An Angiosarcoma in the Left Atrium

Transmyocardial Laser RevascularizationTo the Editor: The placebo effect of surgery for the treatment of angina that is mentioned by Lange and Hillis (Sept.30 issue) 1 in their editorial on transmyocardial laser revascularization deserves elaboration.In the era before direct coronary revascularization, internal-thoracic-artery ligation was touted as a method to increase myocardial blood flow and relieve anginal symptoms.Initial reports by Ellis et al. 2 and Kitchell et al. 3 indicated that 68 to 75 percent of patients had clinical improvement, including approximately 35 percent who had complete relief and 42 percent with objective improvements, as measured electrocardiographically.These investigators were honest and intelligent but not impartial observers.Double-blind studies conducted later by Cobb et al. 4 and Dimond et al. 5 demonstrated that a sham thoracotomy alone could decrease the need for nitroglycerin and increase exercise tolerance and that it produced subjective improvement in more than 35 percent of patients.These results are similar to the improvement among 34 percent of the patients in the British study of transmyocardial laser revascularization mentioned by Lange and Hillis.When perfusion scans do not correlate in time or magnitude with the patient's clinical improvement, the more subjective data on improvement of symptoms must be approached with great skepticism.As was the case in earlier studies, the thoracotomy incision alone may have an important effect directly or indirectly on the patient's perception of pain.The fact that a patient has already undergone the surgical procedure of last resort has a high likelihood of affecting a physician's choice of treatment plans, and thus study outcome.It may be completely correct that transmyocardial laser revascularization benefits patients with chronic angina.However, given the substantial morbidity and mortality (9 to 15 percent mortality among the patients who were crossed over to transmyocardial laser revascularization in the studies by Frazier et al. 6 and Allen et al. 7 discussed by Lange and Hillis), the potential for a placebo effect must always be remembered.