Introduction

Historically, orbital surgery has been performed by general surgeons, ophthalmic surgeons and, in particular, neurosurgeons; indeed, Walter Dandy – the highly influential neurosurgeon – suggested that, ‘all orbital surgery should be performed through the transcranial route’ and that ‘such surgery results in less morbidity than the commonly-used orbital approach’¹. During the second half of the 20th century there was a shift towards less traumatic techniques for orbitotomy, avoiding craniotomy when a disease was solely intraorbital – this leading to the development of incisions and approaches with much better esthetic results. Clearly, a neurosurgical approach remains imperative where a disease straddles the cranio-orbital junction (such as with lesions passing through the superior orbital fissure or optic canal), and a maxillofacial or head-and-neck surgeon is best experienced to deal with orbital diseases involving the paranasal sinuses or extending widely outside the orbit.

Contemporary orbital surgeons will often assist neurosurgical or maxillofacial colleagues with more extensive cases, but disease confined to the orbit is generally amenable to approaches that can be performed by an ophthalmic surgeon with specialist training in this particular branch of oculoplastic surgery. A series of practical approaches that provide good access to all areas of the orbit will be outlined, but specific individual procedures, such as orbital exenteration, will not be considered.

Practical and anatomical considerations

The orbit is a tetrahedral soft tissue structure surrounded by bone – except anteriorly – and so, in the absence of bone removal or displacement, the size of all surgical routes is constrained by the anterior orbital opening (generally no more than 3 cm) and the breadth of such an incision is limited by the ability to displace the globe. An anterior incision of this size means that the maximum theoretical ‘conoid of view’ for an apex lesion is about 50° by 30° – although in practical terms this conoid may be only about half this size, due to orbital fat spilling into the operative field and the presence of many relatively fixed orbital structures. The conoid of view may be increased by two mechanisms: the first is to move the incision nearer to the site of the lesion – that is, to shorten the conoid ‘depth’ – and this is effective during the outer canthotomy approach to lateral masses. The second method is to displace an orbital wall outwards, this being the main rationale for bone-swinging lateral orbitotomy.

Access from the physiological entrance to the orbit also leads to another surgical difficulty that worsens with increasing depth: namely, the problem of increasingly oblique viewing of many structures – both normal and abnormal – within the depths of the orbit. This obliquity, which can cause problems with both the identification and the handling of structures, may be effectively reduced by shifting the incision towards the lateral canthus or by swinging the lateral wall outwards during orbitotomy.

The risk of injury to orbital function varies significantly at different depths within the orbit, there being variations within each zone. The risk of collateral injury is low during surgery in the anterior one-third of the orbit, alongside the globe – the risk probably being greatest in the trochlear area and to the palpebral lobe of the lacrimal gland. The most feared risk whilst operating in the middle-third of the orbit is the (relatively rare) visual loss due to ischemia of the optic nerve or optic nerve head, but the ciliary ganglion is probably the most susceptible structure in this sector – a persistent mydriasis being a not uncommon inconvenience after removal of retrobulbar tumors. Due to crowding of highly important structures in the orbital apex, operating in the posterior third of the orbit carries a high risk of complication – many being serious or permanent ²; complete visual loss is quite common with removal of ‘peanut’ tumors wedged in the apex, especially where the tumor lies below the optic nerve, alongside the ophthalmic artery. Complete ophthalmoplegia is common after operating near the superior orbital fissure, this often recovering dramatically at about 3 months after surgery, but sometimes leaving a selective neuropraxia.

Preoperative and postoperative counseling

The appropriate risks of surgery should be explained to the patient, together with the risks of choosing not to have surgery. Clearly there is no real choice where malignancy is suggested on imaging. However, for well-defined ‘benign’-looking apical lesions that carry a significant risk of serious complication with surgery, it is advisable to wait for documented evidence of progressive visual deterioration before considering intervention. The risks for surgery in the anterior one-third are low and explanation can reasonably be limited to those that are anatomically relevant – such as ipsilateral sensory loss after surgery along the orbital roof – and a brief mention of the miniscule risk of visual loss. With surgery in the middle third and posterior third, however, the risks are much greater and preoperative counseling should cover the danger of complete or partial visual loss, squint and diplopia requiring further intervention, ptosis, periocular sensory loss, and a permanent mydriasis (with mild photophobia or presbyopia).

Apart from difficult access or the encroachment of fat into the operative field, it is intraoperative hemorrhage that creates the greatest difficulty and risk during surgery within the orbital depths; heat generated by the use of bipolar diathermy increases inflammation, the latter probably being contributory to early postoperative visual loss due to vasospasm at the orbital apex ². It is, therefore, very important to take a thorough history for any medications that might affect hemostasis³ – this including over 40 brand-named versions of amino-salicylic acid (ASA) and also many foodstuff and ‘herbal’ remedies. All herbal remedies, garlic, ginger, ginseng and gingko should be avoided completely before surgery. If safe to do so, patients should be encouraged to avoid ASA for at least 3–4 weeks before orbital surgery, non-steroidal anti-inflammatory drugs for at least 2 weeks, and other anti-platelet medications for at least 10–12 days. Special arrangements will be required for patients requiring warfarin anticoagulation for metallic heart valves and for patients with recently implanted arterial stents⁴. Medical conditions, such as systemic hypertension or diabetes, should be well controlled in all patients listed for routine orbital surgery, and a preoperative assessment by the anesthetist is appropriate where hypotensive general anesthesia will be required for surgery.

After surgery, a patient should be nursed semi-recumbent to minimize postoperative swelling and, where accessible, the pupils and vision checked for a few hours after surgery. Early postoperative instruction should insist that medical attention is sought if there is significant and increasing pain, as this can be an important symptom of intraorbital arterial hemorrhage; consideration should be given to urgent release of any sutured closure and drainage of hemorrhage where a ‘rock-hard’ marked proptosis has developed, with poor eye movements and evidence of optic neuropathy. Systemic corticosteroid therapy and, if indicated, antibiotics may be prescribed to limit postoperative inflammation and scarring, or prevent infection.

Patients should be warned that gross swelling of the eyelids is common after deep orbital surgery and that, in some cases, this can take several weeks – or even months – to settle. They can pursue all normal activities, apart from vigorous sports and activities that should be avoided during the first 3 weeks after surgery. New-onset diplopia and (the rare) loss of vision present major practical difficulties and patients will require appropriate counseling in relation to their work and lifestyle.