Author + information
- Received October 15, 2001
- Revision received February 21, 2002
- Accepted February 22, 2002
- Published online May 15, 2002.
- Lars Aaberge, MD*,* (, )
- Kjell Rootwelt, MD, PhD†,
- Svein Blomhoff, MD, PhD‡,
- Kjell Saatvedt, MD, PhD§,
- Michel Abdelnoor, MPH, PhD∥ and
- Kolbjorn Forfang, MD, PhD*
- ↵*Reprint requests and correspondence:
Dr. Lars Aaberge, Feiringklinikken, N-2093 Feiring, Norway.
Objectives The goals of this study were to assess late clinical outcome and left ventricular ejection fraction (LVEF) after transmyocardial revascularization with CO2laser (TMR).
Background During the 1990s TMR emerged as a treatment option for patients with refractory angina not eligible for conventional revascularization. Few reports exist on clinical effects and LVEF >3 years after TMR.
Methods One hundred patients with refractory angina not eligible for conventional revascularization were block-randomized 1:1 to receive continued medical treatment or medical treatment combined with TMR. The patients were evaluated at baseline and after 3, 12 and 43 (range: 32 to 60) months with end points to angina, hospitalizations due to acute myocardial infarctions or unstable angina, heart failure and LVEF. Mortality was registered and MOS 36 Short-Form Health Survey answered at baseline and after 3, 6 and 12 months.
Results Forty-three months after TMR, angina symptoms were still significantly improved, and unstable angina hospitalizations reduced by 55% (p < 0.001). Heart failure treatment (p < 0.01) increased, whereas the number of acute myocardial infarctions, LVEF and mortality was not affected. Quality of life was improved 3, 6 and 12 months after TMR.
Conclusions Forty-three months after TMR, angina symptoms and hospitalizations due to unstable angina were significantly reduced, heart failure treatment increased and LVEF and mortality were seemingly unaffected.
During the 1990s transmyocardial revascularization with CO2laser (TMR) emerged as treatment in patients with severe angina pectoris not amenable for conventional revascularization (1–3). Transmyocardial revascularization with CO2laser is performed by making channels directly through the left ventricular wall with laser energy release (4)and a usual approach through a left anterior thoracotomi. From randomized clinical trials, symptomatic relief has been reported until one year after TMR (5–10). The mechanism is not understood, but angiogenesis, denervation and placebo have been suggested as possible explanations. Clinical and mortality data from randomized clinical studies more than one year after TMR have not been published. Early reduction of global left ventricular function has been reported (11–15), but the effect on left ventricular function in the long run is unknown. We evaluated angina symptoms, hospitalizations and left ventricular ejection fraction (LVEF) three to five years after TMR, quality of life during (QoL) the first year and registered mortality.
Design and study population
The study is a late follow-up of the Norwegian Randomized trial with CO2laser (9)in which 100 eligible patients were block-randomized 1:1 between November 1995 and January 1998 to achieve either continued optimal medical treatment alone (MT group) or in combination with TMR (TMR group) in a noncrossover design. Included were patients suffering from angina pectoris New York Heart Association (NYHA) functional class III or IV despite optimal medical treatment, who were not candidates for percutaneous transluminal coronary angioplasty or coronary artery bypass grafting because of diffuse and peripheral coronary artery disease and who were willing to participate in a randomized study. Exclusion criteria were age >75 years; LVEF <30%; nondemonstrated reversible ischemia as assessed by exercise electrocardiogram, dobutamine stress echocardiography (DSE) or rest/stress technetium 99m-tetrofosmin myocardial perfusion scan (SPECT); overt heart failure and inability to undergo study tests and conditions precluding thoracic surgery. Clinical status and assessments of LVEF were, in the initial study, performed before randomization and after 3 and 12 months. Transmyocardial revascularization with CO2laser has not been offered as treatment against refractory angina in Norway except in clinical trials. After the initial 12 months follow-up, the patients were controlled by their general practitioners or local hospitals. A late follow-up was performed after 43 (range: 32 to 60) months. At baseline, 71% of the patients were classified as angina NYHA functional class III and 29% as NYHA functional class IV patients. Baseline patient characteristics (Table 1) and medications did not differ between the treatment groups. During surgery, one patient was found to have a graftable left anterior descending coronary artery that had not been detected by coronary angiography and was excluded from follow-up analysis. Data were collected from November 2000 until September 2001. One patient in the MT group was not available for assessment. Otherwise, complete clinical data were obtained from all. All patients at baseline and 84% of patients alive at late follow-up had a multiple-gated acquisition (MUGA) performed. No crossover occurred between the treatment groups. The scientific protocol was approved by the regional ethics committee.
In a combined general and thoracic epidural anesthesia and without the use of cardiopulmonary bypass (9)or intra-aortic balloon pump prophylactically, we treated the patients by use of an 800-W CO2laser (The Heart Laser, PLC Medical Systems, Inc., Milford, Massachusetts). The channels were made directly through the left ventricular wall on the beating heart through a left anterior thoracotomy with pulses synchronized with R-wave in the electrocardiogram. The vascular territories to be treated were defined on basis of DSE and SPECT results, and about one channel per cm2of presumed ischemic and viable myocardium was made. An average of 48 (standard deviation ± 7) channels (energy range: 30 J to 50 J) were made. Channel formation was confirmed through a 5-MHz multiplane ultrasound transesophageal probe (VingMed, Horten, Norway).
Clinical information was retrieved with respect to severity of angina symptoms, hospitalizations due to unstable angina, myocardial infarctions, heart failure or other reasons, repeated coronary angiography or coronary interventions, current medications and smoking habits. Clinical assessments were done using standardized telephone interviews. The information was checked against hospital records and the patients’ general practitioners, as was information about causes of death. “Probably cardiac related causes” of death included patients who died from clinically diagnosed heart failure or acute myocardial infarction or who died suddenly. Only the patients who died early postoperatively had an autopsy performed. Before the interviews, data from the National Health Security system were used to update patient information with respect to mortality and time of death. Angina was classified according to the NYHA classification system and was assessed unblinded by one investigator after standardized interviews about symptoms related to graded activities at baseline and follow-up interviews. A change of two or more angina functional classes was considered clinically significant (6).
Investigations were performed consecutively following randomization numbers at Section of Nuclear Medicine, Rikshospitalet University Hospital, and the operators were blinded to clinical patient information. The LVEF reference interval at rest has been established to be 50% to 78% and the standard deviation for same-day repeated measurements 2.1% points (12).
MOS 36 Short-Form Health Survey (SF-36) is a generic questionnaire consisting of 36 questions about the subscales physical functioning, role-physical (role limitations because of physical health problems), body pain, general health, vitality, social functioning, role-emotional (role limitations because of emotional problems) and mental health. A summary physical component scale (PCS) and summary mental component scale (MCS) is computed for the individual patient at each time point. Scoring and interpretation is performed as described by Ware et al. (16). The results are presented in a 100 point scale where 100 is the best and 0 the worst possible score. Summary physical component scale and MCS changes of three or more points are considered clinically relevant (17). The questionnaire was answered by the patients before randomization and after 3, 6 and 12 months. To assess the robustness of the angina scoring, we compared the mean scores between patients assessed to be in different angina functional classes. The Norwegian version of the SF-36 has been validated, and normative data from the Norwegian population have been presented (18).
Results are given as mean ± SD, median (range) or frequency (%). The relative changes of the outcome variables between baseline (OBL) and follow-ups (OFU) were estimated for each patient: RC = [OBL− OFU/OBL] × 100. The evaluation of efficacy of TMR was done by comparing the treatment groups with respect to relative changes from baseline to follow-up for each outcome variable. Normal distribution was evaluated by visual evaluation of the histograms of the end point variables and by performing the one sample Kolmogorow-Smirnov goodness-of-fit test. The data on SF-36 showed approximation of the distribution to the Gaussian curve, and parametric tests were used to compare baseline values and relative changes between the two groups and the mean of scores between the individual angina functional classes (ttest). The data on LVEF were normally distributed but showed heterogeneity of the variance in the two groups, and comparisons were done using the Mann-Whitney Unonparametric test. The chi-square test was used to test significance level for frequency differences. Significance level was defined as p < 0.05 (19).
Clinical outcome and mortality
Median follow-up was 43 (range: 32 to 60) months. Total mortality was 23% (included 4% postoperative mortality) and was not different between the groups (Table 2). Six patients in each treatment group died from “probably cardiac related causes.” A significant improvement in angina symptoms was still present three to five years after TMR (Table 2), and fewer hospitalizations due to unstable angina were registered. Antianginal medication was, however, only insignificantly reduced (Table 3) compared with the MT group. A higher number of hospitalizations due to heart failure (p = NS) and an increased use of diuretics and angiotensin-converting enzyme inhibitors was observed (p = 0.002) in the TMR group (Table 3) compared with the MT group.
Due to health problems, 12 patients were not able to meet for investigation. The analysis was based on 33 patients in the TMR group and 30 patients in the MT group who had complete MUGA investigations at baseline and follow-ups. After TMR, a reduction in LVEF was observed after 3 months (p < 0.05) compared with baseline, whereas the differences were not significant after 12 months and at late follow-up, and the changes were not significantly different from the changes in the MT group (Fig. 1) at any time.
The physical health subscales and PCS of SF-36 were significantly improved after TMR (Fig. 2), whereas the mental and social subscales and MCS did not change significantly. The mean summary PCS scores were clinically and statistically different between patients classified into angina functional class IV, class III and class II at each time point, whereas the difference between functional class I and class II was not significant (Fig. 3). The mean summary MCS were not different between the patients in the respective angina functional classes.
A reduction in angina symptoms could still be observed three to five years after TMR. The lower number of hospitalizations due to unstable angina comply with a continued antianginal effect from TMR. This was, however, not followed by a significant reduction in antianginal medication. An increased use of heart failure medication (p < 0.01) was registered, whereas the number of hospitalizations due to heart failure was not significantly higher than in the MT group. No late changes in LVEF or mortality were observed between the treatment groups during follow-up. However the study was not designed to assess mortality and should be interpreted as such.
The antianginal effect of TMR at late follow-up in this study is consistent with reports from observational studies (4,20,21), whereas late follow-up data have not been published from randomized clinical trials. The improved QoL during the first 12 months after TMR complies with the improved QoL reported from other randomized studies as well (5,6). Angina classification is a soft parameter in the evaluation of treatments in coronary heart disease. Classification may vary between observers and may be affected by patient as well as doctor bias in unblinded studies (5,6). To evaluate the angina classifications, we compared NYHA classification and SF-36 PCS and MCS scores. The PCS scores were significantly different between patients classified into NYHA functional class II, III and IV, respectively, with the lower score in NYHA functional class IV patients, intermediate score in NYHA functional class III patients and higher score in NYHA functional class II patients. No significant differences were observed between NYHA functional class II and NYHA functional class I patients or with respect to MCS score. In summary, the assessment of angina complies with physical components of QoL assessed by SF-36.
The fewer hospitalizations due to unstable angina in the TMR group were statistically significant compared with the MT group after 43 (range: 32 to 60) months but not after 12 months (9). Late biologic effects could explain differences. However, differences may as well be explained by the close contact between the patients and the hospital personnel during the first year, whereas this was not the situation thereafter.
Perioperative mortality has been reported as quite high in some studies (20,22,23)and has not encouraged the use of TMR in high-risk patients. Proper patient selection and perioperative care have reduced risk and mortality (14,15,24). Our selection criteria were based on clinical experience available in the literature until 1995 and excluded high-risk patients who might have benefited more from TMR (age > 75 years, LVEF < 30% or unstable angina) (7,14). Scarring and silent ischemia may be substrates for arrhythmias and sudden death, and the extensive use of beta-blockers in both groups may have contributed to the low overall mortality. An increased incidence of clinical heart failure after TMR is reported in our study. Early detrimental effect on left ventricular function and heart failure has been described after TMR (11,14,15). Early and late detrimental effects on LVEF have been reported three months (5)and three years after TMR (20), and a recently published stress echocardiography study (13)indicated a deterioration after three months. We were not able to detect late differences with respect to LVEF in the present study. A high number of channels were made in our study, and the increased use of heart failure medication could be explained by diastolic dysfunction secondary to ischemia or laser induced necrosis and fibrosis in the myocardium (25). No report on late diastolic effects after TMR has been published. Denervation could as well change symptoms from angina into dyspnea to be misinterpreted as heart failure, which eventually could explain much of the antianginal effects.
The results are based on a single center study of a small patient population. The inclusion and exclusion criteria excluded high-risk patients. Angina symptoms are a soft parameter and their assessment affected by patient’s as well as doctor’s bias, and the results of this study should be interpreted in light of this. A new perfusion scan or DSE was not performed at late follow-up, and the study gives no indication to whether improved myocardial perfusion, placebo effects, denervation or other mechanisms are operating.
Compared with a control group, angina symptoms were still significantly improved three to five years after TMR. An increased incidence of heart failure seemed to occur, but no differences in LVEF or mortality were registered.
☆ Supported by the Norwegian Ministry of Health and Social Affairs.
- dobutamine stress echocardiography
- left ventricular ejection fraction
- summary mental component scale
- medical treatment
- multiple-gated acquisition
- New York Heart Association
- baseline outcome variable
- follow-up outcome variable
- summary physical component scale
- quality of life
- MOS 36 Short-Form Health Survey
- technetium 99m-tetrofosmin myocardial perfusion scan
- transmyocardial revascularization with CO2laser
- Received October 15, 2001.
- Revision received February 21, 2002.
- Accepted February 22, 2002.
- American College of Cardiology Foundation
- Frazier O.H,
- Cooley D.A,
- Kadipasaoglu K.A,
- et al.
- Burkhoff D,
- Schmidt S,
- Schulman S.P,
- et al.
- Aaberge L,
- Nordstrand K,
- Dragsund M,
- et al.
- Hughes G.C,
- Landolfo K.P,
- Lowe J.E,
- Coleman R.B,
- Donovan C.L
- Lutter G,
- Saurbier B,
- Nitzsche E,
- et al.
- Ware J.E
- Ware J.E,
- Kosinski M,
- Keller S.D
- Pocock S
- Nagele H,
- Stubbe H.M,
- Nienaber C,
- Rodiger W
- Burkhoff D,
- Wesley M.N,
- Resar J.R,
- Lansing A.M