A recent article, Long-term functional and radiographic outcomes of anatomic total shoulder arthroplasty using an all-polyethylene cemented glenoid component with a minimum follow-up of 10 years, reported the minimum10 year outcomes for 54 of 129 patients (mean age 78, 55% female) having anatomic shoulder arthroplasty by an individual highly experienced surgeon between 2002 and 2014. Four different implants were used: Global Unite, Global AP, Global Advantage, and the Total Evolutive Shoulder System.
From the authors: "A cemented onlay all-polyethylene pegged glenoid component was implanted, with reaming performed to achieve an inlay-style placement while ensuring osseous support at its borders. To avoid overstuffing, the humeral resection and the selection of the humeral head size were performed according to the native anatomy and tension. The humeral stem was implanted using a press-fit technique when bone quality allowed; otherwise, it was cemented."
Patients with minimum 10 year followup had favorable shoulder function, with a Constant Score of 63, an American Shoulder and Elbow Surgeons score of 83, a Subjective Shoulder Value of 80, and a pain score of 0.
Patients with minimum 10 year followup had forward elevation averaging 160 degrees and internal rotation averaged reaching to T12.
Kaplan-Meier analysis found an implant survival of
89.0% for those patients with 10 year followup.
14 patients (10.4%) underwent a revision after a median of 2.8 years (before the 10 year clinical outcome inclusion mark). Indications for these revisions were instability in 2, overstuffed humerus in 2, cuff failure in 3, infection in 1, humeral loosening in 2, and glenoid loosening in 4. The only risk factor reported for revision was age below 65 yrs.
Complications not requiring revision included arthrofibrosis (2), infection, supraspinatus tear, periprosthetic fracture, scapular fracture, and ulnar nerve palsy. The overall complication rate was 16.4%.
So, the question to ask is :what might have been done differently to avoid these complications?"
(1) Two revisions were performed within the first 6 months after surgery for "overstuffing". As we've discussed in prior posts, overstuffing is avoided by making a proper head cut, avoiding varus or incomplete insertion of the stem, and - most importantly - adjusting the head thickness so that, with the trial components in place, the shoulder has at least 150 degrees of flexion, internal rotation of 60 degrees with the arm in 90 degrees of abduction, 50% posterior translation, and external rotation of 40 degrees with the subcapularis approximated to its insertion site.
2. Two early revisions were required for instability. It is not clear whether in these cases the instability was anterior or posterior. Excessive posterior translation can be identified at surgery and managed by an anteriorly eccentric humeral head (as detailed in prior posts). Anterior instability is often a result of failure of the subscapularis reattachment. This risk can be minimized by a careful subscapularis peel, preserving the capsule on the tendon's deep surface, a complete 360 degree release of the muscle/tendon, avoiding overstuffing, a robust repair to the lesser tuberosity with 6 sutures with an addition suture or two in the lateral rotator interval to reinforce the repair, and avoiding unintended stretching in excessive external rotation.
3. Four revisions were for glenoid compoinent loosening. The risk of this complication can be minimized by using either an all polyethylene glenoid component with ingrowth pegs or a hybrid cage design rather than the all-cemented peg design used in this study and by assuring excellent seating of the component in well-prepared glenoid bone.
It would be helpful to have clarification on a few issues:
(1) What was the rationale for chosing one of the four different humeral components?
(2) What were the technique and implant factors associated with humeral component loosening and periprosthetic fracture?
One other interesting challenge with long term studies such as this one is what's called "immortal time bias". It occurs when there is a period of time in the followup during which the clinical outcome cannot be known because of uncontrollable factors, such as the death of the patient or revision. In this paper there was a 12 year inclusion period (2002-2014) with a required minimum 10 year followup. Patients who died early in the followup period or who had early revision were excluded from the clinical and functional analysis; the patients that died or had early revision would seem more likely to have increased risk of adverse outcomes, but they were removed from the study. In this case 48 patients died and 14 patients had revision before the 10-year followup creating a study cohort of 42% of the initial population that is enriched with cases that are more likely to be successful. This can artifically inflate survival estimates and treatment effects, biasing the results toward favorable outcomes.
The authors appropriately acknowledged this limitation, noting that "due to the high number of deceased patients, a selection bias is inherent" and that "the revision rate may be underestimated due to the possibility that patients who were lost to follow-up may have undergone revision elsewhere."
Clinical research, especially the pursuit of long-term followup is difficult, yet only through long term followup can we answer the patient's question: "how long is this arthroplasty likely to hold up?"
How long?
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133/
Shoulder arthritis - what you need to know (see this link).
How to x-ray the shoulder (see this link).
The ream and run procedure (see this link)
The total shoulder arthroplasty (see this link)
The cuff tear arthropathy arthroplasty (see this link).
The reverse total shoulder arthroplasty (see this link).
The smooth and move procedure for irreparable rotator cuff tears (see this link)