Article Text


Viridans group Streptococcus subretinal abscess
  1. Department of Ophthalmology, San Francisco General Hospital and the University of California, San Francisco, USA
  1. Dr R Y Kim, UCSF Department of Ophthalmology K301, 10 Kirkham Street, San Francisco, CA 94143–0730, USA.

Statistics from

Editor,—Subretinal abscess is an unusual entity, occurring primarily in immunocompromised individuals. Reported causative organisms include fungi1 2 and Gram negative rods.3 4 Here we report a subretinal infection resulting from the Gram positive organism viridans groupStreptococcus in a pancytopenic patient.


A 56 year old man undergoing chemotherapy for IgG multiple myeloma presented with 3 days of fever. He denied injecting drug abuse. He had a temperature of 101.5°F, poor dentition, a quiet indwelling catheter site on the chest, and no cardiac murmur. He was pancytopenic with a haematocrit of 14.2% (normal 39–49%), white blood cell count of 0.2 × 109/l (normal 4.5–11.0), absolute neutrophil count of 0.1 × 109/l (normal 1.8–6.8), and platelet count of 12 × 109/l (normal 150–450). Chestx ray and urinalysis were clear. Empirical treatment with intravenous vancomycin and ceftazidime was initiated.

The next day the patient reported decreased vision in the left eye. Visual acuity in the right eye was 20/25 and light perception in the left. The anterior segments were clear. Several cotton wool spots and blot intraretinal haemorrhages were present in the right eye. In the left eye, yellow white subretinal exudate was observed detaching two thirds of the retina; the vitreous was clear. Vitreous tap was unsuccessful. Since blood cultures had grown Gram positive cocci in pairs and chains that were subsequently identified as viridans groupStreptococcus sensitive to penicillin and vancomycin, the left eye was injected intravitreally with vancomycin 1.0 mg/0.1 ml. Despite two further intravitreal injections of vancomycin, the patient’s left eye rapidly lost light perception and developed increasing vitreous haze (Fig 1). Repeat vitreous tap revealed Gram positive cocci in pairs and chains, although cultures grew no organisms.

Figure 1

Fundus photograph on hospital day 4. Left eye exhibits subretinal exudative material detaching two thirds of the retina. The arrow indicates the optic disc.

The patient remained febrile and his mental status declined during treatment for presumed bacterial endocarditis. (He was too ill to undergo transoesophageal echocardiography.) Head computed tomogram revealed multiple brain lesions consistent with septic emboli. On day 10, the left eye was eviscerated and the indwelling catheter removed because of concern these foci might represent persistent infectious reservoirs. The patient died of sepsis on day 16. Family members declined necropsy.

Histopathological examination of the evisceration specimen revealed fragments of retina with an underlying protein exudate, coagulative necrosis of the inner retina, fibrinous occlusion of some retinal blood vessels, and numerous clumps and individual Gram positive cocci in the subretinal space and inner retina (Fig 2). Inflammatory cells were strikingly absent.

Figure 2

Histological sections of evisceration specimen. (A) Haematoxylin and eosin stain (×40) reveals coagulative necrosis of the inner retina. Arrow indicates outer nuclear layer. Asterisk indicates subretinal exudate with virtually no inflammatory cells. (B) Gram stain (×40) shows numerous Gram positive cocci in clusters and individually in the inner retina (arrow) and the subretinal space.


Viridans group Streptococcus (VGS) is a normal constituent of the oral flora. VGS is the leading cause of native valve endocarditis in non-injecting drug abusers and is also an increasing problem in neutropenic cancer patients.5 In adults undergoing chemotherapy for cancer up to 39% of bacteraemias result from VGS, with mortality rates of up to 30%.5 The most common source of VGS bacteraemia is the oropharynx.

Treatment guidelines for subretinal abscess are unclear. Although one of the most feared complications of vitrectomy surgery for endophthalmitis is retinal break leading to retinal detachment, successful vitrectomy with internal drainage has been described.3 4 We did not pursue vitrectomy surgery because a causative organism was identified, the vitreous was initially clear, and the patient was moribund. In retrospect, the absence of an inflammatory cell response in the retina (Fig 2), presumably because of severe pancytopenia, suggests sterilisation of the eye by any means would have been extremely difficult if not impossible.

The clinical picture of our patient was that of subretinal abscess. The absence of inflammatory cells in the pathological specimen, however, defines the subretinal exudate as non-purulent and primarily proteinaceous in nature.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.