icon-folder.gif   Conference Reports for NATAP  
  10th Conference on Retroviruses and Opportunistic Infections
Boston, Mass, Feb 10-14, 2003
Back grey_arrow_rt.gif
Opportunustic Infections Report from Retrovirus
Written for NATAP by David Margolis, MD, University of Texas Sothwestern Medical Center, VA Medical Center
  Due to the success of antiretroviral therapy, the incidence of opportunistic infections (OIs) has decreased markedly in the last several years. Currently Ois are most often seen in HIV+ individuals not under care. Although it has been said that we should consider removing the "OI" from the conference's title, the 10th CROI had several interesting presentations related to these still-problematic infections.
Absolute CD4 Count May Be More Reliable Than CD4% in Predicting OI Two presentations culled from observations of the urban cohort of HIV+ patients who attend Hopkins' Moore clinic were of note. Gebo et al. (abstr. #792) studied the power of a single absolute CD4 cell count determination or a single CD4 cell % value to predict the occurrence of an OI within the next 2 months. Guidelines suggest the initiation of OI prophylaxis or antiretroviral therapy when these measures fall below certain thresholds. In general, the absolute numerical value of a CD4 cell count can vary more from day to day, and from test to test, than does the CD4 % value. CD4 counts are affected by time of day, intercurrent illness, stress, alcohol intake, and other factors. Gebo et al. studied 15,736 CD4-CD4% pairs from 2,184 pts found in their database after 1995, and 608 OIs documented in their database over the 6 months the CD4 cell test. The median period of observation after a CD4 test was actually 70 days. CD4 counts were classified into groups: < 50, 50-150, 151-250, and 250-350 cells/mm3 for absolute counts, and < 7, 7-14, 15-21, and > 21 for CD4 %. Statistical methods were used to estimate increase in risk of OI (incidence rate ratio or IRR) after adjusting for the factors of sex, race, HIV risk factor, and use of HAART.
For patients with an absolute CD4 count < 50, the risk of an OI was 24.2 times greater (95% confidence interval: 16.0 to 36.5) than patients with a CD4 count between 250-350. For absolute CD4 50-150, the risk was 6.2 times greater (4.0, 9.5), and 2.7 times greater (95% C.I. 1.7 to 4.4) if CD4 151-250. The comparable increase in risk for CD4% < 7% was 14.4 (95% C.I. 9.3 to 22.6), for 7-14% was 3.7 (2.4, 5.9), and 1.9 (1.1, 3.1) for 15-21%, compared to > 21%.
However, absolute CD4 count was a much stronger predictor of risk, and when the predictive power of absolute CD4 count was statistically taken into account, the CD4% provided no additional information. The authors concluded that CD4 count alone may be used in making treatment decisions. However, this finding should be carefully viewed in context. Absolute CD4 counts are generally provided together with CD4%, therefore there is no additional cost in acquiring additional information from the CD4%. Although in a study such as this, CD4 counts provide more information within the entire population under study, a clinician should be wary about initiating antiretroviral therapy based on a single absolute CD4 cell count. Of course, CD4 trends, both absolute and %, should be examined. Decisions about initiating OI prophylaxis might be made less conservatively, as prophylaxis can be stopped without long-term consequences.
Opportunistic Infections in IDUs vs Non-IDUs
Moore presented another analysis from the Hopkins database (abstr. #793) examining the risk of OI in HIV-infected injecting drug users (IDUs) under care in their clinic. Due to the obvious risk of viral hepatitis in the IDU population, this infection was excluded from the comparative analysis, and the occurrence of all other Ois was compared from the time patients started initial HAART.
1077 IDU with CD4 counts < 350 cells/mm3 when they started HAART were compared with 1229 non-IDUs. Due to the availability of ADAP, HAART therapy was reported to be equally accessible to both groups, however the rate of durable undetectable HIV-1 RNA after 2 years of therapy significantly less in the IDU populations (31% IDU vs 47% non-IDU; p < 0.01).
As would be expected, rates of OI declined in both groups between 1996 and 2002. When adjusted for differences in sex, race, initial CD4 count, HIV-1 RNA levels, and type of therapy (PI, PI/r, or NNRTI) OI incidence was equal in the two groups in 1996, not significantly different in 1999 (although numerically less frequent in non-IDU), but significantly more likely in IDUs than non-IDUs after 2000.
There were 23.9 and 19.8 OIs per 100 person-years in 2000 and 2001-2002 (to 7/02) respectively, in the IDU cohort, as compared with 18.4 and 13.1 OIs per 100 patient/years in the non-IDU cohort. This statistically significant difference is likely attributable to suboptimal outcomes in IDUs prescribed HAART. Factors such as poorer adherence, and higher rates of HAART-related toxicity may underlie this outcome.
Ocular Syphilis Risk in MSM
Phadungchai and colleagues from Georgetown (abstr. #794) reported a micro-epidemic of ocular syphilis in treated patients from their clinic in the past year. Although HIV changes the course of syphilis and its response to therapy, ocular syphilis is still rare. Between 1997 and 2002, a total of 11.4% of patients screened (40/350) were diagnosed with syphilis. However, 15 of these diagnoses occurred in the past year (15/111 patients screened). Of these, 40% (6/15) had neurosyphilis and 20% (3/15) had ocular syphilis. This would extrapolate to a point prevalence of (3/40), compared to the prevalence of 0.6% previously estimated. All the cases of ocular syphilis occurred in male homosexuals with a mean age of 42 yrs, mean CD4 of 495 cells/mm3 (range 388-594 cells/mm3), receiving HAART with well-controlled plasma viremia. All responded to a 21-day course of penicillin G 24 million units IV daily. This report suggests a concurrent epidemic of risky sexual behavior under the cover of HAART in male homosexuals in the Dupont Circle area, and suggests that residual immunological abnormalities in this treated cohort may contribute to ocular manifestations upon syphilis infection.
Toxoplasmosis Proliferative Response & HAART
Miro and coworkers form Spain investigated the emergence of immune responses to toxoplasmosis in patients diagnosed with toxoplasma encephalitis subsequently treated with HAART (abstr. #796). 16 pts with IgG antibodies against Toxoplasma gondii (median [IQR] titer of 208 [131-300] UI/mL) were sampled at 3 mos (T3), 8 cases at 6 mos (T6), 13 pts at 9-12 mos (T12), and 6 cases at 16-18 mos (T18). Median CD4+ T-cell counts rose on therapy to 301 (191-369) cells/ÁL over 18 months of therapy. Plasma HIV RNA was < 200 copies/mL in 83% at 18 months.
Laboratory tests of the immune response to toxoplasma increased in these patients, as expected, during HAART. Following 18 months of HAART, proliferation of lymphocytes increased 20-fold when patients' cells were exposed toxoplasma antigens in the laboratory. 83% of subjects studied at 18 months were said to have a positive proliferative response. The drawback of this study is its relatively small sample size, and the fact that not all patients were studied at 18 months. However, the findings suggest, as expected, that suppressive therapy following toxoplasma encephalitis might be discontinued after a substantial period of successful HAART and significant rise of CD4 cell count. The findings of this study give us a timeframe within which to consider studying the discontinuation of such prophylaxis.
Pneumocystis Carinii Resistance Mutations
Kazanjian and colleagues (abstr. 800) reported the incidence of mutations in the dyhydropteroate synthase (DHPS) gene of pneumocystis carinii (PC). This is one of the parasite's genes that is the target of trimethoprim sulfamethoxazole (Bactrim) therapy. The DHPS gene was studied in PC recovered from 145 US patients with PCP between 1983-2000, and 15 patients in China with PCP. There was a significant increase in mutations in the DHPS gene in all AIDS pts from 0% before 1994, to 16% in 1994, to 70% in 2000 (p < 0.01). In patients in China, mutations were present in 1 of the 15 pts (7%). The authors suggested that these and other findings might be explained by the person-to-person transmission of PC. This might be true, however as HIV infection has not existed in the population in China for as long as it has in the US, DHPS mutations may simply appear over time in PC in HIV-infected patients. Of greater significance, there is yet no evidence that mutant strains of P carinii have yet been related to the widespread failure of Bactrim therapy or prophylaxis.
West Nile Virus in HIV-Infected Individuals
West Nile Virus (WNV), a mosquito-borne disease, first appeared in the US in New York a few years ago. WNV has since spread across wide areas of the US, and can cause encephalitis. More than 3000 cases of WNV have been reported since 1999. Del Rio and coworkers reported the 2nd and 3rd cases of WNV complicated by encephalitis in HIV+ patients. By retrospective chart at their medical center in Atlanta they found 3 other cases (HIV-negative) in which the diagnosis of WNV was entertained by the treating physician, and in whom antibody tests showing recent exposure to WNV were positive. One of the HIV+ patients died of encephalitis, while the other recovered without neurological complications. Both HIV+ patients had CD4 counts below 100/uL. The number of white blood cells in the cerebrospinal fluid, a clinical sign that sometimes gives a clue as to the cause of a meningitis or encephalitis, was only slightly elevated (12 cells/mm3) in one patient, and markedly elevated (310 cells/mm3) in the other. These two cases show us that WNV may happen be an entity to consider in HIV+ patients, but more cases will need to be studied before we can know if WNV behaves differently or occurs more frequently in AIDS.