Transmyocardial revascularization is a new surgical procedure that has been shown to be effective in recent clinical trials in reducing angina symptoms. However, the exact mechanism of its angina reducing effect is still controversial. It was hypothesized that blood would flow through open transmural channels, and since laser was thought to be important in maintaining long-term channel patency, it has become the preferred method in recent years. However, as these two pictures from Dr. Burkhoff's group shown here, TMR channels do not stay open, even when they are created with laser.

This is at 24 hours, and the lumen is already filled with fibrinous material. At two weeks, this channel is completely obliterated by scar tissue.

These findings, together with some other observations, have prompted others to propose an alternative hypothesis based on angiogenesis. Angiogenesis occurs as part of normal wound healing process in many tissues. If one were to consider TMR simply as a way of creating controlled myocardial injury, then it is reasonable to expect that such an angiogenesis also occurs in the heart muscle. The following collateral circulation formation may improve myocardial perfusion. Secondly, if you consider angiogenesis as a nonspecific response to tissue injury, then perhaps laser need not be the only method in creating such injuries.

We tested the hypothesis in a model of porcine chronic ischemic model using ameroid constrictors, and the TMR operations are performed at six weeks after.

For this experiment, we used two very different methods of producing TMR channels. One is by laser. This is a general purpose CO₂ laser. The cost of the system is probably somewhere between one hundred and two hundred thousand dollars.

We compared that to a very simple, basic needle mechanical TMR using an 18 gauge hypothermic needle, which probably costs somewhere between one or two dollars.

A total of 20 animals were randomly assigned to one of the four treatment groups. Group 1 had ten punctures by laser. Group 2 had ten punctures by needle, and group 3 had three times as many punctures in the same area using needles. Group 4 had sternotomies only and served as our control. The transmural channels were created in an area between the first and second obtuse marginal branches in an area that measures about 2 x 2 cm.

For this study, we specifically looked at two of the indices of tissue angiogenic response. The first one is the total amount of VGEF expression in the treatment area. One of the observations we had was that tissue injury and inflammation are limited to the immediate vicinity of these puncture sites, and VGEF expression measurements are taken from near these puncture sites. Therefore, the total amount of VGEF expression is shown as the measured amount times the number of punctures. Secondly, we have noticed positive VGEF stains on the endothelial cells of a number of small vascular structures. We believe these are developing blood vessels directly stimulated by VGEF. Densities of these blood vessels are used as an index of angiogenesis.

Here is an example of how VGEF expression is measured using computer-assisted morphometry. Left, you have a section taken from tissue treated with TMR, and the brown color is the positive stains. What is not showing here is that there is very little or almost no stains in areas remote from the treatment site or in the control group. If you look at the left lower corner of this slide, you will see there is almost no stain there, and that is what you will see in the control group. By using an image analysis program, the areas of positive stains are automatically selected, and are calculated to give a semi-quantitative measurement of VGEF expression.

Here is what we found. Again, group 1 had ten laser punctures. Group 2 had ten needle punctures, and group 3 had 30 needle punctures. There are two points to see here. One is the laser group. When same numbers of punctures are made, the laser group had a higher level of VGEF expression, although the difference was not statistically significant. The second is that by far the highest level of VGEF expression is achieved in the 30-needle puncture group and the difference is statistically significant compared to the other groups.

Here are some representative images of VGEF-stained blood vessels. On the right is a section taken from laser treated areas, and on the left is the needle treated segment. You will notice that there are a number of vessels of various sizes. Their endothelium is stained with VGEF in brown color, and the number of these blood vessels are calculated and used in our comparison of vascular density.

Here is the result. Again, there are two points to see here. First is that all three treatment groups had significantly elevated vascular density. The second is that the highest vascular densities are achieved in both the laser group as well as the 30-needle puncture group. There is no difference between these two groups statistically.