Anatomic Total Shoulder Arthroplasty

Indication. Anatomic total shoulder arthroplasty (TSA) replaces the arthritic humeral head with a stemmed component and resurfaces the glenoid, relying on a functioning rotator cuff for stability and movement. The primary indication is glenohumeral osteoarthritis with an intact, repairable rotator cuff that has failed non-operative care. Other accepted indications, all with the cuff still functional, include inflammatory arthritis (rheumatoid, seronegative arthropathies), post-traumatic arthritis with preserved bone stock, and humeral head avascular necrosis (steroids, alcohol, sickle cell) with secondary arthritis. Assess the cuff and the glenoid before committing. Rotator cuff integrity is the deciding factor between anatomic and reverse TSA, and glenoid morphology (the Walch classification) dictates how the glenoid is prepared. Exclude or plan for:

• Rotator cuff tear arthropathy — an irreparable cuff is a contraindication to anatomic TSA; a reverse TSA is used instead.
• Severe glenoid bone loss (vault less than 15mm after reaming), active infection, deltoid dysfunction, or a patient who cannot comply with rehabilitation — all contraindications.
• Glenoid version and posterior erosion (Walch B2/B3) on a preoperative CT — these demand version-correcting strategies, not standard concentric reaming. The one decision that matters. Where the cuff is intact an anatomic TSA restores native anatomy; where it is irreparable a reverse TSA is preferred; in a young patient with avascular necrosis and a pristine glenoid, a hemiarthroplasty may suffice. The choice is driven by the cuff, the glenoid and the patient.

Consent specifically for instability (predominantly anterior, linked to subscapularis failure), stiffness, glenoid component loosening, nerve injury (axillary and musculocutaneous), infection, and a lifelong restriction on repetitive heavy overhead lifting for component longevity. Setup. Beach-chair position 30-45 degrees upright, the head secured in a Mayfield or horseshoe headrest to prevent intraoperative movement, and a bump under the medial scapula to improve posterior glenoid exposure. The arm is free-draped for full mobility. General anaesthesia with controlled hypotension reduces bleeding and improves visualisation; an interscalene block provides postoperative analgesia (caution: phrenic nerve palsy in around 5%, so avoid bilateral blocks). Cefazolin 2g IV is given within 60 minutes of incision (vancomycin 15mg/kg if penicillin-allergic or MRSA risk), redosed every 4 hours, and continued for 24 hours.

The goal: expose the shoulder through the deltopectoral internervous interval, take down the subscapularis, release the capsule and dislocate the humeral head atraumatically, then prepare and implant the glenoid and humeral components, restore soft-tissue tension, and repair the subscapularis securely. The exposure is laid out in full below (and in depth on the deltopectoral approach to the shoulder page). Critical anatomy in the field. The deltopectoral interval is a true internervous plane between deltoid (axillary nerve, C5-C6) and pectoralis major (medial and lateral pectoral nerves, C5-T1) — no muscle is divided, so denervation risk is minimal. The cephalic vein runs in the fatty triangle between them and is the landmark for the interval. The subscapularis (upper and lower subscapular nerves, C5-C6) is the primary internal rotator and the key anterior stabiliser; it inserts on the lesser tuberosity and must be taken down and repaired. The axillary nerve crosses the inferior capsule 5-7mm below the inferior glenoid rim and is the structure most at risk, while the musculocutaneous nerve enters the conjoint tendon 3-8cm distal to the coracoid.

View larger

Rehabilitation The protocol is paced to the security of the subscapularis repair — a secure lesser tuberosity osteotomy moves faster than a peel or tenotomy. | Phase | Timing | Immobilisation | Therapy | |-------|--------|-----------------|---------| | 1 | 0-4 weeks (osteotomy) or 0-6 weeks (peel/tenotomy) | Sling; remove for hygiene; sleep in sling | Passive ROM only — pendulums, therapist-assisted elevation and ER; no active motion (no active IR for a peel/tenotomy) | | 2 | 4-6 weeks / 6-12 weeks | Sling for comfort | Active-assisted ROM with stick or pulley; gradual progression | | 3 | 6-12 weeks / 12-18 weeks | Wean sling | Active ROM; gentle strengthening with resistance bands from 0.5kg | | 4 | 12-24 weeks | — | Progressive strengthening; return to swimming or golf from 4-6 months | Most patients return to desk work by 6-8 weeks. Lifelong restriction on repetitive heavy overhead lifting (greater than 25 pounds) protects component longevity. Low-impact activities (swimming, golf, cycling) are encouraged; high-impact and contact activities are discouraged. Follow-up at 2 weeks (wound), 6 and 12 weeks (radiographs and therapy progress), 6 and 12 months, then annually, monitoring for radiolucent lines (greater than 2mm or progressive is concerning), subsidence, osteolysis and heterotopic bone. Complications

Epidemiology and outcomes. Anatomic TSA is performed for glenohumeral arthritis with an intact, repairable rotator cuff, and primary osteoarthritis is the commonest indication — it also gives the most durable result. Inflammatory arthropathies do worse, driven largely by late rotator cuff failure (Khan et al.: up to 75% cuff failure at 10 years in rheumatoid arthritis) rather than glenoid loosening alone. Around 95% of well-selected patients achieve significant pain relief and 85-90% are satisfied in primary OA with an intact cuff; forward elevation typically improves by 40-60 degrees and function plateaus at 12-18 months. Younger, higher-demand patients have consistently higher revision rates across registries. The commonest reasons for revision are glenoid loosening, instability (predominantly anterior, linked to subscapularis failure) and infection. Biomechanics. Anatomic TSA is unconstrained — stability relies on the rotator cuff, capsule and component conformity (concavity-compression), which is why cuff integrity is essential. Eccentric loading generates a rocking-horse moment on the glenoid: maximum forces reach 2-3 times body weight during elevation, oriented superiorly and posteriorly, driving cement-bone micromotion and loosening. Excessive retroversion increases posterior eccentric load. Overstuffing (oversized head or proud osteotomy) causes pain, stiffness and accelerated loosening; understuffing causes instability. Pathoanatomy and the Walch classification. Posterior humeral head subluxation in primary OA drives asymmetric posterior glenoid wear — the basis of the Walch classification, which governs glenoid preparation.

Component choice. On the glenoid side, a cemented all-polyethylene pegged component is the global default; the alternatives are reserved or avoided:

On the humeral side, a press-fit metaphyseal stem is the modern trend in good bone quality (bone-preserving, easier revision), a cemented stem is preferred in osteoporotic or inflammatory bone and for older smooth stems, and a stemless component maximally preserves bone in carefully selected good-quality, non-dysplastic bone. Subscapularis management. The three take-down techniques differ in the healing interface and mobilisation they offer, but the best comparative evidence does not establish one as superior — a secure repair and protected rehabilitation matter more than the specific method.

Registry survivorship and key studies. Across the National Joint Registry (UK), AJRR (US), AOANJRR (Australia), SHAR (Sweden) and NZJR (New Zealand), a cemented all-polyethylene glenoid gives broadly 90-95% 10-year revision-free survival, while metal-backed designs are substantially worse and avoided. The pivotal nuance is that revision survivorship is high but radiographic loosening is common: Raiss et al. found 36% of glenoids radiographically loose and only 48% survivorship at 13 years using radiographic loosening, despite 100% revision-free survival. Walch et al. defined the glenoid morphology classification (A 59%, B 32%, C 9%); Boileau et al. (RCT) showed 20% metal-backed loosening versus 0% all-polyethylene and abandoned the metal-backed design; Louie et al. (systematic review) showed subscapularis tenotomy and osteotomy have similar outcomes. These are summarised in the references below.