Discontinuation of Secondary Prophylaxis in Patients with Cytomegalovirus Retinitis Who Have Responded to Highly Active Antiretroviral Therapy

Before the introduction of highly active antiretroviral therapy (HAART), cytomegalovirus (CMV) retinitis was a common complication in patients with advanced HIV disease; it was characterized by frequent relapses and complications that eventually led to blindness [1]. Therapy for CMV retinitis was well established in the pre-HAART era; it consisted of an induction phase that aimed to control the infection, followed by a lifelong maintenance phase to avoid or delay relapses [2]. The profound inhibition of viral replication caused by HAART and its beneficial effects on the immune response [3] have led to a dramatic decline in morbidity and mortality among patients with advanced immune depletion [4]. These effects have also paved the way for the discontinuation of secondary prophylaxis against several opportunistic pathogens, such as Mycobacterium avium complex and Pneumocystis carinii [5, 6].

To our knowledge, to date, no randomized study of discontinuation of secondary prophylaxis against CMV retinitis has been published. The feasibility of such a trial is doubtful, given the sharp decline in the incidence of this opportunistic infection. Consequently, it is important to accumulate data from prospective cohorts that are carefully followed. We report the results of an open, nonrandomized, prospective study of discontinuation of secondary prophylaxis against CMV retinitis in a relatively large cohort of patients with increased CD4⁺ cell counts after HAART.

Subjects and methods. Subjects were recruited prospectively from the AIDS units of 4 tertiary teaching hospitals in Madrid, Spain. The ethics committees at each participating center approved the study, and all the patients provided written informed consent.

Subjects were required to have documented HIV infection and CMV retinitis that had been inactive for at least 2 months and, in the case of bilateral eye involvement, no loss of vision. Other eligibility criteria were receipt of HAART (2 nucleoside analogues plus 1 or 2 protease inhibitors) and a sustained response (as determined on 2 occasions at least 12 weeks apart) characterized by a CD4⁺ count of >100 cells/µL together with an HIV virus load of <500 copies/mL, or a CD4⁺ count of >150 cells/µL together with an HIV virus load of <10⁴ copies/mL (as determined by use of the bDNA assay [Chiron]) or <3 × 10⁴ copies/mL (as determined by use of reverse transcriptase—PCR).

In vitro lymphoproliferative assays (LPA) for CMV antigens were performed on a subgroup of patients on discontinuation of prophylaxis. Human CMV strain AD-169 was obtained from the American Type Culture Collection (Manassas, VA). Stocks of ultraviolet—inactivated virus were generated as described elsewhere [7]. Peripheral blood mononuclear cells from HIV-infected patients were isolated by means of centrifugation through Ficoll-Hypaque gradient and resuspended in RPMI (Roswell Park Memorial Institute) medium supplemented with 2 mM L-glutamine, antibiotics, and 10% human AB serum. Cells were seeded in 96-well plates (10⁵ cells/well), incubated with 5 pfu/cell of the ultraviolet-inactivated CMV (based on the original titer) or control supernatants and cultured for 5 days. Cell proliferation was measured by the incorporation of thymidine during the last 18 h of the culture period. Patients were classified into response groups on the basis of the fold induction versus that for control subjects, as follows: <4-fold induction, weak response; 4- to 8-fold induction, moderate response; and >8-fold induction, intense response.

After the discontinuation of secondary prophylaxis against CMV retinitis, patients were evaluated by means of dilated bilateral fundus examinations and CMV antigenemia every 2 weeks for 3 months, once a month up to month 12, and quarterly thereafter. CD4⁺ counts and virus load were measured every 2 months. The study end point was a relapse of CMV retinitis confirmed by findings on retinal photographs.

Results. The study began in March 1998. The characteristics of the 36 patients included in the study are shown in table 1. At the time of discontinuation of prophylaxis, an LPA for CMV antigens was performed for 17 of the 36 patients. Of these 17 patients, the response was weak in 10 patients (58.8%), moderate in 4 (23.5%), and intense in 3 (17.6%).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Clinical characteristics of 35 patients with AIDS who stopped secondary prophylaxis against cytomegalovirus (CMV) retinitis.

After a median follow-up of 90 weeks (interquartile range, 53–97 weeks), relapse of CMV disease occurred in 1 patient with immunological failure. He was a 35-year-old homosexual man who had received the diagnosis of bilateral CMV retinitis. His lowest CD4⁺ count was 0 cells/µL, and his highest HIV virus load was 4.70 log copies/mL. At the time of withdrawal of CMV secondary prophylaxis, his CD4⁺ count was 262 cells/µL, and the HIV virus load was 3.33 log copies/mL. Throughout the study, his CD4⁺ count decreased gradually and fell below 100 cells/µL (to 74 cells/µL) for the first time in week 40, while CMV retinitis was in remission. In week 44, a bilateral relapse of CMV retinitis was detected; the CD4⁺ count was 62 cells/µL and the virus load was 2.85 log copies/mL. The patient received therapy with ganciclovir, and remission of CMV retinitis was achieved. His antiretroviral therapy was changed, and he achieved a virus load of <50 copies/µL, despite which his CD4⁺ count remained <75 cells/µL. At the time of this writing, the patient is medically well, continues to receive ganciclovir (10 mg/kg 3 times a week), and has experienced no further relapses of CMV retinitis. In vitro LPA for CMV antigens demonstrated a weak response.

During follow-up, the CD4⁺ counts in the remaining 35 patients were maintained at levels equal to or higher than they had been at baseline (table 2). Modifications to the initial HAART regimen were made for 13 patients (including the patient who experienced a relapse). The reasons for modification were virological failure in 6 patients, adverse events in 4, and simplification of the regimen in 3. Transient positive CMV antigenemia (⩾3 positive cells/2 × 10⁵ leukocytes) was detected in 3 patients. One patient was lost to follow-up during week 40. Two patients died, 1 as a result of lymphoma during week 24, and the other as a result of end-stage hepatitis C–related liver disease in week 16.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

CD4⁺ cell counts and HIV virus load for patients with AIDS after the discontinuation of secondary prophylaxis against cytomegalovirus retinitis.

After discontinuation of prophylaxis, no new retinal detachments were observed. Infections due to herpes simplex virus were detected in 5 patients and infections by varicella-zoster virus were detected in 3 patients.

Discussion. We undertook a prospective and structured study of whether secondary prophylaxis against CMV can be safely discontinued in patients with AIDS and CMV retinitis after successful treatment with HAART. A total of 36 patients discontinued their prophylaxis after a median of 75 weeks of HAART with a median CD4⁺ count of 287 cells/µL. After a median follow-up of 90 weeks, no reactivation or progression of retinitis was observed in the 35 patients who showed a persistent response to antiretroviral therapy. One patient who experienced immunological failure relapsed during week 44 (CD4⁺ count, 62 cells/µL).

Relatively soon after the introduction of HAART, several studies of small numbers of patients suggested that CMV retinitis might not reactivate after discontinuation of secondary prophylaxis in patients who responded to HAART [8–10]. In the last 2 years, 3 prospective studies of withdrawal of secondary prophylaxis for CMV retinitis have been published (table 3). The first study [11] included 14 patients; no relapses were detected after a median follow-up of 16.4 months. In the second study [12], 3 of 22 patients who discontinued prophylaxis experienced a relapse of CMV retinitis. It is of note that the infections of all 3 patients had failed to respond to HAART, and at the time of reactivation all patients had CD4⁺ counts <50 cells/µL. The third study [13] was multinational and included 48 patients who discontinued prophylaxis, 2 of whom developed recurrent CMV disease. In one of these 2 patients, CMV retinitis relapsed 11 weeks after stopping maintenance therapy. Surprisingly, the CD4⁺ counts at baseline and at the time of CMV retinitis relapse were 302 and 352 cells/µL, respectively. In the other of the 2 patients, peripheral neuropathy developed, which the authors considered was possibly related to CMV despite the absence of laboratory and pathology evidence of CMV disease. The CD4⁺ counts at baseline and at the time of peripheral neuropathy were 140 and 106 cells/µL, respectively [13].

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 3

Characteristics of patients included in the largest prospective studies on the discontinuation of secondary prophylaxis against cytomegalovirus retinitis.

The results of our study and these 3 others [8, 11, 12] support the notion that it is safe to stop secondary prophylaxis against CMV in patients with AIDS and quiescent CMV retinitis who have a sustained increase in the CD4⁺ cell count in response to HAART. An important limitation of the 3 previously mentioned studies, and of our own, is that the CD4⁺ cell count used as an inclusion criterion was substantially lower than the actual CD4⁺ counts at baseline. Thus, we still do not know at which CD4⁺ count it is safe to stop prophylaxis. In an attempt to address this important question, we pooled the data on CD4⁺ counts at the time of withdrawal of prophylaxis from our study and the 3 others that give data for each patient [8, 11, 12] (table 4). The total number of patients included in the 4 studies was 79, and the median CD4⁺ count at the time of discontinuation of prophylaxis was 269 cells/µL (interquartile range, 167–360 cells/µL). More than two-thirds of the patients had CD4⁺ counts of >200 cells/µL, less than one-third had CD4⁺ counts in the range 100–200 cells/µL, and only 3 patients had CD4⁺ counts of <100 cells/µL.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 4

Pooled CD4⁺ cell counts for 79 patients with AIDS at the time of discontinuation secondary prophylaxis for cytomegalovirus retinitis.

Of note, we found that the LPA for CMV at the time of discontinuation of prophylaxis demonstrated weak or moderate responses in ∼80% of tested patients; despite this, none of them had a relapse. Therefore, we cannot recommend the performance of this test to guide clinical decisions for such patients. Nevertheless, studies that have quantified antigen-specific CD4⁺ lymphocyte responses by use of flow cytometry have found that most patients with quiescent CMV-associated end-organ disease after HAART demonstrated strong CMV-specific CD4⁺ lymphocyte responses [14].

In conclusion, our findings and those of other published studies [8, 11, 12] support the discontinuation of secondary prophylaxis against CMV retinitis in patients with AIDS and quiescent CMV retinitis who have a sustained response to HAART that is characterized by a CD4⁺ count of >200 cells/µL. Few of the patients enrolled in the other studies [8, 11, 12] had baseline CD4⁺ counts of 100–200 cells/µL. Nevertheless, it may be reasonable to discontinue secondary prophylaxis in such patients as well, considering that almost all people who subsequently have progressive CMV retinitis had CD4⁺ counts of <100 cells/µL.

After the suppression of secondary prophylaxis, patients may be safely monitored with periodical CD4⁺ cell counts. In those who later developed immunological failure, close observation for evidence of recurrent retinitis or resumption of secondary prophylaxis may be chosen, depending on the risk of irrecoverable loss of vision.

• Received June 20, 2001.
• Revision received September 5, 2001.

• © 2002 by the Infectious Diseases Society of America