Blindness and severe visual impairment in facial fractures: an 11 year review

Transcription

1 British Journal of Plastic Surgery (2002), 55, The British Association of Plastic Surgeons doi: Pojps BRITISH JOURNAL OF [ ~ ] PLASTIC SURGERY Blindness and severe visual impairment in facial fractures: an 11 year review C. A. MacKinnon, D. J. David and R. D. Cooter* Australian Craniofacial Unit, Women and Children's Hospital, North Adelaide, and *Department of Plastic and Reconstructive Surgery, Royal Adelaide Hospital, Adelaide, South Australia, Australia SUMMARY. We present an 11 year review of facial fractures treated in the Australian Craniofacial Unit and the Department of Plastic and Reconstructive Surgery at the Royal Adelaide Hospital, specifically looking at those fractures in adults that resulted in blindness or severe visual impairment. During the period , a total of 2516 patients with facial fractures were treated operatively. From our facial-fracture database, 317 patients were identified as having an ophthalmological complication on presentation. Of these, 19 had severe visual impairment or blindness. The exact fracture pattern and craniofacial disruption score were recorded for each of these 19 cases using the Cooter-David facial-fracture scoring system. Laterally directed forces are implied as major causative factors in these injuries. Recommendations are made based on these findings The British Association of Plastic Surgeons Keywords: facial fracture, blindness, visual impairment. Blindness following facial fracture is rare, and its management is controversial. 13 The actual incidence of severe visual impairment is difficult to ascertain, as there are very few reports with large series of patients. 1,4 Using our database of over 2500 patients with facial fractures, we felt that it was timely to review the incidence of blindness and severe visual impairment in our series. The aim of this review was to examine the data for any correlation between fracture pattern or mechanism and visual outcome. Patients and methods All adult patients with facial fractures treated operatively by members of the Australian Craniofacial Unit and the Department of Plastic and Reconstructive Surgery at the Royal Adelaide Hospital over the 11 year period were included in the study. The details of all patients with facial fractures are encoded into a standardised database on admission, and ongoing management information is recorded perioperatively and through to long-term follow-up. The database records demographic details, history and initial examination findings, the results of radiological and other investigations, alphanumeric coding of the fracture using the Cooter-David facial-fracture scoring system (CDFSS), 5 operative findings, a detailed analysis of the operative procedure, and short, intermediate and long-term follow-up findings. The information is initially recorded by the junior surgical staff, with the operating surgeon providing the operative and follow-up details. Apart from those with minor fractures, all patients have a CT scan of the facial bones preoperatively. Included in this retrospective analysis were all adult patients managed between May 1989 and May A total of 2516 patients were identified. Using the database, 317 patients (13%) were documented as having an ophthalmological complication identified on admission. The database has fields in the history and examination sections where ophthalmological symptoms and signs can be recorded. Keyword searches were applied to other fields throughout the database to identify further ophthalmological symptoms. All of the 317 charts were manually reviewed to identify those patients with severe visual impairment or blindness. Blindness was defined as a visual acuity of 6/60 or worse, or a visual field of 10 ~ or less in the affected eye, or a destroyed globe. Severe visual impairment was defined as a visual acuity of worse than 6/18 in the affected eye. This level was chosen because it is a functionally disabling level of acuity. A useful correlate is that to obtain a new driver's licence in South Australia, one must have a visual acuity of no worse than 6/12 in either eye. For those with an existing driver's licence, one must have a visual acuity of no worse than 6/18 in the best-seeing eye. Hence, a visual acuity of less than 6/18 in one eye severely impairs the patient's ability to drive a motor vehicle, and, after discussion with our ophthalmological colleagues, we felt that this figure represented a reasonable objective measurement for severe visual impairment. All 317 patients with ophthalmological complications were assessed preoperatively and postoperatively by an ophthalmologist, and visual acuity was routinely recorded. A best postoperative visual acuity of less than 6/18 within a 6 month period was defined as severe visual loss. Other ocular disorders, such as diplopia, epiphora, conjunctival haemorrhage, etc., were recorded but not considered further in this study. Review of these 317 cases revealed 19 patients with severe visual loss or blindness (6.0%, or 0.8% of the 2516 patients with facial fractures in this study).

2 2 British Journal of Plastic Surgery The charts for these 19 patients were scrutinised, and the following information was recorded: name, identification number, gender, age, aetiology of injury, type of fracture, alphanumeric code of fracture, craniofacial disruption score, surgical treatment performed, presence or absence of associated injury, Glasgow Coma Score, CT findings, ophthalmological diagnosis, preoperative and postoperative visual acuities, and whether or not any ophthalmological treatment was performed. In addition, the CT scans of 15 out of the 19 patients were reviewed (the other four had been destroyed or reported missing), and a detailed analysis of each fractured bone was recorded for each patient. These were recorded using the minor codes of the CDFSS? The craniofacial disruption score was calculated as per the modification of the CDFSS described by O'Sullivan et al. 6 Briefly, this involves assigning a score to each bone or suture disrupted, depending on the severity of the injury, and totalling them as a percentage. All operations were carried out by consultant craniofacial or plastic surgeons, visiting craniofacial fellows, or advanced trainees in plastic and reconstructive surgery (under supervision). Our principles of operative management include early intervention, wide subperiosteal exposure of all fracture fragments, open reduction and internal fixation using low-profile titanium plates and screws, primary bone grafting where required, and avoidance of long-term intermaxillary fixation. Results Over the 11 year period, 2516 adult facial-fracture patients required surgical management. Of these, 317 (13%) had a recorded ophthalmological complication, of which 19 had severe visual impairment or blindness, as defined previously. There were 15 males and four females, giving a ratio of 3.75:1. The age range was years (mean: 39.8 years). The mean age for males was 43.0 years, and for females it was 27.8 years. Motor-vehicle accidents were the most common aetiology, accounting for nine cases. There were three assaults, three falls, one horse kick to the face, one aeroplane crash, one golf club to the face, and a tree fell on one patient. Eight of the 19 patients suffered multiple injuries in addition to their facial fractures, six suffered from other craniofacial injuries alone, and five had no other injuries (Table 1). The Glasgow Coma Scores were recorded on admission, and most patients had a score of 15 (Fig. 1). Craniofacial disruption scores ranged from 1 to 62 (mean: 15; median: 13). There were three patients with a craniofacial disruption score of 1 (Table 2). Two of these had suffered low-energy blunt injuries (one assault, one struck by a golf club), and each suffered an isolated displaced orbital-floor fracture. The third patient suffered an undisplaced fracture of the anterior wall of the left frontal sinus in a high-speed motor-vehicle accident. Unfortunately, he also had a penetrating eye injury, and, despite scleral repair, he achieved a postoperative visual acuity of only 6/36. The mean craniofacial disruption score was highest in those patients sustaining their injuries in motor-vehicle # 2 o t GC$ Figure 1---The Glasgow Coma Scores of the patients on admission. Table 1 Associated injuries Case Fracture Associated injuries 1 right orbital floor 2 naso-orbito-ethmoid, bilateral zygomas 3 right orbitozygoma, fight mandible 4 left zygoma 5 bilateral condyles, fight angle mandible 6 left orbitozygoma 7 left orbitozygoma 8 mandible 9 left fronto-orbitoethmoid 10 left frontal bone, left zygoma, left maxilla 11 right orbitozygoma, maxilla 12 frontal sinus 13 nasomaxillary, orbital rim 14 LeFort II, bilateral mandible 15 LeFort III 16 left orbital rim 17 LeFort I and III, right mandible 18 naso-orbito-ethmoid, right orbitozygoma 19 mandible, supraorbital bar fractured left tibia and bilateral fibulae, dislocated left humerus and elbow fractured fight humerus and fibula, fractured ribs and dislocated foot bilateral hearing loss and head lacerations left parietal haemorrhage and bilateral wrist fractures fractured ribs and C5, C6 and C7 penetrating facial and eye injury facial lacerations left orbital abscess skull fractures, bilateral subdural haematoma, fractured ribs and clavicles, and ear lacerations fractured C5 and facial lacerations fractured skull base and cerebral contusions fracture dislocation of the left shoulder, fractured ribs, tension pneumothorax fractured right femur, right tibia and right humerus, and cerebral contusions fractured skull base, cerebral contusion, right VIIth cranial nerve palsy, pulmonary contusion

3 Blindness and severe visual impairment in facial fractures 3 Table 2 CT findings and Craniofacial Disruption Score (CDS) Case CT findings CDS blowout fracture of the right orbital floor posteromedially, no canal fracture 1 fractures of the nasoethmoid, left lateral orbit, bilateral zygomas and anterior walls of the maxillae 28 fractures of the right zygoma, right maxilla, right orbital floor and right pterygoid plate 13 fractures of the left zygoma and posterolateral orbit, optic nerve intact, no canal fracture, anterolateral maxilla fracture on left 16 fractured medial orbital walls, bilateral fracture of the mandibular condyles, right angle and symphysis fractures 5 depressed fracture of the left antrum anterior wall, fractured nose, fracture of the left orbital floor inferomedially 3 fracture of the left zygoma and orbital floor 1 right temporal bone fracture into anterior cranial fossa, through lesser wing sphenoid, fractured right zygoma, right subcondylar and left parasymphyseal fractures 7 left frontal bone, orbital roof and ethmoid fractures, left optic canal fracture, left zygomatic arch and lateral wall antrum fractures 8 depressed frontal fracture, bilateral superior orbital rim and orbital roof fractures, left medial orbital fracture, bilateral maxillary fractures, parietal bone fracture, minimally displaced 62 multiple facial fractures, including bilateral orbital lateral walls, three walls right antrum, nasal bones, right zygomatic arch and roof of the right orbit 19 fracture of the anterior wall frontal sinus 1 fractures of the left orbital medial wall, left inferior orbital rim and left nasal bone 4 depressed left frontotemporal fracture and fractures of the right frontal bone and sinus, nasal bones, right orbital medial wall and left body of the mandible 31 naso-orbito-ethmoid fractures, bilateral midface fracture 23 fractured left zygoma, right temporal bone to petrous part including mastoids, bilateral sphenoid, lateral wall of left orbit, left pterygoid plate and multiple left zygomatic fractures 14 bilateral LeFort! and III fractures, right parasymphyseal fracture 30 fractured right orbitozygomatic complex and right frontal bone into sinus, no canal fracture 16 fractures of the floor of the anterior cranial fossa, right frontal bar, right temporal bone to external auditory canal, right clinoid, lateral wall of the right orbit, right zygomatic arch into temporomandibular joint CDS 10 0 Assault MVA Fall Other Aetiology CDS Multitrauma Craniofacial Nil Associated Injury Figure 3---The Craniofacial Disruption Scores of the patients according to the presence of associated injuries. Figure 2--The Craniofacial Disruption Scores of the patients according to aetiology. accidents, and in those who suffered multitrauma (Figs 2 and 3). In the 19 patients, there were a total of 148 fractures: 46 orbital (31%), 35 zygomatic (24%), 30 nasomaxillary (20%), 27 maxillary (18%) and 10 mandibular (7%) (Fig. 4). None of the mandibular fractures occurred in isolation. Each of the major fracture zones have been broken down into their individual minor zones, to give an indication of the numbers of individual bones fractured (Fig. 5A-E). The ophthalmological diagnoses were commonly traumatic optic neuropathy or globe rupture. The diagnoses of optic neuropathy usually received no specific Zygomatic 24% Nasomaxillary 20% Maxillary 18% Orbit Mandibular 31% 7% Figure 4~Facial fractures according to the major zones.

4 4 British Journal of Plastic Surgery Or~- - fl< 2! Inferior orbital rim 20% ial wall 3% Zygornatic arch 31% Orbital roof 9% Lateral wall A 29*. orbital rim 9% s~um B ~5~ Nasomaxillary Antedor ~hmoid Anterior wall Nasofrontal suture 13% ;e Jar bone C 37,/. exnmola 10% D 51% Pterygoid 4% plate 4% Mandibular symphysis Man~ ~dibular cor ~ody 50% 20% E Figure 5--Facial fractures according to the minor zones: (A) orbital major zone; (B) zygomatic major zone; (C) nasomaxillary major zone; (D) maxillary major zone; and (E) mandibular major zone. ophthalmological treatment. Ruptured globes underwent attempted surgical repair, enucleation or evisceration (Table 3). Follow-up lasted for 3 years postoperatively, depending on the type of fracture, complications and patient compliance. Ophthalmological follow-up OCCUlTed in conjunction with craniofacial follow-up, at a separate clinic. Discussion In presenting a study of 2516 adult facial-fracture patients, we acknowledge that a retrospective review of a database is fraught with problems: the database is totally dependent on the enthusiasm and accuracy with which it is recorded. Some patients will be unrecorded, some patients will have been lost to follow-up, or follow-up will be irregular. Various components of the database will have errors, due to inaccurate recording, inter-observer discrepancies and computer input error. Despite these acknowledged shortcomings, a sample size of 2516 patients is significant and compares well with other published series.1'4,6-8 We found that 317 out of the 2516 patients (13%) had an ocular complication. Other reports have noted an incidence ranging from 3.9% to 90.6%! 4,7-9 This extreme

5 Blindness and severe visual impairment in facial fractures Table 3 Ophthalmological findings and treatment Case Preoperative Postoperative Diagnosis Treatment visual acuity visual acuity 1 nil 5/60 retinal haemorrhage and optic nerve contusion nil 2 nil 6/24 ruptured globe scleral repair 3 nil 6/60 traumatic optic neuropathy and abducens palsy nil 4 nil nil traumatic optic neuropathy nil 5 light only 6/12, upper half traumatic optic neuropathy, 90% visual loss nil 6 6/36 6/60 traumatic macular hole nil 7 6/36 6/24 made nil 8 1/60 1/60 Purtcher's ischaemic retinopathy (traumatic optic neuropathy) nil 9 nil nil traumatic optic neuropathy secondary to optic canal fracture dexamethasone and prednisolone 10 nil nil ruptured left globe enucleation 11 count fingers nil ruptured right globe scleral repair, euucleation secondarily 12 count fingers 6/36 penetrating left eye injury scleral repair 13 nil nil retro-orbital abscess drainage of abscess 14 count fingers 1/60 left posterior pole haemorrhage, optic nerve infarct nil 15 nil nil ruptured left globe enucleation 16 nil 6/60 left optic atrophy (traumatic optic neuropathy) nil 17 nil nil ruptured left globe evisceration 18 nil 1/60 vitreoretinal haemorrhage nil 19 nil nil traumatic optic neuropathy nil range of incidence is due to the varied methods of reporting. Some prospective reports may yield a higher incidence of ocular complications, as suggested by Jabaley et al, who compared retrospective and prospective analyses of ocular injuries in orbital fractures, and showed a difference in incidence of 11% versus 29% in retrospective and prospective analyses, respectively. 1~ Other studies have been very selective in their cases. A1-Qurainy et al, who reported an incidence of 90.6%, selected mid-facial fractures only, excluding nasal-bone and mandibular fractures, and neurosurgical patients. 7 We felt that a much more useful approach would be to assess all facial fractures regardless of aetiology, site of fracture or associated injury, as this is how they present daily to a facial-trauma service. In our review, the incidence of blindness or severe visual impairment was 0.8% (19 patients of the 2516 in the study group). This compares with a previously published range of between 0.2% and 6%. t'7-9'11'12 Our male-to-female ratio was 3.5:1. This probably reflects the higher number of males involved in motor-vehicle accidents and assaultive behaviour. The mean age was higher for males (43.0 years) than for females (27.8 years); we do not have any explanation for this discrepancy. Motor-vehicle accident was the most common aetiology (47%), and 14 of the 19 patients (74%) suffered other injuries. These findings reflect the higher energies imparted to the facial skeleton in these patients, and one would expect a higher degree of damage, including an increased risk of ocular trauma. Analysis of the Glasgow Coma Scores revealed that most patients were admitted to hospital with a score of 15, implying a lack of serious brain injury in this group. This supports Le Fort's hypothesis that the process of facial fracturing absorbs energy, thereby preventing further transmission of traumatic forces to the brain. 13 The Craniofacial Disruption Scores ranged from 1 to 62, with a mean of 15. Because this study included all facial fractures, some of the patients with very low fracture scores (and therefore relatively insignificant fractures) had to be included. Two of the patients with a score of 1 had isolated displaced orbital-floor (blowout) fractures. The other patient with a score of 1 had a minimally displaced fracture of the anterior wall of the frontal sinus, but suffered a concurrent penetrating eye injury. This shows that even a relatively insignificant facial fracture can be associated with loss of vision, and the clinician must be alert to this possibility. With such a range of Craniofacial Disruption Scores represented, we were unable clinically to correlate the score with the occurrence of blindness or severe visual disturbance. The mean Craniofacial Disruption Score was highest in patients involved in motor-vehicle accidents, and in those who suffered multitrauma. Again, this would be expected due to the energy imparted to the facial skeleton and body during impact, compared with that imparted in an assault or fall. There were a total of 148 fractures in the 19 patients. This total includes sutural disruption, as per the CDFSS. 5 Sutural disruption is an integral part of this scoring system, as it allows a better quantification of the degree of force, and hence disruption, applied to the facial skeleton. Over half of the fractures (55%) were of the orbitozygomatic region. Analysis of the orbital major zone revealed the lateral orbital wall as the most frequently fractured region (29%). Along with the zygomaticomaxillary buttress, these were the most commonly fractured areas in the 19 patients (14 fractures in each zone). In the zygomatic major zone 35% of the fractures involved the zygomaticofrontal suture, and 31% were fractures of the arch. Hence, 66% of the zygomatic fractures resulted from lateral impacts, and this, combined with the high incidence of lateral-orbital-wall and zygomaticomaxillary (lateral) buttress fractures, implies a laterally directed force as a significant causative factor in these cases of blindness (Figs 6 and 7). Analysis of the maxillary major zone revealed over half (51%) of the fractures involved the zygomaticomaxillary buttress. Analysis of the nasomaxillary zone showed that 37% of the fractures involved the nasal bones. This is not surprising as the nasal bones are the most commonly

6 6 British Journal of Plastic Surgery Figure ~--Preoperative axial CT scan showing comminuted fractures of the right lateral and medial orbital walls. This is an 18-year-old woman who was involved in a motor-vehicle accident as a rear-seat passenger (case 18). She suffered multiple injuries; her visual loss was due to a posterior pole haemorrhage. No visual recovery occurred. Figure 7--Preoperative axial CT scan showing a comminuted fracture of the left lateral orbital wall and a nasoethmoidal fracture. This is an 81-year-old man who was involved in a motor-vehicle accident as a driver (case 15). He suffered LeFort III, C5 and nasoethmoidal fractures. Visual loss was due to a ruptured globe; he underwent enucleation, fractured facial bones, and cadaver studies have shown that nasal fractures occur at impact energies well below those of the other facial bones) 4 Analysis of the mandibular zone revealed that 50% of the fractures involved the condyle, although the total number of mandibular fractures was low. An association between lateral-orbital-wall fractures and blindness has been previously noted. 2,15 Jend and Jend-Rossmann reported sphenotemporal buttress fracturing in five cases, leading to amaurosis in allj 5 Their particular fracture pattern has certainly appeared in our series, and their suggestion that the radiologist and clinician pay particular attention to the presence of this fracture pattern is well noted. These findings contrast with those of Ioannides et al, who analysed 509 fractures over a 9 year period, s They had a total of 15 patients with loss of vision (2.9%), and concluded that patients with nasoethmoidoffontal fractures had a greater chance of being blinded than patients with other fractures. However, their study only assessed fractures involving the orbit. They also report a 26% incidence of ocular injury, yet exclude subconjunctival haemorrhage, corneal abrasion and retinal oedema from this group. The most frequent ophthalmological diagnosis in our series was traumatic optic neuropathy. There were seven patients with this diagnosis, five with a ruptured globe and three with retinal haemorrhage. The incidence of traumatic optic neuropathy following facial injury has been reported as 2%-3%, but it has been suggested that this incidence can be up to 15% if objective pupillary examination is undertaken. 9 Zachariades et al, in their 12 year review of 5936 patients, noted that, of their 19 blinded patients, retrobulbar haemorrhage was the most common cause of blindness. 1 In A1-Qurainy et al's series of selected mid-face fractures traumatic optic neuropathy was the cause of blindness in each instance. 7 The accuracy of the ophthalmological diagnoses will obviously differ from unit to unit and depend on the degree of initial ophthalmological input. Our findings support A1-Qurainy et al, in that traumatic neuropathy is more commonly found as the cause of blindness. Our second most common cause was ruptured globe. We had only one patient with an optic canal fracture (case 9). The patients with a diagnosis of traumatic optic neuropathy usually had no vision or worse than 6/60 visual acuity immediately after injury. In our series, the only one of these patients to be actively treated for his blindness was the patient in whom an optic canal fracture was diagnosed; he was treated with a standard course of highdose dexamethasone intravenously followed by a tapered course of prednisolone. The most common mechanisms thought to be the cause of optic neuropathy are haemorrhage into the optic-nerve sheath secondary to a shear effect, or contusion of the nerve with associated oedema and compression. This oedema would lead to secondary compromise of the microvascular supply to the nerve. 12 The use of steroid treatment is thought to decrease the microcirculatory spasm and oedema, in the hope of preventing nerve-cell compromise. However, it is felt that blindness occurring at the moment of injury has a very poor prognosis, 2 and, as it has been shown that irreversible ischaemic injury to the optic nerve may occur within 1-2h of injury, instituting a high-dose steroid regimen may well be futile unless the treatment is commenced immediately. The single case that we treated medically had an optic canal fracture (case 9). The decision at the time was not to attempt decompression of the canal, and steroid treatment was commenced to try to reduce the nerve oedema associated with the fracture. This patient's visual acuity on admission was nil, and he failed to recover any vision. With hindsight, it could be argued that a decompression procedure should have been attempted. However, the patient, aged 63 years, also had multiple facial and skull fractures, with intracerebral haemorrhage on one side,

7 Blindness and severe visual impairment in facial fractures 7 and may well not have tolerated a craniofacial procedure under general anaesthetic at the time. Five patients had a ruptured globe, and one patient (case 12) sustained a penetrating injury to the left eye. Three of these six patients underwent scleral repair by the ophthalmologist. In two of the three the visual acuity improved (cases 2 and 12); the third patient had no improvement and subsequently underwent enucleation. There is no question that an attempt should be made to salvage the damaged globe, if feasible. In this study, we have not addressed the situation where visual loss occurs after treatment of the facial fracture. Girotto et al report four cases of visual loss after operative treatment of facial fractures, and note that this is usually due to intraorbital haemorrhage, it An interesting study was published by Weymuller, who sought opinions from various experts from different specialties regarding the treatment of an acute LeFort III fracture with visual loss in one eye and a fracture involving the orbital apex on the remaining functional side) The majority of ophthalmologists and oral surgeons would not operate, the otolaryngologists were evenly split, and all of the plastic and reconstructive surgeons would operate. This implies that differing levels of training, expertise and philosophy are important factors in the management of these cases. In summary, we present an 11 year review of facial fractures causing blindness or severe visual loss. The typical patient was a male involved in a motor-vehicle accident, suffering multitrauma. The facial fractures usually involved a lateral impact to the orbitozygomatic region, with fractures involving the lateral orbital wall, zygomatic arch and zygomaticofrontal suture, and zygomaticomaxillary buttress. The cause of blindness was most commonly traumatic optic neuropathy, for which there was no specific treatment. The ruptured globe should undergo early surgical repair, if feasible. We recommend that the clinician managing these patients be alert to the possibility of ophthalmological compromise, no matter how seemingly insignificant the fracture. As shown by our results, special caution should be exercised in managing lateral impact-type fractures, and early identification of these fracture patterns by CT scan can alert the clinician to the possibility of severe visual impairment. Visual assessment is mandatory in these patients. From this series, the Craniofacial Disruption Score, calculated using the modified CDFSS, was not correlated with the occurrence of blindness or severe visual impairment. Acknowledgements The authors thank Mr M. Hammerton FRACS, FRACO, Consultant Ophthalmologist, The Australian Craniofacial Unit, for his advice in the preparation of this manuscript, and the research and secretarial staff of the Department of Plastic and Reconstructive Surgery, Royal Adelaide Hospital, for their assistance with data collection and retrieval. References 1. Zachariades N, Papavassiliou D, Christopoulos E Blindness after facial trauma. Oral Surg Oral Med Oral Pathoi 1996; 81: Lipkin AE Woodson GE, Miller RH. Visual loss due to orbital fracture: the role of early reduction. Arch Otolaryngol Head Neck Surg 1987; 113: Weymuller EA Jr. Blindness and LeFort IlI-fractures. Ann Otol Rhinol Laryngol 1984; 93: Lira LH, Lain LK, Moore MH, Trott JA, David DJ. Associated injuries in facial fractures: review of 839 patients. Br J Plast Surg 1993; 46: Cooter RD, David DJ. Computer-based coding of fractures in the craniofacial region. Br J Plast Surg 1989; 42: O'Sullivan ST, Snyder BJ, Moore MH, David DJ. Outcome measurement of the treatment of maxillary fractures: a prospective analysis of 100 consecutive cases. Br J Plast Surg 1999; 52: A1-Qurainy IA, Stassen LFA, Dutton GN, Moos KF, El-Attar A. The characteristics of midfacial fractures and the association with ocular injury: a prospective study. Br J Oral Maxillofac Surg 1991; 29: Ioannides C, Treffers W, Rutten M, Noverraz E Ocular injuries associated with fractures involving the orbit. J Craniomaxillofac Surg 1988; 16: Li KK, Teknos TN, Lauretano A, Joseph MR Traumatic optic neuropathy complicating facial fracture repair. J Craniofac Snrg 1997; 8: Jabaley ME, Lerman M, Sanders HJ. Ocular injuries in orbital fractures: a review of 119 cases. Plast Reconstr Surg 1975; 56: Girotto JA, Gamble WB, Robertson B, et al. Blindness after reduction of facial fractures. Plast Reconstr Surg 1998; 102: Manfredi SJ, Raji MR, Sprinkle PM, Weinstein GW, Minardi LM, Swanson TJ. Computerized tomographic scan findings in facial fractures associated with blindness. Plast Reconstr Surg 1981; 68: LeFort R. Etude experimentale sur les fractures de la machoire superieure. Rev Chir Paris 1901; 23: 208,360, David DJ, Simpson DA. Craniomaxillofacial Trauma. New York: Churchill Livingstone, 1995: Jend H-H, Jend-Rossmann I. Sphenotemporal buttress fracture: a report of five cases. Neuroradiology 1984; 26: The Authors Craig A. MaeKinnon MB, ChB, FRACS, Honorary Clinical Associate of the University of Hong Kong Department of Surgery, Level $3, Queen Mary Hospital, 102 Pokfulam Road, Pokfulam, Hong Kong. David J. David AC, MD, FRCS, FRACS, Head Australian Craniofacial Unit, Women and Children's Hospital, 72 King William Road, North Adelaide, South Australia 5006, Australia. Rodney D. Cooter MBBS, MD, FRACS, Head of Unit Department of Plastic and Reconstructive Surgery, Royal Adelaide Hospital, Adelaide, South Australia, Australia. Correspondence to Mr David J. David AC, MD, FRCS, FRACS. Paper received 12 January Accepted 24 September 2001, after revision.