Objective. To perform a comprehensive analysis of the integrin forms expressed by normal human articular chondrocytes.Methods. Cartilage sections and collagenasereleased chondrocytes were probed with a comprehen- Conclusion.Normal human articular chondrocytes prominently display substantial quantities of the alp1, a5&, and aVp5 integrin heterodimers, as well as lesser quantities of the a3& and heterodimers. The a,, subunit-containing integrins are detected more readily on the more superficial chondrocytes than on chondrocytes deep within cartilage. These observations provide the basis for analysis of the role of chondrocyte integrins in cartilage homeostasis and in the pathogenesis of joint diseases.

Medial displacement osteotomy of the calcaneous is commonly performed for stage II posterior tibial tendon insufficiency in an effort to improve the valgus deformity of the hindfoot. We performed an anatomic study examining the medial neurovascular anatomy and its relation to the osteotomy in an attempt to determine which structures may be at risk during the procedure. Calcaneal osteotomies were performed through a lateral approach on 22 fresh-frozen cadaver below-knee specimens. Dissection was then performed medially to identify the Medial Plantar Nerve (MPN), the Lateral Plantar Nerve (LPN), the Posterior Tibial Artery (PTA), and their respective branches. Measurements determined either 1) where the structure crossed the osteotomy or 2) if the structure did not cross, the closest perpendicular distance from the osteotomy and at which point along its length this occurred. Perpendicular distances were recorded in millimeters and position along the osteotomy as a percentage of the total length from the posterosuperior aspect. An average of four neurovascular structures crossed each osteotomy site (range 2 to 6), most of which were branches of the LPN or the PTA. The MPN did not cross in any of the specimens studied, the LPN crossed in one specimen, and the PTA crossed in two specimens. The MPN distributed no crossing branches. The calcaneal sensory branch of the LPN was identified and crossed in 86% of the cadavers at 19% (+/- 15%) along the osteotomy length. A more distal second branch of the LPN (Baxter's nerve) was identified and crossed in 95% of the specimens at 61% (+/ 20%) along the osteotomy length. A third crossing branch existed in one specimen. Each PTA distributed from zero to three branches which variably crossed the osteotomy at a point from 2% to 100% along its length. The PTA bifurcated in 77% of the specimens at 49% (+/- 9%) along the osteotomy length. A consistent finding in every specimen was the presence of two veins accompanying the PTA with one on either side. A number of medial neurovascular structures may be at risk when perfoming a calcaneal osteotomy through a lateral approach. A minimum of two structures crossed the osteotomy site at variable positions in this study, although most of these structures represented branches off of the LPN or the PTA, with the LPN and the PTA themselves crossing only infrequently. The authors recommend that the completion of the osteotomy through the medial calcaneal cortex be performed in a carefully controlled manner to reduce the risk of post-operative complications including pain, numbness, and hematoma formation.

Although most authors recommend either 3.5-mm or 4.5-mm cortical screws for syndesmosis fixation, the optimum screw size has yet to be defined. The present study was designed to biomechanically compare syndesmosis fixation with 3.5-mm and 4.5-mm stainless steel screws. Simulated pronation external rotation ankle injuries were created in twelve paired, fresh-frozen cadaveric leg specimens. One limb from each pair received a 3.5-mm tricortical stainless steel screw for syndesmosis fixation (group I), while the contralateral specimen was stabilized using a 4.5-mm screw (group II). Sub-maximal axial ramp (0 to 1200 N) and external rotation/torsional ramp (0 to 5 N-m) loading was performed on each specimen prior to ligament division, following ligament division and following syndesmosis fixation. Axial fatigue testing was then performed at 1.5 Hz for a total of 100,000 cycles (0 to 900 N), and each specimen was subsequently tested to failure in external rotation. Ligament division resulted in syndesmosis widening (p<0.001) and reduced stiffness (p<0.001) during torsional ramp loading. Subsequent syndesmosis screw placement reduced syndesmosis widening (p<0.05) and increased stiffness (p<0.05). Following screw fixation, however, widening remained greater (p<0.005) and stiffness less (p<0.001) than pre-injury levels. No differences between groups I and II were observed during submaximal testing. In external rotation to failure testing, group I failed at a greater angle (38.9 degrees +/- 4.1 degrees vs. 32.0 degrees +/- 3.8 degrees in group II; p<0.05). Failure torque was slightly higher in group I; however, the difference was not statistically significant (17.8 +/- 2.0 N-m vs. 14.3 +/- 2.6 N-m in group II; p=0.082). Five specimens in group I failed by screw pullout and five specimens in group II failed by fibula fracture (p=0.061). The present results suggest that there is no biomechanical advantage of a 4.5-mm screw over a 3.5-mm in fixation of the syndesmosis.

The integrins are a family of adhesion-mediating cell surface receptors that play critical roles in cell-extracellular matrix interactions and have been shown to be important in the healing response in several tissues. We have studied integrin expression in normal human and rabbit anterior cruciate (ACL) and medial collateral (MCL) ligaments of the knee as a preamble to studies of beta 1-integrin expression in healing ligaments. Histologic sections of human and rabbit ACL and MCL were probed for integrin expression utilizing integrin-specific monoclonal antibodies (mAbs) followed by immunoperoxidase detection. Staining of human specimens with mAbs revealed the presence of beta 1-, alpha 1-, and alpha 5-integrin chains on the tissue fibroblasts of both ACL and MCL, while staining of rabbit specimens with rabbit integrin-reactive monoclonals revealed the presence of beta 1- and alpha 5-integrin on these ligaments. Equivalent amounts of the integrins studied were present on normal ACL and MCL. We conclude that the rabbit is an appropriate model for analyzing the expression and functional role of integrins in ligament wound healing.