**7. Investigative studies**

**6. Epidemiology of orbital cellulitis**

130 Common Eye Infections

immediate resolution of his symptoms.

Orbital complications of sinusitis have been reported to range anywhere between 0.5 to 3.9%. [1], [17], [27] However, the incidence of abscess formation vary considerably from 0-25% in the reported series. [15] No cases of abscess formation was reported in the published series from the Children's Memorial Hospital in Chicago including 87 patients with orbital cellulitis and from Children's Hospital in Pittsburgh including 104 orbital celulitis cases. [28], [29] On the other hand, a larger study of 6,770 patients from the Hospital for Sick Children in Toronto revealed that 2.3% developed orbital complications; of which 10.7% had abscess formation. [19] Another study reported 20.8% incidence of abscess formation among the 158 patients admitted for orbital cellulitis. There was a 20.8% incidence of abscess formation. [30] Among other series which has reported orbital complications of sinus disease, the incidence of abscess formation had varied from 6.25 to 20 % to as high as 78.6%. [18], [31], [32] One may attribute differences among these studies due to the inclusion criteria, age group and the severity of the complications studied by these authors. The incidence of major complications following sinusitis may be low, however such complications may be associated with considerable morbidity and mortality. [8], [15] According to the published report, in the pre-antibiotic era, orbital cellulitis resulted in death from meningitis in 17% of cases and blindness in 20%. [6] However, in the antibiotic era, incidence of menengitis was reported as 1.9% in patients with

orbital cellulitis, despite prompt treatment with systemic antibiotics. [33]

**Figure 9.** External photographs as well as CT-scan (axial and coronal views) of a 7-year-old boy who presented with upper respiratory infection followed by painful diplopia, left eye proptosis and decreased vision. His symptoms did not improve with a course of systemic antibiotics. This patient required drainage of his orbital abscess which resulted in

In-spite of systemic antibiotics and surgical intervention, orbital abscesses may have devas‐ tating outcome. [15] According to the series in which final visual results have been reported, a significant percentage of patients have been left with non-seeing eyes ranging anywhere from 7.1% to as high as 23.6%. [10] Visual loss in these cases have been attributed to optic atrophy, central retinal artery occlusion, or exposure keratopathy with corneal ulcer formation. [8], [10], [19] Some of the other hypothesized mechanisms of vision loss are septic optic neuritis, embolic

On orbital ultrasonography (U/S), abscess may show low internal reflectivity and therefore, U/S can be useful as a screening office procedure for patients suspected of having orbital abscess. [8], [27] Computed tomography scan may be necessary to assess the evidence of sinusitis and orbital processes. On CT-scan, orbital abscess may appear as localized, generally homogenous elevation of the periorbita adjacent to an opacified sinus, (Figure 9). On imaging studies, there may be evidence of inflammatory or infective changes in the sinus areas as well as orbital structures. In children, more patients may have subperiosteal abscess as compared to the adult group at the time of their initial presentation. [4] In the series reported by Ferguson and McNab, [1] among children, 29% had inflammatory changes only, while 62% had evidence of a subperiosteal abscess, only 9% had orbital abscesses, compared with 72%, 5% and 22%, respectively, in their adult group. Computed tomography scan may influence the initial therapeutic plan by demonstrating the size and location of the abscess and the specific sinuses involved, features that may be necessary in the approach of surgical drainage. [8], [20], [27] Experience however have shown that the CT-scan characteristics of the subperiosteal collection may not always be predictive of the clinical course. For example, in reports from the patients who recovered with systemic antibiotics alone, findings were similar to the findings in patients who underwent surgical drainage. [22] The imaging studies have shown that the size of an orbital abscess may increase over the first few days of intravenous antibiotics regardless of the bacteriological response to the treatment in these patients. [22] In some patients, the identifi‐ cation of an orbital abscess may be a diagnostic challenge. The reliability of some of the imaging modalities such as CT-scan in demonstrating some orbital abscesses has been questioned. For example from a series of 25 cases of orbital infection, all 15 orbital abscesses were satisfactorily demonstrated only when the CT-scan examination included coronal sections. [34] According to this study, one-third of abscesses would have been missed if coronal sections had not been performed. Magnetic resonance imaging studies have been found to be necessary in some cases where CT-scan have not satisfactorily addressed clinician's concerns.

The development of an orbital abscess does not correlate specifically with visual acuity, proptosis, chemosis, or any other signs. [15], [27] Therefore diagnostic procedures are es‐ sential in evaluating the patient with orbital cellulitis and possible abscess or retained orbi‐ tal foreign bodies. Although sinus X-ray may demonstrate an air-fluid level when present in an abscess cavity, gas-free abscesses may not be readily visible. [15] Ultrasound may de‐ tect an abscess of the anterior orbit or the medial wall with 90% accuracy, [25] although an acute abscess may be poorly delineated. Currently, the investigative procedure of choice to diagnose an orbital infection is the CT-scan, although MRI can be utilized when there is a contra-indications for CT-scan. [8], [27], [35] By CT-scan, orbital walls, extraocular muscles, optic nerve, intraconal area and adipose tissue can be seen clearly. An orbital abscess can be seen as a homogenous, a ring-like, or a heterogenous mass. In these studies, the site of origin, orbital or subperiosteal, and extent of abscess are readily visible. [8], [22] When ad‐ ministered, contrast-media can enhance the surrounding wall of an abscess. Computed to‐ mography scan will not differentiate between preseptal cellulitis and eyelid edema but will differentiate between preseptal and orbital cellulitis. [15] Beside foreign bodies, sinus dis‐ ease and intracranial complications may also be visible on the CT-scan. [8] Our experience has shown that CT-scan may be the most comprehensive source of information about orbi‐ tal infections and the most sensitive means of monitoring resolving orbital or intracranial lesions. [8], [27] Computed tomography scan is indicated in all patients with periorbital in‐ flammation in whom proptosis, ophthalmoplegia, or a decrease in visual acuity develop, in whom a foreign body or an abscess is suspected, severe eyelid edema prevents an ade‐ quate examination, or surgery is contemplated. [8], [15], [20], [22], [35] In our study of the 218 patients with orbital cellulitis, diagnosis was made clinically and confirmed by CTscans or U/S in 90.4% and 36.2% orbits, respectively. [8] Orbital abscesses were identified in 53.2% of orbits. In all cases of orbital cellulitis, there was evidence of inflammatory or infective changes of the orbital structures. Abscess location was found to be medial in 35%, superior in 33%, intraconal in 13%, superomedial in 6%, inferomedial in 6% and lateral in 2% of orbits. [8]

### **8. Bacteriology of orbital infection**

In the reported series, the bacteriology of orbital abscesses has received little attention. In series in which the contents of the abscess cavity have been cultured, a wide range of organisms have been reported. [8], [36] Most commonly reported bacterial species from the abscesses of the orbit and periorbital area include Staphylococcus aureus, Staphylococcus epidermidis, Streptococci, Diphtheroids, Haemophilus influenzae, Escherichia Coli, multiple species of aerobes and anaerobes. There was no growth in up to 25% of abscesses. [15], [17] Microbio‐ logical results from Ferguson and McNab, [1] series varied, with differences in the rate of testing between the pediatric age group and the older age group. In their series, some forms of cultures were performed in 93% of their patients. Fifty percent of their patiens had blood cultures none of which yielded positive results. According to their study, cultures taken from abscesses were more likely to yield positive results. The authors noted that there was no correlation between cultures taken from conjunctival swab and the etiological organisms recovered from the abscesses of those patients with positive cultures. In their study, Staphy‐ lococcus aureus was the most common micro-organism recovered. In their pediatric age group various species of Streptococcus predominated. Among their pediatric patients, 4 patients had anaerobic Streptococcus isolates, two had mixed anaerobes and one had Clostridium bifer‐ mentans. In Ferguson and McNab's, [1] series, orbital cellulitis due to anaerobes was much less common in adults, with only one case of mixed anaerobes identified. In their series, only 5 adults and 4 pediatric patients had multiple organisms isolated from the abscesses. No pathogenic organisms were isolated from their 6 adult and 15 pediatric patients in whom the cultures were performed. [1] Although in the past, H influenza was a major pathogenic bacteria responsible for orbital cellulitis in the pediatric age group, [8], [15] in the series reported by Ferguson and McNab,1 no cases of H. influenza were identified in the pediatric age group and only one case was found in an adult patient. This observation has been attributed to the general immunization of children with H. influenza type B vaccine since the early 1990s. [1], [27] Schramm et al,5 reported 32 cases of orbital abscesses, the predominant microorganisms being Staphylococci, Streptococci and Bacteroids species.

The role of anaerobes, not usually considered pathogens in the sinus disease is unclear, although considerable number of cultures in adult patients have yielded anaerobes. [8], [20], [23] In general, patients during their first decade of life may have infection caused by a single aerobic pathogen which may be responsive to the medical therapy alone. On the other hand, patients older than 15 years of age may have complex infections caused by multiple aerobic and anaerobic organisms that may be slow to clear despite medical and surgical intervention. [23] The virulence of pathogens and responsiveness to anti-microbial agents appear to be agerelated. [20], [37] With enlarging of the size of the sinus cavities, the ostia gets narrow creating optimal condition for anaerobic bacterial growth. As the person ages, there is a trend towards appearance of more complex infections. In mixed infections, aerobes utilize oxygen which encourage growth of more anaerobic microorganisms. On the other hand anaerobes produce B-Lactamase which makes antibiotics less effective. Harris, [20] reviewed microbiology results of 37 patients with orbital abscesses in which one-third were younger than 9 years, 58% were culture negative and the rest had single aerobic pathogen. From his series, 16 patients between ages 9-14 years showed transition towards more complex infections. Among these, 9 patients which were older than 15 years had positive cultures despite being on systemic antibiotics for 3 days. In Harris's, [20] study, older group had more often polymicrobial infections and anaerobes were found in all cases. According to our study the most common microorganisms isolated from the drained abscesses were the most common microorganisms isolated from the drained abscesses were Staphylococci and Streptococci species; less common organisms included Propionibacterium acnes, Haemophilus influenzae, Bacillus, and fungi. [8]
