Measurement of the End-Systolic Pressure-Volume Relation Using Intraaortic Balloon Occlusion

FIG. 1 is an elevation view depicting the intraaortic balloon catheter and the pneumatic system used to control its inflation; also depicted schematically is the circuitry used to control this apparatus.

FIG. 2 depicts typical pressure and flow data collected using the methods and apparatus of the present invention.

FIG. 3-5 are rectilinear plots of end-systolic values of pressure and volume derived from the data collected using the present invention.

FIG. 6 shows the rapid deterioration of E.sub.max in a globally ischemic heart.

FIG. 7 is a partial pictorial representation of the preferred apparatus of the present invention and its placement within a portion of a heart.

Using the methods and apparatus of the present invention ESPVR is determined from a series of ventricular PV loops obtained under varying afterload conditions. Intraventricular pressure is preferably measured by a catheter-tip manometer; ventricular volume is preferably measured by integrating the flow signal from an electromagnetic flow probe located at the aortic root. However, as explained below, other methods of determining ventricular volume are equally applicable to the present invention. In a most preferred embodiment, changes in afterload are produced by a novel method using the intermittent inflation of a small polyethylene balloon positioned in the ascending aorta. When the balloon is inflated for a single heat beat, the aorta is occluded, and ventricular ejection is interrupted. By timing the inflation of the balloon with respect to a cardiac cycle, it is possible to interrupt ejection at varying points within the ejection phase. This produces intervention beats in which systole commences normally, but terminates in isovolumic contraction at varying ventricular volumes. Since the occluding balloon is only inflated transiently, reflex changes in cardiac contractility are essentially eliminated.

Referring now to FIG. 1, the preferred apparatus used to carry out the methods of the present invention is illustrated. A pneumatically driven intraaortic balloon catheter system is shown generally as 10. The system is operated by a driving cylinder 20, the operation of which is regulated by a solenoid valve 22. The solenoid valve 22 regulates the intake by the driving cylinder 20 from a compressed air source 23, or allows compressed air to be exhausted via a suitable exhaust tube 24.

The motion of the driving cylinder 20 is controlled by a four-way solenoid valve 22, for example, Asco oodel 8342. There are two pneumatic ports 25,26 which enter the driving cylinder chamber on either side of the piston. One of these ports is always connected to the source of compressed air or other gas 23; the other 24, serves as an exhaust port to the atmosphere. The four-way valve 22 switches the compressed air from one side of the piston 20 to the other, causing the actuator rod 27 to move outward (balloon inflation) or inward (balloon deflation).

The inflation volume of the intraaortic balloon 36 is adjusted by setting the distance that the actuator rods 27,31 travel. This is accomplished by adjustable mechanical stops 28,30 which limit the motion of the actuator rods. The mechanical stops have both coarse and fine adjustments. The coarse adjustment is governed by positioning a first stop 28 on the compressor-side actuator rod 27. In a preferred embodiment, the stop 28 is an aluminum disc of sufficient diameter to come into contact with the fine adjustment stop 30. The fine adjustment stop 30 makes use of a micrometer head 29 to allow for precise regulation of cylinder travel.

The pressure of the compressed air source is regulated by a 0-100 psi regulator 19, not illustrated. In operation, the pressure is set so that the balloon 36 can be inflated within a specified time period. Time for inflation is measured by using an electrical switch 33 to detect the contact of the actuator rod 27 against the mechanical stop 28. The time between the turn-on signal to the solenoid 22 to the actuator stop signal may be measured with a digital storage oscilloscope, not illustrated. A regulated pressure of 30 psi is preferably used, and produces balloon inflations within 26.7.+-.4.1 msec. The inflation time of the balloon 36 is actually somewhat less than this, since the measured time includes the delay in solenoid activation.

As shown in FIG. 7, the aortic occluding balloon 36 is a 5 cc polyethylene balloon, with 22mm (inflated) outer diameter, attached to a 8.5 Fr. catheter. This was custom-designed and fabricated by DataScope (Oakland, NJ). The balloon catheter 36 is preferably inserted into the heart 50 and is most preferably positioned approximately 1-2 cm distal to the aortic valve 52 in the aortic root, but proximal to the innominate takeoff. Those of ordinary skill will realize that other placements or placements within other biological pumps are possible and useful in practicing the present invention. FIG. 7 also illustrates those portions of the heart 50 necessary to illustrate the placement of the apparatus of the present invention. Shown are the aortic valve 52, the left ventricle 54, and the aorta 56.

Also illustrated in FIG. 7 is the placement of the electromagnetic flow cuff 60, which is preferably used to gather data from which ventricular volume may be derived. Those of ordinary skill will realize that in clinical or other applications, other methods of estimated or measuring ventricular volume may be employed. For example, in a clinical situation, non-surgical methods of measuring volume, such as radiographic imaging, volume conductance catheters, or flow catheters will preferably be employed.

The positioning of a micromanometer 65 used to generate a signal 72 representative of left ventricular pressure is also seen in FIG. 7. As with the measurement of volume, it will be appreciated that alternative positions and embodiments of pressure measuring apparatus may be employed in order to practice the present invention.

A commercial embodiment of the present invention may combine an annular balloon catheter 36, having volume and pressure measuring apparatus 60,65 integrated into the center of the annulus. The resulting device would essentially be a three channel catheter which would provide the data necessary to determine the ESPVR with a single insertion.

The occluding balloon 36 is inflated pneumatically by an electropneumatic actuator 34. The balloon catheter is connected to the output port of a small bore (1.91 cm) air cylinder 34, such as Aurora Air Products style B-12. When the piston of this air cylinder 34, termed the "inflation cylinder", is pushed by actuator rod 31 inward, air is displaced into the balloon. The motion of the actuator rod 31, in turn, is generated by directing compressed air into the driving cylinder 20 described above. This causes the piston of the driving cylinder 20 to be displaced outward. Since the actuator rods 27,31 of the two air cylinders are connected in series, compressed air applied to the driving cylinder 20 causes air to be displaced out of the inflation cylinder 34 and into the balloon 36. For these air cylinders 20, 34, the actuator rod stroke is 11.43 cm, which corresponds to a maximum possible air displacement into the balloon of 32.55 cc.

A microprocessor based timing system 60 allows the balloon 36 to be inflated at designated times within a cardiac cycle. This can be accomplished, for example, by monitoring the electrical impulses which regulate cardiac activity, for example, by using the signals generated by an electrocardiogram (EKG). However, since the coupling between EKG signals and mechanical contraction is not constant, particularly in pathologic states, the preferred embodiment of the control system of the present invention uses, instead, a ventricular pressure signal to establish the timing reference. Essentially, an analog circuit 70 follows left ventricular pressure, and recognizes when ventricular pressure reaches a set fraction of peak pressure. This function is shown schematically in FIG. 1. The ventricular pressure signal 72 goes to a peak detector 74 whose output passes to an adjustable voltage divider 76; the output, or threshold signal 78, is thus a percentage of peak ventricular pressure. The peak detector circuit 70 has a time constant of 1 sec to filter out sudden fluctuations in peak ventricular pressure. The ventricular pressure signal 72 is then compared with the threshold signal 78 by an analog comparator circuit 80. When the rising ventricular pressure reaches the threshold value, the output of the comparator 80 undergoes a step change which is tied to the input port on a microprocessor controller 60. The controller 60 senses this event within the cardiac cycle and initiates balloon inflation after a programmable time delay. The microprocessor timing system 60 is preferably a FORTH-based microcomputer board. In a most preferred embodiment a New Micros board with a Rockwell R65l2 microprocessor is used. The microprocessor controller 60 also preferably includes internal counter/timer circuitry and multiple I/O ports. It will be understood by those of ordinary skill that the time delay software, is thus written in FORTH and assembly language, and is permanently stored in EEPROM.

The microprocessor based timing system 60 also has the capability to receive programming inputs related to timing parameters, as shown functionally at 64. For example, in a preferred embodiment, the timing system is programmed to produce an alternating sequence of balloon occlusions and control periods. Most preferably, the sequence is comprised of three single beat occlusions, with intervening control periods of non-occluded beats.

The intraaortic balloon occluder of the present invention was tested in six Dorsett sheep with normal, enhanced, and depressed contractile states. End-systolic elastance, E.sub.max was measured in each case and compared with measurements of E.sub.max obtained by other techniques.

Six healthy Dorsett sheep (average weight 26.+-.2.4 kg) were studied. Anesthesia was induced with sodium pentothal (25mg/kg) and was maintained with halothane (0.5-1.0%) after endotracheal intubation and mechanical ventilation (Harvard model 607, tidal volume 12-18cc/kg.). Atropine, 1 mg. I.V., was given Q 1 hr. to control ovine bronchial secretions. D5NS was infused at 5cc/kg/hr plus replacement volume. Packed red blood cells (PRBC) were transfused to keep hematocrit greater than 25.

A median sternotomy was performed and the chest opened with the heart suspended in a pericardial cradle. Additional incisions were made in the neck to expose the external jugular vein and the internal carotid artery for placement of the Swan Ganz catheter and the aortic occluder balloon, respectively. The vagus nerve was left intact bilaterally. A groin incision was also made to expose the femoral artery for subsequent cannulation.

Left ventricular cavitary pressure was monitored with a 5F micromanometer-tipped catheter (Millar Instruments, SPC-350) introduced through the left ventricular free wall. The transducer was interfaced with a Millar (TCB-500) control unit. To measure aortic flow, an electromagnetic flow probe (Biotronix BL-5000) was placed around the root of the aorta distal to the coronary ostia, but proximal to the origin of the innominate takeoff. The probe was operated with a Biotronex pulsed-logic blood flowmeter (BL-610). A 7F Swan-Ganz catheter was placed in the pulmonary artery and used to measure core temperature and thermodilution cardiac output (Gould SP1435). Body temperature was maintained between 36.5.degree. C. to 39.degree. C., and the electrocardiogram was continuously monitored using a Gould (ES-1000B)ECG amplifier.

For each sheep, (n=6), end-systolic elastance was measured with the balloon occluder under three conditions associated with different contractile states: control (normal contractility), epinephrine administration (enhanced contractility), and global ischemia (depressed contractility). After each sheep was fully instrumented, occluder triggering and timing parameters were set. First, the trigger threshold (fraction of peak systolic pressure) was set to a point just before ventricular ejection occurred, at roughly 50% of peak systolic pressure. Second, the timing system was programmed to produce an alternating sequence of balloon occlusions and control periods. The sequence consisted of three single-beat occlusions, with intervening control periods of four non-occluded beats. The occlusions were timed so that they occurred just prior to ejection, at peak systolic pressure, and during the late ejection period. These timings are heart-rate dependent; but for an average heart rate of 100 bpm, the occlusions were typically 0, 80, and 125 msec after the trigger point. Inflation duration was set to extend through the early diastolic phase of the following beat; this period was typically 300 msec.

With these settings, with the heart in the control state, pressure-flow data was recorded during the occlusion sequence described above. For each data acquisition run, an additional 10 seconds of data was recorded both before and after the occlusion sequence. If, during data acquisition, a premature contraction occurred, or the occluder timing was not properly set (e.g. because of a change in heart rate), that data was rejected for subsequent analysis and additional runs were made. A minimum of five acceptable data runs were obtained for each sheep in the normal state.

A state of enhanced contractility was obtained by administering a continuous epinephrine drip at 0.5 g/Kg/min. After 15 minutes, balloon occlusion measurements of end-systolic elastance were obtained using the same protocol employed for the normal heart. Since epinephrine increased the heart rate, the inflation times (delay from threshold trigger and inflation duration) were shortened.

A state of depressed contractility was obtained by 15 minutes of warm global ischemia. First, the animal was placed on full cardiopulmonary bypass (CPB). The CPB system consisted of a blood reservoir, an extracorporeal flow probe (Biotronix, BLX-2096), a combined heat exchanger/hollow fiber membrane oxygenator (0.8M2, Maxima-Johnson & Johnson) and a centrifugal blood pump (Biomedicus 520-C). The CPB circuit was primed with 500cc of saline and 1000cc hetastarch (DuPont). The animal was heparinized (15,000 units load then 10,000 units Qlh), and the CPB circuit was completed by placing a 32F venous cannula (Sarns) in the right atrium and a 14F arterial catheter (USCl) into the femoral artery. The aorta was cross-clamped for 15 minutes and the heart was allowed to fibrillate. The left ventricle was vented only during the ischemic period.

After the cross-clamp was removed, the heart was reperfused on full CPB for a minimum of 20 minutes. The reperfusion period was variable as each individual sheep required different levels of post-ischemic resuscitation before it could be weaned off CPB. Resuscitation included small doses of epinephrine (0.2 mg/hr), lidocaine (50mg bolus followed by a 1 mg/min drip), sodium bicarbonate (1/2 ampule) and, if necessary, defibrillation.

When the animals could be weaned off of CPB, measurements of end-systolic elastance were obtained using the same basic protocol employed for the normal heart. Because of the hemodynamic instability of the globally ischemic ventricle, occluder timing had to be continually readjusted to follow the changing state. Timing was reset by performing 1-2 preliminary data runs to confirm that aortic occlusion occurred at the desired points within the cardiac cycle.

At the conclusion of the protocol, the animals were sacrificed. The hearts were excised; the right ventricular free wall, fat, and valvular tissue were dissected away from the left free wall and the septum, and the remaining left ventricle was weighed. In addition, the ascending aorta, with flow cuff still attached, was removed to calibrate the electromagnetic aortic flow pump, and the flow probe value was calibrated against the extracorporeal flow meter.

For each data run, left ventricular pressure, aortic flow, ECG, and occluding balloon timing signals were simultaneously recorded on a 16 channel electrostatic chart recorder, such as a Gould ES 1000 and digitally sampled by a 16 channel 12 bit analog-to-digital converter board, such as a Tecmar Lab Master in a Sperry IT personal computer. All signals were sampled at 100 Hz and later digitally filtered with a 50-Hz low-pass filter.

Data analysis preferably consists of (1) constructing pressure-volume (PV) loops for normal and occluded beats; and (2) computing the end-systolic pressure-volume relationship (ESPVR). The PV loops use ventricular volume expressed as the volume ejected from the ventricle during the systole. Ejected volume is obtained by digitally integrating the aortic flow signal over the period of a single beat, from end-diastole to end-diastole, although in alternative embodiments direct measurement of ventricular volume can be utilized. End-diastole is defined as occurring when dP/dt, the rate of the pressure rise, reaches 25% of the maximum value. During some balloon occlusions, a large negative flow signal was recorded from the electronic flow meter. This may be the result of the close proximity between the balloon and the flow cuff. Accordingly, the flow signal was digitally clipped to eliminate negative values.

The end-systolic elastance may be computed from a number of occlusion loops along with one non-occlusion loop. In the present Example, three occlusion loops are used. End-systolic elastance, E.sub.max, and unstressed volume, V.sub.o, are determined so as to satisfy the equation:

where P.sub.es and V.sub.es are the end systolic pressure and volume, respectively.

To find E.sub.max and V.sub.o a numerical scheme is employed. First, an initial estimate for V.sub.o is made. Then, for each PV loop, the instantaneous elastance, E(t), is calculated using the equation:

For each loop the P(t) and V(t) associated with the maximum value of E(t) is identified. A linear regression of these PV points is carried out. The volume intercept, the point at which the regression line crosses the volume axis, is taken as the new estimate for V.sub.o. The steps of this method are repeated until V.sub.o converges. The final regression line is the ESPVR, with the volume intercept being V.sub.o and the slope being E.sub.max in equation (1).

The operation of the balloon occluder system can be further understood by referring to FIG. 2. Shown in FIG. 2 are left ventricular pressure (FIG. 2A) and aortic flow (FIG. 2B) traces in which four different afterload conditions are observed in a normal sheep heart. Four heart beats are lablled in FIGS. 2A and 2B as 101, 102, 103 and 104. The first three heartbeats 101, 102, 103 correspond to one of the balloon occlusions and the fourth 104 to a nonoccluded beat. A first beat 101 is a nearly isovolumic beat, where the occlusion occurred slightly after ventricular ejection began. In a second beat 102, occlusion occurred at a point which would ordinarily fall in the mid-ejection period. For a third beat 103, occlusion occurred late in the systole. Therefore, for the case of early occlusion 101, there is very little aortic flow and a substantial increase in peak systolic pressure. In the case where occlusion occurred in the middle of the cycle, 102 aorticflow is reduced in both duration and amplitude, with an integrated flow (ejection volume) about one-half of normal. For this beat 102, there was also an increase in peak ventricular pressure. Finally, the case of late occlusion 103 aortic flow and peak systolic flow is slightly increased above normal. The normal beat 104, is taken as a fourth afterload condition.

Referring now to FIG. 3, there is illustrated the pressure-volume loops computed from the pressure-flow traces illustrated in FIG. 2. The reference numerals 101, 102, 103 and 104 correspond to those heartbeats identified in FIG. 2, the end-systolic points are identified by solid triangles. The volume measure, relative LV volume, is defined as the negative integrated aortic flow, one of ordinary skill will understand that other methods of measuring volume are also applicable to the present invention. Thus, end-diastolic volume is associated with a relative volume of zero. The PV loops for the occluded beats 101, 102, 103 initially follow the loop for the unoccluded beat 104, but make a sharp transitions isovolumic systole as the aorta becomes occluded by the balloon. The end-systolic pressure-volume relation, given by the straight line in FIG. 3, was calculated by the iterative method described above. The four end-systolic points fit the line quite closely, with a correlation coefficient of 0.99. The slope of this curve, or peak elastance, is 3.88 mmHg/ml; and the volume intercept, V.sub.i, is -33.13 ml.

To verify that the operation of the occluder did not affect the baseline contractility of the ventricle, a comparison was made between pre- and post-occlusion beats. Pressure and flow data from the five beats immediately preceding the occlusion sequence were averaged together; pressure and flow data from the five beats immediately following the sequence were similarly averaged. Comparisons were made for P.sub.max, stroke volume, and the area under the pressure-volume loop for 6 normal sheep hearts. The range of variations were 0.48-1.98% for P.sub.max, 0.91-11.61% for stroke volume, and 2.61-9.14% for PV area.

Six healthy sheep, with body weights ranging from 23-32 kg (mean 26.8.+-.3.1 kg), were studied according to the protocol previously described. Hemodynamic parameters and end-systolic elastance, E.sub.max, were measured under normal and epinephrine-enhanced contractility. The parameters are summarized in Table I. As is usual with epinephrine administration, the heart rate, peak left ventricular pressure (LVP), maximum dP/dt, cardiac output, and stroke volume all increased. With epinephrine infusion, heart rate increased, on average, from 108.05 to 135.07 bpm. In spite of shortened systolic period with these faster heart rates, the balloon occluder system continues to function without any difficulty. Stroke volume increased from 19.25 to 25.25 ml; this improves volume measurements since the signal-to-noise ratio of the aortic flow is increased. Similarly, the peak LVP increased from 93.82 to 111.70 mmHg. Added to Table I are the body weights and LV weights for each of the experimental animals.

Pressure-volume loops for a typical experiment with enhanced contractility are shown in FIG. 4. This is the same heart as shown in FIG. 3. Four different afterload conditions are shown as before. The end-systolic points are virtually collinear (correlation coefficient=0.99). The end-systolic elastance for this example is 7.74 mmHg/ml, compared to 3.88 mmHg/ml for the normal contractile state, nearly a two-fold increase in contractility.

Table II summarizes the measurement of the end-systolic pressure-volume relationship. For each animal, E.sub.max is given for both contractile states, along with the volume intercept, V.sub.i, and statistical information. Contractile states are compared using two types of regression analyses to calculate ESPVR. In the first, under the heading "Mean" in Table II, E.sub.max, V.sub.i, and the correlation coefficient were computed for each occlusion sequence and the results averaged together. The standard deviation for E.sub.max, therefore, describes the reproducibility of the elastance measurement. Standard deviations were in the range of 0.14 to 0.72 mmHg/ml in the normal heart, indicating that at a given contractile state the measurement is very repeatable.

The second regression method, under the heading "Pooled" in Table II, takes the entire set of end-systolic points from all occlusion sequences associated with a given contractile state and animal, and computes E.sub.max, correlation coefficient, and V.sub.i for the pooled data.

Statistical comparisons were performed on both mean and pooled E.sub.max, and V.sub.i. Each of the distributions were first found to be normal using the Wilk-Shapiro test for normality. Then a two-tailed paired t-test was employed to compare the enhanced state with the normal state. The mean peak elastance, E.sub.max, was significantly different (p=0.004); while the pooled E.sub.max, also significantly different, had a lower p-value (0.039). There was no statistically significant change in V.sub.i between the two contractile states in either mean or pooled cases.

All of the animals studied under normal and enhanced contractile states were subjected to global ischemia, as described above and studied with the aortic balloon occluder. Although the apparatus continued to function for severely depressed ventricles, it was not always possible to determine an ESPVR. In some cases, ventricular ejection was so limited that partial aortic occlusion, in absolute terms, was only marginally reduced. In other cases, although larger changes in ventricular ejection could be produced, only negligible changes in peak ventricular pressure were observed. Even under the best circumstances the range of end-systolic pressure-volume points was small, as shown in FIG. 5. This reduced the quality of the linear regression to the end-systolic points.

FIG. 5 shows results obtained from one sheep approximately 21 minutes after cardio-pulmonary bypass was removed, and epinephrine infusion was stopped. Here, E.sub.max was 1.63 mmHg/ml, compared to 2.56 mmHG/ml in the normal state. This measurement was one of a series of measurements taken after mechanical and pharmacological support of the heart were removed. The entire series is summarized in FIG. 6, which shows E.sub.max over a 25 minute time period. The rapid deterioration in contractility is apparent, and it indicates the difficulty in assigning one particular value to E.sub.max in these severely depressed ventricles. However, this data demonstrates that the balloon occluder measurement technique can be used to follow the time course in E.sub.max even when contractile state is changing quickly.

One of ordinary skill in the art will appreciate, that although particular embodiments of apparatus and certain method steps have been described in detail, variations employing the spirit of the present invention are possible. The full scope of the present invention may be determined by reference to one or more of the appended claims.