The HIV/AIDS pandemic continues to present a major health challenge for sub-Saharan Africa, with an estimated 26.6 million people in this region infected. In Uganda, the prevalence of HIV infection is 6.7%.1
Before highly active antiretroviral therapy (HAART) was widely available in North America and Europe, 50%-75% of the HIV-infected individuals were estimated to develop nonrefractive visual problems at some point during the course of their illness and cytomegalovirus (CMV) retinitis was the leading cause of vision loss.2,3 The incidence of both ocular complications like CMV retinitis and of visual loss has dropped dramatically in these regions since HAART was made available.4
In Uganda, HAART was introduced on a large scale in the year 2004, and currently among the estimated 1 million persons living with HIV, approximately 120,000 are on HAART.5 As HAART becomes more available in this region, we expect the prevalence of HIV-related ocular complications to reduce as seen in the developed world.
In sub-Saharan Africa, the prevalence of ocular disease in HIV-positive individuals has been estimated as between 30% and 45%.6,7 The pattern of ocular diseases affecting patients with HIV infection in Africa, however, is somewhat different from that in North America and Europe in the pre-HAART era, in that ocular tumors and opportunistic infections due to organisms like Mycobacteriumtuberculosis, Cryptococcus neoformans, and Toxoplasma gondii are more frequent causes of ocular morbidity and visual loss.7 However, the most frequent treatable causes of visual loss also are related to opportunistic retinal infections with herpes viruses, predominantly CMV, and less frequently herpes simplex virus and varicella zoster virus.4,8
Given that two-thirds of HIV-infected individuals reside in sub-Saharan Africa, the literature on the prevalence or incidence of visual loss due to HIV-related eye diseases is quite limited. In this study, we report the prevalence of reduced visual acuity and the ocular diseases associated with such visual impairment among HIV-infected individuals in Uganda presenting for HIV care.
MATERIALS AND METHODS
Consecutive patients of 18 years or older with confirmed HIV infection were recruited from a single adult outpatient infectious disease clinic between July 1, 2003 and August 31, 2004. The infectious disease clinic is an HIV research, training, and treatment site at Mulago Hospital in Uganda. Mulago Hospital is Uganda's national referral hospital and a university teaching hospital. Patients were excluded if they presented with a condition expected to impair visual acuity testing, including (1) mental confusion, a mental handicap, or altered mental state or (2) use of anticholinergic drugs at the time of assessment for visual impairment.
This study was approved by the Makerere University Research Ethics Committee and the National Council of Science and Technology in Uganda.
A trained study nurse screened participants for visual impairment by testing visual acuity using an illuminated Snellen chart at 6 meters. Patients wearing refractive distance correction at the time of screening had their vision tested with the available correction. A Landolt C chart was substituted for those who were illiterate. For those whose visual acuity was poor and could not read the chart further assessment was performed in the following order: (1) ability to count fingers, (2) ability to detect hand movement, and (3) ability to perceive light. Each eye was tested separately.
Definition of Loss of Visual Acuity (Visual Impairment)
All patients with a visual acuity of 6/9 or less in 1 or both eyes had a detailed medical, ophthalmological, and laboratory diagnostic evaluation to determine the cause and clinical symptoms associated with the impaired vision.
History Related to Visual Impairment
A history including the time of onset, duration, and awareness of visual impairment was documented.
Patients were queried, and records reviewed for evidence of past or current systemic illnesses and prior use of drugs that potentially could cause visual impairment. These included conditions such as diabetes mellitus, tuberculosis (TB), and meningitis and drugs such as ethambutol, antimalarials, and many others. The patient's latest CD4+ T-cell count, the World Health Organization clinical stage of HIV,9 and antiretroviral treatment (ART) history also were recorded.
A complete ocular examination was performed by the study ophthalmologist (J.O-S.). This included confirmation of the patient's visual acuity, confrontation visual field testing, objective (streak retinoscopy) and manifest refraction. Eye movements were tested in all directions of gaze and pupillary function was assessed. Slit-lamp biomicroscopy was used to evaluate the structures in the anterior and posterior segment of the eye. All patients were dilated and had a detailed fundus assessment by indirect binocular ophthalmoscopy with a 20-diopter lens. Slit-lamp biomicroscopy of the optic nerve head and macula was performed using a 90-diopter lens. Photography documenting clinical findings, visual field plotting, radiological imaging studies, and laboratory investigations were obtained when clinically indicated.
Diagnosis for the cause of visual impairment for each patient in this study was made using a uniform procedure, described in the online table. If more than 1 disease was present, the disease most likely to have significant effect on vision was considered as the cause of visual impairment. A second ophthalmologist independently reviewed fundus findings to confirm each diagnosis. Definite diagnosis could not be made in a small number of cases because of the limited diagnostic resources.
Data was entered in the SPSS version 10 and analyzed using STATA statistical software version 7.0. Categorical variables were compared using the χ2 test or Fisher exact test when expected cell values were less than 5. Continuous variables were compared using Student t test. All reported P values are 2 sided and were considered statistically significant when the type 1 error probability was <0.05. Logistic regression was used to evaluate potential relationships between outcomes and possibly explanatory covariates.
Characteristics of the Study Population
During the 14-month study period, 1225 patients presented for HIV care; of them, 1212 were enrolled in the study. None of the patients were using anticholinergic drugs at the time of assessment for visual impairment. Thirteen patients were excluded because they had an altered mental state and could not be screened for visual loss.
Among the 1212 patients, 29.5% (357) were males and 70.5% (855) were females (Table 1). The mean age in years of the participants was 36 years (SD = 8.2 years, range 18-79 years). The median absolute CD4+ T-cell count was 171 cells per cubic millimeter. According to the World Health Organization clinical staging, 125 (10.3%) were in stage I, 363 (30%) in stage II, 531 (43.6%) in stage III, and 193 (15.9%) in stage IV. Four hundred fifty-three (37.4%) were on ART at the time of screening.
Characteristics of Patients With Visual Impairment
The prevalence of presenting visual acuity of 6/9 or worse in 1 or both eyes was 11.2% [95% confidence interval (CI): 9.50%-13.1%]. Patients with visual impairment were a median of 2 years older than those without visual impairment (P < 0.001), and the risk of a visual acuity of 6/9 or worse was approximately 23% greater for every additional 5 years of age [odds ratio (OR) = 1.23, 95% CI: 1.18-1.29]. Males were more likely than females to have this level of visual impairment (OR = 1.51; 95% CI: 1.04-2.18; P = 0.029).
The mean CD4+ T-cell count was higher in patients without visual impairment (171 cells/mm3) than in patients with visual impairment (118 cells/mm3) (P = 0.015). Individuals with a higher CD4+ T-cell count were therefore less likely to present with a visual acuity of 6/9 or worse and had a approximately 5% less risk of visual impairment for every upward increment of 50 CD4+ T cells per cubic millimeter (OR = 0.95; 95% CI: 0.948-0.953). Patients with visual loss were also less likely to be taking ART (OR = 0.55; 95% CI: 0.38-0.79) and tended to have a more advanced stage of HIV disease as shown in Table 1.
Causes of Visual Loss
Table 2 summarizes the frequency of ocular diseases and the corresponding CD4+ T-cell count distribution among the 136 patients with visual loss. The causes of uveitis included 13 patients with CMV retinitis; 15 with ocular toxoplasmosis; and 20 with other forms of uveitis due to herpes zoster ophthalmicus (7), syphilis (6), TB (1), and idiopathic uveitis (6). “Optic nerve disease” includes 6 patients with ethambutol optic nerve toxicity, 2 patients with optic atrophy secondary to bacterial meningitis, 2 patients had idiopathic optic neuritis, and 10 with optic atrophy secondary to cryptococcal meningitis.
Patients with CMV retinitis and optic nerve diseases (many related to cryptococcosis) were found to have significantly lower CD4+ T-cell counts than patients with the other diagnoses. No CMV retinitis was seen in patients with a CD4+ T-cell count above 50 cells per cubic millimeter. Four patients (30%) with CMV retinitis had both eyes affected at the time of presentation. Females were found to be more likely to have to CMV retinitis than males (P = 0.034). Visual loss caused by CMV infection was primarily due to necrotizing retinitis, with only 3 patients having visual loss due to presumed immune recovery uveitis. Among the 15 patients found with ocular toxoplasmosis, 7 had lesions suggestive of a primary infection, whereas 8 had an ophthalmoscopic appearance consistent with reactivation of a prior infection. Ocular syphilis presented as an anterior nongranulomatous uveitis in 4 patients, as a posterior uveitis in 1 patient, and as a panuveitis in another patient. Three patients presented with an anterior uveitis of unknown origin associated with initiation of HAART with no evidence of CMV retinitis lesions.
Sixteen patients had cataracts as the cause of their visual loss. Four of these patients had evidence of prior intraocular inflammation (complicated cataracts), which may or may not have been related to other opportunistic ocular complications, for example, CMV retinitis with immune recovery uveitis. Aqueous analysis was not performed in these cases due to limited resources. Surgery was performed on 10 patients (63%), and only 1 patient was found to have retinal detachment; the rest had a normal fundus.
Figure 1 depicts the age distribution among patients diagnosed with an HIV-related ocular disease. Patients with visual loss related to retinitis and other forms of uveitis tended to be younger (65% less than 35 years of age), whereas patients with visual loss related to cataracts and optic nerve diseases tended to be older (only 20% and 40%, respectively, less than 35 years of age).
Severity of Visual Impairment
The frequency distribution of patients according to degree of reduced visual acuity, affected eye(s), and ocular diagnosis is given as Table 3. Sixty-six patients (49%) had bilateral disease. The ocular diseases associated with legal blindness (visual acuity of 6/60 or worse) were predominantly related to opportunistic infections and included optic nerve diseases, CMV retinitis, toxoplasmic chorioretinitis, and cataracts.
Visual impairment was not a presenting complaint that patients mentioned during the screening process in most instances. Thirty-five patients (26%) were aware that they had reduced vision and presented the complaint to the attending HIV clinician. Fifty-five additional patients (40%) were aware that they had reduced vision but did not mention it to their physician. Forty-six patients (34%) were not aware that they had reduced vision and only realized this during vision testing of the study. There was no association between awareness of visual loss and bilaterality of the ocular disease [OR (bilateral) = 0.9; 95% CI: 0.4-1.8].
Our results suggest that visual loss among HIV-infected individuals in the sub-Saharan Africa is common, is often severe, and usually is not mentioned as a presenting complaint to HIV care providers. Furthermore, because CMV retinitis and several other opportunistic retinal infections are associated with high annual incidence in Sub-Saharan Africa but brief survival thereafter10 in a severely ill state that may preclude traveling to a clinic (unless successfully treated with HAART), it is likely that the lifetime burden of CMV retinitis is underestimated by our cross-sectional study of patients presenting for HIV care. Our study design also would have missed patients with eye disease that had not yet caused vision loss. Moreover, because patients with mental confusion or altered mental state were excluded, we may have missed patients with opportunistic infections involving the eye and brain. Thus, our results provide a lower bound on the true prevalence of ocular disease in an outpatient HIV care clinic.
In our study, some cases of visual loss resulted from an ocular opportunistic complication of immunodeficiency. A low CD4+ T-cell count of less than 200 cells per cubic millimeter is a widely documented risk factor for developing complications of AIDS including those which cause visual loss.11 In our study cohort, the majority had CD4+ T-cell count less than 200 cells per cubic millimeter, which may explain the high prevalence of visual loss of 11% found. Patients who had severe advanced immunodeficiency or were not taking ART were more likely to present with reduced vision. All of these findings suggest that successful ART is likely to reduce the risk of visual acuity loss. However, because patients often present for care with advanced immunodeficiency and because the coverage of HAART is likely to be incomplete for many years to come, our results suggest that HIV care systems will need to cope with a large number of ocular opportunistic diseases, which require specific anti-infectious treatment over and above ART.
Opportunistic ocular diseases result in visual loss due to direct or indirect damage to the ocular structures or visual pathways. The location of the ocular lesion(s) plays an important role in the predisposition of a patient to developing visual loss. The natural history of diseases such as CMV retinitis and toxoplasmic retinitis in the context of severe immunodeficiency is of relentless progression of the disease eventually leading to irreversible blindness, unless specific treatment or recovery of immunity controls the disease. In resource-limited settings like ours, ocular examinations are not routinely performed in HIV treatment sites, thus predisposing people with ocular opportunistic infections to be missed in early stages with detection of disease in advanced stages after extensive damage to the ocular structures. It is also known that these patients are very sick with other systemic opportunistic infections and therefore may tend to omit reporting of eye symptoms at an early stage or may not present for care at all. Systematic efforts to detect such disease, perhaps involving a screening symptom questionnaire administered by medical assistants, would be needed to prevent presentation after irreversible vision loss that already has occurred.
The opportunistic infections most commonly causing severe visual loss in our patients were CMV retinitis, optic atrophy secondary to cryptococcal meningitis, ethambutol optic nerve toxicity, and ocular toxoplasmosis. These infections are important because of their potential to affect both eyes causing blindness as we demonstrate in our study.
In this study, 1% (n = 13) had visual loss caused by CMV retinitis. CMV retinitis has been the leading cause of visual loss in North America, both before2 and after12 the availability of HAART. The prevalence of CMV retinitis in Africa is reported to be lower ranging from 0% to 8.5%.13 CD4+ T-cell counts less than 50 cells per cubic millimeter and female sex are widely documented risk factors for CMV retinitis worldwide as seen in this study.14 In the HAART era, an additional risk factor for developing visual loss with CMV retinitis is HAART-induced immune recovery uveitis.15,16 A higher prevalence of CMV retinitis would be expected in this cohort if dilated fundoscopy had been performed on all patients rather than just on those with manifest visual loss and if the severely ill patients with altered mental status had been evaluated ophthalmoscopically.
Optic Nerve Disease
The most frequent causes of visual loss among study patients with optic nerve disease were cryptococcal meningitis (0.8% of all patients) and ethambutol optic nerve toxicity (0.5% of all patients). Vision was very poor, typically in both eyes, because of late presentation and advanced nerve damage (Table 3).
In the developed world, visual loss due to cryptococcal meningitis was reported to occur in 1.1%-3.3% of patients with AIDS in the pre-HAART era.17 Visual loss in our study was due to papilledema, optic neuropathy, and optic atrophy; findings that are similar to what has been reported in other studies.18,19 Cryptococcal meningitis has been suggested to occur more frequently in African patients with AIDS20; our study likely underestimates its prevalence because patients with altered mental state, a common complication this disease, were excluded in our study.
Ethambutol toxicity has been documented as an important cause of visual loss among patients on TB therapy.21 To our knowledge, the prevalence of visual loss due to ethambutol toxicity among the African HIV-infected population has not been reported previously. This is a cause of blindness that is preventable if patients on this drug have visual acuity and color vision monitoring performed while on treatment. Although ethambutol toxicity is a more commonly reported cause of visual loss in AIDS patients with TB, ocular infection TB seems to be relatively a rare complication,22 seen in only 1 patient in our study.
Before the access to ART, visual loss due to cryptococcal meningitis or TB in Africa was considered a rare complication in patients with AIDS because both conditions generally rapidly lead to the death of the patient. The relatively high number of visual loss caused by cryptococcal meningitis and TB treatment observed in this study may be related to the improved survival of patients after these opportunistic infections today.
Ocular toxoplasmosis accounted for 11% of patients with visual loss in this study. Their median CD4+ T-cell count was 116 cells per cubic millimeter, slightly higher than what has been documented in other studies. Ocular toxoplasmosis is one of the most frequent causes of uveitis in Africa; however, its prevalence among HIV-infected individuals is not known.23 In North America, it accounts for 1%-3% of ocular diseases in patients with HIV infection.24,25 It is more frequent in Latin America where the prevalence is 8%-13% in the HIV population,26 perhaps related to dietary differences. Ocular toxoplasmosis also contributed significantly to blindness (visual acuity worse than 6/60) in this study because of late presentation and macular involvement. As for CMV retinitis, the prevalence is likely underestimated, lower than it would have been had all patients been studied with ophthalmoscopy.
“Other” Causes of Uveitis
The other causes of uveitis found in our study included 7 patients with herpes zoster ophthalmicus (a finding similar to other reports),27,28 6 patients with syphilitic uveitis, and 6 patients with idiopathic uveitis. For those with idiopathic uveitis, we were limited in performing all necessary tests to fully categorize them as idiopathic like polymerase chain reaction or human leukocyte antigen (HLA) studies. The numbers of patients with ocular syphilis are relatively high compared with reports in other centers.29 Developing countries however report a relatively high prevalence of syphilis among pregnant women,30 and the prevalence of syphilis may also be high among HIV-infected populations like ours where HIV is frequently a sexually transmitted infection. The numbers of cases of ocular syphilis have been reported to be increasing in some studies in the recent past.31,32
In this study, 16 patients had cataracts as the cause of their visual loss. Although these patients were older, on average, than the other patients studied (Fig. 1), they were relatively young compared with the average age of a patient with cataract.
Cataracts have not been documented as a frequent cause of visual loss in other African studies of ocular complications of HIV infection, although it has been reported to be an important cause of visual loss in HIV-infected patients in the HAART era in North America, often related to complications of CMV retinitis.12 They have been reported as a complication of ocular inflammation in immune recovery uveitis or retinal detachment due to CMV retinitis and other necrotizing retinitis like progressive outer retinal necrosis33 or acute retinal necrosis,34 and also in association with long-term use of nucleoside analogue reverse transcriptase inhibitors.35 Surgery was performed on 10 patients (63%), and only 1 patient was found to have retinal detachment; the rest had a normal fundus. We however cannot rule out the role of other retinal diseases like CMV retinitis in all the cataracts because surgery was not performed on all patients. The HIV virus itself may cause cataract formation.36 However, other factors in addition to age such as poor nutrition, exposure to ultraviolet light,37 or other factors also could have contributed to the development of cataract in our patients.
Causes Not Related to Opportunistic Infections and Their Treatment
Our study was limited in identifying causes of visual loss in this category because fluorescein angiography and macular function tests were not performed due to limited resources. HIV-infected patients can develop visual loss due to ischemic maculopathies, which can only be identified using these techniques.38
Refractive errors were the most frequent cause of visual loss not related to opportunistic infections and overall in this study. This is an expected result for an adult population of the similar age group screened in a similar manner for visual loss.39 HIV-infected individuals have however been reported to be predisposed to anomalies of accommodation and early presbyopia.40 In this study, most patients were myopic with a few abnormalities of accommodation.
These observations suggest that concerns previously raised by Kestelyn and Cunningham,8 that hundreds of thousands of patients with HIV infection may suffer vision loss during their lifetime, may be warranted. Presently, the resources allocated to eye care for patients at HIV treatment sites in sub-Saharan Africa and other developing regions are minimal, sometimes absent. Inaccessibility of specialist eye care, lack of routine eye assessment, and lack of awareness about eye disease at the HIV treatment sites form a vicious cycle that likely contributes to preventable blindness among large numbers of patients.
Although full deployment of HAART should remain the priority, to maintain credibility with communities, the HIV treatment programs will need effective measures to deal with a problem of this magnitude and also to deal with the immune recovery inflammatory syndromes commonly occurring in the eye as a side effect of ART.41 It is likely that the majority of patients attending HIV treatment clinics in Africa, like our study site, are presenting at later stages in the clinical course of their HIV infection and are therefore at risk of developing ocular diseases. HIV treatment programs should seek to incorporate methods of ascertaining potentially blinding eye disease at an early treatable point, given that an abundance of studies demonstrate high risk of preventable visual loss in HIV-infected patients. Most treatable visual problems in HIV patients occur with advanced immunodeficiency associated with a CD4+ T-cell count below 200 cells per cubic millimeter. Although identification of patients with eye disorders at a treatable stage is logistically challenging, some simple steps could be implemented with little difficulty. Identification of patients with ocular disorders among those presenting to HIV clinics would be enhanced by evaluation of visual acuity and questioning of patients regarding visual symptoms (floating spots, flashes, and visual loss), which could be performed by medical assistants. Ophthalmic evaluation by binocular indirect ophthalmoscopy of inpatients and hospice patients with advanced AIDS can be done efficiently and also is likely to yield a large number of potentially treatable cases that would result in visual loss. Research on cost-effective methods of identifying patients who are not under care also is needed.
The problem of HIV-related blindness is extensive and deserves more attention as HIV care scales up worldwide.
1. Wolfgang H, Misinguzi J, Kirungi W, et al. The estimated burden of HIV
/AIDS in Uganda
, 2005-2010. AIDS
2. Jabs DA. Ocular manifestations of HIV
infection. Trans Am Opthalmol Soc
3. Hoolbrook JT, Jabs DA, Weinberg DV, et al. Visual loss
in patients with cytomegalovirus retinitis and acquired immunodeficiency syndrome before the widespread availability of highly active antiretroviral therapy. Arch Ophthalmol
4. Ng WT, Franzco PV. Ocular association of HIV
infection in the era of highly active antiretroviral therapy and the global perspective. Clin Experiment Ophthalmol
5. Musinguzi B. Uganda
edges closer to AIDS treatment for all. Bull World Health Organ
6. Schuman JS, Orellana J, Friedman AH, et al. Acquired immunodeficiency syndrome (AIDS). Surv Ophthalmol
7. Lewallen S, Coutright P. HIV
and AIDS and the eye in developing countries. Arch Ophthalmol
8. Kestelyn PG, Cunningham ET. HIV
/AIDS and blindness. Public Health reviews. Bull World Health Organ
9. World Health Organization, UNAIDS and UNICEF. Towards Universal Access: Scaling Up Priority HIV/AIDS Interventions in the Health Service
. Progress Report, April 2007. Geneva, Switzerland.
10. Balo KP, Amoussou YP, Bechetoille A, et al. [Cytomegalovirus retinitis and ocular complications in AIDS patients in Togo]. J Fr Ophthalmol
11. Cunningham ET Jr, Belfort R Jr. HIV/AIDS and the Eye: A Global Perspective
. San Francisco, CA: American Academy of Ophthalmology; 2002.
12. Jennifer ET, Douglas J, John HK, et al. Causes of visual acuity loss among patients with AIDS and cytomegalovirus retinitis in the Era of Highly active antiretroviral therapy. Ophthalmology
13. Kestelyn P. The epidemiology of CMV retinitis in Africa. Ocul Immunol Inflamm
14. Whitecup SM. Cytomegalovirus retinitis in the era of highly active antiretroviral therapy. JAMA
15. Kempen JH, Min YI, Freeman WR, et al. Risk of immune recovery uveitis in patients with AIDS and cytomegalovirus retinitis. Ophthalmology
16. Shelburne SA, Visnegarwala F, Darcourt J, et al. Incidence and risk factors for immune reconstitution inflammatory syndrome during highly active antiretroviral therapy. AIDS
17. Rex JR, Larsen RA, Dismukes WE, et al. Catastrophic visual loss
due to Cryptococcus neoformans
18. Moraes HV. Ocular manifestations of HIV
/AIDS. Curr Opin Ophthalmol
19. Goldberg DE, Smithen LM, Angelilli A, et al. HIV
associated retinopathy in the HAART era. Retina
20. Kestelyn P, Taelman H, Bogaerts J, et al. Ophthalmic manifestations of infections with Cryptococcus neoformans
in patients with the acquired immunodeficiency syndrome. Am J Ophthalmol
21. Goyal JL, De Sarmi, Singh NP, et al. Evaluation of visual functions in patients on ethambutol therapy for tuberculosis: a prospective study. J Commun Dis
22. Kestelyn P, Lewallen S. Ocular problems with HIV
infection and AIDS in Africa. Community Eye Health
23. Ronday MJ, Stilma JS, Barbe RF, et al. Aetiology of uveitis in Sierra Leone, west Africa. Br J Ophthalmol
24. Jabs DA. Ocular manifestations of HIV
infection. Trans Am Ophthalmol Soc
25. Rogers CA, Harris JR. Ocular toxoplasmosis in HIV
infection. Int J STD AIDS
26. Matos KT, Santos MC, Muccioli C. Ocular manifestations in HIV
infected patients attending the department of ophthalmology of Universidade Federal de Sao Paulo. Rev Assoc Med Bras
27. Umeh RE. Herpes zoster ophthalmicus and HIV
infection in Nigeria. Int J STD AIDS
28. Liesegang TJ. Herpes zoster viral infection. Curr Opin Ophthalmol
29. Tran TH, Cassoux N, Bodaghi B, et al. Syphilitic uveitis in patients infected with human immunodeficiency virus. Graefes Arch Clin Exp Ophthalmol
30. Swai RO, Somi GGR, Matee MI, et al. Surveillance of HIV
and syphilis infections among antenatal clinic attendees in Tanzania- 2003/2004. BMC Public Health
31. Dorris JP, Saha K, Jones NP, et al. Ocular syphilis: the new epidemic. Eye
32. Chao JR, Khurana RN, Fawzi AA. Syphilis: reemergence of an old adversary. Ophthalmology
33. Engstrom RE Jr, Holland GN, Margolis TP, et al. The progressive outer retinal necrosis syndrome. A variant of necrotizing herpetic retinopathy in patients with AIDS. Ophthalmology
34. Blumenkranz MS, Culbertson WW, Clarkson JG, et al. Treatment of the acute retinal necrosis syndrome with intravenous acyclovir. Ophthalmology
35. Gaynes BI. “Star” suture opacities of the crystalline lens: an illustrative report in AIDS patients. Med Sci Monit
36. Iwakura Y, Shioda T, Tosu M, et al. The induction of cataracts by HIV
-1 in transgenic mice. AIDS
37. Taylor HR, West SK, Rosenthal FS, et al. Effect of ultraviolet radiation on cataract formation. N Engl J Med
38. Cunninghan ET, Levinson RD, Jampol LM, et al. Ischemic maculopathy in patients with acquired immunodeficiency syndrome. Am J Ophthalmol
39. Mbulaiteye SM, Reeves BC, Mulwanyi F, et al. Incidence of visual loss
in rural south west Uganda
. Br J Ophthalmol
40. Westcott MC, Ward M, Mitchell SM. Failure of accommodation in patients with HIV
41. Otiti-Sengeri J, Meenken C, Van den Horn GJ, et al. Ocular immune reconstitution inflammatory syndromes. Curr Opin HIV AIDS