We would like to thank Dr. Horvath for his insightful comments. The goal of our paper was to present an unbiased review of the topic of myocardial laser revascularization with an emphasis on randomized, controlled clinical trials. Given the word limitations imposed on our manuscript, it would have been inappropriate for us to delve into laser-tissue interactions at the expense of important clinical data. Moreover, an appropriate reference on laser-tissue interactions was made available to the reader.

At no point in our paper did we “equate” transmyocardial revascularization (TMR) and percutaneous myocardial laser revascularization (PMR). These are distinctly separate techniques. Neither do we suggest that the CO₂ and Ho:YAG lasers are of equal value. Any claim, however, that the CO₂ laser is superior to the Ho:YAG laser is speculative, and remains to be shown in a head-to-head randomized clinical trial.

The experimental studies quoted in our paper used animal models of chronic myocardial ischemia, akin to patients with chronic angina. We found no study that entirely explains the mechanism of action of TMR. There are several studies for and against each hypothesis; therein lies the controversy.

We presented the trials in tabular form to emphasize their similarities. Differences were noted in the text. As Dr. Horvath points out, a number of small, nonrandomized studies with short-term follow-up demonstrate enhanced perfusion post-TMR. This could either be related to laser-induced angiogenesis, or to the natural development of collateral vessels in patients with chronic ischemia. The latter explanation emphasizes the danger of relying on the results of uncontrolled studies because, for the most part, enhanced perfusion has not been confirmed in randomized clinical trials. Transmyocardial laser revascularization did not improve myocardial perfusion in four of five trials in which perfusion was assessed before and at various times after enrollment. In the trial by Frazier et al. (1), in which a benefit was seen, there was only a 49% follow-up in the medical arm of the study. Also, the degree of symptomatic improvement was vastly disproportionate to the degree of improvement in perfusion.

In the trial by Frazier et al. (1), 59% of patients initially assigned to maximal medical therapy crossed over to the TMR group. The investigators allowed crossover as an enticement for patients to remain in the study if medical therapy failed and the end point of angina was reached. Angina, unfortunately, is a subjective measure. Irrespective of how “objective” the investigators were in determining the success or failure of antianginal therapy, the use of a subjective end point may have inadvertently introduced bias into the trial. It is difficult to draw proper conclusions from such a trial when large crossover rates are allowed.

With respect to the European trial by Schofield et al. (2), the number of sites with irreversible segments was adjusted for baseline, and for repeated within-patient, between-site measures. Therefore, to suggest that a “doubling” of fixed defects in the medical therapy group implies enhanced perfusion in the TMR group is inappropriate, especially because a subgroup analysis of the same TMR patients showed no improvement in myocardial perfusion with PET scanning (3).

To conclude, we believe that the reported benefits of TMR, even out to five years, may be related to the placebo effect. It is the most plausible mechanism of action, given the lack of concrete evidence to the contrary. A properly powered, blinded, sham-controlled surgical trial of TMR could certainly settle this issue. In the absence of such a trial, however, more studies using new perfusion imaging modalities must be conducted to elucidate the true value of this technique.