A new method of retrograde cardioplegic administration

J

THORAC CARDIOVASC SURG 1990;99:335-44

A new method of retrograde cardioplegic administration Right ventricular protection by right atrial perfusion cooling Retrograde administration of cardioplegic solution via the right atrium with continUOlti cooling of the right ventricular cavity (right atrial perfusion cooling) was assessed for its protective effect in 12 dogs with occlmion of the right coronary artery subjected to global ischemia for 60 minutes. After an initial administration of 4° C crystalloid cardioplegic solution by antegrade aortic perfusion, myocardial protection was established either by right atrial perfusion cooling (group I; n = 6) or by antegrade aortic perfusion alone (group D; n = 6). The right ventricular temperature was approximately 15° C in group I and 20° C in group D. After ischemia for 60 minutes, the adenosine triphosphate content of the right ventricular free wall was significantly higher in group I than in group D (24.4 ± 1.45 versus 13.8 ± 2.34 Itmol/gm dry weight, p < 0.05). The percent recovery of right ventricular contractility, which was evaluated by end-systolic pressure-volume relationships, was significantly better in group I at each reperfusion period (30 minutes: 130.0% ± 9.6% versus 86.1 % ± 11.8%, p < 0.05; 60 minutes: 159.6% ± 12.9% versus 96.5% ± 20.1 %, p < 0.05). Postischemic right ventricular stiffness (reciprocal value of compliance) increased in group D compared with group I, although the difference was not statistically significant There were no major differences in percent recovery of the left ventricular end-systolic pressure-volume relationships between the two groups. The evidence suggests that the right atrial perfusion cooling method produces excellent right ventricular protection.

Yuichirou Nakamura, MD, Kiyotaka Fukamachi, MD, Munetaka Masuda, MD, Toshihide Asou, MD, Yoshihiro Toshima, MD, Masahiro Oe, MD, Atsuo Mitani, MD, Kazuhiko Kinoshita, MD, Yoshito Kawachi, MD, Jiro Tanaka, MD, and Kouichi Tokunaga, MD, Fukuoka, Japan

T

he presence of coronary stenosis restricts adequate deliveryof cardioplegic solution beyond the diseased area of the myocardium. Retrograde cardioplegia-eoronary sinus perfusion!" or right atrial (RA) perfusiont-e-has been proposed to improve this ma1distribution of cardioplegic solution to the region supplied by the diseased left coronary artery. Our previous study clearly showed

From the Division of Cardiovascular Surgery, Research Institute of Angiocardiology, Faculty of Medicine, Kyushu University, Fukuoka, Japan. Received for publication July 13, 1988. Accepted for publication March 13, 1989. Address for reprints: Yuichirou Nakamura, MD, Division of Cardiovascular Surgery. Research Institutee of Angiocardiology, Faculty of Medicine, Kyushu University, Higashi-ku, Fukuoka 812, Japan.

12/1/13088

that retrograde cardioplegia often fails to adequately perfuse the right ventricular (RV) free wall and ventricular septum.' Such maldistribution to the RV myocardium in the presence of right coronary stenosis, therefore, appears to be unavoidable when conventional cardioplegia techniques are used. With this problem in mind, we combined topical cooling of the R V with retrograde cardioplegia to provide a more adequate cardioplegia technique for multiple coronary stenoses. It is not easy to obtain sufficient topical cooling of the RV because of its position.v" Several unique methods have been designed for RV topical cooling.v!? Continuous cooling of the RA and RV cavities was originally advocated by Braimbridge.!' This topical cooling method ensures the protection of both the RV free wall and the septum, which have an area of poor perfusion when retrogradecardioplegia is used. This method was technically applicable to RA perfusion, which is a safe and simple 335

The Journal of

336

Thoracic and Cardiovascular

Nakamura et al.

Surgery

RA Perfusion Cooling (Group I ) A Pressure monitor

RV cooling phase

~JlH.i'h9b"'~'" MPA occlusi on

B Pressure monitor

mmH9C=

RA perfusion phase

20~

I

~l\yJ(r{9i('l'.'o> MPA occlusion

Heat~~ _-~~~~~ Exch~ :L

Cardioplegia

Reservoir

Fig. 1. Schematic diagram of the RA perfusion cooling system (group I). A, RV cooling phase of RA perfusion cooling. The tourniquets on the superior and inferior vein and main pulmonary artery (MPA) are tightened. Irrigation with 4° C cardioplegic solution is performed continuously by the commercial cardioplegia delivery system. During irrigation, the low pressure (5 mm Hg) is maintained by controlling the pump flow. In this condition, cardioplegic solution is not perfused into the coronary sinus. B, RA perfusion phase of RA perfusion cooling. Intermittent administration of cardioplegic solution through the coronary sinus can be easily performed by clamping the R V cannula. The perfusion pressure is raised to around 20 mm Hg. The volume of cardioplegic solution administered is determined by the decreased volume in the cardioplegia reservoir. The aortic cannula used for antegrade aortic perfusion remains as a vent root.

retrograde cardioplegia technique developed by Fabiani and associates." The purpose of the present study is to examine the protective effect of our new method, RA perfusion plus cooling, on the right side of the heart in a right coronary occlusion model while comparing it with antegrade aortic perfusion.

Materials and methods Surgical preparation. Twelve mongrel dogs weighing 12 to 20 kg were anesthetized with intravenous pentobarbital 25 rng/kg and their lungs were mechanically ventilated by a volume-limited ventilator. A catheter was inserted into the femoral artery to monitor the systemic arterial pressure. After standard sternotomy and systemic heparinization, an arterial can-

nula was inserted into the right carotid artery and venous cannulas into the superior and inferior venae cavae. The dogs were placed on total cardiopulmonary bypass with a bubble oxygenator primed with fresh heparinized homologous blood. The left ventricle (LV) was vented via its apex. Experimental protocol. Two experimental groups were designated: Group I (n = 6) was protected during the 60 minutes of cardioplegic ischemia with right coronary occlusion by the RA perfusion cooling technique. Group II (n = 6), the control group, was protected during the same periods by antegrade aortic perfusion alone. The right coronary artery was reversibly occluded I cm distal to its origin during the 60 minutes of cardioplegic arrest. Cardiac arrest was initially induced by antegrade administration of 100 ml of cardioplegic solution through a 14-gauge cannula at a perfusion pressure of 50 mm Hg. Then the dogs were divided into two groups. Intrapericardial topical cooling was not undertaken in either group. In group II (control group), antegrade aortic perfusion alone was used for both the initial administration of 100 ml of cardioplegic solution and the administration of an additional 100 ml at 3D-minute intervals. R V cooling was not performed in this group. In group I, an equivalent volume of cardioplegic solution was administered by RA perfusion cooling. The technique of RA perfusion cooling is shown in Fig. I. Two cannulas, one for infusion and the other for drainage of cardioplegic solution, were inserted into the RA and RV, respectively. A 23-gauge needle was inserted into the R V cavity to monitor pressure. Continuous cooling of the R V was achieved as follows: After the main pulmonary artery was occluded, the cardioplegic solution was infused into the RA with a roller pump and drained through the cannula placed in the RV cavity at an R V pressure of 5 mm Hg (RV cooling phase, Fig. I, A). While the cardioplegic solution was being administered, the cannula at the drainage side was occluded and the RA pressure was elevated to 20 mm Hg (RA . perfusion phase, Fig. I, B). The intermittent cardioplegic dose of 100 ml was perfused at 3D-minute intervals. The RV cavity was cooled continuously except for 10 minutes before reperfusion, to minimize any difference between the myocardial temperature and the perfusate temperature.P Before reperfusion, the RV cavity was emptied by discarding the cardioplegic solution through the drainage cannula, and the tourniquet occluding the main pulmonary artery was released. The aortic root was vented through the cannula used for antegrade aortic perfusion. In both groups, the perfusion temperature was kept at 36° C during the control period, cooling to 30° C, at which point the aorta was clamped. Temperature then fell to 25° C, was maintained at that level during ischemia, and then increased to 30° C 10 minutes before reperfusion. The perfusion pressure was maintained at around 70 mm Hg. The level of pH, oxygen tension, carbon dioxide tension, and base excess were kept within the physiologic range by regulating the gas flow and adding sodium bicarbonate. The room temperature was kept at between 22° and 25° C. After reperfusion, the perfusion temperature was gradually increased from 30° to 36° C for 20 minutes. Cardiopulmonary bypass was temporarily discontinued to measure ventricular function before ischemia and at 30 and 60 minutes after reperfusion. Cardioplegic solution (NaCl 77 mrnol/ L, KCI 20 mmol/L, NaHC03 10 rnmol/L, CaCl z 0.2 mrnol/ L, glucose 25 gm/L, insulin 5 units/L, and osmolarity 360 mOsm/L) at 4° C was used. Pericardial topical cooling was not used. Electrical defibrillation was performed when the heart

Volume 99

RA perfusion cooling

Number 2 February 1990

A

B clamp

!L~

*

Emax (RV or LV)

Pressure

t

Emax

;r~ i J•.•• . ~• •.• . . . •. . • • . . .

Vo

Ved

---.. Volume

Fig. 2. Measurement of Emax. A, Representative recordings displaying the measurement of Emax. The artery is clamped in diastole after ejecting contraction produces two beats with the same preload, ejecting (open circle) and isovolumic (closedcircle). Successful arterial clamping is ascertained by the lack of simultaneous elevation of arterial pressure or arterial flow. B, Schematic diagram of Emax. The pressure-volume loop of two beats (open and closed circles), obtained by the arterial clamp, can be drawn as shown in the figure. The decline of the straight line depicted from the peak of the isovolumic beat to the left upper corner of ejecting beats is Emax. Ved, End-diastolic volume; Vo, volume axis intercept of end-systolic pressure-volume relationship.

*

RV stiffness

(=i=

EDP Ved - Vo )

Pressure

i

ESPVR (Emax )

.. OM

EDP

~--

Vo

*

Ved

--- --- -----+

EDPVR

Volume

Fig. 3. Emax is represented as a monoexponential curve (brokenline). The value of RV stiffness is obtained by calculating EDP (Ved-Vo). Ved, End-diastolic volume; EDP, end-diastolic pressure; Vo, Ved, end-diastolic volume; Va. volume axis intercept of end-systolic pressure-volume relationship (ESPVR).

337

The Journal of Thoracic and Cardiovascular Surgery

3 3 8 Nakamura et al.

35 -0

Group I Group II

30 w

...J

....a:z

25

....

20

u

w

w

a:

:::J f-

«

> :I:

(.:J

a:

a:

15

w

c, ~

10

w

'I'

f...J

35

«

-,

0

5

15

30 35

45

Rep.

0

a: u

« 0

>~

30 w

u

25

zw >

20

...J

....a:

....u, W

...J

15 10

o

5

15

30 35

45

Rep.

Minutes

Fig. 4. Myocardial temperature of both ventricle during 60 minutes of global ischemia in the presence of right coronary occlusion. Group I is the RA perfusion cooling group and group II, the antegrade aortic perfusion group.

fibrillated during reperfusion. After 20 minutes of empty beating, the tourniquets on the superior and inferior venae cavae were released and the vent tube was removed. The hearts were rewarmed while on the partial assist bypass except for temporary interruptions for measurement of cardiac function. Data acquisition. The myocardial wall temperature was monitored with a thermistor-tipped probe inserted into the middle portion of the LV free wall I ern lateral to the left anterior descending artery and the middle portion of the R V free wall in all dogs. Myocardial biopsy specimens from the RV and LV free walls were taken with a Travenol biopsy needle (Travenol Laboratories, Inc., Deerfield, Ill.) during the control state, at the end of ischemia, and both 30 and 60 minutes after reperfusion. A sample was immediately cooled with liquid nitrogen and preserved for later analysis. The myocardial contents of adenosine triphosphate (ATP) were measured by high-performance liquid chromatography (Shimazu, Kyoto, Japanj.P During the control state, as well as at 30 and 60 minutes of reperfusion, functionn of both ventricles was evaluated by ana-

lyzing pressure-volume relationships (Emax). Diastolic function (RV stiffness, a reciprocal index of compliance) was also evaluated. Measurement of Emax. The slope of the end-systolic pressure-volume relationship (Emax) has gained popularity as a load-independent index of contractility'v'" and was used in this study as an index of the myocardial protective effect of cardioplegia. In both groups, RV and LV pressures were measured and their first derivatives (dP/dt) were obtained by a cathetertip micromanometer (MPC-500, Millar Instruments, Inc., Houston) inserted through the R V outflow tract and LV apex. With a 14 to 18 mm electromagnetic flow probe (models FR l40T, 160T, and l80T, Nihon Kohden, Japan) on the main pulmonary artery and ascending aorta, the blood flow in the main pulmonary artery and aorta were measured with a flowmeter (MFV-2100, Nihon Kohden, Japan). While systemic mean pressure was being maintained around 70 mm Hg by adequate volume loads and mechanical ventilation was discontinued, the ascending aorta or main pulmonary artery was occluded in diastole after an ejecting beat with an atraumatic clamp (a tourniquet for occlusion of the main pulmonary artery). A pair of beats was obtained with the same preload. On both ejection and isovolumic contraction (Fig. 2, A). We clamped twice for R V Emax and three timess for LV Emax measurement over a period of 15 seconds. By repeating these procedures, we obtained more than six sets of beats from both ventricles. The data were recorded on an eight-channel, forced ink-pen oscillograph (Nihon Denki Sanei Co. Ltd, Tokyo) at a paper speed of 50 mrn/sec and stored on floppy disk memory with a microcomputer system (PC 9801, NEC, Tokyo). The principle for evaluating Emax is the same as that described by Igarashi and Suga.'? Instantaneous R V and LV volumes during ejection just before arterial clamping were obtained by the integral of aortic flow (main pulmonary flow) and, consequently, a pressure-volume curve can be drawn. Because the isovolumic contraction implies constant end-diastolic volume, a straight line can be drawn from the peak pressure point of isovolumic contractions to the left upper corner of the pressure-volume loop of the ejecting beat. The slope of the line is defined as Emax (Fig. 2, B). The data obtained at the same period were averaged: Any inappropriate data caused by an incomplete arterial clamp were excluded from the study. Measurement of RV stiffness. The end-diastolic pressure-volume relationship was also dctcrmincd.l" The end-diastolic pressure-volume relationship of the R Vat pericardiotomy was expressed as being linear within the low end-diastolic range.'? ED P/ (Ved-Vo) was defined as right ventricular stiffness in each Emax measurement (Fig. 3). Animal care. All animals received human care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society of Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 80-23, revised 1978) Data analysis. The obtained values were expressed as mean ± standard error of the mean. The difference in values between groups was analyzed by Student's t test. Emax values were expressed as percent recovery of control value and RV stiffness as percent change from the control values for comparison and statistical analysis, because these indexes vary individually according to the ventricular geometry. A p value under 0.05 was considered to be statistically significant.

Volume 99 Number 2

RA perfusion cooling 3 3 9

February 1990

RV

LV

*

p

< 0.05

20

....c

....c: Q)

oo

10

£L.

I-

«

10

o

Group I

•

Group II

T

T

c

c

30 min 60 min

end of ischemia

30 min end of ischemia

60 min

Fig. 5. Sequential change of ATP in both ventricular free walls during 60 minutes of global ischemia and 60 minutes of reperfusion in the presence of right coronary occlusion. Group I is the RA perfusion cooling group and group I I, the antegrade aortic perfusion group. *p < 0.05, group I versus group II at the same time.

x

co

(%)

0

Group

~

Group II

E

*

w

> a:

* p <0.05

150

...-

*

o >- 100 ------ -------------------- --- -------------~ Q) >

o o

Q) ~

......

50

C

Q)

o

~

Q)

a..

o

Reperfusion

Repe rfus ion

30 min

60 min

Fig. 6. Percent recovery of RV Emax after reperfusion for 30 minutes and 60 minutes. Group I is the RA perfusion cooling group and group II, the antegrade aortic perfusion group. Recovery of Emax is expressed as percentage of preischemic level. p < 0.05, group I versus group II.

The Journal of

340

Thoracic and Cardiovascular

Nakamura et al.

Surgery

o ><

Group

~ Group II

(%)

ALL N. S.

CO

E

UJ

-

> ..J

150

0

>-

'Q)

100

> o Q)

0

'+-'

C

50

Q)

o

'Q)

n,

o

Reperfusion

Reperfusion

30 min

60 min

Fig. 7. Percent recovery of LV Emax after reperfusion for 30 minutes and 60 minutes. Group I is the RA perfusion cooling group and group II, the antegrade aortic perfusion group. Recovery of Emax is expressed as percentage of preischemic level. N.S., Not significant.

c:::J

Group

I

~

Group

Il

ALL N. S.

(/) (/)

-Q)

c

o -+-' Q) 0)

100

(/)

c:> roo:

..c: o

+-'

50

c: Q) o ~

rf

0 Reperlusion

30

min

Reperfusion

60

min

Fig. 8. Percent change of RV stiffness (reciprocal value of compliance) after reperfusion for 30 minutes and 60 minutes. At reperfusion, RV stiffness was increased (compliance decreased) in both groups. N.S., Not significant.

Volume 99

RA perfusion cooling 3 4 I

Number 2 February 1990

C \

A",:. B ,-.:

Fig. 9. Triple-lumen cannula for RA perfusion cooling. The cardioplegic solution is infused from A and enters into the right atrium from A'. The infused solution irrigates the Rt/cavity, is collected from the RV cavity (B'), and is returned to the cardioplegic reservoir from B. The irrigation or perfusion pressure of the cardioplegic solution is monitored by the right atrial pressure line (C). Results Myocardial temperature (Fig. 4). The temperature of the RV in group I dropped below 15 Cat 30 minutes of coronary occlusion because of R V irrigation. It was about 20° to 22° C in group II throughout the ischemic period. The temperature of the LV did not differ between groups. Without intrapericardial topical cooling, LV temperature ranged from 20 to 24° C during cardioplegic arrest. Thus RA perfusion cooling exerted a marked cooling effect on the R V in the presence of right coronary occlusion. Myocardial ATP content (Fig. 5). The ATP content of the RV was significantly higher in group I at the end of ischemia (24.4 ± 1.45 versus 13.8 ± 2.43 ~mol/ gm dry weight, p < 0.05). The ATp content of the LV at the end of ischemia showed no significant differences between them. On reperfusion, the ATP content dropped in all hearts, but not to a statistically significant degree. Ventricular function (Figs. 6, 7, and 8). The percent recovery of R V Emax was significantly higher in group I than in group II after reperfusion: at 30 minutes, 130.0% ± 9.6% versus 86.1% ± 11.8% (p < 0.05); at 60 minutes, 159.6% ± 12.9% versus 96.5% ± 20.1% (p < 0.05) (Fig. 6). Hence RA perfusion cooling provided a significant RV protective effect despite the presence of right coronary occlusion. The percent recovery of LV Emax was not statistically different, but tended to be higher in group I: at 30 minutes, 122.8% ± 11.5% versus 89.4% ± 12.0%); at 60 minutes, 137.8% ± 21.0%versus 97.5% ± 11.3% (Fig. 7). The percent change of R V stiffness during reperfusion was +64.4% ± 11.9% at 30 minutes and +58.1% ± 12.4% at 60 minutes in group I, the R V stiffness increasing by about 1.6 times that of the control value at reperfusion. In group II these figures were 0

0

+96.0% ± 10.7% at 30 minutes and +85.6% ± 20.8% at 60 minutes, with the RV stiffness increasing by about 1.9 times that of the control value at reperfusion. There were no statistically significant differences between the two groups (Fig. 8). Thus the diastolic component was impaired in both groups during reperfusion. RA perfusion cooling, however, showed no deleterious effects on the R V diastolic properties under study. Discussion The difficulty of adequate right ventricular protection in the presence of right coronary stenosis has been mentioned by Rabinovitch.P Christakis," and their associates in their clinical reports of the occurrence of selective R V failure after multiple coronary artery bypass grafting. There are two reasons for this problem: (l) It is difficult to perfuse the distribution of the diseased right coronary artery by conventional cardioplegia techniques; (2) adequate topical cooling of the R V during the operative procedure is not easy to achieve. In the present study, we have proposed a new cardioplegic technique with an adequate RV protective effect in cases of high-degree right coronary stenosis. Retrograde cardioplegia has been proposed for the distribution of the diseased left coronary artery. 1,2 We 3 previously demonstrated the superior LV protective effect of this method in the presence of an occluded left coronary artery in a dog model. However, its application for diseased areas in the right coronary distribution might be meaningless, because this method per se often fails to provide adequate perfusion to the R V free wall and ventricular septum. The coronary blood flow of the R V free wall and a large part of the ventricular septum is not drained into the coronary sinus but directly into the R V and LV cavities via the anterior cardiac vein or thebesian

The Journal of

342 Nakamura et al.

vein individually.t Thus an adequate protectiveeffect on the distribution of the diseased right coronary artery might be best achievedby using a sufficient topical cooling technique. On the other hand, the RV temperature tends to increaseand effective topicalcoolingis not easy to attain, because ofthe anterior locationof the RV.6-8 Most techniques intended for RV cooling require special equipment.v'? In the case of coronary bypass, however, someof these methods interfere with the operativeprocedures. In contrast, topicalcoolingof the RV cavity,advocated by Braimbridge,11 appears to be the most applicable method. This coolingmethod does not disturb the operative procedure,and irrigation is still performedduring situations such as ventricular elevation for bypass grafting to the circumflexartery. This method wouldimprove protection of both the ventricular septum and RV free wall by retrograde cardioplegia. The RV irrigation method is easilycombinedwith the retrograde RA perfusion cardioplegic technique. RA perfusionis an alternative retrograde method,introduced by Fabiani and colleagues" as a safeand simpletechnique. A preliminary assessmentof this method of RV cooling showed it was still inadequate, because the temperature of the solutionin the RV increasedup to the environmental or systemictemperature beforethe next infusion. This rewarmingof the cardioplegicsolutionsometimesseemed to result in poor ventricular protection, as Salter22 suggests. It was assumed that continuous irrigation of the cardioplegicsolutionmight reduce the rewarmingof cardioplegic solution stored in the RV cavity. The present study showed the stable topical coolingeffect on the RV by the continuous coolingof the RV cavity. RV temperature decreased below 15° C. This coolingeffect is controllable by the appropriate use of an adequate apparatus or irrigation cannula. The significant RV protection afforded by RA perfusion cooling was assessed by the pressure-volume relationship.Emax is a load-independentindex of contractility, whereas other indexes, such as cardiac output, ejection fraction, stroke work, and dP jdt, are largely influencedby loadingconditions. The difficult part of this measurement, which includesthe end-diastolicpressurevolume relationship,was making an accurate estimation of ventricular volume. It was especially difficult since there is still no ideal means of measuring RV volume, because of the geometric complexityof the ventricle.P The method developed by Igarashi and Suga 17 permitted us to estimate Emax without any direct measurement of ventricular volume in the in situ heart. Ernax is such a sensitive and specific index of contractility that it reflects

Thoracic and Cardiovascular SurQEllY

subtle damage to the myocardium. By means of this index, the superiority of RA perfusion cooling has been clearly established. The diastolicproperty of the RV is another important determinant of ventricular pump function. In the present study, the diastolic property of the RV was impaired by reperfusion even with RA perfusion cooling. This suggested a discrepancybetween the recovery factor of contractility and compliance. On reperfusion, the compliance decreases (stiffness increases) by the mechanical effect (garden hose effect caused by coronary hyperemia) and metabolicchanges at reperfusion (increasedintracellular calcium or failure of ATP synthesisj.i" It was assumed that protection from reperfusion injury was essential for the preservation of postischemic diastolic function. Some interesting implications can be presumed from the simultaneousassessmentof LV Emax in the present study. The percent recovery of LV Emax in group I tendedto be better than in group II. This wouldclarifythe adequate LV protectiveeffectof retrograde cardioplegia. The improved protection of the LV by RA perfusion coolingin the present study is consideredto be due to the influence ofa significant protectiveeffectof the RVon LV function. Functionof one ventriclealwaysinfluences that of the other, so that failure of one is alwaysaccompanied by failure of the other. The significant septal protection provided by RA perfusion coolingmay be another factor. The ventricular septum is a componentof both ventricles and plays an important role in ventricular interdependence. Injury to the septum makes it easy to transmit failure from one ventricle to the other and cause biventricular failure.25-27 Successful protection of the LV is considered to be the result of adequate protectionby RV cooling, not onlyfor the RV free wall but alsofor the ventricular septum. RV distentionis caused by the infusionof cardioplegic solution. Salter-?claimedin hisexperimentalstudyofRA perfusion that RV functionmight becompromised by RV distention,probablybecauseofa disruptionof myocardial sarcomeres. However, he used a perfusionpressureof 60 mm Hg for his RA perfusion. This is too high a pressure for retrograde cardioplegiaand is likelyto injure the coronary venous system.28. 29 We regard the optimal perfusionpressurefor retrograde cardioplegiato be 20 mm Hg, whichshould be enoughto deliveran adequate amount of cardioplegic solution into the LV myocardium in the presenceof left coronary stenosis.' We applied the same perfusion pressure for RA perfusion in RA perfusion cooling. Our evaluation of RV function with pressurevolume relationships showed no significant deleterious change in RV contractility or compliance with RA per-

Volume 99 Number 2 February 1990

fusion cooling. It was assumed that the myocardial injury was induced by the ischemia insult rather than mechanical stretching. For clinical application, we designed a triple-lumen catheter (Fig. 9). This cannula may be convenient when RA perfusion cooling is being performed by single cannulation through the RA. Monitoring the perfusion pressure of the cardioplegic solution is necessary not only to avoid RV overdistention but also to avoid inadequate protection resulting from minor leakage of cardioplegic solution, such as from a patent foramen.P An adequate perfusion pressure ensures the delivery of retrograde cardioplegia to the LV myocardium. If adequate pressure cannot be obtained, some other technique must be used. We conclude that RA perfusion cooling protects the RV by providing stable cooling of the R V free wall and ventricular septum. This method is simple and may be applicableto multiple coronary bypass operations complicated by high-grade stenosis in the right coronary artery. We gratefully acknowledge the helpful suggestions and support of Dr. M. V. Braimbridge, St. Thomas' Hospital, London, in completing the manuscript, and the assistance of Miss M. Nakanishifor biochemical analysis. We also thank Mr. B. T. Quinn, Kyushu University, for helpful comments. REFERENCES I. Bolling SF, Flaherty JT, Bulkley BH, Gott VL, Gardner TJ. Improved myocardial preservation during global ischemia by continuous retrograde coronary sinus perfusion. J THORAC CARDIOVASC SURG 1983;86:659-66. 2. Gundry SR, Kirsh MM. A comparison of retrograde cardioplegia versus antegrade cardioplegia in the presence of coronary artery obstruction. Ann Thorac Surg 1984;38: 124-7. 3. Masuda M, Yonenaga K, Shiki K, Morita S, Kohno H, Tokunaga K.. Myocardial protection in coronary occlusion by retrograde cardioplegic perfusion via the coronary sinus in dogs. J THORAC CARDIOVASC SURG 1986;92:255-63. 4. Fabiani IN, Deloche A, Swanson J, Carpentier A. Retrograde cardioplegia through the right atrium. Ann Thorac Surg 1986;41:101-2. 5. Shiki K, Masuda M, Yonenaga K, Asou T, Tokunaga K. Myocardial distribution of retrograde flow through the coronary sinus of the excised normal dog heart. Ann Thorae Surg 1986;41:265-71. 6. Fisk RL, Ghaswalla D, Guibeau EJ. Asymmetrical myocardial hypothermia during hypothermic cardioplegia. Ann Thorac Surg 1982;34:318-23. 7. Gonzalez AC, Brandon TA, Fisk RL, et al. Acute right ventricular failure is caused by inadequate right ventricular hypothermia. J THORAC CARDIOVASC SURG 1985; 89:386-99.

RA perfusion cooling 3 4 3

8. Daily PO, Pfeffer TA, Wisniewski JB, et al. Clinical comparison of methods of myocardial protection. J THORAC CARDIOVASC SURG 1987;93:324-36. 9. Hurley EJ, Lower RR, Dong E Jr, Pillsburg RC, Shumway NE. Clinical experience with local hypothermia in elective cardiac arrest. J THORAC CARDIOVASC SURG 1959; 109:750-4. 10. Rosenfeldt FL, Fambiato A, Pastoriza-Pinol J, Stirling GR. A recirculating cooling system for improved topical cardiac hypothermia. Ann Thorac Surg 1981;32:401-5. 11. Braimbridge MV, Cankovic-Oarracott S, Hearse OJ. Crystalloid cardioplegia: experience with the St. Thomas solution. In: Engelman MR, Levitsky S, ed. A textbook of clinical cardioplegia. New York: Futura, 1982:177-98. 12. Kohda Y, Tominaga R, Ueno Y, Andoh H, Nakano E, Tokunaga K. Effects of cardioplegia on myocardial metabolism during hypothermic ischemic: components and mode of administration. Jpn Circ J 1985;49:75-80. 13. Karniike W, Watanabe F, Kawashima Y, et al. Change in cellular levels of ATP and its catabolites in ischemic rat liver. J Biochem 1982;91:1349-56. 14. Suga H, Sagawa K. Instantaneous pressure-volume relationships and their ratio in the excised, supported canine left ventricle. Circ Res 1974;35:117-26. 15. Suga H, Sagawa K, Shoukas AA. Load independence of the instantaneous pressure-volume ratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio. Circ Res 1973;32:314-22. 16. Moughan WL, Shoukas AA, Sagawa K, Weisfeldt ML. Instantaneous pressure-volume relationship of the canine right ventricle. Circ Res 1979;44:309-15. 17. Igarashi Y, Suga H. Assessment of slope of end-systolic pressure-volume line of in situ dog heart. Am J Physiol 1986;250:685-92. 18. Asou T, Morita S, Tanaka J, Tokunaga J, Sunagawa K. Measurement of end-diastolic pressure-volume relation of in vivo canine heart: a new method for estimating effective left ventricular volume. Circulation I986;74(Pt 2):II443. 19. Janicki JS, Weber KT. Factors influencing the diastolic pressure-volume relation of the cardiac ventricles. Fed Proc 1980;39:133-9. 20. Rabinovitch MA, Elstein J, Chiu RCJ, Rose CP, Arzoumanian A, Burgess JH. Selective right ventricular dysfunction after coronary artery bypass grafting. J THORAC CARDIOVASC SURG 1983;86:444-50. 21. Christakis GT, Fremes SE, Mclaughlin PR, et al. Right ventricular dysfunction following cold potassium cardioplegia. J THORAC CARDIOVASC SURG 1985;90:243-50. 22. Salter DR. Ventricular function after atrial cardioplegia. Circulation 1987;76(Pt 2):VI29-40. 23. Morris JJ III, Wechsler AS. Right ventricular function: the assessment of contractile performance. In: Fisk RL, ed. The right heart. Cardiovascular clinics, 1987. Phildelphia: FA Davis, 1987:3-18. 24. Apstein CS, Grossman W. Opposite initial effect of supply and demand ischemia on left ventricular diastolic compli-

3 44

Nakamura et al.

ance: the ischemia-diastolic paradox. J Mol Cell Cardiol 1987;19:119-28. 25. Fixler DE, Monroe GA, Wheeler JM. Hemodynamic alterations during septal or right ventricular ischemia in dogs. Am Heart J 1977;93:210-5. 26. Tani M. Roles of the right ventricular free wall and ventricular septum in right ventricular performance and influence of the parietal pericardium during right ventricular failure in dogs. Am J Cardiol 1983;52:196-202. 27. Maughan WL, Sunagawa K, Sagawa K. Characterization of ventricular interdependence using a three-element volume elastance model studied in isolated canine ventricles. Am J Physiol [In press].

The Journal of Thoracic and Cardiovascular Surgery

28. Poirier RA, Guyton RA, McIntosh CL. Drip retrograde coronary sinus perfusion for myocardial protection during aortic cross-damping. J THORAC CARDIOVASC SURG 1975;70:966-73. 29. Hammond GL, Davies AL, Austen WG. Retrograde coronary sinus perfusion: a method of myocardial protection in the dog during left coronary artery occlusion. Ann Surg 1967;166:39-47. 30. Gundry SR. Retrograde cardioplegia. Ann Thorac Surg 1987;43:121.

Bound volumes available to subscribers Bound volumes of THE JOURNAL OFTHORACIC AND CARDIOVASCULAR SURGERY are available to subscribers (only) for the 1990 issues from the Publisher, at a cost of$58.00 ($75.00 international) for Vol. 99 (January-June) and Vol. 100 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact The C.V. Mosby Company, Circulation Department, 11830 Westline Industrial Drive, St. Louis, Missouri 63146-3318, USA; phone (800) 325-4177, ext. 351. • Subscriptions mest be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.