- Email: [email protected]
Percutaneous transluminal coronary angioplasty
J
THoRAc CARDIOVASC SlJRG
92:847-852, 1986
Percutaneous transluminal coronary angioplasty A growing surgical problem The incidence of prior percutaneous transluminal coronary angioplasty in surgical cases is nearly doubling yearly. In 1985, 11.4% of our bypass patients bad one or more prior angioplasties.One hundred thirty-five patients with prior angioplasty are compared to 2,205 patients without angioplasty undergoing surgicalrevascularization. The mortality is 3.2 times higher in the angioplasty patients than in the control patients and the perioperative infarction rate is 2.5 times higher. Forty-four patients were taken directly to the operating room from the catheterization laboratory, 50 were operated on within 10 days, and 41 underwent operation more than 10 days after angioplasty. All of these late failures were of the lesion previomly dilated. The infarction rate was less in patients taken immediatelyto the operating room on an emergency basis than in those whoseoperation was delayed up to 10 days (30% versus70 %). All patients whodied bad angioplasty of the anterior descending coronary artery. Angioplastyof this artery increases operative mortality should surgical treatment become necessary acutely. Patients should be informed before angioplasty of the increased surgical risks after a failed angioplasty procedure.
u. Scott Page, M.D. (by invitation), J. Edward Okies, M.D., Leon Q. Colburn, PA (by invitation), John C. Bigelow, M.D. (by invitation), Neal W. Salomon, M.D. (by invitation), and Albert H. Krause, M.D., Portland. Ore.
In 1985, 11.4%of our patients having pure revascularization had had a prior percutaneous transluminal coronary angioplasty (PTCA). The yearly increase in PTCA rate is nearly doubling (Fig. 1). The majority of these patients were low risk surgical candidates before PTCA. The higher perioperative myocardial infarction and mortality rates led to this comparison.
Patients Between August 1981 and December 1985, 135 patients (PTCA group) had bypass grafting after one or more PTCA procedures. These are compared to 2,205 patients (control group) who had bypass alone during the same period. Patients treated during an acute (0 to 6 hours) myocardial infarction were excluded from both groups. Table I compares age, ejection fraction, number of grafts, reoperation for bleeding, and the use of the internal mammary artery (IMA) as a graft. Use of the From the Division of Cardiovascular and Thoracic Surgery, Good Samaritan Hospital and Medical Center, Portland, Ore. Read at the Sixty-sixth Annual Meeting of The American Association for Thoracic Surgery, New York, N. Y, April 28-30, 1986. Address for reprints: U. Scott Page, M.D., Northwest Surgical Associates, 2226 N.W. Pettygrove, Portland, Ore. 97210.
IMA, age, and ejection fraction lacked statistical significance. Table II compares mortality and myocardial infarction rates in these two groups. The myocardial infarction rate is 2.5 times higher in the PTCA group and the death rate is 3.2 times higher, both differences highly significant (p < 0.005). Myocardial infarction was judged by a creatine kinase MB fraction of 40 or greater 12 or 24 hours postoperatively or by postmortem examination demonstrating myocardial necrosis. In our experience the MB fraction is a much more sensitive and reliable measure of myocardial damage than electrocardiographic change, and for that reason it was used as the criterion of infarction. In most patients undergoing PTCA, either "formal" or "informal" standby was used during PTCA. Formal standby includes pre-PTCA surgical and anesthesia consultations and an operating room completely set up. Informal standby holds the operating room available until the PTCA is completed, usually between scheduled cardiac procedures, without visits by a surgeon or anesthesiologist. A number of our emergency cases occurred at odd hours without any standby, although the surgeon was usually notified that a PTCA was in progress. Twenty-two of the 135 PTCA patients had two 847
The Journal of
8 4 8 Page et al.
Thoracic and Cardiovascular Surgery
Fig. 1. These bars represent the total number of patients undergoing bypass in the years 1981 through 1985. The line and percentages indicate the percentage of total cases represented by patients with a prior PTCA.
Table I. Comparison of parameters for the PTCA and no PTCA groups Parameter
PTCA
Average age (yr) Average ejection fraction (%) Average No. of grafts Use of IMA Reoperation for bleeding (%)
59.0 65.0 1.79 48/88 (54%) 10.4%
I
NoPTCA 61.3 63 2.99 1,178/1885 (62%) 5.7%
Table II. Mortality and morbidity
I p Value NS NS <0.005 0.1387 0.0270
Legend: NS, Not significant.
PTCA procedures. Myocardial infarction developed in two of these 22 patients and both died at operation. They were both elderly (72 and 79 years) women not believed to be surgical candidates initially. The clinical condition at operation differed dramatically between the PTCA and control groups. Operations were divided into four categories: elective-no pain or electrocardiographic changes; urgent-eritical anatomy and controlled unstable angina; emergent---evolving infarction with ongoing pain and/or ST elevations; and desperate--need for cardiopulmonary resuscitation, pressors, or intra-aortic balloon pumping before the operation. There was a shift toward higher risk situations in the PTCA group. There wre more patients in desperate condition (cardiopulmonary resuscitation at some time before bypass) in the PTCA group than in the non-PTCA group (6/135 versus 5/2,205) (Table III). Fig. 2 shows the clinical condition of the PTCA
PTCA
No PTCA
(n= 135)
(n = 2,205)
%
No·1 %
p Value
193 38
<0.005 <0.005
No·1 Myocardial infarction" Operative death
30 8
22.2 5.9
8.1 1.8
*Thc criteria for defining myocardial infarction were a creatine kinase MB fraction of more than 40 or necrosis at postmortem examination.
patients by year. In 1985, 30% of our PTCA patients were in emergent or desperate condition. The PTCA patients have been divided into three groups according to interval from PTCA to operation: Group I, 44 patients taken directly to operation; Group II, 50 patients operated on 0 to 10 days after PTCA; and Group III, 41 patients operated on more than 10 days after PTCA. Group I. Results. When Group I is categorized by character of operation (Table IV), it becomes obvious that the infarction and death rates increase as urgency increases. Infarctions developed in 30% of patients requiring emergency operation. Duration of ischemia before operation was impossible to determine accurately from the records reviewed. Often a PTCA would result in acute occlusion or dissection of the coronary artery. In many of these patients, the operator would continue to redilate or manipulate the lesion for several hours before turning the patient over to the surgeons. In many cases flow was reestablished intermittently during this time, so that the onset of ischemia was difficult to determine. The
Volume 92 Number 5 November 1986
PTCA
849
Fig. 2. In the years 1981 through 1985 the absolute numberof patients with prior PTCA are shown. Each bar is divided into four ranges showing the character of operation.
Table
m. Character of operation-all patients* PTCA (n = 135)
Character of operation
No.
Elective Urgent Emergent Desperate
79 13 37 6
I
Table IV. Character of operation-Group I patients
NoPTCA (n = 2.205)
%
No.
58.5 9.6 27.4 4.4
1801 295 104 5
I
%
Character of operation
No. of patients
81.7 13.4 4.7 0.2
Elective Urgent Emergent Desperate
7 6 26 5
Total
44
"p < 0.001 (2 X 4 X').
Postop. MI
I
Postop deaths
I
%
No.
0 1 8 5
0 16.6 30.8 100
0 0 2 1
0 0 7.7 20.0
14
29.5
3
6.8
No.
%
Legend: MI. Myocardial infarction.
average time from leaving the catheterization laboratory to discontinuing bypass was 100 minutes in single graft patients, 150 minutes in double grafts, and 182 minutes in triple graft patients. All five of the patients whose condition was termed desperate had an infarction, only one died (20%). It appears that the surgical team moved faster in desperate and emergency situations. There was no correlation between these times and infarction rates in this series. Coronaryartery dilated. Table V sorts these Group I patients by coronary site treated by PTCA. The left anterior descending (LAD) was the most common vessel dilated. These 23 patients (52%) had an infarction rate of 35% and a mortality of 13%. Seventeen of these 23 patients had dissection or occlusion of the LAD at PTCA, and there were eight infarctions (47%) and three deaths (I7.6%) after immediate bypass grafting. Dilatation of other coronary vessels was associated with myocardial infarctions but no deaths. In one interesting case a dissection of the right coronary artery (RCA) developed which led to the urgent placement of a vein
Table V. Coronary artery dilated in Group I patients
Vessel
LAD RCA Circumflex Obtuse marginal Total
No. of patients 23 19 2
Postop. M1 No. 8t 6+
It
3
o
47*
15
I % 34.8 31.6 50.0
Postop. deaths No.
3
0
o o o
31.9
3
I
%
13.0
o o o
6.4
'Three patients had three simultaneous PTCAs. tOne of these three patients who had three simultaneous PTCAs had two arteries close acutely. tOne patient needed urgent placement of a graft to the RCA and then had spasm of the RCA and LAD with myocardial infarction and lowoutput (treated by intra-aortic balloon counterpulsation).
graft. Several hours later an anterior ST elevation developed, and on restudy both the LAD and RCA were in severe spasm. The woman survived with intra-aortic balloon pumping and coronary dilators (MB frac-
The Journal of Thoracic and Cardiovascular Surgery
8 5 0 Page et af.
Table VI. Character of operation-Group II patients Character of operation
No. of patients
Elective Urgent Emergent Desperate
32 7 10 I
3 2 7
Total
50
Postop. Ml
Postop. deaths
I
I
%
%
No. I 1*
1
9.4 28.6 70 100
1 I
3.1 14.3 10.0 100
13
26.0
4
8.0
No.
*This patient died of a cerebrovascular accident.
Table VII. Delay to operation in Group II patients Delay to operation (days)
No. of patients
0 I 2 3 4 5-10
3 18 8 6 4 II 50
Total
Postop. deaths
Postop. Ml No.
%
No.
2 4 I 0 3 3
67 22 12.5 0 75 27.3
0 2 0 0 I 1*
13
26.0
4
I
I
Table VIII. Coronary artery dilated in Group II patients Postop. deaths
Postop. Ml
Vessel
No. of patients
LAD RCA Circumflex Obtuse marginal Vein graft
25 18 8 2 2
6 3 3 0 I
24 16.7 37.5 0 50
Total
56*
13
24.1
No.
I %
No.
I %
4t
16
It
12.5
o
o o
o o o
"Two patients had two simultaneous PTCAs and two patients had three arteries dilated. t Onc patient had PTCA of the circumflex and LAD arteries after three previous myocardial infarctions. Five days postoperatively the patient had a severe cerebrovascular accident and died.
% 0 0 25 9.1 8.0
tion = 144). In Group I patients there were 71 grafts (1.61 per patient). Group II. Results. Fifty patients in Group II left the catheterization laboratory and then became unstable or were scheduled for bypass because of a failed PTCA. When these patients are categorized by character of operation (Table VI), the infarction and death rates increase with emergency status. Only 30% of the emergent group were spared infarction compared with 70% in Group I (p = 0.0224). One of the four deaths was due to a stroke on the fifth postoperative day in a patient who did not have a perioperative myocardial infarction but had had three infarctions previously. Attempts to dilate the LAD and circumflex arteries were unsuccessful and both were grafted, without further infarction. This was the only death in Group II in which the delay from PTCA to operation exceeded 4 days (6 days). Table VII shows the interval from PTCA to operation in these 50 Group II patients. Coronary artery dilated. Group II patients are categorized by site of PTCA in Table VIII. Again, the LAD was the most common site and was associated with a high infarction rate and all four cardiovascular deaths. There were 85 grafts in these 50 patients, 1.5 grafts per patient.
Group III. Results. In Group III (41 patients operated on more than 10 days after the last PTCA), 16 patients had had two PTCA procedures (39%). All second PTCA procedures were performed because of restenosis. Eighty-two percent of these repeat PTCA procedures involved the LAD. These 57 PTCA procedures were followed by four infarctions in three patients before bypass grafting. The patient who had two acute infarctions required an RCA graft and anterior aneurysm resection. None of these patients required operation on an emergent or desperate basis, and the mortality and infarction rates are not significantlydifferent from those of the control group. The average time from PTCA to operation in Group III was 101 days with a range of 11 to 780 days. Thirty-two of these patients (78%) had the operation within 90 days of PTCA and nine (22%) more than 90 days after PTCA. These 41 patients had 84 bypass grafts (2.04 grafts/patient). Very low risk patients. Of the total patients, 32 of 135 of the PTCA patients and 230 of 2,205 control patients were judged to be at very low risk when first treated. The criteria for low risk were an ejection fraction greater than 60%, age less than 61 years, and fewer than three grafts. Twenty-three percent of the PTCA group met these criteria and 10.4% of the control group (p < 0.005). Even in these very low risk patients the myocardial infarction rate was five times higher in the PTCA than in the control group (6/32 versus 8/230) (p < 0.005) and the death rate was 7.2 times higher 0/32 versus 1/230), although the difference between groups was not significant (p < 0.094).
Volume 92 Number 5 November 1986
Discussion These data do not demonstrate the risks and benefits of PTCA because the total PTCA experience is not known. We estimate that 750 PTCA procedures were done by our referral sources during this time. One cardiologist has done approximately 300 cases. Twentyeight cardiologists shared the remaining 450 procedures over 5 years. Group I patients then represent a 5.8% (44/750) immediate failure rate and Group II patients another 6.4% (50/750) early failure rate. This is consistent with other reports on early PTCA efficacy, although the Group II patients are often included in the "late failure" category."? It does not appear that the frequency of emergency and desperate situations emanating from our catheterization laboratory is decreasing with experience. In fact, the need to operate on an emergency basis is increasing linearly as PTCA procedures increase. Risks may increase further if neither the surgeons nor anesthesiologists have had the opportunity to review these patients before operating in emergency conditions. In reviewing the PTCA literature, it is particularly disturbing that the inference is made that the need to operate on a patient is only a PTCA failure with slightly higher surgical risk. These data show that PTCA is injurious and a successful operation cannot always reverse this damage. As noted in other studies of this subject, some PTCA failures are associated with distal dissection of the coronary arteries, which makes bypass grafting very difficult or occasionally impossible.v' Three such dissections were present in this review, two in Group I and one in Group II. All three patients had infarctions and one Group I patient died. This must be reflected in PTCA analysis. As shown, the PTCA group reflects the lowest risk surgical patients twice as often as the control group. To compare PTCA against all surgical cases will result in inappropriate bias toward PTCA. Only a randomized study could obtain a valid comparison. In those patients who survive occlusion and operation without infarction, a vein graft was often substituted for an IMA with a known decrease in long-term success. Emergency and desperate operations involve reestablishing blood flow as quickly as possible. IMA dissection would slow this process by 15 to 20 minutes. In addition, the IMA may not be capable of supplying the high flow that ischemic muscle may need early after reperfusion. It is not sufficient to show that PTCA is cheaper than bypass grafting. It must be shown that a PTCA, especially of the LAD, is as safe as grafting and results in no significant deterioration of long-term survival.
PTCA 851
Until this can be shown, indications for PTCA should not be expanded. The PTCA literature suggests experience is essential in selecting patients and performing PTCA safely. When 28 cardiologists are performing, on average, less than three procedures each year, true proficiency is going to be hard to achieve or maintain. Attempting to dilate a second vessel after failing to cross the primary target seems hard to justify. Repeating the PTCA to "tack down" the intima of a dissected vessel suggests a lack of understanding of vascular anatomy and disease. The risks of PTCA do not end when the patient leaves the catheterization laboratory. Analysis of PTCA must include surgical results that are altered by PTCA. It may be possible to reduce infarction and mortality rates if patients with less than an ideal result from PTCA are immediately taken to the operating room.v" Conclusions I. The incidence of prior PTCA is increasing rapidly in patients undergoing bypass grafting. 2. PTCA leading to urgent or emergent coronary bypass increases surgical mortality and morbidity. 3. Patients undergoing PTCA should understand the risks of an emergency operation, especially if the LAD is involved. The damage caused by PTCA cannot always be reversed. 4. Long-term comparison of patients treated initially with IMA-LAD shunting versus PTCA is badly needed. REFERENCES Bredlau CE, Roubin GS, Leimgruber PP, Douglas JS, King SB, Gruentzig AR: In-hospital morbidity and mortality in patients undergoing elective coronary angioplasty. Circulation 72: I044-1052, 1985 2 BlockPC: Percutaneous transluminal coronary angioplasty. Role in the treatment of coronary artery disease. Circulation 72:Suppl 5:161-168, 1985 3 Roubin GS, Gruentzig AR: Coronary angioplasty. Changing indications. Primary Cardiol September 1985, pp 5969
4 Brahos GJ, Baker NH, Ewy G, Moore PJ, Thomas JW, Sanfelippo PM, McVicker RF, Fankhauser DJ: Aortocoronary bypass following unsuccessful PTCA. Experience in 100 consecutive patients. Ann Thorac Surg 40:7-10, 1985 5 Killen DA, Hamaker WR, Reed WA: Coronary artery bypassfollowing percutaneoustransluminal coronary angioplasty. Ann Thorac Surg 40: 133-138, 1985 6 Reul GJ, Cooley DA, Hallman GL, Duncan JM, Livesay JJ, Frazier OH, Ott DA, Angelini P, Massumi A, Mathur VS: Coronary artery bypass for unsuccessful percutaneous transluminal coronary angioplasty. J THoRAc CARDIOVASC SURG
88:685-694, 1984
8 5 2 Page et al.
Discussion DR. ANTHONY J. ACINAPURA New York. N. Y.
The use of PTCA has been increasing rapidly. Review of our early data included 198 PTCAs. PTCA patients were younger and had better left ventricular function than patients having elective coronary artery bypass patients. PTCA was successful in 142 of 198 patients, or 72%. Review of our more recent PTCAs shows a similar vessel distribution and a similar rate of success, 73%, best results occurring in stenotic saphenous vein grafts. Summarizing our entire series, 101 of 404 PTCAs, or 25%, were unsuccessful, with emergency coronary bypass being necessary in 32, or 8%, of the patients. Eleven, or 33%, of these patients requiring emergency coronary artery bypass had perioperative myocardial infarctions. There were no operative deaths. This increased operative morbidity in PTCA patients requiring emergency coronary artery bypass and the continued high restenosis rate of 22% in our series and 30% in the multi-institutional report from the National Heart, Lung, and Blood Institute certainly suggests that a more critical appraisal of the surgical support structure for PTCA be made. I have two questions for Dr. Page. What was the time interval from the onset of acute ischemia to the complete resurgical revascularization, and what were your indications for insertion of a preoperative intra-aortic balloon for counterpulsation? DR. GIORGIO ARU Milwaukee. Wis.
I would like to describe a new device whose use is connected to this topic. This is a very simple tool made for us by Cadman & Shurtleff, Inc., Randolph, Massachusetts (Ref. C. D. No. 04553) from the tip of a coronary probe (I, 1.5, and 2 mm), hollow inside and welded to a spinal needle (respectively 23, 21, and 19 gauge). The tip is extremely smooth and can be easily inserted into the coronary arteries or the IMA). We have found several interesting applications for the needle probe. First, it can make the surgical approach to the failed angioplasty faster and more logical. Most of the time failed angioplasty is an emergency of total ischemia of a major portion of the myocardium, and surgical treatment has to deliver oxygen to the myocardium in the shortest span of time. In addition, we also believe that cardioplegic solution cannot be delivered through a dissected and totally obstructed coronary artery. Therefore, what we propose is to insert the needle probe directly beyond the occlusion and infuse blood before the distal anastomosis is begun. The distal anastomosis is then performed around the probe. At the end of the distal anastomosis the needle probe is now inserted into one side branch of the vein graft, and a bulldog clamp is applied proximally on the vein graft. The proximal anastomosis can then be performed during total perfusion of the heart with the aorta unclamped. Another useful application for the needle probe is the
The Journal of Thoracic and Cardiovascular Surgery
injection of diluted papaverine into the IMA. I have personally seen this technique performed in several institutions in the United States with a No. 22 Angiocath catheter. Magnification of the tip of an Angiocath catheter clearly shows that the edge is very sharp, possibly too sharp for the delicate structure of the IMA. By contrast, the tip of the needle probe appears very smooth. It can be inserted easily into the lumen of the IMA without creating any dissection, which is not an uncommon complication with the Angiocath catheter. Another very interesting application of the needle probe is for delivery of cardioplegic solution beyond a very tight stenosis. Imagine a multiple coronary bypass operation in which saphenous vein grafts are used for three or four vessels and finally an IMA for the LAD. If the LAD has a very tight stenosis or is totally occluded, the protection of such an important area by cardioplegia is marginal and relies only on the collateral circulation during most of the operation. With the needle probe available, the vessel can now be opened and cardioplegic solution injected into the lumen at any time. Finally, the needle probe can be a life-saving tool in the case of accidental transection of the IMA or of a vein graft during redo opening, because the graft can be perfused with blood first and then cardioplegic solution during aortic crossclamping. For this technique one must insert a 16 gauge Angiocath catheter into the aorta, prime a tubing and the needle probe with blood, clamp the tubing, insert the needle probe into the severed graft, open the clamp, and perfuse. Fortunately, we have not had the opportunity to use the needle probe in this situation, but if this situation arises, at least we will be ready. Other applications of the needle probe may be devised, as well. DR. PAGE (Closing) I will answer the questions in reverse order. There were two patients in whom a balloon pump was inserted. In both cases there was no operating room available immediately and there was about an hour's delay. Most of the other patients could be taken to the operating room quickly enough that it did not seem reasonable to stop and insert a balloon. The first question concerned the interval of ischemia. This matter was one of the main reasons that I undertook this investigation. The problem was that we could not determine the ischemic interval because of the situation in the catheterization laboratory. An artery would become occluded and cause ischemia; a balloon would be inserted to open the artery and relieve the ischemia; the artery would close again, again causing ischemia. It was difficult to convince the cardiologists to release the patient for bypass grafting. They kept wanting to try one more time. In some instances they were actually successful during that interval, and the patient was in more stable condition when the operation began than he would have been had the cardiologists not persisted with PTCA. Thus it is very difficult to determine, from old records, when the ischemic interval began. In most patients we instituted cardiopulmonary bypass within about 10 minutes after arriving in the operating room.
THoRAc CARDIOVASC SlJRG
92:847-852, 1986
Percutaneous transluminal coronary angioplasty A growing surgical problem The incidence of prior percutaneous transluminal coronary angioplasty in surgical cases is nearly doubling yearly. In 1985, 11.4% of our bypass patients bad one or more prior angioplasties.One hundred thirty-five patients with prior angioplasty are compared to 2,205 patients without angioplasty undergoing surgicalrevascularization. The mortality is 3.2 times higher in the angioplasty patients than in the control patients and the perioperative infarction rate is 2.5 times higher. Forty-four patients were taken directly to the operating room from the catheterization laboratory, 50 were operated on within 10 days, and 41 underwent operation more than 10 days after angioplasty. All of these late failures were of the lesion previomly dilated. The infarction rate was less in patients taken immediatelyto the operating room on an emergency basis than in those whoseoperation was delayed up to 10 days (30% versus70 %). All patients whodied bad angioplasty of the anterior descending coronary artery. Angioplastyof this artery increases operative mortality should surgical treatment become necessary acutely. Patients should be informed before angioplasty of the increased surgical risks after a failed angioplasty procedure.
u. Scott Page, M.D. (by invitation), J. Edward Okies, M.D., Leon Q. Colburn, PA (by invitation), John C. Bigelow, M.D. (by invitation), Neal W. Salomon, M.D. (by invitation), and Albert H. Krause, M.D., Portland. Ore.
In 1985, 11.4%of our patients having pure revascularization had had a prior percutaneous transluminal coronary angioplasty (PTCA). The yearly increase in PTCA rate is nearly doubling (Fig. 1). The majority of these patients were low risk surgical candidates before PTCA. The higher perioperative myocardial infarction and mortality rates led to this comparison.
Patients Between August 1981 and December 1985, 135 patients (PTCA group) had bypass grafting after one or more PTCA procedures. These are compared to 2,205 patients (control group) who had bypass alone during the same period. Patients treated during an acute (0 to 6 hours) myocardial infarction were excluded from both groups. Table I compares age, ejection fraction, number of grafts, reoperation for bleeding, and the use of the internal mammary artery (IMA) as a graft. Use of the From the Division of Cardiovascular and Thoracic Surgery, Good Samaritan Hospital and Medical Center, Portland, Ore. Read at the Sixty-sixth Annual Meeting of The American Association for Thoracic Surgery, New York, N. Y, April 28-30, 1986. Address for reprints: U. Scott Page, M.D., Northwest Surgical Associates, 2226 N.W. Pettygrove, Portland, Ore. 97210.
IMA, age, and ejection fraction lacked statistical significance. Table II compares mortality and myocardial infarction rates in these two groups. The myocardial infarction rate is 2.5 times higher in the PTCA group and the death rate is 3.2 times higher, both differences highly significant (p < 0.005). Myocardial infarction was judged by a creatine kinase MB fraction of 40 or greater 12 or 24 hours postoperatively or by postmortem examination demonstrating myocardial necrosis. In our experience the MB fraction is a much more sensitive and reliable measure of myocardial damage than electrocardiographic change, and for that reason it was used as the criterion of infarction. In most patients undergoing PTCA, either "formal" or "informal" standby was used during PTCA. Formal standby includes pre-PTCA surgical and anesthesia consultations and an operating room completely set up. Informal standby holds the operating room available until the PTCA is completed, usually between scheduled cardiac procedures, without visits by a surgeon or anesthesiologist. A number of our emergency cases occurred at odd hours without any standby, although the surgeon was usually notified that a PTCA was in progress. Twenty-two of the 135 PTCA patients had two 847
The Journal of
8 4 8 Page et al.
Thoracic and Cardiovascular Surgery
Fig. 1. These bars represent the total number of patients undergoing bypass in the years 1981 through 1985. The line and percentages indicate the percentage of total cases represented by patients with a prior PTCA.
Table I. Comparison of parameters for the PTCA and no PTCA groups Parameter
PTCA
Average age (yr) Average ejection fraction (%) Average No. of grafts Use of IMA Reoperation for bleeding (%)
59.0 65.0 1.79 48/88 (54%) 10.4%
I
NoPTCA 61.3 63 2.99 1,178/1885 (62%) 5.7%
Table II. Mortality and morbidity
I p Value NS NS <0.005 0.1387 0.0270
Legend: NS, Not significant.
PTCA procedures. Myocardial infarction developed in two of these 22 patients and both died at operation. They were both elderly (72 and 79 years) women not believed to be surgical candidates initially. The clinical condition at operation differed dramatically between the PTCA and control groups. Operations were divided into four categories: elective-no pain or electrocardiographic changes; urgent-eritical anatomy and controlled unstable angina; emergent---evolving infarction with ongoing pain and/or ST elevations; and desperate--need for cardiopulmonary resuscitation, pressors, or intra-aortic balloon pumping before the operation. There was a shift toward higher risk situations in the PTCA group. There wre more patients in desperate condition (cardiopulmonary resuscitation at some time before bypass) in the PTCA group than in the non-PTCA group (6/135 versus 5/2,205) (Table III). Fig. 2 shows the clinical condition of the PTCA
PTCA
No PTCA
(n= 135)
(n = 2,205)
%
No·1 %
p Value
193 38
<0.005 <0.005
No·1 Myocardial infarction" Operative death
30 8
22.2 5.9
8.1 1.8
*Thc criteria for defining myocardial infarction were a creatine kinase MB fraction of more than 40 or necrosis at postmortem examination.
patients by year. In 1985, 30% of our PTCA patients were in emergent or desperate condition. The PTCA patients have been divided into three groups according to interval from PTCA to operation: Group I, 44 patients taken directly to operation; Group II, 50 patients operated on 0 to 10 days after PTCA; and Group III, 41 patients operated on more than 10 days after PTCA. Group I. Results. When Group I is categorized by character of operation (Table IV), it becomes obvious that the infarction and death rates increase as urgency increases. Infarctions developed in 30% of patients requiring emergency operation. Duration of ischemia before operation was impossible to determine accurately from the records reviewed. Often a PTCA would result in acute occlusion or dissection of the coronary artery. In many of these patients, the operator would continue to redilate or manipulate the lesion for several hours before turning the patient over to the surgeons. In many cases flow was reestablished intermittently during this time, so that the onset of ischemia was difficult to determine. The
Volume 92 Number 5 November 1986
PTCA
849
Fig. 2. In the years 1981 through 1985 the absolute numberof patients with prior PTCA are shown. Each bar is divided into four ranges showing the character of operation.
Table
m. Character of operation-all patients* PTCA (n = 135)
Character of operation
No.
Elective Urgent Emergent Desperate
79 13 37 6
I
Table IV. Character of operation-Group I patients
NoPTCA (n = 2.205)
%
No.
58.5 9.6 27.4 4.4
1801 295 104 5
I
%
Character of operation
No. of patients
81.7 13.4 4.7 0.2
Elective Urgent Emergent Desperate
7 6 26 5
Total
44
"p < 0.001 (2 X 4 X').
Postop. MI
I
Postop deaths
I
%
No.
0 1 8 5
0 16.6 30.8 100
0 0 2 1
0 0 7.7 20.0
14
29.5
3
6.8
No.
%
Legend: MI. Myocardial infarction.
average time from leaving the catheterization laboratory to discontinuing bypass was 100 minutes in single graft patients, 150 minutes in double grafts, and 182 minutes in triple graft patients. All five of the patients whose condition was termed desperate had an infarction, only one died (20%). It appears that the surgical team moved faster in desperate and emergency situations. There was no correlation between these times and infarction rates in this series. Coronaryartery dilated. Table V sorts these Group I patients by coronary site treated by PTCA. The left anterior descending (LAD) was the most common vessel dilated. These 23 patients (52%) had an infarction rate of 35% and a mortality of 13%. Seventeen of these 23 patients had dissection or occlusion of the LAD at PTCA, and there were eight infarctions (47%) and three deaths (I7.6%) after immediate bypass grafting. Dilatation of other coronary vessels was associated with myocardial infarctions but no deaths. In one interesting case a dissection of the right coronary artery (RCA) developed which led to the urgent placement of a vein
Table V. Coronary artery dilated in Group I patients
Vessel
LAD RCA Circumflex Obtuse marginal Total
No. of patients 23 19 2
Postop. M1 No. 8t 6+
It
3
o
47*
15
I % 34.8 31.6 50.0
Postop. deaths No.
3
0
o o o
31.9
3
I
%
13.0
o o o
6.4
'Three patients had three simultaneous PTCAs. tOne of these three patients who had three simultaneous PTCAs had two arteries close acutely. tOne patient needed urgent placement of a graft to the RCA and then had spasm of the RCA and LAD with myocardial infarction and lowoutput (treated by intra-aortic balloon counterpulsation).
graft. Several hours later an anterior ST elevation developed, and on restudy both the LAD and RCA were in severe spasm. The woman survived with intra-aortic balloon pumping and coronary dilators (MB frac-
The Journal of Thoracic and Cardiovascular Surgery
8 5 0 Page et af.
Table VI. Character of operation-Group II patients Character of operation
No. of patients
Elective Urgent Emergent Desperate
32 7 10 I
3 2 7
Total
50
Postop. Ml
Postop. deaths
I
I
%
%
No. I 1*
1
9.4 28.6 70 100
1 I
3.1 14.3 10.0 100
13
26.0
4
8.0
No.
*This patient died of a cerebrovascular accident.
Table VII. Delay to operation in Group II patients Delay to operation (days)
No. of patients
0 I 2 3 4 5-10
3 18 8 6 4 II 50
Total
Postop. deaths
Postop. Ml No.
%
No.
2 4 I 0 3 3
67 22 12.5 0 75 27.3
0 2 0 0 I 1*
13
26.0
4
I
I
Table VIII. Coronary artery dilated in Group II patients Postop. deaths
Postop. Ml
Vessel
No. of patients
LAD RCA Circumflex Obtuse marginal Vein graft
25 18 8 2 2
6 3 3 0 I
24 16.7 37.5 0 50
Total
56*
13
24.1
No.
I %
No.
I %
4t
16
It
12.5
o
o o
o o o
"Two patients had two simultaneous PTCAs and two patients had three arteries dilated. t Onc patient had PTCA of the circumflex and LAD arteries after three previous myocardial infarctions. Five days postoperatively the patient had a severe cerebrovascular accident and died.
% 0 0 25 9.1 8.0
tion = 144). In Group I patients there were 71 grafts (1.61 per patient). Group II. Results. Fifty patients in Group II left the catheterization laboratory and then became unstable or were scheduled for bypass because of a failed PTCA. When these patients are categorized by character of operation (Table VI), the infarction and death rates increase with emergency status. Only 30% of the emergent group were spared infarction compared with 70% in Group I (p = 0.0224). One of the four deaths was due to a stroke on the fifth postoperative day in a patient who did not have a perioperative myocardial infarction but had had three infarctions previously. Attempts to dilate the LAD and circumflex arteries were unsuccessful and both were grafted, without further infarction. This was the only death in Group II in which the delay from PTCA to operation exceeded 4 days (6 days). Table VII shows the interval from PTCA to operation in these 50 Group II patients. Coronary artery dilated. Group II patients are categorized by site of PTCA in Table VIII. Again, the LAD was the most common site and was associated with a high infarction rate and all four cardiovascular deaths. There were 85 grafts in these 50 patients, 1.5 grafts per patient.
Group III. Results. In Group III (41 patients operated on more than 10 days after the last PTCA), 16 patients had had two PTCA procedures (39%). All second PTCA procedures were performed because of restenosis. Eighty-two percent of these repeat PTCA procedures involved the LAD. These 57 PTCA procedures were followed by four infarctions in three patients before bypass grafting. The patient who had two acute infarctions required an RCA graft and anterior aneurysm resection. None of these patients required operation on an emergent or desperate basis, and the mortality and infarction rates are not significantlydifferent from those of the control group. The average time from PTCA to operation in Group III was 101 days with a range of 11 to 780 days. Thirty-two of these patients (78%) had the operation within 90 days of PTCA and nine (22%) more than 90 days after PTCA. These 41 patients had 84 bypass grafts (2.04 grafts/patient). Very low risk patients. Of the total patients, 32 of 135 of the PTCA patients and 230 of 2,205 control patients were judged to be at very low risk when first treated. The criteria for low risk were an ejection fraction greater than 60%, age less than 61 years, and fewer than three grafts. Twenty-three percent of the PTCA group met these criteria and 10.4% of the control group (p < 0.005). Even in these very low risk patients the myocardial infarction rate was five times higher in the PTCA than in the control group (6/32 versus 8/230) (p < 0.005) and the death rate was 7.2 times higher 0/32 versus 1/230), although the difference between groups was not significant (p < 0.094).
Volume 92 Number 5 November 1986
Discussion These data do not demonstrate the risks and benefits of PTCA because the total PTCA experience is not known. We estimate that 750 PTCA procedures were done by our referral sources during this time. One cardiologist has done approximately 300 cases. Twentyeight cardiologists shared the remaining 450 procedures over 5 years. Group I patients then represent a 5.8% (44/750) immediate failure rate and Group II patients another 6.4% (50/750) early failure rate. This is consistent with other reports on early PTCA efficacy, although the Group II patients are often included in the "late failure" category."? It does not appear that the frequency of emergency and desperate situations emanating from our catheterization laboratory is decreasing with experience. In fact, the need to operate on an emergency basis is increasing linearly as PTCA procedures increase. Risks may increase further if neither the surgeons nor anesthesiologists have had the opportunity to review these patients before operating in emergency conditions. In reviewing the PTCA literature, it is particularly disturbing that the inference is made that the need to operate on a patient is only a PTCA failure with slightly higher surgical risk. These data show that PTCA is injurious and a successful operation cannot always reverse this damage. As noted in other studies of this subject, some PTCA failures are associated with distal dissection of the coronary arteries, which makes bypass grafting very difficult or occasionally impossible.v' Three such dissections were present in this review, two in Group I and one in Group II. All three patients had infarctions and one Group I patient died. This must be reflected in PTCA analysis. As shown, the PTCA group reflects the lowest risk surgical patients twice as often as the control group. To compare PTCA against all surgical cases will result in inappropriate bias toward PTCA. Only a randomized study could obtain a valid comparison. In those patients who survive occlusion and operation without infarction, a vein graft was often substituted for an IMA with a known decrease in long-term success. Emergency and desperate operations involve reestablishing blood flow as quickly as possible. IMA dissection would slow this process by 15 to 20 minutes. In addition, the IMA may not be capable of supplying the high flow that ischemic muscle may need early after reperfusion. It is not sufficient to show that PTCA is cheaper than bypass grafting. It must be shown that a PTCA, especially of the LAD, is as safe as grafting and results in no significant deterioration of long-term survival.
PTCA 851
Until this can be shown, indications for PTCA should not be expanded. The PTCA literature suggests experience is essential in selecting patients and performing PTCA safely. When 28 cardiologists are performing, on average, less than three procedures each year, true proficiency is going to be hard to achieve or maintain. Attempting to dilate a second vessel after failing to cross the primary target seems hard to justify. Repeating the PTCA to "tack down" the intima of a dissected vessel suggests a lack of understanding of vascular anatomy and disease. The risks of PTCA do not end when the patient leaves the catheterization laboratory. Analysis of PTCA must include surgical results that are altered by PTCA. It may be possible to reduce infarction and mortality rates if patients with less than an ideal result from PTCA are immediately taken to the operating room.v" Conclusions I. The incidence of prior PTCA is increasing rapidly in patients undergoing bypass grafting. 2. PTCA leading to urgent or emergent coronary bypass increases surgical mortality and morbidity. 3. Patients undergoing PTCA should understand the risks of an emergency operation, especially if the LAD is involved. The damage caused by PTCA cannot always be reversed. 4. Long-term comparison of patients treated initially with IMA-LAD shunting versus PTCA is badly needed. REFERENCES Bredlau CE, Roubin GS, Leimgruber PP, Douglas JS, King SB, Gruentzig AR: In-hospital morbidity and mortality in patients undergoing elective coronary angioplasty. Circulation 72: I044-1052, 1985 2 BlockPC: Percutaneous transluminal coronary angioplasty. Role in the treatment of coronary artery disease. Circulation 72:Suppl 5:161-168, 1985 3 Roubin GS, Gruentzig AR: Coronary angioplasty. Changing indications. Primary Cardiol September 1985, pp 5969
4 Brahos GJ, Baker NH, Ewy G, Moore PJ, Thomas JW, Sanfelippo PM, McVicker RF, Fankhauser DJ: Aortocoronary bypass following unsuccessful PTCA. Experience in 100 consecutive patients. Ann Thorac Surg 40:7-10, 1985 5 Killen DA, Hamaker WR, Reed WA: Coronary artery bypassfollowing percutaneoustransluminal coronary angioplasty. Ann Thorac Surg 40: 133-138, 1985 6 Reul GJ, Cooley DA, Hallman GL, Duncan JM, Livesay JJ, Frazier OH, Ott DA, Angelini P, Massumi A, Mathur VS: Coronary artery bypass for unsuccessful percutaneous transluminal coronary angioplasty. J THoRAc CARDIOVASC SURG
88:685-694, 1984
8 5 2 Page et al.
Discussion DR. ANTHONY J. ACINAPURA New York. N. Y.
The use of PTCA has been increasing rapidly. Review of our early data included 198 PTCAs. PTCA patients were younger and had better left ventricular function than patients having elective coronary artery bypass patients. PTCA was successful in 142 of 198 patients, or 72%. Review of our more recent PTCAs shows a similar vessel distribution and a similar rate of success, 73%, best results occurring in stenotic saphenous vein grafts. Summarizing our entire series, 101 of 404 PTCAs, or 25%, were unsuccessful, with emergency coronary bypass being necessary in 32, or 8%, of the patients. Eleven, or 33%, of these patients requiring emergency coronary artery bypass had perioperative myocardial infarctions. There were no operative deaths. This increased operative morbidity in PTCA patients requiring emergency coronary artery bypass and the continued high restenosis rate of 22% in our series and 30% in the multi-institutional report from the National Heart, Lung, and Blood Institute certainly suggests that a more critical appraisal of the surgical support structure for PTCA be made. I have two questions for Dr. Page. What was the time interval from the onset of acute ischemia to the complete resurgical revascularization, and what were your indications for insertion of a preoperative intra-aortic balloon for counterpulsation? DR. GIORGIO ARU Milwaukee. Wis.
I would like to describe a new device whose use is connected to this topic. This is a very simple tool made for us by Cadman & Shurtleff, Inc., Randolph, Massachusetts (Ref. C. D. No. 04553) from the tip of a coronary probe (I, 1.5, and 2 mm), hollow inside and welded to a spinal needle (respectively 23, 21, and 19 gauge). The tip is extremely smooth and can be easily inserted into the coronary arteries or the IMA). We have found several interesting applications for the needle probe. First, it can make the surgical approach to the failed angioplasty faster and more logical. Most of the time failed angioplasty is an emergency of total ischemia of a major portion of the myocardium, and surgical treatment has to deliver oxygen to the myocardium in the shortest span of time. In addition, we also believe that cardioplegic solution cannot be delivered through a dissected and totally obstructed coronary artery. Therefore, what we propose is to insert the needle probe directly beyond the occlusion and infuse blood before the distal anastomosis is begun. The distal anastomosis is then performed around the probe. At the end of the distal anastomosis the needle probe is now inserted into one side branch of the vein graft, and a bulldog clamp is applied proximally on the vein graft. The proximal anastomosis can then be performed during total perfusion of the heart with the aorta unclamped. Another useful application for the needle probe is the
The Journal of Thoracic and Cardiovascular Surgery
injection of diluted papaverine into the IMA. I have personally seen this technique performed in several institutions in the United States with a No. 22 Angiocath catheter. Magnification of the tip of an Angiocath catheter clearly shows that the edge is very sharp, possibly too sharp for the delicate structure of the IMA. By contrast, the tip of the needle probe appears very smooth. It can be inserted easily into the lumen of the IMA without creating any dissection, which is not an uncommon complication with the Angiocath catheter. Another very interesting application of the needle probe is for delivery of cardioplegic solution beyond a very tight stenosis. Imagine a multiple coronary bypass operation in which saphenous vein grafts are used for three or four vessels and finally an IMA for the LAD. If the LAD has a very tight stenosis or is totally occluded, the protection of such an important area by cardioplegia is marginal and relies only on the collateral circulation during most of the operation. With the needle probe available, the vessel can now be opened and cardioplegic solution injected into the lumen at any time. Finally, the needle probe can be a life-saving tool in the case of accidental transection of the IMA or of a vein graft during redo opening, because the graft can be perfused with blood first and then cardioplegic solution during aortic crossclamping. For this technique one must insert a 16 gauge Angiocath catheter into the aorta, prime a tubing and the needle probe with blood, clamp the tubing, insert the needle probe into the severed graft, open the clamp, and perfuse. Fortunately, we have not had the opportunity to use the needle probe in this situation, but if this situation arises, at least we will be ready. Other applications of the needle probe may be devised, as well. DR. PAGE (Closing) I will answer the questions in reverse order. There were two patients in whom a balloon pump was inserted. In both cases there was no operating room available immediately and there was about an hour's delay. Most of the other patients could be taken to the operating room quickly enough that it did not seem reasonable to stop and insert a balloon. The first question concerned the interval of ischemia. This matter was one of the main reasons that I undertook this investigation. The problem was that we could not determine the ischemic interval because of the situation in the catheterization laboratory. An artery would become occluded and cause ischemia; a balloon would be inserted to open the artery and relieve the ischemia; the artery would close again, again causing ischemia. It was difficult to convince the cardiologists to release the patient for bypass grafting. They kept wanting to try one more time. In some instances they were actually successful during that interval, and the patient was in more stable condition when the operation began than he would have been had the cardiologists not persisted with PTCA. Thus it is very difficult to determine, from old records, when the ischemic interval began. In most patients we instituted cardiopulmonary bypass within about 10 minutes after arriving in the operating room.