Classifications of shoulder arthritis describe either a cuff tear arthropathy (CTA) in the coronal or a primary osteoarthritis (OA) in the cross-sectional plane. None combine the two directions, which may neglect a mixed excentricity.
130 standardised CTs of patients which underwent shoulder arthroplasty between 2009-2020 in our department were available for analysis. These were evaluated according to existing classifications (Walch, Favard and Hamada) and then measured according to a self-developed protocol: The glenoid centre, trigonum and inferior angle of the scapula were aligned in one plane using a 3-dimensional software. The subluxation of the humeroscapular alignment (HSA) was measured as the distance from the centre of rotation of the humeral head to the scapula axis (line from trigonum through glenoid centre). This was then expressed as a percentage of the radius of the humeral head in both the antero-postereor (A-P) and supero-inferior (S-I) direction. To evaluate reproducibility, a second independent observer measured a random selection of 50 of the CTs.
The diagnosis recorded in the operative report was CTA in 83 cases and OA in 47. The evaluation of the existing classifications showed a mixed excentricity in 76 instances. The subluxation of the HSA had a range of -74.1% (posterior) to +23.5% (anterior) A-P and -17.2% (inferior) to +68.6% (superior) S-I. The mean diameter of the head was 44.4 mm in the cross-sectional and 47.0 mm in the coronal plane, paired t-test <0.001. The interrater correlation coefficient showed a high degree of reproducibility (0.864 – 0.985).
A central HSA was defined as between -20% to +5% A-P and -5% to +20% S-I, after a graphical analysis of the values displayed in ascending order. The alignment was the grouped in both planes and described in combination as posterior (P), central (C) or anterior (A), as well as superior (S), central (C) or inferior (I). Cases with posterior subluxation of >60% of the radius (equivalent to 80% of the diameter of the humeral head) or with static acetabularization, were additionally labelled as extra-posterior (XP) or extra-superior (XS) respectively. 22 patients had central-central (CC), 40 central-superior (CS) and 1 central-inferior alignment. Of 60 shoulders with posterior subluxation, 33 were postero-central (PC), 24 postero-superior (PS) and 3 postero-inferior (PI). There were 3 patients with antero-central (AC) and 4 antero-superior (AS) arthritis. 5 cases of XP and 17 of XS malalignment were identified. Erosion was graded as 1-3 (1: No substantial erosion, 2: Biconcavity or central erosion/crater, 3: Neoglenoid covers the entire glenoid surface in one plane).
Our results show that to comprehensively describe the pathology of degenerative arthritis of the shoulder, a 3-dimensional classification based on the HSA should be employed. Such a classification could greatly facilitate preoperative planning when performing arthroplasty.