Reverse prosthesis: Rationale, biomechanics and medium-term results in cuff tear arthropathy

Authors: F. Postacchini

References: SECEC 2005, Rome

Cuff tear arthropathy is characterized by degenerative changes of the humeral; head, massive tear of the rotator cuff and upward migration of the humeral head, which looses the normal relationship with the glenoid.  The latter may undergo arthritic changes due to abnormal reciprocal motion with the head of the humerus.  Inverse prosthesis is able to restore a functional relationship between humerus and glenoid and allow better functioning of the deltoid.

In the past, several models of inverse prosthesis have been conceived, but all failed at long term because the centre of rotation of the joint was left as lateral as in the normal joint.  As a result, the force passing through the centre of the “inversing head” caused a bending moment, which led to mobilisation of the glenoid component.  The Grammont’s intuition was to conceive a prosthetic model which medialises the centre of rotation, thus reducing the mechanical stress on the glenoid.  In addition, such a model involves humeral lowering, which leads the centre of rotation to coincide with the centre of the glenosphere.  However, lowering of the centre of the rotation does not occur when the humeral head is not displaced proximally, as in fractures of the upper humerus.  In both cases, the arm lengthens by 1-2 cm due to tensioning of the deltoid.  The strength of the deltoid, in cuff tear arthropathy with upward migration of the humeral head, increased due to lowering of the centre of rotation (increased tension of the muscle) and medialization of the latter because a larger portion of the muscle can be activated during shoulder flexion and abduction.

We analysed the results of surgery in 18 patients (mean age, 69y) with cuff tear arthropathy 3.1 years on average (range 2 to 6y) after implantation of an inverse prosthesis.  At F-U, the patients were evaluated with the Constant-Murley method and radiographs of the shoulder were obtained in each case.  The mean score for pain, ADL, forward flexion, abduction, internal rotation and external rotation had increased by 7.6, 9.5, 4.9, 1.2, 3.8 and 0.9 points, respectively.  The mean Constant score was 33% preoperatively and 61% at F-U.  In 8 patients we found a glenoid notch.  We compared the results in these patients with those in a series of 11 patients who had undergone hemiarthroplasty for cuff tear arthropathy a mean of 5 years before final evaluation.  At F-U, the mean Constant score in this group had increased by 9 points less compared with the inverse prosthesis group.  The difference was essentially due to less improvement in active motion in the hemiarthroplasty group.  It should be noted, however, that preoperatively none of the latter patients had less than 80° forward flexion and abduction.

In a multicentre study including 6 hospitals, we evaluated the results of surgery in 98 patients who had undergone total shoulder arthroplasty using Delta prosthesis a mean of 4 years (range 2.6-7.4 years) after operation.  The mean age at surgery was 71 years.  The superior approach had been used in 39 cases and the deltopectoral in 59.  All patients were assessed clinically using the Constant method and radiographed.  The mean pain score at F-U (12.9) was 5 points higher than the preoperative score.  The mean score for ADL had increased by 10 points (from 5.4 to 15.4).  The mean preoperative score for forward flexion (2.8), abduction (2.7), internal rotation (2.5) and external rotation (2.1) had increased by 5.1, 4, 1 and points, respectively.  The mean preoperative and postoperative Constant scores were 23.9 and 62.7, respectively.  The rate of complication was 10%.  Radiographs showed a glenoid notch in many patients, but only a minority complained of shoulder pain.