Post-Operative Endophthalmitis following Routine Photorefractive Keratectomy Open Access

Endophthalmitis following laser corneal refractive surgery is an exceptionally rare but devastating complication, with only two previous cases reported in the peer-reviewed literature [1, 2]. The incidence of post-operative infectious keratitis is reported by a 2020 meta-analysis to be 0.0006% and 0.0001% following laser-assisted in situ keratomileusis and photorefractive keratectomy (PRK), respectively [3]. Combined with the infrequent rates of endophthalmitis secondary to infectious keratitis reported to be between 0.3 and 0.5% [4, 5], this would suggest a very low overall risk of post-operative infectious keratitis progressing to endophthalmitis following routine corneal refractive surgery.

In the current report, a rare case of P. aeruginosa endophthalmitis and corneal perforation following routine myopic PRK and mitomycin C (MMC) application is presented, which required urgent pars plana vitrectomy (PPV) facilitated by temporary keratoprosthesis (KPro) and subsequent therapeutic penetrating keratoplasty (PKP). The case was further complicated by the management of acute anaphylaxis at anaesthetic induction, which culminated in aggressive intra-operative posterior positive pressure (PPP) and spontaneous extrusion of the iris and crystalline lens.

A 34-year-old immunocompetent female was initially referred to our tertiary ophthalmic service for emergency management of infectious keratitis following routine PRK performed in another centre. The referring clinician reported an unremarkable pre-operative assessment except for meibomian gland dysfunction which was treated for 1 month before surgery. The patient underwent routine myopic PRK correction (−4 dioptres) with MMC 0.02% application for 12 s. Both eyes were treated on the same day with the same protocol. She was discharged home with a bandage contact lens, topical dexamethasone 0.1% and chloramphenicol 0.5% 4 times daily. As the case was performed outside of our institution, we did not have access to the pre-operative tomography. The operating surgeon reported to us that it was within normal limits.

The day 1 post-operative review was reported to be within normal limits. She was subsequently referred to our tertiary ophthalmic service with severe, limbus to limbus, suspected infectious keratitis 4 days post-operatively. Table 1 outlines the clinical timeline.

Corneal scrape cultured P. aeruginosa, with intermediate sensitivity to ceftazidime, ciprofloxacin, and piperacillin-tazobactam. Following 48 h of intensive hourly topical fortified cefuroxime 5% and tobramycin 1.36% around the clock, ongoing clinical deterioration was noted. A B-scan ultrasonography exam was performed. This demonstrated significant vitreous echoes consistent with potential endophthalmitis, despite an intact cornea. Subsequent vitreous tap cultured P. aeruginosa and intravitreal broad-spectrum vancomycin 1 mg and ceftazidime 2 mg were promptly administered. On day 7 post-operatively, the cornea had perforated with iris plug formation (Fig. 1).

The decision to perform a PPV with repeat intravitreal antibiotic injection was made early on day 7, which required placement of a temporary KPro followed by therapeutic PKP. The patient received succinylcholine as part of a rapid sequence induction due to inadequate pre-operative fasting. Ten minutes post-induction, significant hypotension and loss of end-tidal carbon dioxide were detected, signifying life-threatening acute anaphylaxis that was most likely secondary to succinylcholine. The patient was managed according to the local anaphylaxis protocol with lowering of the head of the bed, and administration of 2 L intravenous colloid fluids, adrenaline, and hydrocortisone. Intra-operative serum tryptase was elevated to >200 ng/mL (normal range, 1–15 ng/mL) and reduced to 27 ng/mL 1 day post-operatively.

Twenty minutes after achieving haemodynamic stability, the decision was made to proceed with surgery. Partial thickness manual trephination was performed. Upon full-thickness excision of the first quadrant of the host cornea, a very high level of PPP was encountered, leading to significant iris prolapse. Although the orbit was not tight, the PPP persisted despite lifting the speculum. Intra-operative B-scan ultrasonography demonstrated no evidence of choroidal haemorrhage or effusion. Temporary sutures were placed to close the corneal wound and blind partial PPV was completed by our vitreoretinal colleagues to alleviate the PPP. However, on completion of the host corneal excision, the PPP persisted with spontaneous extrusion of the native crystalline lens, associated with ongoing anterior iris bowing and visualisation of a dense retrolental plaque. Manual pressure was placed on the globe to prevent further extrusion of ocular contents. The temporary KPro was sutured in place and a full PPV was performed which finally alleviated the PPP. No choroidal effusions or haemorrhages were detected. The retina showed evidence of necrosis and silicone oil was placed, and therapeutic PKP was performed (Fig. 2, 3).

At 6-month follow-up, the best-corrected visual acuity (BCVA) was light perception with the therapeutic graft in situ. Electrodiagnostic tests show poor visual potential. The graft is opacified, but the eye has not become phthisical. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000546142).

To our knowledge, there have only been two earlier reports of endophthalmitis following laser corneal refractive surgery in the peer-reviewed literature [1, 2]. Mulherne et al. [1] described a case of myopic laser-assisted in situ keratomileusis complicated by Streptococcus pneumoniae keratitis and subsequent endophthalmitis, which recovered to a final BCVA of 6/7.5. In a second case reported by Karth et al. [2] endophthalmitis occurred in an elderly 81-year-old PRK patient with pre-existing radial keratotomies. Methicillin-resistant Staphylococcus aureus was cultured from the vitreous sample in the absence of infectious keratitis, leading the authors to suggest that the pathogen entered the eye via the radial keratotomies. BCVA recovered to 6/30 at final follow-up [2].

The current case presented was particularly unusual because the rapid progression from infectious keratitis to endophthalmitis occurred in a young and immunocompetent patient, and before the development of corneal perforation [6]. It is likely that the use of the MMC and the aggressive nature of P. aeruginosa might have contributed to the very rapid disease progression observed [6, 7]. P. aeruginosa resistance rates are increasing worldwide to many different commonly used antibiotics (ciprofloxacin [9%], gentamicin [22%], and ceftazidime [13%]) [8]. To the best of our knowledge, the patient reported adherence to the post-operative protocol of bandage contact lens, chloramphenicol and dexamethasone 0.1% therapy, and the referring clinician recorded an unremarkable day 1 post-operative examination.

Our case was also further complicated by aggressive intra-operative PPP that may potentially be secondary to a combination of severe pre-operative intraocular inflammation, as well as intra-operative succinylcholine anaphylaxis and its management [9‒12]. PPP develops when a high-to-low pressure gradient forms from the posterior to the anterior chamber. This can be secondary to an increase in the posterior pressure, a reduction in the anterior chamber pressure or a combination of both [10].

Succinylcholine is a muscle relaxant used to facilitate tracheal intubation in rapid sequence induction [13, 14]. It has many advantages, including rapid onset of action and intense blockade; however, there is a recognised risk of anaphylaxis [15]. Succinylcholine causes a transient elevation in intraocular pressure (IOP) of approximately 6–8 mm Hg, with peak effect at 2–4 min and normalisation occurring within 6 min [14]. The physiological mechanisms of this IOP rise remain controversial. Succinylcholine causes contraction of the extraocular muscles (EOMs), which could elevate IOP through external scleral pressure [13]. This could create a PPP gradient in an open eye, and the use of succinylcholine has, therefore, been discouraged by some clinicians in cases of open globe injury [13]. However, in a fellow-eye comparative study of patients undergoing unilateral enucleation, there was no significant difference in IOP increase in diseased eyes following EOM detachment and normal eyes with attached EOMs, suggesting that the ocular hypertensive effects of succinylcholine might be unrelated to EOM contraction [16]. Nevertheless, in the current case, due to the anaphylactic reaction, surgery was commenced well outside the typical duration of action of succinylcholine on IOP [14].

Anaphylactic reactions can be associated with significant swelling of the conjunctiva and periocular tissue secondary to altered vascular permeability and fluid shifts [11]. The mechanical effects on the external sclera might have contributed to the development of a pressure gradient between the posterior and anterior chambers, thereby partially contributing to the PPP observed in the current case [9‒11].

The treatment administered for the anaphylaxis event is also likely to have contributed to the PPP. As per the local emergency protocol, the head of the bed was lowered. This increases orbital venous and choroidal perfusion, which may both cause PPP [9]. Although initially hypotensive, haemodynamic stability was restored following administration of epinephrine and resuscitation with 2 L of colloid fluids over 45 min. This increase in intravascular volume may have contributed to increased choroidal perfusion [9, 12]. Straining on the endotracheal tube during the recusation may have also caused significant spikes in IOP [12].

PKP is a major risk factor for PPP. The incidence of PPP in PKP is estimated to be 40–50% [17, 18] with younger patients being at increased risk [18]. The removal of the host cornea lowers the anterior chamber pressure rapidly and a pressure gradient forms between the posterior and anterior chambers [18]. In addition, prolonged hypotony increases the risk of PPP developing [18]. In the current case, the corneal perforation with iris plugging had occurred prior to the initial incision. Significant PPP was already detected upon making the first full-thickness incision into the cornea, before a significant pressure gradient attributable to prolonged surgically induced hypotony could have developed. This would suggest that, in the current case, the main driving force behind the PPP was present before the removal of the donor cornea.

Endophthalmitis is an exceptionally rare complication following laser corneal refractive surgery. Infectious keratitis has a low risk of progression to endophthalmitis, particularly in younger immunocompetent patients. MMC use following PRK is likely to have contributed to the rapid progression encountered in this case. B-scan ultrasonography was useful in the diagnosis of endophthalmitis with an opaque cornea. Additionally, we report that acute anaphylaxis and its management can lead to PPP. Intraocular surgery should be delayed in such cases, and measures should be taken to reduce the risk of PPP before full-thickness incisions are made.

The patient was very disappointed with the development of this exceptionally rare complication. Her main concern at present is the appearance of the eye. She will meet with oculoplastics to discuss the cosmetic options available.

This study was performed in accordance with the Declaration of Helsinki. Ethics approval for this human study was waived by Greenlane Clinical Centre Ethics Committee. All adult participants provided written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for publication of the details of their medical case and any accompanying images. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

B.P. is the primary author. M.W., A.Z.C., and S.S. helped prepare the manuscript and read and approved the final manuscript.

The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.