TO THE EDITOR:
I would like to comment on "Tendon-Healing to Cortical Bone Compared with Healing to a Cancellous Trough. A Biomechanical and Histological Evaluation in Goats" (77-A: 1858—1866, Dec. 1995), by St. Pierre et al. The authors suggested that this model allowed evaluation of the biomechanical value of tendon repair to a cancellous trough as compared with that of repair to cortical bone. However, at twelve weeks, the tendon inside the cancellous trough was well healed and failure occurred proximal to it, level with the surface of the cortical bone. The tendon that was fixed to cortical bone failed at the same location. Thus, in this goat model, the tendon substance failed first so that differences in the strength of tendon-healing to the bone tissue would not have been demonstrated even if they had been present. There was no failure of the substance of the normal, control tendons, which failed at a threefold higher load either by pull-out at the clamp or by avulsion fracture. Thus, something due to the tendon repair induced this weakening of the tendon substance. The authors noted that the site of failure was probably the zone of edema and vascular proliferation seen in the healing tendon of animals used for histological evaluation. It is certainly possible that the histological findings represent a generic response to injury and repair and that they may have weakened the tendon. However, it is also possible that the type of tendon-gripping stitch that was used by the authors strangulated the tendon or that the tendon was damaged by abrasion against the cortical edges of the trough or by other factors.
Rodeo et al.4 placed a tendon through a bone tunnel in dogs and found that at eight weeks the tendon failed under load by pull-out from the tunnel, while at twelve weeks the tendon failed either by pulling out of the clamp or by mid-substance rupture. Although several eight-week specimens left a thin membrane of tissue along the surface of the bone, Rodeo et al. did not report a consistent site of failure at the entrance into the cancellous tunnel in their model, in which a tendon-gripping suture had not been employed. Gerber et al.1, comparing different suture-repair techniques for the rotator cuff, did not recommend the technique of tendon-grasping that yielded the greatest in vitro pull-out strength but rather the technique that provided one of the greater strengths but was least likely to cause strangulation of the tendon.
With respect to clinical relevance, soft-tissue repair is likely to be quicker, stronger, and more reliable in healthy young animals than in older, degenerated, poor-quality human tissue such as is often seen with large tears of the rotator cuff. Furthermore, the bone at the tuberosity is often sclerotic and covered with bursal tissue. While many of us no longer create deep cancellous troughs, in order to avoid weakening the bone that must hold sutures, most believe that it is important to freshen the bone to some extent to stimulate healing. A healthy goat tuberosity from which the tendon has just been removed is essentially prefreshened.
St. Pierre et al. are to be commended for critically investigating an area where clinical practice is more often guided by tradition and authoritative statements than by science. However, as in all basic-science studies, caution must be exercised when drawing conclusions for clinical practice.
Evan L. Flatow, M.D.: The Shoulder Service, New York Orthopaedic Hospital Associates, Columbia-Presbyterian Medical Center, 161 Fort Washington Avenue, New York, N.Y. 10032
Lieutenant Colonel St. Pierre, Major Olson, Major Elliott, Major O'Hair, Lieutenant Colonel McKinney, and Colonel Ryan reply:
We thank Dr. Flatow for critically reviewing our manuscript. His first comment was in regard to the location of tendon failure in our model. He concluded that the small amount of tissue that was left on the bone after failure at twelve weeks indicated that we were not testing the strength of the tendon repair to bone. Our conclusion is that the small amount of tissue that remained on the bone indicates the development of Sharpey fibers connecting the tendon to the bone. This conclusion was supported histologically by the presence of organized parallel fibers extending from the tendon to the bone. The fact that these fibers remained in continuity with the bone, despite failure of the tendon immediately proximal to its insertion, demonstrates the strength of this connection. The membrane that Rodeo et al.4 described at the surface of the bone in their eight-week specimens and the tissue that remained in our twelve-week specimens appear to be identical.
We do not believe that there is any basis for Dr. Flatow's comment about strangulation of the tendon with the suture technique. If the looping passes of suture were strangulating the tendon, then failure should have occurred at that site instead of distally at the bone-tendon junction. There was no gross or histological evidence of avascular tendon between the grasping suture proximally and the insertion of the tendon into the bone. In fact, we noted vascular proliferation distal to where the grasping sutures were located. Clinically, this technique has been used by a number of experienced shoulder surgeons without evidence of increased failure.
Failure of the controls by fracture of the tuberosity is similar to what happens clinically in fractures of the proximal aspect of the humerus, as described by Neer2,3. It makes sense that, in the shoulders of young healthy goats, the uninjured, unrepaired tendon is stronger than the repaired tendon at six and twelve weeks.
With respect to clinical relevance, we were very careful not to make any recommendations for clinical application. We believe that clinical studies should be performed to evaluate the findings in our study before widespread clinical application. Our findings do not support the repair of tendon to the thickened bursal tissue often noted during repair of the rotator cuff. We noted only that there was no difference between repair to cortical bone and repair to cancellous bone. The fact that Dr. Flatow and others only freshen the cortical bone, instead of making a trough to reach cancellous bone, indicates that they may be applying the findings of our study to their clinical practice. We hope to support our study with an ongoing prospective randomized clinical trial.
Lieutenant Colonel Patrick St. Pierre: Department of Orthopaedics, Madigan Army Medical Center, Tacoma, Washington 98431
Major Eric J. Olson: Orthopaedic Service, Walter Reed Army Medical Center, 6825 16th Street, Washington, D.C. 20307
Major James J. Elliott; Major Kevin C. O'Hair: United States Army Medical Research Institute for Infectious Diseases, Fort Detrick, Maryland 21701
Lieutenant Colonel Lu Ann McKinney: Department of Veterinary Pathology, Armed Forces Institute of Pathology, 14th and Alaska Avenue, N.W., Washington, D.C. 20306
Colonel Jack Ryan: United States Army Institute for Surgical Research, Fort Sam Houston, Texas 78234