Case Reports

Top Case Reports Discussion Author & Article Info References

The demographic characteristics, clinical presentations, and outcomes of patients are shown in Table 1.

All patients presented with acute-onset eye pain and diminished visual acuity; ophthalmologic findings included injected conjunctiva, keratitic precipitates, and cells in the anterior chamber. Symptoms resolved with topical therapy. In patients 1 and 3, rifabutin was continued and a second episode of uveitis occurred within 2 months in the opposite eye. In patients 3 and 4, arthralgias developed near the time of uveitis and resolved when rifabutin therapy was discontinued.

Rifabutin levels were measured in stored serum specimens at 2, 4, 6, 8, and 10 months Table 1 [8]. Steady state levels of rifabutin (77 to 353 ng/mL) and the equipotent metabolite LM565 (9.3 to 40.5 ng/mL) did not increase over time.

Discussion

Top Case Reports Discussion Author & Article Info References

Anterior uveitis, a rare ocular finding in patients infected with HIV, has recently been recognized as a toxicity of rifabutin administration [4-7]. In a treatment trial involving patients with the acquired immunodeficiency syndrome and disseminated M. avium complex infection, 38% of patients receiving a combination of rifabutin (600 mg daily), clarithromycin, and ethambutol developed uveitis [5]. We report the occurrence of uveitis in four patients participating in an M. avium complex and fungal prophylaxis trial. Two patients received 300 mg of rifabutin daily in combination with fluconazole, and two patients remained blinded to study therapy but were presumed to be receiving rifabutin because of the orange discoloration of urine that coincided with the initiation of M. avium complex prophylaxis. Because all four patients had no evidence of any other diseases commonly associated with uveitis and no recurrence after discontinuation of prophylaxis, we attribute uveitis in these cases to rifabutin toxicity.

In contrast to previously described patients, our patients developed anterior uveitis after at least 7 months of rifabutin therapy. The putative relation of rifabutin to uveitis may be overlooked because of the delay in clinical symptoms after drug initiation and because uveitis can be successfully treated in the face of ongoing rifabutin therapy. To illustrate this point, rifabutin was continued after the onset of uveitis in three of our patients. When patients 1 and 3 developed a second episode in the opposite eye, the relation of rifabutin to this toxicity was recognized, the drug was discontinued, and no further episodes occurred.

After initiation of topical prednisolone and atropine, our patients had rapid improvement in visual acuity and eye pain and resolution of clinical findings within 1 month. The rate of recurrence with the continuation of rifabutin therapy is not known, although our experience suggests it may be high. Whether dose reduction would prevent recurrence is also unclear, but the use of lower doses of rifabutin for M. avium complex prophylaxis has not been evaluated.

The rifabutin serum levels measured in patients 1 and 2 are similar to those seen during a controlled phase I study assessing the interaction between fluconazole and rifabutin [9]. All four of our patients were receiving fluconazole. Although the higher levels of rifabutin achieved in the presence of fluconazole may potentiate the protective effect of rifabutin against M. avium complex [10], this commonly administered drug combination may increase the risk for uveitis. Sequential serum rifabutin and LM565 levels in our two patients did not increase over time, suggesting that the late development of this toxicity cannot be predicted by serum levels. Although rifabutin did not accumulate in the serum, the possibility of deposition in other intracellular compartments cannot be excluded.

Uveitis was not reported in two trials of rifabutin prophylaxis (300 mg/day), trials that involved more than 1000 patients [1]; however, only 566 of these patients were randomly assigned to rifabutin, fluconazole use was not universal, and the average duration of follow-up in these studies (231 days and 190 days) may not have been long enough to allow for detection of uveitis. Our four cases occurred among 365 study patients followed for at least 7 months, two thirds (n = 240) of whom were randomly assigned to rifabutin, suggesting that the incidence of this toxicity is between 1% and 2%.

These cases show that uveitis is a treatable complication of rifabutin when administered in patients infected with HIV. The morbidity associated with this toxicity, along with the likelihood of recurrence, argues for the prompt discontinuation of rifabutin as prophylactic therapy if uveitis develops.