To report the incidence and clinical outcomes of non-cytomegalovirus (non-CMV) ocular opportunistic infections in patients with acquired immunodeficiency syndrome (AIDS) in the era of highly active antiretroviral therapy.
Multicenter, prospective, observational study of patients with AIDS.
Medical history, ophthalmologic examination, and laboratory tests were performed at enrollment and every 6 months subsequently. Once an ocular opportunistic infection was diagnosed, patients were seen every 3 months for outcomes.
At enrollment, 37 non-CMV ocular opportunistic infections were diagnosed: 16 patients, herpetic retinitis; 11 patients, toxoplasmic retinitis; and 10 patients, choroiditis. During the follow-up period, the estimated incidences (and 95% confidence intervals [CI]) of these were: herpetic retinitis, 0.007/100 person-years (PY) (95% CI 0.0004, 0.039); toxoplasmic retinitis, 0.007/100 PY (95% CI 0.004, 0.039); and choroiditis, 0.014/ 100 PY (95% CI 0.0025, 0.050). The mortality rates appeared higher among those patients with newly diagnosed or incident herpetic retinitis and choroiditis (rates = 21.7 deaths/100 PY [ P = .02] and 12.8 deaths/100 PY [ P = .04]), respectively, than those for patients with AIDS without an ocular opportunistic infection (4.1 deaths/100 PY); toxoplasmic retinitis did not appear to be associated with greater mortality (6.4/100 PY, P = .47). Eyes with newly diagnosed herpetic retinitis appeared to have a poor visual prognosis, with high rates of visual impairment (37.9/100 PY) and blindness (17.5/100 PY), whereas those outcomes in eyes with choroiditis appeared to be lower (2.3/100 PY and 0/100 PY, respectively).
Although uncommon, non-CMV ocular opportunistic infections may be associated with high rates of visual loss and/or mortality.
Ocular opportunistic infections represent the most common ocular complications of acquired immunodeficiency syndrome (AIDS) and occur primarily among patients with severe immune compromise. As a consequence of highly active antiretroviral therapy (HAART) and immune recovery, the incidence of cytomegalovirus (CMV) retinitis has declined by over 80% over the last 15 years. Nevertheless, CMV retinitis remains the most common attributable cause of visual impairment and blindness among patients with AIDS. However, other non-CMV-related ocular opportunistic infections occur, including herpetic retinitis (attributable to either herpes simplex virus or varicella-zoster virus), toxoplasmic retinitis, and choroiditis (attributable to Pneumocystis carinii , Cryptococcus neoformans , Mycobacterium tuberculosis , and Mycobacterium avium complex infections). Case series describing these infections in the pre-HAART era have been published; however, the extent to which these infections occur among patients with AIDS and their effect on visual outcomes and association with mortality in the HAART era have not been described. The purpose of this study was to estimate the incidence of and to describe the clinical outcomes associated with these non-cytomegalovirus (non-CMV) ocular opportunistic infections in the HAART era, using data from the Longitudinal Study of Ocular Complications of AIDS (LSOCA) cohort.
The LSOCA is a multicenter, prospective, observational study of patients with AIDS. Patients aged 13 years and older with a diagnosis of AIDS according to the 1993 Centers for Disease and Prevention Revised Surveillance Case Definition were eligible to participate in the study. Patients with and without ocular opportunistic infections were recruited, and the cohort was enriched with patients with CMV retinitis in order to determine the outcomes of this ocular opportunistic infection. The characteristics of LSOCA patients with CMV retinitis have been described previously.
Data collection in LSOCA has been described previously. Briefly, medical and ophthalmologic histories and ophthalmic examinations were performed at enrollment and at in-person follow-up visits every 6 months following enrollment. Once an ocular opportunistic infection was diagnosed, in-person follow-up was increased to every 3 months. Herpetic retinitis, toxoplasmic retinitis, and choroiditis were diagnosed on the basis of the ophthalmic examination performed by experienced study-certified ophthalmologists at each clinical center. Herpetic retinitis was diagnosed using published criteria: by the presence of characteristic confluent, full-thickness, yellow necrotizing retinitis with little or minimal hemorrhage, with or without an overlying vitritis (acute retinal necrosis) or by a multifocal, rapidly progressive yellow necrotizing retinitis with early posterior pole involvement without vitritis (progressive outer retinal necrosis). Toxoplasmic retinitis was diagnosed on the basis of a focal, yellow-white necrotizing retinitis with fluffy borders and no or few scattered intraretinal hemorrhages, with or without adjacent chorioretinal scar. Choroiditis was diagnosed by the presence of acquired choroidal lesions without associated retinitis. When possible, the etiology of the choroiditis was inferred based on both the clinical appearance and the associated systemic opportunistic infections.
Best-corrected visual acuity was measured using logarithmic visual acuity charts according to the protocol developed for the Early Treatment of Diabetic Retinopathy Study (ETDRS). The visual acuity score is expressed as the number of letters read correctly. Quantitative Goldmann perimetry was performed as previously described. The Goldmann visual field (GVF) score was obtained by summing the degrees of field seen along each 30-degree meridian using the V-4 test object; the normal GVF score is ∼800 degrees.
Three subsets of patients with non-CMV ocular opportunistic infections were identified: those with longstanding ocular opportunistic infections at enrollment, those with newly diagnosed ocular opportunistic infections at enrollment, and those with newly diagnosed ocular opportunistic infections during follow-up (incident cases). Longstanding ocular opportunistic infections at enrollment were defined as those cases diagnosed more than 45 days before enrollment. Newly diagnosed at enrollment ocular opportunistic infections were defined as cases diagnosed within 45 days prior to enrollment. Incident ocular opportunistic infections were those cases not present at enrollment that developed during follow-up. The incident subset was used to estimate the incidence of ocular opportunistic infections. The newly diagnosed at enrollment and incident subsets were combined to evaluate the outcomes of newly diagnosed ocular opportunistic infections.
Clinically significant thresholds for loss of visual acuity and visual field score from GVF testing were chosen based on recommendations for reporting in this area or on previous publications from LSOCA. Incidence rates of developing non-CMV ocular opportunistic infections were calculated as the number of patients with the event divided by the person-years (PY) at risk. Mortality rates were calculated as the number of deaths per 100 PY of follow-up. Follow-up time was calculated as time from enrollment to event or end of follow-up, which occurred when a patient ended participation in LSOCA (eg, death, lost to follow-up, withdrawal from study) or the close of the database for this manuscript (December 31, 2010). Exact Poisson confidence intervals for rates were calculated using the Blaker method. Because of the small number of events, eye-level analyses do not adjust for eyes from the same individual. Data management and analyses were performed in SAS (SAS/STAT software, version 9.2 of the SAS System for Windows; SAS Institute Inc, Cary, North Carolina, USA) and R (R version 2.13.1; The R Foundation for Statistical Computing, Vienna, Austria).
Characteristics of the Study Population
As of December 31, 2010, 2362 patients (4722 eyes) were enrolled in the LSOCA cohort. Among these, 1835 patients (4011 eyes) had no evidence of ocular opportunistic infections at the time of enrollment ( Table 1 ). CMV retinitis was the most common ocular opportunistic infection and was present in 505 patients at enrollment; an additional 29 patients developed CMV retinitis during follow-up. Thirty-seven non-CMV ocular opportunistic infections were diagnosed at enrollment. Sixteen patients (26 eyes) were diagnosed with herpetic retinitis at enrollment. Of these 16 patients, 11 patients (18 eyes) had longstanding disease at enrollment, and 5 patients (8 eyes) were newly diagnosed at enrollment. One additional patient (2 eyes) developed herpetic retinitis during follow-up (incident case). Of these 17 patients, 10 patients had acute retinal necrosis and 7 patients had progressive outer retinal necrosis. At enrollment, 11 patients (12 eyes) were diagnosed with toxoplasmic retinitis. Of these 11, 8 patients (9 eyes) had longstanding disease and 3 patients (3 eyes) were newly diagnosed at enrollment. One additional patient (1 eye) developed toxoplasmic retinitis during follow-up (incident case). At enrollment, 10 patients (15 eyes) were diagnosed with choroiditis. Of these10, 6 patients (8 eyes) had longstanding disease and 4 patients (7 eyes) were newly diagnosed at enrollment. Two additional patients (3 eyes) developed choroiditis during follow-up (incident cases). Of the 12 patients with choroiditis at enrollment or during follow-up, 3 patients had pneumocystosis, 6 patients had cryptococcosis, 1 patient had a suspected fungal choroiditis, 1 patient had Mycobacterium avium complex choroiditis, and 1 patient had choroiditis of unknown etiology.
|Characteristic||Herpetic Retinitis||Toxoplasmic Retinitis||Choroiditis||No Enrollment OOI|
|Longstanding||Newly Diagnosed a||Incident a||Longstanding||Newly Diagnosed a||Incident a||Longstanding||Newly Diagnosed a||Incident a|
|Interquartile range||37, 46||33, 49||NA||37, 47||32, 43||NA||34, 42||33, 44||37, 47||38, 50|
|Sex [n (% ) men]||8 (73)||4 (80)||1 (100)||6 (75)||3 (100)||0 (0)||5 (83)||4 (100)||1 (50)||1477 (80)|
|Times since AIDS diagnosis (years)|
|Interquartile range||0, 2||0.3, 3||NA||-0.7, 5.1||0.1, 0.6||NA||1, 2||0.5, 3||5, 10||NA|
|Current HAART use [n (%)]||8 (73)||3 (60)||1 (100)||6 (75)||2 (67)||1 (100)||5 (83)||3 (75)||2 (100)||1553 (85)|
|Time from OOI diagnosis to enrollment (years)|
|Interquartile range||0.2, 3||0.08, 0.1||NA||1, 5||0, 0||NA||2, 4||0, 0.1||NA||NA|
|Nadir CD4+ T cells (cells/μL)|
|Interquartile range||3, 143||10, 170||NA||10.5, 77||2, 150||NA||6, 71||6, 43.5||18, 40||12, 106|
|Enrollment CD4+ T cells|
|Interquartile range||32, 174||28, 213||NA||52, 265||2, 189||NA||39, 354||13, 70||40, 179||80, 357|
|HIV load (log 10 [copies/mL])|
|Interquartile range||2.25, 3.90||2,23, 4.99||NA||2.89, 5.04||3.39, 5.00||NA||2.06, 2.60||2.06, 4.57||4.94, 5.19||1.88, 4.61|
|Bilateral disease [n (%)]||7 (64)||3 (60)||NA||1 (12)||0||NA||2 (33)||3 (75)||NA||NA|
|Percent retina involved|
|Interquartile range||15, 80||20, 80||NA||5, 10||5, 15||NA||NA||NA||NA||NA|
|Zone 1 involvement [n (%)]||11 (61)||2 (25)||NA||3 (33)||1 (33)||1 (100)||7 (87)||4 (57)||NA||NA|
|Visual acuity [n (%)]|
|20/50 or worse||15 (83)||4 (50)||0||4 (44)||1 (33)||1 (100)||2 (25)||0||0||118 (3)|
|20/200 or worse||4 (22)||1 (12.5)||0||3 (33)||1 (33)||0||0||0||0||56 (1)|
|Visual field score [n (%)]|
|600 or less (25 % loss)||15 (83)||8 (100)||0||4 (44)||2 (67)||1 (100)||3 (38)||1 (15)||0||342 (8)|
|400 or less (50 % loss)||14 (78)||6 (75)||0||3 (33)||1 (33)||0||3 (37.5)||0||0||80 (2)|
Patients with non-CMV ocular opportunistic infections had demographic characteristics similar to the entire LSOCA cohort: patients tended to be young (median age 42 years at enrollment), male, and white, non-Hispanic. Patients with newly diagnosed non-CMV ocular opportunistic infections at enrollment typically had low CD4+ T cells; medians were 89 cells/μL for herpetic retinitis, 33 cells/μL for toxoplasmic retinitis, and 43 cells/μL for choroiditis. For comparison, among those participants without ocular opportunistic infections, the median CD4+ T cell count at enrollment was 196 cells/μL. However, the range of CD4+ T cells was wide for herpetic and toxoplasmic retinitis, with 2 newly diagnosed patients with herpetic retinitis having a CD4+ T cell count >200 cells/μL and 1 patient with toxoplasmic retinitis having a CD4+ T cell count ≥150 cells/μL, a result consistent with the occurrence of these diseases in immunologically normal hosts. HIV loads typically were elevated, though not obviously different from those observed among those without ocular opportunistic infections.
Incidence of Non-CMV Ocular Opportunistic Infections
Among those participants without an ocular opportunistic infection at enrollment, herpetic retinitis, toxoplasmic retinitis, and choroiditis occurred in 1, 1, and 2 patients, respectively, for estimated incidence rates of 0.007/100 PY (95% CI 0.0004, 0.039), 0.007/100 PY (95% CI 0.0004, 0.039), and 0.014/100 PY (95% CI 0.0025, 0.050), respectively. Rates among participants with lower CD4+ T cells at presentation did not appear to be substantially greater. For those with a CD4+ T cell count <200 cells/μL, the incidence rates (and 95% CIs) for herpetic retinitis, toxoplasmic retinitis, and choroiditis were 0 (0, 0.04), 0.01/100 PY (0.0007, 0.08), and 0.03/100 PY (0.005, 0.09), respectively. For those with a CD4+ T cell count <100 cells/μL, the incidence rates (and 95% CIs) for herpetic retinitis, toxoplasmic retinitis, and choroiditis were 0 (0, 0.08), 0.02/100 PY (0.001, 0.14), and 0.02/100 PY (0.001, 0.14), respectively.
Table 2 presents mortality data for patients with non-CMV ocular opportunistic infections. Mortality rates for patients with longstanding ocular opportunistic infections were fairly similar among the different types of ocular opportunistic infections (range = 2.1 to 4.3 deaths/100 PY) and were similar to the mortality rate for patients in LSOCA without ocular opportunistic infections (4.1 deaths/100 PY). Mortality rates, however, appeared higher among patients with newly diagnosed ocular opportunistic infections, with the highest mortality observed among patients with newly diagnosed herpetic retinitis (rate = 21.7 deaths/100 PY, P = 0.02 vs patients without ocular opportunistic infections) followed by newly diagnosed choroiditis (12.8 deaths/100 PY, P = 0.04 vs patients without ocular opportunistic infections) and newly diagnosed toxoplasmic retinitis (6.4 deaths/100 PY, P = 0.47 vs patients without ocular opportunistic infections). For comparison, the mortality rates for those without ocular opportunistic infections were 6.0 deaths/100 PY for those with CD4+ T cells <200 cells/μL (95% CI 5.4, 6.7) and 7.8 deaths/100 PY for those with CD4+ T cells <100 cells/μL (95% CI 6.8, 8.9). Because of the potential survivor bias in the longstanding group, their mortality rates are difficult to interpret. However, the data from the recently diagnosed group suggest a higher mortality among patients recently diagnosed with these non-CMV ocular opportunistic infections.
|Group||n/N a||Mortality Rate (/100 PY) b||95% CI b||P Value c|
|Longstanding at enrollment||2/11||3.9||0.7, 13.8||>.99|
|Newly diagnosed d||3/6||21.7||5.9, 61.8||.02|
|Longstanding at enrollment||2/8||4.3||0.8, 15.3||>.99|
|Newly diagnosed d||1/4||6.4||0.3, 35.2||.47|
|Longstanding at enrollment||1/6||2.1||0.1, 11.6||.73|
|Newly diagnosed d||4/6||12.8||4.4, 32.2||.04|
|No longstanding or newly diagnosed ocular opportunistic infection at enrollment|
|Entire cohort without ocular opportunistic infections||465/1835||4.1||3.7, 4.5|
|CD4+ T cells <200 cells/μL at enrollment||328/922||6.0||5.4, 6.7|
|CD4+ T cells <100 cells/μL at enrollment||230/526||7.8||6.8, 8.9|
Visual Outcomes of Non-CMV Ocular Opportunistic Infections
Table 3 presents the visual outcomes among the patients with non-CMV ocular opportunistic infections. Because of the small numbers of patients at risk, the event rates have broad 95% CIs. Nevertheless, herpetic retinitis appeared to have a poor visual prognosis, with high rates of visual impairment (20/50 or worse) and blindness (20/200 or worse). All of the patients at risk suffered loss of at least 25% of the visual field and one-half lost 50% of the visual field on Goldmann perimetry. Conversely, loss of acuity and field were uncommon among patients with choroiditis, with only 1 patient suffering visual impairment and none blindness. Similarly, choroiditis was not associated with substantial loss of visual field on Goldmann perimetry. Toxoplasmic retinitis appeared to have an intermediate course; however, the number of events was low.
|Outcome||Herpetic Retinitis||Toxoplasmic retinitis||Choroiditis|
|Longstanding||Newly Diagnosed a||Longstanding||Newly Diagnosed a||Longstanding||Newly Diagnosed a|
|20/50 or worse|
|Rate (/100 EY)||13.5||37.9||0||0||14.5||2.3|
|95% confidence interval b||(0.7, 74.4)||(12.9, 95.3)||(–, 9.6)||(–, 37.5)||(0.7, 80.1)||(0.1, 12.9)|
|20/200 or worse|
|Rate (/100 EY)||16.3||17.5||0||11.5||3.8||0|
|95% confidence interval||(7.1, 35.4)||(3.1, 61.7)||(–, 7.6)||(0.6, 63.6)||(0.2, 20.9)||(–, 7.0)|
|Goldmann visual field score|
|<600 degrees (25% loss)|
|Rate (/100 EY)||68.7||30.4||13.6||63.0||33.3||0|
|95% confidence interval||(18.7, 196.1)||(5.4, 107.2)||(2.4, 47.9)||(3.2, 348.0)||(1.7, 183.8)||(–, 10.2)|
|<400 degrees (50% loss)|
|Rate (/100 EY)||12.2||17.0||0||11.6||0||0|
|95% confidence interval||(2.2, 43.0)||(3.0, 60.0)||(–, 13.0)||(0.6, 64.0)||(–, 23.4)||(–, 10.2)|