doi: 10.1161/01.CIR.0000036844.40992.40
(Circulation. 2002;106:e9021.)
© 2002 American Heart Association, Inc.
Cardiovascular News
Circulation Newswriter
News From the 2002 Congress of the European Society of Cardiology: The Hotlines
BERLIN, Germany—In a city decorated with gaily painted bears and on the verge of the 10th anniversary of its reunification, the 2002 Congress of the European Society of Cardiology opened its sessions with the latest news on percutaneous coronary interventions (PCI) and their use in acute coronary syndromes. As Patrick Serruys, MD, of Erasmus University in Rotterdam, the Netherlands, told attendees at the first hotline session on September 1, "As an interventional cardiologist, I am pleased to see that the interventional approach is gaining ground in the field of acute coronary syndromes."
That message was repeated again and again during the first Hotlines dealing with acute coronary syndromes, but the message was altered with the study reported by Francisco Fernández-Avilés, MD, and his colleagues. They found no difference in mortality at 30 days in the GRupo de Analisis de la Cardiopatia Isquemica Aguda (GRACIA) that enrolled 500 subjects. In the trial, 50% of patients were randomized to classical drug-based treatment and 50% to an early interventional strategy that included cardiac catheterization within 24 hours of thrombolysis followed by stent when appropriate. Patients who could not benefit from the stent received coronary artery bypass grafting or intensive drug therapy.
|
In the group assigned to intervention, stent implantation was accomplished in 81% of cases, bypass in 3%, and drug treatment in the remaining 16%. The incidence of nonfatal cardiac events was 3.7% in the conventional arm and 0.8% in the interventional arm. The length of hospital stay was significantly shorter in the interventional arm as well—7.6 days on average versus 11.2 days, according to Dr Avilés.
Dr Avilés said he and his colleagues undertook the study because of their conviction that "thrombolysis is likely to remain a mainstay of reperfusion therapy." However, the frequent failure of that treatment to open the arteries or of opened arteries to reocclude means that some treatment is needed after the use of the antithrombus medication. The impact of his study will become clearer when the 1-year follow-up is complete, he said. Christian Hamm, MD, PhD, of the Kerckhoff Heart Center in Bad Nauheim, Germany, who was called upon to discuss the presentation, agreed with that assessment, saying that the true results will only be known next year, when follow-up is complete and all assessments are in.
"The main result of the study was that the hospital stay was shorter among patients in the interventional arm," he said. He said GRACIA shows that the strategy of combining thrombolysis and early catheterization is safer than previously thought. However, he said, further studies are needed to assess the efficacy of the study, reminding the audience that the difference was not in deaths but in nonfatal events.
Rushing patients to hospitals set up to perform percutaneous interventions appears to be safe for most patients, but the differences in mortality at 30 days between patients who received thrombolysis and standard medical treatment in community hospitals and those who were sent to PCI centers did not reach statistical significance, said Petr Widimsky, MD, of the Cardiocenter at the University Hospital Vinohrady in Prague, Czech Republic. He presented results on behalf of the group that organized the PRAGUE-2 (PRimary Angioplasty after transport of patients from General community hospitals to catheterization Units with/without Emergency thrombolytic infusion) study.
|
Overall mortality was 10% in the thrombolysis group versus 6.8% in the treatment group, Dr Widimsky said. Mortality was virtually the same (7.4% in the thrombolysis group versus 7.3% in the PCI group) for those who arrived at the hospital <3 hours after the onset of symptoms. However, he said, the differences in mortality among patients who arrived at the hospital >3 hours after the onset of symptoms was significant and prompted the study’s ethics board to halt the study prematurely. In that group of subjects, mortality was 15.3% in the thrombolysis group versus 8.4% in the PCI group.
Dr Widimsky said the study shows that interhospital transport is safe and feasible and improves the 30-day prognosis. "Primary PCI should be the preferred reperfusion therapy for acute myocardial infarction, especially in patients presenting after 3 hours. The best approach is if the emergency ambulance can take the patient directly to the PCI center," he said during a press conference preceding his presentation. "The new guidelines of the Czech Heart Society say this is the best approach."
However, A. Betriu Gibert, MD, of the Institute of Cardiovascular Diseases of the Corporaci Sanitaria Clinic in Barcelona, Spain, said that the most positive side of the study seemed to be the safety of transporting patients to centers for PCI.
"The primary end point was not achieved," he said. The most likely reason was that the study was underpowered to achieve that result, Dr Gibert said. He also questioned the significance of the subset analysis that showed such a dramatic difference in mortality among patients who arrived at the hospital late. "The patients were not randomized by time to treatment. It is interesting as a hypothesis-generating finding, but it probably should not be a general recommendation."
He said PRAGUE-2 adds to the summary of trials that favor the routine use of PCI in patients with acute myocardial infarction. Included among those is Danish Multicenter Randomized Study on Thrombolytic Therapy versus Acute Coronary Angioplasty in Acute Myocardial Infarction-2 (DANAMI-2) reported on September 2 in a recent clinical trial update session. Early information from that trial indicated that primary PCI, even when patients had to be transported as far as 150 km to a center, was more effective and better for patients than thrombolysis in community hospitals. Six months later, nothing has changed. "There was no subgroup in which PCI was not superior to lysis," said Henning Rud Andersen, MD, of Skejby University Hospital in Aarhus, Denmark. Maarten Simoons, MD, the outgoing President of the European Society, asked if the studies suggest not to use thrombolysis. "We are comparing two effective treatments. We shouldn’t be ashamed to use them, although PTCA [percutaneous transluminal coronary angioplasty] is better in specific settings."
In areas where individuals can be taken to PCI centers without long delays, that procedure might be preferred. However, a person should not wait 12 hours to get PCI when thrombolytic agents are available, said Dr Simoons. "Many patients don’t get any reperfusion therapy." He advised communities to streamline their systems for getting patients into care quickly and efficiently, developing a coordinated approach to reperfusion therapy.
In the Randomized Intervention Treatment of Angina (RITA-3) study in the United Kingdom, researchers from 45 study sites randomized 1810 patients with unstable angina or non-ST–segment elevation myocardial infarction to receive a conservative, ischemia-driven revascularization treatment or to go directly to intervention. Those in the intervention group received coronary angiography followed by revascularization by either PCI or bypass. The conservative group received intervention when their symptoms seemed to warrant it.
At 4 months, "There was no significant difference in death or myocardial infarction," said Kenneth A. Fox, MD, who presented the study. However, there was a major difference in the rates of angina. At that time, 9.6% of patients in the intervention group had died, had a heart attack, or had suffered from significant angina, he said. By comparison, 14.5% of patients in the conservative treatment group had reached one of those end points. However, the rate for death or myocardial infarction in the two groups was essentially the same: 7.2% for the intervention group and 8.2% for the conservative treatment group.
|
"In patients with moderate risk, we now have for the first trial to report a halving of refractory angina," said Dr Fox. "There was no impact on deaths. However, it increases the quality of life and decreases the incidence of subsequent angina and the need for therapy." The trial is significant because it was the first large study conducted in patients with moderate heart disease risk, he said. More definitive answers on the effect of the interventional therapy on death or myocardial infarction will be available when the trial has completed its 5-year follow-up.
Dr Serruys agreed that the main impact of the strategy was on angina. "I am always suspicious of combined end point data when only one end point is affected," he said as a discussant of the presentation.
The use of intravascular ultrasound imaging, which gives the cardiologist a picture of the lumen and walls of the coronary arteries, improves balloon angioplasty and reduces stent use, according to François Schiele, MD, of the Department of Cardiology, University Hospital, Besançon, France. Dr Schiele discussed the Balloon Equivalent to Stent (BEST) trial that compared intravascular ultrasound-guided angioplasty with stent. In the study, 250 patients were randomized—approximately half to receive stents and the others to receive angioplasty guided by the ultrasound procedure.
|
"The stent group had better immediate results," said Dr Schiele. However, he said, at 6 months, 16% of patients in the angioplasty plus ultrasound group and 20% of patients in the stent group had experienced a clinical event (death, infarction, unstable angina, or revascularization). The difference was not statistically significant. Recurrent narrowing of the artery was 18.1% in the stent group and 16.8% in the angioplasty group. Forty-five percent of the group that had the ultrasound-guided angioplasty was eventually given stents because physicians felt they were needed to insure optimal results for the patients.
Dr Schiele said his study proved the "noninferiority" of the new technique to the use of stents. However, the technique is more complex and takes longer than stent placement and cannot be used in some patients, he said.
An even newer technique tested in the EUROpean Sonotherapy Prevention of Arterial Hyperplasia (EURO-SPAH) produced equivocal results because the decrease in late stent loss that was sought was not achieved, according to Dr Serruys, who presented the study for the group. In the study, researchers at 23 centers in 9 European countries used a special catheter designed to emit 1 MHz of sound. Previous studies had shown that sound at that level has a profound (though transient) effect on the cytoskeleton and shape of myoblasts, said Dr Serruys.
In the study, 403 patients with multiple first-time stented lesions were randomized to receive either sonotherapy or sham. Even though the intended prevention of late stent loss was not achieved, the study confirms the feasibility and safety of sonotherapy. However, Dr Serruys said, there was a significant reduction in the need for revascularization and reduction of major adverse cardiac events.
However, Karl R. Karsch, MD, Chair of Cardiology at the University of Bristol in the United Kingdom, was brusque. "It was a negative trial, wasn’t it?" he said. He questioned the basic and animal science behind the study. "In my opinion, all of this has not been shown and analyzed thoroughly enough for clinical experiments," he said.
Dr Serruys also presented the ACTinomycin-D Eluting Stent Improves Outcomes by Reducing Neointimal Hyperplasia (ACTION) trial that was halted when the Data Safety Monitoring Board decided the study would not meet its efficacy end points, a decision that he said was confirmed by the angiographic outcomes of the 360 patients enrolled from Europe, Australia, New Zealand, and Brazil.
A routine internal review showed a high target lesion revascularization rate, he said. At that time, accelerated patient follow-up began. They found that patients who received the bare stent were doing better than those who received the drug-eluting stent that used actinomycin, an anticancer drug.
Willibald Maier, MD, of Switzerland said the study "shows that drug elutions are not a guaranteed road to clinical success with stents," although he was puzzled by the lack of effect of actinomycin.
PRAGUE 4, presented by Dr Widimsky on September 2 in a hotline session, put a positive mark in the column in favor of off-pump coronary artery bypass. He concluded that the procedure is applicable in 84% of patients to whom no special selection criteria had been applied, reduces costs, and is at least as clinically effective and safe as coronary artery bypass surgery that is done on the heart-lung bypass pump.
To prove his point, Dr Widimsky and his colleagues randomized 400 patients to on- and off-pump bypass procedures. Eventually, 204 patients underwent the on-pump procedure and 184 the off-pump surgery. There was significant crossover in the study in the off-pump group going to the on-pump surgery and 12 in the on-pump going to off-pump surgery for a variety of reasons.
"There were a significant number of differences," said Dr Widimsky. The off-pump group had less blood loss and shorter intubation time. However, there was no significant difference in the length of hospital stay. There was no significant difference in major adverse cardiac events including death, myocardial infarction, stroke, or hemodialysis within 30 days, said Dr Widimsky.
Peter P.T. de Jaegere, MD, of the University Medical Center Utrecht, Utrecht, the Netherlands, who was asked to discuss the presentation said, "The primary objective was to test the feasibility and safety for bypass patients, and you have been successful."
However, he questioned whether patient selection was not subject to some exclusion because the demographics of the groups shown were significantly different from those usually seen in Europe. "The high crossover rate may be related to selection," Dr de Jaegere said. The event rate was low, he said, at 3.8% in the on-pump group compared with 2.9% in the off-pump group. "That brings up a point of the power calculations of all these studies (of off-pump surgery). They all may be too low to detect a difference."
Dr de Jaegere said the move to off-pump surgery has the laudable goals of enhancing safety and efficacy as well as improving cost-effectiveness and decreasing patient discomfort. "Should we now offer it to all patients or define those who would most benefit?" he asked.
He warned that liberal use of off-pump surgery has potential for affecting the quality of surgical technique, making it more difficult to accomplish careful anastomosis of delicate arteries.
In the Optimal Trial in Myocardial Infarction with the Angiotensin II Antagonist Losartan (OPTIMAAL), angiotensin II inhibitors did no better than ACE inhibitors in preventing mortality and other serious cardiac events in patients with acute myocardial infarction and heart failure, said Keith Dickstein, MD, of the University of Bergen in Norway.
|
The study enrolled 5477 patients who were randomized to either 50 mg of losartan or 50 mg of captopril 3 times a day. The patients were monitored for an average of 2.7 years. There were a total of 946 deaths—499 in the losartan group and 447 in the captopril group. There was also no difference for stroke or reinfarction, said Dr Dickstein.
In their Lancet article, Dr Dickstein and his colleagues concluded: "Since we saw a nonsignificant difference in total mortality in favor of captopril, ACE inhibitors should remain first-choice treatment in patients after complicated acute myocardial infarction. Losartan cannot be generally recommended in this population."
One major problem is that the study used a very low dose of losartan, he said. At that time, it was not known how high a dose could be used in this population. Helmut Drexler, MD, of the Medical School in Hanover, Germany, agreed. "If you look at the dose, 50 mg of losartan is just too low."
In the Anticoagulation for Cardioversion using Enoxaparin (ACE) trial, investigators found that low-molecular-weight heparin is at least as effective and safe as unfractionated heparin and oral phenprocoumon in the conventional manner to treat patients who undergo cardioversion for atrial fibrillation, according to Christoph Stellbrink, MD, of the University Hospital in Aachen, Germany.
|
Dr Stellbrink said conventional anticoagulation is often underused for this purpose because of the fear of bleeding and the need for frequent monitoring. Low-molecular-weight heparin can be used in ambulatory settings and does not require frequent monitoring. The low-molecular-weight heparin is given subcutaneously.
There were 212 patients in the heparin-phenprocoumon arm of the study and 216 in the low-molecular-weight heparin arm. The study actually had 2 strata, he said. A small number underwent cardioversion without transesophageal echocardiography, although in most, the cardioversion was performed under guidance of the transesophageal cardioversion.
Twelve of the patients in the heparin group and 7 in the low-molecular-weight heparin group suffered one of the combined primary end points of the study—stroke or peripheral emboli, major bleeding, or death. Thus, the low-molecular-weight heparin met the test for "noninferiority," according to Dr Stellbrink.
In the Trans European Network (TEN-HMS) initiative, investigators attempted to measure the effect of high tech to provide low-tech comfort and monitoring to patients with severe heart failure. A total of 428 patients from the Netherlands, the United Kingdom, and Germany were randomized to regular care, close monitoring by a nurse on the telephone, or telemonitoring. The telemonitoring consisted of a scale, automated blood pressure machine, and an automated ECG that transmitted their findings via the telephone. Patients took their measurements in the morning and at teatime, said Dr John G.F. Cleland of the University of Hull, Castle Hill Hospital in Kingston-upon-Hull in the United Kingdom.
|
The major comparison was between the nurse care group and those who were telemonitored, he said. At 400-day follow-up, he said, those in usual care lived on average 295 days, those in the nurse group lived 340 days, and those in the telemonitoring group, 347 days.
Hospital admissions were increased in the telemonitoring group, but that might mean that problems were likely to be picked up more quickly, Dr Cleland said. Overall, he said, there was no significant difference between the telemonitoring and the nurse telephone group.
The nurse plan is good because the nurse can give advice over the telephone. However, when the patient is admitted to the hospitals, doctors are more confident sending them home sooner if they can be telemonitored. "This is the first time this technology has been used in a large, randomized controlled trial. Everyone is learning from this system," he said.
The Carvedilol ACE inhibitor Remodeling Mild CHF Evaluation (CARMEN) tested the combination of a ß-blocker (carvedilol) and an ACE inhibitor (enalapril) head-to-head and in combination for their abilities to prevent progression of mild heart failure. The combination of the two drugs was the most potent, said Willem J. Remme, MD, of the STICARES Cardiovascular Research Foundation in the Netherlands, leading to highly significant improvement of left ventricular remodeling and function. "It made the heart smaller and reversed the remodeling process," said Dr Remme.
|
The trial, conducted at 65 sites in 13 European countries, enrolled 572 patients with what Dr Remme called "mild heart failure." During the 18-month, double-blind, randomized trial, patients were randomized into one of 3 arms: carvedilol with enalapril, enalapril and placebo, or carvedilol and placebo. The study medication was given in addition to diuretics, digitalis, and nitrates patients were already taking for their disease.
Carvedilol alone also had a significant effect on remodeling and left ventricular function, but it was less than that of the combination of drugs, Dr Remme said. "The CARMEN results confirm the current heart failure treatment guidelines and provide an additional mandate for not delaying the prescription of an ACE inhibitor and carvedilol in mild heart failure patients," he said. He pointed out in this case that the ß-blocker was started before the ACE inhibitor, which "challenges the historical practice of initiating treatment with the ACE inhibitor alone." The carvedilol effect was not a class of drugs phenomenon, he said. "The results of CARMEN apply to carvedilol only."
The trial showed no difference in mortality among the groups, but Dr Remme said he was not looking for clinical end points such as that. Philip A. Poole-Wilson, MD, of the United Kingdom, said this is the first study "of reasonable size" to compare a ß-blocker against an ACE inhibitor. "We already know that a ß-blocker plus an ACE inhibitor has advantages," he said.
He questioned, however, Dr Remme’s statement that enalapril had no effect on ventricular remodeling. Before the study, 64% of patients had taken ACE inhibitors at some time. "Remodeling from that might have been still present and may reduce the validity of the comparison," said Dr Poole-Wilson. In addition, he said, the trial uses surrogates for clinical end points and is open to question for that reason.
Reaching for a macrolide antibiotic after a heart attack is unlikely to do any good, said Ralf Zahn, MD, of the Heart Center in Ludwigshafen, Germany, as he presented the results of the Antibiotic Therapy after an Acute Myocardial Infarction (ANTIBIO) study during Hotline presentations on September 3. He and his colleagues in the study were testing the hypothesis that bacterial infections, particularly Chlamydia pneumoniae, might play a role in the inflammation that is associated with the pathogenesis of arteriosclerosis. If that were so, it would seem logical to treat the patient with the appropriate antibiotic, he said.
|
In this prospective, randomized, double-blind, placebo-controlled trial, 872 patients with acute myocardial infarction were randomized to 6 weeks of either roxithromycin or placebo. They came from 68 participating centers in Germany. The study, which had planned to include nearly 4000 subjects, was forced to quit at 872 because of slow enrollment. There were 433 patients in the placebo group and 439 in the antibiotic group.
Analysis at 12 months showed no difference in mortality (6.5% in the antibiotic group and 6% in placebo). Again there was no significant difference in the combined end points of death, myocardial infarction, resuscitation, stroke, postinfarction angina, or in the percentage of patients needing percutaneous intervention or coronary artery bypass grafting.
"Routine treatment with roxithromycin in acute myocardial infarction patients cannot be advised," said Dr Zahn. He said it is now time for researchers to consider whether infectious agents might not contribute as much to the inflammatory process as had been thought before.
Sandeep Gupta, MD, from St Bartholomew’s Hospital in London said that although he was not ready to give up on the possibility of an infectious agent in acute myocardial infarction altogether, he agreed that the time is not right for the routine use of antibiotics in treating patients who suffer this event.
Nine men and 1 woman have been recruited to a Polish study involving the transplantation of skeletal myoblasts into their hearts that have been damaged by heart attack to the point that they are suffering postinfarction heart failure. Tomasz Siminiak, MD, of the University School of Medical Sciences in Poznan, Poland, said the cells were taken in a biopsy from the vastus lateralis and expanded to 20 million cells in culture. The myoblasts were introduced into the postinfarction scar tissue during coronary artery bypass grafting. One patient died 7 days postoperatively. At autopsy, researchers found a new myocardial infarction in a part of the heart that had been previously normal, he said.
|
Although the trial was designed to prove safety and feasibility, Dr Siminiak said it appears that ejection fraction has increased in every patient in the 5 months of the trial and the contractility of segments of infarcted tissue was positively affected.
One concern is whether the new cells will couple with native cardiac myocytes and the lack of gap junctions, he said. Worries about arrhythmias in such patients are valid. Episodes of serious tachycardia arose in the first 2 patients. Now all patients receive infusions of amiodarone after the procedure and 2 patients continued on oral drug for 2 months. At 5 months, no patient was taking amiodarone, he said.
Philippe Menasche, MD, of Groupe Hospitalier Bichat-Claude Bernard in Paris, is conducting a similar study and congratulated Dr Siminiak on his trial. He said he was concerned about the fact that some segments of heart muscle receiving skeletal myoblasts were not connected and that others were dyskinetic, making it harder for the cell therapy to work. He also recommended the use of >20 million cells. His group is planning a trial in which the use of 400 million cells is tested against 800 million. He also advised consideration of long-term care of the patients, noting that some experts have recommended implantation of an implantable cardiac defibrillator in all of them.
Device implantation was at the heart of the next study in the Hotline presentations, the Eastbourne Syncope Assessment (EaSyAS) study. Neil Sulke, MD, of Eastbourne General Hospital in the United Kingdom, said all suitable patients (201) with syncope at the National Health Service facility were randomized to receive either an implantable loop recorder called the Reveal Plus or conventional investigation. With the loop recorder, patients could either trigger the device themselves when they experienced an episode of syncope or the machine could be set to record it automatically.
|
There was no difference in mortality between the groups, but use of the loop recorder decreased the time to electrocardiogram diagnosis for patients in that arm of the trial. It also dramatically reduced the time to ECG-guided therapy, he said. However, the record did not note the first episode of syncope in nearly half the patients and in 12% of those who experienced a second episode. Twenty-one percent of patients were diagnosed by the device’s automatic feature, he said.
Georg Noll, MD, of the University Hospital in Zurich, Switzerland, was concerned about the failure of the loop recorder to diagnose the first and second episodes of syncope in significant numbers of patients. He also questioned the need to go to such lengths to diagnose such patients when there are few viable treatments for many patients with a disorder that can have a variety of etiologies. "Implantable loop recorder may be a promising tool, but I think further studies are needed."
Endothelial-A receptor antagonists failed to move the Endothelial A Receptor Antagonist Trial in Heart Failure (EARTH) into the plus column, Thomas F. Lüscher, MD, of the University Hospital in Zurich, told Hotline session attendees. In this study, 650 patients were randomized to test the endothelial-A receptor antagonist darunsentan, made by Abbott Pharmaceuticals, which sponsored the study.
|
Patients were randomized to receive 10, 25, 50, 100, or 300 mg darunsentan each day or placebo. The study was designed to determine if the drug would affect left ventricle remodeling as a surrogate end point for improving the prognosis in heart failure. The measurements were made by magnetic resonance imaging and measured in a single center.
The trial found no dose-dependent effect on left ventricular remodeling, according to Inder S. Anand, MD, of the University of Minnesota School of Medicine in Minneapolis. There was no change in hypertension or heart rate. No excess mortality was found in any of the groups, although patients at higher drug doses were more likely to drop out of the trial.
Dr Lüscher said he wonders whether cardiologists have blocked as many hormones as possible in heart failure. "To move further, we need to treat the heart rather than block more neuroendothelial systems," he said during a press conference at the meeting.
Study Finds the Progress of Mitral Valve Prolapse Widely Variant
The future for patients diagnosed with mitral valve prolapse is widely variant, with prognoses ranging from normal life expectancy and few problems in half of patients to groups of patients who have health problems or even die from causes directly related to their heart condition and its major complication, mitral regurgitation. The study led by Jean-François Avierinos, MD, of the Mayo Clinic in Rochester, Minnesota, appears in this week’s issue of Circulation (Circulation. 2002;106:1355–1361).
The researchers monitored 833 patients diagnosed with asymptomatic mitral valve prolapse. Patients were monitored for an average of 5.4 years with end points of mortality and cardiovascular morbidity as well as mitral valve prolapse–related events. The most frequent risk factors associated with cardiovascular mortality were mitral regurgitation and ejection fraction of <50%.
In an accompanying editorial (Circulation. 2002;106:1305–1307), Martin St John Sutton, MD, of the University of Pennsylvania Medical Center in Philadelphia and Arthur E. Weyman, MD, of Massachusetts General Hospital in Boston, noted that an earlier Framingham study of the disorder had called it a "benign" condition.
They note that although the authors call their study one of "community," it is not randomly selected and does not include the whole population of Olmsted County, where the clinic resides. It "describes patients with MVP [mitral valve prolapse] identified either clinically or echocardiographically" and thus identifies a hospital or referral population that is subject to some kind of selection bias. They note that the authors reported 96 deaths during 4581 years of patient follow-up. Forty-one of these are cardiovascular and of these, 31 were related to mitral valve prolapse.
They wrote: "In the report of the Framingham study, the choice of the term benign was unfortunate, particularly when close to 10% of the affected patients have had or can be expected to require valve repair or replacement. The report of Avierinos et al supports these concepts, but because of the method of patient selection, the data are not directly applicable to the general population."
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||