heart has been the cornerstone of much of the work on the no reflow phenomenon.9

Should hearts be paced?

Like so many other facets of heart perfusion, this decision is based on a compromise between protocol requirements and quality of data. Left to contract at its spontaneous rate, the isolated heart undergoes a small progressive time-dependent decline in heart rate. Furthermore, spontaneous heart rate in a perfused heart preparation is usually significantly below the physiological norm. For rat hearts (which have an in vivo rate of 350-400 beats/minute), heart rates of 250-320 beats/minute, can be expected. This changing baseline will add additional variability to the data not only as a consequence of the fall in heart rate plus individual-to-individual variability in heart rate but also in the value for left ventricular pressure which ( depending upon whether the species under study has a positive or negative staircase) will either increase or decrease as a consequence of the fall in rate. Some investigators attempt to compensate for this by expressing function in terms of pressure x rate product.

In vivo, the atria and sino-atrial node are not perfused by coronary vessels but by extracardiac vessels, which are severed when the heart is excised for perfusion. The affected tissue is therefore dependent for its oxygen, nutrients and temperature control upon either: (i) fortuitous superfusion by coronary effluent flowing out of the right atrium or (ii) deliberate superfusion from a cannula delivering warm perfusion fluid. The implementation of an atrial superfusion line certainly helps in maintaining temperature and a more stable heart rate. However, it complicates the assessment of coronary flow. An alternative approach is to immerse the entire heart in oxygenated, temperature-regulated perfusion fluid.

Isolated hearts are easily paced to most required levels and in studies where low heart rates are required (and escape to a higher spontaneous rate may be threatened), the sino-atrial node may be crushed to prevent natural impulse generation. Pacing is not recommended in any studies of arrhythmogenesis (where the pacing may modify the nature or incidence of any arrhythmia) or the study of agents that influence heart rate. During studies involving severe ischemia many investigators choose to terminate pacing during the ischemic interval and may delay restoration of pacing until several minutes after the onset of reperfusion.

What is the maximum duration of perfusion?

Perhaps the most frequent question posed by lay visitors to a heart perfusion laboratory relates to the longevity of the preparation. Clearly, from the moment an ex vivo preparation is established, it will begin to deteriorate and the rate will depend on a large number of factors including the skill of the operator (avoiding contusion injury), the species, the composition of the perfusion fluid, the presence or absence of various drugs, age, heart rate and work load and the temperature at which the studies are carried out. Certainly, investigators should always undertake their own stability studies and establish their own exclusion criteria. However, it is our experience that, with both the Langendorff and working heart preparation, a deterioration of contractile function (e.g. left ventricular developed pressure or cardiac output) of 5-10%/hour can be expected. However, this can be influenced greatly such that, in some species, periods of hypothermic arrest with tissue preservation for up to 24 hours or longer may be followed by a return to approaching initial levels of function. The rate at which a preparation deteriorates can be critical in the design and interpretation of some studies where it may be necessary to use time-matched controls with corrections for baseline deterioration when comparing groups.

Should exclusion criteria be used in perfused heart studies?

No self-respecting investigator who uses in vivo preparations would design and undertake a protocol without pre-defining criteria for the exclusion of preparations that are, for one reason or another, unacceptable for use. Unfortunately, only a minority of publications involving heart perfusion report prospective exclusion criteria or policy for replacement of excluded hearts (the latter being a potentially difficult issue with the study of arrhythmias10). Those few publications that do cite exclusion criteria often only state the lower limits of a functional index (usually spontaneous heart rate or left ventricular developed pressure), apparently accepting very high values however extreme. The failure to apply rigorous prospective exclusion criteria in heart perfusion studies probably has its origins in the remarkable reproducibility of these preparations - however, this is not an acceptable excuse for ignoring a fundamental requirement of good protocol design.

What is the best perfusion fluid composition?

The majority of studies in the literature (with the exception of those involving NMR where phosphate often has to be removed from the solution) are based on a bicarbonate perfusion fluid as defined by Krebs and Henseleit.11 This perfusion fluid, which was supposed to mimic the key ionic content of blood or plasma and have a pH of 7.4 at 37C, has the following composition (in mM): NaCl 118.5, NaHCO3 25.0, KCl 4.7, MgSO4 1.2, KH2PO4 1.2, glucose 11.0 and CaCl2 2.5. Unfortunately, in formulating this ‘physiological’ solution, Krebs and Henseleit failed to take account of the fact that much of the calcium in blood is bound to proteins and the realistic plasma ionised calcium concentration is approximately half of the recommended value of 2.5mM. As a consequence, for decades, heart perfusers have used excessively high concentrations of calcium in their perfusion fluid which usually do not contain protein or molecules capable of binding calcium. This has meant that, in many studies, hearts have been in a state of continuous inotropic challenge, probably working near the upper limits of the calcium/function curve. Nowadays, many investigators are rectifying this problem by using ionised calcium concentrations which are in the range of 1.2-1.8mM.

In discussing calcium, it is important to stress that, in preparing perfusion solutions containing both calcium and phosphate ions, there is a risk of precipitation of calcium phosphate particles which will occlude coronary arteries and destroy a preparation. This is readily circumvented in bicarbonate-buffered solutions by ensuring that the calcium component is the last to be added and that the pH is lowered by gassing with 95%O2 + 5%CO2 before adding the calcium. Perfusion fluids should always be filtered (5m filter) before use to remove the remarkable number of particulate impurities that can be present in even the purest commercial chemical.

In discussing perfusion fluid composition, it is important to consider the substrates that are provided to support the large energy requirements of cardiac contractile function. Traditionally, high concentrations (approximately 10mM) of glucose are usually added as the sole substrate. The use of ‘diabetic’ levels of glucose reflects the fact that normal physiological concentrations of glucose are unable to sustain adequate function in the perfused heart in the absence of insulin. Insulin can be added to overcome this problem, however, it is rarely done - probably because of the other complex cellular effects of insulin. The choice of glucose as the sole substrate in most heart perfusion relies on the heart’s ability to utilise almost any substrate as an efficient energy source - this is despite the fact that, in vivo, fatty acids are the predominant energy source. The general practice of avoiding fatty acids as an energy source results primarily from the difficulty of dissolving these agents in aqueous solutions and the complication of frothing when the fatty acid-containing solutions are gassed. The same reason, together with cost, probably explains the general reluctance to add albumin, or other oncotic agents, to perfusion solutions - a decision that undoubtedly contributes to the severe edema which characterises most perfused heart preparations. Drugs or other agents can be added to any perfusion solutions; this is normally achieved by adding them, in the desired concentration, to the basic perfusion medium. However, in some cases, (e.g. unstable compounds) concentrates can be infused via a side arm to the aortic or left atrial cannula. In this instance, the concentration of the infusate is set so as to achieve the desired circulating concentration when the infusate is mixed with the perfusion fluid as it enters the cannula.

What is the most common cause of unstable or failing preparations?

Other than mishandling the heart during excision, making an error in the formulation of a perfusion solution, adding a toxic agent to a perfusion solution or the occurrence of arrhythmias, the most common cause of contractile failure in both working and Langendorff hearts is contamination - either particulate or bacterial. The source of bacterial contamination may be the perfusion apparatus itself where inadequate washing at the end of an experiment or poor apparatus design (small pockets between oversize tubing and a glass or plastic connector which create spaces where the substrate-containing perfusion fluid can be trapped) allows bacteria to thrive. Alternatively, the perfusion solution may be the cause with particulate impurities in reagents or bacterial contamination that occurs during preparation or storage (it is recommended not to store substrate-containing solutions for more than a few hours). In both instances the problem can be alleviated by: (i) filtration (5m filter) in the course of preparation and inclusion of filters in recirculating perfusion circuits, (ii) making up fresh and not storing glucose- or substrate-containing buffers which are good bacterial growth media and (iii) thoroughly washing the perfusion apparatus with either detergent or distilled water followed by (with appropriate safety precautions) boiling water after every day of use. Some investigators resort to the inclusion of broad spectrum antibiotics in their perfusion media but this is not recommended. If contamination of the perfusion apparatus does occur it may be possible to remove it by washing with acid or detergent but usually it is better to dismantle, wash and sterilise all components. However, a well designed and properly washed apparatus with well prepared perfusion solutions can be used for years without contamination occurring.

Oxygen delivery: gassed perfusion solution, perfluorochemicals, erythrocytes or blood?

Central to the survival of any perfused organ is the continuous provision of oxygen in quantities sufficient to support normal metabolism, the maintenance of transmembrane ion gradients and, in the case of the heart, the large amounts of energy needed to support contraction. In conventional Langendorff and working heart preparations, oxygen is provided by gassing the perfusion fluid with high concentrations of oxygen, typically 95%O2 + 5%CO2 (the CO2 being required to achieve the correct pH in bicarbonate-based buffers). If there are no fatty acid or protein additions to the buffer then this is a straight forward procedure with no complications from foaming. If foaming is a problem then anti-foam agents can be added (but this is not recommended) or membrane oxygenators may be employed.

Gassed asanguinous perfusion fluids have a relatively low oxygen carrying capacity but a very high pO2 (> 500mmHg). The fact that they are able to provide sufficient oxygen to the heart is testified by: (i) reducing the pO2 by using gassing mixtures containing only 70% oxygen does not result in a loss of contractile performance, (ii) the venous pO2 is still relatively high indicating a reserve in oxygen availability, (iii) perfusion fluid perfused preparations will work for several hours without any great deterioration in function and (iv) when challenged with inotropic agents perfusion fluid perfused hearts can increase their work for sustained periods without injury. However, it should be stressed that asanguinous perfusion fluids can only provide sufficient oxygen if the coronary flow rates are several times the physiological norm.

Driven by concerns over the transmission of infections during blood transfusion, there has been considerable research into developing blood substitutes and especially perfluorochemical oxygen-carrying hemoglobin substitutes. These agents, when used as emulsions with aqueous media, are able to deliver more than twice as much oxygen.12 As a consequence perfluorochemicals have been used by some investigators in perfusion solutions but this practice has not been widely adopted. This probably reflects the general satisfaction that oxygenated aqueous media are able to deliver sufficient oxygen to most isolated heart preparations. As a consequence, most work in the perfluorochemical area has resolved around using these agents to provide oxygen to tissue which is deprived of adequate flow. In this connection there are many studies of these agents as additives to cardioplegic solutions or as agents to limit the evolution of myocardial infarction. However, again, they have not been introduced into widespread use.

The artificially high coronary flow rates and high pO2 of asanguinous perfusion solutions, together with the absence of the many key components of blood, makes the concept of perfusing isolated hearts with blood an attractive option. Certainly, blood perfusion is used successfully with many other organs; it would be expected to allow near normal coronary flow rates; its protein content should help alleviate the unwanted edema that characterises perfusion fluid-perfused hearts and it should also allow the heart to be exposed to blood-borne elements such as neutrophils. The widespread use of blood in heart perfusion has probably been discouraged by: (i) the volume of blood needed (perfusing one rat heart would require the blood of several rats), (ii) the inability to use donor blood from some (but not all) larger species because they have larger erythrocytes than the rat and cannot traverse rat heart capillaries and (iii) the difficulty in oxygenating the blood since conventional gassing creates extensive foaming and damage to blood cells. However, all of these problems can be circumvented either through the use of parabiotic preparations with support rats or the use of membrane oxygenators. As a consequence, a number of blood or erythrocyte preparations have been developed13,14,15,16 and used with great success.

Blood perfusion: The first report of an attempt to develop an isolated blood perfused rat heart was published by Gamble et al in 1970.14 A ventilated support rat was placed in a chamber containing humidified air at 37C and the left carotid artery and right jugular vein were cannulated for the supply and return of blood to an isolated Langendorff heart. Blood from the carotid artery of the support rat first flowed into a latex bag, the volume of which was maintained constant by a feedback device. The blood was then pumped to the aortic cannula of the isolated heart and returned to the jugular vein of the support rat. In 1988, Abraham et al15 modified the preparation by removing the pump.

Blood perfusion with a support animal for oxygenation: The preparation currently used in our laboratory for rats and rabbits, is based on the procedure described by Qiu et al13,16, which is a combination of the two methods mentioned above. The details for establishing this preparation (Figure 4) in the rat are outlined below.

Support rats (350-400g) are anesthetised with sodium pentobarbitone (60mg/kg, ip) and anticoagulated with heparin (1000 IU/kg). The left femoral artery and right femoral vein are exposed by blunt dissection and cannulated for the eventual supply and return of blood to an isolated Langendorff perfused heart from another animal of the same species. An extracorporeal circuit between these two vessels, primed with a Gelofusine solution (a plasma substitute consisting of gelatin, sodium chloride and water) is constructed and the flow of blood through this is gradually established and maintained for a short time (usually 20 minutes) so as to ensure that the priming solution is adequately mixed with the blood of the support animal and the entire preparation is stable. It is advantageous to decrease the hematocrit of the perfusing blood to approximately 28-30% by mixing it with Gelofusine. This decrease in the viscosity allows for better flow and also helps minimise the possibility of clotting. The gradual establishment of the extracorporeal circuit is achieved via an in-line peristaltic pump which is positioned on the arterial outflow of the support animal and flow through this circuit is gradually increased over 10 minutes to its final required value (which depends on the species used and the size of the heart but for a 1g rat heart the flow rate would normally be in the range 2-3 ml/minute). The gradual increase in flow through the extracorporeal circuit is necessary to prevent the sudden drop in arterial pressure that would have occurred if the final flow rate of had been established abruptly. The feedback system described earlier can be used with this preparation to perfuse the isolated heart either under conditions of constant flow or constant pressure. The blood is pumped through a cannula (to which the aorta of the perfused heart will subsequently be attached) and returned, by gravity, via a reservoir (comprising of the bottom half of a large syringe positioned in a water-jacketed heart chamber) and filter (200m filter) to the venous inflow line of the support animal. Meanwhile, the isolated rat heart is harvested and prepared exactly as described in earlier sections. The heart can also be instrumented with an intraventricular balloon and other devices exactly as described earlier. The extracorporeal circuit then flows from the left femoral arterial supply line of the support animal to the isolated heart. The coronary effluent from the perfused heart is then returned, by gravity, via a reservoir and filter to the venous inflow line of the support animal. During periods of global ischemia, the blood can be diverted from the isolated heart by opening the bypass line (which has been positioned to flow into the reservoir) and clamping the line to the heart.

It is possible to artificially ventilate the support rat, however, it is usually sufficient to allow the animal to spontaneously breathe a mixture of 95% O2 + 5% CO2 through a Venturi face mask, the flow rate of which is adjusted to maintain blood pO2 and pCO2 within the physiological range. Before placing the face mask over the support animal, the tongue should be gently pulled slightly out of the mouth and placed to one side to ensure unobstructed breathing. Body temperature, monitored by a rectal thermometer, should be stabilised by means of a thermostatically-controlled heating pad. Blood pressure should be monitored by a pressure transducer attached to the arterial line. Blood gas and electrolyte levels of the support rat should also be checked constantly and corrected when necessary. During the course of the experiment, donor blood from another rat can be transfused as required to maintain the volume and stability of the preparation. Additional anticoagulant and anesthetic can be administered as required.

The above method of blood perfusion requires that very careful attention be paid to the hemodynamics and blood chemistry of the support animal, with constant monitoring of vital signs, breathing and the colour of mucous membranes. An air-filled syringe (Figure 4) above the perfusion cannula acts as a compliance chamber, which serves to dampen oscillations in perfusion pressure which occur as a consequence of the contraction of the isolated heart and the peristaltic action of the pump.

Advantages and disadvantages of the preparation: There are a number of major advantages with this preparation. Firstly, in our experience, it is even more stable than the isolated perfusion fluid perfused heart, with left ventricular developed pressure deteriorating less than 5%/hour. The blood perfused heart suffers far less edema and excellent pressure development is observed (in the range 130-180 mmHg when end diastolic pressure is in the range 4-8 mmHg). Coronary flow rate (2-3 ml/minute) is much closer to the physiological range than that with perfusion fluid perfused preparations and the preparation allows one to study the effect of blood elements such as neutrophils on cardiac function. By using two support animals attached to one heart it is possible to study the effect of transient exposure (or removal) of factors such as neutrophils. The preparation can be used to study global or regional ischemia at zero or reduced flow rates exactly as a perfusion fluid-perfused heart and is amenable to the same wide range of biochemical, morphological, pharmacological and functional assessments. With careful attention to the support animal, it is our experience that one support rat can be used for several hours and allow the sequential study of more than one isolated heart.

A potential disadvantages of the support animal method of blood perfusion is that any substance released by, or administered to, the isolated heart will normally be transmitted to the support animal with possible deleterious consequences. On the other hand, in some instances, the support animal may exert a beneficial action by metabolising or excreting such substances. Another related disadvantage of this preparation is that any deterioration of the support animal will threaten the survival of the isolated heart. However, with care, a replacement support animal can be introduced into the perfusion circuit. When perfusing with either whole blood or a washed red cell preparation, care must be taken to ensure that the perfusate does not come into contact with glass as this will promote red cell hemolysis.

Erythrocyte perfusion: If the use of a support animal is not feasible then an alternative approach is to undertake isolated heart perfusion with washed red blood cells using a simple membrane oxygenator constructed from coils of thin walled silicone rubber tubing. In 1979 Bergmann et al17 reported a preparation in which isolated rabbit hearts were perfused with perfusion fluid containing red cells at a hematocrit of 25% or 40%. A key feature of this preparation was the use of blood cells from a different species (sheep) which are small enough to traverse the capillaries of the rabbit heart. They found that, in addition to exhibiting an enhanced stability when compared to crystalloid perfused hearts, the red cell perfused hearts were less likely to develop significant edema. The model currently used in our laboratory18 is essentially similar (Figure 5). A red cell (bovine blood) suspension is kept in a reservoir which is stirred continuously to prevent red cell sedimentation. From this reservoir the red cell suspension is pumped in an artificial ‘lung’, comprising of gas permeable silicon tubing wound into a coil. This tubing is housed in a water-jacketed chamber, which maintains the perfusate at 37C. We find it to be advantageous to pump the red cell perfusate both into and out of the oxygenator since pumping on one side of the circuit can lead to stretching and distortion of the oxygenator tubing. However, it is essential that the pump speeds are identical to prevent emptying or pooling of blood in the oxygenator. In addition, 95%O2 + 5%CO2 is pumped into the reservoir so as to maintain oxygenation of the red cells, whilst keeping pH within the physiological range. The perfusate is then passed through a white cell filter, which has two functions: (i) the large surface area of the filter and its immersion in heated water provides an efficient method of temperature regulation and (ii) the filter protects the heart against micro-emboli.

Excision of the heart from the donor animal: The procedure for anesthesia and excision is identical to that described in the section on blood perfusion with a support animal.

Preparation of a red cell perfusate: Preparation of a red cell perfusate is more time-consuming than the preparation of a simple perfusion solution. Fresh blood (most commonly ovine or bovine) is collected into vessels containing 5,000 IU heparin/litre of blood and 0.2 MU benzylpenicillin /litre of blood. After filtering (200m filter) the blood is centrifuged at 1000g for 20 minutes, after which the supernatant and white cell layer are removed and discarded and the red cells resuspended in nominally calcium-free perfusion fluid. This process is repeated until the supernatant is clear. If the cells are to be used the next day, it is advisable to omit the final resuspension, instead storing the packed cells in a refrigerator (4C) and washing twice on the day of use. Following the final wash, the red cell concentrate is mixed 1:1 with a dextran/albumin solution, made by adding 250ml of deionised water to 500 ml of sterile Gentran 70 (a plasma expander consisting of 6% dextran and 0.9% sodium chloride). Electrolytes and glucose are added in the following concentrations (mM): NaCl 118.5, KCl 3.8, KH2PO4 1.2, NaHCO3 25.0, MgSO4 1.0 and glucose 10.0. After filtration (5m porosity) the solution is warmed and 3g of albumin (Fraction V, Sigma) added. Prior to perfusing, the ionic composition of the solution should be checked using a blood gas/electrolyte analyser and corrected if necessary (e.g., the final calcium concentration should approximate the ionised calcium concentration in blood). Dextran is used to prevent microagglutionation and gentamicin (2 mg/litre) can be added to retard bacterial growth.


Advantages and disadvantages of the preparation: The advantages with this preparation are similar to those for the blood perfused preparation which uses a support animal: (i) it is more stable than the isolated perfusion fluid perfused heart, with pressure development remaining stable for extended periods of perfusion (approximately 5 % loss/hour), (ii) the red cell perfused heart suffers far less edema than the perfusion fluid perfused heart, and (iii) coronary flow rate (2-3 ml/minute) is much closer to the physiological range than that with the perfusion fluid perfused preparation. Pressure development in the erythrocyte perfusion preparation is usually not as great as that seen in the blood perfused heart and is similar to that seen in the buffer perfused heart.

Perfusing hearts with blood from another species raises the possibility of an adverse immunological reaction, but as the blood is normally depleted of white cells and plasma proteins during its preparation, significant reactions are unlikely. However, progressive hemolysis of the red cells during use is unavoidable, care should be taken to avoid contact with glass (use plastic only) during the preparation or usage of the blood, as this will greatly accelerate red cell haemolysis.


How can ischemia be best induced?

Whether perfused with an asanguinous solution, washed red cells or blood in the Langendorff or the working mode, many investigators use the isolated heart for the study of regional or global ischemia. The isolated preparation is ideally suited to such studies since pump failure or lethal arrhythmias do not necessarily terminate an experiment as would be the case with in vivo studies. Global (whole heart) zero-flow ischemia is readily induced in the Langendorff and the working heart simply by occluding the perfusion inflow lines. Graded whole heart ischemia at various degrees of flow can also be readily induced in the Langendorff preparation but is difficult to achieve in the working heart (as a consequence, investigators often switch the working heart temporarily back to the Langendorff mode for such treatment). Regional ischemia can also be induced in both preparations by ligating a coronary artery (usually the left main) and the size of the ischemic zone can be influenced by the positioning of the occlusion point. To prevent tearing of tissue and also to facilitate reperfusion, the occluding ligature is usually tied against a small length of plastic tubing. Reperfusion can be achieved by cutting the ligature. The relatively uniform coronary artery anatomy of the rat heart offers a great advantage in that ischemic zones of very similar sizes are usually created. The size of the ischemic zone can be easily estimated by the percent fall in coronary flow (in constant pressure perfusions) or increase in coronary vascular resistance (in constant flow preparations). Validation of occlusion is often made by transient infusion of a coloured dye or fluorescent particles. In the Langendorff heart a novel dual lumen perfusion cannula19 (Figure 2D) can be used to facilitate independent perfusion of the left and right coronary ostia. This cannula is a powerful tool for investigating the regional effects of drugs - it also offers the unique opportunity of creating regional low flow ischemia in the rat heart. Isolated hearts also provide the opportunity for studying hypoxia and reoxygenation, again either regional or global.

Concluding comments.

Having selected the isolated perfused heart as a model to investigate a cardiovascular phenomenon, the choice of preparation is very wide. Blood perfused or perfusion solution perfused, Langendorff or working, each has its own advantages and disadvantages, both of which can be established in the quest for a better understanding of cardiac function and malfunction. Although the investigative power of the preparation is great, as with all experimental models, they are fraught with potential pitfalls, which must be recognised and addressed. This article has attempted to address many of these issues in the hope that it will assist investigators to make the best possible use of this experimental model. 


1. Hearse DJ and Sutherland FJ. Experimental models for the study of cardiovascular function and disease. Pharmacological Research 2000 (in press).

2. Galinanes M and Hearse DJ. Species differences in susceptibility to ischemic injury and responsiveness to myocardial protection. Cardioscience 1990;1: 127-143.

3. Langendorff O. Untersuchungen am uberlebenden Saugethierherzen. Pflugers Archives fur die Gesamte Physiologie des Menschen and der Tiere 1895;61:291-332.

4. Neely JR, Liebermeister H, Battersby EJ and Morgan HE. Effect of pressure development on oxygen consumption by isolated rat heart. American Journal of Physiology 1967; 212:H804-H814.

5. Shattock MJ, Miller JIA, Bray DG and Waldron CB. An electronic feed-back circuit to control a peristaltic pump for constant-pressure perfusion of isolated hearts or other organs. Journal of Physiology 1997; 505:4P.

6. Bretschneider HJ, Hubner G, Knoll D, Lohr B, Nordbeck H and Spieckermann PG. Myocardial resistance and tolerance to ischemia: physiological and biochemical basis. Journal of Cardiovascular Surgery 1975; 16 ,3:241-260.

7. Awan MM, Taunyane C, Aitchison KA, Yellon DM and Yellon DM. Normothermic transfer times up to 3 min will not precondition the isolated rat heart. Journal of Molecular and Cellular Cardiology 1999; 31:503 - 511.

8. Fukunami M and Hearse DJ. The inotropic consequences of cooling:studies in the isolated rat heart. Heart and Vessels 1989; 5:1-9.

9. Kloner RA and Alker KJ. The effect of streptokinase in intramyocardial hemorrage, infarct size, and the no-reflow phenomenon during coronary reperfusion. Circulation 1984; 3:513-521.

10. Curtis MJ. Models for the study of arrhythmias in myocardial ischaemia and infarction: the use of the rat. Journal of Molecular and Cellular Cardiology 1987; 19:399-419.

11. Krebs HA and Henseleit K. Untersuchungen ueber die Harnstoffbildung im Tierkoerper. Hoppe-Seyler’s Zeitschrift fur Physiologische Chemie 1932; 210:33-36.

12. Lowe KC. Perfluorinated blood substitutes and artificial oxygen carriers. Blood Reviews 1999; 13:171-184

13. Qiu Y and Hearse DJ. Comparison of ischemic vunerability and responsiveness to cardioplegic protection in crystalloid-perfused versus blood-perfused hearts. Journal of Thoracic and Cardiovascular Surgery 1992; 103:960-968.

14. Gamble WJ, Conn PA, Kumar AE, Plenge R and Monroe RG. Myocardial oxygen consumption of blood-perfused, isolated, supported, rat heart. American Journal of Physiology 1970; 219:H604-H612.

15. Abraham RGM, Merserau WA and Chiu RC-J. A simplified alloperfused rat heart model for studying myocardial protection. Journal of Investigative Surgery 1988:107-116.

16. Qiu Y, Manche A and Hearse DJ. Contractile and vascular consequences of blood versus crystalloid cardioplegia in the isolated blood-perfused rat heart. European Journal of Cardio-thoracic Surgery 1993; 7:137-145.

17. Bergmann SR, Clarck RE and Sobel BE. An improved isolated heart preparation for external assessment of myocardial metabolism. American Journal of Physiology 1979; 236:H644-H651.

18. Connaughton M. Aspects of ischaemia and reperfusion in the isolated rat heart. A thesis presented for the University of London in fulfilment of the regulations for the degree of Doctor of Medicine 1997.

19. Avkiran M and Curtis MJ. Independent dual perfusion of left and right coronary arteries in isolated rat hearts. American Journal of Physiology 1991; 261:H2082-H2090.

Return to the HELP Table of Contents