Butterfly fragments in diaphyseal tibial fractures heal unpredictably: Should we be adopting other strategies for these high risk fractures?

The natural history of diaphyseal tibial butterfly fragments is poorly documented. Numerous studies have analyzed risk factors for nonunions in the tibial shaft with known factors including Gustilo classification, ASA class, and cortical contact. However, the healing potential and ideal management of nonsegmental butterfly fragments in this setting remains unknown. The aim of this study was to determine the nonunion rate of diaphyseal tibial fractures with a butterfly fragment. A performed a retrospective review of patients at a single academic Level 1 Trauma Center from 2000-2020 who underwent intramedullary nailing of tibial shaft fractures. Those with non-segmental butterfly fragments (OTA/AO: 42-B) and minimum 12 month follow up were included. Morphologic measurements of butterfly fragments were performed to measure location, size, and displacement, and mRust scores at final follow up were calculated. Outcome measures were surgery to promote union, and mRust scores. A total of 99 patients were included with 21 patients requiring revision surgery to promote union. Thirty six patients had open fractures and 77% of patients were male with a mean age of 34 (range: 12-80). Average follow up was 19 months (3 months – 12 years). The most common location of the butterfly fragment was the anterior cortex (42%), with a mean length of 7.8cm (SD: 3.3) and width of 1.8cm (SD: 0.5cm). At final follow-up 37% of fractures had persistent lucency without callus at the site of the butterfly while only 31% of fractures had remodeled cortex. Average time to complete healing was 13.3 months. Open fractures with butterfly fragments were more likely to go on to nonunion than closed (44% vs 9.2%, p=