Author + information
- Received March 23, 1999
- Revision received February 16, 2000
- Accepted March 27, 2000
- Published online July 1, 2000.
- Frits Bär, MD, PhDa,* (, )
- Jindra Vainer, MDa,
- Jeroen Stevenhagen, MSc∗,
- Kars Neven, MSc∗,
- Rob Aalbregt, MSc∗,
- Ton Oude Ophuis, MD, PhD†,
- Vincent van Ommen, MD, PhDa,
- Hans de Swart, MDa,
- Ebo de Muinck, MD, PhDa,
- Willem Dassen, PhDa and
- Hein Wellens, MD, PhD, FACCa
- ↵*Reprint requests and correspondence: Dr. Frits Bär, Department of Cardiology, University Hospital Maastricht, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands
How effective and safe is rescue percutaneous transluminal coronary angioplasty [PTCA] compared with primary PTCA, and is it cost effective?
In acute myocardial infarction (AMI), primary PTCA has been shown to be beneficial in terms of clinical outcome. In contrast, the value of rescue PTCA has not been established.
In a retrospective analysis, we compared the angiographic and clinical outcomes of 317 consecutive patients who had rescue PTCA ∼90 min after failed thrombolysis and 442 patients treated with primary PTCA. An estimation of interventional costs was compared with the strategies of primary and rescue PTCA or with the strategy of thrombolysis with rescue PTCA, when indicated.
Baseline characteristics between primary and rescue PTCA were comparable for most variables. Treatment delay was longer for patients who had rescue PTCA: 240 min. versus 195 min. Coronary patency after PTCA was comparable: 90.2% for rescue PTCA and 91.4% for primary PTCA (p = 0.67, power 71.9%). In-hospital mortality rates were 4.7% and 6.6%, respectively (p = 0.37). Also, the other complications were fairly similar during the in-hospital phase and during one-year follow-up. Predictors of death were age, infarct size, localization of AMI, failed PTCA and left main stem occlusion. The estimated interventional costs during one-year follow-up were $7,377 for primary PTCA and $8,246 for rescue PTCA: difference $869 (11.7%).
In this retrospective analysis of 759 patients with AMI, rescue angioplasty early after failed thrombolysis seems to be as effective and safe as primary PTCA. In the present evaluation, interventional costs of primary PTCA are less than those of rescue PTCA (p = 0.0001).
Thrombolytic therapy and primary percutaneous transluminal coronary angioplasty (PTCA) have been shown to be beneficial in improving the angiographic and clinical outcome of patients with acute myocardial infarction (AMI) (1–4). The value of rescue PTCA early after failed thrombolysis is still controversial (5–8).
At the University Hospital of Maastricht, Maastricht, The Netherlands, all three forms of these therapies have been used since 1987. This retrospective analysis will show the angiographic and clinical outcomes, including one-year follow-up of rescue PTCA (n = 317), which will be compared with the accepted form of therapy: primary PTCA (n = 442). Over a 10-year period (from 1987 to 1997), all 759 patients had angioplasty in the acute phase of their AMI.
In our hospital, patients with clinical and electrocardiographic (ECG) signs of AMI are treated with thrombolytic therapy or PTCA if the chest pain lasts ≥30 min. Initially, only patients up to 70 years of age had such therapy in case treatment delay between chest pain and intervention was <4 h. At a later stage (since 1989), in principle, no upper age criterion was used, and the treatment delay increased to 6 h. In case of persistent pain and ST segment elevation, patients with longer delays could also undergo an intervention. Electrocardiographic criteria for an intervention were ST segment elevation of at least 1 mm in two or more extremity leads or ST segment elevation of at least 2 mm in two or more precordial leads. During the 10-year period of the analysis, the number of patients at the Maastricht Hospital with the final diagnosis AMI was 2,310. Of these patients, 901 (39%) were treated with thrombolysis, 196 (21.8%) of whom underwent rescue PTCA (9,10). In addition, 121 patients were referred from five other hospitals for rescue PTCA (11,12). In the same period, 287 of the patients with AMI (12.4%) who were directly admitted to the Maastricht Hospital had primary PTCA, and 155 patients were referred for such an intervention. Two hundred forty-one of the patients who had rescue PTCA were included in thrombolysis studies (11,13–17), and 75 patients with primary PTCA were involved in the LImburg Myocardial Infarction (LIMI) trial (11). The reasons to decide for thrombolytic therapy or primary PTCA cannot accurately be described in this retrospective analysis. Initially, the primary choice was thrombolysis, unless a contraindication was present. For example, in 1987, 20 patients had rescue PTCA and only eight patients underwent primary PTCA; in 1996, the numbers were reversed to 72 patients and 111 patients, respectively. Because of the excellent results of primary PTCA, as reported in the published data, a gradual change in attitude toward primary PTCA was observed. In addition, in 1996, the LIMI study was under way (11). In our institution, catheterization service is available on a 24-h basis.
Routine treatment of the AMI study group was as follows: First, patients were given intravenous nitroglycerin and a bolus of 5,000 U heparin, followed by an intravenous heparin drip titrated according to the activated partial thromboplastin time (2 to 2.5 times the normal value). Since 1988, patients were also pretreated with aspirin. Patients treated with thrombolysis received one of the following thrombolytic agents: streptokinase, alteplase, anistreplase, reteplase, lanotaplase or saruplase, in the recommended dosages. None of the patients was treated with a IIb/IIIa receptor blocker.
The median delay between admission and cardiac catheterization of the patients directly admitted to the Maastricht Hospital was ∼90 min. At that time another 5,000 U heparin was given. Patients who were involved in angiographic thrombolysis studies typically had early catheterization at 60 to 90 min. In the other patients who received thrombolytic therapy, angiography was performed if ST segment resolution (≥50%) was not observed (18,19). Patients who had cardiac catheterization with normal coronary perfusion grades after (apparently successful) thrombolytic therapy were not considered to be candidates for acute PTCA because of the results of the Thrombolysis In Myocardial Infarction (TIMI) II-A trial and the European Cooperative Study (20,21). Angiographically proven complete reperfusion (TIMI flow grade 3) after thrombolytic therapy was observed in 50% to 70% of the patients who were included in thrombolytic trials in which early angiography was required (13–15). If thrombolytic therapy failed or if primary PTCA was the therapy of choice, coronary angioplasty was performed. In 5% of patients, PTCA was also not performed if it appeared unlikely that the patient would benefit from PTCA, because only the occluded artery had perfused a small area. In another 5% of patients, PTCA was not done because of technical reasons or because of the complexity of the coronary anatomy.
Successful PTCA was defined as complete restoration of flow (TIMI flow grade 3) and residual stenosis <50% (qualitative assessment) 10 min after the last balloon inflation. After successful PTCA, an additional bolus of 5,000 U heparin and 300 mg intravenous aspirin were given.
During recent years, in an increasing number of patients, a stent was implanted in case of insufficient or moderate angiographic results. Initially, such patients were treated with anticoagulant therapy. Since 1996, ticlopidine was given instead of Coumadin in case of a stent implantation. This drug was continued for four weeks in a dosage of 500 mg/day.
Except for 11 patients who were lost to follow-up, data are available for the first year after myocardial infarction. Data on outcome during admission and one-year follow-up will be given separately.
Costs were calculated using estimates of unit costs of the interventions during the first year after AMI. The methods and calculated interventional costs of Zijlstra et al. (22) were used. Admission duration was similar for both groups. Medication at discharge and during follow-up was also comparable. For this reason, these two components were not introduced in the model. Also, indirect costs were not taken into account.
Continuous variables are expressed as the mean value ± SD or as the median value (range), depending on their distribution. The mean and median values of two groups were compared using the Student unpaired t test and the Mann-Whitney U test, respectively. Categoric variables were compared using the chi-square test, with a continuity correction in case of 2 × 2 tables. Survival was depicted using the Kaplan-Meier curve, with the log-rank test statistics for equality of survival distribution. A p value <0.05 was considered to be statistically significant.
The success rate in the primary PTCA group is 91.4%. The aim of this study was to demonstrate that rescue PTCA treatment has the same success rate or differs by no more than 5% as compared with standard treatment. This difference of ≤5% is considered as clinically not relevant. On the basis of the actual number of patients in both groups and a p-value of 0.05 (one-sided), this study has a power of 71.9% to reject the hypothesis that the difference is >5% (23).
The baseline clinical and angiographic characteristics of the 759 patients who underwent rescue or primary PTCA are presented in Table 1. Rescue PTCA was performed in 317 patients (41.8%) and primary PTCA in 442 patients (58.2%). In total, 483 patients were primarily admitted to the Maastricht Hospital, whereas 276 patients were referred from other hospitals (Table 2). Most baseline data of the two types of treatment were comparable. Of the known risk factors, diabetes was more frequent in the primary PTCA group. More patients with a left circumflex artery (LCx) (14.9% vs. 9.8%) and main stem lesions (1.1% vs. 0.3%) had primary PTCA rather than rescue PTCA. Note that the median delays between chest pain and PTCA were acceptable (204 min for primary PTCA and 269 min for rescue PTCA). Obviously, the delay between onset of pain and PTCA was longer for patients who were transferred from other hospitals, especially for patients who underwent rescue PTCA (Table 2).
Before the intervention, 8 of the 759 patients had an open vessel and TIMI flow grade 2, whereas all of the other 751 patients had a total occlusion and TIMI flow grade 0 or 1. The angiographic results showed that after the PTCA, 90.9% of patients had a patent culprit vessel and good TIMI flow grade 3 (Table 3). In this study, the effect size of rescue PTCA as compared with primary PTCA is −1.2%, with a 90% confidence interval of −4.7% to −2.3%.
Success rate was independent of pretreatment with thrombolytic agent. The complication rates of the invasive procedure were also not different. In patients who underwent rescue PTCA, somewhat more stents were used to achieve an optimal result (12.9% vs. 8.2% for primary PTCA). The number of stents implanted increased gradually over time. The first stent was implanted in 1993. In 1997, 29.2% of patients received a stent.
During the hospital period, the mortality rate was 4.7% for rescue PTCA and 6.6% for primary PTCA (Table 4). Maximal aspartate aminotransferase was highest in patients who had rescue PTCA. Reinfarction and repeat intervention rates were comparable in both groups. Because of pretreatment with a thrombolytic agent, more blood transfusions were given in the rescue PTCA group (p < 0.05). However, hospital stay was comparable. Causes of death during the hospital stay are mentioned in Table 5. The main reason for death was cardiogenic shock. Age, infarct size, localization of myocardial infarction and failure of PTCA were predictors of long-term mortality (Table 6). In this cohort of patients, the few patients with acute main stem occlusions had a poor outcome (five of six patients died).
After discharge, one-year mortality was slightly higher in the primary PTCA group than in the rescue PTCA group (3.7% vs. 2.7%) (p = 0.63) (Fig. 1). However, during the complete one-year follow-up, no statistical differences were found between the two groups (p = 0.24). Somewhat more patients who had primary PTCA had heart failure (p = 0.39) (Table 7). In 77 patients, a stent was implanted. As compared with the patients who had no stents, no differences during the one-year follow-up were observed for reinfarction, presence of angina, repeat PTCA, coronary artery bypass graft surgery or mortality. These findings were consistent for both primary PTCA and rescue PTCA.
Angioplasty failed in 69 patients. Their clinical outcome was clearly worse compared with patients having successful angioplasty and independent of treatment allocation (Table 8).
Total interventional costs of rescue PTCA during the first year after AMI were $869 (11.7%) more expensive than those of primary PTCA (p = 0.0001) (Table 9).
Mortality and patency
Is rescue PTCA as safe and effective as the accepted form of therapy: primary PTCA? This series of 759 patients treated with PTCA in the acute phase of AMI is the largest (single-center) cohort published so far. The results of this analysis indicate that the aggressive approach of our institution to reopen the culprit coronary artery promptly results in comparable patency rates (power 71.9%) and an acceptable overall in-hospital mortality rate (5.8%), as well as one-year mortality rate (9.1%). At least in this analysis, the outcome of rescue PTCA is fairly similar to that of primary PTCA. It is likely that the short delay between the start of thrombolytic therapy and rescue PTCA (±90 min) is the most important factor explaining this. The outcomes of the few and small-sized rescue PTCA studies that have been previously reported were usually worse. In the group presented here, the rescue PTCA group represented those patients in whom thrombolytic therapy failed. The mortality of the patients who had primary PTCA is clearly higher than that in the initial randomized trials of Grines et al. (2), Zijlstra et al. (3) and Gibbons et al. (4). In contrast, in the larger sized, “more real world” Global Use of Strategies To Open occluded arteries (GUSTO) IIb study (24), patients who had primary PTCA had an in-hospital mortality rate of 5.7%, more or less identical to that in our group of patients.
In the present analysis, no significant in-hospital and one-year mortality differences were observed between rescue PTCA and primary PTCA. However, one has to realize that a comparison between these two groups has limitations, as will be discussed in the last paragraph. In our previous study, we reported on the poor outcome of patients with failed angioplasty (25). This finding is well known for failed rescue PTCA. Table 8 shows that failed primary PTCA carries comparable risks. More patients in the latter group had bypass surgery, which, in contrast to rescue PTCA, could be done early because of the normal fibrinolytic state. Patients with successful thrombolysis were not included in the angiographic and clinical part of the analysis. The outcome of patients with successful thrombolysis is usually excellent (9–12), indicating that the clinical outcome of both groups (patients with successful thrombolysis and patients with rescue PTCA) must be favorable. None of the patients had IIb/IIIa receptor blocker treatment. The outcome of the rescue PTCA group, as well as of the primary PTCA group, is likely to improve with this type of therapy (26–28).
In the randomized LIMI study, a first attempt was made to compare three treatment strategies in patients admitted to non-PTCA hospitals (11): 1) thrombolytic therapy and local routine treatment; 2) thrombolytic therapy, followed by transfer to the PTCA center where rescue PTCA was performed in case of impaired coronary flow; and 3) transfer for primary PTCA. In that study, no important differences were detected in the angiographic success rates between patients randomized to thrombolytic therapy and rescue PTCA, when indicated, and patients randomized to primary PTCA (90% vs. 97%). The total number of patients in that study (n = 224) was too low to draw firm conclusions on the efficacy of the three strategies. The results of the Prague study were presented during the European Congress of Cardiology in Barcelona in 1999 (29). That study had an identical setup to that of the LIMI trial. Their findings were that mortality rate in the thrombolysis-only group was 15%, as compared with 12% in the rescue PTCA group and 7% in the primary PTCA group. It is unclear, so far, why such high mortality rates were found in the first two groups. Recently, the data of the Plasminogen activator Angioplasty Compatibility Trial (PACT) were presented (30). In that trial, patients were treated with a bolus of 50 mg alteplase or with placebo, followed by immediate angiography. In patients with TIMI flow grade 3, an additional bolus of 50 mg alteplase was given, and these patients were treated medically. In patients with impaired flow, immediate PTCA was performed. In that study, the angiographic and clinical outcomes in patients with primary and rescue PTCA were similar: the angiographic success rates were 93% and 95%, whereas the mortality rates were 3.4% and 3.0%, respectively.
Rescue PTCA and primary PTCA have been used routinely in our hospital for >10 years. Previously, we published the outcome of the initial 176 patients who underwent early angioplasty (25). This retrospective analysis already suggested that the angiographic and clinical outcomes of primary and rescue PTCA in our hospital were comparable. The present study confirmed this in a much larger patient population. In a study of 133 patients, McKendall et al. (31) found no significant advantage of rescue PTCA as compared with thrombolytic treatment only. They suggested that the moderate success rate (82%) of the rescue PTCA procedure and the reocclusion rates (6%) of their study were responsible for the rather high mortality (12%). In that study, the delay between onset of pain and intervention was short (180 ± 54 min). In contrast, in the present study, procedural success of the rescue PTCA procedure is good (90.9%), and the reinfarction rate is acceptable (5.1%). This might partially explain the differences between McKendall’s and our findings.
Delays and need for angiography
When comparing the delays (pain to start of PTCA) of patients admitted to our hospital and patients sent from other hospitals, we found that the median difference was 75 min in the rescue PTCA group; in the primary PTCA group, this difference was only 25 min. Frequently, the decision to transfer patients requires much more time (Table 2). The main reason is that, not uncommonly, the result of thrombolytic therapy is awaited. In case of no reperfusion, the patient will be transferred. In patients who are candidates for primary or rescue PTCA, the shortest possible delay between chest pain and starting the procedure should be achieved. The delay can only be short if logistics are optimal: the catheterization team has to be available on a 24-h basis. In our setting, outside office hours, the team is immediately called when a patient with an extensive AMI is admitted to the Maastricht Hospital. This approach differs importantly from that of many other centers, where the decision to catheterize after failed thrombolytic therapy is frequently postponed and only done if the condition of the patient deteriorates further. This probably explains why, in such a setting, the outcome is usually poor.
Noninvasive reperfusion criteria can be used with reasonable accuracy before starting the invasive procedure to determine whether angiography is still needed (18,19). In a subgroup analysis, Oude Ophuis et al. (12) found that 66 patients (40%) who were treated with thrombolytic therapy and referred from another hospital showed these noninvasive reperfusion signs. None of these 66 patients had early catheterization. Nearly all patients immediately returned to the referring hospital and none of them died in the hospital, indicating that such an approach is safe. In case of no or questionable reperfusion (based on the noninvasive criteria), we performed angiography immediately, followed by rescue PTCA in case of inadequate flow.
In this analysis, only the unit costs of the interventions are taken into account. The reason is that no significant differences were observed in admission duration and medication. Our analysis shows that primary PTCA is less expensive than rescue PTCA. However, one has to keep in mind that the patients in the rescue PTCA group were only selected because of failed thrombolytic therapy. If the total thrombolysis group were analyzed, interventional costs would be lower. First, in our hospital, patients with clear ECG signs of reperfusion usually will not undergo early angiography. Second, in the remaining patients in whom early acute angiography is performed, one-third of the them will have TIMI flow grade 3; therefore, no early intervention will be performed (18). The price to pay for this relatively conservative approach is that at a later stage, more subacute interventions, as compared with primary PTCA, must be performed because of reischemia. However, the advantage is that these procedures usually can be scheduled.
This analysis is nonrandomized and retrospective in nature, and it is a single-center experience. Patients who have primary versus rescue PTCA are not completely similar. Primary PTCA usually is preferred over thrombolytic agents in case ECG findings suggest extensive ischemia or a poor hemodynamic state. Such an approach was also used in our patients. For example, this resulted in an overrepresentation of main stem occlusions in the primary PTCA group (five of six patients). All five patients died in the hospital. In addition, baseline data showed that more patients had LCx occlusion in the primary PTCA group (p < 0.05). This is most likely explained by the absence of characteristic ECG changes of an occlusion in that location. Patients with true posterior AMI not uncommonly show only precordial ST segment depression with no or minimal ST segment elevation. These ECG changes and the persistence of pain are usually the reason to perform angiography without thrombolytic pretreatment to differentiate between the diagnosis of unstable angina and AMI. When during angiography it becomes apparent that the LCx is occluded, primary PTCA will be done. This group of patients with LCx lesions usually has a reasonably good prognosis.
The need for rescue PTCA after failed thrombolysis is more obvious in patients with an extensive AMI than in those with a small AMI. In addition, of all patients who had thrombolysis, the patients who have rescue PTCA represent the worst group because of pharmacologic treatment failure. Enzymatic infarct size was higher in the rescue PTCA group than in the primary PTCA group. Either the area at risk was larger in the rescue PTCA group or the treatment with thrombolytic therapy and rescue PTCA resulted in more muscle loss. Also, the longer delay between onset of pain and PTCA in the rescue group will have influenced the amount of muscle loss.
The calculation of unit costs was based on an estimation of the percentage of interventions and prices. Other costs for medication and admission duration were not taken into account, because these were comparable. Indirect costs have not been calculated.
Although this retrospective study has several limitations, the results suggest that from a clinical perspective, rescue PTCA after failed thrombolysis is a good alternative to the accepted form of therapy: primary PTCA. This procedure is safe and effective, but is somewhat more expensive. The costs found may differ in a prospective study and will also differ locally. (30)
We thank Mrs. Margriet Dirkx for typing the manuscript and the personnel of the catheterization laboratory and the Coronary Care Unit for their active participation.
- acute myocardial infarction
- Global Use of Strategies To Open occluded arteries trial
- left circumflex coronary artery
- LImburg Myocardial Infarction trial
- Plasminogen activator Angioplasty Compatibility Trial
- percutaneous transluminal coronary angioplasty
- Thrombolysis In Myocardial Infarction trial
- Received March 23, 1999.
- Revision received February 16, 2000.
- Accepted March 27, 2000.
- American College of Cardiology
- Vermeer F.,
- Simoons M.L.,
- Bär F.W.,
- et al.
- Primary Angioplasty in Myocardial Infarction Study Group,
- Grines C.L.,
- Browne K.F.,
- Marco J.,
- et al.
- Mayo Coronary Unit and Catheterization Laboratory Groups,
- Gibbons R.J.,
- Holmes D.R.,
- Reeder G.S.,
- Bailey K.R.,
- Hopfenspirger M.R.,
- Gersh B.J.
- Califf R.M.,
- Topol E.J.,
- Stack R.S.,
- Lee T.H.
- RESCUE Investigators,
- Ellis S.G.,
- Ribeiro-de Silva E.,
- Heyndrickx G.,
- Talley D.,
- Cernigliari C.,
- Topol E.J.
- Widdershoven J.W.M.G.,
- Gorgels A.P.M.,
- Vermeer F.,
- et al.
- Doevendans P.A.,
- Cheriex E.,
- van der Zee R.,
- et al.
- ↵Vermeer F, Oude Ophuis A, Van den Berg E, et al. Prospective randomized comparison between thrombolysis rescue PTCA and primary PTCA in patients with extensive myocardial infarction admitted to a hospital without PTCA facilities: the LIMI study, a safety and feasibility study. Heart 1999;82:436–31.
- InTIME Investigators,
- Den Heijer P.,
- Vermeer F.,
- Ambrosioni E.
- ↵Oude Ophuis AJM, Bär FW, Janssen WJ, et al. The prospective use of non-invasive reperfusion parameters in patients with acute myocardial infarction can limit acute coronary angiography. Heart 2000. In press.
- Verstraete M.,
- Bernard R.,
- Bory M.,
- et al.
- Zijlstra F.,
- De Boer M.J.,
- Beukema W.P.,
- et al.
- Meinert C.L.
- Lefkovits J.,
- Ivanhoe R.J.,
- Califf R.M.,
- et al.
- RESTORE Investigators
- Groch L.,
- Widemski P.,
- Aschermann M.,
- Zelisko M.,
- Bednar F.,
- Hlinomaz O.
- Ross A.M.,
- Coyne K.S.,
- Reiner J.S.,
- et al.
- Thrombolysis in Myocardial Infarction Investigators,
- McKendall G.R.,
- Forman S.,
- Sopko G.,
- Braunwald E.,
- Williams D.O.