The Humeral Bone Loss (HUMBL) classification identifies patterns of humeral bone loss associated with various revision, instability, and humeral loosening rates in primary and revision shoulder arthroplasty.

BACKGROUND

Humeral Bone Loss (HUMBL) is commonly encountered in primary and revision shoulder arthroplasty and can lead to complications including humeral loosening or instability. The HUMBL classification categorizes bone loss into 5 major patterns of increasing severity: 1) Intact/No substantial bone loss (including bone compromised by prior implants/hardware or abnormal bone quality) 2) isolated greater tuberosity loss or compromise (including acute fracture, nonunion and malunion) 3) segmental bone loss (bone loss distal to the surgical neck) 4) ectatic bone loss (expansion of the cortex compared to an intact segment) and 5) humeral discontinuity (separation of the humeral shaft from a prior implant or where a new implant would be placed). The purpose of this study was to describe patterns of bone loss in primary and revision shoulder arthroplasty and stratify outcomes based on bone loss pattern.

METHODS

This retrospective review identified shoulders with HUMBL undergoing primary or revision shoulder arthroplasty between 2012 and 2022. There were 1,083 shoulders included. Patterns of bone loss were identified with preoperative radiographs according to the HUMBL classification. When patients had a combination of bone loss, they were grouped with the most severe pattern present. Four hundred sixty-nine shoulders (43%) had no substantial bone loss, 114 (11%) had isolated greater tuberosity loss or compromise, 361 (33%) had segmental bone loss, 84 (8%) had ectatic bone loss and 55 (5%) had humeral discontinuity. Primary outcomes were all-cause reoperation, instability, and humeral loosening.

RESULTS

The all-cause revision rate was 16%. Compared to bone intact shoulders (9% revision rate), shoulders with segmental (21%, odds ratio (OR) 2.8, P < .001), ectatic (38%, OR 6.6, P < .001), or humeral discontinuity bone loss (27%, OR 4.0, P < .001) had higher revision rates. The overall rate of post-operative instability was 11%. The rate of instability was higher in shoulders with compromised greater tuberosity (7%, OR 2.5 P = .049), segmental (15%, OR 5.6, P < .001), ectatic (35%, OR 17.1, P < .001) or humeral discontinuity (25%, OR 11.1, P < .001). The overall rate of humeral loosening was 6% and was higher in shoulders that had segmental (9%, OR 15.1, P < .001), ectatic (21%, OR 42.4, P < .001) or humeral discontinuity bone loss (16%, OR 30.4, P < .001) (Table 2).

CONCLUSION

The HUMBL classification demonstrates that certain patterns of bone loss are associated with increased rates of reoperation, instability, and humeral loosening. The presence of ectatic bone loss, not previously described in shoulder arthroplasty, led to high rates of revision, instability, and humeral loosening.