2017 Sec 1 Green Book

Volume 137, Number 4 • Management of Orbital Fractures

injuries, contributing to a paradoxically elevated rate of surgical intervention in orbital fracture patients with low energy mechanisms. Losee et al. reported a series of 74 pediat- ric orbital fracture patients, 25 of which were isolated orbital fractures. 2 Only three of the 25 patients (12 percent) underwent surgical treat- ment, two of which were because of entrapment. Mild detectable enophthalmos was seen in one of the surgically treated fractures (33 percent) and three of the conservatively treated fractures (14 percent). However, among “significant fractures” ( n  = 12), defined as fractures that involved more than 50 percent of the area or displacement mea- suring greater than three times the cortex width, the enophthalmos rate was 30 percent with con- servative treatment. The authors questioned the practical significance of mild enophthalmos and advocated a conservative approach to manage- ment in the absence of entrapment or early globe malposition. This is clearly an area where surgeon preference and informed consent by parents as to the risks and benefits of intervention versus obser- vation is essential. Attempts to correlate radiographic fracture size with the development of late complications began shortly after the widespread adoption of computed tomographic scanning. In 1983, Hawes and Dortzbach postulated that a greater than 50 percent defect of the orbital floor on computed tomographic scan was necessary to cause enoph- thalmos. 11 This number has been widely adopted as a standard for “critical fracture size” in the ophthalmology literature. For example, Hatton et al. published a series of 96 pediatric patients, inclusive of multiple facial fracture patterns, from the Massachusetts Eye and Ear Hospital following this criterion. 12 Forty-nine patients (51 percent) underwent surgery; however, only four patients (4 percent) had surgery based on the size of the fracture. Bansagi and Meyer published a review of a 34-patient experience, of which eight were trap- door fractures. 13 They noted that this subset of patients had better recovery of ocular motility if operated on immediately (within 48 hours) ver- sus later (3 to 14 days after injury). Specifically, these patients demonstrated long-term limitation in supraduction. In contrast, Egbert et al. 14 con- cluded that delay of up to 1 week was acceptable. In our experience, these patients tend to do bet- ter the earlier they undergo reduction. Accord- ingly, our preference is to perform surgery for this subgroup within 24 hours of presentation when- ever possible.

Despite this, there is little in the way of consensus regarding the full breadth of indications that war- rant surgical repair. At least four different surgical specialties manage these injuries, and recommen- dations for treatment vary across the literature. The greatest areas of disagreement focus on (1) optimal timing of surgery, for both entrapped and nonentrapped fractures; (2) ophthalmologic symptoms necessitating intervention; and (3) frac- ture characteristics that warrant early intervention before the development of globe malposition. We sought to specifically answer the question of surgical intervention in orbital fractures with an emphasis on the outcomes of surgical versus conservative treatment. Although it represents the largest series of isolated pediatric orbital fractures and the largest operative series in the literature, our cohort is also unique in that we excluded patients with any other type of facial fracture or injury that could potentially confound the decision to operate. Thus, patients were strat- ified between surgery and conservative manage- ment cohorts based on their clinical indications. Both cohorts also included only patients who returned after their injury, with nearly 80 per- cent of patients completing their recommended follow-up visits and released to follow-up on an as-needed basis. Limitations of our study include the potential for bias by excluding patients who did not return for follow-up. Our results may also have limited generalizability to high-energy cra- niofacial trauma with multiple concomitant facial fractures. We previously reported a large cohort of pedi- atric orbital roof fractures, with a significant rate of intracranial injury and multisystem trauma. 10 These patients (who were not included in this series) were admitted to the Johns Hopkins Pedi- atric Trauma Center and had high rates of multi- ple trauma by means of high-energy mechanisms. In contrast, the patients in this series were man- aged through the state-designated Wilmer Eye Institute Ocular and Orbital Trauma Center and were, by inclusion criteria, isolated injuries. This is congruent with the low incidence of roof frac- tures seen in this series. Examination of the demo- graphics and fracture patterns of the two groups offers an interesting comparison between low- and high- energy facial injuries in the pediatric popu- lation. Unlike the high rate of operative repair in this series, less than 10 percent of patients with orbital roof fractures had surgery, 10 and their management is typically more analogous to a skull fracture than an orbital floor fracture. Trapdoor fractures are much less common in high-energy

27