A thought began to brew in my mind last June at the ESHRS meeting in Brussels. Jean Devroye, MD introduced this thought to me. He noted that density studies done to date involved very irregular patterns. He suggested that his silicone sheets along with his regular patterns offered a solution to create a regular density. I agreed that some sort of standardization was needed.
As I contemplated his concerns further, I realized the obvious. Density studies without regular patterns create areas of high densities and areas of low densities. This was confirmed in my own patients where I average 45 grafts per square centimeter. Multiple measurements of post operative density confirmed that in some areas I created 60 grafts per sq. cm, but in other areas, the density was 30, 40, or 50 grafts per sq. cm. In other words, one has to take an average density to derive an average density and my average density has been 45 grafts per square centimeter for some time.
Recent studies suggest that the ideal yield at 10 months is with a density of 20 grafts per square centimeter. The yield at 20 grafts per square centimeter is greater than 90%. This is an ideal yield since typically 9 to 10% of all hairs on the scalp are in a resting phase. Yields begin to drop as densities exceed 20 grafts per square centimeter. At 30 grafts per square centimeter they drop to 73% and this low yield continues at 40 grafts per square centimeter. This is alarming. These studies, however raise some other very real concerns.
Imagine that you draw a 1 cm sq. box. Suppose you do your best to create a density of 10, 20, 30, 40, etc per sq cm. As the density increases the probability that isolated areas will have a higher density than other areas increases. How might isolated areas of increased density affect data? Well, I don't think you need to venture far beyond the recent Korean study in Derm Surg. or the preceding three studies by Mayer, Keene, and Beehner. Perhaps, however, the results of these studies are skewed by the probabilities that isolated densities exceed the study parameters, while other areas within the same square cm. are far less.
My concern has been that a cumulative threshold of trauma triggers some sort of inflammatory response that impairs hair graft survival. Isolated areas of higher density might indeed be the source of such a cascade threshold of pathological sequelae. In other words, you might get great growth at 30 grafts per sq. cm, but potential pathological affects at 40 or 50 per sq. cm. This supposition is supported in some degree by Beehner who showed that in at least one instance higher densities produce ideal yields in some individuals. Might it be that if the yeild at 40 grafts per sq cm is higher provided that the overall density is consistent?
Jean and I met recently to derive a study that creates a consistent density. We used his stencil to create a very regular pattern of recipient sites. Tree farmers among other mathematicians such as Ruifernandez and Francisco Jimenez, MD have shown that the highest densities are possible from triangular patterns as opposed to rectangular patterns. We used a triangular pattern in our study. We created a very regular pattern to study recipient site density and yield. In our study, however we used body hair.
Jean and I suggest that future studies measuring density and yield should utilize his stencils in an effort to create a consistent and regular density.
The enclosed photos show a regular pattern for body hair grafts in a sagital and coronal plane that include single hair anagen body hairs. Such regular patterns avoid the potential for isolated areas of a much higher density. Once again, Jean and I suggest that future density studies should utilize such stencils to insure more useful data.