Critical Degree of Orbital Floor Displacement Drives Operative Repair of Zygomaticomaxillary Complex Fractures: Findings from a 10-Year Retrospective Study

Introduction

The prominent position of the zygoma makes it particularly vulnerable to blunt force trauma. From a clinical standpoint, ZMC fractures may result in trismus, malar depression, diplopia, enophthalmos, globe displacement and paresthesias of the cheek, nose, upper lip and lower eyelid.¹ Asymptomatic and minimally displaced fractures without functional sequelae may be observed. However, even minimally displaced fractures may result in aesthetic and functional deficits ultimately requiring surgical intervention.² For fractures requiring surgical management, prior to the adoption of rigid internal fixation, 3-point wire fixation was historically employed but proved challenging to maintain nondisplaced fracture ends and to reduce and fixate certain small fragments.³ Current surgical technique most often entails miniplate fixation at one or more fracture sites with repair or reconstruction of the orbital floor as required.^2,4

The wide body of literature on ZMC fractures suggests that regional geography plays an important role in prevalence, severity, etiology and management of fractures, though few studies have described specific variables associated with surgical versus nonsurgical management of zygomatic complex fractures.^2,5 This retrospective study describes all ZMC fracture patients managed at an academic medical center in New York City between 2008 – 2017. The purpose of this study was to characterize the distinction between operative and non-operative management of ZMC fractures in an urban setting. The investigators hypothesize that there will be clinically significant differences in the characteristics of ZMC fractures managed operatively versus non-operatively. The specific aims of this study were to describe patient and injury characteristics, to determine clinical and radiographical variables associated with surgical versus nonsurgical management of ZMC fractures, and to identify general trends in operative management among various subspecialty services.

Methods:

This was a single-center, retrospective study of patients diagnosed with a zygomatic complex (ZMC) fracture at New York-Presbyterian/Weill Cornell Medical Center between January 2008-December 2017. We identified all patients with maxillofacial fractures who presented to New York-Presbyterian/Weill Cornell and were managed by one of three consulting services including Otolaryngology–Head and Neck Surgery, Oral and Maxillofacial Surgery, and Plastic Surgery. Data regarding the patients was captured via electronic medical record (EMR) query, which was verified and supplemented by manual chart review. We included all patients who had fractures of the zygomatic complex, defined as fractures involving all of the following sites: the lateral orbital wall (either at the zygomaticofrontal suture superiorly or zygomaticosphenoid suture inferiorly), the anterior maxilla and inferior orbital rim, the zygomatic arch, and the orbital floor. Patients were excluded if they did not have a ZMC fracture or left the hospital prior to evaluation. Those with isolated fractures of the zygoma, orbital floor, zygomaticofrontal suture, or lateral orbital wall were also excluded as they did not involve the entire tripod or terapod complex.

Data collected included age, sex, demographic characteristics, mechanism of injury, initial clinical examination, radiographic findings including degree of fracture displacement, presence of muscle and/or fat entrapment, comminution, and type of repair if surgical treatment was pursued. Radiology reports from all available computed tomography (CT) scans were reviewed and any missing relevant data were noted. Nominal data were presented as frequencies and metric data as means and standard deviations. The N-1 chi-square test and Fischer’s Exact tests were employed for comparisons between treatment groups for nominal data. P values < 0.05 were considered statistically significant. Statistics were performed with R (R Core Team 2020 for Mac, R Foundation for Statistical Computing, Vienna, Austria, www.R-project.org) and GraphPad Prism (GraphPad Prism version 9.0.0 for Mac, GraphPad Software, San Diego, California USA, www.graphpad.com) software. All elements of this study were approved by the Weill Cornell Medicine Institutional Review Board.

Results

Between 2008 and 2017, 1914 patients with 3944 maxillofacial fractures presented to New York-Presbyterian/Weill Cornell. Of these, 196 patients with fractures of the zygomatic complex (5.0%) were evaluated. The average age was 45.7 ± 22.2 years, and 138/196 (70.4%) of patients were male (Figure 1). The most common mode of presentation was by ambulance to the emergency room (37.9%) followed by direct referral from a physician office (28.7%). Patients who presented with ZMC fractures had an average of 1.84 maxillofacial fractures. The most common associated injury on presentation with a ZMC fracture was a facial laceration (62/196, 31.6%), which was similar to the overall facial fracture cohort (622/1914, 32.5%). Intracranial complications (including intracranial hemorrhage/contusion, cerebrospinal fluid leak, or cranial vault fracture) were the second most common associated injury (31/196, 15.8%), compared to the entire facial fracture cohort incidence of 24.8% (475/1914 patients). High kinetic energy trauma, defined as falls >6ft or motor vehicle accidents, comprised 42/196 (21.4%) of all ZMC fractures (Table 1). Of patients with low kinetic energy injury related fractures, 90/154 (58.4%) required surgery, compared to 29/42 (69.0%) of patients with high kinetic energy trauma requiring surgical intervention. There was no significant difference in the proportion of patients who underwent surgery in the high vs. low kinetic energy related ZMC fracture groups.

Figure 1.

Associated Injuries of Patients Presenting with ZMC Fractures

Table 1.

Demographic Data

Demographics	N(%)
Total # patients	196
Total # facial fractures	3944
Total #ZMC fractures (%)	196 (5.0)
Total # zygomatic arch only fractures (%)	222 (5.6)
Average age	45.7 ± 22.2
#Male (%)	138 (70.4)
Mode of Presentation (%)
Brought in by Ambulance	74 (37.9)
ED Walk-In	45 (23.1)
Transfer from local hospital	1 (0.5)
Transfer from outside hospital	12 (6.1)
Referral from MD office	56 (28.7)
Unknown	7 (3.6)
Associated Injuries (%)
Facial Laceration	48 (31.6)
Intracranial Complications	24 (15.8)
Injury Mechanism
High Kinetic Injury: Falls>6ft +MVA	42 (21.4)
Low Kinetic Energy Injury	154 (78.6)

Three hospital services managed patients with ZMC fractures: Plastic Surgery (PS) (n=62, 31.6% of cases). Otolaryngology–Head and Neck Surgery (OtoHNS) (n=37,18.9%) and Oral and Maxillofacial Surgery (OMFS) (n=60, 30.6 %). Collaborative management between multiple teams occurred for 34 cases (17.5%), and consult data were unavailable for three cases. The PS service had the highest rates of independent operative management (46/62, 74.2%) compared to OtoHNS (23/37, 62.1%, p=0.20) and OMFS (34/60, 49.2%, p=0.004). The PS service had the highest number of patients referred from the office for evaluation (23/62, 37.1%) compared to the OtoHNS (12/37, 32.4%) and the OMFS (7/60, 11.4% p<0.05) services.

A total of 121 ZMC fractures were treated surgically (61.7%). The average age of patients undergoing surgery was significantly lower than those observed (38.9 ±18 years vs. 56.1 ±23.5 years respectively, p<0.0001). There were no significant gender differences between surgical and non-surgical groups with ZMC fractures. All patients who presented with significant concurrent ophthalmologic exam findings (globe injury, blindness, retrobulbar complications, entrapment/restricted gaze, or enophthalmos) with ZMC fracture (n=17, 8.7% cases) were managed surgically.

Cases managed surgically used a range of five different approaches for open reduction/internal fixation (ORIF). The most common approach to the ZMC was the sublabial/gingivobuccal (31.9% of all approaches) followed by lateral brow incision (20.1%) and transconjunctival approaches (18.1%, Figure 1). A total of three approaches were most common (44/121 cases), and the most common approaches employed (17/44) were the combined lateral brow, transconjunctival, and sublabial incisions (Figure 2). When a single approach was used, the sublabial (5/16) or Gilles/temporal (4/16) incisions were most often cited. Fourteen patients reported V2 paresthesia, which was present in 9/14 patients pre-operatively and noted to be new in 5/14 patients post-operatively. One patient exhibited a prolonged postoperative ventilator requirement (>48 hours), and another patient developed a deep venous thrombosis postoperatively. Seven patients (7/121, 5.8%) experienced wound complications ranging from superficial surgical site infections to wound dehiscence. Thirty-two patients were lost to follow up. There were no other significant immediate postoperative complications reported.

Figure 2A.

Distribution of surgical Approaches to ZMC Fracture Repair

Oculomotor muscle herniation, muscle entrapment, and the presence of fracture comminution were examined from available CT maxillofacial imaging reports. A total of 137 written reports were available for review. A significant proportion of patients (71%) had radiologic reports or clinical notes with transcribed relevant radiologic findings but were lacking raw imaging data which could be reviewed to measure degree of displacement and percentage of the orbital floor involved. Of the 106 patients who underwent surgery with available imaging reports, six patients had CT findings of muscle entrapment, whereas 0/31 patients who were observed had CT findings of muscle herniation (Table 2). In the surgery group, 55/106 (52%) patients with available imaging reports had comminuted fractures, whereas only 8/31 (26%, p=0.011) in the observation group had comminuted fractures.

Table 2.

Radiographic Findings of ZMC Fractures and Clinical Management: Features of CT maxillofacial imaging between surgery and observation groups.

	Observation N (%total observed)	Surgery N (%total surgery)	P value
Orbital Floor Comminution	8 (26)	55 (52)	0.011*
Muscle Entrapment	0 (0)	6 (6)	0.180
Orbital Floor Displacement (any)	17 (54)	76 (72)	0.079
Orbital Floor Displacement (>=4mm)	4 (21)	18 (49)	0.045*

^*Indicates statistical significance as detailed by predetermined p-value <0.05.

Among ZMC fractures the extent of orbital floor involvement was analyzed as well as the average degree of displacement on both coronal and sagittal CT images (Figure 3). In those with raw imaging data available, the absolute maximum displacement of the orbital floor was 5.7 ±3.2mm in groups undergoing surgery, vs 2.9 ±1.5mm in the observation group (p=0.001). On subgroup analysis, orbital floor displacement ≥4mm was predictive of surgical intervention. Specifically, 18/22 (81.8%) patients with at least 4mm displacement underwent surgical repair in contrast to the 19/34 (56%) patients with floor displacement of less than 4mm who underwent surgical repair.

Figure 3:

Sample markedly comminuted left ZMC fracture with depression and rotation of the posterior zygoma fracture fragment. The left lamina papyracea is fractured. There is a depressed fracture through the inferior orbital floor medially. Orbital floor displacement was calculated by raters using the maximum vertical distance between the orbital floor along its axis and the fracture segment.

Discussion

The purpose of this study was to characterize the distinction between operative and non-operative management of ZMC fractures in an urban setting. The specific aims of this study were to describe patient and injury characteristics, to determine clinical and radiographic variables associated with surgical vs. nonsurgical management of ZMC fractures, and to identify general trends in operative management among various subspecialty services. Of the 196 patients with zygomatic complex fractures in this study, the majority of patients were males in their 40s, similar to other cross sectional ZMC studies described in the literature.^5,6 Falls were the most common etiology observed in our population, which differs from other geographic regions in which the predominant etiologies are motor vehicle accidents or physical assault.^2,6–8 No clear trends in fracture incidence were noted by year between 2008 and 2017. The average age of patients with ZMC fractures undergoing surgery was significantly lower than those managed with observation alone (38.9 ±18.3 vs. 56.1 ±23.5 years, p< 0.0001). The higher mean age of patients managed by observation alone may be secondary to an overall higher rate of medical comorbidities or a higher threshold to address aesthetic concerns in this older population.

Among ZMC fractures, the PS service was more likely to operate (74.2% of total ZMC fracture patients) than OMFS (49.2%, p=0.004) or OtoHNS (62.1%). There was a relatively high proportion of patients in this cohort directly referred to our institution from a physician’s office (28.1%), which may have increased the operative rate given that patients referred to PS may have been referred specifically for operative management. Additionally, this may reflect a higher proportion of patients with aesthetic concerns interested in surgical intervention presenting to the PS’s office relative to the other two services.

While there was a significant proportion of patients with ZMC fractures with associated intracranial complications (15.8%), this is lower than the incidence of concomitant intracranial complications in the entire cohort of patients with maxillofacial fractures (24.8%). In our study cohort, there was a relatively low proportion of high kinetic injury ZMC fractures (21.4%) which may explain the lower incidence of associated intracranial complications. This may reflect New York City’s highly urban landscape and associated relative dearth of high-speed motor vehicle accidents compared to other regions. In particular, the scarcity of interstate highways, regional policies such as a comprehensive speed limit reduction to 25 miles per hours throughout New York City in 2014, and New York’s status as the first state to mandate seat belt use for drivers and front seat passengers, may all contribute to the relatively low proportion of high kinetic injury ZMC fractures.^9,10 However, there did not appear to be a significant difference in the rate of surgery in high vs. low kinetic energy related ZMC fractures. There is sparse literature reviewing low vs. high kinetic energy related ZMC fractures, but our data show that more than half of these low energy injuries required surgery.

Of the 121 patients treated surgically, 12% (n = 14) reported V2 paranesthesia either pre- or post-operatively. The incidence of V2 paresthesia seen in this cohort is lower than previously reported data on trigeminal nerve dysfunction in ZMC fractures which varies widely from 52–100% at presentation and 7.7–55% in patients treated surgically.^11,12

While several authors describe clinical signs including persistent diplopia, increased orbital pressure, enophthalmos, visual impairment, extraocular movement limitations and hypoesthesia of the infraorbital nerve as indications for operative management, radiographic criteria for operative repair are limited.¹³ Using 3-dimensional CT analysis, Pau et al. showed that patients who required operative repair of ZMC fractures had greater cumulative displacement of the malar eminence than those who did not require operative repair.¹⁴ Based on a retrospective review examining fracture patterns in 94 patients with isolated unilateral ZMC fractures managed surgically, Anehosur et al. advocated for exploration of the orbital floor when the fracture line passes medial to the infraorbital foramen.¹⁵ Even so, very few studies have critically sought to determine a “cutoff” displacement for which operative intervention is recommended. In our study, the average displacement of the orbital floor was 5.7 ±3.2 mm in groups undergoing surgery, vs 2.9 ±1.5mm in the observation group (p=0.001). Furthermore, patients with greater than or equal to 4mm of displacement of the orbital floor on CT were more likely to receive surgical treatment than observation (81.8% vs. 56%, p=0.045). Many additional clinical and radiographic factors may also be predictive of operative intervention along with orbital floor displacement, as 56% of patients with <4mm displacement also required surgery. Traditional indications for isolated orbital floor fractures include disruption of greater than 50% of the orbital floor or at least 2mm of pre-operative enophthalmos. Alinasab et al. additionally considered the size of the floor disruption, the volume of soft tissue herniation and the distance to the posterior-most edge of the fracture in predicting delayed enophthalmos.¹⁶ The lateral rotational movement of the ZMC often required would be expected to further worsen orbital floor defects seen on pre-operative imaging, so the more conservative standard of >= 4mm orbital floor displacement in ZMC fractures in our study is quite reasonable. It is likely that several contributing factors on presentation are positively predictive of open repair, and future multivariate modeling is needed to identify relative predictive power of these covariates.

Limitations of this study include the substantial proportion of imaging unavailable for secondary review and measurements. In addition, as our study is limited to the immediate hospital and outpatient follow up records, it is impossible to accurately determine what percentage of the early non-operative patients ultimately required surgery for late ophthalmic or aesthetic concerns. As a tertiary care center, most patients did not follow up long term within our system and may have sought additional care at a later date within the New York metropolitan area or beyond. While there is a lack of standardized follow up, the reported complications at known follow up were minimal, which is in line with several other retrospective ZMC fracture studies.^8,13,17 Additionally it has been noted that non-algorithmic techniques for measuring orbital floor defects may incur higher interrater variability when compared to algorithm driven methods. Orbital floor displacement was measured in the greatest dimension in the coronal and sagittal image planes by two independent reviewers, however this methodology was subject to the bias of the reviewers. ^18–20 3D volumetric analysis would perhaps more accurately assess the degree of orbital volume change and might more accurately predict the need for surgery in more complex cases.

Conclusion:

Established surgical indications for ZMC fractures include ophthalmologic complications and concerns for long lasting cosmetic deformities. In this cohort, patients more likely to undergo surgical reduction were young patients with ophthalmologic symptoms on presentation and >4mm displacement of the orbital floor. The data from this review also suggest that low kinetic energy ZMC fractures may warrant surgical management as often as high energy ZMC fractures in urban centers, which is a considerable finding given the nature of maxillofacial injury in large urban communities. While orbital floor comminution has been shown to be a predictor for operative reduction, in this study we also demonstrated a difference in rate of reduction based upon severity of bony floor displacement. This may have significant implications in both the triage and selection of patients most suitable for operative reduction.

Figure 2B.

Distribution of No. Surgical Approaches per Patient