Humeral head necrosis after proximal humeral nailing: what are the reasons for bad outcomes?

Introduction: Humeral head necrosis (HHN) remains a major problem in fracture care. Neither its occurrence, its extend, nor its impact on clinical outcomes is predictable on the long term. This study was designed to evaluate clinical and radiological outcomes in patients depending on the influence of HHN.

Patients And Methods: 32 patients with a 3-6 year follow up participated in this study. Their humeral fractures had been stabilized with a standard Targon PH nail (Aesculap, Tuttlingen, Germany) for an acute humeral head fracture. Constant score (CS), DASH score, UCLA shoulder rating scale, and Neer score were assessed. Range of motion (ROM) as well as pain during exercise was documented (VAS). HHN was detected radiologically and graded in stages 0-5.

Results: All fractures had healed. HHN was found in 10 cases (31.3%). 4 patients (12.5%) showed interlocking screw perforation as part of the head collapse caused by HHN. Median CS was 73 (range: 24-85). There was no association detectable between number of fracture fragments and CS (p ≥ 0.631). The median DASH score was 16.4 (range: 0-74.1), UCLA score 30 (range: 9-35), Neer score 80 (range: 29-100). Three (37.5%) of the patients with a stage IV or V osteonecrosis reported about pain (twice VAS grade 4, once VAS grade 5). All patients suffering from pain were affected by high grade HHN and screw perforation. CS was nonsignificantly affected by HHN (75.5 vs. 63.5; p = 0.12), however massively diminished if additional implant protrusion was present (63.5 vs. 25; p = 0.02). Findings for normalised CS, relative CS, DASH score, UCLA shoulder rating scale, Neer score, and ROM were analogous.

Discussion: Whereas HHN itself seems to contribute only mildly to functional outcome, we identified screw protrusion as major predictor for bad clinical results. The high rate of HHN found in our study (31.3%) may be attributed to the inclusion of mild HHN and our long follow-up period, as it is known that late-onset HHN may occur more than 3 years after trauma.

Conclusions: HHN may lead to screw perforation, resulting in poorest outcomes. We recommend regular clinical and radiographic follow-up for at least five years in order to detect impending screw perforation and plan screw removal in time.