Treatments for CMV retinitis include systemic ganciclovir therapy [4, 8-12], foscarnet [9, 10, 13], and the ganciclovir intraocular device [14]. Cidofovir, previously known as (S)-1-[3-hydroxy-2-(phosphonylmethoxy)propyl]cytosine (HPMPC), is a nucleotide analogue with activity against CMV in humans [15-21]. In contrast to nucleoside analogues, such as ganciclovir and acyclovir, cidofovir contains a phosphonate group that enables it to bypass the initial virus-dependent phosphorylation. Cellular enzymes convert cidofovir to the active diphosphate form, which has a long intracellular half-life and a prolonged effect against CMV [16, 17, 21]. The HPMPC Peripheral Cytomegalovirus Retinitis Trial (HPCRT) was a randomized, controlled clinical trial done to evaluate the efficacy and safety of intravenous cidofovir for the treatment of CMV retinitis in patients with AIDS. To minimize the risk for loss of visual acuity, only patients with small, peripheral CMV retinitis lesions were enrolled in the trial.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

Patients enrolled in the study had AIDS and untreated, "small, peripheral" CMV retinitis, which was defined as the presence of lesions involving less than 25% of the total retinal area and located at least 1500 µm from the margin of the optic disk and at least 3000 µm from the center of the fovea (that is, entirely in zones 2 and 3, in accordance with the previously published classification scheme [22]). Other inclusion criteria were a Karnofsky performance score of 60 or more [23], a serum creatinine level of 133 µmol/L (1.5 mg/dL) or less, proteinuria of less than 1+ on urinalysis, a total bilirubin level of 34 µmol/L (2.0 mg/dL) or less, hepatic aminotransferase levels less than five times the upper limit of normal, an absolute neutrophil count of 0.750 x 10⁹ cells/L (750 cells/micro L) or more, a platelet count of 50 x 10⁹ cells/L (50 000 cells/micro L) or more, and a hemoglobin concentration of 1.16 mmol/L (7.5 g/dL) or more. Patients who had retinal detachments, extraocular CMV disease, renal disease, cardiac disease, or allergy to probenecid or were receiving therapy with a nephrotoxic drug were ineligible. Diagnosis of AIDS was made on the basis of the Centers for Disease Control and Prevention 1993 revised surveillance case definition [24]. Cytomegalovirus retinitis was diagnosed if the characteristic white necrotizing retinitis was seen on examination by an ophthalmologist [22]. The treatment protocol and consent procedures were reviewed and approved by the institutional review boards at each participating institution, and all participants signed consent statements.

Randomization

Because cidofovir can cause nephrotoxicity, our trial was done in two stages. In stage 1, 29 patients were randomly assigned to receive either deferred therapy or immediate therapy with low-dose cidofovir. In stage 2, an additional 35 patients were randomly assigned on a 1:1:1 basis to receive deferred therapy, immediate therapy with low-dose cidofovir, or immediate therapy with high-dose cidofovir. Once the retinitis had progressed, as assessed by the clinician, patients were treated according to the clinicians' best medical judgment; when appropriate, patients assigned to receive deferred therapy were offered cidofovir therapy. Treatment assignments were generated by computer and arranged in random permuted blocks. Randomization was administered centrally after the completion of an eligibility checklist. Patients were counted as enrolled when their treatment assignments were revealed to clinic personnel.

Sample Size

The planned sample size was 90 patients: 35 in the deferral group, 35 in the low-dose cidofovir group, and 20 in the high-dose cidofovir group. With a two-sided type I error of 0.05, a power of 0.80, and an estimate of 10% of patients lost to follow-up, the minimal detectable differences in the median time to progression were a 100% increase for the low-dose cidofovir group compared with the deferral group and a 144% increase for the high-dose cidofovir group compared with the deferral group. The study was not designed to detect differences between the low-dose and high-dose cidofovir groups and had low power to compare them.

Treatment Administration

The administration of treatment was unmasked. Patients assigned to receive either low-dose or high-dose cidofovir were given identical induction therapy, an intravenous infusion of 5 mg/kg of body weight once weekly for 2 weeks. The low-dose cidofovir group then received maintenance therapy with cidofovir, 3 mg/kg once every 2 weeks; the high-dose cidofovir group received maintenance therapy with cidofovir, 5 mg/kg once every 2 weeks. To minimize nephrotoxicity [21], all cidofovir infusions were administered in conjunction with oral probenecid and intravenous saline hydration. A total of 4 g of probenecid was given orally with the infusion of cidofovir: Two grams were given 3 hours before the infusion, 1 g was given 2 hours after the infusion, and 1 g was given 8 hours after the infusion. Hydration consisted of 2 L of normal saline given intravenously: One liter was given immediately before the cidofovir infusion, and 1 L was started during or immediately after the cidofovir infusion. Because of a possible interaction between probenecid and zidovudine, which would cause an increased rate of cutaneous reactions [25], patients taking zidovudine did not take it on the days when probenecid and cidofovir were administered.

Cidofovir therapy was stopped for renal toxicity, which was defined as any one of the following: 1+ proteinuria on urinalysis with an increase in the serum creatinine level of 44 µmol/L (0.5 mg/dL) or more from baseline; persistent proteinuria of 2+ or more; serum creatinine levels of 177 µmol/L (2.0 mg/dL) or more; or a calculated creatinine clearance of 40 mL/min. If urinalysis done before cidofovir administration indicated proteinuria of 2+ or more, a second urine specimen was collected and analyzed after the patient had received 1 L of saline. If proteinuria of 2+ or more persisted, cidofovir therapy was discontinued. Cidofovir therapy was also discontinued in patients who developed intolerance to probenecid.

Data Collection

Data collection visits were scheduled at enrollment (baseline), 1 week after enrollment, then every 2 weeks through week 11, and then every 4 weeks after week 11. At each visit, a history was taken and an ophthalmologic examination and fundus photography were done. Cultures of blood and urine for CMV were done at baseline, week 3, and week 15; CD4⁺ T-cell counts were measured at baseline. Before each infusion of cidofovir, a complete blood count with differential and platelet count, serum chemistry panel, and urinalysis were done. Follow-up continued until death or the common study close-out date.

Visual Acuity

Best-corrected visual acuity was measured using logarithmic charts [22, 26]. Visual acuity was expressed as the number of letters read correctly (visual acuity score). A visual acuity score of 85 letters is equal to a Snellen acuity score of 20/20.

Fundus Photography

At each visit, retinal photography was done with a wide-angle camera according to a standardized protocol [10, 22, 27]. Eight fields of nonstereoscopic photographs were taken, as was a pair of stereoscopic photographs of the disc and macula; this technique documents all of the retina with the exception of the anterior periphery. Two sets of fundus photographs were taken. One set was forwarded to the fundus photograph reading center for the evaluation of retinitis progression, and the second set was used by the clinicians caring for the patient.

Assessment of Photographs

At the fundus photograph reading center, graders compared follow-up photographs with baseline photographs to determine whether progression had occurred [10, 22, 27, 28]. Graders at the reading center were masked to treatment assignment. The percentage of retinal area involved by CMV was determined at baseline and 5, 15, and 23 weeks after randomization [27].

Progression

The primary evaluation of retinitis progression, which was defined as 1) the movement of the border of a CMV lesion 750 µm or more along a front 750 µm or more in length or 2) the occurrence of a new lesion in either eye that covered 25% or more of the disc area [10, 27], was done at the fundus photograph reading center. The results of this evaluation were not communicated to the clinical centers. At each follow-up visit, the clinic ophthalmologist also estimated the degree of retinitis progression. Decisions about the reinduction or the initiation of therapy were made by the clinician.

Laboratory Data

Serious or life-threatening adverse events were reported to the Coordinating Center. Data from hematologic testing, chemistry testing, and urinalysis were reported to the Coordinating Center at scheduled data collection visits. Lymphocyte subset analysis was done by flow cytometry at the individual clinical centers. Cultures of blood and urine for CMV were done at the individual clinical centers using standard techniques [22, 27].

Database and Analytic Procedures

This report includes data on the 64 patients who were assigned to treatment before 1 March 1996. Unless otherwise indicated, all analyses were done according to each patient's original treatment assignment. Data collected at the clinic before 1 March 1996 and reported to the Coordinating Center before 29 March 1996 were included. P values for the comparison of CMV culture results were based on the Fisher exact test [29]. Group means were reported for continuous variables with symmetrical distributions, whereas medians were reported for discrete variables and for variables with asymmetrical distributions. Time-to-event curves were derived by using the Kaplan-Meier method [30]. The method of Brookmeyer and Crowley [31] was used to estimate 95% CIs. Event rates were calculated as the number of events divided by the number of person-years at risk. Relative risks, event rates, and mean changes from baseline were estimated by using the Cox proportional-hazards method, Poisson regression, and generalized estimating equations, respectively [32-34]. Variances were estimated by robust techniques. When one group had no events and the other had at least one event, P values were calculated by using an exact test [35].

The Policy and Data Monitoring Board reviewed the accumulating data five times: at three meetings scheduled at 6-month intervals and at two interim conference calls. Clinic personnel caring for patients, with the exception of the Study Chair, did not see interim results. There were no formal stopping rules. The Policy and Data Monitoring Board recommended suspension of the treatment protocol on 1 March 1996 on the basis of the efficacy of cidofovir and a conditional power analysis suggesting that the outcome was unlikely to change if the trial continued to the designed end. The outcome data were first reported to the Research Group on 8 March 1996. Patients were informed of the results after the 8 March meeting.

Results

Top Methods Results Discussion Author & Article Info References

Patient Enrollment, Baseline Characteristics, and Follow-up

Of the 64 patients enrolled in the trial, 26 were assigned to the deferral group, 26 to the low-dose cidofovir group, and 12 to the high-dose cidofovir group. The three groups were reasonably similar at baseline, a finding consistent with the nature of the randomization process (Table 1). Only 14% of the 599 follow-up visits were missed (12% in the deferral group, 15% in the low-dose cidofovir group, and 16% in the high-dose cidofovir group).

Comparison of the low-dose cidofovir group with the deferral group is presented in Table 2, and comparison of the high-dose cidofovir group with the deferral group is shown in Table 3. Comparisons between the low-dose cidofovir group and the deferral group included patients enrolled during the entire course of the trial, whereas comparisons between the high-dose cidofovir group and the deferral group were restricted to patients in stage 2 of the trial because patients were assigned to receive high-dose cidofovir therapy only during stage 2.

Mortality

Mortality rates were similar among the three groups. For the evaluation of low-dose cidofovir, mortality rates were 0.34 per person-year for the deferral group and 0.37 per person-year for the low-dose cidofovir group (P > 0.2). For the evaluation of high-dose cidofovir, mortality rates were 0.24 per person-year for the deferral group and 0.60 per person-year for the high-dose cidofovir group (P > 0.2).

Visual Outcomes

Progression of Retinitis

Both low-dose and high-dose cidofovir were effective in controlling CMV retinitis, as evidenced by their ability to delay the progression of retinitis. For the evaluation of low-dose cidofovir, the median times to progression, as assessed by the fundus photograph reading center, were 21 days for the deferral group and 64 days for the low-dose cidofovir group (P = 0.052, log-rank test) (Figure 1, top). For the evaluation of high-dose cidofovir, the median times to progression, as assessed by the fundus photograph reading center, were 20 days for the deferral group and were not reached for the high-dose cidofovir group (P = 0.009, log-rank test) (Figure 1, bottom) because fewer than 50% of the patients in the high-dose cidofovir group had had progression by the time of data closure. For the evaluation of low-dose cidofovir, the clinicians' evaluations of the median times to progression were 16 days for the deferral group and 82 days for the low-dose cidofovir group (P < 0.001); for the evaluation of high-dose cidofovir, the clinicians' evaluations of these times were 17 days for the deferral group and were not reached for the high-dose cidofovir group (P = 0.013). Clinicians' estimates of time to progression are typically longer than those from a fundus photograph reading center [10]; in our trial, however, the clinicians' estimates of time to progression were shorter for the deferral group and longer for the cidofovir-treated groups than those from the fundus photograph reading center. Because of this difference between treatment groups in the clinicians' evaluation of time to progression, several patients in the deferral group were treated before the fundus photograph reading center had assessed progression. As a result, an intention-to-treat analysis of readings from the fundus photograph reading center tended to underestimate the difference in progression between the deferral group and the cidofovir groups. Therefore, we did a censored analysis in which progressions identified by the fundus photograph reading center were counted only when the patient in question was receiving assigned treatment (Table 2 and Table 3). In this analysis, data from patients in the deferral group who began receiving therapy before a fundus photograph reading center identified progression were censored after the initiation of therapy, and patients assigned to receive cidofovir who discontinued cidofovir therapy were censored at that point. This analysis (Table 2 and Table 3) strengthened the conclusion that cidofovir was effective, as evidenced by the lower P values for the comparisons of the deferral group with the low-dose cidofovir group (P = 0.002) and the high-dose cidofovir group (P = 0.005).

View larger version (10K): [in this window] [in a new window]   Figure 1. Cumulative probability of progression of cytomegalovirus retinitis.Top. Cidofovir (CDV), at a maintenance dose of 3 mg/kg of body weight, compared with observation alone. Bottom. Cidofovir, at a maintenance dose of 5 mg/kg, compared with observation alone.

Other Visual Outcomes

Analysis of the amount of retinal area affected by CMV also showed the efficacy of cidofovir. For the evaluation of low-dose cidofovir, the increases in retinal area involved by CMV were 1.81% per month for the deferral group and 0.80% per month for the low-dose cidofovir group (P = 0.012). For the evaluation of high-dose cidofovir, the increases in retinal area involved by CMV were 2.20% per month for the deferral group and 0.34% per month for the high-dose cidofovir group (P = 0.008). This analysis incorporated information from all follow-up visits. Because patients assigned to the deferral group usually began receiving therapy by week 5, an analysis at the week 5 visit was done to account for the arrest of retinitis progression after the start of therapy. For the evaluation of low-dose cidofovir, the increases in retinal area involved by CMV at week 5 were 2.96% per month for the deferral group and 0.77% per month for the low-dose cidofovir group (P = 0.0008). For the evaluation of high-dose cidofovir, the increases in retinal area involved by CMV at week 5 were 2.98% per month for the deferral group and 0.51% per month for the high-dose cidofovir group (P = 0.003).

No differences were noted in the rate of loss of visual acuity between the deferral group and the cidofovir treatment groups (Table 2 and Table 3), whether the loss was defined as a loss of 15 letters or more or a loss of 30 letters or more. A loss of 15 letters (3 lines) represents a doubling of the visual angle. For the evaluation of low-dose cidofovir, the rate of a loss of 15 letters or more and the rate of a loss of 30 letters or more were 0.25 per person-year and 0.11 per person-year, respectively, for the deferral group and 0.56 per person-year (P > 0.2) and 0.34 per person-year (P > 0.2), respectively, for the low-dose cidofovir group. For the evaluation of high-dose cidofovir, the rate of a loss of 15 letters or more and the rate of a loss of 30 letters or more were 0.32 per person-year and 0.00 per person-year, respectively, for the deferral group and 0.51 per person-year (P > 0.2) and 0.00 per person-year, respectively, (P value not calculable) for the high-dose cidofovir group.

Cytomegalovirus Culture Results

Baseline blood cultures for CMV were positive in approximately one third of patients, and baseline urine cultures were positive in two thirds of patients, regardless of treatment assignment (Table 4). Neither low-dose nor high-dose cidofovir appeared to have much effect on CMV viremia when patients receiving these regimens were compared with patients in the deferral group at the week 3 visit. However, the frequency of CMV viruria was halved at the week 3 visit in both the low-dose and the high-dose cidofovir groups, whereas it remained unchanged in the deferral group (deferral group compared with low-dose cidofovir group, P = 0.092; deferral group compared with high-dose cidofovir group, P = 0.182). In no patients in the deferral group was therapy still being deferred at 15 weeks, and no comparison between the deferral group and either cidofovir group was done for week 15. However, data suggested that the frequency of viremia in treated patients was lower at week 15 than at baseline or at week 3.

Morbidity

Morbidity rates while patients were receiving assigned treatment (Table 5) were calculated from the period of time during which the patient was receiving the assigned treatment. Rates of proteinuria of 2+ or more were 2.6 per person-year for the deferral group, 2.8 per person-year for the low-dose cidofovir group (P > 0.2), and 6.8 per person-year for the high-dose cidofovir group (P = 0.135). No patient developed 4+ proteinuria. No patient in the deferral group developed an elevation of the serum creatinine level to more than 133 µmol/L (1.5 mg/dL) while therapy was being deferred, whereas the rates of this elevation were 0.3 per person-year for the low-dose cidofovir group and 0.5 per person-year for the high-dose cidofovir group. No patient in the deferral group developed neutropenia with neutrophil counts less than 0.500 x 10⁹ cells/L (500 cells/micro L) while therapy was being deferred, whereas the rates of this degree of neutropenia were 0.8 per person-year for the low-dose cidofovir group (P > 0.2) and 1.1 per person-year for the high-dose cidofovir group (P > 0.2). For all patients receiving cidofovir (the low-dose and high-dose groups combined), the rate of reactions to probenecid was 0.70 per person-year. Of the 26 patients assigned to the deferral group, 81% received cidofovir after clinical progression of retinitis. At some time, therefore, a total of 39 patients were treated with low-dose cidofovir and a total of 19 patients were treated with high-dose cidofovir. To estimate rates of cidofovir-related morbidity from a larger denominator, we analyzed morbidity that occurred while patients were receiving cidofovir (regardless of initial treatment group assignment) (Table 6). From this analysis, the rates of proteinuria of 2+ or more were 4.3 per person-year for low-dose cidofovir and 6.7 per person-year for high-dose cidofovir, and the rates of an elevated serum creatinine level to 133 µmol/L or more were 0.7 per person-year for low-dose cidofovir and 0.8 per person-year for high-dose cidofovir. During the study period, 10 patients developed an elevated serum creatinine level to 133 µmol/L or more; 6 of the 10 were receiving low-dose cidofovir and 4 of the 10 were receiving high-dose cidofovir. In 6 of these 10 patients, the elevated serum creatinine levels had resolved as of the last follow-up visit before data closure. As of data closure, serum creatinine levels remained at 133 µmol/L or more in 2 patients treated with low-dose cidofovir (one level was 150 µmol/L [1.7 mg/dL] and one was 168 µmol/L [1.9 mg/dL]) and 2 patients treated with high-dose cidofovir (one level was 150 µmol/L [1.7 mg/dL] and one was 195 µmol/L [2.2 mg/dL]). After data closure, 1 additional patient treated with low-dose cidofovir had an increase in the serum creatinine level to 301 µmol/L (3.4 mg/dL), and this level remained at 204 µmol/L (2.3 mg/dL) as of 6 May 1996 (1 month after the initial increase), despite discontinuation of cidofovir therapy.

In this larger data set, the overall rate of reactions to probenecid was 0.86 per person-year. Of the 10 patients who had reactions to probenecid, 8 discontinued probenecid therapy and, therefore, cidofovir therapy. Reactions leading to the discontinuation of therapy included fever, gastrointestinal upset, skin rash, and angioedema.

Discussion

Top Methods Results Discussion Author & Article Info References

Trials of "small, peripheral" CMV retinitis (involving a comparison of immediate therapy with deferred therapy) are an established way to evaluate promising new drugs for the treatment of CMV retinitis. Trials of this type have documented the efficacy of ganciclovir [8], foscarnet [13], and the ganciclovir intraocular device [14]. Patients whose CMV retinitis lesions are far from the optic nerve or fovea do not have an imminent risk for the loss of visual acuity and may be observed for several weeks without therapy. The threshold chosen for the definition of progression (750 micro m) is sufficiently small that untreated patients "progress," on average, within 3 weeks [8, 13, 14]. The efficacy of a drug can be determined by its ability to prolong the time to progression. Patients assigned to deferred therapy are observed closely and, when retinitis progresses, they are offered therapy, usually with the investigational drug. The safety of these trials is shown by the similar long-term visual outcomes among the three groups in our trial, whether the groups had therapy deferred or received immediate treatment with cidofovir.

Our results indicate that cidofovir, at either dosage, is effective in controlling CMV retinitis. Both levels significantly prolonged the time to progression and reduced the rate of increase in the retinal area involved by CMV. Because of the small number of patients enrolled, this trial had low power to distinguish between outcomes in the low-dose and high-dose cidofovir groups; no attempt was made to compare these outcomes. The fundus photograph reading center estimate of the median time to progression for low-dose cidofovir was similar to previously published photographic estimates of time to progression for intravenous ganciclovir and foscarnet [10, 13]. However, this comparison must be made with caution because of changes in the treatment of AIDS and other factors that might influence the time to progression [10].

The comparison of CMV culture results in patients receiving cidofovir therapy and patients receiving deferred therapy did not show a substantial effect of cidofovir on CMV viremia at the week 3 visit. Conversely, the daily systemic administration of either ganciclovir or foscarnet suppresses CMV viremia [5, 11-13]. The relatively short duration of follow-up in our trial did not allow for comparison of our rates of contralateral ocular and visceral disease with those rates in patients treated with systemic ganciclovir or foscarnet therapy or determination of whether poorer suppression of viremia with cidofovir has an adverse effect on clinical outcomes. Our data suggest that cidofovir reduces viruria, and suppression of CMV viruria with cidofovir has been reported [21]. The differential effect of cidofovir on viruria as opposed to viremia may be due to cidofovir's greater concentration in the urine, which occurs because the mode of elimination of cidofovir is renal. The effectiveness of cidofovir in controlling retinitis, as opposed to suppressing viremia, may be due to the need for diphosphorylation of cidofovir to the active compound, the prolonged intracellular half-life of cidofovir, and the intermittent method of cidofovir administration.

Nephrotoxicity has been the major side effect of treatment with cidofovir. In the early stages of the drug's development, cases of irreversible renal damage requiring dialysis were reported [21]. Intermittent administration (infusions once weekly for induction and once every 2 weeks for maintenance), hydration, and use of probenecid were introduced to reduce the incidence of nephrotoxicity [21]. Furthermore, proteinuria appears to be a precursor of more serious nephrotoxicity, and the development of proteinuria of more than 2+ prompted the discontinuation of cidofovir therapy. Because of the short period during which therapy was deferred, trials of "small, peripheral" CMV retinitis similar to ours provide limited comparative data on long-term morbidity and on mortality. However, within these limitations, our data suggest that the current method of administering cidofovir reduced but did not prevent the development of nephrotoxicity.

Studies done in animals have reported an association between subcutaneous cidofovir therapy and local neoplasms in female rats (Gilead Sciences, Inc. Unpublished data). In our trial, the rate of neoplasms in either of the cidofovir groups was similar to that in the deferral group. Uncontrolled case series [36, 37] have reported that the intravitreous administration of cidofovir is effective in controlling CMV retinitis. However, high-dose intravitreous cidofovir (100 micro g) has caused ocular hypotony (low intraocular pressure) and visual loss [36]. Cases of ocular hypotony have also been reported with the intravenous administration of cidofovir. In our trial, no differences were seen in the rates of ocular hypotony between either the low-dose cidofovir or the high-dose cidofovir group and the deferral group. However, our small sample size limited our ability to detect rare or unusual side effects.

In conclusion, the results of the HPMPC Peripheral Cytomegalovirus Retinitis Trial suggest that both high-dose and low-dose cidofovir, given as intermittent intravenous infusions, are effective in controlling CMV retinitis. The use of probenecid, hydration, and monitoring for proteinuria appeared to minimize but not prevent nephrotoxicity. Cidofovir appears to be a valuable addition to the list of therapies available for CMV retinitis.

Appendix

Clinical Centers Participating in the HPMPC Peripheral Cytomegalovirus Retinitis Trial

Baylor College of Medicine and Cullen Eye Institute (Houston, Texas): Richard A. Lewis, MD, MS (Director); Louise M. Carr, CRA; Kay Doyle, RN; Victor Fainstein, MD; Ronald Gross, MD; Silvia Orengo-Nania, MD; Tobias C. Samo, MD; James W. Shigley, CRA; Steven S. Spencer, COMT; and Mary Weinert, MD.

The Johns Hopkins University School of Medicine (Baltimore, Maryland): J.P. Dunn, MD (Director); John Bartlett, MD; Rebecca Becker, PA-C; Judith Feinberg, MD; Douglas A. Jabs, MD; Daniel A. Johnson, MD; Susan LaSalvia, RN; Tracy Miller, LPN, COT; Laura G. Neisser, COT; Richard D. Semba, MD; Mei-Ling Tay-Kearney, MD; and Pamela Tucker, MD.

Louisiana State University Medical Center (New Orleans, Louisiana): Bruce Barron, MD (Director); Christine Jarrott, RN; Gholam Peyman, MD; and Dennis Swenie, MD.

Mount Sinai School of Medicine (New York, New York): Alan H. Friedman, MD (Director); Robin Ginsburg, MD; Henry Sacks, MD, PhD; Colette Severin, MS; Steven Teich, MD; and Fran Wallach, MD.

New Jersey Medical School (Newark, New Jersey): Ronald Rescigno, MD (Director); Jay Cowan, MD; Catherine Horan, COMT; Patricia Kloser, MD; and Maxine Wanner.

New York University Medical Center (New York, New York): Dorothy N. Friedberg, MD, PhD (Director); Adrienne Addessi, MA, RN; Abraham Chachoua, MD; Douglas Dieterich, MD; Jason Hill; Richard Hutt, RN; Jonathan Ligh, MD; Monica Lorenzo-Latkany, MD; Maria Pei; Therese Powers, MS; and Carol Scoppe.

Northwestern University (Chicago, Illinois): David V. Weinberg, MD (Director); Lee M. Jampol, MD; Alice T. Lyon, MD; Annmarie Munana, RN; Robert Murphy, MD; Frank Palella, MD; Len Richine; Zuzanna Strugala, OMA; and Gloria Valadez.

University of California, Los Angeles (Los Angeles, California): Gary N. Holland, MD (Director); Margrit E. Carlson, MD; Suzette Chafey, RN, NP; W. David Hardy, MD; Ann K. Johiro, RN, MN, FNP; Lesley Macarthur-Chang, MEd; Maureen A. Martin, RN, MN, FNP; Ardis A. Moe, MD; Catherine A. Strong, MT, ASCP; Adnan Tufail, MB, BS, FRCOphth; Prabha S. Ugalat; and James M. Weisz, MD.

University of California, San Diego (San Diego, California): William R. Freeman, MD (Director); J. Fernando Arevalo-Colina, MD; Tom Clark, CRA; Cheryl L. Jarman; Linda Meixner, RN; Tze Chiang Meng, MD; Stephen Spector, MD; Ibrahim Taskintuna, MD; and Francesca J. Torriani, MD.

University of California, San Francisco (San Francisco, California): James O'Donnell, MD (Director); Pierre Alfred, MD, Fermin Ballesteros; David Clay; Rebecca Coleman, PharmD; Kathleen Gordon, MD; Deborah Gumbley; Jacqueline Hoffman; Alexander Irvine, MD; Mark Jacobson, MD; James Larson, COT; Leonardo Macalalag; Michael Narahara; Mary Payne, RN; Stuart Seiff, MD; Scarlette Wilson, MD; and Harlan Woodring.

University of Miami School of Medicine (Miami, Florida): Janet Davis, MD (Director); Paul Mendez, MD; Timothy Murray, MD; and Tom Simmons.

University of North Carolina (Chapel Hill, North Carolina): Charles van der Horst, MD (Director); Jan Kylstra, MD; David Wohl, MD; and Kimberlee Ziman, BA.

University of South Florida (Tampa, Florida): Peter Reed Pavan, MD (Director); Gary A. Bergen, MD; Steven M. Cohen, MD; Jayne Ann Craig, OT; Richard L. Dehler, MD; Elizabeth Elbert, MD; Roger W. Fox, MD; W. Sanderson Grizzard, MD; Mark E. Hammer, MD; Lizette S. Hernandez, MD; Stacey Herrera; Douglas Holt, MD; Stephen Kemp, MD; Julie A. Larkin, MD; Dennis K. Ledford, MD; Richard F. Lockey, MD; Matthew M. Menosky, MD; Sharon Millard, RN, COT; Jeffrey P. Nadler, MD; Robert P. Nelson Jr., MD; Dorece Norris, MD; L. David Ormerod, MD; Scott E. Pautler, MD; Sarah Jessica Poblete, MD; Dena Rodriguez, COT; Kevin P. Rosenbach, MD; Daniel W. Seekins, MD; and John R. Toney, MD.

Resource Centers

Chairman's Office, The Johns Hopkins University School of Medicine (Baltimore, Maryland): Douglas A. Jabs, MD (Study Chairman); Joan M. Dodge; Joan L. Klemstine; Tracey A. Schuerholtz; and Maria Stevens.

Coordinating Center, The Johns Hopkins University School of Hygiene and Public Health (Baltimore, Maryland): Curtis L. Meinert, PhD (Director); Debra Amend-Libereci; Laura Coleson, RN, MPH; Karen L. Collins; Betty J. Collison; Chris Dawson; John Dodge; Michele Donithan, MHS; Cathleen Ewing; Nancy Fink, MPH; Charlotte Gerczak, MLA; Judith Harle; Janet T. Holbrook, MS, MPH; Robert Huffman; Milana R. Isaacson, BS; Adele M. Kaplan Gilpin, JD, PhD; Madelyn Lane; Charlene R. Levine, BS; Barbara Martin; Jill Meinert; Deborah J. Nowakowski; Rosetta M. Owens; Bonnie Piantadosi, MSW, MPH; Alfred Saah, MD, MPH; Michael Smith; James Tonascia, PhD; and Mark L. Van Natta, MHS.

Fundus Photograph Reading Center, University of Wisconsin (Madison, Wisconsin): Matthew D. Davis, MD (Director); Jane Armstrong; Judith Brickbauer; Rosemary Brothers; Marika Chop; Larry Hubbard, MAT; Dolores Hurlburt; Linda Kastorff; Yvonne Magli; Michael Neider; Jim Onofrey; Vicki Stoppenbach; and Marilyn Vanderhoof-Young.

Drug Distribution Center, McKesson Bioservices Corporation (Rockville, Maryland): Mark Walls and Robert Hughes.

National Eye Institute (Bethesda, Maryland): Natalie Kurinij, PhD, and Richard L. Mowery, PhD.

National Institute of Allergy and Infectious Diseases (Bethesda, Maryland): Beverly Alston, MD, and Mary Foulkes, PhD.

Study Committees

Officers of the Study: Douglas A. Jabs, MD (Chair); Matthew D. Davis, MD; Natalie Kurinij, PhD; Curtis L. Meinert, PhD; and Richard L. Mowery, PhD.

Steering Committee: Douglas A. Jabs, MD (Chair); Adrienne Addessi, MA, RN; Beverly Alston, PhD; Tom Clark, CRA; Matthew D. Davis, MD; Judith Feinberg, MD; William Freeman, MD; Janet Holbrook, MS, MPH; Gary N. Holland, MD; Larry Hubbard, MAT; Mark Jacobson, MD; Natalie Kurinij, PhD; Richard A. Lewis, MD, MS; Lesley Macarthur-Chang, MEd; Curtis Meinert, PhD; Richard Mowery, PhD; Robert Murphy, MD; Bruce Polsky, MD; and James Tonascia, PhD.

Studies of Ocular Complications of AIDS-AIDS Clinical Trials Group Joint Executive Committee: Douglas A. Jabs, MD (Chair); Matthew D. Davis, MD; William R. Duncan, PhD; Judith Feinberg, MD; Harold Kessler, MD; Natalie Kurinij, PhD; A. Garey Lambert; Curtis L. Meinert, PhD; Richard L. Mowery, PhD; William Powderly, MD; Steve Schnittman, MD; Steven Spector, MD; and James Tonascia, PhD.

Policy and Data Monitoring Board: Byron W. Brown Jr., PhD (Chair); Brian Conway, MD; James Grizzle, PhD; Robert Nussenblatt, MD; John P. Phair, MD; Harmon Smith, PhD; and Richard Whitley, MD. Nonvoting members: Beverly Alston, MD; Matthew D. Davis, MD; Mary Foulkes, PhD; Douglas A. Jabs, MD; Natalie Kurinij, PhD; Curtis L. Meinert, PhD; Richard L. Mowery, PhD; and James Tonascia, PhD.

Community Advisory Board: Ben Cheng, Kevin Frost, A. Garey Lambert, and Michael Marco.