Opportunistic Infections: Down and Out in London, Paris and New York (but not in Harare, Jo'berg or Kinshasa)

     David Alain Wohl, MD - Clinical Assistant Professor, The University of North Carolina

Perhaps the 9th Conference on Retroviruses and Opportunistic Infections should be renamed simply the '9th Conference on a Retrovirus'? The pithy moniker would not be completely inappropriate considering the minimal discussion regarding opportunistic infections (OI's) had at the 8th conference last week in Chicago. Over the last two years, talk of OI's has largely given way to 'co-pathogens' such as hepatits C and B which, strictly speaking, are not true OI's. For example, at a slide session dedicated to co-pathogens and OI's, viral hepatitis was the subject of 4 of the 7 presentations - one of the remaining OI talks reported on treatment for visceral leishmaniasis (not exactly a burning OI issue even south of the Rio Grande).

This, of course, reflects the epidemiology of OI's in the era of HAART. While we all know persons living with HIV infection in places where HAART is unavailable continue to be plagued by OI's without abatement, this conference has never really been about HIV in the developing world - that is left to the World AIDS Conference - and discussion of problems facing these people was largely absent in Chicago. In the resource-rich nations we continue to see low levels of OI's, mostly in those for whom HAART has failed or forsaken. A telling example was the presentation by Milos Opravil, who described the outcomes of patients in the Swiss HIV Cohort Study who initiated antiretroviral therapy at a CD4 cell count above 350/cu mm and matched controls beginning treatment a year later (Abstract LB6). Among the controls a significant number of AIDS-associated diseases were observed. A few presentations describing causes of hospitalizations and deaths among persons with HIV in the era of HAART point to the low level simmering of OI's. However, much of what was reported regarding OI's related to the impact of HAART such as immune reconstitution phenomena and experiences with OI treatment discontinuation.

CMV Recovery Uveitis and Reactivation
The incidence of CMV retinitis has dropped precipitously over the past six years, yet many of those who have benefited from HAART and survived CMV retinitis have experienced an immune mediated uveitis that in some cases has led to loss of sight. The etiology of this inflammatory condition has been unclear and attempts to treat it with systemic and/or intraocular steroids have met with limited success. Rachael Schrier and colleagues at University of California, San Diego (UCSD) examined a variety of immune parameters and the clinical histories of 18 patients with a history of CMV retinitis but off CMV therapy following response to HAART (Abstract 31). Of these, 14 patients had immune recovery uveitis. Among those with uveitis, systemic T-cell lymphoproliferative responses to CMV were greater than the responses observed in patients without uveitis. Levels of IL-12 but not IL-6 were higher in the eyes of uveitis patients compared to patients with active retinitis. Clinically, patients with uveitis tended to have had a shorter course of anti-CMV maintenance therapy following diagnosis of retinitis. This, and the immunologic pattern described has led the investigators to postulate that persistent CMV or CMV proteins remain in the eye even after immune recovery and act as a stimulus for inflammation. However, to date, culture and PCR of the eyes of those afflicted with uveitis has failed to demonstrate the presence of CMV.

A related poster compared the activity of CD4 cells dedicated to combat CMV in patients with active CMV end-organ disease and those with a history of CMV retinitis but free from disease despite discontinuation of anti-CMV therapy for at least 2 years (Abstract 554). Two assays were used to assess the function of these T-cells: a CMV-specific CD4 a proliferation assay and a cytokine expression assay employing flow cytometry. The idea here is to determine whether measurement of the functional ability of CD4 cells dedicated to respond to CMV identifies which patients may reactivate CMV. While the great majority of the patients with active retinitis did not exhibit responses and all of the CMV survivors did have a response detected by one or both assays during 6 months of study follow-up, responses in the latter group were inconsistent. Although some of these CMV survivors had negative responses by the assays, there were no instances of CMV reactivation.

The Viral Activation Transfusion Study (VATS), a National Heart, Lund and Blood Institute sponsored study of white cell depleted blood transfusion in patients with AIDS, collected blood specimens from over 500 patients for CMV and HIV viral loads. Correlation between CMV viral load and clinical outcomes was reported in a poster presentation (Abstract 555). Among the cohort, the median CD4 cell count was 15/cu mm. A quarter of the patients had a history of CMV end organ disease and 22% had evidence of CMV in their blood at baseline by PCR. It should be appreciated that most people with HIV infection have antibodies to CMV in their blood indicating past exposure.

Like other viruses in the family, CMV remains latent and does not activate and replicate to cause disease unless the immune system falters. When this happens, CMV itself may be detected in the blood. In this study, as in others, increasing CMV viral load was clearly associated with risk of CMV disease, such as retinitis. Patients with CMV viremia were over 5 times more likely to develop CMV disease than those who did not have viremia (RH=5.66 CI 5.3-9.71). Likewise, CMV viremia was associated with a drop in Karnofsky performance score indicating an effect on daily functioning. Initiation of HAART led to a decline in CMV viremia but not at once. There was an average 90 day delay in decline in CMV virus levels in the blood following the start of HAART.

These data add to the mounting evidence that CMV in the blood is not a good thing. HAART, however, even in the absence of CMV therapy can drop CMV blood levels and the risk of CMV mediated threats. The AIDS Clinical Trials Group has initiated a study to test whether giving the oral anti-CMV drug valganciclovir to patients with detectable CMV viral load and who have not benefited from HAART will reduce the risk of CMV disease or death. For information contact the ACTG site nearest you:
(http://aactg.s-3.com/pub/reports/unitprin.htm).

Can Cryptococcal Meningitis Therapy be Stopped in HAART Responders?
The safety of discontinuation of CMV maintenance therapy and prophylaxis for PCP and MAI have been demonstrated in several published papers. However, whether the same holds true for suppressive therapy of cryptococcus is not known. According to an Italian study of patients with a history of cryptococcal meningitis who responded to HAART and had their anti-fungals stopped, the answer is 'Si!' (Abstract 546). A total of 46 patients with cryptococcal meningitis were studied. Of these 24 discontinued their anti-fungals a median of 25 months from the diagnosis of their meningitis, while the other 22 continued cryptococcal therapy. The mean CD4 cell count at the time of anti-fungal cessation was 80/cu mm and the average viral load was 2.55 log10 copies/mL. No episodes of recurrent cryptococcal meningitis occurred in either group. Serum cryptococcal antigen was detected in 5% (one patient) who stopped fungal therapy compared to 62% of those who maintained therapy. This observational, non-randomized study seems to indicate that cryptococcus could be added to the heap of OI's for which response to HAART confers protection.

In summary, opportunistic infections are fading as a major concern among HIV infected persons and their clinicians although for those not benefiting from HAART these diseases continue to be a threat. It is evident that HAART is the best prophylactic against development or recurrence of OI's and despite the high failure rate of combination therapies a clear upswing in the incidence of OI's has not been observed.