Endophthalmitis is a vision-threatening complication of cataract surgery and penetrating trauma (PT). Endophthalmitis vitrectomy study (EVS)1 reported a best corrected visual acuity (BCVA) of < 20/40 in 44% patients following successful treatment of post cataract surgery (PCS) endophthalmitis. In the EVS,1 vitreoretinal causes of decreased vision included pigmentary degeneration of the macula (33%), macular edema (32%), no apparent cause (26%), preretinal membrane (15%), presumed optic nerve damage (13%), retinal detachment (RD) (8%), macular ischemia (6%) and vitreous opacification (3%). Choroidal neovascularization,2 macular hole,3 epiretinal membrane (ERM)4 and traction and/or rhegmatogenous retinal detachment (RRD)5 have also been reported in endogenous candida endophthalmitis patients.
In the present series, we prospectively performed optical coherence tomography (OCT) in 20 eyes (20 patients) of endophthalmitis who obtained a BCVA of < 20/40 and clear media following surgery with successful anatomical outcome, to establish the treatable and nontreatable vitreoretinal causes of poor vision.
Materials and Methods
All patients presenting at our center from May 2003 to April 2004 with clinical features of endophthalmitis, following cataract surgery or PT, who had a BCVA of < 20/40 and clear media following surgery with successful anatomical outcome were included. Diagnosis of endophthalmitis was based on the following symptoms and signs: decreased vision, pain, hypopyon, fibrin in anterior chamber and vitreous inflammation leading to loss of red reflex.
Post cataract surgery endophthalmitis patients were managed according to EVS.6 The protocol was revised if smear or culture was positive for fungus. In these eyes immediate pars plana vitrectomy (PPV) with injection of amphotericin B (5 mg in 0.1 ml) with dexamethasone (400 mg in 0.1 ml) was performed.
Patients with PT endophthalmitis underwent primary repair and intravitreal antibiotics (IVAB) (vancomycin 1 mg in 0.1 ml and ceftazidime 2.25 mg in 0.1 ml) in emergency setting. These patients received intravenous antibiotics (ciprofloxacin 500 mg twice daily for five days). Nonresponding patients were subjected to PPV with IVAB guided by culture report if positive.
Postoperatively, at all visits, the ocular examination included BCVA, intraocular pressure, slit-lamp biomicroscopy and indirect ophthalmoscopy.
OCT scanning (using StratusTM OCT, Carl Zeiss, Dublin. CA) of the macula was done atleast three months postoperatively in the horizontal and vertical meridians. Fundus fluorescein angiography (FFA) was done to confirm the diagnosis, wherever required.
There were 106 eyes of as many patients (56 PCS endophthalmitis; 40 PT endophthalmitis) in the study period. Of these 50 eyes (40 PCS endophthalmitis and 10 PT endophthalmitis) achieved a BCVA ³ 20/40. Fifty-six eyes achieved a BCVA < 20/40. Of these, 20 eyes (12 PCS endophthalmitis; eight PT endophthalmitis) of as many patients (13 men, seven women; range 12-80 years; mean 42.65 years) fulfilled inclusion criteria. The rest of the 36 eyes were excluded from the study because of poor media clarity or retinal detachment. Two eyes of PCS endophthalmitis recovered with IVAB alone, while 18 eyes underwent PPV.
Presenting BCVA ranged from light perception (PL) to 20/100. Final BCVA ranged from hand movements close to face (HMCF) to 20/50 at three months. The BCVA of HMCF to £ 20/400 was seen in three eyes; > 20/400 to £ 20/200 in two eyes; > 20/200 to < 20/40 in 15 eyes.
At three months, on clinical examination alone, a BCVA of < 20/40 was due to "no apparent cause" in nine (45%) eyes, cystoid macular edema (CME) in five (25%) eyes, ERM in four (20%) eyes, pigmentary changes in the macula and atrophic macula in one (5%) eye each.
In clinically "no apparent cause" eyes, OCT findings revealed normal fovea in four eyes, neurosensory atrophy in three eyes, ERM and serous foveal detachment (SFD) in one eye each. [Figs. 1 and 2] The BCVA in these eyes ranged from 20/400 to 20/50. Three eyes with neurosensory atrophy had retinal thickness at the fovea of 85 m, 80 m and 60 m respectively. The FFA in these neurosensory atrophy eyes revealed normal findings in two, while one had macular ischemia. The eye with SFD showed mild delayed leakage in the fovea. In the remaining four clinically normal eyes (BCVA range 20/60 to 20/50), neither OCT nor FFA could detect the cause of poor vision.
Clinical diagnosis of CME in five eyes was confirmed on FFA. OCT in these eyes revealed increased retinal thickness with large full thickness cystoid spaces. Additionally, ERM was identified by OCT in three eyes along with CME.
Clinically seen ERM in four eyes was confirmed with OCT in all eyes and additionally detected CME in two eyes was seen on FFA as well.
Eyes with clinical pigmentary changes and atrophic macula showed decreased retinal thickness i.e. neurosensory atrophy on OCT. The FFA revealed normal findings in retina with pigmentary changes and macular ischemia in cases with atrophic macula.
EVS1 reported "no apparent cause" for decreased vision in 26% of cases. However, in our series nine of 20 (45%) eyes had no apparent cause of poor visual gain. OCT in these eyes revealed neurosensory atrophy in three eyes, ERM and SFD in one eye each. In the remaining four eyes, we failed to find any cause. We did not, however, do any electrophysiological studies7 to detect any physiologic dysfunction of the macula. The OCT was most useful in prognosticating the visual outcome of patients with neurosensory atrophy. Of the total five cases of neurosensory atrophy, clinical suspicion of atrophy was made in one eye each with pigmentary changes and macular ischemia on FFA respectively.
SFD was associated with low vision in one patient that was not apparent either clinically or angiographically. Presence of such SFD (though fluorescein angiography did not show any leakage into foveal detachments) has been demonstrated in retinal detachment eyes following buckle surgery, where authors89 have demonstrated slow gradual resorption of fluid. Over a follow-up of eight weeks in our patient, however, no resolution was seen.
The ERM was clinically apparent in four eyes, while OCT revealed ERM in an additional four eyes. We diagnosed CME in five eyes clinically and angiographically. OCT remains a useful noninvasive tool to monitor the progress of the CME.
In conclusion, OCT was helpful in detecting nontreatable causes like neurosensory atrophy in postendophthalmitis patients. These cases are likely to be missed on clinical examination and angiography alone.
1. Microbiologic factors and visual outcome in the endophthalmitis vitrectomy study. . The Endophthalmitis Vitrectomy Study Group Am J Ophthalmol. 1996;122:830–46
2. Jampol LM, Sung J, Walker JD, Folk JC, Townsend-Pico WA, Lowder CY, et al Choroidal neovascularisation secondary to Candida albicans chorioretinitis Am J Ophthalmol. 1996;121:643–9
3. Kusaka S, Hayashi N, Ohji M, Ikuno Y, Gomi F, Tano Y. Macular hole secondary to fungal endophthalmitis Arch Ophthalmol. 2003;121:732–3
4. McDonald HR, De Bustros S, Sipperly JO. Vitrectomy for epiretinal membrane with Candida chorioretinitis Ophthalmology. 1990;97:466–9
5. Pesin SR, Thomas MA, Smith ME. Combined rhegmatogenous -traction retinal detachment following successful treatment of Candida chorioretinitis Arch Ophthalmol. 1992;110:1051–2
6. Doft BH. The endophthalmitis vitrectomy study (editorial) Arch Ophthalmol. 1991;109:487–9
7. Rowsey JJ, Newsom DL, Sexton SJ, Harms WK. Endophthalmitis: Current approaches Ophthalmology. 1982;89:1055–66
8. Wolfensberger TJ. Foveal reattachment after macula-off retinal detachment occurs faster after vitrectomy than after buckle surgery Ophthalmology. 2004;111:1340–3
9. Wolfensberger TJ, Gonvers M. Optical coherence tomography
in the evaluation of incomplete visual acuity recovery after macula-off retinal detachments Graefes Arch Clin Exp Ophthalmol. 2002;240:85–9