Heterotopic ossification following primary reverse total shoulder arthroplasty

The available data on heterotopic ossification (HO) following implantation of reverse total shoulder arthroplasty (rTSA) is limited. It was the aim of this study to determine occurrence of HO following rTSA and the impact on clinical outcome. A prospectively followed rTSA database of a tertiary referral center was retrospectively analyzed between September 2009 and February 2018 to evaluate the prevalence of HO in primary rTSA cases at minimum follow-up of 2 years. All basic demographic data, clinical outcomes (Constant-Murley score (CS), Subjective Shoulder Value, range of motion), complications and reinterventions were collected. HO was classified as type 1a (islands of bone within the soft tissue), type 1b (bone spurs on the inferior scapular neck), type 1c (incomplete ankylosis), type 2 (combination of bone spur and islands of bone), and type 3 (complete ankylosis). Further, an association between complications/reinterventions and occurrence of HO was evaluated. A total of 663 rTSAs (627 patients, mean age 72 &#xb1; 9 years, 60% female) were available for analysis at a mean follow-up of 6 &#xb1; 4 years. Of those, 16% (104/663) had at least one complication and 8% (53/663) needed a reintervention. At 2 years, HO was seen in 59% (392 of 663) of the cases with type 1a in 108 (16%), type 1b in 226 (34%), type 1c in 19 (2.9%), and type 2 in 39 (5.9%). Overall, occurrence of HO had minimal impact on absolute CS with values of 68 &#xb1; 15 (no HO), 70 &#xb1; 13 (type 1a), 67 &#xb1; 13 (type 1b), 64 &#xb1; 13 (type 1c), and 68 &#xb1; 14 (type 2) (P = .043). The most relevant effect was seen on abduction with 117 &#xb1; 35&#xb0; in type 1c vs. 136 &#xb1; 30&#xb0; without HO (P = .02). The effect on other functional parameters was not statistically significant with a tendency for inferior results in type 1c. Further risk factor analysis revealed that complications and reinterventions were not associated with occurrence of HO (P = .443 and P = .930) but female gender (coefficient -0.37, P < .001), moderate alcohol consumption (coefficient 0.35, P = .001), secondary fracture treatment (coefficient 2.96, P < .001), and osteoarthritis (-0.48, P < .001). The occurrence of HO is very common with progress over time. Male gender, moderate alcohol consumption, and secondary fracture treatment seem to be associated with higher HO rates. It affects mainly abduction and to some extent the overall clinical outcome, with inferior results especially for type 1c. There was no association with other complications and revision surgery.