The most important finding is that the technically feasible glenoid allograft for resurfacing of a degenerated glenoid surface combined with anatomical HA does not work in-vivo.
We report the results of five consecutive osteochondral glenoid allograft resurfacings of the glenoid for the treatment of glenohumeral OA young patients performed in a single institution. After unsatisfactory results with 3 other types of biological resurfacings (i.e. capsular interposition, meniscal allograft interposition, and glenohumeral GraftJacket™ interposition arthroplasty) [13], we aimed at trying to resurface the glenoid using a glenoid allograft. The concept was experimentally tested, the design of the construct was optimized and feasibility of glenoid allografting was validated in-vitro [15]. After successful experimental validation, a limited pilot study for patients with OA and an intact rotator cuff was set up [21]. As OA of the young can be either primary or occur after previous trauma or instability we decided to use allograft resurfacing for primary or secondary OA provided the rotator cuff was intact and there was no major posttraumatic malposition of the proximal humerus or the glenoid. At the time of this trial, the use of HA or TSA in Type B glenoids was accepted. Today, primary replacement with RTSA is preferred by many although good mid- to long-term results have also been observed with TSA [27].
A 2- to 15-year review presents the results of soft tissue glenoid resurfacing in 34 patients (36 shoulders) with a mean age of 51 years treated between 1988 and 2003. Various biologic surfaces, including anterior capsule (7 cases), autogenous fascia lata (11 cases) and Achilles tendon allograft (18 cases) were used, reflecting the evolution of the technique over time. All scores performed improved postoperatively, and using Neer’s criteria, the outcome was excellent in 18 shoulders, satisfactory in 13, and insatisfactory in 5 shoulders. Unsatisfactory results were related to infection, early re-injury, and the use of the anterior capsule as interpositional material. Erosion of the glenoid averaged 7.2 mm but appeared to stabilize after approximately 5 years [28]. This contrasts with results of a study that published lateral meniscus allograft or human acellular dermal tissue matrix in 42 patients at an intermediate-term follow-up of 2.8 years. The lateral meniscal allograft cohort had a failure rate of 45.2%, with a mean time to failure of 3.4 years. Human acellular dermal tissue matrix interposition had a failure rate of 70.0%, with a mean time to failure of 2.2 years [10].
All procedures in this study were performed by a single surgeon (C.G.) using a consistent, previously laboratory tested technique. The results of 5 out of 5 patients are known and only 1 result is clinically durably good but radiographically associated with advanced glenoid erosion, a combined clinical and radiographic result well known from HA without any resurfacing [1]. A statistical analysis is not feasible as early failure with severe pain and poor function requiring revision within 3 years, one case requiring revision at 9 years and one clinically good result with an unsatisfactory radiographic course are considered prohibitive for further use of this technique. Unfortunately, no difference in clinical or radiographic parameters could be identified which differentiated the two results which were satisfactory for at least 8 years and the three poor results.
As the main limitations, the range of indications can be criticized but this corresponds to the type of OA that is seen in rather young individuals who have an intact rotator cuff. Secondly, we used commercially available, size and side matched deep-frozen allografts which were thawed in Ringer’s solution at the beginning of the surgery, and we cannot exclude that some specific preparation of the graft material could have yielded better results, but the grafts used correspond to the standard used in other joints or indications. In addition, it may be criticized that there was no control group for this pilot study. The study, however, was designed to determine the safety and suitability of the procedure for a further RCT.
The early mid- and long-term results of HA and TSA for OA in relatively young patients are not yet where patients and surgeons would like them to be, but they are well established [1]. With a follow-up of 3 to 13 years, the overall implant survivorship is between 60 and 90%. Glenoid loosening is the most common reason for revision (52%). A reliable improvement of the subjective as well as objective outcome can be expected, which possibly decreases between mid- and long-term follow-up [29, 30]. Glenoid resurfacing using allograft is a theoretically attractive solution which can technically reproducibly be executed in-vitro. In-vivo, however, it led to early revision in three out of 5 cases due to severe pain, and to a result no better than the appropriately documented results of HA and TSA in the two best cases so that we have abandoned this concept.