Sir,

We report an unusual case of recurrent endogenous Klebsiella endophthalmitis.

Case report

A 58-year-old HIV-negative male with diabetes mellitus, intravenous drug use, and Klebsiella pneumoniae sepsis complained of floaters of the right eye following 6 days of intravenous ceftriazone. Visual acuity of the right eye was 6/120 with 3+ anterior chamber cells, posterior synechiae, but no hypopyon. Fundal examination showed vitritis and choroidal abscesses. The left eye was normal.

Intravitreal vancomycin and ceftazidime were given following vitreous tap that showed no bacterial growth. Intravenous ceftriaxone was continued for another 4 weeks with topical ciprofloxacin 0.3%. Dexamethasone 0.1% was added at 4 weeks. The inflammation improved and topical medications were stopped at 3 months. At 4 months, visual acuity of the right eye was 6/30, with quiet anterior chamber, mild vitreous haze, a small choroidal scar, but no active lesions.

After 1 month, he developed severe right panopththalmitis (Figure 1a and b). Visual acuity of the right eye was hand motions with oedematous cornea, 4+ anterior chamber cells, with a hypopyon. IOP was 34 mm Hg. Fundal examination revealed multiple vitreous opacities with no choroidal abscesses. B-scan ultrasound showed thickened sclera. He was given intravitreal vancomycin and ceftazidime, topical ciprofloxacin, glaucoma medications, and intravenous ceftriaxone following vitreous tap, which was positive for Klebsiella pneumoniae, sensitive to ceftriaxone and ciprofloxacin. The septic workup was negative. Unfortunately, he rapidly worsened and the right eye was eviscerated 8 days later due to increasing pain, proptosis, and refractory glaucoma.

Figure 1
figure 1

(a) Recurrent episode showing proptosis of the right eye. (b) Slit lamp examination showing conjunctival chemosis.

Full size image

Comment

Recurrence following successful treatment of endogenous endophthalmitis (EE) is unknown. Our patient was well for 2 months before presenting as a relentless panophthalmitis. Although corticosteroids can decrease tissue destruction from the host inflammatory response, its use in EE remains controversial with visual outcomes ranging from no perception of light to better than 6/60.1, 2, 3 In our patient, the use of topical steroids as well as his diabetes mellitus may have allowed the organism to become sequestered rather than being totally eradicated, hence permitting a recurrence, which proved to be more fulminant than the first episode.

In conclusion, despite apparently successful treatment of EE, ophthalmologists should be aware of possible recurrence, especially in patients who had received corticosteroids.