- Email: [email protected]
Internal orbital fractures in the pediatric age group
Internal Orbital Fractures in the Pediatric Age Group Characterization and Management Zsolt C. Bansagi, BA, Dale R. Meyer, MD, FACS Objective: To evaluate the specific characteristics and management of internal orbital fractures in the pediatric population. Design: Retrospective observational case series. Participants: Thirty-four pediatric patients between the ages of 1 and 18 years with internal orbital (“blowout”) fractures. Methods: Records of pediatric patients presenting with internal orbital fractures over a 5-year period were reviewed, including detailed preoperative and postoperative evaluations, surgical management, and medical management. Main Outcome Measures: Ocular motility restriction, enophthalmos, nausea and vomiting, and postoperative complications. Results: Floor fractures were by far the most common fracture type (71%). Eleven of 34 patients required surgical intervention for ocular motility restriction. Eight were trapdoor-type fractures with soft-tissue incarceration; five had nausea and vomiting. Early surgical intervention (⬍2 weeks) resulted in a more complete return of ocular motility compared with the late intervention group. Conclusions: Trapdoor-type fractures, usually involving the orbital floor, are common in the pediatric age group. These fractures may be small with minimal soft-tissue incarceration, making the findings on computed tomography scans quite subtle at times. Marked motility restriction and nausea/vomiting should alert the physician to the possibility of a trapdoor-type fracture and the need for prompt surgical intervention. Ophthalmology 2000;107:829 – 836 © 2000 by the American Academy of Ophthalmology. Orbital fractures may occur in many patients who have blunt trauma to the face and skull.1–11 The management of orbital fractures has been debated and remains controversial. Although there are multiple studies on orbital fracture management in adult patients, there are relatively few published studies evaluating the characteristics and management of internal orbital fractures in the pediatric population. Previous studies have suggested that pediatric patients differ in the pattern of injury, the incidence of so-called “trapdoor” fractures, and clinical symptoms associated with these fractures.1–13 Trapdoor fractures occur when a circular segment of the bony orbit fractures and becomes displaced but remains attached on one side.5 A study by de Man et al1 evaluated 15 pediatric orbital fracture cases and reported an astounding 14 (93%) of these cases had “trapdoor fractures.” Koltai et al3 described the mechanism and patterns of
orbital-facial trauma in the pediatric population with respect to facial morphology and craniofacial development. Possibly because of the greater elasticity of the orbital bones in the pediatric population, a greater potential exists for development of trapdoor-type fractures, which in turn allows prolapsed orbital tissue to be caught and trapped in the fracture site. Entrapment of the soft tissue and muscle may not only limit eye movement but may also result in early tissue necrosis caused by impaired blood supply caused by compression of the tissue within the confines of the fractured bony orbit.1,4,12,13 The purpose of our study was to further evaluate the characteristics and management of orbital fractures in the pediatric population, focusing specifically on internal (“blowout”) fractures.
Methods Originally received: July 23, 1999. Accepted: January 13, 2000. Manuscript no. 99411. From the Orbital and Ophthalmic Plastic Surgery Service, Lions Eye Institute, Department of Ophthalmology, Albany Medical College, Albany, New York. Presented in part at the Annual Meeting of the American Academy of Ophthalmology, Orlando, Florida, October, 1999. Reprint requests to Dale R. Meyer, MD, FACS, Lions Eye Institute, 35 Hackett Boulevard, Albany, NY 12208. © 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
From May 1993 until May 1998, 113 patients with internal orbital fractures were seen by one of the authors (DRM) at Albany Medical Center. All patients were identified from a computergenerated database with the diagnosis of internal orbital fracture. Patients older than 18 years at the time of their injury were excluded, leaving a total of 38 patients in the pediatric age group. Four patients with extensive craniofacial injuries managed by other surgical services were excluded because of limited patient cooperation for initial ophthalmologic evaluation and subsequent ISSN 0161-6420/00/$–see front matter PII S0161-6420(00)00015-4
829
Ophthalmology Volume 107, Number 5, May 2000 Table 1. Specific Causes of Orbital Fractures in Study Group Causes of Injury
Patients (n ⴝ 34)
Percentage of Total
12 9 6 5 2
35 26 18 15 6
Assault* Sport injury Fall MVA Other
MVA ⫽ motor vehicle accident. * The assault category includes all injuries that individuals inflicted upon each other regardless of the intent (i.e. injuries that occurred during horseplay).
Table 2. Type of Fractures and Specific Orbital Wall Involvement Types of Fracture Orbital Wall Involvement Floor Medial wall Roof Lateral wall Total*
Pure Internal Orbital Fractures (No Orbital Rim Involvement)
Impure Internal Orbital Fractures (Orbital Rim Involvement)
Total*
20 1 1 0 22
7 3 4 2 16
27 4 5 2 38
* Note that the totals do not add up to n ⫽ 34, due to four patients having multiple orbital wall injuries. Two patients presented with combined roof and medial wall fractures in addition to two patients with floor and lateral wall fractures.
not require surgery were discharged with the recommendation to be followed-up within a month by our service. Other services participated in medical/surgical management and follow-up if the patient had additional injuries requiring treatment. Our indications for surgical intervention included significant (i.e., ⬎2 mm; clinically obvious) enophthalmos, extraocular muscle restriction with symptomatic diplopia (within 30 degrees fixation), and/or CT findings of large orbital wall defect (⬎50% wall with high potential for enophthalmos). Surgeries described in this study for internal orbital fracture repair were performed with the patients under general anesthesia by one of the authors (DRM), who also performed the initial and follow up evaluations. Forced duction tests were performed at the beginning of surgery. Exposure to the inferior and medial orbital walls was usually obtained by a transconjunctival approach with lateral canthotomy and cantholysis. Dissection was carried down to the inferior orbital rim periosteum, which was incised and reflected. Once the fracture was fully exposed, the entrapped tissue was gently teased out of the fracture site using blunt dissection. Every effort was made to preserve the infraorbital neurovascular bundle. Once the tissue was teased out of the fracture site, forced duction testing was repeated. Depending on the size and character of the fracture, the fracture site was either covered by a Gelfilm implant (small fracture, minimal displacement) or Medpor (1.5-mm sheet) implant (large fracture, displaced). After closure, antibiotic ointment was placed in the eye along with a light dressing. Most patients were treated with perioperative intravenous antibiotics continued orally for 7 days. Patients who underwent surgical repair were advised to follow-up with our service at 1 week, 3 months, and 6 months after surgery. All surgical patients completed at least the 3-month follow-up.
Results loss to follow-up. The remaining 34 patients who make up the study group all had internal orbital fractures that were confirmed by axial and coronal computed tomography (CT) scans, clinical findings, and, in several cases, intraoperative observation. All patients underwent a coronal CT evaluation to determine the presence or absence of fracture, the involvement of orbital walls, and the possibility of muscle or soft tissue entrapment. Coronal CT scans were initially read by a radiologist and were subsequently reviewed by an ophthalmologist with subspeciality training in orbital and ophthalmic plastic surgery (DRM). Initial clinical evaluations were performed on all 34 patients, including history and full ophthalmic examination with measurement of visual acuity, enophthalmos, diplopia, ocular alignment, and ductions. Ocular motility limitation was graded on a numerical system of 0 to ⫺4, where 0 equals no limitation (normal) and each increment reflects a 25% reduction (i.e., ⫺1 ⫽ 25%, ⫺2 ⫽ 50%, ⫺3 ⫽ 75%, and ⫺4 ⫽ 100% limitation). Forced duction testing was performed when patient cooperation allowed. Patients who did
The mean age of our patients was 11.8 years (⫾5.45 SD; range, 1–18 years old). The sex distribution was 24 males and 10 females. The causes of the orbital fractures are presented in Table 1. Assault was the most common cause of injury in this group (12 of 34 [35%]); the assault category included injuries inflicted by another individual, whether intentional or accidental (e.g., “horseplay”). This was followed by injuries sustained during sports (9 of 34 [26%]). Twenty-two of the 34 (65%) patients were identified with “pure” internal orbital fractures (i.e., without involvement of the orbital rim), whereas 12 of 34 (35%) had “impure” internal orbital fractures (i.e., those that extended through the orbital rim). Table 2 summarizes the type of orbital fracture (“pure” vs. “impure”) and specific orbital walls involved. The results show that floor fractures were by far the most common location; 27 of 38 (71%) walls were involved in these 34 patients. The frequency of other walls involved was roof, 5 of 38 (13%); medial wall, 4 of 38 (11%); and lateral wall 2 of 38 (5%).
Table 3. Relationship of Ocular Motility Limitation to Orbital Wall Fractures Orbital Walls Involved Floor Roof Medial wall Lateral wall Total*
Supraduction Limitation
Infraduction Limitation
Abduction Limitation
Adduction Limitation
No Movement Limitation
13 0 0 0 13
11 0 1 0 12
2 0 1 0 3
1 0 1 0 2
13 5 2 2 22
* Note that the totals do not add up to n ⫽ 34, since some patients had multiple wall fractures and a single fracture may result in multi-directional ocular motility restriction (i.e. one patient with a medial wall fracture had both horizontal and vertical motility restrictions.
830
Bansagi and Meyer 䡠 Pediatric Orbital Fractures Table 4. Initial Examination Findings Pure Internal Orbital Fractures n ⴝ 22
Impure Internal Orbital Fractures n ⴝ 12
Initial Examination Findings*
All Patients n ⴝ 34
With Trapdoor, n ⫽ 8
W/o Trapdoor, n ⫽ 14
With Trapdoor, n ⫽ 0
W/o Trapdoor, n ⫽ 12
Ocular motility restriction Enophthalmos (⬎2 mm) Nausea/vomiting
15 (44%) 1 (3%) 7 (21%)
8 0 5
7 1 2
0 0 0
0 0 0
* Initial examination findings stratified according to fracture type. The 15 patients with ocular motility restriction represented 13 floor fractures, and 2 medial wall fractures (one with horizontal motility restriction, the other with combined horizontal and vertical motility restriction). None of the patients with impure fractures had motility restrictions or required surgical intervention by our service. The seven patients with nausea/vomiting included 5 patients with trapdoor fractures (all floor fractures) and two patients with non-trapdoor fractures (one roof and one floor fracture). The three trapdoor fractures without nausea/vomiting included two floor fractures and one medial wall fracture.
Ocular motility restriction noted at the initial ophthalmic examination and the relationship to the specific orbital walls involved was also compared. The average timing of the initial examination was 7.18 days (range, 0 – 43 days; SD, ⫾9.27). Thirty-one of 34 (91%) patients had their initial examination within the first 2 weeks after their injuries. The three remaining patients had initial examinations at 3 weeks, 4 weeks, and 7 weeks, respectively. The relationship of orbital wall fracture location to ocular motility restriction is presented in Table 3. Vertical motility impairment (either supraduction or infraduction limitation) was by far the most common ocular motility disturbance and was, with one exception, associated with fracture of the orbital floor. Of the 27 patients with floor fractures, 13 had supraduction limitation, with 11 of these also having infraduction limitation. To further quantitatively evaluate the presenting clinical characteristics of the specific types of fractures, we compared initial ocular motility limitation, enophthalmos (⬎2 mm), and nausea/ vomiting, with fracture type (i.e., pure vs. impure; trapdoor vs. non-trapdoor). Table 4 summarizes these findings. Trapdoor-type fractures were identified by CT findings and/or intraoperative observations. A total of 8 patients was found to have trapdoor-type fractures, and all of them underwent surgical intervention to free up the entrapped extraocular muscle and restore the orbital wall defect. Characteristics of the group of patients with trapdoor fractures included a mean age of 12.0 years, with eight of eight (100%) having a pure internal orbital fracture with involvement of either the floor or the medial wall (7 and 1, respectively). As shown in Table 4, all patients with trapdoor-type fracture had motility restrictions. Enophthalmos ⬎2 mm was only noted in one patient with a large non-trapdoor floor fracture. The relationship between trapdoor fractures and nausea/vomiting and motility restriction, for the
pure internal orbital fracture subgroup and the entire group (pure and impure internal orbital fractures) was evaluated by two-tailed Fisher exact tests. Table 5 shows the results of this analysis. A significant relationship was noted for both nausea/vomiting and ocular motility restriction to trapdoor-type fractures. Nausea and vomiting was a presenting symptom in 7 of 34 patients (21%). All seven of these patients had pure internal orbital fractures, five had trapdoor-type fractures and two did not. Nausea/vomiting was 63% (5 of 8 patients) sensitive for the presence of trapdoor fractures and 92% (24 of 26 patients) specific for the entire group of 34 patients with internal orbital fractures. The positive predictive value of nausea/vomiting for identifying trapdoor fractures in this group was 71% (5 of 7 patients). Ocular motility restriction was noted in 15 patients, 8 with trapdoor fractures and 7 with non-trapdoor fractures (all pure internal orbital fractures). Recognizing the association between supraduction limitation and floor fractures described earlier (Table 3), we compared the mean preoperative supraduction motility restriction for floor fractures with and without trapdoor type. From the total of 27 patients with floor fractures, 7 had trapdoor fractures and 20 had non-trapdoor fractures. The mean preoperative supraduction motility restriction was ⫺2.7 in the trapdoor group and was ⫺0.1 in the non-trapdoor group. Eleven patients underwent surgical exploration and repair because of ocular motility restriction. Four patients with initial ocular motility limitation improved sufficiently within 2 weeks and did not require surgery. The average time of surgical intervention was 13.4 days after the injury (⫾SD 11.4; range, 0 – 40 days). The timing of surgical intervention was categorized as immediate (0 – 48 hours), early (3–14 days), and late (⬎15 days) repair. There were no intraoperative complications and no patient sustained vision loss or worsening of ocular motility as a result of surgery.
Table 5. Relationship of Trapdoor Fractures to Nausea/Vomiting and Motility Restrictions* Pure Internal Fractures
All Fractures
Trapdoor
Non-Trapdoor
With nausea/vomiting Without nausea/vomiting Total
5 3 8
2 12 14 P ⫽ 0.05
7 15 22
With motility restriction Without motility restriction Total
8 0 8
7 7 14 P ⫽ 0.02
15 7 22
Trapdoor
Non-Trapdoor
With nausea/vomiting Without nausea/vomiting Total
5 3 8
2 24 26 P ⬍ 0.01
7 27 34
With motility restriction Without motility restriction Total
8 0 8
7 19 26 P ⬍ 0.01
15 19 34
* Statistical relationship (contingency tables) for presence of nausea/vomiting and ocular motility restriction compared with presence of trapdoor fracture (for pure internal orbital fractures n ⫽ 22 and all fractures n ⫽ 34).
831
Ophthalmology Volume 107, Number 5, May 2000 Table 6. Mean Changes in Supraduction Limitation Based on the Time of Intervention* Time of Intervention (From Date of Injury in Days)
No. Patients
Mean Preop Supraduction Limitation
Mean Postop Supraduction Limitation
Mean Change in Supraduction Limitation
2 3 5
⫺3.0 ⫺0.6 ⫺3.4
0 0 ⫺1.7
3.0 0.6 1.3
0–2 3–14 15 & over (range 15–40, mean 21)
* Comparison of mean supraduction limitation preoperatively and postoperatively, stratified by timing of surgery for orbital floor fractures. Limitations in supraduction are indicated with a negative value, therefore a positive mean change means improvement in duction limitation.
The mean estimated blood loss was 10 ml (⫾SD 8). There were nine surgeries performed for the repair of floor fractures, one for medial wall fracture and one for combined floor and lateral wall fractures. An implant was placed over the fracture site in all cases (7 Gelfilm [Pharmacia-Upjohn, Kalamazoo, MI] and 4 Medpor) [Porex Surgical, College Park, GA]. Gelfilm was generally used when the fracture was narrow or the displaced segment of orbital bone could be repositioned in nearly anatomic location after release of entrapped orbital tissues (typical of most trapdoor fractures). Medpor was used for manifestly larger defects that required structural support. Although all patients showed improvement at the 3-month follow-up in their ocular motility, for analysis we evaluated the quantitative amount of supraduction limitation in the 10 patients with floor fractures who underwent surgery to compare the benefit of surgery based on the time of intervention. Table 6 summarizes the mean change in supraduction limitation using the same quantitative scale (0 to ⫺4), which we described earlier in the study. All patients having surgery within 14 days had complete recovery of supraduction, although the mean preoperative supraduction limitation and postoperative improvement was greater in the immediate (0 –2 day) surgical group. The late surgery group demonstrated some persistent (mean) supraduction limitation. Table 7 lists the characteristics of the patients who underwent surgery. Two illustrative cases from this series are described.
Horizontal ductions were intact. The coronal CT was reported by the radiologist to show a left orbital floor fracture. One of us (DRM) believed that there was distortion and probable entrapment of the inferior rectus muscle on the CT scan (Fig 2). The patient underwent orbital exploration the same day. Forced duction testing at the beginning of the surgery revealed marked restriction in attempted manual supraduction. At surgery a hinged, trapdoortype fracture was discovered in the medial aspect of the orbital floor, with soft-tissue pulled into the fracture site (Fig 3). The incarcerated tissue was freed and the inferiorly displaced portion of the fracture “sprang back” into nearly normal anatomic position. The fracture site was covered with a Gelfilm implant. Forced duction tests were repeated and the results were completely nor-
Case 1 An 8-year-old boy slipped on the kitchen floor and fell sustaining blunt trauma to his left eye. In the emergency room the boy had nausea and an episode of vomiting. On examination there was normal visual acuity, diplopia in primary gaze, and ⫺4 limitation of both infraduction and supraduction (Fig 1) in the left eye. Table 7. Surgical Group* Patient
Age (yrs)
Trapdoor (w/entrapment)
Timing (Days)
Floor
1 2 3 4 5 6 7 8 9 10 11
14 13 18 5 8 7 18 14 18 16 13
Yes Yes Yes Yes Yes No Yes No Yes Yes No
7 15 15 6 0 17 40 24 1 7 15
Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes
* Characteristics of patients who underwent surgery for internal orbital fractures. Note that 10/11 patients had orbital floor fractures.
832
Figure 1. Case 1: 8-year-old boy who fell, sustaining left orbital trauma. A, eyes in primary position. Note relative absence of orbital edema. B, note marked limitation of left eye in attempted supraduction.
Bansagi and Meyer 䡠 Pediatric Orbital Fractures
Figure 2. Case 1: Orbital coronal CT. A, anterior orbit shows distortion of inferior rectus muscle. B, posterior orbit shows trapdoor-type fracture with inferior rectus entrapment.
mal. There was minimal vertical duction limitation at the 1-week follow-up. By the third week after surgery, the patient had full extraocular muscle movements without diplopia (Fig 4).
Case 2 An 18-year-old boy sustained blunt trauma to his orbit after being struck in the left eye during a fight. He was evaluated in the emergency department by an emergency physician, treated, and released. An orbital x-ray film was reported as normal. The patient was referred for ophthalmologic evaluation the day after injury. The patient complained of persistent nausea and vomiting, photophobia, diplopia, and severe pain in the left eye on attempted upgaze. He had normal visual acuity, no enophthalmos, a left hypotropia in primary gaze with ⫺3 limitation in supraduction and ⫺2 limitation of infraduction in the left eye (Fig 5). A coronal CT scan was obtained, which showed a small left orbital floor fracture with soft-tissue entrapment and distortion of the left inferior rectus (Fig 6). It was suspected that the nausea and vomiting was the result of vagal reaction caused by pain and the traction on the inferior rectus muscle. The patient underwent urgent orbital exploration and repair. Forced duction tests were performed at the beginning of surgery, which showed marked restriction on attempted manual supraduction. A slightly displaced, narrow fracture was discovered on the medial aspect of the left orbital floor, with soft-tissue entrapment. The tissue was gently removed and the fractured was covered with a Gelfilm implant. Repeated forced
Figure 3. Case 1: Orbital fracture repair. A, markedly positive forced duction test. B, trapdoor fracture with soft-tissue incarceration in medial aspect of the orbit floor was confirmed.
ductions showed no residual restriction. The patient felt better immediately after surgery, and had no further nausea or vomiting. Mild residual extraocular motility restriction in supraduction and infraduction was noted at the 1-week follow-up, which fully resolved by the 1-month follow-up (Fig 7).
Discussion Our understanding and management of internal orbital (“blowout”) fractures is evolving.5–10 Previous management philosophies have ranged from immediate surgery on all blowout fractures to surgery on none. Current generally accepted indications for surgical repair after an internal orbital fracture are muscle entrapment resulting in ocular motility restriction with diplopia, early enophthalmos (⬎ 2 mm), and orbital defects involving more than 50% of the floor or medial wall.5,6,9 Traditional recommendations for orbital blowout fracture management in the adult population include a 2-week observation period to allow edema to subside to help determine the relative contribution of orbital edema to ocular motility restriction before considering surgical intervention.7–9 More recent recommendations indi-
833
Ophthalmology Volume 107, Number 5, May 2000
Figure 6. Case 2: Coronal CT shows soft-tissue entrapment and distortion of the inferior rectus muscle within a trapdoor-type fracture of the medial orbital floor.
Figure 4. Case 1: 3 weeks postoperatively. A, good globe position in primary gaze. B, full return of ocular motility.
cate the need for careful evaluation and the consideration of earlier surgery with evidence of symptomatic diplopia and CT evidence of entrapment. In cases with obvious clinical signs and CT evidence of frank muscle entrapment, repair within 48 hours has been advocated.5,6,9 Recent publications from several disciplines evaluating the pediatric pop-
Figure 5. Case 2: 18-year-old boy struck in the left eye had pain on attempted upgaze and persistent nausea and vomiting. Marked restriction in supraduction of the left eye was noted.
834
ulation have suggested an increased potential for trapdoortype fractures in this young age group that may require expeditious repair to restore motility and reduce the possibility of tissue ischemia and necrosis.1– 4,12,13 The frequency of this phenomenon, however, is not well-established. In our study we found that trapdoor fractures occurred in 8 of 34 patients (24%) with internal orbital fractures in the pediatric population. All trapdoor fractures occurred in the subgroup of patients with pure internal orbital fractures (i.e., those not involving the orbital rim), and the frequency for this subgroup was thus higher (8 of 22 [36%]). These percentages are clinically important but are less than the 93% reported by de Man and coworkers.1 Most trapdoor fractures involved the orbital floor (7 of 8 patients [88%]). The orbital floor was also the most common wall of the orbit involved in the pediatric population with internal orbital fractures (27 of 38 [71%]). We found that trapdoor-type fractures were strongly associated with ocular motility restriction. Indeed, all eight patients with trapdoor fractures demonstrated reduced ocular motility. In the subset of patients with floor fractures, where trapdoor fractures tended to occur, there was much greater supraduction limitation in patients with trapdoor
Figure 7. Case 2: 1 month postoperatively full ocular motility was noted.
Bansagi and Meyer 䡠 Pediatric Orbital Fractures fractures than in those without trapdoor-type fractures. The limitation in supraduction was usually quite marked, with horizontal ductions preserved or minimally affected. Typically, orbital edema was relatively minimal in these cases and not sufficient to explain the motility pattern. These findings are in agreement with the observations of previous authors.1– 4,12,13 The risk of soft-tissue entrapment in trapdoor fractures is likely increased in pediatric patients with pure internal orbital fractures, where the more elastic, flexible bone can crack, forming a narrow circular or elliptical fracture, hinged on one side, that traps soft tissue within the depressed segment. We have found that trapdoor fractures of the orbital floor tend to be hinged on the medial side, with the depressed portion of the fracture adjacent to the course of the infraorbital neurovascular bundle (Figs 2 and 6). With complex or impure internal orbital fractures, involving both the internal orbit and the orbital rim or other facial bones, the dispersion of force over a larger area probably results in less likelihood of soft tissue entrapment. We also found a strong association between nausea and vomiting and trapdoor fractures. We identified seven patients with nausea and/or vomiting, all with pure internal orbital fractures; five were of the trapdoor type. Although nausea and vomiting alone is certainly not an indication for urgent surgical intervention, this finding coupled with markedly restricted extraocular motility should alert the physician to the increased possibility of trapdoor fracture with extraocular muscle entrapment in the pediatric population. We believe that nausea and vomiting may be a vagally mediated response to pain or other sensory feedback associated with extraocular muscle traction. Sires et al4 described three pediatric patients with trapdoor fractures and entrapment of the inferior rectus muscle, resulting in oculocardiac reflex and the triad of bradycardia, nausea, and syncope. One patient showed hypoxia of the entrapped muscle after only 7 hours from the time of injury. Another patient whose surgical intervention was delayed 72 hours showed ischemic necrosis of the entrapped muscle.4 The degree of hypoxia or ischemic necrosis of the entrapped tissue can be difficult to assess during orbital surgery. We cannot say with certainty whether any of our patients experienced these consequences. Although statistical comparison was not possible because of the small number of patients in the surgical subgroups, we found that the degree of recovery of ocular motility was more complete in the group of patients who had surgical intervention in less than 2 weeks compared with the delayed surgical intervention group (⬎15 days.). Those operated on immediately (within 2 days of injury) had greater ocular motility restriction than those operated on early (3–14 days after injury), yet still had complete recovery. None of the patients in our study reported syncope. Aside from the initial emergency department evaluation of vital signs that included pulse rate, we did not monitor our patients’ cardiac rhythm. Therefore, we cannot say with certainty that any of our patients demonstrated complete oculocardiac reflex. Nevertheless, the mechanism is probably closely related, and physicians managing pediatric orbital trauma should be aware of these phenomena and their implications for evaluation and treatment. Coronal CT scans are an accepted and useful method of identifying orbital fractures and possible extraocular mus-
cle/soft-tissue entrapment. The sensitivity of coronal CT scans in identifying tissue entrapment depends on the quality of the scan, the position and size of fracture, and the acumen of the radiologist or the physician evaluating the CT scan. Bower Wachler and Holds12 reported two cases of inferior rectus muscle entrapment with orbital trapdoor fractures in young patients that were missed radiographically. Jordan and coworkers,13 in a study of 20 pediatric patients with orbital fractures from four institutions, also noted the relative scarcity of CT findings with small or trapdoor fractures, despite clinical motility restriction and surgically confirmed entrapment. Our study emphasizes the importance of careful review of coronal CT scans for the presence of both fractures and entrapment, especially in the pediatric population, that appear predisposed to narrow or trapdoortype fractures. These fractures may be small and the amount of tissue incarceration minimal, making the findings on CT quite subtle at times. CT findings should be carefully evaluated and combined with additional clinical information before ruling out fracture and entrapment. Signs of significant ocular motility restriction and nausea/vomiting in the pediatric population should alert the physician to the increased possibility of a narrow or trapdoor-type fracture with entrapment. A markedly positive forced duction test is suggestive of entrapment but can be difficult to perform in young poorly cooperative patients. Like most retrospective trauma studies, the limitations of our study must be acknowledged. Because of the close association of our ophthalmology department with a level I trauma center (Albany Medical Center), our study likely reflects referral bias and may include more severe cases of orbital trauma based on the type of patients transferred to such trauma centers. There are also other surgical subspecialties (e.g., oral surgery, otolaryngology, general plastic surgery, and neurosurgery) involved in the care of patients with craniomaxillofacial trauma that influence the initial evaluation (“triage”) and follow-up of such patients. At our institution, however, the ophthalmology service is primarily responsible for the evaluation and management of all patients with pure internal orbital fractures, the largest subgroup and focus of this study and the group most ophthalmologists elsewhere are called on to evaluate and help manage. Our study does confirm the suggestion that many young patients who sustain internal orbital fractures frequently have trapdoor-type fractures with soft-tissue entrapment and ocular motility impairment. These trapdoor fractures most commonly involve the orbital floor and demonstrate a corresponding limitation in vertical motility (particularly supraduction) that is typically disproportionate to the amount of orbital edema and limitation in other fields of gaze. Because edema is usually not a prominent feature, there tends to be little improvement in ocular motility after 1 to 2 weeks. We also found that nausea and vomiting was a common finding highly predictive of trapdoor-type fracture in the setting of pediatric orbital trauma. Although this study cannot unequivocally answer the question of ideal surgical timing, we agree with the recent suggestions of others that in the setting of documented or high-probability trapdoor fractures with restricted ocular motility accompanied by nausea/vomiting or complete oculocardiac reflex, earlier surgical intervention should be considered in this pediatric
835
Ophthalmology Volume 107, Number 5, May 2000 population. Prompt intervention may protect these patients from ischemic injury to incarcerated orbital soft tissues,4,12,13 and, as described by Sires and coworkers,4 in cases of complete oculocardiac reflex, may prevent rare, but potentially fatal, arrhythmias. Acknowledgment We thank Dr. Sanjayant R. Chamakura for his assistance in the data collection for this manuscript.
6. 7. 8.
References
9.
1. de Man K, Wijngaarde R, Hes J, de Jong PT. Influence of age on the management of blow-out fractures of the orbital floor. Int J Oral Maxillofac Surg 1991;20:330 – 6. 2. Anderson PJ, Pole MD. Orbital floor fractures in young children. Craniomaxillofac Surg 1995;23:151– 4. 3. Koltai PJ, Amjad I, Meyer DR, Feustel PJ. Orbital fractures in children. Arch Otolaryngol Head Neck Surg 1995;121:1375–9. 4. Sires BS, Stanley RB, Levine LM. Oculocardiac reflex caused by orbital floor trapdoor fracture: an indication for urgent repair. Arch Ophthalmol 1998;116:955– 6. 5. Meyer DR. Orbital fractures. In: Tasman W, Jaeger EA, eds.
10. 11. 12.
13.
Duane’s Clinical Ophthalmology, rev. ed. Philadelphia: Lippincott-Raven, 1996; chap. 48. Rubin PA, Bilyk JR, Shore JW. Management of orbital trauma: fractures, hemorrhage, and traumatic optic neuropathy. Am Acad Ophthalmol Focal Points 1994;XII(7):1– 8. Levin LM, Kademani D. Clinical considerations in the management of orbital blow-out fractures. Compendium 1997;6: 593– 600. Wilkins RB, Havins WE. Current treatment of blow-out fractures. Ophthalmology 1982;89:464 – 6. Dutton JJ, Manson PN, Iliff N, Putterman AM. Management of blow-out fractures of the orbital floor. Surv Ophthalmol 1990;35:279 – 8. Hawes JM, Dortzbach RK. Surgery on orbital floor fractures. Ophthalmology 1983;90:1066 –70. Carr RM, Mathog RH. Early and delayed repair of orbitozygomatic complex fractures. J Oral Maxillofac Surg 1997;55:253–8. Bower Wachler BS, Holds JB. The missing muscle syndrome in blowout fractures: an indication for urgent surgery. Ophthalmic Plast Reconstr Surg 1998;14:17–18. Jordan DR, Allen LH, White J, et al. Intervention within days for some orbital floor fractures: the white-eyed blowout. Ophthalmic Plast Reconstr Surg 1998;14:379 –90.
Discussion by Mark R. Levine, MD The overall indication for surgical intervention for blowout fractures remains controversial. Indications for conservative surgery within 2 weeks of the trauma include symptomatic diplopia with positive forced ductions, computed tomographic (CT) evidence of muscle entrapment with no clinical improvement over 1 to 2 weeks, enophthalmos of 3 mm or more, significant hypoophthalmos, and an orbital wall defect involving more than 50% of the floor.1 Conversely, the need for immediate surgery does arise occasionally, as in the case of globe herniation into the maxillary sinus. Another study found that small orbital fracture defects with incarcerated extraocular muscles lead to a compartment syndrome that causes muscle ischemia, fibrosis, and motility restriction.2 Bansigi and Meyer describe an important subset of 34 pediatric patients (1 to 18 years of age), 71% of whom had orbital floor fractures. Trapdoor fractures were associated with more severe motility restriction than non-trapdoor fractures, and nausea and vomiting were associated presenting symptoms in 21% of cases. Interestingly, CT evidence was not reported by the radiologist in nearly half the patients with trapdoor fractures. The authors conclude that ocular motility recovery was more complete in patients who had surgical intervention in less than 2 weeks compared with those who had surgery after 2 weeks. Although they caution that their study cannot unequivocally answer the question of ideal surgical timing, they do agree that early surgical intervention should be considered when patients have the previously mentioned criteria. Bansagi and Meyer make a strong case for early surgical intervention in pediatric patients with trapdoor fractures, motility restriction, nausea and vomiting. The small number of patients (n ⫽ 2) operated on within 48 hours had a more rapid motility recovery than those within 2 weeks or longer. They note that
From the University Hospitals of Cleveland, Case Western Reserve University School of Medicine, Cleveland, Ohio. Address correspondence to Mark R. Levine, MD, Mt. Sinai Medical Building, 26900 Cedar Road, Suite #311, Beachwood, OH 44122.
836
earlier evaluation might have been better able to address appropriateness of surgical timing if they could have evaluated the patients more expeditiously (initial examination 7.18 days with surgical intervention 13.4 days after injury) and surgically intervened much earlier. Therefore, the conclusion of early surgical intervention was based on less than the 2-week period that many of us already accept as appropriate surgical intervention. Their article does, however, support articles written by Sires3 and Jordan for early intervention in the pediatric patient to improve surgical outcomes and prevent muscle ischemia resulting in fibrosis and motility restriction. We have learned that, because of their softer, more flexible orbital bones, trapdoor fractures are more likely in the pediatric age group compared with adults. The harder, more brittle orbital bones of the adult usually result in breaks in several areas, causing a buckle effect. Moreover, in the pediatric patient in whom tissue (fat, connective tissue, and muscle) is incarcerated in a flexible trapdoor fracture, the potential space around the extraocular muscles is disrupted, and a high compartment pressure occurs around the inferior rectus muscle resulting in ischemia. Therefore, surgical intervention within 24 to 72 hours in the pediatric patient who has marked motility restriction, trapdoor fracture, and nausea and vomiting may restore circulation to the rectus muscle and perimuscular tissue, preventing fibrosis and enhancing extraocular muscle recovery. A randomized study of trapdoor fractures with tissue incarceration operated on at different time intervals might be helpful. References 1. Jordan DR, Allen LH, White J, et al. Intervention within days for some orbital floor fractures: the white-eyed blowout. Ophthalmic Plast Reconstr Surg 1998;14:379 –90. 2. Smith B, Lisman RD, Simonton J, et al. Volkman’s contracture of the extraocular muscles following blowout fractures. Plast Reconstr Surg 1984;74:200 –5. 3. Sires BS, Stanley RB, Levine LM. Oculocardiac reflex caused by orbital floor trapdoor fracture: an indication for urgent repair. Arch Ophthalmol 1998;116:955– 6.
orbital-facial trauma in the pediatric population with respect to facial morphology and craniofacial development. Possibly because of the greater elasticity of the orbital bones in the pediatric population, a greater potential exists for development of trapdoor-type fractures, which in turn allows prolapsed orbital tissue to be caught and trapped in the fracture site. Entrapment of the soft tissue and muscle may not only limit eye movement but may also result in early tissue necrosis caused by impaired blood supply caused by compression of the tissue within the confines of the fractured bony orbit.1,4,12,13 The purpose of our study was to further evaluate the characteristics and management of orbital fractures in the pediatric population, focusing specifically on internal (“blowout”) fractures.
Methods Originally received: July 23, 1999. Accepted: January 13, 2000. Manuscript no. 99411. From the Orbital and Ophthalmic Plastic Surgery Service, Lions Eye Institute, Department of Ophthalmology, Albany Medical College, Albany, New York. Presented in part at the Annual Meeting of the American Academy of Ophthalmology, Orlando, Florida, October, 1999. Reprint requests to Dale R. Meyer, MD, FACS, Lions Eye Institute, 35 Hackett Boulevard, Albany, NY 12208. © 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
From May 1993 until May 1998, 113 patients with internal orbital fractures were seen by one of the authors (DRM) at Albany Medical Center. All patients were identified from a computergenerated database with the diagnosis of internal orbital fracture. Patients older than 18 years at the time of their injury were excluded, leaving a total of 38 patients in the pediatric age group. Four patients with extensive craniofacial injuries managed by other surgical services were excluded because of limited patient cooperation for initial ophthalmologic evaluation and subsequent ISSN 0161-6420/00/$–see front matter PII S0161-6420(00)00015-4
829
Ophthalmology Volume 107, Number 5, May 2000 Table 1. Specific Causes of Orbital Fractures in Study Group Causes of Injury
Patients (n ⴝ 34)
Percentage of Total
12 9 6 5 2
35 26 18 15 6
Assault* Sport injury Fall MVA Other
MVA ⫽ motor vehicle accident. * The assault category includes all injuries that individuals inflicted upon each other regardless of the intent (i.e. injuries that occurred during horseplay).
Table 2. Type of Fractures and Specific Orbital Wall Involvement Types of Fracture Orbital Wall Involvement Floor Medial wall Roof Lateral wall Total*
Pure Internal Orbital Fractures (No Orbital Rim Involvement)
Impure Internal Orbital Fractures (Orbital Rim Involvement)
Total*
20 1 1 0 22
7 3 4 2 16
27 4 5 2 38
* Note that the totals do not add up to n ⫽ 34, due to four patients having multiple orbital wall injuries. Two patients presented with combined roof and medial wall fractures in addition to two patients with floor and lateral wall fractures.
not require surgery were discharged with the recommendation to be followed-up within a month by our service. Other services participated in medical/surgical management and follow-up if the patient had additional injuries requiring treatment. Our indications for surgical intervention included significant (i.e., ⬎2 mm; clinically obvious) enophthalmos, extraocular muscle restriction with symptomatic diplopia (within 30 degrees fixation), and/or CT findings of large orbital wall defect (⬎50% wall with high potential for enophthalmos). Surgeries described in this study for internal orbital fracture repair were performed with the patients under general anesthesia by one of the authors (DRM), who also performed the initial and follow up evaluations. Forced duction tests were performed at the beginning of surgery. Exposure to the inferior and medial orbital walls was usually obtained by a transconjunctival approach with lateral canthotomy and cantholysis. Dissection was carried down to the inferior orbital rim periosteum, which was incised and reflected. Once the fracture was fully exposed, the entrapped tissue was gently teased out of the fracture site using blunt dissection. Every effort was made to preserve the infraorbital neurovascular bundle. Once the tissue was teased out of the fracture site, forced duction testing was repeated. Depending on the size and character of the fracture, the fracture site was either covered by a Gelfilm implant (small fracture, minimal displacement) or Medpor (1.5-mm sheet) implant (large fracture, displaced). After closure, antibiotic ointment was placed in the eye along with a light dressing. Most patients were treated with perioperative intravenous antibiotics continued orally for 7 days. Patients who underwent surgical repair were advised to follow-up with our service at 1 week, 3 months, and 6 months after surgery. All surgical patients completed at least the 3-month follow-up.
Results loss to follow-up. The remaining 34 patients who make up the study group all had internal orbital fractures that were confirmed by axial and coronal computed tomography (CT) scans, clinical findings, and, in several cases, intraoperative observation. All patients underwent a coronal CT evaluation to determine the presence or absence of fracture, the involvement of orbital walls, and the possibility of muscle or soft tissue entrapment. Coronal CT scans were initially read by a radiologist and were subsequently reviewed by an ophthalmologist with subspeciality training in orbital and ophthalmic plastic surgery (DRM). Initial clinical evaluations were performed on all 34 patients, including history and full ophthalmic examination with measurement of visual acuity, enophthalmos, diplopia, ocular alignment, and ductions. Ocular motility limitation was graded on a numerical system of 0 to ⫺4, where 0 equals no limitation (normal) and each increment reflects a 25% reduction (i.e., ⫺1 ⫽ 25%, ⫺2 ⫽ 50%, ⫺3 ⫽ 75%, and ⫺4 ⫽ 100% limitation). Forced duction testing was performed when patient cooperation allowed. Patients who did
The mean age of our patients was 11.8 years (⫾5.45 SD; range, 1–18 years old). The sex distribution was 24 males and 10 females. The causes of the orbital fractures are presented in Table 1. Assault was the most common cause of injury in this group (12 of 34 [35%]); the assault category included injuries inflicted by another individual, whether intentional or accidental (e.g., “horseplay”). This was followed by injuries sustained during sports (9 of 34 [26%]). Twenty-two of the 34 (65%) patients were identified with “pure” internal orbital fractures (i.e., without involvement of the orbital rim), whereas 12 of 34 (35%) had “impure” internal orbital fractures (i.e., those that extended through the orbital rim). Table 2 summarizes the type of orbital fracture (“pure” vs. “impure”) and specific orbital walls involved. The results show that floor fractures were by far the most common location; 27 of 38 (71%) walls were involved in these 34 patients. The frequency of other walls involved was roof, 5 of 38 (13%); medial wall, 4 of 38 (11%); and lateral wall 2 of 38 (5%).
Table 3. Relationship of Ocular Motility Limitation to Orbital Wall Fractures Orbital Walls Involved Floor Roof Medial wall Lateral wall Total*
Supraduction Limitation
Infraduction Limitation
Abduction Limitation
Adduction Limitation
No Movement Limitation
13 0 0 0 13
11 0 1 0 12
2 0 1 0 3
1 0 1 0 2
13 5 2 2 22
* Note that the totals do not add up to n ⫽ 34, since some patients had multiple wall fractures and a single fracture may result in multi-directional ocular motility restriction (i.e. one patient with a medial wall fracture had both horizontal and vertical motility restrictions.
830
Bansagi and Meyer 䡠 Pediatric Orbital Fractures Table 4. Initial Examination Findings Pure Internal Orbital Fractures n ⴝ 22
Impure Internal Orbital Fractures n ⴝ 12
Initial Examination Findings*
All Patients n ⴝ 34
With Trapdoor, n ⫽ 8
W/o Trapdoor, n ⫽ 14
With Trapdoor, n ⫽ 0
W/o Trapdoor, n ⫽ 12
Ocular motility restriction Enophthalmos (⬎2 mm) Nausea/vomiting
15 (44%) 1 (3%) 7 (21%)
8 0 5
7 1 2
0 0 0
0 0 0
* Initial examination findings stratified according to fracture type. The 15 patients with ocular motility restriction represented 13 floor fractures, and 2 medial wall fractures (one with horizontal motility restriction, the other with combined horizontal and vertical motility restriction). None of the patients with impure fractures had motility restrictions or required surgical intervention by our service. The seven patients with nausea/vomiting included 5 patients with trapdoor fractures (all floor fractures) and two patients with non-trapdoor fractures (one roof and one floor fracture). The three trapdoor fractures without nausea/vomiting included two floor fractures and one medial wall fracture.
Ocular motility restriction noted at the initial ophthalmic examination and the relationship to the specific orbital walls involved was also compared. The average timing of the initial examination was 7.18 days (range, 0 – 43 days; SD, ⫾9.27). Thirty-one of 34 (91%) patients had their initial examination within the first 2 weeks after their injuries. The three remaining patients had initial examinations at 3 weeks, 4 weeks, and 7 weeks, respectively. The relationship of orbital wall fracture location to ocular motility restriction is presented in Table 3. Vertical motility impairment (either supraduction or infraduction limitation) was by far the most common ocular motility disturbance and was, with one exception, associated with fracture of the orbital floor. Of the 27 patients with floor fractures, 13 had supraduction limitation, with 11 of these also having infraduction limitation. To further quantitatively evaluate the presenting clinical characteristics of the specific types of fractures, we compared initial ocular motility limitation, enophthalmos (⬎2 mm), and nausea/ vomiting, with fracture type (i.e., pure vs. impure; trapdoor vs. non-trapdoor). Table 4 summarizes these findings. Trapdoor-type fractures were identified by CT findings and/or intraoperative observations. A total of 8 patients was found to have trapdoor-type fractures, and all of them underwent surgical intervention to free up the entrapped extraocular muscle and restore the orbital wall defect. Characteristics of the group of patients with trapdoor fractures included a mean age of 12.0 years, with eight of eight (100%) having a pure internal orbital fracture with involvement of either the floor or the medial wall (7 and 1, respectively). As shown in Table 4, all patients with trapdoor-type fracture had motility restrictions. Enophthalmos ⬎2 mm was only noted in one patient with a large non-trapdoor floor fracture. The relationship between trapdoor fractures and nausea/vomiting and motility restriction, for the
pure internal orbital fracture subgroup and the entire group (pure and impure internal orbital fractures) was evaluated by two-tailed Fisher exact tests. Table 5 shows the results of this analysis. A significant relationship was noted for both nausea/vomiting and ocular motility restriction to trapdoor-type fractures. Nausea and vomiting was a presenting symptom in 7 of 34 patients (21%). All seven of these patients had pure internal orbital fractures, five had trapdoor-type fractures and two did not. Nausea/vomiting was 63% (5 of 8 patients) sensitive for the presence of trapdoor fractures and 92% (24 of 26 patients) specific for the entire group of 34 patients with internal orbital fractures. The positive predictive value of nausea/vomiting for identifying trapdoor fractures in this group was 71% (5 of 7 patients). Ocular motility restriction was noted in 15 patients, 8 with trapdoor fractures and 7 with non-trapdoor fractures (all pure internal orbital fractures). Recognizing the association between supraduction limitation and floor fractures described earlier (Table 3), we compared the mean preoperative supraduction motility restriction for floor fractures with and without trapdoor type. From the total of 27 patients with floor fractures, 7 had trapdoor fractures and 20 had non-trapdoor fractures. The mean preoperative supraduction motility restriction was ⫺2.7 in the trapdoor group and was ⫺0.1 in the non-trapdoor group. Eleven patients underwent surgical exploration and repair because of ocular motility restriction. Four patients with initial ocular motility limitation improved sufficiently within 2 weeks and did not require surgery. The average time of surgical intervention was 13.4 days after the injury (⫾SD 11.4; range, 0 – 40 days). The timing of surgical intervention was categorized as immediate (0 – 48 hours), early (3–14 days), and late (⬎15 days) repair. There were no intraoperative complications and no patient sustained vision loss or worsening of ocular motility as a result of surgery.
Table 5. Relationship of Trapdoor Fractures to Nausea/Vomiting and Motility Restrictions* Pure Internal Fractures
All Fractures
Trapdoor
Non-Trapdoor
With nausea/vomiting Without nausea/vomiting Total
5 3 8
2 12 14 P ⫽ 0.05
7 15 22
With motility restriction Without motility restriction Total
8 0 8
7 7 14 P ⫽ 0.02
15 7 22
Trapdoor
Non-Trapdoor
With nausea/vomiting Without nausea/vomiting Total
5 3 8
2 24 26 P ⬍ 0.01
7 27 34
With motility restriction Without motility restriction Total
8 0 8
7 19 26 P ⬍ 0.01
15 19 34
* Statistical relationship (contingency tables) for presence of nausea/vomiting and ocular motility restriction compared with presence of trapdoor fracture (for pure internal orbital fractures n ⫽ 22 and all fractures n ⫽ 34).
831
Ophthalmology Volume 107, Number 5, May 2000 Table 6. Mean Changes in Supraduction Limitation Based on the Time of Intervention* Time of Intervention (From Date of Injury in Days)
No. Patients
Mean Preop Supraduction Limitation
Mean Postop Supraduction Limitation
Mean Change in Supraduction Limitation
2 3 5
⫺3.0 ⫺0.6 ⫺3.4
0 0 ⫺1.7
3.0 0.6 1.3
0–2 3–14 15 & over (range 15–40, mean 21)
* Comparison of mean supraduction limitation preoperatively and postoperatively, stratified by timing of surgery for orbital floor fractures. Limitations in supraduction are indicated with a negative value, therefore a positive mean change means improvement in duction limitation.
The mean estimated blood loss was 10 ml (⫾SD 8). There were nine surgeries performed for the repair of floor fractures, one for medial wall fracture and one for combined floor and lateral wall fractures. An implant was placed over the fracture site in all cases (7 Gelfilm [Pharmacia-Upjohn, Kalamazoo, MI] and 4 Medpor) [Porex Surgical, College Park, GA]. Gelfilm was generally used when the fracture was narrow or the displaced segment of orbital bone could be repositioned in nearly anatomic location after release of entrapped orbital tissues (typical of most trapdoor fractures). Medpor was used for manifestly larger defects that required structural support. Although all patients showed improvement at the 3-month follow-up in their ocular motility, for analysis we evaluated the quantitative amount of supraduction limitation in the 10 patients with floor fractures who underwent surgery to compare the benefit of surgery based on the time of intervention. Table 6 summarizes the mean change in supraduction limitation using the same quantitative scale (0 to ⫺4), which we described earlier in the study. All patients having surgery within 14 days had complete recovery of supraduction, although the mean preoperative supraduction limitation and postoperative improvement was greater in the immediate (0 –2 day) surgical group. The late surgery group demonstrated some persistent (mean) supraduction limitation. Table 7 lists the characteristics of the patients who underwent surgery. Two illustrative cases from this series are described.
Horizontal ductions were intact. The coronal CT was reported by the radiologist to show a left orbital floor fracture. One of us (DRM) believed that there was distortion and probable entrapment of the inferior rectus muscle on the CT scan (Fig 2). The patient underwent orbital exploration the same day. Forced duction testing at the beginning of the surgery revealed marked restriction in attempted manual supraduction. At surgery a hinged, trapdoortype fracture was discovered in the medial aspect of the orbital floor, with soft-tissue pulled into the fracture site (Fig 3). The incarcerated tissue was freed and the inferiorly displaced portion of the fracture “sprang back” into nearly normal anatomic position. The fracture site was covered with a Gelfilm implant. Forced duction tests were repeated and the results were completely nor-
Case 1 An 8-year-old boy slipped on the kitchen floor and fell sustaining blunt trauma to his left eye. In the emergency room the boy had nausea and an episode of vomiting. On examination there was normal visual acuity, diplopia in primary gaze, and ⫺4 limitation of both infraduction and supraduction (Fig 1) in the left eye. Table 7. Surgical Group* Patient
Age (yrs)
Trapdoor (w/entrapment)
Timing (Days)
Floor
1 2 3 4 5 6 7 8 9 10 11
14 13 18 5 8 7 18 14 18 16 13
Yes Yes Yes Yes Yes No Yes No Yes Yes No
7 15 15 6 0 17 40 24 1 7 15
Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes
* Characteristics of patients who underwent surgery for internal orbital fractures. Note that 10/11 patients had orbital floor fractures.
832
Figure 1. Case 1: 8-year-old boy who fell, sustaining left orbital trauma. A, eyes in primary position. Note relative absence of orbital edema. B, note marked limitation of left eye in attempted supraduction.
Bansagi and Meyer 䡠 Pediatric Orbital Fractures
Figure 2. Case 1: Orbital coronal CT. A, anterior orbit shows distortion of inferior rectus muscle. B, posterior orbit shows trapdoor-type fracture with inferior rectus entrapment.
mal. There was minimal vertical duction limitation at the 1-week follow-up. By the third week after surgery, the patient had full extraocular muscle movements without diplopia (Fig 4).
Case 2 An 18-year-old boy sustained blunt trauma to his orbit after being struck in the left eye during a fight. He was evaluated in the emergency department by an emergency physician, treated, and released. An orbital x-ray film was reported as normal. The patient was referred for ophthalmologic evaluation the day after injury. The patient complained of persistent nausea and vomiting, photophobia, diplopia, and severe pain in the left eye on attempted upgaze. He had normal visual acuity, no enophthalmos, a left hypotropia in primary gaze with ⫺3 limitation in supraduction and ⫺2 limitation of infraduction in the left eye (Fig 5). A coronal CT scan was obtained, which showed a small left orbital floor fracture with soft-tissue entrapment and distortion of the left inferior rectus (Fig 6). It was suspected that the nausea and vomiting was the result of vagal reaction caused by pain and the traction on the inferior rectus muscle. The patient underwent urgent orbital exploration and repair. Forced duction tests were performed at the beginning of surgery, which showed marked restriction on attempted manual supraduction. A slightly displaced, narrow fracture was discovered on the medial aspect of the left orbital floor, with soft-tissue entrapment. The tissue was gently removed and the fractured was covered with a Gelfilm implant. Repeated forced
Figure 3. Case 1: Orbital fracture repair. A, markedly positive forced duction test. B, trapdoor fracture with soft-tissue incarceration in medial aspect of the orbit floor was confirmed.
ductions showed no residual restriction. The patient felt better immediately after surgery, and had no further nausea or vomiting. Mild residual extraocular motility restriction in supraduction and infraduction was noted at the 1-week follow-up, which fully resolved by the 1-month follow-up (Fig 7).
Discussion Our understanding and management of internal orbital (“blowout”) fractures is evolving.5–10 Previous management philosophies have ranged from immediate surgery on all blowout fractures to surgery on none. Current generally accepted indications for surgical repair after an internal orbital fracture are muscle entrapment resulting in ocular motility restriction with diplopia, early enophthalmos (⬎ 2 mm), and orbital defects involving more than 50% of the floor or medial wall.5,6,9 Traditional recommendations for orbital blowout fracture management in the adult population include a 2-week observation period to allow edema to subside to help determine the relative contribution of orbital edema to ocular motility restriction before considering surgical intervention.7–9 More recent recommendations indi-
833
Ophthalmology Volume 107, Number 5, May 2000
Figure 6. Case 2: Coronal CT shows soft-tissue entrapment and distortion of the inferior rectus muscle within a trapdoor-type fracture of the medial orbital floor.
Figure 4. Case 1: 3 weeks postoperatively. A, good globe position in primary gaze. B, full return of ocular motility.
cate the need for careful evaluation and the consideration of earlier surgery with evidence of symptomatic diplopia and CT evidence of entrapment. In cases with obvious clinical signs and CT evidence of frank muscle entrapment, repair within 48 hours has been advocated.5,6,9 Recent publications from several disciplines evaluating the pediatric pop-
Figure 5. Case 2: 18-year-old boy struck in the left eye had pain on attempted upgaze and persistent nausea and vomiting. Marked restriction in supraduction of the left eye was noted.
834
ulation have suggested an increased potential for trapdoortype fractures in this young age group that may require expeditious repair to restore motility and reduce the possibility of tissue ischemia and necrosis.1– 4,12,13 The frequency of this phenomenon, however, is not well-established. In our study we found that trapdoor fractures occurred in 8 of 34 patients (24%) with internal orbital fractures in the pediatric population. All trapdoor fractures occurred in the subgroup of patients with pure internal orbital fractures (i.e., those not involving the orbital rim), and the frequency for this subgroup was thus higher (8 of 22 [36%]). These percentages are clinically important but are less than the 93% reported by de Man and coworkers.1 Most trapdoor fractures involved the orbital floor (7 of 8 patients [88%]). The orbital floor was also the most common wall of the orbit involved in the pediatric population with internal orbital fractures (27 of 38 [71%]). We found that trapdoor-type fractures were strongly associated with ocular motility restriction. Indeed, all eight patients with trapdoor fractures demonstrated reduced ocular motility. In the subset of patients with floor fractures, where trapdoor fractures tended to occur, there was much greater supraduction limitation in patients with trapdoor
Figure 7. Case 2: 1 month postoperatively full ocular motility was noted.
Bansagi and Meyer 䡠 Pediatric Orbital Fractures fractures than in those without trapdoor-type fractures. The limitation in supraduction was usually quite marked, with horizontal ductions preserved or minimally affected. Typically, orbital edema was relatively minimal in these cases and not sufficient to explain the motility pattern. These findings are in agreement with the observations of previous authors.1– 4,12,13 The risk of soft-tissue entrapment in trapdoor fractures is likely increased in pediatric patients with pure internal orbital fractures, where the more elastic, flexible bone can crack, forming a narrow circular or elliptical fracture, hinged on one side, that traps soft tissue within the depressed segment. We have found that trapdoor fractures of the orbital floor tend to be hinged on the medial side, with the depressed portion of the fracture adjacent to the course of the infraorbital neurovascular bundle (Figs 2 and 6). With complex or impure internal orbital fractures, involving both the internal orbit and the orbital rim or other facial bones, the dispersion of force over a larger area probably results in less likelihood of soft tissue entrapment. We also found a strong association between nausea and vomiting and trapdoor fractures. We identified seven patients with nausea and/or vomiting, all with pure internal orbital fractures; five were of the trapdoor type. Although nausea and vomiting alone is certainly not an indication for urgent surgical intervention, this finding coupled with markedly restricted extraocular motility should alert the physician to the increased possibility of trapdoor fracture with extraocular muscle entrapment in the pediatric population. We believe that nausea and vomiting may be a vagally mediated response to pain or other sensory feedback associated with extraocular muscle traction. Sires et al4 described three pediatric patients with trapdoor fractures and entrapment of the inferior rectus muscle, resulting in oculocardiac reflex and the triad of bradycardia, nausea, and syncope. One patient showed hypoxia of the entrapped muscle after only 7 hours from the time of injury. Another patient whose surgical intervention was delayed 72 hours showed ischemic necrosis of the entrapped muscle.4 The degree of hypoxia or ischemic necrosis of the entrapped tissue can be difficult to assess during orbital surgery. We cannot say with certainty whether any of our patients experienced these consequences. Although statistical comparison was not possible because of the small number of patients in the surgical subgroups, we found that the degree of recovery of ocular motility was more complete in the group of patients who had surgical intervention in less than 2 weeks compared with the delayed surgical intervention group (⬎15 days.). Those operated on immediately (within 2 days of injury) had greater ocular motility restriction than those operated on early (3–14 days after injury), yet still had complete recovery. None of the patients in our study reported syncope. Aside from the initial emergency department evaluation of vital signs that included pulse rate, we did not monitor our patients’ cardiac rhythm. Therefore, we cannot say with certainty that any of our patients demonstrated complete oculocardiac reflex. Nevertheless, the mechanism is probably closely related, and physicians managing pediatric orbital trauma should be aware of these phenomena and their implications for evaluation and treatment. Coronal CT scans are an accepted and useful method of identifying orbital fractures and possible extraocular mus-
cle/soft-tissue entrapment. The sensitivity of coronal CT scans in identifying tissue entrapment depends on the quality of the scan, the position and size of fracture, and the acumen of the radiologist or the physician evaluating the CT scan. Bower Wachler and Holds12 reported two cases of inferior rectus muscle entrapment with orbital trapdoor fractures in young patients that were missed radiographically. Jordan and coworkers,13 in a study of 20 pediatric patients with orbital fractures from four institutions, also noted the relative scarcity of CT findings with small or trapdoor fractures, despite clinical motility restriction and surgically confirmed entrapment. Our study emphasizes the importance of careful review of coronal CT scans for the presence of both fractures and entrapment, especially in the pediatric population, that appear predisposed to narrow or trapdoortype fractures. These fractures may be small and the amount of tissue incarceration minimal, making the findings on CT quite subtle at times. CT findings should be carefully evaluated and combined with additional clinical information before ruling out fracture and entrapment. Signs of significant ocular motility restriction and nausea/vomiting in the pediatric population should alert the physician to the increased possibility of a narrow or trapdoor-type fracture with entrapment. A markedly positive forced duction test is suggestive of entrapment but can be difficult to perform in young poorly cooperative patients. Like most retrospective trauma studies, the limitations of our study must be acknowledged. Because of the close association of our ophthalmology department with a level I trauma center (Albany Medical Center), our study likely reflects referral bias and may include more severe cases of orbital trauma based on the type of patients transferred to such trauma centers. There are also other surgical subspecialties (e.g., oral surgery, otolaryngology, general plastic surgery, and neurosurgery) involved in the care of patients with craniomaxillofacial trauma that influence the initial evaluation (“triage”) and follow-up of such patients. At our institution, however, the ophthalmology service is primarily responsible for the evaluation and management of all patients with pure internal orbital fractures, the largest subgroup and focus of this study and the group most ophthalmologists elsewhere are called on to evaluate and help manage. Our study does confirm the suggestion that many young patients who sustain internal orbital fractures frequently have trapdoor-type fractures with soft-tissue entrapment and ocular motility impairment. These trapdoor fractures most commonly involve the orbital floor and demonstrate a corresponding limitation in vertical motility (particularly supraduction) that is typically disproportionate to the amount of orbital edema and limitation in other fields of gaze. Because edema is usually not a prominent feature, there tends to be little improvement in ocular motility after 1 to 2 weeks. We also found that nausea and vomiting was a common finding highly predictive of trapdoor-type fracture in the setting of pediatric orbital trauma. Although this study cannot unequivocally answer the question of ideal surgical timing, we agree with the recent suggestions of others that in the setting of documented or high-probability trapdoor fractures with restricted ocular motility accompanied by nausea/vomiting or complete oculocardiac reflex, earlier surgical intervention should be considered in this pediatric
835
Ophthalmology Volume 107, Number 5, May 2000 population. Prompt intervention may protect these patients from ischemic injury to incarcerated orbital soft tissues,4,12,13 and, as described by Sires and coworkers,4 in cases of complete oculocardiac reflex, may prevent rare, but potentially fatal, arrhythmias. Acknowledgment We thank Dr. Sanjayant R. Chamakura for his assistance in the data collection for this manuscript.
6. 7. 8.
References
9.
1. de Man K, Wijngaarde R, Hes J, de Jong PT. Influence of age on the management of blow-out fractures of the orbital floor. Int J Oral Maxillofac Surg 1991;20:330 – 6. 2. Anderson PJ, Pole MD. Orbital floor fractures in young children. Craniomaxillofac Surg 1995;23:151– 4. 3. Koltai PJ, Amjad I, Meyer DR, Feustel PJ. Orbital fractures in children. Arch Otolaryngol Head Neck Surg 1995;121:1375–9. 4. Sires BS, Stanley RB, Levine LM. Oculocardiac reflex caused by orbital floor trapdoor fracture: an indication for urgent repair. Arch Ophthalmol 1998;116:955– 6. 5. Meyer DR. Orbital fractures. In: Tasman W, Jaeger EA, eds.
10. 11. 12.
13.
Duane’s Clinical Ophthalmology, rev. ed. Philadelphia: Lippincott-Raven, 1996; chap. 48. Rubin PA, Bilyk JR, Shore JW. Management of orbital trauma: fractures, hemorrhage, and traumatic optic neuropathy. Am Acad Ophthalmol Focal Points 1994;XII(7):1– 8. Levin LM, Kademani D. Clinical considerations in the management of orbital blow-out fractures. Compendium 1997;6: 593– 600. Wilkins RB, Havins WE. Current treatment of blow-out fractures. Ophthalmology 1982;89:464 – 6. Dutton JJ, Manson PN, Iliff N, Putterman AM. Management of blow-out fractures of the orbital floor. Surv Ophthalmol 1990;35:279 – 8. Hawes JM, Dortzbach RK. Surgery on orbital floor fractures. Ophthalmology 1983;90:1066 –70. Carr RM, Mathog RH. Early and delayed repair of orbitozygomatic complex fractures. J Oral Maxillofac Surg 1997;55:253–8. Bower Wachler BS, Holds JB. The missing muscle syndrome in blowout fractures: an indication for urgent surgery. Ophthalmic Plast Reconstr Surg 1998;14:17–18. Jordan DR, Allen LH, White J, et al. Intervention within days for some orbital floor fractures: the white-eyed blowout. Ophthalmic Plast Reconstr Surg 1998;14:379 –90.
Discussion by Mark R. Levine, MD The overall indication for surgical intervention for blowout fractures remains controversial. Indications for conservative surgery within 2 weeks of the trauma include symptomatic diplopia with positive forced ductions, computed tomographic (CT) evidence of muscle entrapment with no clinical improvement over 1 to 2 weeks, enophthalmos of 3 mm or more, significant hypoophthalmos, and an orbital wall defect involving more than 50% of the floor.1 Conversely, the need for immediate surgery does arise occasionally, as in the case of globe herniation into the maxillary sinus. Another study found that small orbital fracture defects with incarcerated extraocular muscles lead to a compartment syndrome that causes muscle ischemia, fibrosis, and motility restriction.2 Bansigi and Meyer describe an important subset of 34 pediatric patients (1 to 18 years of age), 71% of whom had orbital floor fractures. Trapdoor fractures were associated with more severe motility restriction than non-trapdoor fractures, and nausea and vomiting were associated presenting symptoms in 21% of cases. Interestingly, CT evidence was not reported by the radiologist in nearly half the patients with trapdoor fractures. The authors conclude that ocular motility recovery was more complete in patients who had surgical intervention in less than 2 weeks compared with those who had surgery after 2 weeks. Although they caution that their study cannot unequivocally answer the question of ideal surgical timing, they do agree that early surgical intervention should be considered when patients have the previously mentioned criteria. Bansagi and Meyer make a strong case for early surgical intervention in pediatric patients with trapdoor fractures, motility restriction, nausea and vomiting. The small number of patients (n ⫽ 2) operated on within 48 hours had a more rapid motility recovery than those within 2 weeks or longer. They note that
From the University Hospitals of Cleveland, Case Western Reserve University School of Medicine, Cleveland, Ohio. Address correspondence to Mark R. Levine, MD, Mt. Sinai Medical Building, 26900 Cedar Road, Suite #311, Beachwood, OH 44122.
836
earlier evaluation might have been better able to address appropriateness of surgical timing if they could have evaluated the patients more expeditiously (initial examination 7.18 days with surgical intervention 13.4 days after injury) and surgically intervened much earlier. Therefore, the conclusion of early surgical intervention was based on less than the 2-week period that many of us already accept as appropriate surgical intervention. Their article does, however, support articles written by Sires3 and Jordan for early intervention in the pediatric patient to improve surgical outcomes and prevent muscle ischemia resulting in fibrosis and motility restriction. We have learned that, because of their softer, more flexible orbital bones, trapdoor fractures are more likely in the pediatric age group compared with adults. The harder, more brittle orbital bones of the adult usually result in breaks in several areas, causing a buckle effect. Moreover, in the pediatric patient in whom tissue (fat, connective tissue, and muscle) is incarcerated in a flexible trapdoor fracture, the potential space around the extraocular muscles is disrupted, and a high compartment pressure occurs around the inferior rectus muscle resulting in ischemia. Therefore, surgical intervention within 24 to 72 hours in the pediatric patient who has marked motility restriction, trapdoor fracture, and nausea and vomiting may restore circulation to the rectus muscle and perimuscular tissue, preventing fibrosis and enhancing extraocular muscle recovery. A randomized study of trapdoor fractures with tissue incarceration operated on at different time intervals might be helpful. References 1. Jordan DR, Allen LH, White J, et al. Intervention within days for some orbital floor fractures: the white-eyed blowout. Ophthalmic Plast Reconstr Surg 1998;14:379 –90. 2. Smith B, Lisman RD, Simonton J, et al. Volkman’s contracture of the extraocular muscles following blowout fractures. Plast Reconstr Surg 1984;74:200 –5. 3. Sires BS, Stanley RB, Levine LM. Oculocardiac reflex caused by orbital floor trapdoor fracture: an indication for urgent repair. Arch Ophthalmol 1998;116:955– 6.