Helen Chung,* an 83-year-old Chinese woman, had just finished celebrating her granddaughter’s high school graduation when she began to notice pain and redness in her right eye. As she had never had any eye problems in the past, she thought little of her symptoms and assumed the discomfort would resolve on its own. Over the next few days the pain and redness progressed, however, and she rapidly lost vision in her right eye. Subsequently, she developed pain, redness and loss of vision in the left eye. Six days after the initial onset of her symptoms, Ms. Chung sought medical attention at the ER.
We Get a Look
Ms. Chung had no light perception bilaterally. Her eyelids were diffusely red and swollen with marked chemosis of the bulbar conjunctiva in each eye (Fig. 1).
Both eyes had a fibrinous anterior chamber reaction, and the left eye showed 2-mm hypopyon formation. There was no view to the fundus in either eye. B-scan ultrasonography demonstrated the presence of vitreous debris and choroidal thickening bilaterally. Her IOP was 50 mmHg in the right eye and 52 mmHg in the left.
Systemically, her physical examination was unremarkable—she was afebrile and her vital signs were within normal limits.
A CT scan of the brain and orbits revealed marked inflammation within both globes and orbits (Fig. 2). There was no evidence of orbital hemorrhage or mass in either eye, nor was there any evidence of central nervous system involvement.
To decompress the orbits and normalize IOP, we performed bilateral lateral canthotomies and cantholyses. We then performed vitreous aspiration for Gram stain and culture followed by intravitreous injection of ceftazidime (2.25 mg/0.1 ml), vancomycin (1 mg/0.1 ml), dexamethasone (400 µg/0.1 ml) and amphotericin B (5 µg/0.1 ml). The aspirated vitreous fluid was turbid and dark yellow. The Gram stain showed numerous inflammatory cells, but revealed no organisms.
Ms. Chung was admitted to the hospital and started on topical moxifloxacin every hour, as well as topical prednisolone acetate four times per day. She was also placed on intravenous levofloxacin, 500 mg daily.
What's Your Diagnosis?
(Figure 1) Photo of the patient’s left eye taken in ER. Note the marked lid swelling and chemosis.
|(Figure 2) CT scan shows bilateral lid swelling, increased radiodensity of the globes, fat stranding in the retrobulbar space and circumferential soft-tissue swelling around the globe, all consistent with prominent intraocular and orbital inflammation. |
We Make a Diagnosis
Given the apparent lack of systemic illness, a workup for an underlying source of infection was initiated. Blood and urine cultures taken in the ER were negative for bacterial organisms. A CT scan of the abdomen and pelvis showed multiple hypodense lesions in the liver that were consistent with small abscesses (Fig. 3). Gastrointestinal consultation was obtained and the patient underwent endoscopic retrograde cholangiopancreatography, which revealed sludge in the common bile duct. Under ultrasound guidance, one of the hepatic lesions was drained and the fluid was sent to the laboratory for culturing. After three days, cultures from both the vitreous aspirates in each eye (Fig. 4) and the liver abscess grew out Klebsiella pneumoniae. The organisms were resistant to ampicillin, but otherwise sensitive to all antibiotics tested.
Over the next few days, the patient’s eyes became less painful and swollen. She began to develop corneal thinning in each eye, however, and her vision remained no light perception bilaterally. Follow-up vitreous aspirations followed by injection of ceftazidime (2.25 mg/0.1 ml) and dexamethasone (400 µg/0.1 ml) were performed on her fourth day in hospital. The cultures demonstrated persistent K. pneumoniae in the right eye and no organisms in the left. In spite of continued aggressive antibiotic and corticosteroid therapy, her right cornea progressively thinned and ultimately perforated. After much debate, Ms. Chung’s family refused any further care and the patient was discharged to home care.
We began a work-up for an underlying source of infection. (Figure 3) Enhanced CT scan through the liver shows one of many hepatic abscesses (arrow) evident on serial sections.
|(Figure 4) Gram stain of organisms cultured from a vitreous aspirate taken at presentation reveals multiple Gram-negative rod-shaped bacteria. |
Endogenous endophthalmitis is an uncommon but well-described complication of Klebsiella sepsis, which has been noted to occur more frequently in individuals with diabetes mellitus and hepatic disease.1,2 Most reports of endogenous Klebsiella endophthalmitis have involved East Asians, as with our patient.3 Bilateral endogenous endophthalmitis has been described as well, but is quite rare.1-3 Prognosis for visual outcome in patients with endogenous Klebsiella endophthalmitis is generally quite poor.4 A recent case series from Taiwan reviewed the records of 75 patients with endogenous endophthalmitis treated over a 10-year period. Of these patients, 61 percent had Klebsiella as the causative organism and liver abscesses were the source of infection in 87 percent of patients.
Most infections were quite aggressive, with only 12.5 percent of eyes having a final visual acuity of 20/400 or better.5 While the reason(s) for the preponderance of endogenous Klebsiella endophthalmitis in East Asian patients is probably multifactorial, it is worth noting that East Asians tend to have a higher incidence of cholangiohepatitis, as compared with Caucasians,6 and Klebsiella is one of the more frequent causes of liver abscess and bacteremia in the setting of hepatic infection.7
When a patient presents with endophthalmitis in the absence of recent ocular trauma or surgery, clinicians should search for an endogenous source. For East Asian patients, clinicians should be particularly aware of the increased prevalence of Klebsiella as the infective organism, and should strongly consider a hepatic source. Given the poor visual prognosis in these patients, early diagnosis with prompt, aggressive treatment is critical for minimizing the risk of vision loss.
* Patient name is fictitious.
Drs. Zelefsky and Teng are ophthalmology residents at New York University School of Medicine. Dr. Cunningham is currently director of the uveitis service at California Pacific Medical Center and an adjunct clinical professor of ophthalmology at Stanford University.
The authors thank Chee Soon Phaik, MD, for thoughtfully reviewing an early draft of this article.
This submission was supported in part by The San Francisco Retinal Foundation and The Pacific Vision Foundation.
1 Margo, C. E. et al. Ophthalmology 1994;101:1298–1301.
2 Ang, L. P. et al. Eye 2000;14:855–860.
3 Wong, J. S. et al. Ophthalmology 2000;107:1483–1491.
4 Scott, I. U. et al. Am J Ophthalmol 2004;138:662–663.
5 Chen, Y. J. et al. Retina 2004;24:383–390.
6 Nakayama, F. et al. Dig Dis Sci 1986;31:21–26.
7 Frey, C. F. et al. Medicine (Baltimore) 1996;75:99–113.
NEXT MONTH: The case of “sinusitis” that antibiotics couldn’t treat.