Changes in the natural history of cytomegalovirus retinitis following the introduction of highly active antiretroviral therapy

Deayton, Jane Ra; Wilson, Paulineb; Sabin, Caroline Ac; Davey, Clare Cb; Johnson, Margaret Ad; Emery, Vincent Ca; Griffiths, Paul Da


Objective: To determine the effect of highly active antiretroviral therapy (HAART) on the natural history of cytomegalovirus (CMV) retinitis.

Design and participants: Retrospective analysis of 103 consecutive patients diagnosed with CMV retinitis between 1990 and 1998.

Setting: Specialist HIV medicine department of a London hospital.

Main outcome measures: Incidence of CMV retinitis, time to death following diagnosis, episodes of progression, incidence of inflammatory complications. The date of first use of HAART was January 1995. Data were censored on 30 June 1998.

Results: The incidence of CMV retinitis has declined dramatically following the introduction of HAART. Survival following CMV retinitis increased from a median of 0.65 years prior to 1995 to a median of 1.07 years after this date (P  = 0.004). In multivariate analyses HAART was independently associated with improved survival (P  = 0.02) and the association with year of diagnosis was no longer significant, suggesting that this effect is predominantly due to HAART. None of the patients receiving HAART experienced progression after 6 months of treatment. Complications of retinitis such as retinal detachment, uveitis and optic atrophy occurred in 39% of patients. The rare inflammatory complications of vitritis and cystoid macular oedema ocurred only in recipients of HAART.

Conclusions: The introduction of HAART has had a major impact on the natural history of CMV retinitis with improved survival time and decreased risk of progression following diagnosis. However, immune reconstitution may be associated with inflammatory complications which can result in significant visual loss in the absence of active CMV disease.

Author Information

From the Departments of aVirology, bOphthalmology, cPrimary Care and Population Sciences and dThoracic/HIV Medicine, Royal Free Hampstead National Health Service Trust and Royal Free and University College School of Medicine, London, UK.

Received: 6 July 1999;

revised: 27 January 2000; accepted: 7 February 2000.

Sponsorship: Supported by Medical Research Council (MRC) Training Fellowship G84/5139 & Grant 1R01 AI41687-01 from the National Institutes of Health.

Correspondence to Dr J Deayton, MRC Research Fellow, Department of Virology, Royal Free and University College School of Medicine, Royal Free Campus, Rowland Hill Street, London NW3 2PF, UK. Tel: 0207 830 2997/fax 0207 830 2854; e-mail:

Article Outline
Back to Top | Article Outline


Until recently, cytomegalovirus (CMV) retinitis was the most important cause of visual impairment in persons infected with HIV. CMV retinitis was characterized by relapsing retinal disease resulting in progressive visual loss and increased risk of death despite the use of repeated-induction anti-CMV therapy and maintenance treatment between episodes. However, the introduction of highly active antiretroviral therapy (HAART) has resulted in a dramatic decline in opportunistic infections, including CMV retinitis, and a reduction in the death rate from AIDS [1–3]. The risk of developing CMV retinitis following HAART remains uncertain; there are reports of appearance or relapse of retinitis despite an apparently good response to HAART [4–6] but also several reports of remission of disease following HAART in the absence of specific anti-CMV therapy [7,8]. Inflammatory complications of opportunistic infections are recognized increasingly as a consequence of immune restoration following HAART and there is evidence that the clinical manifestations of retinitis may be modified if it does occur.

We report here the cases of CMV retinitis seen at our institution to document changes in the natural history following the introduction of HAART. We specifically wished to determine the effect of HAART on: (1) the incidence of retinitis; (2) survival after diagnosis of retinitis; (3) risk of retinitis progression; and (4) complications of retinitis.

Back to Top | Article Outline



All patients attending the Ian Charleson Day Centre at the Royal Free Hospital who were diagnosed with CMV retinitis between November 1990 and January 1998 were studied. Patients with asymptomatic CMV viraemia or other CMV end-organ disease were excluded from the analysis. All patients who had a CD4 cell count of 100 × 106 cells/l or less were examined monthly by indirect fundoscopy conducted by an experienced ophthalmologist. Initial and recurrent episodes of retinitis were diagnosed using standard clinical criteria. Following a diagnosis of retinitis, patients were evaluated by the ophthalmologist weekly while on induction anti-CMV therapy and monthly thereafter. The decision to discontinue maintenance therapy was made by the ophthalmologist in consultation with the HIV physician.

Clinical findings were obtained from case notes, the hospital's HIV database and from laboratory records. The following data were recorded: sex, date of birth, dates of initial diagnosis of retinitis and subsequent episodes of progression, clinical features at presentation, anti-CMV therapy, complications of retinitis, CD4 cell counts, plasma HIV RNA where available, receipt of HAART and date of death. HAART was defined as combination antiretroviral therapy containing a protease inhibitor (PI); there were no patients in this cohort receiving combination anti-retroviral therapy inclusive of a non-nucleoside reverse transcriptase inhibitor (NNRTI). The date of introduction of HAART was defined as 1 January 1995, this being the earliest date that protease inhibitors became available at our centre although they did not become widely used until mid-1996.

Back to Top | Article Outline
Statistical analysis

The most recent CD4 cell count prior to CMV diagnosis was used in all analyses. Where this was taken more than 6 months before diagnosis the value was considered missing. Comparisons of CD4 cell counts at the time of diagnosis in different subgroups of the patients were performed using the Mann–Whitney U test.

Survival and time to first progression of retinitis were analysed using standard survival methods. Survival was considered from the date of CMV diagnosis until the date of death. Progression was considered from the date of CMV diagnosis until the date of the first clinical progression requiring re-induction anti-CMV therapy. For both analyses patient follow-up was right-censored at the end of June 1998 if the patient remained alive on that date. Two patients left the UK before the end of the study period; these patients were right-censored on the date of departure from the country.

Survival and time to progression were illustrated using Kaplan–Meier methods. Differences in rates between subgroups were tested for significance using the log-rank test. Univariate and multivariate Cox proportional hazards regression models were used to quantify the independent effects of a number of factors on survival and progression. Age at diagnosis, CD4 cell count at diagnosis, date of diagnosis (before or after 1 january 1995), CMV status at diagnosis (symptomatic or asymptomatic) and the use of HAART before diagnosis were modelled as fixed covariates at baseline. The use of subsequent HAART was included as a time-updated covariate which took the value 0 before the date that the patient started to receive HAART and 1 thereafter. Those patients who had received HAART before CMV diagnosis therefore took the value 1 at all times. The impact of progressions of CMV disease on survival was also considered as a time-dependent covariate whose value increased by one each time the patient experienced another progression of retinitis.

Back to Top | Article Outline



The cohort consists of 103 patients diagnosed with CMV retinitis between 1990 and 1998. Follow-up on these patients ranged from 0.04 to 3.62 years (median 0.69 years). Demographic and clinical details of these patients are summarized in Table 1. Date of first diagnosis of retinitis was available for 101 patients. There has been a steady decline in the incidence of CMV retinitis following the introduction of HAART despite the fact that the total number of patients attending this centre has increased yearly over this time. The decline in incidence is most marked after the middle of 1996 when HAART had become widely prescribed in the clinic (Fig. 1).

The diagnosis of retinitis was made in 58 patients (56.3%) after the development of visual symptoms, most commonly blurred vision or loss of visual field (36.9%) but also floaters or flickering (19.4%). As clinical awareness of retinitis increased, screening for retinitis was introduced. Initially this was conducted in individuals with low CD4 cell counts, progressive HIV disease or extra-ocular CMV disease. Routine screening for CMV viraemia by polymerase chain reaction (PCR) was introduced to this clinic in 1997 as it had been shown that this identifies individuals at increased risk of CMV disease [9]. Patients with CMV viraemia detected by PCR underwent more intensive screening for retinitis. Therefore, in this cohort 35 patients (34%) were asymptomatic at the time of diagnosis.

CD4 cell counts in the 6 months prior to diagnosis were available for 96 (93.2%) of the patients; CD4 cell counts at the time of diagnosis were low (median 10 × 106 cells/l). There was no evidence that CD4 cell counts at the time of diagnosis were changing over time (P  = 0.31 for comparison of CD4 cell counts of diagnoses made before and after January 1995).

Back to Top | Article Outline
Antiviral therapies received

Almost all patients received ganciclovir (100 patients, 97.1%). In addition, 33 patients (32.0%) received treatment with foscarnet and 10 (9.7%) received treatment with cidofovir. Twenty-one patients (20.4%) received HAART; six of these patients had started HAART before their diagnosis of retinitis (from 11 days to 2.97 years before diagnosis of CMV retinitis). The remaining patients started HAART between 3 days and 2.14 years after diagnosis of retinitis.

Back to Top | Article Outline

Over the study period 87 of the patients died, a Kaplan–Meier rate of 90.0% by 3.62 years. The median survival time was 0.76 years. Patients diagnosed since 1995 had a longer median survival time of 1.07 years compared to those diagnosed before 1995, in whom the median survival was 0.65 years (P  = 0.004 log-rank test) (Fig. 2).

Univariate Cox regression analyses showed that individuals diagnosed after 1 January 1995 had a 48% reduction in the hazard of death compared with those diagnosed before this date (P  = 0.004) (Table 2). In addition, the receipt of HAART was associated with a 61% reduction in the hazard of death (P  = 0.009). Each additional progression of CMV disease necessitating re-induction with ganciclovir was associated with a 23% increase in the hazard of death (P  = 0.03). Neither age nor CD4 cell count at diagnosis were associated with survival in univariate analyses. In multivariate analyses treatment with HAART remained independently associated with an improved survival (P  = 0.02) (Table 3). After adjusting for all other factors, patients who had commenced HAART before their CMV diagnosis had an increased hazard of death (P  = 0.08). However, year of diagnosis was no longer significantly associated with survival (P  = 0.52). Increased age at diagnosis was associated with a poorer prognosis (P  = 0.04). Being symptomatic at diagnosis was associated with a reduction in the hazard of death although this was marginally non-significant (P  = 0.06). Neither the number of progressions of CMV retinitis nor the CD4 count at the time of diagnosis remained significant after adjusting for the other factors in the model.

Back to Top | Article Outline
Progression of retinitis

Over the follow-up period, 66 patients experienced a progression of their CMV retinitis, a Kaplan–Meier progression rate of 84.0% by 2.79 years after initial diagnosis (Fig. 3). In these patients the number of progressions ranged from one to eight (median, two progressions). The median time to first progression was 0.31 years. In univariate Cox regression analysis none of the factors, including the use of HAART, were associated with time to progression (Table 2). Consequently, no multivariate analyses were performed. However, no progressions of retinitis occurred in patients who had received more than 6 months of HAART.

Back to Top | Article Outline
Complications of retinitis

Information on complications was available for 100 (97.1%) patients. 39 patients experienced a total of 46 complications of retinitis. The most common complication recorded was retinal detachment (19 patients, seven in the left eye only, 10 in the right eye only and two bilaterally) although uveitis (11 patients) and optic atrophy (seven patients) were also relatively common.

Eight of these 39 patients experienced inflammatory complications of retinitis (Table 4). Vitritis (inflammation of the posterior segment of the eye) was documented in five patients and cystoid macular oedema (CMO) (swelling of the macula caused by the development of fluid-filled spaces in the sensory retina) in four patients, all of whom were receiving HAART at the time. No inflammatory complications occurred in patients not receiving HAART (P  < 0.00001). There was no association between the development of vitritis or CMO and prior or current receipt of any particular anti-CMV agent. All patients in this cohort were still receiving maintenance therapy for CMV at the time of development of the inflammatory complication.

Vitritis was diagnosed a median of 105 days (range 91 to 237 days) after starting HAART. At the time of starting HAART, CD4 cell counts ranged from 0 to 100 × 106 cells/l (median 5 × 106 cells/l). At the time of development of vitritis CD4 cell counts had increased in four patients (increases of 29, 35, 97 and 120 × 106 cells/l) and decreased in one patient (decrease of 90 × 106 cells/l). CMO was diagnosed 162, 193, 231 and 651 days after starting HAART. CD4 cell counts were available for three of these patients. Prior to HAART, the CD4 cell counts were 3, 10 and 100 × 106 cells/l; at the time of diagnosis of CMO the counts had increased in two patients (by 113 and 197 × 106 cells/l) and decreased in one (by 46 × 106 cells/l). There were no significant differences in maximum CD4 cell counts achieved in the first 12 months after HAART between patients who developed inflammatory complications (CD4 cell range, 51–503 × 106/l; median 197 × 106/l) and those who did not (CD4 range, 8–330 × 106/l; median 142 × 106/l). Plasma HIV RNA levels were available for six of eight patients at the time of first diagnosis of vitritis or CMO. In five of these patients the HIV RNA was undetectable (< 400 copies/ml) and it was > 750 000 copies/ml in the sixth.

Back to Top | Article Outline


The incidence of CMV retinitis has declined dramatically since the introduction of HAART. The majority of recent new cases of retinitis at our hospital have occurred in individuals presenting with untreated HIV infection. This is reflected in the finding of this study that there has been no significant change in the CD4 cell count at the time of first diagnosis of retinitis over time. It can be concluded therefore that an increased CD4 cell count following HAART is protective against the development of retinitis even in patients that were previously at high risk.

Survival following a diagnosis of retinitis has improved significantly and these results suggest that this is predominantly due to the introduction of HAART. Other groups have reported similar findings [10]. Although patients who commenced HAART after a diagnosis of retinitis did experience progression of disease, no progressions were seen after HAART had been taken for more than 6 months. This observation may be important when considering discontinuation of maintenance therapy for retinitis in patients responsive to HAART, a practice which has been successfully adopted by several groups [11–13].

The previously uncommon complications of vitritis and CMO only occurred in association with the use of HAART in this cohort. CMV retinitis in the pre-HAART era was characterized by only mild and often asymptomatic vitreous and anterior chamber inflammation. These newer manifestations of CMV retinitis are important as they can cause considerable loss of visual acuity in the absence of active retinitis [14]. Although a direct drug effect has been suggested as a possible aetiology for these complications, it is now thought that these are inflammatory conditions resulting from immune restoration. One patient in our series developed vitritis and CMO despite a fall in the total CD4 cell count and there was no difference in maximum CD4 cell counts attained from those without inflammatory complications, suggesting that patients at risk may not be identifiable simply by considering the CD4 cell count. Our findings do not support the hypothesis that the risk of inflammatory complications is increased following the discontinuation of maintenance therapy for CMV.

The protective effect of HAART on CMV disease may be explained by the recently described reconstitution of T-lymphocyte responses to a variety of pathogens, including CMV, following HAART [15–17]. Similar effects on the incidence of other opportunistic infections such as progressive multifocal leucoencephalopathy, Kaposi's sarcoma, cryptosporidiosis and microsporidiosis have been reported [18–23]. In parallel, paradoxical inflammatory complications attributed to immune reconstitution have been described in Mycobacterium tuberculosis infection following HAART [24]. Patients who develop vitritis have enhanced proliferative responses to antigen in vitro compared with controls and it is possible that it is the degree of recovery of these specific responses rather than the actual cell count which determines the degree of inflammatory response to intra-ocular CMV antigen [25]. Currently, the optimal management of vitritis and CMO remains uncertain. Patients may respond to oral or ocular corticosteroids without reactivation of CMV retinitis [26]. Alternative therapies such as oral acetazolamide and topical non-steroidal agents are of limited effectiveness [14]. However, recovery may occur in the absence of anti-inflammatory therapy which suggests that these complications may be transient phenomena [27].

The introduction of HAART has made a dramatic impact on the management of HIV infection. However, it has recently become apparent that the longer-term receipt of PIs is associated with severe metabolic complications which may limit their future use in some patients [28]. Adherence to complicated therapeutic regimens is difficult and treatment failure secondary to antiretroviral resistance is increasingly problematic in clinical practice [29]. Interest in `PI-sparing' HAART regimens containing NNRTIs is increasing. This cohort did not contain any patients receiving such a regimen as NNRTIs came into use later than PIs. There are therefore no data on the effect of NNRTI-containing HAART regimens on CMV disease.

It remains unknown how long the beneficial effect of HAART on both the development of retinitis and risk of progression after previous disease may be sustained. It is probable that the natural history of CMV disease will continue to evolve and so further monitoring is required to document these changes.

Back to Top | Article Outline


The authors thank Dr Amanda Mocroft, Professor Clive Loveday, Dr Helen Devereux and staff in immunology for help in the collection of data for this study.

Back to Top | Article Outline


1. Centres for Disease Control and Prevention. Update: trends in AIDS incidence. MMWR 1997, 46: 861 –867.
2. Palella FJ, Delaney KM, Moorman AC. et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998, 338: 853 –860.
3. Mocroft A, Vella S, Benfield T. et al. Changing patterns of mortality across Europe in patients infected with HIV-1. Lancet 1998, 352: 1725 –1730.
4. Jacobson MA, Zegans M, Pavan PR. et al. Cytomegalovirus retinitis after initiation of highly active antiretroviral therapy. Lancet 1997, 349: 1443. 1443.
5. Gilquin J, Piketty C, Thomas V. et al. Acute cytomegalovirus infection in AIDS patients with CD4 counts above 100 ×106 cells/l following combination antiretroviral therapy including protease inhibitors. AIDS 1997, 11: 1659 –1660.
6. Mallolas J, Arrizabalga J, Loncá M. et al. Cytomegalovirus disease in HIV-infected patients treated with protease inhibitors. AIDS 1997, 11: 1785 –1787.
7. Whitcup SM, Fortin E, Nussenblatt RB. et al. Therapeutic effect of combination antiretroviral therapy on cytomegalovirus retinitis. JAMA 1997, 277: 1519 –1520.
8. Reed JB, Schawb IR, Gordon J. et al. Regression of cytomegalovirus retinitis associated with protease-inhibitor treatment in patients with AIDS. Am J Ophthalmol 1997, 124: 199 –205.
9. Bowen EF, Sabin C, Wilson P. et al. Cytomegalovirus viraemia detected by polymerase chain reaction identifies a group of patients at high risk of CMV disease. AIDS 1997, 11: 889 –893.
10. Walsh JC, Jones CD, Barnes EA. et al. Increasing survival in AIDS patients with cytomegalovirus retinitis treated with combination antiretroviral therapy including HIV protease inhibitors. AIDS 1998, 12: 613 –618.
11. Tural C, Romeu J, Sirera G. et al. Long-lasting remission of cytomegalovirus retinitis without maintenance therapy in human immunodeficiency virus-infected patients. J Infect Dis 1997, 177: 1080 –1083.
12. Macdonald JC, Torriani FJ, Morse LS. et al. Lack of reactivation of cytomegalovirus (CMV) retinitis after stopping CMV maintenance therapy in patients with sustained elevations in CD4 T cells in response to highly active antiretroviral therapy. J Infect Dis 1998, 177: 1182 –1187.
13. Vrabec TR, Baldassano VF, Whitcup SM. Discontinuation of maintenance therapy in patients with quiescent cytomegalovirus retinitis and elevated CD4 counts. Ophthalmology 1998, 105: 1259 –1264.
14. Silverstein BE, Smith JH, Hasenyager Smith J. et al. Cystoid macular edema associated with patients with the acquired immunodeficiency syndrome. Am J Ophthalmol 1998, 125: 411 –415.
15. Autran B, Carcelain G, Li TS. et al. Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science 1997, 277: 112 –116.
16. Li TS, Tubiana R, Katlama C. et al. Long-lasting recovery in CD4 T-cell function and viral-load reduction after highly active antiretroviral therapy in advanced HIV-1 disease. Lancet 1998, 351: 1682 –1686.
17. Komanduri KV, Viswanathan MN, Wieder ED. et al. Restoration of cytomegalovirus-specific CD4+ T-lymphocyte responses after ganciclovir and highly active antiretroviral therapy in individuals infected with HIV-1. Nat Med 1998, 4: 953 –956.
18. Elliot B, Aromin I, Gold R. et al. 2. :5 year remission of AIDS-associated progressive multifocal leukoencephalopathy with combined antiretroviral therapy. Lancet 1997, 349: 850. 850.
19. Murphy M, Armstrong D, Sepkowitz KA. et al. Regression of AIDS-related Kaposi's sarcoma following treatment with an HIV-1 protease inhibitor. AIDS 1997, 11: 261 –262.
20. Conant MA, Opp KM, Poretz D, Mills RG. Reduction of Kaposi's sarcoma lesions following treatment of AIDS with ritonavir. AIDS 1997, 11: 1300 –1301.
21. Blum L, Pellet C, Abgalika F. et al. Complete remission of AIDS-related Kaposi's sarcoma associated with undetectable human herpesvirus-8 sequences during anti-HIV protease therapy. AIDS 1997, 11: 1653 –1655.
22. Goguel J, Katlama C, Sarfati C. et al. Remission of AIDS-associated intestinal microsporidiosis with highly active antiretroviral therapy. AIDS 1997, 11: 1658. 1658.
23. Carr A, Marriott D, Field A. et al. Treatment of HIV-1 associated microsporidiosis and cryptosporidiosis with combination antiretroviral therapy. Lancet 1998, 351: 256 –261.
24. Race EM, Adelson-Mitty J, Kriegel GR. et al. Focal mycobacterial lymphadenitis following initiation of protease-inhibitor therapy in patients with advanced HIV-1 disease. Lancet 1998, 351: 252 –255.
25. Torriani F, Durand FD, Karavellas M, et al. Evidence that HAART induced immune recovery vitritis in CMV retinitis patients is immune mediated. XIIth World AIDS Conference. Geneva, 1998. [Abstract 15/22240].
26. Karavellas MP, Lowder CY, Macdonald C. et al. Immune recovery vitritis associated with inactive cytomegalovirus retinitis: a new syndrome. Arch Ophthalmol 1998, 116: 169 –175.
27. Zegans ME, Walton RC, Holland GN. et al. Transient vitreous inflammatory reactions associated with combination antiretroviral therapy in patients with AIDS and cytomegalovirus retinitis. Am J Ophthalmol 1998, 125: 292 –300.
28. Carr A, Samaras K, Burton S. et al. A syndrome of peripheral lipodystrophy, hyperlipidaemia and insulin resistance in patients receiving HIV protease inhibitors. AIDS 1998, 12: F51 –F58.
29. Ledergerber B, Egger M, Opravil M. et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Lancet 1999, 353: 863 –868.

Cited By:

This article has been cited 5 time(s).

Current Opinion in Infectious Diseases
The role of observational studies in assessing the impact of antiviral therapies
Sabin, CA
Current Opinion in Infectious Diseases, 13(6): 631-635.

PDF (74)
Current Opinion in Infectious Diseases
When can cytomegalovirus prophylaxis and maintenance therapy be stopped in HIV disease?
Deayton, JR; Griffiths, PD
Current Opinion in Infectious Diseases, 13(6): 637-641.

PDF (71)
Immune Reconstitution Inflammatory Syndrome: Emergence of a Unique Syndrome During Highly Active Antiretroviral Therapy
Medicine, 81(3): 213-227.

PDF (1356)
Retina, 22(3): 268-277.

PDF (659)
HIV-Associated Retinopathy in the Haart Era
Retina, 25(5): 633-649.

PDF (1525)
Back to Top | Article Outline

cytomegalovirus,; retinitis,; natural history,; highly active antiretroviral therapy,; vitritis,; cystoid macular oedema

© 2000 Lippincott Williams & Wilkins, Inc.