Silicone oil-associated orbital cellulitis with lipogranulomatous inflammation in the setting of HIV: a management challenge and clinicopathological correlation

Abstract

A 58-year-old Indian man presented with pain and redness of the left eye (OS) for one day. Patient had undergone silicone oil removal in OS for emulsified oil following vitrectomy and oil tamponade six months ago when he was diagnosed with retinal detachment in both eyes due to HIV retinopathy. Retinal detachment in the right eye (OD) was inoperable and had turned prephthisical at presentation, while his vision in OS was finger counting. Intraocular pressure in OD was 8 mm Hg and unrecordably elevated in OS. Extraocular movements were limited by periorbital oedema and proptosis. Slit-lamp examination revealed corneal haze, cells 2+/flare 1+ with pseudophakia, and attached retina. Histopathology showed lipogranulomatous inflammation, hitherto unreported in association with silicone oil. The index case posed a management challenge since his only functional eye had potentially been compromised by glaucoma and orbital cellulitis with compartment syndrome, against the backdrop of an immunocompromised status.

Background

Modern vitrectomy techniques have significantly improved the outcomes in cases of retinal detachment repair and have allowed surgeons to successfully fix eyes that were previously deemed unsalvageable.¹ Two important factors which made this possible are advanced vitrectomy machines with microsurgical instruments and various tamponade agents. Depending on the clinical scenario, that is, patient factors, complexity of the case and duration of tamponade required to achieve reattachment of retina, various agents are currently used - namely purified air, expansile gases and silicone oil. Silicone oil is a synthetic compound made of repeating units of ‘siloxane’ (–Si–O) alongside chains of hydrocarbon radicals. Therefore, silicone oil made of dimethyl siloxane is stable, highly biocompatible and biologically inert. Due to its viscosity and polymeric chemical structure, it is immiscible with the aqueous. The viscosity of silicone oil, which is used in intraocular tamponade, ranges from 1000 cs (37 kDa) to 5000 cs (65 kDa).^{2 3} Silicone oil tamponade (SOT) is commonly used in retinal detachment surgery, where it tends to emulsify into tiny globules after having remained inside the eye for a prolonged period.^{4 5} These tiny globules remain suspended in the aqueous, fill the postvitrectomy eye, scatter light and impair vision. Emulsified silicone oil may also clog the trabecular meshwork, leading to raised IOP and secondary open-angle glaucoma.⁶ Many factors contribute to silicone oil emulsification at variable rates in different patients. In one study, impurities such as sterilisation chemicals and detergents present in surgical tubing gained intraocular access and accelerated emulsification.⁷ Hence, reuse of surgical instruments should be avoided. The rate of emulsification depends on two cardinal factors, that is, viscosity and molecular weight. Less viscous and lower molecular weight oils with fewer long-chain branching tend to emulsify earlier, in contrast to their heavier counterparts.^{8 9} The duration of oil in-situ is yet another consistently observed factor associated with quicker emulsification.¹⁰ Proteins and phospholipids hasten emulsification by lowering its surface tension.¹¹ Plasma proteins and red blood cells' membrane have been observed to increase the rate of emulsification in a study by Savion et al.¹² Oshika¹³ observed that patient age was a predominant factor governing emulsification rates, where younger patients with higher aqueous protein concentration developed accelerated emulsification. Silicone oil (in-situ) can lead to cataract and retinal changes such as reproliferation of the epiretinal and subretinal fibrous membrane, retinal toxicity and so on.^{14 15} Moreover, aseptic intraocular inflammation can occur after SOT especially when silicone oil remains in the eye for several months.¹⁶ This warrants its removal from the eye once the primary objective of retinal reattachment has been achieved.⁵

Management of the index case posed a challenge due to the following reasons. First, the patient was ambulatory using only a single functional eye, which had potentially been compromised by secondary glaucoma due to emulsified oil. Second, orbital cellulitis with compartment syndrome combined with unrecordably high IOP mandated an urgent clinical decision making in order to prevent irreversible optic neuropathy. Decision to treat entailed ruling out an infectious aetiology. Our team was faced with a new dilemma once aseptic aetiology was established and the decision to use high-dose steroid therapy was taken. This dilemma was the patient’s HIV status superimposed with immunosuppression by corticosteroids. However, in the backdrop of highly active antiretroviral therapy (HAART) with adequate CD4 counts and absence of opportunistic infections, infectious disease consultation advised going forward with corticosteroid therapy.

Case presentation

A 58-year-old Indian man presented with forward protrusion of the left eye (OS) for one day associated with pain, watering, redness and periocular swelling. The patient had undergone silicone oil removal (SOR) in OS for emulsified oil and developed the symptoms on first postoperative day. He had undergone pars-plana vitrectomy and SOT six months ago, having been diagnosed with retinal detachment in both eyes due to HIV retinopathy. The right eye (OD) was diagnosed with inoperable retinal detachment and was left alone. On examination OD was prephthisical, with finger counting vision in OS. IOP of OD was 8 mm Hg and unrecordably elevated in OS. IOP of OS prior to SOR was recordable at 52 mm Hg with inverse hypopyon. Extraocular movements were limited by periorbital oedema and conjunctival congestion. Proptosis (figure 1) was non-pulsatile, non-reducible and without postural variation. Hertel exophthalmometry measured 18 mm (OD) and 22 mm (OS). No bruit was heard on auscultation. Extraocular movements were limited in all gazes. Tense periorbital oedema with conjunctival congestion was present, with visible subconjunctival cysts/globules (figure 2A, B). Slit lamp examination revealed corneal haze, cells 2+/flare 1+ and pseudophakia, and the fundus was clear with attached retina. In a multidisciplinary liaison approach, otorhinolaryngology and infectious disease consultations were sought to rule out the involvement of sinuses for pre-existing primary foci causing cellulitis, as well as to establish the absence of opportunistic infections in the background of HIV. Thereafter systemic and appropriate laboratory/diagnostic work-up were completed.

Figure 1.

Clinical photograph at presentation showing proptosis of the left eye in frontal view. Lid oedema and conjunctival congestion can be appreciated.

Figure 2.

(A, B) Globules of silicone oil in the superior and inferior subconjunctival space of the nasal conjunctiva with ciliary injection.

Investigations

Physical examination and laboratory work-up did not reveal any systemic pathology. Proptosis was evident with an air bubble in the anterior chamber on the first postoperative day after SOR (figure 3A), with straightening of the intraorbital portion of the optic nerve on axial CT scans (figure 3B). Contrast-enhanced CT scan of the orbits revealed soft tissue stranding around the inferior rectus muscle in the left orbit (arrow in figure 3C). The presence of infective pathology, mass lesion or abscess appeared unlikely. Once the working diagnosis of aseptic cellulitis was established, infectious disease consultation cleared the patient for high-dose corticosteroid therapy, in the backdrop of HAART with adequate CD4 counts and absence of opportunistic infections. Concurrently a swab was taken from the subconjunctival space along with tissue biopsy after SOR for definitive diagnosis. Gram stain, KOH mount and culture did not reveal any micro-organisms. Repeat CT scan after pulsed steroid therapy showed significant reduction in signs of inflammation and proptosis (figure 3D). Histopathology report showed presence of clear vacuoles in the conjunctival and subconjunctival tissues, which were attributable to silicone oil-filled cavities, as also visualised on clinical examination at the site of the tissue biopsy. This was accompanied by lipogranulomatous inflammation consisting of histiocytes and macrophages surrounding the clear vacuoles (figure 4A). Multiple granulomas along with multinucleated giant cells were present (figure 4B, C).

Figure 3.

CECT of brain and orbit. (A, B) Axial scan on day 1 post SOR showing proptosis of the left eye with an air bubble in the anterior chamber (yellow arrow) and straightening of the intraorbital portion of the optic nerve (red arrow). (C) Coronal scan on day 1 post SOR showing soft tissue stranding around the inferior rectus muscle (arrow) in the left orbit. (D) Axial scan at 1 week post SOR showing improvement in proptosis after receiving intravenous methylprednisolone. SOR, silicone oil removal.

Figure 4.

Histopathology of conjunctival biopsy (H&E stain). (A) Under 10× view showing clear vacuoles in the conjunctival and subconjunctival tissues which were silicone oil-filled cavities (arrow). Lipogranulomatous inflammation (arrowhead) consisting of histiocytes and macrophages, surrounds these clear spaces. (B, C) Under 40× and 400× views showing granuloma (arrow) consisting of multinucleated giant cell (red arrowhead).

Differential diagnosis

Based on presentation and history, the following differential diagnoses were considered: orbital cellulitis (septic/aseptic), cavernous sinus thrombosis, orbital haemorrhage and caroticocavernous fistula. Radiological investigations ruled out the presence of a mass lesion. Caroticocavernous fistula usually presents with rapidly progressive pulsatile proptosis, while cavernous sinus thrombosis presents with headache, fever and periorbital oedema, hence were unlikely causes in this case. After SOR, orbital haemorrhage was likely, which does present with pain and rapid loss of vision and evolving relative afferent pupillary defect (RAPD); however, CT scan findings were not consistent with the same. The patient’s afebrile status combined with radiological and clinical findings also ruled out the possibility of orbital cellulitis secondary to an infectious aetiology. CT imaging findings suggestive of true orbital cellulitis with orbital compartment syndrome, air bubble in the anterior chamber and histopathological evidence of lipogranulomatous inflammation ruled out the possibility of pseudo-orbital cellulitis. Gas leakage to the periocular tissue may be a potential cause of orbital cellulitis. However, except for the anterior chamber tamponade by air, gas was not used in any stage of the surgery. Also the gas tamponade of the anterior chamber carries the possibility of a raised IOP by blocking of aqueous circling and thus might have a superadded effect on the IOP elevation, in addition to the compartment syndrome secondary to orbital cellulitis.

Treatment

Broad-spectrum antibiotics (ciprofloxacin and tinidazole) were prophylactically started in view of the compromised immune status. CD4 counts were reviewed and were found adequate and stable during the last year; therefore pulsed steroid (methylprednisolone 500 mg) therapy could be safely administered to treat the acute orbital cellulitis. Antiglaucoma medications were introduced promptly to bring the IOP under control. After only a single dose of high-dose corticosteroids, the patient showed significant improvement and therapy was continued for three days until complete resolution of proptosis was seen and extraocular movements had returned to normal. Topical corticosteroids (prednisolone acetate 1%) QID and cycloplegic agent (homatropine 2%) TID were also added as a standard postoperative regimen and to deal with intraocular inflammation. Resultant response was good with controlled anterior chamber inflammation.

Outcome and follow-up

After three days of pulsed steroid therapy, the patient had fully recovered. IOP was 18 mm Hg on topical antiglaucoma (brimonidine and timolol) drugs and RAPD had resolved. The periorbital swelling subsided and the extraocular movements returned (figure 5). The patient was discharged after a week and continued regular follow-up thereafter. IOP is controlled with the above drugs and visual acuity is maintained. Regular follow-up with the physician has been ensured to monitor HIV status and HAART.

Figure 5.

Clinical picture at 1 week showing significant reduction in proptosis of the left eye with improvement in lid oedema and conjunctival congestion.

Discussion

SOT is an effective adjunct for vitreoretinal surgery in the management of retinal detachment with proliferative vitreoretinopathy, diabetic vitrectomy, repair of giant retinal tears and other complex vitreoretinal procedures.¹⁷ However, there are a number of associated complications, such as subconjunctival and orbital migration of silicone oil, keratopathy and corneal decompensation, secondary glaucoma, retinal changes, cataractous changes and intraocular emulsification, which have been reported with silicone oil left in-situ for a prolonged period of time. This was especially true in our patient since he was lost to follow-up after primary surgery. To minimise such complications, oil removal should subsequently be endeavoured around three months.¹⁸ Moreover, when present, the severity of inflammation has been observed to be greater with emulsified silicone oil.¹⁹

We believe that, in our case, emulsified silicone oil migrated into the subconjunctival space through peritomy, causing a foreign body-triggered, exaggerated inflammatory response and resulting in orbital cellulitis. An individual case of a patient with HIV with silicone oil migration into the brain has been described in a study by Eller et al.²⁰ We report an atypical case of aseptic orbital cellulitis in an HIV-infected patient with subconjunctival migration of silicone oil. To our knowledge such an association has not been reported earlier and the clinicopathological correlation performed in this case did provide us with a definitive answer towards the underlying pathological process, which eventually helped us achieve successful outcome in this clinical management challenge.

Silicone oil migration into the subconjunctival space subsequent to pars plana vitrectomy with SOT, SOR and even via an aqueous drainage implant (Ahmed Glaucoma Valve - AGV and Baerveldt implant) has been reported in the literature; however, its association with orbital cellulitis and lipogranulomatous inflammation is an unlikely phenomenon due to its biocompatible and inert nature. Histopathology of the eviscerated orbital tissue revealed multiple vacuolated spaces containing oil encapsulated by fibrous septae. Fibrous septae with skeletal muscle were accompanied by foreign body giant cells.⁶

Silicone oil intrusion into the subconjunctival space through the Molteno glaucoma valve was first reported by Hyung and Min²¹ in 1998. Histopathological examination showed non-specific inflammation along with fibrous septa encapsulating the cystic cavities. Similar cases of silicone oil migration into the subconjunctival space and orbit have been reported through the AGV by Nazemi et al in 2001⁶ and via the Baerveldt implant by Chan et al in 2005.²² Subconjunctival and ‘orbital silicone-oil granuloma’ (siliconoma) was also reported by Lee et al²³ in another case who underwent pars plana vitrectomy and SOT for endogenous endophthalmitis.

In all these reports, subconjunctival/orbital migration of silicone oil remains a common denominator. The presentation of orbital compartment syndrome in the setting of HIV infection is a feature that differentiates the index case from previously published reports. Silicone oil possibly triggered an exuberant inflammatory foreign body reaction and rapid deterioration. This case warrants further distinction in terms of management of silicone oil-associated orbital cellulitis using high-dose pulsed steroids therapy in the background of HIV infection. Orbital lipogranulomatous inflammation has been reported by Gonzalez-Fernandez and Kaltreider²⁴ in a case of iatrogenic Mycobacterium abscessus orbital infection following a blepharoplasty procedure in a 71-year-old woman. This was a dimorphic reaction with superficial purulent and deeper granulomatous process. Mustafa et al¹⁵ also reported findings in a case of a 20-year-old man who underwent PPV and SOT following traumatic globe rupture and retinal detachment. However, unlike our report, the former did not develop any orbital cellulitis.

Cases of silicone oil migration into the subconjunctival space have been reported where the oil had leaked via sutureless sclerotomies.^{25 26} In our case, sclerotomy ports were sutured; however, oil spillage might have occurred during closure, having found its way into the subconjunctival space. Nevertheless, some oil spill is expected during wound closure, which mandates a thorough saline wash before closing the Tenon’s and conjunctival peritomy. The patient did not undergo timely SOR. Henceforth, we presumed that following silicone oil emulsification, emulsified oil bubbles had migrated into the subconjunctival space, leading to chronic lipogranulomatous inflammation. Later on, the postoperative episode of surgically induced acute inflammation over the pre-existing chronic granulomatous inflammation triggered this exuberant acute on chronic reaction.

This rare presentation of orbital compartment syndrome due to silicone oil-associated orbital cellulitis in one-eyed patient with HIV infection presented a unique management challenge involving the use of high-dose steroid therapy, multidisciplinary liaison and satisfactory outcome. The stakes remained very high throughout, in view of the patient’s single-eyed status, previous ocular morbidity, hyperoleon and secondary glaucoma. However, the involvement of the patient in clinical decision making, combined with regular counselling sessions with the family members, helped us sail through this seemingly tumultuous storm. Lipogranulomatous inflammation, a hitherto unreported entity, in association with silicone oil migration was a unique finding which questions the extraocular biocompatibility of silicone oil.

Patient’s perspective.

I had already been on HAART for a few years, when I began having problem with my vision. I was diagnosed with retinal detachment in both eyes. That was a massive shock to me especially at a time when I was beginning to adapt to the challenges of HIV infection and the social stigma associated with the disease. The side effect of daily medications and a constant fear of contracting opportunistic infections had altered my day to day living. Losing vision would devastate me completely, I thought to myself, and could potentially take away my will to continue with life. While the medical team was discussing treatment options, it was revealed to me that my right eye is inoperable, while there is still some hope for the left eye. The medical team and I put all the eggs in one basket. Left eye was operated for retinal detachment repair and silicone oil was filled to keep the retina in place. The risk was enormous and the chance of losing vision completely was real. Post operatively the surgeons informed me that the surgery was difficult however they could achieve the intended outcome. I was happy to have gained some vision after surgery and was asked to follow up afterwards for routine postoperative care. I was also clearly instructed that the removal of silicone oil will be performed after 3 months. I followed up for one month and was glad to notice a gradual progressive improvement in vision with the left eye. Many of the medications could be stopped however a lubricant and another drug to lower eye pressure was continued. I returned to my routine and was able to perform daily chores, with some difficulty in reading. However, I was grateful for whatever could be salvaged. But due to my financial constraints, I didn’t follow up after one month as advised to plan for silicone oil removal. Six months after the surgery, I noticed progressively increasing pain and vision drop in my left eye, which had been fine until now. I rushed to the hospital and doctors told me to undergo another surgery for removal of silicone oil, as was planned earlier after 3 months of initial surgery. They said that due to the delay, the oil had gone bad and caused the eye pressure to rise abnormally. In order to control eye pressure before surgery, medications had to be started and only after that surgery could be undertaken. Next day, after surgery, I developed intense pain in the left eye with redness, watering and swelling around my eye. I was truly frightened because of my condition and felt guilty for not having followed medical advice regarding timely follow up and second surgery for oil removal. I was preparing myself to pay the heavy price for my recklessness with total blindness. I asked to see the same doctor who performed my surgery the first time. I was really please to meet him. He consoled me and pacified me, although I was afraid and uncertain of the outcome. The diagnosis at that moment was not definitive but they were suspecting orbital cellulitis or swelling around and behind the eye ball. I was informed that this problem can be seen following any kind of eye surgery but whether it is due to an infectious agent needs to be ascertained before definitive management is initiated. It was especially important in my case because I suffered from HIV which had already weakened my immune system and made me vulnerable to infections. I understood this well given the history of my HIV status. Another issue that loomed over me was the fact that the orbital swelling was compressing the optic nerve which connects my eye to brain and is essential to vision. My eye was already compromised with previous retinal detachment that required surgery and glaucoma caused by the oil that had emulsified because of my reckless attitude. I was informed that another insult to whatever is left of my eye could prove a final blow, stripping any useful vision that remained. The decision had to be made quickly. The team began examining my eye in detail and carrying out tests such as CT scan and swabs from the eye. ENT specialist was also consulted to locate any focus of infection from the nose or sinuses. It was ascertained that the swelling was not due to an infection and I would be treated with high dose corticosteroids. However, the cause of swelling was still unclear. The decision was made the same day and I was explained the side-effects of the corticosteroid’s infusion, given my immunosuppressive status due to HIV. Opinion of my infectious disease physician was also sought on the issue of treating my condition with corticosteroids. His opinion was swift and there was a green light considering the urgency of my condition. I was convinced by his decision and gave my consent to start corticosteroids. After 3 days, my vision improved, swelling gradually subsided. I was also informed about the biopsy report and that I suffered an unusual kind of swelling in my eye which could be associated with the oil that had turned bad. I happily gave consent to use my data for academic and research purposes and for reporting this finding. I was able to move my eye freely. I was so relieved and grateful to get back vision in my only seeing eye. The gratitude I felt for being able to see enough to go about my daily business, cannot be put into words. I am and will always remain immensely thankful to the team, that helped me through a very difficult phase of my life. Not only the management of the problem was efficient, but I truly appreciate the manner in which I and my family members were involved in decision making throughout. It is often not easy for patients and their loved ones to make tough medical decisions based on risk probabilities and statistics. When the treating physician becomes part of the decision-making process as much as a patient’s family, it not only makes decision making easier but also gives confidence to the doctors to give their best.

Learning points:

• Silicone oil migration into the subconjunctival space can occur via glaucoma drainage devices or following pars plana vitrectomy, leading to foreign body inflammatory response or lipogranulomatous inflammation.

• Aseptic orbital cellulitis and orbital compartment syndrome can be rare presentations associated with silicone oil migration into the subconjunctival space/orbit and warrant urgent intervention to prevent irreversible optic nerve damage.

• Effective management using pulsed steroid therapy may be achieved even in HIV-positive cases provided the patient is on regular HAART, has adequate CD4 counts and without opportunistic infections.

• Variable ‘oil-spill’ is expected and often occurs during vitreoretinal procedures; hence a thorough saline wash is necessary after sclerotomy suturing and before closure of conjunctival peritomy.

• The patient’s involvement in the clinical decision making and management can go a long way towards achieving satisfactory outcomes, especially when the stakes are high and prognosis is guarded.