Tendon end separation with loading in an Achilles tendon repair model: comparison of non‐absorbable vs. absorbable sutures

Michael R. Carmont, Jan Herman Kuiper, Karin Grävare Silbernagel, Jón Karlsson, Katarina Nilsson‐Helander

Rupture of the Achilles tendon often leads to long-term morbidity, particularly calf weakness associated with tendon elongation. Operative repair of Achilles tendon ruptures leads to reduced tendon elongation. Tendon lengthening is a key problem in the restoration of function following Achilles tendon rupture. A study was performed to determine differences in initial separation, strength and failure characteristics of differing sutures and numbers of core strands in a percutaneous Achilles tendon repair model in response to initial loading. Nineteen bovine Achilles tendons were repaired using a percutaneous/minimally invasive technique with a combination of a modified Bunnell suture proximally and a Kessler suture distally, using non-absorbable 4-strand 6-strand repairs and absorbable 8-strand sutures. Specimens were then cyclically loaded using phases of 10 cycles of 100 N, 100 cycles of 100 N, 100 cycles of 190 N consistent with early range of motion training and weight-bearing, before being loaded to failure. Pre-conditioning of 10 cycles of 100 N resulted in separations of 4 mm for 6-strand, 5.9 mm for 4-strand, but 11.5 mm in 8-strand repairs, this comprised 48.5, 68.6 and 72.7% of the separation that occurred after 100 cycles of 100 N. The tendon separation after the third phase of 100 cycles of 190 N was 17.4 mm for 4-strand repairs, 16.6 mm for 6-strand repairs and 26.6 mm for 8-strand repairs. There were significant differences between the groups (p < 0.0001). Four and six strand non-absorbable repairs had significantly less separation than 8-strand absorbable repairs (p = 0.017 and p = 0.04 respectively). The mean (SEM) ultimate tensile strengths were 4-strand 464.8 N (27.4), 6-strand 543.5 N (49.6) and 8-strand 422.1 N (80.5). Regression analysis reveals no significant difference between the overall strength of the 3 repair models (p = 0.32) (4 vs. 6: p = 0.30, 4 vs. 8: p = 0.87; 6 vs. 8: p = 0.39). The most common mode of failure was pull out of the Kessler suture from the distal stump in 41.7% of specimens. The use of a non-absorbable suture resulted in less end-to-end separation when compared to absorbable sutures when an Achilles tendon repair model was subject to cyclical loading. Ultimate failure occurred more commonly at the distal Kessler suture end although this occurred with separations in excess of clinical failure. The effect of early movement and loading on the Achilles tendon is not fully understood and requires more research.