Treatment of proximal humeral fractures with reverse shoulder arthroplasty in elderly patients

Proximal humeral fractures in four or even only three parts, with metaphyseal hinge distances of <8 mm, represent a serious and widely debated problem. Reduction is complex and plating is often instable, especially in elderly patients. Failures, sometimes involving necrosis of the head, are frequent. Hemiarthroplasty has long been used for 3- or 4-part complex fractures, even in young patients, although often with sub-optimal results, due to reabsorption of tuberosities. This complication has partly been overcome with reverse shoulder prostheses which, although more invasive than partial ones, may lead to less disappointing results, even in cases of reabsorption of tuberosities. We have data on a homogeneous series of patients treated with reverse shoulder arthroplasty for proximal fractures, with a maximum follow-up of 10 years. The aim of this study was mainly to identify which cases can be selected for effective treatment and which technical aspects are best to adopt.

There were 33 patients in this study, mean age 76.6 years (range 54–85). Fractures were classified according to Neer. Surgery was undertaken on average 4.4 days after trauma. The deltopectoral approach was used. Sutures were hooked over the major and lesser tubercles for later reduction and fixation after the prosthesis had been applied. This passage was sometimes not possible in cases of serious degeneration of the rotator cuff. One day after surgery, a shoulder brace providing an abducted angle of 15° was applied for 30 days. Patients were re-assessed with DASH and Constant scores (CS), and the ratio between healthy and operated shoulders was calculated. Physical examination was followed by X-rays, mainly to evaluate and classify any infraglenoid scapular notching according to Nerot.

Mean follow-up was 42.3 months (range 10–121). According to the CS, mean pain was 12.6/15 (range 3–15/15), activities of daily living 16.3/20 (range 8–20/20), ROM 21.8 (range 8–32/40) and power 5.4/25 (range 2–12/25). Total mean CS was 56.4 (range 23–80/100). The mean DASH score was 49.7 (range 32–90). The ratio of the CS parameters between opposite and operated shoulders was on average 72.8 % (range 28–90 %). Long-term complications were eight cases of scapular notching (24.2 %) of which four of grade 2 (12.1 %) and four of grade 1 (12.1 %).

Total reverse prostheses are more invasive because they also compromise the glenoid surface of the scapula, but they do offer good stability, even in cases of damage to the rotator cuff. Reverse prostheses have great advantages as regards to ROM, allowing functional recovery, which is good in cases with re-insertion of tuberosities, and acceptable in cases when tuberosities are not re-inserted or resorbed. In our cases, the first 3 reverse prostheses lasted 10, 8.3 and 7.3 years, and we believe that they will become increasingly long-lived, so that applying them in cases of complex fractures becomes more feasible. We prefer the deltopectoral approach because it can reduce and stabilize possible intra-operative diaphyseal fractures. Possible scapular notching must be foreseen when inserting the glenosphere. We had eight cases (24.2 %), of which four were Nerot grade 1 and four were grade 2. Applying the Kirschner wire in an infero-anterior position allows the glenosphere to be lowered with a tilt of 10°. Reverse prostheses are suitable for 3- or 4-part complex proximal humeral fractures in patients over 65. Prolonged physiokinesitherapy is essential.