| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 1998;14:480-487
© 1998 Elsevier Science NL
a Cardiac Surgery Department, Gasthuisberg University Hospital, Katholieke Universiteit, Herestraat 49, 3000 Leuven, Belgium
b Division of Cardiothoracic Surgery, University of Alabama, Birmingham, AL, USA
Received 30 September 1997; received in revised form 19 July 1998; accepted 28 July 1998.
* Corresponding author. Tel.: +32 16 344339; fax: +32 16 344616; e-mail: paul.sergeant@uz.kuleuven.ac.be
 |
Abstract |
|---|
 Top Abstract Introduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
Key Words: Coronary bypass surgery • Reintervention • Arterial grafting • Multivariable analysis
|
Introduction |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
We, therefore, studied the nature of reintervention after CABG in order to dissect, if possible, this confounding and to provide thereby a basis for assessing the costs and benefits of reintervention.
|
Methods |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
Surgical technique
One, two, three, four or five bypass grafts were constructed respectively in 12, 42, 38, 7 and 1% of the patients (total N=9600). One and two mammary artery grafts revascularized, in situ, the coronary circulation in 50.9 and 12.4% of the patients. Bilateral in situ mammary artery grafts were constructed to a single coronary system in 204 patients and to two coronary systems in 987 patients but the use and method of use of arterial in situ grafts has been inconsistent over time.
One, two, three and four in situ mammary anastomoses were present in 45.3, 16.6, 1.9 and 0.03% of the patients. Free mammary and gastro-epiploic artery grafts were used in addition in 122 and 43 patients.
Follow-up
The K.U. Leuven Coronary Surgery Data Base is updated through a regular stream of follow-up reports by the referring specialists. These reports are sent after each event, suspicion of event or regular visit. In addition to these non-systematic follow-up reports, a formal cross-sectional follow-up of known surviving patients was undertaken between 1 January 1993 and 1 July 1994, as previously described
[1]. The common closing date for outcome information was set at 1 January 1993, all events after that date were excluded from analysis. The follow-up was complete for 99.9%.
The dataset was further refined for intervals without any information, for time frames around and or immediately following important clinical events.
Event
The event was defined as the first reintervention for ischemic heart disease. This reintervention could be a cardiological intervention such as a PTCA or similar procedure on a cardiosurgical intervention such as isolated or combined CABG, mitral valve repair or replacement for ischemic valve disease and cardiac transplantation. These events were recorded and analyzed whenever and wherever this event took place, irrespective of the institution, country or continent. Patients not experiencing the event were censored at the common closing date, at the end of follow-up for those with incomplete follow-up or a death, whichever occurred earliest.
Statistical method
Non-parametric estimates of overall non-risk-adjusted freedom from reintervention were obtained by the method of Kaplan and Meier
[3]. A completely parametric method
[4] was used to resolve the number of hazard phases, identify the form of the equation for each phase, and estimate the parameters that characterize the distribution of times until first reintervention. The general methods used to identify incremental risk factors for reintervention have been previously described
[1]
[2], including the variables and their organization for analysis, the exploratory analyses accompanying the multivariable analyses conducted in the parametric, multiphase hazard-function domain. In the directed stepwise entry of variables into the multivariable risk factor model, the P-value criterion used for retention of variables in the final analysis was 0.05. Exploration of the influence of risk factors in the parametric multivariable analysis was performed by constructing nomograms representing the solution of the equation for the values of each risk factor of a patient of median risk (see Appendix A) in the last year of the study. To better understand the nature of reintervention, further study was made of the time-related freedom from reintervention after post-CABG return of angina and infarct, survival and freedom from angina after reintervention.
|
Results |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
|
From 1973 to 1980 these reinterventions were exclusively surgical, the cardiological reinterventions increased from that date, gradually up to 1987 to form ±45% of all first reinterventions. This ratio has stabilized since 1987.
Influence of incremental risk factors after risk-adjustment
Table 1 lists the incremental risk factors associated with reintervention after CABG and the direction of their relation. In multi-phase hazard function analysis, the influence of a given risk factor is related both to the magnitude of the risk factor coefficient and to the magnitude of the hazard phase it modifies. All variables related to the late phase, have therefore a possible large impact on the event incidence rate.
|
|
|
|
The 1-month and 10-year non-risk-adjusted freedom from any cardiological or cardiosurgical reintervention after post-CABG return of angina was 96 and 59%, and after post-CABG infarct was 79 and 63%. This freedom from any reintervention after return of angina is very strongly influenced, before risk-adjustment, by the degree of angina at first return (P=0.0001), the presence or absence of an arterial graft at first CABG procedure (P=0.0026) and by increasing number of in situ internal mammary artery distal anastomoses (P=0.0003) ( Fig. 5 ). Reintervention is common after none and one distal arterial graft but rare in the presence of two or more arterial distal anastomes.
|
The 1-month and 10-year non-risk-adjusted survival after a first post-CABG cardiological or cardiosurgical reintervention was 95 and 73% ( Fig. 6 ). The type of reintervention (surgical or cardiological), without risk-adjustment, influenced significantly (P=0.0003) the form of the survival curve ( Fig. 7 ) but both curves crossed one another at 6 years after reintervention.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
The analysis concerns the clinical practice of reintervention after primary CABG and not the actual need for reintervention. The decision of the clinician might have been influenced by his appreciation of the patient's physical reserves, his life-expectancy, the ventricular mass at risk, the possible risks and benefits of a particular intervention and the patient's appreciation of an intervention. Any analysis of the variables influencing the prevalence of reintervention could therefore be biased.
The drive towards early reintervention
The dramatic circumstances of the early (first 30 h) reinterventions, sometimes massive ischemia, rarely unannounced ventricular tachycardia or fibrillation, were indicative of an imminent life-threatening situation. These symptoms created the drive towards reintervention and left little room for bias.
Fast-track surgery where the patients are discharged to the normal wards, 6–12 h after the start of the procedure, may have these events happen in a lesser monitored environment with a possible additional risk. Related to this small early hazard is the extremely unstable anginal syndrome, also procedural variables and demographic variables increasing the procedural difficulty. Surgeons' experience seems therefore logical in reducing this hazard. Fast track makes economical and medical sense in patients with stable anginal symptoms, without a previous myocardial infarct and having a complete revascularization by an experienced surgeon.
The drive towards and against late reintervention
Seven-hundred and thirty-eight PTCA- and isolated CABG-reinterventions (95% of all first reinterventions) after primary CABG were performed to resolve or attenuate anginal symptoms and or possibly to improve survival. An exploration of the prevalence and consequences of return of angina, survival after the return of angina and freedom from reintervention after return of angina is mandated.
Not all patients with return of angina underwent a reintervention, even in the first 10 years after the return of symptoms. Fig. 2 superimposes, for a median-risk patient with varying age, the 10-year freedom from angina and the 10-year freedom from any reintervention. If all return of angina was treated by reintervention, the two curves would be superimposed. The separation of the two lines indicates that a large proportion of patients were not intervened upon after return of angina, but that nearly all patients in the 45–55 year age range at the time of surgery underwent a reintervention after post-CABG return of angina.
Appropriateness of medical care restricts isolated and combined repeat-CABG procedures to the septuagenarians and octogenarians with residual physical and mental reserves. The further separation of the two lines with increasing age above the median age explains the additional physicians reluctance of submitting older patients to reinterventions. Freedom from reintervention is therefore certainly biased by the increasing age of the patient.
Whereas the effect of older age could easily be explained, the fewer reinterventions (but with wide confidence limits, due to the limited number of patients at risk) of any type in younger patients was a surprise. Stratified Kaplan–Meier analysis confirmed this younger and older information.
The presence of severe left main stenosis obstructs the pathway for cardiological reinterventions on the native coronaries of the left side of the heart and may create a possible technical bias against these cardiological reinterventions.
Since medical technology improves and medical decision changes, it was interesting to study the influence of date of surgery on the prevalence of reinterventions. Some of the types (PTCA and transplantation) of reinterventions were not available in the first decade of coronary surgery at our institution; the increasing interventional approach of cardiologists and cardiac surgeons in the treatment of symptoms after CABG can be observed daily in the clinical practice. Therefore the presence (Table 1) of a major relation between the date of surgery and the prevalence of reintervention, late in follow-up, is comprehensible, even after risk-adjustment. This relation ( Fig. 4) brings another bias into the scientific analysis of reintervention after CABG.
The increasing prevalence of reinterventions after primary CABG in the late eighties and seemingly further in time, identified by ourselves in this study but also by others [5], as well as the economical aspects of these interventions, focuses our attention on the possible cost-benefit in survival. If there would be an improved survival with this more recent evolution then this should be apparent in the analysis of post-CABG survival. The late survival should improve for the patients with a more recent date of surgery.
There is an important relation between more recent date of surgery and improved late survival [1] (after correction for variability by multivariable time-related parametric analysis). Fig. 4 is the depiction of this relation for a median patient with complete risk-adjustment for all patient-, procedure- and institutional variability. This effect is most conspicuous in the 1970–1980 intervals after which there is no improved survival. Thus there is no evidence of improved late survival in parallel with the considerable increase in late reinterventions. This absence of benefit for a median patient could be due to the 5% first month survival cost, uncorrected for type of reintervention and patient-variability ( Fig. 6 Fig. 7) of any first reintervention.
Return of angina after primary CABG [2] is rarely an immediate threat for survival, with an initial survival cost of 5% in the first year. The 10-year survival after return of angina is still 68%. Cardiosurgical reinterventions, due to their substantial initial survival cost [6] [7] [8], ( Fig. 7) require an expected major benefit in survival. They should probably only be performed for anginal relief in patients with low periprocedural risk, because of the good survival after return of angina. A proximal stenosis of the anterior descending artery, with considerable area at risk and perfused by a severely stenotic venous graft or no graft at all, increases the potential benefit of a cardiosurgical intervention, preferably with arterial anterior reconstruction.
Patient-variables, such as severely damaged ventricular function and older age, with their impact on early and late survival [1] will influence negatively the cost-benefit ratio of this surgical alternative. This same finding was observed by Brener et al. [9]. A prophylactic cardiosurgical reintervention in asymptomatic patients after 10 or 15 years for `upgrading' the grafts seems to have an unfavourable cost-benefit ratio in most patients.
The lesser early jeopardy for cardiological reinterventions ( Fig. 7) goes in parallel with lesser late benefit [10] ( Fig. 8), certainly for lesions in venous grafts [11]. In the presence of an open arterial graft to the antero-septal region and enough muscular area at risk, seems to warrant a cardiological reintervention, even repeated if feasible.
The interaction of arterial grafting on the drive towards and on the occurrence of reintervention
Reintervention is strongly reduced with an increasing number of internal mammary artery anastomoses (
Fig. 3). Please note that the confidence limits for the difference with and without a single arterial anastomosis (preferably to the LAD) overlap for a median patient, thereby indicating that in this low risk median patient the difference could be due to chance alone.
Differences in the occurrence of reintervention with increasing arterial grafting is no proof of diverging need for reintervention. The possible lines of thinking explaining this relationship between multiple arterial grafting and lesser late reinterventions are explored. (1) Lesser return of angina after CABG with multiple arterial grafting, thereby creating a lesser need for reintervention. (2) Angina arising from new territories not previously grafted, but possibly such that the surgeon does not have another arterial graft to place, or is fearful of damaging the functioning grafts. (3) Angina arising from new territories not previously grafted, but possibly such that the attending physician estimates the areas at risk insufficient to warrant the risk of the reintervention. (4). Angina arising from areas revascularized by venous grafts in the presence of multiple arterial grafts, for which PTCA or CABG may be possible, but where its effectiveness might be so questionable that it is not done. (5) Angina arising from areas supplied by arterial grafts, which generally would not be amenable to PTCA, and where the surgeon may think that he cannot improve the distal revascularization. (6) General reluctance by surgeons to attempt revascularization in the face of multiple arterial grafts they wish not to disturb.
No demonstrable benefit in survival [1] or return of angina [2] for extensive arterial grafting could be identified after adequate correction for patient-variability. So other mechanisms are responsible for this interaction between an increasing number of arterial anastomoses and a reduced occurrence of reinterventions. An interesting investigation into this maze is the analysis of reintervention after the return of angina. Fig. 5 depicts the reintervention after return of angina, stratified (univariably) by increasing the number of arterial distals. This plot corrects, therefore, already for variability in the return of angina. Reintervention overall is rare after the return of angina, certainly during the first years. It is rather common after one distal arterial graft but rare in the presence of two or more arterial distal anastomoses. This was further confirmed after multivariate time-related risk adjustment.
Thus, the observation remains: the more extensive the arterial grafting, the less likely it is that reintervention will be done after the return of angina. This supports the idea that physicians and surgeons are more reluctant to reintervene after arterial grafting despite return of angina, even though Lytle [12] could not identify an increased periprocedural risk in the presence of patent arterial grafts. Our institutional appreciation agrees with these findings but the dataset studied is representative for the physician's appreciation wherever these original K.U. Leuven patients were treated.
|
Conclusions |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Present address: Cleveland Clinic Foundation, Cleveland OH, USA.
|
Patient-profile of the median patient used for the nomograms and patient-specific predictions |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
|
Appendix A. Conference discussions |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
Dr U. Althaus (Bern, Switzerland): You have found that the use of several arterial grafts reduced the incidence of redo surgery in spite of the minimal impact on reoccurrence of clinical symptoms. That would mean that the clinical criteria you have applied in the indication of reintervention have not been the same, have not been identical in the two subsets of patients, in the group of extensive arterial grafting and in the other group where patients received predominantly venous grafts. Could you comment on that question?
Dr. Sergeant: Yes. First, I would like to clarify that the patients, the primary 9600, were treated at our institution. The reintervention, will be cardiological or cardiosurgical, took place in our institution, in another institution, in Belgium, or in any other institution in the world. So this means that the decision-making process about reintervention was not only taking place at our own institution but was taken into account wherever that has happened.
We have been able to identify that the event reintervention does not behave in parallel with the event return of angina.
This bias can also be due to the surgeon. Some surgeons have been reluctant to go back in as soon as there was one arterial graft. At our own institution we definitely have not been reluctant to go in when there was one arterial graft present. Maybe there is a higher step once there are two or three arterial grafts present. It might have been that the area at risk was too small, that the possible benefit was too small. So it's a very complex reasoning.
It might be that, in fact, as normal surgery should have been performed, if you have done two or three mammary artery anastomosis, you've anastomosed to the most important vessels, so the return of angina can be due to a smaller vessel in a smaller area. But for us what has been the most important finding is that the return of angina is not a lethal disease. Also in the presence of, at least, one arterial graft to the anterior wall, we should be very reluctant to intervene, certainly for the surgical reintervention. The cardiological intervention goes with one-fifth of the risk of the surgical reintervention. But, of course, repeated cardiological reinterventions accumulate to the same amount of risk of a cardiosurgical reintervention with less benefits.
Dr J. Pepper (London, UK): I think that in the context of reintervention, survival is important. But most of us operate because of severe symptoms and therefore data on the quality of life would arguably, be as or even more important than survival. And do you have any of that data for reintervention?
Dr Sergeant: We do not yet have these particular data and I agree completely with you because that has been our own philosophy. Quality of life comes after survival. This 5% mortality needs to be corrected for patient variability. This means that it will go from 1% for a particular patient, but up to 20–30% for another patient related to their condition.
Improving the quality of life at a cost of 5–10% mortality is probably something which gets very questionable, whatever that improvement will be, because if you're not there it cannot be improved. So for myself, who has been doing quite a lot of reinterventions I have become more critical, and I think the conclusion from this presentation, is that I have probably performed reinterventions, which I should not have done. In some of these patients I have not improved the quality of the patients because they have not survived the procedure. What is acceptable mortality to improve quality of life?
|
References |
|---|
|
TopAbstractIntroduction
MethodsResultsDiscussionConclusions
Patient-profile of the median...Appendix A. Conference...References |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. M. van Eck, L. Noyez, F. W.A. Verheugt, and R. M.H.J. Brouwer Changing profile of patients undergoing redo-coronary artery surgery Eur. J. Cardiothorac. Surg., February 1, 2002; 21(2): 205 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. B. Luciani, G. Montalbano, G. Casali, and A. Mazzucco Predicting long-term functional results after myocardial revascularization in ischemic cardiomyopathy J. Thorac. Cardiovasc. Surg., September 1, 2000; 120(3): 478 - 489. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. H. Blackstone and B. W. Lytle COMPETING RISKS AFTER CORONARY BYPASS SURGERY: THE INFLUENCE OF DEATH ON REINTERVENTION J. Thorac. Cardiovasc. Surg., June 1, 2000; 119(6): 1221 - 1232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Dion, D. Glineur, D. Derouck, R. Verhelst, P. Noirhomme, G. El Khoury, E. Degrave, and C. Hanet Long-term clinical and angiographic follow-up of sequential internal thoracic artery grafting Eur. J. Cardiothorac. Surg., April 1, 2000; 17(4): 407 - 414. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
