Article Text


Cytomegalovirus anterior uveitis: outcome of treatment
  1. S-P Chee1,2,3,
  2. A Jap1,4
  1. 1Ocular Inflammation and Immunology Service, Singapore National Eye Centre, Singapore, Singapore
  2. 2Department of Ophthalmology, Yong Loo Lin School of Medicine National University Health System, Singapore, Singapore
  3. 3Singapore Eye Research Institute, Singapore, Singapore
  4. 4Division of Ophthalmology, Changi General Hospital, Simei, Singapore, Singapore
  1. Correspondence to Soon-Phaik Chee, Singapore National Eye Centre, 11 Third Hospital Avenue, Singapore 168751, Singapore; chee.soon.phaik{at}


Aim To determine the outcome of antiviral treatment of cytomegalovirus (CMV) anterior uveitis.

Methods A retrospective review of patients from Singapore National Eye Centre with CMV anterior uveitis diagnosed by aqueous polymerase chain reaction. Ganciclovir treatment consisted of systemic, topical, intravitreal injections or intravitreal implant. The main outcome measure was resolution of anterior chamber inflammation.

Results 72 eyes of 70 patients were positive for CMV DNA. 35 eyes were treated (23 eyes with acute recurrent anterior uveitis and 12 eyes with chronic anterior uveitis). Eyes that did not respond or recurred with one treatment may receive another course of treatment. There were 47 treatment episodes, 36 (76.6%) of which resulted in a response. However, there were 27 (75.0%) episodes of recurrences after stopping treatment. Systemic and intravitreal ganciclovir and ganciclovir implant had good response rates but also had very high recurrence rates. Ganciclovir gel had moderate response rates, but its recurrence rates were also lower than those of the other modalities.

Conclusions Ganciclovir gel had lower recurrence rates than the systemic ganciclovir and the implant and should be considered as an option for treatment of CMV anterior uveitis.

  • Cytomegalovirus anterior uveitis
  • valganciclovir
  • polymerase chain reaction
  • anterior chamber
  • infection
  • inflammation
  • treatment medical

Statistics from

Cytomegalovirus (CMV) anterior segment infection in the immunocompetent usually presents as an acute recurrent or chronic anterior uveitis associated with ocular hypertension or corneal endotheliitis.1 2 In an earlier study, we found that 52.2% of eyes with acute recurrent iritis (presumed Posner–Schlossman syndrome) and 41.7% of eyes with chronic anterior uveitis (which we had previously termed presumed Fuchs uveitis syndrome because they had features that were similar to this clinical entity) were CMV positive.3

To date, 110 eyes of 106 non–human immunodeficiency virus (HIV)–positive patients with CMV anterior segment infection have been described,1 3–18 of which 27 eyes of 24 patients had endotheliitis, 57 eyes (57 patients) had acute recurrent anterior uveitis and the remaining had chronic anterior uveitis. All the 83 anterior uveitis eyes and 24 of the endotheliitis eyes had ocular hypertension. Hitherto, there have been no studies that focus specifically on the outcomes of treating these eyes with antiviral agents. The results of treatment have been derived from case reports or small case series. The main mode of antiviral therapy in these eyes has been systemic ganciclovir or valganciclovir.1 2 5–8 12 Some authors have also used intravitreal injections of ganciclovir10 or ganciclovir gel2 with variable outcomes. In many cases, the inflammation resolved with therapy, only to recur on cessation of treatment.2 5 7 8 In this study, we describe the response of CMV anterior uveitis to four modes of ganciclovir therapy.

Methods and materials

We retrospectively reviewed the case records of all patients with CMV anterior uveitis seen in the Singapore National Eye Centre uveitis clinic from 1 January 2002 to 31 August 2008. Data were collected on patient demographics; best-corrected visual acuity, intraocular pressure (IOP), the clinical presentation and mode and outcome of antiviral therapy. The institutional review board approved this study.

All the patients with anterior uveitis associated with ocular hypertension presenting to the uveitis clinic after the year 2004 routinely had their aqueous sampled as part of the uveitis workup protocol. The taps were performed after informed consent, only when anterior chamber activity was present, before starting any anti-inflammatory agents. The aqueous was tested for CMV DNA by nested and/or real-time polymerase chain reaction.2 19 20 The nested polymerase chain reaction also screens for herpes simplex virus, varicella zoster virus and Toxoplasma gondii.2 The patients who were positive for CMV DNA subsequently had their sera tested for HIV antibodies, CMV antigen and antibodies (immunoglobulins G (IgG) and M (IgM)).

All the CMV-positive patients were offered antiviral therapy and were reviewed by an infectious diseases physician at the Singapore General Hospital, with a view to treatment with systemic antiviral drugs. The first choice of treatment was systemic ganciclovir, as this was then the criterion standard. If the patient could not afford the systemic medication or had contraindications to systemic therapy, the other alternatives were intravitreal injections of ganciclovir or topical ganciclovir. The ganciclovir implant was another option for those whose conditions recurred after successful systemic therapy but wished to minimise adverse effects. The patients who refused antiviral treatment were treated with 0.5% ketorolac tromethamine and glaucoma medications and were followed up as indicated.

Initially, systemic ganciclovir was administered intravenously at 5 mg/kg of body weight twice a day for 6 weeks, followed by 1 g oral ganciclovir three times a day for another 6 weeks. When oral valganciclovir subsequently became available in Singapore, the patients on systemic therapy received it at 900 mg twice a day for 6 weeks, followed by 450 mg twice a day for another 6 weeks. Full blood count and serum creatinine were monitored every 2 weeks during the entire course of the systemic antiviral treatment.

The ganciclovir implant (Vitrasert; Bausch & Lomb Inc, Claremont, California, USA) contains 4.5 mg of ganciclovir in a non-biodegradable drug delivery system and is sutured into place at the pars plana. It releases ganciclovir over 5 to 8 months, but for the purpose of analysis, the duration of its action is taken as 8 months.

Ganciclovir at 2 mg/0.1 ml was administered weekly for 12 weeks. Ganciclovir gel (0.15% Virgan gel; Laboratories Thea, France) was applied four times a day for a minimum of 3 months.

Eyes on systemic antiviral therapy also received 0.12% prednisolone acetate twice a day. The eyes that were treated with intravitreal injections or ganciclovir 0.15% gel received either ketorolac tromethamine 0.5% four times a day or prednisolone acetate 0.12% twice a day. The patients with an implant were treated with 1% prednisolone acetate four times a day during the immediate postoperative period. In all the patients, the anti-inflammatory agents were tapered off as the inflammation resolved. They also received glaucoma medications as required.

Response to therapy in eyes with chronic anterior uveitis was defined as the clinical resolution of anterior chamber cells and keratic precipitates (KPs) and good IOP control with or without glaucoma medications while on treatment. In eyes with acute recurrent anterior uveitis, response was defined as absence of recurrence of inflammation and normalisation of IOP without glaucoma medications other than that given during the acute episode while on treatment. Failure of therapy in eyes with chronic anterior uveitis was defined as persistence of anterior chamber cells and KPs and/or elevated IOP despite glaucoma medications during the treatment period. Failure of therapy in eyes with acute recurrent anterior uveitis was defined as recurrence of anterior chamber cells and KPs with elevated IOP during the treatment period.

The patients were reviewed at weekly intervals for the first month and at monthly intervals thereafter, with more frequent reviews as necessary. Their best-corrected visual acuity, IOP and anterior chamber inflammation were assessed at each visit by one of the authors (SPC). The infectious diseases physician also reviewed them regularly.

Each eye may have received more than one mode of therapy. When the first treatment failed or relapses occurred, they were offered a repeat of the first treatment or an alternative mode of therapy if cost or adverse effects were a consideration. Statistical analysis was performed using the Mann–Whitney test for quantitative variables and χ2 test for binomial data on SPSS V.13, and statistical significance was set at p<0.05.


Seventy and two eyes of 70 patients were positive for CMV DNA. They included 50 eyes of 49 patients with acute recurrent anterior uveitis, 21 eyes of 20 patients with chronic anterior uveitis and one eye of a patient with sectoral iris atrophy. All were seropositive for CMV IgG and seronegative for CMV IgM, CMV antigen and HIV.

Thirty and three patients (35 eyes, 57%) agreed to an antiviral therapy after discussion of the risks and benefits of treatment with the infectious diseases physician. These included 23 eyes of 22 patients with acute recurrent anterior uveitis and 12 eyes of 11 patients with chronic anterior uveitis. Twenty and seven patients had completed the treatment course and were included in subsequent analysis, with a total of 47 treatment episodes. Of the 19 treatment episodes with systemic ganciclovir, only the first four patients received intravenous ganciclovir. All the later treatment episodes consisted of oral valganciclovir. The other eight patients were still on treatment at the end of the study. The patient with sectoral iris atrophy refused treatment and was quiet 7 months post tap. The median duration of the follow-up after stopping the treatment was 27.7 months (range 4.9–54.8 months). None of the patients had any serious adverse events, and none of the patients receiving systemic ganciclovir needed to stop treatment.

Thirty and six (76.6%) of 47 treatment episodes resulted in a response to antiviral therapy, but there were 27 (75.0%) episodes of recurrences after stopping the treatment (table 1).

Table 1

Outcome of various modalities of ganciclovir therapy in eyes with CMV-positive anterior uveitis

Among the 27 treatment episodes in eyes with acute recurrent anterior uveitis, 21 (77.8%) resulted in a response, but 16 (76%) recurred subsequently. Fifteen (75.0%) of the 20 treatment episodes in eyes with chronic anterior uveitis showed a response to treatment, but 11 (73.0%) had recurrent inflammation. The systemic ganciclovir and the ganciclovir implant had good response rates, but they also had high recurrence rates after cessation of therapy, and the recurrence rate was just as high with a second course of the treatment. Intravitreal injections had good response rates, but the recurrence rate was high. No patient had a second course of intravitreal injections. Although ganciclovir gel had moderate response rates, it also had low recurrence rates. The response to ganciclovir gel was unpredictable. Some eyes that did not initially respond responded to a subsequent course and vice versa.

Acute recurrent anterior uveitis

Twenty and seven (54%) patients chose not to receive antiviral treatment. These untreated eyes were less likely to have glaucoma and cataracts than the treated eyes and also had a shorter follow-up period post tap (table 2).

Table 2

Comparison of clinical features of acute recurrent anterior uveitis eyes in patients who chose (treated eyes) or refused antiviral treatment (untreated eyes)

Three (27%) eyes had no further attacks with a follow-up period of 27.1 to 40.8 months. Eight eyes continued to have recurrent attacks with a follow-up post tap of 1.3 to 44.5 months. The remaining 16 eyes did not have further episodes of iritis with a follow-up period post tap of <1 year.

Chronic anterior uveitis

Nine (42.9%) patients did not wish to receive antiviral medications. The untreated eyes were more likely to have good visual acuity than the treated eyes but were similar in their likelihood of having cataract and glaucoma (table 3).

Table 3

Comparison of clinical features of chronic anterior uveitis eyes in patients who chose (treated eyes) or refused antiviral treatment (untreated eyes)

Five of the untreated eyes had at least 1 year follow-up post tap, and all had 6/12 or better visual acuity at this time. Four of these five eyes had only mild anterior chamber cells and good IOP control during the follow-up period. The fifth eye had significant inflammation with elevated IOP despite maximal medical therapy and required glaucoma filtration surgery 2 months after the tap, with subsequent good IOP control and resolution of inflammation.


The prevalence of CMV infection in our eyes with anterior uveitis associated with ocular hypertension is higher than in the West. This could be because of the apparently higher seroprevalence of CMV in Asian countries (87% to 100%), as compared with the West (51.5% to 54.4%).21–24 In fact, of the 106 patients with CMV anterior segment infection reported thus far, only 18 were from Western countries.1 3–18

All the CMV-positive eyes were offered antiviral therapy, as they were at risk of visual loss from glaucomatous damage with repeated attacks in the case of acute recurrent disease or refractory glaucoma in chronic disease. It was anticipated that with successful control of the CMV infection, recurrences could thus be prevented. Not surprisingly, because antiviral treatment has its own risks, patients with less severe disease declined treatment. Systemic ganciclovir had been previously used in a small number of cases with some success in controlling CMV anterior uveitis; hence, it was the first option offered to the patients. The intravitreal mode of delivery was also initially offered to patients who had cost constraints or contraindications to systemic therapy. However, owing to high recurrence rates, it was discontinued after the first seven patients. The ganciclovir implant was an alternative mode of drug delivery for patients who had recurrences after previous positive response to systemic treatment but wished to avoid the adverse effects of systemic medications.

We found that CMV anterior uveitis responded well to systemic and implant ganciclovir therapy, but the recurrence rates after termination of the therapy were also high. The overall recurrence rate (75.0%) after antiviral therapy was not very different from the outcome in eyes that did not receive antiviral treatment, where the overall recurrence rate was 76.4%. Among the chronic anterior uveitis eyes that did not receive ganciclovir, four of five had only mild inflammation and good IOP control at 1 year post tap. The recurrence rate in the acute recurrent anterior uveitis eyes that did not receive ganciclovir was 72.7%. The failure of ganciclovir to eliminate the CMV infection may be partly due to the fact that ganciclovir is virostatic and not virucidal. Furthermore, the phenomenon of ocular immune privilege may play a role in preventing total eradication of the infection.

Our results and those of other authors suggest that a much longer duration of therapy may be required in the treatment of CMV anterior uveitis (table 4). Although only two-thirds of the CMV anterior uveitis eyes responded to ganciclovir gel, its recurrence rates were also low at 57.1% in eyes with acute recurrent anterior uveitis and 25.0% in eyes with chronic anterior uveitis. Animal models25 have shown that the mean aqueous concentration of ganciclovir after one application of 0.2% ganciclovir gel was 0.39 mg/l. The mean levels in the iris (1162 ng/g) though were higher than that in the aqueous.25 Because CMV has been found in the iris and ciliary body26 in HIV-positive patients, this may account for the effectiveness of ganciclovir gel in the treatment of CMV anterior uveitis despite its relatively low aqueous levels as compared with the mean inhibitory dose for treatment of CMV retinitis, which is 0.25–1.22 mg/l. We also found that the patients' responses varied from one episode of treatment to another. This could be, in part, because of lack of compliance despite the fact that we had deliberately simplified the dose to a four-times-a-day dosage regime. This suggests perhaps that if patients could comply with the recommended five-times-a-day dosage regime, the outcome may be more favourable and consistent. Ganciclovir gel may thus have a potential role as long-term maintenance therapy for the patients who had responded to a short course of systemic ganciclovir, as it is has minimal adverse effects and is much more affordable.

Table 4

Literature review of outcomes of treatment of CMV anterior uveitis in non–HIV-positive patients

On the other hand, a prolonged course of systemic ganciclovir has a greater potential for morbidity and would be too costly for most patients. Hence, we used a relatively short duration of systemic ganciclovir in this cohort of non–HIV-positive patients, as our aim had been to reduce the viral load with the anticipation that their own innate immunity would subsequently be able to suppress the residual infection. Based on a similar rationale, we used a once-a-week dosage of intravitreal ganciclovir injections rather than the recommended twice-weekly dosage. This may have contributed to the apparently worse outcome of intravitreal administration as compared with systemic ganciclovir. With a single intravitreal injection of 2 mg of ganciclovir, the mean vitreous concentration of ganciclovir is 23.4 mg/l at 72 h, and by 7 days, it is 0.6 mg/l.27 The main route for clearance of intravitreal ganciclovir is via the retina, and it has been observed that only 1% of it is eliminated via the aqueous humour.28 Hence, the amount of ganciclovir in the aqueous at 72 h would be only 0.23 mg/l and by 7 days would be only 0.006 mg/l. The mean vitreous level of ganciclovir with the implant is 4.1 mg/l,29 giving an estimated level of 0.04 mg/l in the aqueous. The aqueous level of ganciclovir after systemic administration has been reported to be 0.4 times30 that of the concurrent plasma level (8 mg/l),31 resulting in an aqueous concentration of approximately 3.2 mg/l. However, as was seen with the gel, the iris levels of ganciclovir after these modalities of treatment may be higher than those in the aqueous because of its greater affinity for cellular tissue. The levels would also be more consistent with the implant and systemic routes; hence, they could have a higher likelihood of achieving a response than a once-a-week dosage of intravitreal injections.

The concurrent use of anti-inflammatory agents with anti-infective medication is a well-accepted concept in the management of infectious uveitis. In this study, the anti-inflammatory agents were administered at very low doses and were tapered off as the inflammation resolved, whereas the antivirals were continued for a longer period as described above. Furthermore, the patients who chose not to receive antiviral treatment were treated with ketorolac tromethamine 0.5%. Hence, the use of these anti-inflammatory agents was unlikely to be a confounding factor.

This is a retrospective non-randomised study where the treated eyes were more likely to have severe disease and some of the treatment groups were small. Hence, the outcomes of this study should be interpreted with caution and validated by other trials. Nevertheless, to date, this is the only study that reviewed the outcome of the various options in the treatment of CMV anterior uveitis.

In view of the high prevalence of CMV infection in eyes with anterior uveitis associated with ocular hypertension, these eyes should have their aqueous analysed for CMV during periods of activity. In particular, eyes with features resembling Posner Schlossman, older males presenting with Fuchs uveitis syndrome-like chronic anterior uveitis and nodular endothelial lesions or eyes whose clinical manifestations are atypical of the varicella zoster virus/herpes simplex virus (including a failure to respond to maximal treatment with acyclovir) should be investigated for CMV.3 Patients with CMV-positive chronic anterior uveitis can be offered ganciclovir gel, but they need to be counselled that they may need long-term treatment. If CMV diagnostic tools are not available or the patient declines antiviral therapy, topical non-steroidal anti-inflammatory agents may be considered rather than steroids. If these eyes that are not tested do not respond to the non-steroidal anti-inflammatory agents, they may be given a trial of ganciclovir gel.

In conclusion, although CMV anterior uveitis generally responded well to ganciclovir, recurrence rates were high regardless of treatment regimen. Furthermore, the cost and the potential adverse effects of systemic ganciclovir and the ganciclovir implant limit their use on a long-term basis. On the other hand, eyes treated with ganciclovir gel had low recurrence rates. Ganciclovir gel also has the least issues in cost and potential for adverse reactions for long-term use. Hence, it is a treatment modality that should be considered in future prospective randomised therapeutic trials. Until such data are available, our approach for antiviral therapy is to offer ganciclovir gel as the first treatment option. For patients who fail to respond or who show frequent recurrence while on topical treatment and have progressive visual loss, we would then offer them a 3-month course of oral valganciclovir after informed discussion of its risks and benefits. If they continue to relapse, with progressive visual loss, and have no contraindications, they would be offered a much longer course of oral valganciclovir. Alternatively, if the patient chooses local therapy, higher twice-weekly dosages of intravitreal ganciclovir may be used as a therapeutic trial to predict if the patient would be a suitable candidate for the ganciclovir implant.


View Abstract


  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Singapore Eye Research Institute institutional review board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.