Keywords

FormalPara Learning Objectives
  • The term subperiosteal abscess is a misnomer. It should be called more appropriately subperiosteal empyema.

  • For Chandler’s classification, there seems to be little support.

  • Orbital abscess formation as a complication of orbital cellulitis is rare.

  • If during inspection, the eye itself cannot be visualized because of severe swelling and tightness of the eyelids, the condition must be regarded as a retroseptal orbital cellulitis and managed accordingly.

Introduction: Classifications and Fallacies

Rootman [1] and others distinguish several types of orbital inflammation:

1. Specific inflammations, characterized by a specific pathogen or a specific clinical constellation and/or a specific histopathology:

  1. (a)

    Caused by bacteria, viruses, fungi or parasites

  2. (b)

    With predominant vasculitis (granulomatosis with polyangiitis or Wegener’s disease)

  3. (c)

    With predominant granulomatosis (sarcoidosis, xanthogranulomatosis)

  4. (d)

    Transitional (Kimura’s disease, Sjögren’s disease)

  5. (e)

    Autoimmune (Graves’ disease, IgG4-related disease)

2. Non-specific inflammation:

  1. (a)

    Idiopathic orbital inflammation

In this chapter, we will mainly focus on infectious orbital inflammations caused by bacteria, which are called orbital cellulitis (OC). Two types of OC can be distinguished: (1) the rather innocent preseptal orbital cellulitis (pOC) and (2) the potentially life-threatening retroseptal orbital cellulitis (rOC). The orbital septum is a thin, but firm, fibrous structure stretched out between the orbital rim and the transition of the eyelid retractors to the tarsal plate (Fig. 19.1). The orbital septum separates the orbital contents from the skin. A pOC is hence no more than an inflammation of the eyelid, whereas a rOC is an inflammation of the orbit itself.

Fig. 19.1
A well-labeled diagram of the anterior border of the orbital cavity. It depicts the preaponeurotic fat pocket, Whitnall's ligament, peripheral arterial arcade, superior tarsus, and others.

Orbital septum, the anterior border of the orbital cavity

Patients with OC are seen by many different medical doctors: general practitioners, pediatricians, ear, nose and throat specialists, ophthalmologists, oromaxillary surgeons, neurosurgeons, and radiologists. The literature on OC appears in all of their specialized journals, and this may contribute to the erroneous ideas that exist in this field of medicine.

The most important fallacy is that no distinction is made between an orbital abscess and a subperiosteal empyema, for which then the misnomer subperiosteal abscess is used. As will be explained, subperiosteal empyema and an orbital abscess are extensions and complications of OC, but differ significantly in their nature and require different therapeutical approaches.

An abscess is a collection of pus that has been accumulating within a tissue [2]. It is a defensive reaction to prevent the spread of infectious material to other parts of the body. An abscess is composed of an abscess wall (or: capsule), which is formed by adjacent healthy cells in an attempt to keep the pus from infecting neighboring structures. However, such an encapsulation may prevent penetration of antibiotics and immune cells from attacking bacteria present within the pus or from reaching the causative organism or foreign object. Therefore, treatment of an abscess consists of opening the abscess wall and draining its contents. In contrast, an empyema is a collection of pus in an extant space, such as the space between the bone and the periosteum. There is no capsule that hampers antibiotics to reach their target.

Another misconception, in our understanding, is the classification of the progression of the inflammation as proposed by Chandler et al. [3]: I. preseptal cellulitis > II. diffuse infiltration of the orbit (OC) > III. subperiosteal abscess > IV. orbital abscess > V. cavernous sinus thrombosis. In our series, only 2 out of 68 patients with pOC developed a rOC [4], i.e., 2.9%. It is more plausible to assume that bacteria spread to the most adjacent structures. So, infections of the ethmoid sinuses spread primarily to the orbit, whereas infections of the sphenoid sinus spread to the cavernous sinus and infections of the frontal sinus spread preferentially to the orbit and to the brain.

We have defined the several presentations of OC as follows [4]

  1. 1.

    Diffuse OC: Diffuse or localized infiltration of the orbital fat without fluid collection or rim enhancement (Fig. 19.2).

  2. 2.

    Orbital subperiosteal empyema: Fluid collection between the orbital wall and the periorbita (Fig. 19.3).

  3. 3.

    Orbital abscess: Fluid collection with peripheral enhancement and intraorbital extension through a defect of the periorbita (Fig. 19.4).

Fig. 19.2
An axial view C T scan depicts the diffuse orbital cellulitis. It represents swelling on the right side of the face.

Diffuse infiltration of the medial part of the orbit secondary to ethmoidal sinus disease

Fig. 19.3
A front view C T scan of the skull. The arrow marks the location of the collection of the fluid.

Orbital subperiosteal empyema: Fluid collection between the orbital wall and the periorbita

Fig. 19.4
A lateral view C T scan of the skull. The arrow marks the location of the collection of the fluid.

Orbital abscess: fluid collection with peripheral enhancement and intraorbital extension through a defect of the periorbita

Epidemiology and Presentation

OC is a disease of all ages. Over a period of 26 years, 53 adults with OC were seen at a university clinic in Split, Croatia [5]. For 6 years, 28 children with OC were seen at a university clinic in Marrakech, Morocco [6]. In the University Medical Center of Amsterdam, we have seen 116 patients, both children and adults, with OC during a period of 7 years [4]. Sixty-eight of them had pOC, 48 had rOC. Seventeen patients (25%) with pOC were younger than 9 years of age; another 17 (35%) with rOC fell in the same age-group. In conclusion, OC—albeit pOC or rOC—is rather rare. The reported incidence is 0.1–3.5 per 100,000 individuals [7]. It is assumed that children suffer from OC more often than adults. This is not in accordance with our experience, but this may be due to referral bias. In our series, 47% of the patients with pOC and 77% of the patients with rOC were male.

Clinical symptoms of pOC are diffuse eyelid swelling, eyelid redness, chemosis, and conjunctival redness. Symptoms of rOC are in addition: pain, exophthalmos, eye motility impairment, a relative afferent pupil defect, retinal vascular involvement, loss of visual functions, and a swollen optic nerve head as the result of compression of the optic nerve. If the eyelids are swollen and tight to such an extent that the eye itself cannot be inspected and the abovementioned symptoms cannot be assessed, the condition must be regarded as a pOC (Fig. 19.5).

Fig. 19.5
A front-view photograph of the face of a young child with swelling on the right eyelid.

Young child with swollen ptotic and red eyelid suggestive for orbital cellulitis

Fig. 19.6 and 19.7
2 axial view C T scans of the skull depict the swelling on the right side of the face.

CT scans showing orbital cellulitis and cerebral involvement

Etiology, Differential Diagnosis and Complications

The most frequent cause of OC is an extension of paranasal sinus disease. Inflammation of the ethmoid sinus can co-exist with signs of pOC or with a rOC. The medial wall of the orbit is paper-thin (hence, its name lamina papyracea). Extension of infectious material along vessels through the openings in this wall or retrograde spread by the interconnecting valveless venous system of the orbit and the sinuses easily results in OC. pOC is also seen as a complication of dermatitis. The diagnosis is based upon the presence of the clinical symptoms in combination with (a history of) upper respiratory disease or dermal disease. MRI scans or CT scans of the orbit and paranasal sinuses are indicated in case of doubt and to differentiate between the possible sources of origin and the different orbital presentations. A peculiar form is cellulitis of odontogenic origin, which is feared for its bad outcome [8].

In general, a distinction is made between OC in adults and in young children. According to Harris et al. [9] the populations of causative bacteria in children under the age of nine are usually monomicrobial in nature and less aggressive than those found in older patients. Therefore, they would better respond to antibiotic therapy. In odontogenic cellulitis, in contrast, a polymicrobial infection consisting of both aerobic and anaerobic bacteria is responsible for the bad outcome [10].

The eyelid swelling and redness as seen in pOC must be differentiated from severe conjunctivitis, secondary to severe uveitis or endophthalmitis, and secondary to eyelid or orbital surgery. A mosquito bite, allergic reactions, or malignancies such as a rhabdomyosarcoma or non-Hodgkin lymphoma can also be mistaken for OC.

Apart from blindness, extension to the cavernous sinus resulting in thrombosis and extension to other intracranial loci are feared complications of OC. Cavernous sinus thrombosis will be discussed at the end of this chapter. In our series, we have seen three children and one adult with intracranial spread of OC. Interestingly, they had a short history of disease (less than 4 days) and presented with manifestations of cerebral spread, already at their admission to the hospital (Figs. 19.6 and 19.7). They all survived and none had long-term complications.

Other long-term complications are diplopia due to motility impairment and eyelid abnormalities (retraction, entropion, ptosis, and lasting swelling.

Treatment and Outcome

Since classical times, the adage “ubi pus, ibi evacua” (pus must be surgically drained wherever it shows itself) has been cherished in western medicine and through empowerment of time modern-age, medical doctors act almost automatically to the term abscess with an intention to operate. In our series, we found many cases in which surgery had been performed without clear indications. In the era of antibiotics, the need for a surgical approach of OC has dramatically decreased. Souliere and Harris demonstrated that medical treatment alone is safe in most children younger than 9 years of age [11, 12]. In our retrospective series, 68 patients with pOC were treated with antibiotics (74% orally, and 26% intravenously administered). Five patients (7%) underwent orbital surgery (incision and drainage, the indication could not be retrieved). All patients with pOC recovered completely without any complication. We therefore suggest ambulatory oral antibiotics as an initial treatment for patients with pOC plus careful watching. We found that only two patients (3%) with pOC developed a rOC.

Of 48 patients with rOC, 83% had co-existing paranasal sinus disease. All patients were admitted to the hospital. Nineteen patients (40%) were treated with intravenously administered antibiotics alone. The remaining 29 patients (60%) underwent surgery as well (21 underwent external orbitotomy with or without sinus surgery; 8 underwent sinus surgery alone). Retrospective evaluation between the two groups (antibiotics alone versus antibiotics plus surgery) revealed no significant changes in age, gender, duration of disease prior to admission, presumed origin of the disease, intraorbital localization of the inflammation, and cultured bacteria. In other words, the criteria for surgery could not be clarified. Was it the term subperiosteal abscess that prompted surgery?

In our initial radiological reports, orbital abscess was mentioned in 16 patients. Re-examination of the MRI scans and CT scans, using the abovementioned criteria, resulted in 12 subperiosteal empyema’s and only four true orbital abscesses in a series of 48 patients with rOC (8%).

Three out of 48 patients (6.5%) experienced permanent blindness due to the inflammation. In contrast, one 58-year-old woman without any light perception for 24 subsequent hours left the hospital with normal visual functions. Other long-term complications were reduced motility in five patients, ptosis in two patients, and lagophthalmos in another two patients. These complications seemed to be caused by excessive fibrosis secondary to the inflammation. A prudent use of adjuvant systemic steroids may therefore be considered [13].

In conclusion, in patients with rOC we advise to start with intravenously administered antibiotics. If the response is not sufficient within the next 48–72 h, surgery can be considered. Immediate surgery is advised in cases with: (1) a tight orbit with signs of compressive opticopathy, due to raised intraorbital pressure [14] as a consequence of space-taking swellings in the orbit, such as a very large subperiosteal empyema (compare with retrobulbar hemorrhage), (2) true orbital abscess, and (3) odontogenic OC.

Cavernous Sinus Thrombosis and Mucormycosis

Cavernous sinus thrombosis (CST) is notorious for its bad prognosis in terms of mortality and morbidity, but frequencies of these are based on old series. Using the files of the University Medical Center at Amsterdam during the period 2005–2017, we found only 12 patients with CST. Eleven survived and nine recovered without any permanent deficits [15]. Patients presented with eyelid swelling, chemosis, proptosis, high rates of impaired ocular motility, and signs of optic neuropathy. In contrast to “simple” OC, patients with CST more often had bilateral eye symptoms, high fever, and severe headaches. CT scans—or preferably contrast-enhanced MRI scans—show direct and indirect signs of CST. Direct signs are expansion of and filling defects in the cavernous sinus. Indirect signs, caused by venous obstruction, are dilatation of the superior ophthalmic vein, exophthalmos, and increased dural enhancement of the border of the cavernous sinus. An associated sign is secondary thrombosis. In our series, 50% of the CST was caused by extension of an infection of the sphenoid sinus. Other causes were otitis media, fungal rhinosinusitis, pharyngeal infection, and meningitis. Treatment consists of intravenously administered antibiotics in combination with surgical (endoscopic) drainage of the initially infected spaces or paranasal sinuses. In addition, selective anticoagulation therapy should be considered.

Rhino-orbital mucormycosis is a rare and opportunistic fungal infection, which is seen in patients with diabetes mellitus or ketoacidosis as well as in immunocompromised individuals. At the early stage of the infection, occlusion of blood vessels and subsequent thrombosis and necrosis can lead to a blind and immobile eye. Immediate treatment of the underlying disease together with extensive surgical debridement (often including orbital exenteration) and systemic antifungal drugs can be life-saving (Fig. 19.8).

Fig. 19.8
A photograph of a female patient with C S T. It depicts the swelling, chemosis, proptosis, high rates of impaired ocular motility in the eyelid.

A 20-year-old woman with status after eyelid-sparing exenteration for mucormycosis after ketoacidosis in diabetes mellitus