The bicipital groove as a landmark for humeral version reference during shoulder arthroplasty: a computed tomography study of normal humeral rotation

Accurate reproduction of humeral version is important in shoulder arthroplasty. Traditional referencing relative to the transepicondylar axis (TEA) is prone to error as it is absent on preoperative imaging and inaccurately reproduced intraoperatively. The bicipital groove is present on preoperative imaging and in the operative field and thus may be a useful landmark for accurate reproduction of humeral version. Two trained observers analyzed 101 full-humerus computed tomography scans of patients undergoing a myeloma screening protocol. Measurements of humeral retroversion relative to the TEA (angle A), humeral articular axis retroversion relative to the bicipital groove (angle B), and the bicipital groove axis relative to the TEA (angle C) were made with comparison of the measurement properties of each. Humeral retroversion relative to the TEA was 23.7° ± 8° (range, 0.2°-48.7°; 95% confidence interval, 22°-26°). The humeral articular axis was retroverted to the bicipital groove axis (angle B) by 33.5° ± 9.4° (range, 15.5°-61.7°; 95% confidence interval, 32°-35°). Overall inter-rater reliability was 0.88. Measurement of humeral head retroversion relative to the bicipital groove is not inferior to the gold-standard measurement. The bicipital groove is present both on preoperative imaging and in the operative field, making it a potential reference landmark for accurate reproduction of humeral version in shoulder arthroplasty.