- Email: [email protected]
Improved myocardial preservation during global ischemia by continuous retrograde coronary sinus perfusion
J
THoRAc CARDIOVASC SURG
86:659-666, 1983
Improved myocardial preservation during global ischemia by continuous retrograde coronary sinus perfusion To investigate whether retrograde continuous low-pressure perfusion of the coronary sinus could deliver cardioplegic solutions with oxygen and substrate beyond stenoses and result in improved myocardial preservation. we subjected 41 canine hearts to 90 minutes of ischemia with an occlusion on the circumflex coronary artery. There were four groups: Group 1. antegrade (aortic root) crystalloid cardioplegia every 30 minutes during ischemia; Group 11. antegrade plus topical cooling; Group Ill. continuous retrograde perfusion; Group IV, same as Group Ill. with an oxygenated perfluorocarbon. All solutions had a Po; of 400 to 500 mm Hg. Intramyocardial oxygen and carbon dioxide tensions (PO; and Peo;) and mean myocardial temperatures were monitored during ischemia. and left ventricular (LV) function was assessed before ischemia and after reperfusion. After global ischemia. the circumflex occlusion was released and the hearts reperfused. Following 60 minutes of reperfusion, isovolumic developed pressure returned to 36% ± 4% and 41 % ± 5% of preischemic levels. respectively. in Groups I and 11. By contrast. Groups 111 and IV (retrograde perfusion) had a significantly greater percent of recovery (78% ± 5% and 73% ± 5%). Circumflex area intramyocardial Po; fell 20 and 25 mm Hg below preischemic levels in Groups I and 11 during ischemia. whereas in Group Ill. intramyocardial Po; in the circumflex region remained near preischemic levels. and in Group IV, it rose 19 mm Hg. Mean myocardial temperature during ischemia in the circumflex area was significantly higher in Group I than in Groups 11. Ill, and IV. Peak intramyocardial Pco, in the circumflex region was significantly less in the retrogradely perfused hearts. Retrograde coronary sinus perfusion resulted in significant improvement in recovery of LV function. uniform myocardial cooling. normal intramyocardial Po; and less intramyocardial Pco, accumulation. despite the presence of a total circumflex coronary artery occlusion.
Steven F. Bolling, M.D. (by invitation), John T. Flaherty, M.D. (by invitation), Bernadine H. Bulkley, M.D. (by invitation), Vincent L. Gott, M.D., and Timothy J. Gardner, M.D., Baltimore. Md.
CardiOPlegiC arrest and myocardial cooling are accepted techniques to provide myocardial protection during surgically induced ischemia. Although these protective effects of arrest and hypothermia are additive,1 a minimal energy requirement. persists in the cold, arrested heart, and there is continued utilization of energy stores during global ischemia.' Critical coronary artery stenoses may result in nonhomogeneous distribution of the cardioplegic solution and poor myocardial
From the Division of Cardiac Surgery. The Johns Hopkins Medical Institutions, Baltimore, Md. Supported by Ll.S, Public Health Service Grant 2 RO HLl94l4-07 from the National Heart, Lung and Blood Institute and by the Robert H. Maier Family Cardiovascular Surgical Research Fund. Read at the Sixty-third Annual Meeting of The American Association for Thoracic Surgery, Atlanta. Ga., April 25-27, 1983.
cooling distal to such coronary arterial narrowings." 4 There is an increase in myocardial energy needs in these poorly protected areas during ischemia and depression of left ventricular (LV) function following reperfusion, as has been demonstrated experimentally and clinical-
ly.5.7 This study was undertaken to assess the effectiveness of continuous retrograde coronary sinus perfusion to provide myocardial hypothermia, metabolic substrate, and oxygen delivery as well as by-product washout, in the presence of a critical coronary artery stenosis.
Material and methods After the induction of general anesthesia, 41 dogs (20 to 25 kg) had catheters inserted and underwent sternotomy. A Harvey Model H200 oxygenator was primed with heparinized homologous blood and saline. The animals were placed on bypass (mean aortic pressure 80
659
660
The Journal of Thoracic and Cardiovascular Surgery
Bolling et al.
+20 +10 CJl
J:
E E (\J
o
E
a..
I.. 5
15
Fig. 1. The intramyocardial oxygen tensions (Pmo;) during 90 minutes of global ischemia and 60 minutes of reperfusion in all experimental groups in the occluded circumflex distribution (l-antegrade, II-antegrade with topical cooling, Ill-retrograde with crystalloid, and IV-retrograde with pertluorocarbon)..
mm Hg) using the left subclavian artery for arterial inflow and a two-stage venous return cannula. After the sinoatrial node was crushed, ventricular pacing was begun at 120 beats/min. The pH was maintained near 7.4 by appropriate buffering. Hematocrit values during bypass were 31% ± 6%. Perfusate oxygen and carbon dioxide tensions (Po, and Pco.) were maintained at 250 mm Hg and 40 mm Hg, respectively. Three needle thermistors were placed at midmyocardial depths in the right ventricle (RV) and the left anterior descending (LAD) and circumflex areas of the LV to continuously record myocardial temperatures. Two 22 gauge, Teflon-coated stainless steel cannulas were inserted adjacent to the temperature probes in the LAD and circumflex regions and connected to a mass spectrometer (Medspec S-8) to continuously monitor intramyocardial PO z and Pco; The mass spectrometer data were corrected for temperature by the usual method. 8 The heart was isolated in situ by ligation of the descending thoracic aorta, the common brachiocephalic trunk, the left subclavian artery, the superior and inferior venae cavae, and the azygos vein, as described previously." A large-bore cannula with a constrictor was inserted retrogradely in the descending thoracic aorta. Perfusion pressure was maintained at 80 mm Hg by adjusting the constriction. The mitral valve was excised through a left atriotomy and a latex balloon mounted on a button was inserted into the LV. The balloon catheter was connected to a pressure transducer (Statham Db 23) and was filled to produce a baseline LV end-diastolic pressure (L VEDP) of 10 mm Hg. Isovolumic LV pressure was recorded with a Brush Model II 4307-04 amplifier and a Brush Mark 200 direct-writing recorder.
Maximum positive rate of rise of LV pressure (dP/ dt max) was obtained by electronic differentiation with a Brush Model 13-4212-02 differentiator; LV developed pressure (systolic-EDP) and dP/dt max were used as indices of LV function. After control measurements of LV function were obtained and preischemic intramyocardial gas tensions and myocardial temperatures were recorded, the circumflex coronary artery was occluded with a nontraumatic clamp. Global ischemia was induced by discontinuing aortic root perfusion, and with the circumflex artery occlusion present, arrest was achieved by injection of 250 ml of hyperkalemic (25mEq/L) crystalloid cardioplegic solution into the aortic root. Ischemia was maintained for 90 minutes, and the 41 animals were divided into four groups. Group I received 250 ml of cardioplegic solution every 30 minutes during ischemia. Group II received 250 ml of cardioplegic solution every 30 minutes and, in addition, had continuous pericardial irrigation with chilled normal saline to maintain a constant mean myocardial temperature of 16° C in all three areas monitored. Group III underwent continuous retrograde perfusion of the coronary sinus at a mean pressure of 15 mm Hg with a crystalloid solution containing 10 mEq of potassium chloride. Perfusion was achieved with a catheter inserted into the coronary sinus through the venous return cannula. This catheter obviated the need for a separate right atriotomy. The coronary sinus catheter was connected to a Statham Db 23 pressure transducer and coronary sinus pressure was recorded continuously. Coronary sinus perfusion pressure was maintained at 15 mm Hg by manipulation of the flow rate of a separate, constant infusion pump. The crystalloid solutions infused in Groups I, II, and III were
Volume 86 Number 5
Retrograde coronary sinus perfusion 6 6 1
November, 1983
+60 +40 ~ +20 E
E
o'" u E
a. -20
-40
Fig. 2. The intramyocardial carbon dioxide tensions (PmCO;) during 90 minutes of global ischemia and 60 minutes of reperfusion in all experimental groups in the occluded circumflex distribution (l-antegrade, II-antegrade with topical cooling, Ill-retrograde with crystalloid, and IV-retrograde with perfluorocarbon). bubbled with oxygen to a POz from 400 to 500 mm Hg, as was the hyperkalemic (10 mEqjL) perfluorocarbon (Fluosol-DA) solution in Group IV, which also received retrograde coronary sinus perfusion. The time to electrical arrest after onset of ischemia was noted in the circumflex area by a myocardial electrode. Throughout the ischemic period, continuous recordings were made of intramyocardial P0 2 and Pco, and myocardial temperature. During global ischemia, the intraventricular balloon was empty. Postischemic reperfusion was commenced at 37° C and at 80 mm Hg after 90 minutes of ischemia. The circumflex coronary obstruction was removed just prior to reflow. The length of time to electromechanical activity, the occurrence of spontaneous defibrillation, or the need for electrical defibrillations was recorded. For the initial 15 minutes of reperfusion, a beating, nonworking state was maintained by leaving the intraventricular balloon deflated, after which the balloon was reinflated with the same preischemic control volume. I:V function was measured at 15 minute intervals during reperfusion, and myocardial gas tensions and temperatures were recorded. After 60 minutes of reperfusion, the experiment was terminated and the animals were put to death humanely. Hearts were inspected for areas of ecchymosis or injury from the coronary sinus cannula or retrograde perfusion. A portion of the LV was weighed, after which this sample was dried for 48 hours at 80° C and then reweighed. Myocardial water content (%H 20 ) was calculated by the formula: (1 - dry weight/wet weight) X 100 = %HzO In addition, a transverse section of the LV myocardi-
urn (circumflex area) was excised for light microscopy. This tissue was fixed in acetate-buffered 10% formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin and with phosphotungsten and hematoxylin. Histologic evidence of ischemic damage was identified by myocardial cell contraction band necrosis and interstitial hemorrhage. Additional samples, excised for electron microscopy, were fixed in cold 3% glutaraldehyde, washed with 0.1M phosphate buffer, postfixed with osmium tetroxide and, after dehydration, embedded in epoxy resin. Ultrathin sections stained with lead citrate and uranyl acetate were examined for subcellular damage. Injury was judged as mild, moderate, or severe based on the degree of separation and disorganization of the myofibrils, nuclear chromatin clumping, mitochondrial swelling, disruption, mineralization, and contraction band formation. All sections were examined randomly. Statistical analysis was performed by using Student's t test, and a p value of less than 0.05 was considered significant. All results are expressed as mean ± the standard error of the mean. Results Myocardial metabolism. The mean preischemic, intramyocardial P0 2 in the circumflex area for all hearts was 37 ± 2 mm Hg. The time course of intramyocardial P0 2 during ischemia and reperfusion in the circumflex area is shown in Fig. 1. Within 5 minutes of ischemia, intramyocardial P0 2 in all four groups fell significantly. However, in Groups I and II, despite the subsequent cardioplegic injections at 30 and 60 minutes
The Journal of Thoracic and Cardiovascular Surgery
6 6 2 Bolling et af.
90
*-p<.05vsI6:n: *
80 70 ...J
60
0
0::
f-
z
50
0
u ~ 0
o, 0
40 30 20
Fig. 3. The recovery of isovolumic developed pressure (DP) after 90 minutes of global ischemia and 60 minutes of reperfusion in all experimental groups (l-antegrade, llantegrade with topical cooling, llI-retrograde with crystalloid, and IV-retrograde with perfluorocarbon).
of ischemia, intramyocardial P0 2 remained depressed during ischemia and returned to preischemic levels only between 5 and 15 minutes after reperfusion. In Group III there was a 12 mm Hg drop in intramyocardial P0 2 in the circumflex area after 5 minutes of ischemia, but with continuous coronary sinus perfusion using an oxygenated, crystalloid solution, there was recovery of intramyocardial P0 2 to baseline levels by 30 minutes of ischemia. After 60 minutes of ischemia, Group III showed a 5 mm Hg rise in intramyocardial Po 2; during reperfusion, this value returned to baseline. In Group IV there was only a 6 mm Hg drop in intramyocardial P0 2 after 5 minutes of ischemia, following which intramyocardial P0 2 rose to a significant peak of 19 mm Hg above the preischemic baseline. During reperfusion in Group IV, although intramyocardial P0 2 declined toward baseline levels, it continued to be higher than in all other groups. The mean preischemic intramyocardial P0 2 in the LAD area was 39 ± 2 mm Hg for all hearts. In the LAD region, intramyocardial P0 2 at the end of ischemia in Group I decreased to the same level observed for intramyocardial P0 2 in the circumflex region (20 ± 2 mm Hg). In Group II, however, LAD intramyocardial P0 2 decreased only 9 ± 8 mm Hg, as opposed to a
circumflex intramyocardial P0 2 decrease of 16 ± 5 mm Hg in the same hearts. In both Groups III and IV, intramyocardial P0 2 in the LAD area was not significantly different from that in the circumflex area. Mean preischemic intramyocardial Pco, in all groups was 64 ± 4 mm Hg. As seen in Fig. 2, circumflex intramyocardial Pco, during ischemia and reperfusion did not increase significantly in Group III or IV. The maximal increases of circumflex intramyocardial Pco, in Groups I and II, however, measured at the time of' cross-clamp release, were 18 ± 6 mm Hg and 33 ± 11 mm Hg, respectively. Additionally, peak intramyocardial Pco, levels during early reperfusion were significantly higher in Groups I and II than in Groups III and IV. In Group I peak LAD intramyocardial Pco, at the time of cross-clamp release showed only a 4 ± 2 mm Hg rise, as opposed to the 18' ± 6 mm Hg rise in the circumflex area. A similar differential in peak intramyocardial Pco, levels between the LAD and the circumflex areas was also seen in Group II but was not seen in either Group III or IV. Electromechanical arrest occurred in each study with the injection of the initial cardioplegic solution. All hearts remained in flaccid arrest throughout ischemia. The circumflex area distal to the obstruction, however, continued to fibrillate in Group I significantly longer than in the other groups, with fibrillation lasting up to 4 minutes. There were no significant differences noted between any hearts in the time to onset of electrical activity after reperfusion, the rate of spontaneous defibrillation, or the number of electrical shocks required. LV function. Measurements of developed pressure and dP j dt max made after 15, 30, 45, and 60 minutes of reperfusion are expressed as percent return of function compared to the prearrest isovolumic control values. There were no significant differences in the prearrest developed pressure or dP j dt max values between any of the groups. The mean preischemic developed pressure for all hearts was III ± II mm Hg and the mean preischemic dPjdt max was 2,275 ± 180 mm Hgjsec. In Group I, after 90 minutes of ischemia and 60 minutes of reperfusion, hearts recovered 36% ± 4% of developed pressure and 37% ± 5% of dPjdt max. In Group II there was better functional recovery, with 41% ± 5% recovery of developed pressure and 43% ± 5% recovery of dPjdt max. With retrograde coronary sinus perfusion utilizing a crystalloid solution (Group III), there was a very significant (p < 0.05) improvement in LV function compared to Groups I and II, with 78% ± 5% return of developed pressure and 76% ± 6% of positive dP j dt max. Group IV demonstrated a similar
Volume 86 Number 5 November, 1983
enhancement of functional recovery, with 75% ± 6% recovery of developed pressure and 82% ± 6% recovery of dP/dt max (Fig. 3). Measurement of LVEDP after 90 minutes of ischemia and 60 minutes of reperfusion documented significantly lower values in Groups III and IV than in Groups I and II (p < 0.05). In Group I LVEDP was 16.4 ± 1.5 mm Hg, and Group II had an LVEDP of 17.2 ± 2.4 mm Hg. Groups III and IV showed significantly less increase in LVEDP (12.6 ± 1.6 mm Hg and 12.5 ± 1.5 mm Hg, respectively). Myocardial water content. The myocardial water content was 80.0% ± 0.6% for Group I, 80.4% ± 0.3% for Group II, 81.0% ± 1.0% for Group III, and 80.3% ± 0.4% for Group IV. There are no statistical differences. Mean myocardial temperature. Mean myocardial temperature was determined for the RV, the unobstructed LAD region, and the obstructed circumflex region. In the circumflex area, mean myocardial temperature measured during ischemia (Fig. 4) was significantly higher only in Group I (22.9° ± 0.6° C), as opposed to Groups II, III, and IV (15,70 ± 0.5°, 17.4° ± 0.4°, and 16.0° ± 0.4° C, respectively.) In the LAD region, the mean myocardial temperatures in all groups showed no significant differences. Temperatures measured in the RV of all hearts were similar to the LAD temperatures, and only Group I had any significant difference compared to Groups III and IV. Myocardial morpbology. By light microscopy, biopsy specimens showed evidence of focal hemorrhage and contraction band formation, but differences between groups could not be identified. By electron microscopy, Group I demonstrated the most injury, with all but one of the hearts showing moderate to severe injury. Group III and IV hearts showed the least injury, with over half of the specimens showing normal myocardium. Group II hearts were in better condition than Group I hearts but were not as well preserved as those in Groups III and IV. Shown in Fig. 5, A is an electron micrograph of a biopsy specimen in which there is evidence of severe myocardial injury, characterized by disruption and disorganization of the myofibrils and focal mitochondrial swelling. By contrast, the myocardium in Fig. 5, B, which comes from a heart in Group IV, is entirely normal. All specimens were taken from the circumflex area and were magnified X4,OOO. Discussion Retrograde perfusion of the coronary sinus to sustain the myocardium was first attempted in 1898 by Pratt."
Retrograde coronary sinus perfusion 6 6 3
25
u
a, w 20
a::
*
*-p<.05 vs II,m a nz:
::J
~
a:: w
15
w
:E
10
5
Q..
IZ
:E
Fig. 4. The mean myocardial temperature during 90 minutes of global ischemia in the occluded circumflex distribution in all experimental groups (l-antegrade, ll-antegrade with topical cooling, Ill-retrograde with crystalloid, and IV-retrograde with perfluorocarbon).
He was able to maintain ventricular contraction by perfusing the coronary sinus with blood in a completely devascularized feline heart. In the late 1940s, Beck 11 used retrograde coronary sinus perfusion to deliver oxygenated blood from the aorta to the entire coronary venous system. Although this technique did not support the beating, working heart, it did restore attention to the use of coronary sinus retroperfusion. With the development of ischemic arrest and cardioplegia for use in cardiac operations, interest in retrograde coronary sinus perfusion appeared to diminish. Reports of recent studies, however, have reintroduced the concept of using retrograde coronary sinus perfusion, in this case for the delivery of cardioplegic solutions. 12. 13 The presence of critical coronary artery stenoses may result in the nonhomogenous distribution of cardioplegia when the solution is administered in the usual way via the aortic root. Poor myocardial cooling and inadequate protection distal to these arterial narrowings have occurred.v' Cardioplegic administration by means of retrograde coronary sinus perfusion may obviate this maldistribution problem in patients with extensive coronary artery disease. It has been demonstrated that the coronary venous system is not affected by atherosclerosis, regardless of the extent of the disease in the coronary arterial system. This fact has led some investigators to examine selective arterialization of the coronary venous system.":" From these studies, the endocardial/epicardial flow ratios of retrograde venous perfusion were 1.4:1, indicating effective delivery of solutions to the subendocardium via venous channels. In 1972, retrograde insufflation of the coronary sinus
664
Bolling et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 5. Electron micrographs taken from the occluded circumflex area. A, Biopsy specimen from Group I, showing severe myocardial injury. B. Biopsy specimen from group IV, showing entirely normal ultrastructure. (Original magnifications X4,OOO; reduced 28%).
with gaseous oxygen was shown to provide enhanced myocardial protection as compared to that in ischemic controls. 17 Further studies have shown significant improvement in return of ventricular function after ischemia in hearts protected with coronary sinus injection of a cardioplegic solution. 18,19 More recently, studies utilizing experimental coronary artery occlusion to create a model of regional ischemia have shown a decrease in the ischemic zone with diastolic retroperfusion of blood through the coronary sinus during the ischemic period. 20, 2\ The normal coronary venous pressure is 0 to 6 mm Hg. Gott, Gonzalez. and Zuhdi " demonstrated experimentally that coronary sinus pressures of 40 mm Hg or greater result in subepicardial and myocardial ecchymoses, whereas coronary sinus pressures in excess of 80 mm Hg may lead to widespread hemorrhage and significant myocardial damage. It has been suggested that 30 mm Hg is the highest pressure appropriate for the coronary sinus and coronary venous system. As has been demonstrated in this present study, coronary sinus perfusion at lower pressure is apparently safe, since there was no evidence of any macroscopic or microscopic damage to the heart. In addition , even at low pressure, coronary sinus flow rates were adequate to allow
effective delivery of cardioplegic solutions in these canine hearts. In the present study , intramyocardial P02 remainedat or above baseline levels in both Groups III and IV, the retrogradely perfused hearts. This finding is not surprising for Group IV hearts, since perfluorocarbon compounds have been demonstrated to provide effective oxygen delivery during hypothermic ischemia." In Group III, intramyocardial P0 2 is maintained at or above baseline levels despite the fact that the amount of oxygen delivered in a crystalloid solution is small. Since LV functional recovery was similar in both Groups III and IV, it is clear that the continuous infusion of an oxygenated, crystalloid solution is able to deliver sufficient oxygen to meet the much reduced metabolic demands of the cold, arrested heart.": 25 Retrograde coronary sinus perfusion was shown in this study to be effective in achieving myocardial cooling in the presence of a critical coronary artery stenesis Neither antegrade delivery alone nor antegrade perfusion in conjunction with topical hypothermia was as effective in maintaining myocardial cooling or in preserving LV function. Because RV venous drainage docs not occur primarily via the coronary sinus, some have questioned whether there might be inadequate delivery
Volume 86 Number 5 November, 1983
of solutions to the RV with coronary sinus retroperfusion and, therefore, inadequate protection of the right side of the heart with this technique.e- 27 Although the RV venous system does drain directly into the right atrium and RV, there are apparently numerous venovenous connections, as has been shown by the infusion of a silicone rubber solution (Microfil) into the coronary sinus. This material did appear in the coronary veins of the RV when injected into the coronary sinus." Furthermore, in the present study, the injection of methylene blue into the coronary sinus resulted in staining of the RV. In addition, the RV myocardium achieved the same level of cooling in the retroperfused hearts, with no difference in temperature noted between the RV, LV, and circumflex regions. Although myocardial cooling is an important mechanism for myocardial protection during global ischemia, substrate delivery, metabolic by-product washout, and the prevention of intramyocardial acidosis likely play important additional roles. In the present study, if one compares the degree of myocardial protection afforded in Group II with that in Group III, even though both groups received the same basic solutions and achieved the same degree of myocardial cooling, LV functional recovery was significantly better in Group III. There was, in addition, significantly less intramyocardial Pco, accumulation in Group III as compared to Group II, suggesting better metabolic washout with continuous retrograde perfusion. With the use of this method of myocardial protection, intramyocardial P0 2 was maintained within normal limits, improved myocardial metabolism and/or metabolic end-product washout was demonstrated, and enhanced myocardial cooling was seen even in the presence of a completely obstructing coronary lesion. Based on the results of this study, the technique of low pressure, continuous, retrograde coronary sinus perfusion appears to be a promising new approach to improving myocardial protection during prolonged global ischemia. REFERENCES Schaff HV, Dombroff R, Flaherty JT, Bulkley BH, Hutchins GM, Goldman RA, Gott VL: Effect of potassium cardioplegia on myocardial ischemia and post arrest ventricular function. Circulation 58:240-249, 1978 2 Flaherty JT, Jacobus WE, Weisfeldt ML, Hollis DL, Gardner TJ, Gott VL: Use of "phosphorous nuclear magnetic resonance to assess myocardial protection during global ischemia. Am J Cardiol 43:362-273, 1979 3 Heinemann FW, MacGregor DC, Wilson JG, Ninomiya 0: Regional and transmural myocardial temperature
Retrograde coronary sinus perfusion 6 6 5
distribution in cold chemical cardioplegia. J THORAC CARDIOVASC SURG 81:851-859, 1981 4 Landymore RW, Tice D, Trehan N, Spencer F: Importance of topical hypothermia to ensure uniform myocardial cooling during coronary artery bypass. J THORAC CARDIOVASC SURG 82:832-836, 1981 5 Grondin CM, Helias J, Vouhe PR, Robert P: Influence of critical coronary artery stenosis on myocardial protection through cold potassium cardioplegia. J THORAC CARDIoVASC SURG 82:608-615, 1981 6 Hilton CJ, Teubl W, Acker M, Levinson HJ, Millard RW, Riddle R, McEnany MT: Inadequate cardioplegic protection with obstructed coronary arteries. Ann Thorac Surg 28:323-334, 1979 7 Vander Salm TJ, Okike ON, Cutler BS, Paraskos JA, Ferulo J, Daggett W: Improved myocardial preservation by improved distribution of cardioplegic solutions. J THORAC CARDIOVASC SURG 82:767-771, 1982 8 Holness DE, Brock AG: Temperature-dependent characteristics of Teflon membranes used in mass spectrometry. Med Instrum 9:23-33, 1975 9 Lucas SK, Gardner TJ, Flaherty JT, Bulkley BH, Elmer EB, Gott VL: Beneficial effects of mannitol administration during reperfusion after ischemic arrest. Circulation 62:Suppl 1:34-41, 1980 10 Pratt FH: The nutrition of the heart through the vessels of Thebesius and the coronary veins. Am J Physiol 1:86-103, 1898 II Beck CS: Revascularization of the heart. Ann Surg 128:854-870, 1948 12 Menasche P, Kural S, Fauchet M, Lavergne A, Commin P, Bercot M, Touchot B, Georgiopoulos G, Piwnica A: Retrograde coronary sinus perfusion. A safe alternative for ensuring cardioplegic delivery in aortic valve surgery. Ann Thorac Surg 34:647-657, 1982 13 Solorzano J, Taitelbaum G, Chiu R: Retrograde coronary sinus perfusion for myocardial protection during cardiopulmonary bypass. Ann Thorac Surg 25:201-208, 1978 14 Andreadis P, Natsikas N, Arealis E: The aorto coronary venous anastomosis in experimental acute myocardial ischemia. Vase Surg 8:45-53, 1974 15 Bhayana J, Olsen D, Bryn EJ: Reversal of myocardial ischemia by arterialization of the coronary vein. J THORAC CARDIOVASC SURG 67:125-132, 1974 16 Hochberg M, Roberts W, Morrow A, Austen W: Selective arterialization of the coronary venous system. J THORAC CARDIOVASC SURG 77: 1-11, 1979 17 Brown A, Niles N, Braimbridge EM, Austen W: Retrograde insuflation of gaseous oxygen into the coronary sinus as a means of myocardial maintenance. Arch Surg 105:622-627, 1972 18 Poirier R, Guyton R, MacIntosh C: Drip retrograde coronary sinus perfusion for myocardial protection during aortic cross-clamping. J THORAC CARDIOVASC SURG 70:966-973, 1975 19 Lolley D, Hewitt R, Drapanas T: Retroperfusion of the
The Journal of Thoracic and Cardiovascular Surgery
6 6 6 Bolling et al.
20
21
22
23
24
25
26 27
heart with a solution of glucose, insulin, and potassium during anoxic arrest. J THORAC CARDIOVASC SURG 67:364-369, 1974 Berdeaux A, Farcto J, Bourdarias J, Barry M, Bardet J, Giudicelli J: Effects of diastolic synchronized retroperfusion on regional coronary blood flow in experimental myocardial ischemia. Am J Cardiol 47:1033-1040, 1981 Gundry S: Modification of myocardial ischemia in normal and hypertrophied hearts using diastolic retroperfusion of the coronary veins. J THORAC CARDIOVASC SURG 83:659669, 1982 Gott VL, Gonzalez J, Zuhdi M: Retrograde perfusion of coronary sinus for direct vision aortic surgery. Surg Gynecol Obstet 104:319-328, 1957 Kanter KR, Jaffin JH, Ehrlichman RJ, Flaherty JT, Gott VL, Gardner TJ: Superiority of perfluorocarbon cardioplegia over blood or crystalloid cardioplegia. Circulation 64:Suppl 2:75-80, 1981 Donovan T, Mahkeri VB, Owens G: Myocardial perfusion and metabolism at normothermic and hypothermic levels. Arch Surg 110:208-219, 1975 Bodenhamer RM, DeBoer LW, Geffin GA, O'Keefe DO, Fallon JT, Aretz TH, Haas GS, Daggett WM: Enhanced myocardial protection during ischemic arrest. Oxygenation of a crystalloid cardioplegic solution. J THORAC CARDIOVASC SURG 85:769-780, 1983. Hochberg M, Austen W: Selective retrograde coronary venous perfusion. Ann Thorac Surg 29:578-588, 1980 Bates R, Toscano M, Balderman S, Anagnostopoulos C: The cardiac veins in retrograde coronary venous perfusion. Ann Thorac Surg 23:83-90, 1977
Discussion DR. WILLARD M. DAGGETT Boston. Mass.
The authors' study comprised four groups, two of which had retrograde and two antegrade perfusion. I think the improved preservation in Group tV can be explained purely on the basis of oxygen delivery by the oxygenated perfluorocarbon rather than the route of delivery.
If I understood correctly, Group III had retrograde perfusion with an oxygenated crystalloid solution. I was not sure whether Groups I and II (antegrade perfusion) were given an oxygenated or an unoxygenated crystalloid solution. If the first two groups were not oxygenated, then it would be possible to explain all of the fmdings, not on the route of perfusion, but by delivery of oxygen to the cross-clamped heart. Also, if the volumes of cardioplegic solution delivered to the different groups were not equal, then oxygen delivery would differ. DR. FRANK C. SPENCER New York. N. Y.
Dr. Bolling, may I ask you a question? You will have noticed that in the big puzzle of myocardial preservation is coronary washout with the collaterals. If blood can be perfused retrogradely in the coronary sinus, do you suppose that by putting suction on the right atrium and emptying it out, noncoronary flow can be aspirated the other way, so that it does not wash the potassium out? DR. BOLLING (Closing) Thank you for the comments. Dr. Daggett, all the solutions, including the crystalloid and the perfluorocarbon compounds, were oxygenated to an approximate P0 2 between 400 and 500. In Group II (perfused antegradely) the crystalloid solution was delivered oxygenated, so that Groups II and III both received the same oxygenated crystalloid cardioplegic solution. They both achieved the same mean myocardial temperature in the circumflex area, one via the means of topical cooling and one by retrograde perfusion. Yet Group III, the retroperfused group, demonstrated significantly enhanced return of function. Dr. Spencer, I am not sure of the answer to your question. I think one of the things that would inhibit the clinical use of this method is the ease of use itself. Some have pointed out that a separate atriotomy is necessary. In one of the initial phases of this study, we are developing a balloon catheter that is an integral part of the two-stage venous cannula, which allows very easy entrance into the coronary sinus without a separate atriotomy.
THoRAc CARDIOVASC SURG
86:659-666, 1983
Improved myocardial preservation during global ischemia by continuous retrograde coronary sinus perfusion To investigate whether retrograde continuous low-pressure perfusion of the coronary sinus could deliver cardioplegic solutions with oxygen and substrate beyond stenoses and result in improved myocardial preservation. we subjected 41 canine hearts to 90 minutes of ischemia with an occlusion on the circumflex coronary artery. There were four groups: Group 1. antegrade (aortic root) crystalloid cardioplegia every 30 minutes during ischemia; Group 11. antegrade plus topical cooling; Group Ill. continuous retrograde perfusion; Group IV, same as Group Ill. with an oxygenated perfluorocarbon. All solutions had a Po; of 400 to 500 mm Hg. Intramyocardial oxygen and carbon dioxide tensions (PO; and Peo;) and mean myocardial temperatures were monitored during ischemia. and left ventricular (LV) function was assessed before ischemia and after reperfusion. After global ischemia. the circumflex occlusion was released and the hearts reperfused. Following 60 minutes of reperfusion, isovolumic developed pressure returned to 36% ± 4% and 41 % ± 5% of preischemic levels. respectively. in Groups I and 11. By contrast. Groups 111 and IV (retrograde perfusion) had a significantly greater percent of recovery (78% ± 5% and 73% ± 5%). Circumflex area intramyocardial Po; fell 20 and 25 mm Hg below preischemic levels in Groups I and 11 during ischemia. whereas in Group Ill. intramyocardial Po; in the circumflex region remained near preischemic levels. and in Group IV, it rose 19 mm Hg. Mean myocardial temperature during ischemia in the circumflex area was significantly higher in Group I than in Groups 11. Ill, and IV. Peak intramyocardial Pco, in the circumflex region was significantly less in the retrogradely perfused hearts. Retrograde coronary sinus perfusion resulted in significant improvement in recovery of LV function. uniform myocardial cooling. normal intramyocardial Po; and less intramyocardial Pco, accumulation. despite the presence of a total circumflex coronary artery occlusion.
Steven F. Bolling, M.D. (by invitation), John T. Flaherty, M.D. (by invitation), Bernadine H. Bulkley, M.D. (by invitation), Vincent L. Gott, M.D., and Timothy J. Gardner, M.D., Baltimore. Md.
CardiOPlegiC arrest and myocardial cooling are accepted techniques to provide myocardial protection during surgically induced ischemia. Although these protective effects of arrest and hypothermia are additive,1 a minimal energy requirement. persists in the cold, arrested heart, and there is continued utilization of energy stores during global ischemia.' Critical coronary artery stenoses may result in nonhomogeneous distribution of the cardioplegic solution and poor myocardial
From the Division of Cardiac Surgery. The Johns Hopkins Medical Institutions, Baltimore, Md. Supported by Ll.S, Public Health Service Grant 2 RO HLl94l4-07 from the National Heart, Lung and Blood Institute and by the Robert H. Maier Family Cardiovascular Surgical Research Fund. Read at the Sixty-third Annual Meeting of The American Association for Thoracic Surgery, Atlanta. Ga., April 25-27, 1983.
cooling distal to such coronary arterial narrowings." 4 There is an increase in myocardial energy needs in these poorly protected areas during ischemia and depression of left ventricular (LV) function following reperfusion, as has been demonstrated experimentally and clinical-
ly.5.7 This study was undertaken to assess the effectiveness of continuous retrograde coronary sinus perfusion to provide myocardial hypothermia, metabolic substrate, and oxygen delivery as well as by-product washout, in the presence of a critical coronary artery stenosis.
Material and methods After the induction of general anesthesia, 41 dogs (20 to 25 kg) had catheters inserted and underwent sternotomy. A Harvey Model H200 oxygenator was primed with heparinized homologous blood and saline. The animals were placed on bypass (mean aortic pressure 80
659
660
The Journal of Thoracic and Cardiovascular Surgery
Bolling et al.
+20 +10 CJl
J:
E E (\J
o
E
a..
I.. 5
15
Fig. 1. The intramyocardial oxygen tensions (Pmo;) during 90 minutes of global ischemia and 60 minutes of reperfusion in all experimental groups in the occluded circumflex distribution (l-antegrade, II-antegrade with topical cooling, Ill-retrograde with crystalloid, and IV-retrograde with pertluorocarbon)..
mm Hg) using the left subclavian artery for arterial inflow and a two-stage venous return cannula. After the sinoatrial node was crushed, ventricular pacing was begun at 120 beats/min. The pH was maintained near 7.4 by appropriate buffering. Hematocrit values during bypass were 31% ± 6%. Perfusate oxygen and carbon dioxide tensions (Po, and Pco.) were maintained at 250 mm Hg and 40 mm Hg, respectively. Three needle thermistors were placed at midmyocardial depths in the right ventricle (RV) and the left anterior descending (LAD) and circumflex areas of the LV to continuously record myocardial temperatures. Two 22 gauge, Teflon-coated stainless steel cannulas were inserted adjacent to the temperature probes in the LAD and circumflex regions and connected to a mass spectrometer (Medspec S-8) to continuously monitor intramyocardial PO z and Pco; The mass spectrometer data were corrected for temperature by the usual method. 8 The heart was isolated in situ by ligation of the descending thoracic aorta, the common brachiocephalic trunk, the left subclavian artery, the superior and inferior venae cavae, and the azygos vein, as described previously." A large-bore cannula with a constrictor was inserted retrogradely in the descending thoracic aorta. Perfusion pressure was maintained at 80 mm Hg by adjusting the constriction. The mitral valve was excised through a left atriotomy and a latex balloon mounted on a button was inserted into the LV. The balloon catheter was connected to a pressure transducer (Statham Db 23) and was filled to produce a baseline LV end-diastolic pressure (L VEDP) of 10 mm Hg. Isovolumic LV pressure was recorded with a Brush Model II 4307-04 amplifier and a Brush Mark 200 direct-writing recorder.
Maximum positive rate of rise of LV pressure (dP/ dt max) was obtained by electronic differentiation with a Brush Model 13-4212-02 differentiator; LV developed pressure (systolic-EDP) and dP/dt max were used as indices of LV function. After control measurements of LV function were obtained and preischemic intramyocardial gas tensions and myocardial temperatures were recorded, the circumflex coronary artery was occluded with a nontraumatic clamp. Global ischemia was induced by discontinuing aortic root perfusion, and with the circumflex artery occlusion present, arrest was achieved by injection of 250 ml of hyperkalemic (25mEq/L) crystalloid cardioplegic solution into the aortic root. Ischemia was maintained for 90 minutes, and the 41 animals were divided into four groups. Group I received 250 ml of cardioplegic solution every 30 minutes during ischemia. Group II received 250 ml of cardioplegic solution every 30 minutes and, in addition, had continuous pericardial irrigation with chilled normal saline to maintain a constant mean myocardial temperature of 16° C in all three areas monitored. Group III underwent continuous retrograde perfusion of the coronary sinus at a mean pressure of 15 mm Hg with a crystalloid solution containing 10 mEq of potassium chloride. Perfusion was achieved with a catheter inserted into the coronary sinus through the venous return cannula. This catheter obviated the need for a separate right atriotomy. The coronary sinus catheter was connected to a Statham Db 23 pressure transducer and coronary sinus pressure was recorded continuously. Coronary sinus perfusion pressure was maintained at 15 mm Hg by manipulation of the flow rate of a separate, constant infusion pump. The crystalloid solutions infused in Groups I, II, and III were
Volume 86 Number 5
Retrograde coronary sinus perfusion 6 6 1
November, 1983
+60 +40 ~ +20 E
E
o'" u E
a. -20
-40
Fig. 2. The intramyocardial carbon dioxide tensions (PmCO;) during 90 minutes of global ischemia and 60 minutes of reperfusion in all experimental groups in the occluded circumflex distribution (l-antegrade, II-antegrade with topical cooling, Ill-retrograde with crystalloid, and IV-retrograde with perfluorocarbon). bubbled with oxygen to a POz from 400 to 500 mm Hg, as was the hyperkalemic (10 mEqjL) perfluorocarbon (Fluosol-DA) solution in Group IV, which also received retrograde coronary sinus perfusion. The time to electrical arrest after onset of ischemia was noted in the circumflex area by a myocardial electrode. Throughout the ischemic period, continuous recordings were made of intramyocardial P0 2 and Pco, and myocardial temperature. During global ischemia, the intraventricular balloon was empty. Postischemic reperfusion was commenced at 37° C and at 80 mm Hg after 90 minutes of ischemia. The circumflex coronary obstruction was removed just prior to reflow. The length of time to electromechanical activity, the occurrence of spontaneous defibrillation, or the need for electrical defibrillations was recorded. For the initial 15 minutes of reperfusion, a beating, nonworking state was maintained by leaving the intraventricular balloon deflated, after which the balloon was reinflated with the same preischemic control volume. I:V function was measured at 15 minute intervals during reperfusion, and myocardial gas tensions and temperatures were recorded. After 60 minutes of reperfusion, the experiment was terminated and the animals were put to death humanely. Hearts were inspected for areas of ecchymosis or injury from the coronary sinus cannula or retrograde perfusion. A portion of the LV was weighed, after which this sample was dried for 48 hours at 80° C and then reweighed. Myocardial water content (%H 20 ) was calculated by the formula: (1 - dry weight/wet weight) X 100 = %HzO In addition, a transverse section of the LV myocardi-
urn (circumflex area) was excised for light microscopy. This tissue was fixed in acetate-buffered 10% formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin and with phosphotungsten and hematoxylin. Histologic evidence of ischemic damage was identified by myocardial cell contraction band necrosis and interstitial hemorrhage. Additional samples, excised for electron microscopy, were fixed in cold 3% glutaraldehyde, washed with 0.1M phosphate buffer, postfixed with osmium tetroxide and, after dehydration, embedded in epoxy resin. Ultrathin sections stained with lead citrate and uranyl acetate were examined for subcellular damage. Injury was judged as mild, moderate, or severe based on the degree of separation and disorganization of the myofibrils, nuclear chromatin clumping, mitochondrial swelling, disruption, mineralization, and contraction band formation. All sections were examined randomly. Statistical analysis was performed by using Student's t test, and a p value of less than 0.05 was considered significant. All results are expressed as mean ± the standard error of the mean. Results Myocardial metabolism. The mean preischemic, intramyocardial P0 2 in the circumflex area for all hearts was 37 ± 2 mm Hg. The time course of intramyocardial P0 2 during ischemia and reperfusion in the circumflex area is shown in Fig. 1. Within 5 minutes of ischemia, intramyocardial P0 2 in all four groups fell significantly. However, in Groups I and II, despite the subsequent cardioplegic injections at 30 and 60 minutes
The Journal of Thoracic and Cardiovascular Surgery
6 6 2 Bolling et af.
90
*-p<.05vsI6:n: *
80 70 ...J
60
0
0::
f-
z
50
0
u ~ 0
o, 0
40 30 20
Fig. 3. The recovery of isovolumic developed pressure (DP) after 90 minutes of global ischemia and 60 minutes of reperfusion in all experimental groups (l-antegrade, llantegrade with topical cooling, llI-retrograde with crystalloid, and IV-retrograde with perfluorocarbon).
of ischemia, intramyocardial P0 2 remained depressed during ischemia and returned to preischemic levels only between 5 and 15 minutes after reperfusion. In Group III there was a 12 mm Hg drop in intramyocardial P0 2 in the circumflex area after 5 minutes of ischemia, but with continuous coronary sinus perfusion using an oxygenated, crystalloid solution, there was recovery of intramyocardial P0 2 to baseline levels by 30 minutes of ischemia. After 60 minutes of ischemia, Group III showed a 5 mm Hg rise in intramyocardial Po 2; during reperfusion, this value returned to baseline. In Group IV there was only a 6 mm Hg drop in intramyocardial P0 2 after 5 minutes of ischemia, following which intramyocardial P0 2 rose to a significant peak of 19 mm Hg above the preischemic baseline. During reperfusion in Group IV, although intramyocardial P0 2 declined toward baseline levels, it continued to be higher than in all other groups. The mean preischemic intramyocardial P0 2 in the LAD area was 39 ± 2 mm Hg for all hearts. In the LAD region, intramyocardial P0 2 at the end of ischemia in Group I decreased to the same level observed for intramyocardial P0 2 in the circumflex region (20 ± 2 mm Hg). In Group II, however, LAD intramyocardial P0 2 decreased only 9 ± 8 mm Hg, as opposed to a
circumflex intramyocardial P0 2 decrease of 16 ± 5 mm Hg in the same hearts. In both Groups III and IV, intramyocardial P0 2 in the LAD area was not significantly different from that in the circumflex area. Mean preischemic intramyocardial Pco, in all groups was 64 ± 4 mm Hg. As seen in Fig. 2, circumflex intramyocardial Pco, during ischemia and reperfusion did not increase significantly in Group III or IV. The maximal increases of circumflex intramyocardial Pco, in Groups I and II, however, measured at the time of' cross-clamp release, were 18 ± 6 mm Hg and 33 ± 11 mm Hg, respectively. Additionally, peak intramyocardial Pco, levels during early reperfusion were significantly higher in Groups I and II than in Groups III and IV. In Group I peak LAD intramyocardial Pco, at the time of cross-clamp release showed only a 4 ± 2 mm Hg rise, as opposed to the 18' ± 6 mm Hg rise in the circumflex area. A similar differential in peak intramyocardial Pco, levels between the LAD and the circumflex areas was also seen in Group II but was not seen in either Group III or IV. Electromechanical arrest occurred in each study with the injection of the initial cardioplegic solution. All hearts remained in flaccid arrest throughout ischemia. The circumflex area distal to the obstruction, however, continued to fibrillate in Group I significantly longer than in the other groups, with fibrillation lasting up to 4 minutes. There were no significant differences noted between any hearts in the time to onset of electrical activity after reperfusion, the rate of spontaneous defibrillation, or the number of electrical shocks required. LV function. Measurements of developed pressure and dP j dt max made after 15, 30, 45, and 60 minutes of reperfusion are expressed as percent return of function compared to the prearrest isovolumic control values. There were no significant differences in the prearrest developed pressure or dP j dt max values between any of the groups. The mean preischemic developed pressure for all hearts was III ± II mm Hg and the mean preischemic dPjdt max was 2,275 ± 180 mm Hgjsec. In Group I, after 90 minutes of ischemia and 60 minutes of reperfusion, hearts recovered 36% ± 4% of developed pressure and 37% ± 5% of dPjdt max. In Group II there was better functional recovery, with 41% ± 5% recovery of developed pressure and 43% ± 5% recovery of dPjdt max. With retrograde coronary sinus perfusion utilizing a crystalloid solution (Group III), there was a very significant (p < 0.05) improvement in LV function compared to Groups I and II, with 78% ± 5% return of developed pressure and 76% ± 6% of positive dP j dt max. Group IV demonstrated a similar
Volume 86 Number 5 November, 1983
enhancement of functional recovery, with 75% ± 6% recovery of developed pressure and 82% ± 6% recovery of dP/dt max (Fig. 3). Measurement of LVEDP after 90 minutes of ischemia and 60 minutes of reperfusion documented significantly lower values in Groups III and IV than in Groups I and II (p < 0.05). In Group I LVEDP was 16.4 ± 1.5 mm Hg, and Group II had an LVEDP of 17.2 ± 2.4 mm Hg. Groups III and IV showed significantly less increase in LVEDP (12.6 ± 1.6 mm Hg and 12.5 ± 1.5 mm Hg, respectively). Myocardial water content. The myocardial water content was 80.0% ± 0.6% for Group I, 80.4% ± 0.3% for Group II, 81.0% ± 1.0% for Group III, and 80.3% ± 0.4% for Group IV. There are no statistical differences. Mean myocardial temperature. Mean myocardial temperature was determined for the RV, the unobstructed LAD region, and the obstructed circumflex region. In the circumflex area, mean myocardial temperature measured during ischemia (Fig. 4) was significantly higher only in Group I (22.9° ± 0.6° C), as opposed to Groups II, III, and IV (15,70 ± 0.5°, 17.4° ± 0.4°, and 16.0° ± 0.4° C, respectively.) In the LAD region, the mean myocardial temperatures in all groups showed no significant differences. Temperatures measured in the RV of all hearts were similar to the LAD temperatures, and only Group I had any significant difference compared to Groups III and IV. Myocardial morpbology. By light microscopy, biopsy specimens showed evidence of focal hemorrhage and contraction band formation, but differences between groups could not be identified. By electron microscopy, Group I demonstrated the most injury, with all but one of the hearts showing moderate to severe injury. Group III and IV hearts showed the least injury, with over half of the specimens showing normal myocardium. Group II hearts were in better condition than Group I hearts but were not as well preserved as those in Groups III and IV. Shown in Fig. 5, A is an electron micrograph of a biopsy specimen in which there is evidence of severe myocardial injury, characterized by disruption and disorganization of the myofibrils and focal mitochondrial swelling. By contrast, the myocardium in Fig. 5, B, which comes from a heart in Group IV, is entirely normal. All specimens were taken from the circumflex area and were magnified X4,OOO. Discussion Retrograde perfusion of the coronary sinus to sustain the myocardium was first attempted in 1898 by Pratt."
Retrograde coronary sinus perfusion 6 6 3
25
u
a, w 20
a::
*
*-p<.05 vs II,m a nz:
::J
~
a:: w
15
w
:E
10
5
Q..
IZ
:E
Fig. 4. The mean myocardial temperature during 90 minutes of global ischemia in the occluded circumflex distribution in all experimental groups (l-antegrade, ll-antegrade with topical cooling, Ill-retrograde with crystalloid, and IV-retrograde with perfluorocarbon).
He was able to maintain ventricular contraction by perfusing the coronary sinus with blood in a completely devascularized feline heart. In the late 1940s, Beck 11 used retrograde coronary sinus perfusion to deliver oxygenated blood from the aorta to the entire coronary venous system. Although this technique did not support the beating, working heart, it did restore attention to the use of coronary sinus retroperfusion. With the development of ischemic arrest and cardioplegia for use in cardiac operations, interest in retrograde coronary sinus perfusion appeared to diminish. Reports of recent studies, however, have reintroduced the concept of using retrograde coronary sinus perfusion, in this case for the delivery of cardioplegic solutions. 12. 13 The presence of critical coronary artery stenoses may result in the nonhomogenous distribution of cardioplegia when the solution is administered in the usual way via the aortic root. Poor myocardial cooling and inadequate protection distal to these arterial narrowings have occurred.v' Cardioplegic administration by means of retrograde coronary sinus perfusion may obviate this maldistribution problem in patients with extensive coronary artery disease. It has been demonstrated that the coronary venous system is not affected by atherosclerosis, regardless of the extent of the disease in the coronary arterial system. This fact has led some investigators to examine selective arterialization of the coronary venous system.":" From these studies, the endocardial/epicardial flow ratios of retrograde venous perfusion were 1.4:1, indicating effective delivery of solutions to the subendocardium via venous channels. In 1972, retrograde insufflation of the coronary sinus
664
Bolling et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 5. Electron micrographs taken from the occluded circumflex area. A, Biopsy specimen from Group I, showing severe myocardial injury. B. Biopsy specimen from group IV, showing entirely normal ultrastructure. (Original magnifications X4,OOO; reduced 28%).
with gaseous oxygen was shown to provide enhanced myocardial protection as compared to that in ischemic controls. 17 Further studies have shown significant improvement in return of ventricular function after ischemia in hearts protected with coronary sinus injection of a cardioplegic solution. 18,19 More recently, studies utilizing experimental coronary artery occlusion to create a model of regional ischemia have shown a decrease in the ischemic zone with diastolic retroperfusion of blood through the coronary sinus during the ischemic period. 20, 2\ The normal coronary venous pressure is 0 to 6 mm Hg. Gott, Gonzalez. and Zuhdi " demonstrated experimentally that coronary sinus pressures of 40 mm Hg or greater result in subepicardial and myocardial ecchymoses, whereas coronary sinus pressures in excess of 80 mm Hg may lead to widespread hemorrhage and significant myocardial damage. It has been suggested that 30 mm Hg is the highest pressure appropriate for the coronary sinus and coronary venous system. As has been demonstrated in this present study, coronary sinus perfusion at lower pressure is apparently safe, since there was no evidence of any macroscopic or microscopic damage to the heart. In addition , even at low pressure, coronary sinus flow rates were adequate to allow
effective delivery of cardioplegic solutions in these canine hearts. In the present study , intramyocardial P02 remainedat or above baseline levels in both Groups III and IV, the retrogradely perfused hearts. This finding is not surprising for Group IV hearts, since perfluorocarbon compounds have been demonstrated to provide effective oxygen delivery during hypothermic ischemia." In Group III, intramyocardial P0 2 is maintained at or above baseline levels despite the fact that the amount of oxygen delivered in a crystalloid solution is small. Since LV functional recovery was similar in both Groups III and IV, it is clear that the continuous infusion of an oxygenated, crystalloid solution is able to deliver sufficient oxygen to meet the much reduced metabolic demands of the cold, arrested heart.": 25 Retrograde coronary sinus perfusion was shown in this study to be effective in achieving myocardial cooling in the presence of a critical coronary artery stenesis Neither antegrade delivery alone nor antegrade perfusion in conjunction with topical hypothermia was as effective in maintaining myocardial cooling or in preserving LV function. Because RV venous drainage docs not occur primarily via the coronary sinus, some have questioned whether there might be inadequate delivery
Volume 86 Number 5 November, 1983
of solutions to the RV with coronary sinus retroperfusion and, therefore, inadequate protection of the right side of the heart with this technique.e- 27 Although the RV venous system does drain directly into the right atrium and RV, there are apparently numerous venovenous connections, as has been shown by the infusion of a silicone rubber solution (Microfil) into the coronary sinus. This material did appear in the coronary veins of the RV when injected into the coronary sinus." Furthermore, in the present study, the injection of methylene blue into the coronary sinus resulted in staining of the RV. In addition, the RV myocardium achieved the same level of cooling in the retroperfused hearts, with no difference in temperature noted between the RV, LV, and circumflex regions. Although myocardial cooling is an important mechanism for myocardial protection during global ischemia, substrate delivery, metabolic by-product washout, and the prevention of intramyocardial acidosis likely play important additional roles. In the present study, if one compares the degree of myocardial protection afforded in Group II with that in Group III, even though both groups received the same basic solutions and achieved the same degree of myocardial cooling, LV functional recovery was significantly better in Group III. There was, in addition, significantly less intramyocardial Pco, accumulation in Group III as compared to Group II, suggesting better metabolic washout with continuous retrograde perfusion. With the use of this method of myocardial protection, intramyocardial P0 2 was maintained within normal limits, improved myocardial metabolism and/or metabolic end-product washout was demonstrated, and enhanced myocardial cooling was seen even in the presence of a completely obstructing coronary lesion. Based on the results of this study, the technique of low pressure, continuous, retrograde coronary sinus perfusion appears to be a promising new approach to improving myocardial protection during prolonged global ischemia. REFERENCES Schaff HV, Dombroff R, Flaherty JT, Bulkley BH, Hutchins GM, Goldman RA, Gott VL: Effect of potassium cardioplegia on myocardial ischemia and post arrest ventricular function. Circulation 58:240-249, 1978 2 Flaherty JT, Jacobus WE, Weisfeldt ML, Hollis DL, Gardner TJ, Gott VL: Use of "phosphorous nuclear magnetic resonance to assess myocardial protection during global ischemia. Am J Cardiol 43:362-273, 1979 3 Heinemann FW, MacGregor DC, Wilson JG, Ninomiya 0: Regional and transmural myocardial temperature
Retrograde coronary sinus perfusion 6 6 5
distribution in cold chemical cardioplegia. J THORAC CARDIOVASC SURG 81:851-859, 1981 4 Landymore RW, Tice D, Trehan N, Spencer F: Importance of topical hypothermia to ensure uniform myocardial cooling during coronary artery bypass. J THORAC CARDIOVASC SURG 82:832-836, 1981 5 Grondin CM, Helias J, Vouhe PR, Robert P: Influence of critical coronary artery stenosis on myocardial protection through cold potassium cardioplegia. J THORAC CARDIoVASC SURG 82:608-615, 1981 6 Hilton CJ, Teubl W, Acker M, Levinson HJ, Millard RW, Riddle R, McEnany MT: Inadequate cardioplegic protection with obstructed coronary arteries. Ann Thorac Surg 28:323-334, 1979 7 Vander Salm TJ, Okike ON, Cutler BS, Paraskos JA, Ferulo J, Daggett W: Improved myocardial preservation by improved distribution of cardioplegic solutions. J THORAC CARDIOVASC SURG 82:767-771, 1982 8 Holness DE, Brock AG: Temperature-dependent characteristics of Teflon membranes used in mass spectrometry. Med Instrum 9:23-33, 1975 9 Lucas SK, Gardner TJ, Flaherty JT, Bulkley BH, Elmer EB, Gott VL: Beneficial effects of mannitol administration during reperfusion after ischemic arrest. Circulation 62:Suppl 1:34-41, 1980 10 Pratt FH: The nutrition of the heart through the vessels of Thebesius and the coronary veins. Am J Physiol 1:86-103, 1898 II Beck CS: Revascularization of the heart. Ann Surg 128:854-870, 1948 12 Menasche P, Kural S, Fauchet M, Lavergne A, Commin P, Bercot M, Touchot B, Georgiopoulos G, Piwnica A: Retrograde coronary sinus perfusion. A safe alternative for ensuring cardioplegic delivery in aortic valve surgery. Ann Thorac Surg 34:647-657, 1982 13 Solorzano J, Taitelbaum G, Chiu R: Retrograde coronary sinus perfusion for myocardial protection during cardiopulmonary bypass. Ann Thorac Surg 25:201-208, 1978 14 Andreadis P, Natsikas N, Arealis E: The aorto coronary venous anastomosis in experimental acute myocardial ischemia. Vase Surg 8:45-53, 1974 15 Bhayana J, Olsen D, Bryn EJ: Reversal of myocardial ischemia by arterialization of the coronary vein. J THORAC CARDIOVASC SURG 67:125-132, 1974 16 Hochberg M, Roberts W, Morrow A, Austen W: Selective arterialization of the coronary venous system. J THORAC CARDIOVASC SURG 77: 1-11, 1979 17 Brown A, Niles N, Braimbridge EM, Austen W: Retrograde insuflation of gaseous oxygen into the coronary sinus as a means of myocardial maintenance. Arch Surg 105:622-627, 1972 18 Poirier R, Guyton R, MacIntosh C: Drip retrograde coronary sinus perfusion for myocardial protection during aortic cross-clamping. J THORAC CARDIOVASC SURG 70:966-973, 1975 19 Lolley D, Hewitt R, Drapanas T: Retroperfusion of the
The Journal of Thoracic and Cardiovascular Surgery
6 6 6 Bolling et al.
20
21
22
23
24
25
26 27
heart with a solution of glucose, insulin, and potassium during anoxic arrest. J THORAC CARDIOVASC SURG 67:364-369, 1974 Berdeaux A, Farcto J, Bourdarias J, Barry M, Bardet J, Giudicelli J: Effects of diastolic synchronized retroperfusion on regional coronary blood flow in experimental myocardial ischemia. Am J Cardiol 47:1033-1040, 1981 Gundry S: Modification of myocardial ischemia in normal and hypertrophied hearts using diastolic retroperfusion of the coronary veins. J THORAC CARDIOVASC SURG 83:659669, 1982 Gott VL, Gonzalez J, Zuhdi M: Retrograde perfusion of coronary sinus for direct vision aortic surgery. Surg Gynecol Obstet 104:319-328, 1957 Kanter KR, Jaffin JH, Ehrlichman RJ, Flaherty JT, Gott VL, Gardner TJ: Superiority of perfluorocarbon cardioplegia over blood or crystalloid cardioplegia. Circulation 64:Suppl 2:75-80, 1981 Donovan T, Mahkeri VB, Owens G: Myocardial perfusion and metabolism at normothermic and hypothermic levels. Arch Surg 110:208-219, 1975 Bodenhamer RM, DeBoer LW, Geffin GA, O'Keefe DO, Fallon JT, Aretz TH, Haas GS, Daggett WM: Enhanced myocardial protection during ischemic arrest. Oxygenation of a crystalloid cardioplegic solution. J THORAC CARDIOVASC SURG 85:769-780, 1983. Hochberg M, Austen W: Selective retrograde coronary venous perfusion. Ann Thorac Surg 29:578-588, 1980 Bates R, Toscano M, Balderman S, Anagnostopoulos C: The cardiac veins in retrograde coronary venous perfusion. Ann Thorac Surg 23:83-90, 1977
Discussion DR. WILLARD M. DAGGETT Boston. Mass.
The authors' study comprised four groups, two of which had retrograde and two antegrade perfusion. I think the improved preservation in Group tV can be explained purely on the basis of oxygen delivery by the oxygenated perfluorocarbon rather than the route of delivery.
If I understood correctly, Group III had retrograde perfusion with an oxygenated crystalloid solution. I was not sure whether Groups I and II (antegrade perfusion) were given an oxygenated or an unoxygenated crystalloid solution. If the first two groups were not oxygenated, then it would be possible to explain all of the fmdings, not on the route of perfusion, but by delivery of oxygen to the cross-clamped heart. Also, if the volumes of cardioplegic solution delivered to the different groups were not equal, then oxygen delivery would differ. DR. FRANK C. SPENCER New York. N. Y.
Dr. Bolling, may I ask you a question? You will have noticed that in the big puzzle of myocardial preservation is coronary washout with the collaterals. If blood can be perfused retrogradely in the coronary sinus, do you suppose that by putting suction on the right atrium and emptying it out, noncoronary flow can be aspirated the other way, so that it does not wash the potassium out? DR. BOLLING (Closing) Thank you for the comments. Dr. Daggett, all the solutions, including the crystalloid and the perfluorocarbon compounds, were oxygenated to an approximate P0 2 between 400 and 500. In Group II (perfused antegradely) the crystalloid solution was delivered oxygenated, so that Groups II and III both received the same oxygenated crystalloid cardioplegic solution. They both achieved the same mean myocardial temperature in the circumflex area, one via the means of topical cooling and one by retrograde perfusion. Yet Group III, the retroperfused group, demonstrated significantly enhanced return of function. Dr. Spencer, I am not sure of the answer to your question. I think one of the things that would inhibit the clinical use of this method is the ease of use itself. Some have pointed out that a separate atriotomy is necessary. In one of the initial phases of this study, we are developing a balloon catheter that is an integral part of the two-stage venous cannula, which allows very easy entrance into the coronary sinus without a separate atriotomy.