Radiographic Indications for Orbital Blowout Fracture Reconstruction: A Critical Review of 2 Decades of Literature

Abstract

Background:

Clinical criteria (eg, persistent diplopia, enophthalmos) are clear indications for surgery in orbital blowout fractures (OB-OFs). A variety of additional radiographic indications are often described (eg, fracture size, orbital volume changes, inferior rectus muscle [IRM] rounding and displacement) but their significance varies across studies. This systematic review aimed to evaluate the reliability of these radiographic criteria.

Methods:

Following Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines, a PubMed search (2000–2024) was conducted to assess studies analyzing radiographic indications for surgery in OB-OFs. We included only studies where surgery was based exclusively on clinical grounds with radiographic correlation established retrospectively, or cohorts followed up conservatively until the development of symptoms. Identified studies were assessed for aim, design, methods, bias, radiographic criteria, surgery, and study outcomes.

Results:

Out of 683 articles, only 6 were eligible. Fracture size was found to be weakly predictive of symptoms or the need for surgery, with variable cutoff values (≥42%, >2.3 cm², >3 cm²). Orbital volume changes showed conflicting results. Herniated volume demonstrated contradictory outcomes with variable cutoff values (>0.531, 1.3–2.2, or >2.2 cm³). IRM rounding was either weakly predictive or strongly associated with symptoms. IRM displacement, however, was found to be consistently predictive of symptomatic fractures.

Conclusions:

Clinical criteria remain the most reliable indications for OB-OF reconstruction. Caudal IRM displacement was consistently predictive of symptomatic fractures. Until a consensus is reached on a consistent methodology to calculate volume and fracture area, the reported cutoff values may be insufficient to guide treatment independently.

Takeaways

Question: Are radiographic criteria (eg, fracture size, orbital volume, inferior rectus muscle [IRM] rounding and displacement) reliable to guide treatment?

Findings: Except for IRM displacement, all other criteria lacked enough evidence to guide treatment independently.

Meaning: Except for IRM displacement below the orbital floor, other published radiographic criteria lack enough evidence to guide treatment. Clinical criteria continue to be the most reliable to guide treatment consistently.

INTRODUCTION

Orbital fractures constitute the chief cause of eye trauma hospitalizations in the United States^1–3 and form 13% to 15% of all facial fractures.^4,5 Orbital blowout fractures (OB-OF), in particular, can result in diplopia, limitation of gaze, and blurring of vision secondary to inferior rectus muscle (IRM) entrapment, enophthalmos (Fig. 1), and hypoglobus.⁶

Fig. 1.

Worm’s-eye view of a male patient with left orbital floor fracture. Note significant enophthalmos of the left eye.

Persistent diplopia, failure of upward gaze, and symptomatic enophthalmos are established “clinical” indications for surgery. When these are absent and minimal displacement is evident on computed tomography (CT) scans, patients are managed conservatively. Nevertheless, clinical signs are frequently masked by initial swelling (Fig. 2) and at times take 2 months or longer to manifest.^7,8 Waiting for confirmatory clinical signs risks fracture healing and scarring of prolapsed tissues, rendering surgery challenging and potentially hazardous (Fig. 3).^6,9 Thus, “radiographic” findings predictive of subsequent symptom development are pursued to guide treatment. Nevertheless, inaccurate radiographic indications may expose patients to unnecessary surgery and potential complications.^10–12

Fig. 2.

Male patient with a left orbital floor fracture documented by CT scan. Note significant left orbital swelling, precluding accurate assessment of clinical criteria.

Fig. 3.

Intraoperative view of a right orbital floor fracture. Note the floor defect (white arrow) visible after easy dissection of prolapsed orbital tissue, given early intervention.

Despite substantial literature, published radiographic indications vary across studies (Table 1). The most frequently cited include fracture size,^13,15,16,19 degree of herniation into the maxillary sinus,^17,22 orbital volume changes,^17,22 IRM rounding,^23,21 and caudal IRM displacement (Fig. 4).^9,22 Furthermore, fracture sizes of 1–2 cm^2,13,24 greater than 2 cm^2,16,17 and fractures involving more than 50% of the orbital floor¹⁹ have all been described in publications across surgical disciplines (ophthalmology, plastic surgery, craniomaxillofacial surgery).^{13,15,19,24,14} These imaging-based parameters are presented as predictive of subsequent development of enophthalmos and employed as indications for surgery.^16–18 Nevertheless, multiple studies report conflicting evidence regarding their accuracy and usefulness.^{10–12,25,26}

Table 1.

Radiographic Indications for Surgery Described in the Literature

Criterion	Reference	Notes	Origin
Fracture size 1–2 cm2	Aldekhayel et al 201413	Review and survey	Canada
Fracture size > 2 cm2	Patel et al. 202214 Shokri et al 201915 Gart and Gosain 201416 Manson et al 198617	Review Review Review CT scan analysis study	United States United States United States United States
Defect > 50% of the wall behind the equator of the globe	Kim and Jeong 201618	Review	South Korea
More than one-half of the orbital floor	Hawes and Dortzbach 198319 Patel et al 202214	Retrospective study Review	United States United States
Orbital prolapse > 1 cm in coronal CT scan	Polligkeit et al 201320	Prospective study of an endoscopy-assisted combined subciliary/transantral approach	Germany
Orbital volume expansion of 5%	Manson et al 198617 Patel et al 202214	CT scan analysis study Review	United States United States
Rounding of IRM: height-to-width ratio of 1.00 or greater	Matic et al 200721	Retrospective study	Canada
Caudal displacement of IRM below the orbital floor (entrapped) Grade III IRM displacement (below orbital floor)	Gilbard et al 198522 Scolozzi et al 20209	Prospective study Prospective study	United States Switzerland

Fig. 4.

Coronal CT scan of a left orbital floor fracture. Note radiographic criteria: size of fracture, herniation of orbital contents into the maxillary sinus, orbital volumetric expansion compared with the contralateral orbit, IRM rounding compared with the contralateral side (red arrows), and caudal displacement of the IRM compared with the level of the orbital floor (red dashed line).

Thus, there is a need for consensus on precise and reliable radiographic criteria that can guide treatment consistently.^{10,11,15,14,27} Hence, this article aimed to systematically review the last 2 decades of literature to understand the predictive value of different radiographic indications for OB-OF reconstruction. We sought to include only studies in which the decision to operate was based exclusively on clinical grounds, with radiographic correlation established retrospectively. This specific design allows for a precise vetting of radiographic parameters against established clinical criteria.

METHODS

The research strategy was developed in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines. A search of the PubMed database sought original, peer-reviewed studies investigating OB-OF in patients aged 16 years or older from 2000 to 2024. Eligibility criteria included articles with a sample size greater than 10 patients in which the decision to operate was based exclusively on clinical grounds, with radiographic correlation established retrospectively. Additionally, articles reporting cohorts managed conservatively and followed up clinically to analyze imaging criteria against the development of clinical signs were sought and included. Exclusion criteria involved articles in which indications for surgical treatment were based on CT criteria or were not explicitly reported. Moreover, editorials, reviews, letters, non-English language publications, and reports of pediatric cohorts (<16 y) were excluded. The search strategy was as follows: orbit*[tiab] AND (surgery[tiab] OR reconst*[tiab] OR repair[tiab]) AND (indication*[tiab] OR criteria[tiab] OR treatment[tiab]) AND fracture[tiab] NOT (Review[pt] OR Systematic Review[pt]).

Two reviewers independently screened articles for inclusion. Discrepancies were reconciled through discussion or by consultation with the senior author. Articles were included if they met the following:

• 1. Described patients with OB-OF assessed for surgery based exclusively on clinical criteria (eg, persistent diplopia, significant enophthalmos),

or

• 2. Reported cohorts with OB-OF followed clinically without surgical intervention,

and

• 3. Retrospectively analyzed radiographic findings to identify those predictive of the need for surgery or symptom development.

These criteria were used to reduce bias and confirm the predictive ability of different radiographic parameters. The identified articles were then analyzed to extract aim, design (prospective versus retrospective), methodology, risk of bias, sample size, clinical indications (diplopia, enophthalmos), radiographic findings, timing of surgery, follow-up, and outcome.

Statistical Methods

Eligible studies were analyzed to describe categorical and numerical data and identify trends and frequencies. Categorical variables were described as percentages, whereas numerical variables were described as means. Eligible studies displayed considerable variability in scope, objectives, and methodology, precluding inferential and associative statistical analysis.

Results

Out of 683 articles, 6 met the inclusion criteria (Fig. 5). (See table, Supplemental Digital Content 1, which displays an overview of study characteristics and outcomes, https://links.lww.com/PRSGO/E815.) Three studies were prospective. A total of 520 patients (65% men; mean age: 47 y) were reported. Surgery was performed in 57 (11%) patients. The median follow-up was 8 months.

Fig. 5.

Preferred Reporting Items for Systematic Reviews and Meta-Analysis flowchart showing the steps of the retrieval of the included studies. ZMG, zygomaticomaxillary complex.

Clinical Indications for Surgery

The following criteria were considered clinical indications for surgery:

• Persistent diplopia^7,9,10,28

• Hypoglobus²⁸

• Superior sulcus deformity²⁸

• Enophthalmos^9–11,28

• Motility restriction^7,9,11

• Persistent oculocardiac reflex.¹¹

Fracture Size/Area

Fracture size/area was evaluated in 4 studies (3 prospective).^9–11,29 These included 336 patients (76% men, mean age: 47 y). Diplopia occurred in 51 (15%) patients. Enophthalmos of at least 2 mm was encountered in 38 (11%) patients. Approximately 17% (57 patients) underwent surgery. Variable methods were used to estimate the fractured area, and studies used different software. Moreover, CT scan slice thickness varied across studies (2–3, 1, and ≥1 mm).^9,10,28 In general, studies found the absolute fracture area to be either weakly predictive¹¹ or less predictive of symptomatic fractures when compared with the fracture ratio (fracture area/orbital wall area)²⁸ and orbital volume, respectively.¹⁰ The cutoff value at which the fracture area would be predictive of symptom development and the need for surgery varied across studies as follows:

• At least 42% of the floor or a fracture area of at least 2.3 cm^2,28

• 3 cm^2,10

• At least 1.5 or at least 2.0 cm².¹¹

Orbital Volume Ratio

Orbital volume ratio (OVR), defined as “the relative volume difference between the fractured and nonfractured orbit,”²⁹ was calculated as a percentage using the equation: “volume of the traumatized orbit/orbital volume of normal side × 100.”⁷ It was assessed in 2 retrospective studies.^7,29 These included 184 patients (96% men; mean age: 48 y). Diplopia occurred in 15 (8%) patients. Enophthalmos of 2 mm or more developed in 26 (14%) patients. No patient underwent surgery.^7,29

The studies reported conflicting results. Although 1 study failed to identify a correlation,²⁹ the other found an OVR of 106.85% to accurately predict enophthalmos greater than 2 mm in 71.7% of patients.⁷

Herniated Orbital Volume

Herniated orbital volume was evaluated in 3 retrospective studies.^7,10,29 These included 305 patients (76% men; mean age: 47 y). Diplopia occurred in 33 (11%) patients. Enophthalmos of 2 mm or more developed in 37 (12%) patients. Fifteen (5%) patients underwent surgery, and 17 (5.6%) patients declined surgery. In 5 more symptomatic patients, surgical treatment was not reported.

Contradictory results were reported. Either no correlation was identified²⁹ or a herniated orbital volume of 0.531 cm³ or more in 1 study⁷ and greater than 1.3 cm² in another¹⁰ was found to be associated with a higher risk of symptom development.

IRM Rounding

IRM rounding was evaluated in 2 studies (1 prospective).^10,11 These included 149 patients (61% men; mean age: 45.7 y). Diplopia occurred in 14 (9%) patients. Enophthalmos of 2 mm or more developed in 14 (9%) patients. Fifteen patients underwent surgery (10%).

IRM was defined as: “normal (flat and thin) or displaying clear loss of typical features with a rounded appearance compared with the unaffected orbit.”¹⁰Alternatively, an IRM rounding index was analyzed. This index was defined as the height-to-width ratio of the muscle on the coronal slice exhibiting the maximal fracture width.¹¹ Contradictory results were reported. In 1 study, IRM rounding was found to be predictive of symptomatic fractures when considered as part of a risk stratification model consisting of additional parameters (caudal IRM malposition, orbital volume > 1.3 cm³, fracture area > 3 cm²).¹⁰ In the other study, the IRM rounding index of at least 1 was weakly predictive of enophthalmos of at least 2 mm at 6 months or more after injury (0% positive predictive value).¹¹

Caudal IRM Displacement

Caudal IRM displacement was evaluated in 2 studies (1 prospective).^9,10 These included 229 patients (68.5% men; mean age: 46.7 y). Diplopia occurred in 48 (21%) patients. Enophthalmos of at least 2 mm developed in 17 (7.4%) patients. Forty-nine patients underwent surgery (21%).

In 1 study, IRM displacement was classified relative to the level of the fracture as:

1. Mild (IRM within the orbit);

2. Moderate (most IRM within the orbit);

3. Severe (most IRM within the maxillary sinus).⁹

Alternatively, a transverse line bisecting the IRM was used to differentiate its position above, within, or below the level of the orbital floor.¹¹ Both studies reported a significant association between the severity of IRM displacement and symptom development and the decision for surgery⁹ (odds ratio, 3.7).¹¹

DISCUSSION

Our findings suggested that clinical criteria continue to be the most reliable in guiding the treatment of OB-OFs. Although CT scans remain the gold standard to confirm diagnosis and determine severity, there is no clear consensus on appropriate slice thickness, software processing, and calculation techniques to measure the different radiographic parameters. Except for IRM displacement below the orbital floor, contradictory conclusions were reported for all other radiological criteria. Of note, an absolute fracture size or volume alteration assumes that orbital size and volume are consistent across all individuals, irrespective of age, gender, and ethnicity. Nevertheless, orbital morphometric studies report variable measurements across ethnic groups and even among the same ethnicities, questioning the accuracy of a consistent cutoff value that is valid for all.^30–35

Fracture Size/Area

Alinasab et al²⁸ found that the risk for late deformity (superior sulcus deformity and/or hypoglobus and/or enophthalmos ≥ 3 mm) was highest for fractures involving at least 42% of the floor or a fracture area of at least 2.3 cm². Scolozzi et al⁹ concluded that the fracture size alone is “predictive only for enophthalmos but not for surgical repair or annoying diplopia.” Basta et al¹⁰ found symptomatic fractures associated with a fracture area greater than 3 cm². They found orbital volume changes of greater predictive accuracy. Billig et al¹¹ identified an orbital floor defect area of at least 1.5 or at least 2.0 cm² to be only weakly predictive of enophthalmos.

Of note, previous studies found that a fracture size of 1 cm² and 50% of the orbital wall, values commonly described as indications for surgery, are not always equivalent.³⁶ Thus, Schouman et al³⁶ found a 50% fracture ratio (surface of fracture/surface of fractured wall) to correlate with a variable fracture area of 2.47–3.74 cm² across their study subjects. Hence, the ratio of the fractured orbital wall rather than the absolute surface area is more accurate. Nevertheless, Alinasab et al²⁸ report that “calculation is time consuming and challenging in daily clinical applications.” The commonly cited indication, “fracture of 50% of the orbital wall” probably stems from the Hawes and Dortzbach¹⁹ report in 1983. In this retrospective study (51 patients), large fractures were defined as those greater than half of the orbital floor. They recommend early surgery (within 2 wk) in this cohort, even in the absence of clinical signs. Of note, not all patients in this study were evaluated with “tomography”; not all patients had isolated orbital fractures; the technique for fracture size determination was not disclosed; follow-up was limited to 6 weeks; and details of surgical interventions were lacking.¹⁹ Thus, it is difficult to generalize their recommendations to contemporary practice. Manson and Illif³⁷ suggest that “fractures with more than 2–3 cc (2-3 milliliters) of orbital volume difference begin to produce globe positional changes, and result when over one-half of the orbital floor is involved in fracture displacement.” They recommend early surgery in these cases.³⁷ Again, no evidence is presented to support the “one-half” determination.³⁷ The “2 cm² fracture area” commonly cited as an indication for surgery may have stemmed from a study by Manson et al¹⁷ in 1986, in which they analyzed bony and soft tissue volumes in 10 patients with posttraumatic orbital injuries.¹⁷ All but 1 patient had enophthalmos.¹⁷ Their study suggests bony volumetric expansion as the principal mechanism of enophthalmos.¹⁷ They state, based on their clinical experience, that fractures most likely to result in enophthalmos are those “where the orbital floor disruption exceeds a total area of 2 cm² diameter, where the bone volume change exceeds 1.5 cc (1.5 milliliters) (5% of orbital volume), and where significant fat and soft-tissue displacement occurs.”¹⁷ The 2-cm² cutoff metric is not substantiated by any data.¹⁷ Thus, the evidence to support a specific fracture dimension (1 cm², 2 cm², 50%, one-half, etc.) to guide treatment remains questionable (Fig. 6).

Fig. 6.

CT scan (coronal) of left OB-OF (floor and medial wall). The fractured segment is indicated by a blue dashed line. Despite a sizable fracture, displacement is minimal, expansion is mild, and the IRM (green arrows) remains within the orbit. Fracture size or surface area in isolation may not be reliable.

Orbital Volume Ratio

Alinasab et al²⁹ failed to identify a correlation between changes in orbital volume and the development of enophthalmos. Furthermore, they found a mean volume difference of 0.6 mL (2.5%) between right and left orbits in a control group of 18 nonfractured subjects.²⁹ They concluded that the OVR was insufficient to guide surgery.²⁹ Conversely, Yang et al⁷ found an OVR of 106.85% to accurately predict enophthalmos greater than 2 mm in 71.7% of patients.

The literature likewise displays contradictory conclusions. Thus, although Manson et al¹⁷ suggest enophthalmos to occur when bony orbital volume increase “exceeds 1.5 cc (1.5 milliliters)(5% of orbital volume),” Schönegg et al³⁸ found orbital volume changes to be poorly predictive of enophthalmos and diplopia.

Herniated Orbital Volume

Alinasab et al²⁹ failed to identify “statistical correlation between the herniated volume and enophthalmos.” Meanwhile, Yang et al⁷ and Basta et al¹⁰ independently found the volume of herniated tissue to be predictive of symptomatic fractures. Nevertheless, the cutoff values were different (0.531 cm³ versus 1.3–2.2 or ≥2.2 cc/milliliters).^7,10 Other studies beyond those included in our review suggest different cutoff values. Thus, Kovar et al³⁹ recommend surgery when the herniated volume exceeds 1400 mm³ in anteroposterior fractures and 500 mm³ in anterior or posterior fractures.” Ploder et al⁴⁰ found a herniated volume of 1.62 mL to be associated with enophthalmos greater than 2 mm.

IRM Morphology and Displacement

Basta et al¹⁰ found muscle rounding to be predictive of symptomatic fractures. Conversely, Billig et al¹¹ found an IRM rounding index of at least 1 to be only weakly predictive (0% positive predictive value).

Nevertheless, both Scolozzi et al⁹ and Basta et al¹⁰ independently found caudal IRM displacement to be predictive of symptomatic fractures. This conclusion was corroborated in other studies beyond those included in our review.^36,41 Thus, Schouman et al³⁶ found the “severity of IRM displacement” to be “the most important independent radiologic factor predictive of surgery.” Colangeli et al⁴¹ similarly identified a statistically significant correlation between the severity of IRM displacement and preoperative and postoperative diplopia. Compared with other radiological criteria, to the best of our knowledge, the predictive value of IRM displacement has remained uncontested across studies so far.

LIMITATIONS

Our review was limited by the small number of studies included. This was primarily due to our inclusion criteria, which restricted the analysis to reports in which surgical decisions were based exclusively on clinical grounds, with radiographic correlation established retrospectively. Nevertheless, this design was critical to confirm the impact of variable radiographic criteria and reduce bias. Furthermore, 2 studies originated from the same group (Karolinska Institute, Sweden) published 7 years apart, further limiting the pool of reports.^28,29 This institute adopts surgery for OB-OF when more than 1.5 mL of herniation is identified radiologically. Thus, the studies from this group comprised only patients managed conservatively who had less than 1.5 mL of herniation. Therefore, their reports, specifically those related to volume alterations, have an inherent selection bias that undermines the reliability of their conclusions. In addition, 3 studies were retrospective^7,10,29; thus, potentially liable to recall bias, missing data, and confounding variables. In addition, the extreme variation in objectives and methodology made it difficult to perform an aggregate assessment. This precluded inferential and associational statistical testing. This could have increased the significance of our findings.

Based on our review, except for caudal IRM displacement, other radiographic criteria reported (fracture size, orbital volume changes, herniated volume, rounding of IRM) may not accurately inform the treatment of OB-OF when considered in isolation. Clinical signs remain the most precise means to guide surgery. Nevertheless, imaging remains vital in confirming the diagnosis, understanding fracture morphology (eg, isolated orbital floor, medial wall, or combined floor and medial wall), ruling out associated fractures, planning surgery (eg, patient-specific implants), evaluating postoperative outcomes, and analyzing complications.

Of note, patient-specific implants have emerged as valuable tools to enhance precision in orbital reconstruction (Figs. 7, 8).^42–44 (See Video [online], which displays a postoperative CT scan of a patient with a patient-specific implant.)

Fig. 7.

Orbital reconstruction with patient-specific implants. A, Front view of the patient with a left orbital floor fracture showing enophthalmos of the left eye. B, Front view of the patient with a left orbital floor fracture showing 3 months postoperative appearance after reconstruction of the left orbit with a patient-specific titanium implant and correction of enophthalmos of the left eye. C, Worm’s-eye view of the same patient with a left orbital floor fracture showing enophthalmos of the left eye. D, Worm’s-eye view of the same patient with a left orbital floor fracture showing 3 months postoperative appearance after reconstruction of the left orbit with a patient-specific titanium implant and correction of enophthalmos of the left eye.

Fig. 8.

Orbital reconstruction with patient-specific implants. A, CT scan (coronal view) of the same patient in Figure 7 with a left orbital floor fracture, showing the fracture, volumetric orbital expansion, and rounding with caudal displacement of IRM. B, CT scan (coronal view) of the same patient showing satisfactory reconstruction with a patient-specific titanium implant.

Video 1. Postoperative video of CT scan coronal view of same patient in Figure 7 after satisfactory reconstruction with titanium patient specific implant.

Download video file ^{(58.6MB, mp4)}

Several systematic reviews have confirmed their role in delivering comparable^42,43 or superior⁴⁴ outcomes in orbital fracture management. Despite higher costs, their use may be associated with a decreased incidence of postoperative enophthalmos and diplopia, fewer revisions, and reduced operative times compared with conventional implants.^43,44

CONCLUSIONS

• Clinical criteria continue to be the most reliable indications for OB-OF reconstruction.

• Displacement of IRM below the orbital floor was the only radiological criterion with consistent predictive potential across studies. Nevertheless, more quality research is needed to confirm its predictive ability.

• Other CT scan parameters lack enough evidence to guide treatment independently and report variable cutoff values, limiting their usefulness.

• Specifically, until there is a consensus on the most appropriate determination of relative volume changes and fractured areas, cutoff values presented in the literature may only suggest but remain insufficient to direct management independently.

• Future directions should involve high-quality, multicenter prospective studies implementing standardized imaging protocols and processing software to scrutinize radiological parameters and establish accurate indications. Subsequent development of artificial intelligence algorithms considering the contralateral orbit may facilitate the rapid recognition of surgical candidates.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

PATIENT CONSENT

The patients provided written consent for the use of their images.

ACKNOWLEDGMENT

The authors thank Samuel Mansour, MD (Department of General Surgery, Wayne State University), for helping with the manuscript preparation and identifying relevant references.