Opportunistic Infections in the Era of HAART - A Report from the 4th Conference on Retroviruses
HIV RNA as guide to risk
Protease inhibitors (ritonavir, indinavir and nelfinavir) in combination with reverse transcriptase inhibitors (commonly referred to as "HAART," for highly active antiretroviral therapy) have been shown to dramatically decrease HIV replication and increase short-term survival in people with AIDS. A significant amount of data reported at the 4th Conference on Retroviruses and Opportunistic Infections earlier this year documented that HAART also has a profound effect on a myriad of AIDS-related opportunistic infections as TAG's "OI guy," Michael Marco, explains in the following report.
This year's national retrovirus conference featured several presentations which furthered our knowledge in three active areas of opportunistic infection research: 1) the use of HIV viral load to determine those at risk for developing opportunistic infections; 2) the elevation of HIV expression and immune destruction which results from an opportunistic infection; and 3) the changing natural history of opportunistic infections after initiating HAART. Much of the opportunistic infection data generated at the conference did not address specific treatments for the various opportunistic pathogens, but the effect HAART has on certain opportunistic infections. It is important to note that much of the data discussed were from small case series and non-randomized, uncontrolled clinical trials. Likewise, many of the treatment results were on persons with fewer than 6 months of follow-up.
HIV Viral Load as a Prognostic Factor
For many years we have known that a person's CD4 count is predictive of developing certain opportunistic infections (i.e., the majority of all cytomegalovirus (CMV) retinitis cases occur when a person's CD4 count is below 50). Now, however, HIV viral load - and changes in it after initiating antiretroviral therapy - might further help us to identify those at risk for developing specific opportunistic infections. Williams, Swindells and Currier from the AIDS Clinical Trials Group (ACTG) presented Data Analysis Concept Sheet (DACS) 071, a retrospective data analysis of 813 participants from four separate ACTG antiviral studies (pre-protease inhibitors: ACTG 116A, ACTG 116B/117, ACTG 175, ACTG 241) which looked at the predictive value of plasma HIV RNA and CD4 cells on the development of CMV, Pneumocystis carinii pneumonia (PCP) and Mycobacterium avium complex (MAC). DACS 071 documented that both baseline and post-treatment HIV RNA levels - as well as CD4 cell count - were strong predictors of these three opportunistic infections. Persons with baseline HIV RNA over 100,000 copies/ml had twice the risk of PCP and five to six times the risk of CMV and MAC compared to those with less than 100,000 copies/ml. The relative risk for PCP, CMV and MAC was, 2.29, 5.64, and 4.74, respectively. Likewise, persons with less than 75 CD4 cells had four to six times the risk of PCP, CMV and MAC. Alarmingly, persons with over 100,000 copies/ml of virus and below 75 CD4 cells had approximately 28 times the risk of MAC.
Viral Load As Indicator of OI Risk
|HIV RNA >100,000 copies /ml|
|CD4 <75 cells/mm3|
|Source: DACS 071, Williams, Swindells, Currier et al. ACTG 1997|
At eight weeks, a decrease in plasma HIV RNA of 0.5 log significantly reduced the risk of developing both CMV and PCP by approximately 70%. Any decrease in plasma RNA was shown to significantly reduce the risk of developing MAC. An increase of just 50 CD4 cells reduced the risk of developing all three opportunistic infections by 30-35%. After 24 weeks on antiretroviral therapy, this sustained decrease of 0.5 log in plasma HIV RNA further reduced the risk of developing CMV and MAC by approximately 85% and PCP by 57%.
These viral load data - albeit from older, non-protease inhibitor studies - may aid us in identifying specific thresholds or cutoff values for targeting prophylaxis regimens in individuals at highest risk for developing an opportunistic infection. And, most importantly, data from the DACS study suggests that active antiretroviral therapy (in this case mostly permutations of AZT, ddI and ddC) can have a profound protective effect in preventing these three opportunistic infections. Using approved opportunistic infection prophylaxes (e.g., TMP-SMX (Bactrim) for PCP and a macrolide for MAC) might not be enough if a person's HIV viral load is skyrocketing.
Chaisson and colleagues presented data which further illustrated the fact that opportunistic infections increase one's risk of death. In their analysis of 2,081 HIV-positive individuals with a mean follow-up of 30 months, the development of PCP, CMV, MAC, esophageal candidiasis, Kaposi's sarcoma, non-Hodgkin's lymphoma, progressive multifocal leukoencephalopathy (PML), dementia, wasting syndrome, toxoplasmosis and cryptosporidiosis was found to be independently associated with death while cryptococcal meningitis and herpes zoster were not. Moreover, the development of PCP, CMV, MAC and toxoplasmosis was associated with an increased risk of death regardless of CD4 count (p< 0.001 adjusted for CD4). For MAC, CMV, PCP, and toxoplasmosis, the relative hazard of death was 2.56, 1.63, 1.29, and 1.85, respectively.
The fact that these opportunistic infections increase the risk of death regardless of CD4 count could be explained by the fact that the development of an opportunistic infection probably increases HIV expression and immune damage by causing immune activation. Thus, an opportunistic infection might not simply be an annoying infection that warrants treatment, but actively affecting the natural history of HIV disease and resulting in significantly shorter survival. This alone is cause for initiating effective opportunistic infection prophylaxis in individuals who warrant them. Havlir and colleagues documented that patients infected with MAC had higher RNA viral loads than those not infected with MAC. Cases of disseminated MAC (dMAC) and controls (those without) were matched for baseline CD4 count, prior antiretroviral therapy, MAC prophylaxis regimens, and the length of follow-up. The baseline HIV RNA levels were found to be higher for those with dMAC (4.8 logs) than the controls (4.65 logs, p= 0.08). With the development of disease, levels increased 0.14 log for those with dMAC and 0.04 for the controls (p= 0.11).
Cooper and colleagues reported on a group of five HIV-infected children with dMAC who had elevated HIV RNA at the diagnosis. Four of the five children had approximately a one log drop in their HIV RNA within two month of initiating three or four-drug anti-MAC therapy. Bush and colleagues reported similar findings from their retrospective analysis of ten patients whose viral load was monitored before, at time of diagnosis, and after the resolution of PCP. Seven of the patients were antiretroviral naïve and three continued antiretroviral monotherapy during their course of PCP. The medium serum HIV RNA prior to diagnosis of PCP (median time before onset= 81 days) was 113,850 copies/ml, compared with 231,450 copies/ml at the time of PCP diagnosis (p= 0.03). Nine of these ten patients had marked elevations of their HIV RNA upon developing PCP - five of whom experienced rises to three-times that of baseline values. Seven of the ten patients had a decrease in their HIV RNA upon resolution of PCP (median HIV RNA 198,500 copies/ml).
Orenstein and Wahl biopsied lymph nodes of individuals with and without opportunistic infections to look for co-expression of HIV. By using in situ hybridization, they found that unprecedented levels of HIV production were evident in the tissues of those with active opportunistic infections. Moreover, Orenstein found that the pathogens - namely PCP and MAC - were localized in macrophages and not lymphocytes. This elevation in HIV RNA seen with the development of PCP and MAC (and possibly other opportunistic infections we do not have data on) seriously calls into question the routine discontinuation or interruption of antiretroviral therapy when a patient develops an opportunistic infection. This is often done to simplify the management of the toxicities which can occur with taking multiple concomitant medications. From these data, however, it appears that maximizing antiretroviral therapy during the duration of the opportunistic infection is necessary in order to counter such elevations in HIV expression.
Changing the Natural History of opportunistic infections with HAART
Two posters were presented on case series of patients with microsporidiosis and cryptosporidiosis who were administered HAART. In the first case series from Carr and colleagues, all of the five patients with cryptosporidiosis and seven patients with microsporidiosis (all antibiotic-resistant) who were administered HAART had a remission of their diarrhea within 12 weeks. Moreover, only one person was still observed to have parasites in his stools and only one person relapsed. In the second poster from Benhamou and colleagues, 27 patients with chronic cryptosporidiosis and microsporidiosis were administered HAART. In all but two patients (the two with microsporidiosis), no identifiable parasites were observed upon examination of stools. Two additional posters documented the resolution of a single patient's PML after the initiation of HAART. Both patients - one using indinavir and the other ritonavir - were said to have had a dramatic improvement in their physical functions and MRI findings.
These results seen with HAART are not completely surprising: Flanigan and colleagues showed in 1992 that patients whose CD4 counts went above 180 while receiving AZT were able to overcome cryptosporidiosis within one month. This suggests that improved immune function (a decrease in viral load and an increase in CD4 cells) can effectively overcome and kill these parasites. Hopefully, these results will be corroborated when data sets from Phase III HAART studies become available.
While HAART seems to have a profound effect on the resolution of cryptosporidiosis and PML as well as on the prevention of other OIs, it might not benefit people with AIDS in preventing CMV retinitis. Jacobson and colleagues presented data on 5 patients with CD4 counts over 200 who developed CMV retinitis just 4 to 8 weeks after initiating HAART. Four to 24 weeks before initiating HAART, all five patients had CD4 counts 85. This rise in CD4 cells obviously did not help these individuals immunologically fight off the CMV; it is almost unheard of for an HIV-infected person to develop CMV with over 100 CD4 cells, let alone 200. It is possible, Jacobson speculated, that these patients were CMV DNA PCR positive and that the CMV virus had already seeded the eye before they started HAART. Once the CMV is established within the sanctuary of the eye, it is possible that only effective anti-CMV would be capable of halting its progression.
Restoration of Immune Function with HAART
Finally, the restoration of immune function with HAART has been reported to cause a relapse of sulfa allergy. Race and colleagues presented data on four individuals who entered the hospital with TMP-SMX-related fever just 7 to 21 days after starting indinavir. All four individuals had previously required dose reduction or desensitization to trimethoprim-sulfamethoxazole (TMP-SMX) because of hypersensitivity reactions. Upon discontinuation of TMP-SMX, all fevers resolved, and the four patients were able to continue their indinavir with their TMP-SMX replaced with dapsone.
HIV viral load monitoring and the advent of the protease inhibitors promise not only to change our approach to treating opportunistic infections, but the complete clinical management of HIV-infected people. Admittedly, additional data are needed to validate HIV RNA's usefulness as a guide to the treatment and prophylaxis of AIDS-related opportunistic infections. Long-term data from people on HAART, in particular, will help us to understand whether these changes in the natural history of opportunistic infections will be merely transient or more enduring.