The multiple fixation techniques described in literature indicate that the optimal treatment for displaced proximal humeral fractures continues to be controversial [7,8,11,19,20,21]. The preferred operation technique depends on fracture type, patient age, bone quality, and functional expectation. Surgery, using an angular and sliding stable antegrade nail (Targon PH), is a standard treatment option which can provide good functional results ༻10,22,23༽. Supplemental tension band sutures are recommended for proximal humeral fracture treatment in a myriad of literature with favorable clinical results༻12,13,14,15,16,17༽. Badman et al. advocated that effectiveness derives from the counterforce to the natural deforming forces of the rotator cuff༻12༽. According to Park et al., tension band sutures placed between the rotator cuff and the head of the interlocking screw or washer, using No.5 Ethibond suture material, increase the stability of the bone fragment with good postoperative shoulder function༻15༽. Badman et al. and Shukla et al. reported that locked plating with tension band rotator cuff fixation using a minimum of four or five No.2 FiberWire sutures can prevent fixation failure and result in favorable clinical outcomes༻12,17༽. Micic et al. emphasized the importance of applying a tension band suture over the tuberosity for additional stability; they report that negligence of this procedure is a risk factor for revision surgery༻24༽. On the other hand, there is a contradictory result which reports the invalidity of the tension band suture. Arvesen et al. performed a cadaveric study and concluded that tension-relieving rotator cuff sutures with No.5 FiberWire do not add stability to the repair of 3-part proximal humeral fractures༻25༽. Furthermore, Voigt et al. also reported no contribution to reduce interfragmentary motion by additive fiber-cerclages in unstable 3-part fracture model with an intact rotator cuff༻26༽. The necessity of the tension band suture is yet controversial and heterogeneity of surgical indication exists. Moreover, suture materials, artifacts, and threading methods for tension band suture vary in the literature, which hinders discussions of its effectiveness. In this study, we attempted to present an ideal method from a mechanical viewpoint by focusing on the combination of three essential elements of the tension band suture: the suture material, threading angle, and washer.
FiberWire, a representative nonabsorbable suture made of multi-strand, long-chain, ultra-high molecular weight polyethylene (UHMWPE) demonstrated higher strength than the other conventional sutures as Barber et al. reported; its superiority showed remarkable statistical significance in our study (P < 0.001) [27]. Moreover, the rupture pattern differed between FiberWire and the other suture materials. This might be ascribed to its structural composition and loading type. FiberWire consists of a UHMWPE core with a braided jacket of polyester and UHMWPE; whereas Ethibond and Surgilon are made of polyester and nylon with a braided structure coated with polybutylate and silicone. Wright et al. verified that FiberWire’s non-braided core, protected in its polyester jacket, resists elongation and enables it to maintain strength, even when the suture is partially cut༻28༽. In most previous biomechanical experiments, the load to failure tensile tests are performed by mere continuous traction to the suture ༻18༽. We performed cyclic loading in this study to replicate the type of load for which the tension band suture is considered to be exposed after surgery by the motion of the shoulder joint. Frictional force occurs repeatedly between the suture material and washer, in addition to tensile force. We think the gradual erosion of FiberWire by frictional force might be a consequence of the structural characteristics mentioned above. In spite of this distinctive property of FiberWire, Abbi et al. and Barber et al. reported that knot slippage occurred more frequently with FiberWire than Ethibond, which must be considered another mode of tension band suture failure༻27,29༽. In our study, there was no knot slippage, regardless of suture materials, utilizing 5 square knots tied on each end.
Theoretically, when loading an identical tensile force to suture material, the normal force at the contact point between the washer and the suture material increases as the threaded suture makes an acute angle. As a result of this larger dynamic friction force, the 15 degrees group had a tendency to be vulnerable; although no statistical significance existed between angle groups. We assume that threading the suture to the washer at a wider angle is desirable for rupture prevention. However, in clinical settings, the threading angle is affected by multiple conditions, such as design of artifact, bone fragments, and soft tissue.
Artifacts used for tension band sutures also play an important role. Generally, when treating with plates, dedicated eyelets in the plate are used to thread the suture [12,17]. However, Cho et al. illustrated the difficulty in providing tension to sutures using eyelets in the plates because the knots might eventually loosen༻13༽. There are plural methods for tension band sutures using IMN. Hao et al. introduced a technique to augment tuberosity fixation by threading suture holes on the interlocking screws༻30༽. This might potentially have the same issue as threading to eyelets in the plate; additionally, the contraction of the rotator cuff can lead to screw backout. Park et al. performed tension band and locking sutures in addition to IMN and reported good clinical outcomes༻16༽. They hung the sutures only at the head of the interlocking screw, which we consider technically difficult with a potential risk of suture slippage or knot failure. To prevent these risks, washers were introduced. Cho et al. used 2 washers with plates to interpose the suture material and transmit the tension through the sutures༻13༽. Kim et al. employed a washer to secure the suture and compensate for the shortcomings of the tension band sutures with IMN༻14༽. We advocate this technique and are attempting to refine the method. When threading the suture to the washer, frictional force becomes a problem. The type of washer did not significantly affect the result when threading to a single hole; however, threading both holes of the washer ring in succession (washer ring-2) militated against the durability. Thus, engendering frictional wear at two points is a risk for early rupture.
Our study has several limitations. This is an in vitro study, so our model does not completely replicate the in vivo environment. Different external forces might act on the tension band suture when using a bone model of proximal humeral fracture. Additional cadaveric study might reveal those dynamics. Also, the sample size for each combination was limited to three, because we used a brand-new washer for each trial to ensure a uniform environment.