Silicone oil tamponade in managing recalcitrant endophthalmitis after cataract surgery secondary to Clostridium intestinale

  1. Sara E Francomacaro 1,
  2. Janani Singaravelu 2,
  3. Rithwick Rajagopal 2 and
  4. Albert S Li 2 , 3
  1. 1 Department of Ophthalmology & Visual Sciences, University of Maryland School of Medicine, Baltimore, Maryland, USA
  2. 2 John F. Hardesty, MD, Department of Ophthalmology & Visual Sciences, Washington University in St Louis School of Medicine, St Louis, Missouri, USA
  3. 3 Vitreoretinal Consultants of New York, Northwell Health Eye Institute, Department of Ophthalmology, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Great Neck, New York, USA
  1. Correspondence to Dr Albert S Li; AlbertLiMD@gmail.com

Publication history

Accepted:03 Nov 2022
First published:01 Dec 2022
Online issue publication:01 Dec 2022

Case reports

Case reports are not necessarily evidence-based in the same way that the other content on BMJ Best Practice is. They should not be relied on to guide clinical practice. Please check the date of publication.

Abstract

A woman presented two weeks after uncomplicated cataract surgery with decreased visual acuity from endophthalmitis. One week after initial management with intravitreal antibiotics, her visual acuity decreased further, undergoing pars plana vitrectomy with intravitreal and intravenous antimicrobial coverage with preliminary improvement. Three days after vitrectomy, her vision decreased with recurrent inflammation. Initial cultures grew Clostridium intestinale. She underwent repeat vitrectomy with silicone oil tamponade with no subsequent recurrence. The silicone oil was removed after 4 months and her visual acuity returned to 20/20 after 1 month and through 1 year of follow-up. Postoperative endophthalmitis is rare, with cases due to Clostridium species particularly destructive. In this first reported case of C. intestinale endophthalmitis, conventional management did not achieve lasting quiescence until silicone oil tamponade was employed. Pars plana vitrectomy with silicone oil tamponade should be considered in the management of recurrent endophthalmitis or endophthalmitis secondary to a recalcitrant microbe.

Background

Cataract surgery is the most performed operation in the USA, with the number of surgeries performed annually on the rise.1 Complications after cataract surgery are rare, but potentially devastating as in cases of postoperative endophthalmitis. The incidence of postoperative endophthalmitis in cataract surgery averages around 2 in 1000 cases or 0.2%.1

The source of infection in postoperative endophthalmitis is predominantly organisms of the lids and conjunctiva that gain access to the eye during surgery or through the corneal wound postoperatively.2 Most isolated organisms are gram positive (94%) with the majority of those being coagulase-negative Staphylococci (70%).3 Rarer organisms, not typically found on the eye surface or adnexal structures, have also been reported to cause postoperative endophthalmitis, including exogenous organisms from the operating room.4 These can include gram-positive spore-forming bacilli such as the aerobic Bacillus species or the anaerobic Clostridium species. More commonly, however, these bacilli are associated with cases of endophthalmitis resulting from penetrating injuries to the eye, particularly those that involve organic material.5

Clostridium species are a rare cause of endophthalmitis and tend to be associated with poor visual outcomes.5 For example, endophthalmitis secondary to Clostridium perfringens often fails medical and surgical management, ultimately resulting in loss of the eye.6 In this report, we describe a case of endophthalmitis secondary to C. intestinale, an aerotolerant gram-positive spore-forming bacilli. It is believed to be an organism of gastrointestinal colonisation, and to date has only been reported as the source of one human infection, a case of bacteraemia.7 To our knowledge, this is the first report in the literature of C. intestinale causing endophthalmitis. Herein, we describe the clinical course of this patient, which was initially refractory to conventional management of endophthalmitis but resolved after pars plana vitrectomy (PPV) and silicone oil tamponade with excellent visual outcomes.

Case presentation

A woman in her seventh decade of life was referred to the retina service for evaluation of acutely decreased vision in the right eye, 2 weeks after uncomplicated sutureless cataract extraction by phacoemulsification through a clear corneal incision, and monofocal one-piece Tecnis ZCB00 intraocular lens (Abbott Medical Optics, Santa Ana, California, USA) placement in the capsular bag. She had no medical history and ocular history was only notable for refractive error and presbyopia, she had no preoperative corneal abnormality. Following surgery, she was using gatifloxacin 0.5% and prednisolone acetate 1% four times daily as directed. Her examinations on postoperative day 1 and week 1 were as expected, without signs of infection, and she achieved a visual acuity (VA) of 20/15.

On presentation to the retina service on postoperative week 2, VA in the right eye was decreased to 20/50 with normal intraocular pressure, marked anterior chamber fibrinoid reaction without hypopyon, and 3+ vitreous cell with debris limiting posterior view. Given poor posterior view, the patient underwent ophthalmic B-scan ultrasonography to rule out alternate diagnoses and to monitor the course of the vitritis over time (figure 1). The patient did not have any associated eye pain.

Figure 1

B-scan on presentation. B-scan ultrasonography through the macula, with the probe in longitudinal positioning at 9 o’clock, of the patient’s right eye. Images from day of presentation to the retina service, 2 weeks after cataract surgery. Examination notable for 3+ vitreous cell with debris.

Treatment

The fulminant inflammation after recent intraocular surgery was concerning for endophthalmitis, so an anterior chamber paracentesis and vitreous tap were obtained and sent for gram stain, aerobic, anaerobic and fungal cultures. The patient was empirically treated at that time with intravitreal vancomycin (1 mg/0.1 mL) and ceftazidime (2.25 mg/0.1 mL) and monitored daily for changes in ophthalmic examination. The next day, her vision was 20/300 with development of a small hypopyon, which resolved the following day with slight improvement of VA to 20/200.

The gram stain from the anterior chamber paracentesis resulted positive for rare polymorphonuclear leucocytes, without organisms, and the patient’s examination over the next couple of days remained stable. On postoperative week 3, 1 week after intravitreal antibiotics, the patient’s VA worsened to 20/400, with increased anterior chamber inflammation, a recurrent hypopyon, and vitritis (figure 2). Anterior chamber paracentesis and vitreous tap were repeated and again sent for gram stain and culture. Intravitreal injections of vancomycin (1 mg/0.1 mL) and ceftazidime (2.25 mg/0.1 mL) were also repeated, with the addition of voriconazole (100 μg/0.1 mL) due to lack of improvement and concern for possible fungal infection. At this point, all cultures were without growth. Two days later, vision worsened to counting fingers and a 25-gauge PPV (Alcon Constellation, Fort Worth, Texas, USA) with partial air-fluid exchange was performed, thoroughly removing the vitreous, with repeated administration of intravitreal vancomycin (1 mg/0.1 mL), ceftazidime (2.25 mg/0.1 mL) and voriconazole (100 μg/0.1 mL). Given concern for vision loss in a patient with an undiagnosed panuveitis suspicious for endophthalmitis, a surgical vitreous sample was sent for diagnostic testing covering a significantly broadened differential. The sample was sent for cytology, aerobic, anaerobic and fungal cultures, in addition to herpes simplex viruses 1 and 2, varicella zoster virus and cytomegalovirus PCR testing, acid-fast bacilli smear and culture, and mycobacteriology.

Figure 2

B-scan one week after presentation. B-scan ultrasonography through the macula, with the probe in longitudinal positioning at 9 o’clock, of the patient’s right eye. Images from 1 week after initial intravitreal antibiotics, 3 weeks after cataract surgery. Examination notable for drop in vision.

The same day, after the patient underwent vitrectomy, 9 days after the initial ocular sample was taken, preliminary culture results from the initial anterior chamber paracentesis grew non-perfringens Clostridium species, which was sent to an outside laboratory for confirmation, identification and sensitivity testing. All other anterior chamber and vitreous sample testing resulted negative. On post vitrectomy day 1, the patient had a marked improvement in VA to 20/30, and only mild residual anterior chamber inflammation without vitritis. However, on post vitrectomy day 3, the patient’s VA dropped to 20/200, with recurrent anterior chamber and vitreous inflammation concerning for recurrence of endophthalmitis. The patient was admitted to the hospital for an infectious disease consultation and systemic intravenous ceftazidime, vancomycin and metronidazole were initiated to target the unknown Clostridium species. The following day, she underwent repeat 25-gauge PPV, during which minimal residual vitreous was noted, with injection of intravitreal vancomycin (1 mg/0.1 mL), amikacin (200 μg/0.05 mL) and tamponade with 5000-centistoke silicone oil injected in the typical fashion with a viscous fluid injector. Two days later, the patient’s vision improved to 20/30 with only mild residual intraocular inflammation and new fibrotic membrane over the intraocular lens (IOL). She was discharged on a 2-week course of oral metronidazole per infectious disease recommendations.

Outcome and follow-up

Confirmatory testing identified the pathogen as C. intestinale, and the isolate was found by minimum inhibitory concentration to be susceptible to vancomycin, piperacillin/tazobactam, ertapenem, clindamycin and metronidazole, indicating that these antibiotics at reasonable doses prevented microbial growth, and had intermediate susceptibility to penicillin, which had a minimum inhibitory concentration of 1 μg/mL. The patient was followed closely for 4 months without infectious recurrence, after which silicone oil was removed. One month after silicone oil removal, the eye remained quiet, and uncorrected VA remained 20/20 (figure 3). At her most recent follow-up, at 13 months after her cataract surgery, the affected eye remains quiescent, without new ocular diagnoses, with an uncorrected VA of 20/20.

Figure 3

Postoperative fundus photograph. Fundus photograph of the patient’s right eye 1 week after silicone oil removal. The dilated fundus examination was normal with uncorrected visual acuity of 20/20.

Discussion

Clostridium species are fulminant bacteria that have proven challenging to treat in endophthalmitis, with many cases resulting in enucleation after unsuccessful medical and surgical management.5 In this first reported case of endophthalmitis secondary to C. intestinale, our patient responded minimally to intravitreal broad-spectrum antibiotics alone, despite later sensitivities showing susceptibility to vancomycin, which was injected intravitreally multiple times. As a result, she was managed with PPV, with subsequent significant improvement in VA and inflammation, but the recurrence 3 days later suggested the infection was not yet completely controlled. She obtained lasting improvement with quiescence of intraocular inflammation only after a repeat PPV with silicone oil tamponade. The confluence of multiple approaches applied in a stepwise fashion, including local and systemic antibiotics, debulking the vitreous via PPV, and silicone oil tamponade, ultimately resulted in sterilisation of the eye. Our patient ultimately achieved an excellent visual outcome after silicone oil removal and remains at an uncorrected VA of 20/20 without signs of infectious recurrence more than a year after cataract surgery.

C. intestinale was initially discovered by Lee et al in 1989 while performing identification studies on the faecal flora of healthy cattle.8 This strain was also found to match Clostridia isolated from grassy plants in a study by Minamiswa et al in 2004, presumed to be ubiquitous in soil.9 C. intestinale has only been previously reported in the literature as causing a human infection once. Elsayed and Zhang published in 2004 on a 17-year-old girl without medical history, who presented to the emergency department with right lower quadrant pain.7 The patient’s initial lab workup, gynaecological examination and abdominal ultrasound were unremarkable. Two separate blood cultures collected on admission resulted positive within 48 hours with large gram-positive, boxcar-like, rod-shaped bacteria, and were ultimately classified by partial gene sequencing to be C. intestinale.7 The strain was sensitive to penicillin G, clindamycin, metronidazole and vancomycin, and the patient received intravenous ampicillin for 48 hours followed by 7 days of oral amoxicillin.7 Our patient’s medical history was also unremarkable, without recent trauma, gastrointestinal disease, malignancy or organic matter exposure. It is likely that our patient’s infection was exogenous rather than endogenous, in the setting of recent intraocular surgery; however, the exact source of her endophthalmitis is unclear. The recalcitrant course of this endophthalmitis suggests an atypical organism as the cause, lending credence to C. intestinale as the aetiology of the infection, rather than a possible contaminant of the culture. The difficulty in culturing C. intestinale, coupled with the prior administration of empiric antibiotics to which this strain was sensitive, likely led to our inability to isolate the organism on subsequent ocular fluid cultures.

The natural history of our patient’s C. intestinale endophthalmitis was similar to other reported cases of refractory endophthalmitis caused by gram-positive spore-forming bacilli. Chen and Roy reported in 2000 on four cases of recurrent Bacillus circulans endophthalmitis after cataract surgery with contaminated viscoelastic.10 Of 42 patients exposed to the viscoelastic, 14 were diagnosed with endophthalmitis, 11 did not respond to intravitreal antibiotics and underwent core PPV with apparent resolution, and 4 presented with recurrent endophthalmitis which was refractory to repeated PPV and intravitreal antibiotics.10 Three cases ultimately underwent IOL explantation and capsule removal prior to ultimate resolution of the infection.10 Of note, none of these cases were managed with silicone oil tamponade.

Given that our patient did not fully respond to intravitreal injections and core PPV, it is possible she had a non-clearing nidus of infection, though such a nidus was not apparent on examination. Prior reports of recurrent postoperative endophthalmitis, after multiple failed intravitreal antibiotic injections and core PPV, describe successful control of the infection with IOL removal, with or without capsulotomy, whether or not a non-clearing nidus of infection is apparent.11 Many of these reported cases were secondary to Propionibacterium acnes,11 though other causative microbes have also been reported.12 Should she have not responded to repeat vitrectomy with silicone oil tamponade, IOL removal was considered to control the infection. While the disadvantages of silicone oil tamponade include the need for a subsequent surgery to remove the silicone oil, removal of the IOL and capsule also assumes the need for a subsequent surgery if optimal refractive outcomes are desired without the need for aphakic contact lens.

Silicone oil was initially found to be bactericidal in 1993 by Ozadamar et al.13 There have been a few examples in the literature since then of silicone oil being a useful adjunct in the treatment of endophthalmitis.14–16 In a randomised controlled study, by Kaynak et al, of 54 patients who met criteria for primary vitrectomy according to the Endophthalmitis Vitrectomy Study: 24 eyes were randomised to management with core PPV versus 28 eyes randomised to complete PPV with silicone oil, encircling band and endolaser.14 Three eyes in the core PPV group had recurrent endophthalmitis compared with none in the silicone oil group.14 Their finding of adjunctive silicone oil success in preventing recurrent endophthalmitis has been repeated in studies where initial intravitreal antibiotics were attempted and in studies including more diverse bacteria.15 16 Azad et al performed a randomised controlled study including 24 cases of post-traumatic endophthalmitis who had failed primary management with intravitreal injection of vancomycin and amikacin.15 Twelve patients underwent core vitrectomy alone versus 12 patients who underwent complete vitrectomy with silicone oil tamponade. One patient of the core vitrectomy group had recurrent exudates compared with none in the silicone oil group.15 Pinarci et al performed a non-comparative study of PPV with silicone oil tamponade in eight patients who presented with acute endophthalmitis after intravitreal injection and failed initial management with intravitreal antibiotics.16 Endophthalmitis was controlled in all eight patients after silicone oil tamponade.16 These studies suggest a role of silicone oil tamponade in combination with PPV in the management of endophthalmitis, given the bactericidal effect of silicone oil.13–16

In summary, we present a case of postoperative endophthalmitis secondary to C. intestinale, which did not respond to initial management with intravitreal injections of vancomycin, to which the C. intestinale isolate was sensitive, and debulking of the vitreous via PPV. Quiescence was achieved only after repeat vitrectomy and intravitreal and systemic antibiotics with silicone oil tamponade, indicating how resistant this infection was to treatment. Our patient obtained an excellent final visual outcome after silicone oil removal, suggesting that long-term silicone oil tamponade can be combined with a repeat PPV to successfully treat cases of endophthalmitis secondary to C. intestinale, as well as being another option in the treatment of other recalcitrant endophthalmitides.

Learning points

  • Postoperative endophthalmitis secondary to Clostridium can be particularly destructive and refractory to treatment.

  • Clostridium intestinale is a coloniser of the gastrointestinal tract and has been identified as a cause of bacteraemia in humans in one case in the literature; this report describes a case of postoperative endophthalmitis secondary to C. intestinale, which has not been previously known to cause ocular infections.

  • Initial improvement of postoperative endophthalmitis after injection of broad-spectrum antibiotics does not preclude worsening later in the clinical course; in such cases that fail to improve or that worsen after initial improvement, other etiologies should be considered such as fungal, viral or recalcitrant microbes such as spore-forming bacteria.

  • Pars plana vitrectomy with silicone oil tamponade should be considered in the management of recurrent endophthalmitis or endophthalmitis secondary to a recalcitrant microbe; this demonstrated efficacy in the definitive management of this case of endophthalmitis secondary to C. intestinale.

Ethics statements

Patient consent for publication

Acknowledgments

The authors wish to acknowledge Drs. Brigid Marshall and Harris Sultan for sharing in the care of this patient.

Footnotes

  • Twitter @AlbertLiMD

  • Contributors SEF performed the literature review, created the figures, was a major contributor in writing the manuscript. JS acquired data for the manuscript and was a major contributor in writing the manuscript. RR participated in the care of this patient and was involved in revising the manuscript. ASL participated in the care of this patient, made substantial contributions to the conception of the manuscript, and substantively revised the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

References

Use of this content is subject to our disclaimer