Almost 30 years ago Gilles Walch called our attention to arthritic glenoid retroversion. Morphologic study of the glenoid in primary glenohumeral osteoarthritis

From that point on, there has been great interest in the version of the arthritic glenoid: what is it?, how should it be measured?.

And in the execution of an anatomic total shoulder (aTSA), should the version of the glenoid component be "corrected" to some particular value?, if so what value?, how should this correction be achieved? and does changing the preoperative version affect the clinical outcome of aTSA?

It goes without saying that a lot of time and money can go into the evaluation and management of arthritic glenoid retroversion. Perhaps it's time to see how much glenoid component matters to the patient.

About 15 years ago, Patterns of loosening of polyethylene keeled glenoid components after shoulder arthroplasty for primary osteoarthritis: results of a multicenter study with more than five years of follow-up pointed out that posterior tilting of the glenoid component was associated with preoperative posterior decentering and with excessive reaming. The authors suggested that preserving subchondral bone may be important for long-term longevity of the glenoid component.

Glenoid component retroversion is associated with osteolysis found that osteolysis around the center peg of a glenoid component was correlated with component retroversion of ≥15°, the paper clearly stated that "the presence of osteolysis around the center peg was not correlated with a worse clinical outcome defined by shoulder scores or a reoperation due to glenoid loosening".

Nevertheless, achieving component retroversion of <15° has become a goal for many surgeons and an opportunity for orthopaedic companies who have made substantial investments in three-dimensional planning platforms, patient specific instrumentation, navigation, augmented / virtual reality and robotic assisted glenoid preparation.

A recent paper, Does postoperative glenoid component retroversion following anatomic total shoulder arthroplasty affect clinical outcomes? A systematic review and meta-analysis assessed the clinical importance of implanting the glenoid component in <15° of retroversion. After screening 2,457 articles, 15 studies comprising 1,190 shoulders met inclusion criteria. Patients were stratified by whether postoperative glenoid component retroversion was <15° or ≥15°. The principal finding was that no clinically significant differences were observed between the two groups in patient-reported outcome scores, range of motion, or complications.

An unexpected and important finding in this meta-analysis was that shoulders with ≥15° of postoperative retroversion were actually more likely to have no radiolucency (a Lazarus grade 0 radiographic score) than those with <15° retroversion (76.9% vs. 55.6%; P = .00021). This finding argues against the presumption that retroversion promotes loosening.

Does Postoperative Glenoid Retroversion Affect the 2-Year Clinical and Radiographic Outcomes for Total Shoulder Arthroplasty? directly addressed the question in a case-control study. At 2-year follow-up, patients with glenoid components implanted in ≥15° retroversion had similar improvement in Simple Shoulder Test (SST) scores, similar final outcome scores, and similar radiographic findings compared to those with <15° retroversion. Notably, none of the patients with retroverted glenoid components underwent revision surgery, compared to 3 of 50 patients in the non-retroverted group.

Anatomic Total Shoulder Arthroplasty with All-Polyethylene Glenoid Component for Primary Osteoarthritis with Glenoid Deficiencies reported outcomes of aTSA with conservative, noncorrective reaming in shoulders with glenoid deficiencies, including those with significant retroversion. Mean postoperative retroversion in this series was 16°, yet mean postoperative SST was 9 out of 12, consistent with the excellent results achieved in shoulders without glenoid deformity. The revision rate was zero in 143 shoulders at mean follow-up of 34 months.

Anatomic total shoulder arthroplasty for posteriorly eccentric and concentric osteoarthritis: a comparison at a minimum 5-year follow-up compared outcomes of aTSA for posteriorly eccentric (Walch B) versus concentric osteoarthritis at minimum 5-year follow-up, finding no significant difference in ASES scores, revision rates, or radiographic loosening between groups. These durable results were achieved without attempting retroversion correction.

Does glenoid version and its correction affect outcomes in anatomic shoulder arthroplasty? A systematic review analyzed 16 studies and 1,211 shoulders finding that 8 of 11 reports found no significant association between pre- or postoperative glenoid retroversion and clinical results, including patient-reported outcomes, range of motion, and revision rates.

Total shoulder arthroplasty outcomes after noncorrective, concentric reaming of B2 glenoids reported a 95% implant survivorship at a mean of 4.9 years in a series treated with noncorrective reaming with a mean postoperative retroversion of 19°.

The Effect of Version Correction Techniques

If clinical outcomes are equivalent regardless of postoperative retroversion, the question becomes whether correction efforts add value without adding risk. Does postoperative glenoid component retroversion following anatomic total shoulder arthroplasty affect clinical outcomes? A systematic review and meta-analysis compared three correction strategies: posteriorly augmented glenoids, eccentric (“high-side”) reaming, and conservative noncorrective reaming. Eccentric reaming was associated with a significantly higher complication rate than noncorrective reaming (9.3% vs. 3.1%; P = .043, OR 3.22) and a significantly higher revision rate (7.4% vs. 1.2%; P = .015, OR 6.18).

Loss of the dense subchondral bone layer from corrective reaming may result in reduced glenoid component support, increased micromotion, and greater loosening risk over time.

Implications for 3D CT Planning and Technology Transfer

CT-based three-dimensional preoperative planning, patient specific instrumentation, intraoperative navigation, augmented/virtual reality, and robotic glenoid preparation are all predicated on the same clinical logic: that achieving a glenoid component position closer to neutral version will improve patient outcomes. As reviewed above, the available evidence indicates that differences in postoperative glenoid component retroversion do not produce clinically detectable differences in pain relief, function, or implant survival at the follow-up intervals studied.

Use of Preoperative CT Scans and Patient-Specific Instrumentation May Not Improve Short-Term Adverse Events After Shoulder Arthroplasty: Results from a Large Integrated Health-Care System compared aTSA with and without preoperative CT scanning and PSI. These technologies expose patients to additional radiation from CT scanning and incur substantially greater costs of care. Use of CT scans and PSI did not reduce the rate of short-term adverse events following shoulder arthroplasty. Patients receiving PSI may be at greater risk of deep vein thrombosis or deep infection, possibly reflecting the additional operative time this technology requires.

Three-dimensional computed tomography analysis of pathologic correction in total shoulder arthroplasty based on severity of preoperative pathology analyzed 152 shoulders with 3D CT postoperatively and found that while glenoid component shift occurred in 51% of patients, neither component shift nor central peg osteolysis was associated with worse clinical outcomes at minimum 2-year follow-up.

What the Evidence Does Suggest About Optimizing aTSA outcomes?

The available evidence points to factors other than retroversion that drive aTSA results.

See: Below left-poor glenoid seating with cement interposed between the glenoid component and the bone.

Below right-posterior decentering due to poor glenoid preparation and ill-advised use of a posteriorly eccentric humeral head component.

Surgeons may wish to consider 6 aspects of aTSA that are relevant to the glenoid side of the arthroplasty.

(1) conservative reaming to retain the maximal amount of quality host bone

(2) rather than "correcting" glenoid retroversion (A, below), consider "accepting" it (B, below)

(3) component seating — good carpentry with complete backside contact of the glenoid implant against prepared bone; no cement between the component and the bone.

(4) humeral head centering on the glenoid achieved through soft tissue balancing and the possible us of an anteriorly eccentric humeral component.

(5) awareness that technologies can lead surgeons to prioritize postoperative glenoid component retroversion <15° retroversion (without acknowledging the potential risks)

(6) recognition that specialized (e.g. augmented) glenoid components may have downstream risks (see chart below from the 2025 AOANJRR)

Here is a thought provoking case in which substantial glenoid retroversion was accepted

Function at 14 years

Conclusion:

Glenoid component version may not be as critical to the outcome of aTSA as secure seating of the glenoid component on quality host bone and centering of the humeral head on the prosthetic glenoid.

Rather than improving outcomes, there is evidence that "corrective reaming" can be associated with a significantly higher complication rate than noncorrective reaming (9.3% vs. 3.1% ) and a significantly higher revision rate (7.4% vs. 1.2%).

Seating and balance