Postoperative necrotizing scleritis: a report of four cases.

Postoperative necrotizing scleritis should be considered in cases of persistent localized postoperative inflammation following all forms of surgical trauma. We present the history, clinical findings, and follow-up data of four patients with postoperative necrotizing scleritis. The clinical records of four patients who developed scleritis following ocular surgery were retrospectively reviewed. The first step in managing necrotizing scleritis is to rule out infectious etiology. Surgically induced necrotizing scleritis is an immune-mediated condition that can coexist with concomitant infectious condition, i.e. endophthalmitis, but response to immunosuppression leads to resolution of the disease and verifies the diagnosis.

INTRODUCTION

Scleral inflammation and necrosis can complicate the postoperative course of ocular surgery and can lead to sight-threatening complications.1234 Scleral inflammation and necrosis can occur at variable time intervals after different anterior and posterior segment ocular surgeries including pterygium, cataract extraction, trabeculectomy, vitrectomy, and penetrating keratoplasty.1234567891011 Although the exact etiology of surgically induced necrotizing scleritis (SINS) has not conclusively identified, it has been noted that SINS usually occurs in eyes that have undergone multiple surgeries.1 Hence, some have postulated that SINS may be a result of a hypersensitivity reaction against an antigen revealed or altered due to multiple surgeries.121314 For appropriate management of postoperative scleritis, it is imperative to rule out infectious etiology first and immunosuppressives can be considered in non-responsive cases where infectious etiology has been ruled out.

The present case series reports the history, clinical findings, and management of four cases of SINS at a tertiary eye care centre.

CASE REPORTs

The clinical records of four patients who developed scleritis following ocular surgery were retrospectively reviewed. The details of surgery, risk factors in the previous medical history, features of the scleritis, and the treatment were noted for the four eyes. Findings were reviewed of the slit-lamp examination and indirect ophthalmoscopy, anterior segment photographs, and laboratory investigations, i.e., complete hematology and immunology profiles such as rheumatoid arthritis (RA), antinuclear antibody (ANA), and anti-neutrophil cytoplasmic antibodies (ANCA) and serology testing such as venereal disease research laboratory (VDRL) and treponema pallidum hemagglutination (TPHA). In all four cases, infective scleritis was excluded by wound swabs cultured on infusion broths and agar plates (blood, chocolate, and Sabouraud's).

Case 1

A 62-year-old hypertensive female presented with sudden pain and reduced vision in her right eye four weeks after undergoing surgery for macular hole. She had undergone an uneventful phacoemulsification, intraocular lens implantation, 23-gauge vitrectomy, internal limiting membrane peeling, and perfluoropropane (C3F8) tamponade in her right eye. There was no wound leak or hypotony postoperatively.

The best corrected visual acuity (BCVA) was 20/200, N36 and 20/40, N6 in right and left eyes, respectively. Slit-lamp examination revealed localized episcleral congestion at the site of the infusion port along with anterior chamber inflammation and 0.5 mm of hypopyon [Figure 1a]. Fundus examination with indirect ophthalmoscopy revealed an attached retina with visualization out to the third order vessels. An aqueous tap was performed and patient was treated with intravitreal vancomycin (1 mg/0.1 ml) and ceftazidime (2.25 mg/0.1 ml) in the right eye. She was started on fortified vancomycin and ceftazidime eye drops along with topical corticosteroids and mydriatics. Aqueous aspirate revealed Pseudomonas aeruginosa on culture. Subsequently, the intravitreal injections were repeated twice over the following week. At one week follow up, the patient had improved symptomatically but the sclerotomy site showed scleral thinning and necrosis along with deep stromal infiltration in the adjoining peri-limbal cornea. Microbiological assessment of scleral scrapings was negative for microorganisms. At two weeks follow up, the infusion cannula site showed progressive thinning [Figure 1b]. Serum RA factor, ANA c-ANCA, and p-ANCA were negative. The patient was diagnosed with SINS and was started on oral prednisolone (1 mg/Kg) and weekly oral methotrexate (7.5 mg). After six weeks of oral immunosuppressive therapy, the vision in the right eye vision improved to 20/120, N 8 with no pain and no sign of active scleral necrosis with corneal infiltration.

Figure 1

(a) Localized episcleral congestion and hypopyon in case 1, (b) Scleral thinning and necrosis at the site of sclerotomy in case 1

Case 2

A 48-year-old diabetic female developed acute pain in her left eye, seven weeks after undergoing surgery for diabetic vitreous hemorrhage. She had undergone phacoemulsification with intraocular lens implantation along with 20-gauge vitrectomy and endolaser. BCVA was 6/12, N8 in both eyes. Slit-lamp examination revealed scleral necrosis at the superotemporal sclerotomy site. Serum RA factor, ANA c-ANCA and p-ANCA were negative. Microbiological assessment of scleral scrappings did not reveal any microorganisms. The patient was started on oral prednisolone (1 mg/Kg) along with topical corticosteroids and ciprofloxacin. At one week follow up, the area of scleral necrosis was noted to have increased. The patient was started on oral methotrexate (15 mg/week). At three weeks follow up, the patient was asymptomatic with a small patch of scleral thinning persisting at the sclerotomy.

Case 3

A 55-year-old diabetic and hypertensive male presented with gradually increasing pain, redness, and swelling in his left eye seven weeks after undergoing surgery for diabetic vitreous hemorrhage. He had undergone underwent 20-gauge vitrectomy, membrane peeling, and endolaser.

Slit-lamp examination revealed localized conjunctival congestion and whitish deposits in the subconjunctival space suggestive of exudates at the superonasal sclerotomy site [Figure 2a]. Fundus examination with indirect ophthalmoscopy revealed an attached retina with visualization of the third order vessels. The patient was requested to start topical steroids and topical and systemic antibiotics. At one week follow up, his pain recurred with increasing intensity and surgery was performed to drain an abscess and to perform an antibiotic wash in the operating room. The culture was positive for Pseudomonas species. Topical fortified vancomycin and tobramycin drops were added. During the following two weeks, the pain and redness persisted and scleral ectasia was noticed at the site of sclerotomy along with adjacent corneal scarring [Figure 2b]. Suspecting it to be a case of surgically induced necrotizing scleritis, the patient oral prednisolone and cyclophosphamide (75 mg daily) therapy was initiated. Serum RA factor, ANA c-ANCA, and p-ANCA were negative. There was reduction in pain and redness a day after starting oral steroids. Cyclophosphamide had to be stopped due to deteriorating nephropathy. Follow up could not be completed as the patient's general condition worsened and had to be shifted to a general hospital.

Figure 2

(a) Congestion at superonasal sclerotomy site in case 3, (b) Scleral thinning and adjacent corneal scarring in case 3

Case 4

A 61-year-old hypertensive female with advanced glaucoma in both eyes with BCVA 20/40, N8 in right and 20/30, N12 in left was scheduled to undergo trabeculectomy with mitomycin-C in the right eye. She was pseudophakic bilaterally and had a history of YAG peripheral iridectomy. Intraoperatively, subconjunctival hemorrhage was noted after passing a bridle suture beneath the superior rectus muscle. Over next few minutes, a subconjunctival hematoma was noted at the 12 o'clock limbus. Further complicated by the intraoperative rise of blood pressure, the surgery had to be cancelled. Intraoperative fundus examination after dilatation did not reveal retinal or subretinal hemorrhage in the superior periphery. There was no hypotony. The patient was requested to start topical antibiotics and was referred to a general physician to control hypertension. Two weeks later, the patient presented to us with increased pain in her right eye localized subconjunctival abscess near the superior limbus.

Drainage of abscess was performed and a sample was sent for microbiological analysis. A circumscribed area of scleral necrosis was noted starting from limbus and extending beneath the insertion of the superior rectus. Microbiological analysis did not reveal any microorganism. Patient was maintained on topical corticosteroids and ciprofloxacin eye drops. At one week follow up, the necrotic scleral area continued to enlarge. At this point, we suspected this was a case of SINS. The patient was prescribed oral prednisolone (1 mg/Kg) and methotrexate (15 mg/week). At next follow up after one week, there was no increase in the size of the scleral necrosis. Over the following weeks there was reduction in pain and redness. At two months follow up, the active inflammation reduced with formation of a localized staphyloma [Figure 3]. Her final BCVA was 20/60, N10.

Figure 3

Localized staphyloma at two months follow up in case 4

DISCUSSION

The current theory is that classic SINS involves a delayed hypersensitivity response that can be precipitated by mild surgical trauma or ischemia-exposed tissue antigens resulting in sensitization of the immune system.56 Alternate theories include molecular mimicry possibly triggered by infection or generalized immune complex deposition.146 The fact that SINS responds to and is controlled by immunosuppressive therapy supports these hyopotheses.146

The present group of cases differ from the description of classic SINS in view of the early presentation (3-8 weeks versus 9 months).1 There was no history of previous ocular surgery in one of the patients in whom surgery was cancelled after passing the bridle suture (patient 4). Although SINS is a noninfectious inflammatory process, it can be preceded by an infection.15 Bacterial superinfection is also known to complicate the presentation and adds to the diagnostic dilemma.9 The inherent limitations in the term SINS, including difficulty to determine the relative contributions of trauma, epithelial breakdown, immune activation, and infection makes the diagnosis of SINS tenuous at best.16 One of our patients was positive for Pseudomonas aeruginosa from aqueous aspirate and another patient was positive from scleral abscess samples. However, both of patients improved clinically only after starting immunosuppressive therapy. Therefore, exclusion of infection, and treatment of an infection, if present, is critical prior to diagnosing SINS. The other two patients (Cases 2 and 4) had single microbial cultures which were negative.

Our series of patients were either diabetic or hypertensive, which could have contributed to the local ischemia at the site of surgical wound. Polyglactin suture has been described to increase the local inflammation and enhance the antigen presentation in eyes developing SINS following pars plana vitrectomy.10 This might be related to the predilection of the sclerotomy site for the development of SINS in our series of patients.

Immunosuppressive therapy in the form of oral corticosteroids, methotrexate, and cyclophosphamide are the mainstay of treatment for SINS.139 In our series, three patients had resolution of scleritis after institution of immunosupressives. Two of them had scleral thinning at the site of scleritis. A high index of suspicion may help in early diagnosis and treatment, thereby preventing adverse outcomes such as the formation of staphyloma and scleral melting. Persistence of pain or sudden reappearance of pain postoperatively, which would not improve despite adequate topical or systemic anti-inflammatory medication should lead us to consider scleritis as one of the possibilities. While immunosuppression clearly plays a role in the management of postoperative necrosis that occurs in the setting of systemic immune disorders and often plays a role in the absence of such underlying conditions, it is extremely important to consider and rule out infection prior to placing patients on long-term immunosuppression therapy. Although our patients responded well to immunosuppressives a single culture may not have been sufficient to completely rule out infection and this is a drawback of our case series. Our case series has limited follow up and that may not be sufficient to rule out infection completely or to confirm the benefit of immunosuppression conclusively. Treating or excluding the ocular infection and infectious scleritis prior to starting potent anti-inflammatory medications and immunosuppressives may be the correct approach in the management of postoperative scleritis.