CONFERENCE ON RETROVIRUSES AND OPPORTUNISTIC INFECTIONS OVERVIEW
The 14th Conference on Retroviruses and Opportunistic Infections was held in Los Angeles, Calif, on February 25 to 28, overlapping with the Academy Awards weekend. Fifty-seven percent of abstract submissions were accepted (totaling 921) and, along with webcasts and pedcasts, are accessible through the conference Web site www.retroconference.org/2007). The following is a review of highlights of the meeting attended by over 3800 leading researchers and clinicians from around the world; the material presented is referenced by author and abstract number.
Although it has been well established that simian immunodeficiency virus-Pan troglodytes troglodytes (SIVcpzPtt)-in west central African chimpanzees was the source of human cross-species spread of human immunodeficiency virus (HIV) (HIV-1 group M [pandemic], HIV-1 group N [nonpandemic in Cameroon], and HIV-1 group O [west central Africa]), recently, SIVgor in gorillas has been linked to a new HIV-1 group O-like virus also in Cameroon (M. Peeters, abstract 4). In addition, the simian foamy virus found in 75% to 100% of nonhuman adult primates with high concentrations in saliva has caused infection in 1% to 4% of humans occupationally exposed in zoos and primate centers in North America and humans engaged in hunting activities (A. Gessain et al, abstract 5). Emerging classification schemata include primate T-lymphotrophic virus 1 that encompasses simian T-lymphotrophic virus 1 and human T-lymphotrophic virus 1 (severe lymphoproliferative diseases in 1% to 2% of infected people) and primate T-lymphotrophic virus 2 that contains simian T-lymphotrophic virus 2 and human T-lymphotrophic virus 2 (W. Switzer, abstract 6). The association of HIV-2 with long-term nonprogressors or so-called elite controllers is in part explained by preserved HIV-specific T-helper responses characterized by interleukin 2 production and proliferation and strong responses toward a highly conserved region of the HIV-2 gag protein (S. Rowland-Jones, abstract 7).
There seems to be a more rapid disease progression with HIV subtypes D and A/D and circulating recombinant forms versus clade A (J. Baeten et al, abstract 68; and N. Kiwanuka et al, abstract 307). The cause of this difference is unclear, but receptor tropism may be a contributing factor; there is a higher prevalence of CXCR4 HIV-1 coreceptors (CXCR) 4 in subtype D versus A and a lower prevalence in subtype C (D. Kuritzkes, abstract 108).
Dr Kevin De Cock of the World Health Organization in the opening workshop for new investigators and trainees discussed changing patterns of US and Global HIV epidemiology. He used the World Bank life expectancy data to emphasize the extreme heterogeneity of the epidemic, stating that prevalence is greater than incidence, prevalence can increase rapidly but decrease slowly, prevalence increases most rapidly with high incidence (eg, in Eastern Europe), and that prevalence decreases most rapidly with high mortality (eg, acquired immune deficiency syndrome [AIDS] deaths now exceed new infections in Kenya).
The European epidemic is being driven by regional trends in economics, drug trafficking patterns, migration, sexual behavior (including sex tourism), public health care practices, sexually transmitted disease (STD) rates, and availability of antiretrovirals (A. Johnson, abstract 54).
Disturbingly, after almost a decade of approximately 40,000 new infections annually in the United States, there was more than 10% increase in 2005 (the last year for which numbers are available). In addition, death rates in the United States are 2 times higher than those in Western Europe and 10 times higher than those in the United Kingdom (H. Jaffe, abstract 63). Rates of HIV infection are 8 times higher among African Americans versus whites. Abstract 55 used modeling to assess the incidence and future prevalence of HIV among men who have sex with men (MSM) in the United States; with an estimated incidence of 1.9% and assuming an 8% prevalence in MSM aged 20 years with no mortality due to HIV, more than 50% of all MSM in the United States could be HIV infected by the age of 60 years. Among African American MSM with an incidence of 4%, almost 80% are projected to be HIV positive by age of 60 years.
The latest statistics, tables, and slides can be found at the following Web sites: www.worldbank.org, www.who.int, and www.cdc.gov (recently redesigned to improve usability and performance including a new search engine).
One workshop focused on the emerging field of genomics and host-HIV interactions. A practical application is the association of HLA B 5701 with the abacavir hypersensitivity reaction (HSR) that occurs in more than 5% of patients. This acute (within the first 6 weeks with a median onset of 9 days) systemic (fever, rash, and gastrointestinal and/or respiratory symptoms) adverse reaction is most common in whites (African Americans and men having half the risk). All patients with HSR have HLA B 5701 (plus or minus other haplotypes) that can be ordered commercially for between $100 and $200. Only 4% of patients with HLA B 5701 will not experience HSR. Some argued that obtaining this as a screening test was similar to ordering a G6PD level.
Globally, only 24% of eligible HIV-infected patients now receive antiretroviral therapy (ART; 1.65 of 6.8 million as per the World Health Organization as of June 2006). In resource-limited countries, primarily Africa and Asia, there are few regimens available (the most common being the fixed combination of stavudine, lamivudine, and nevirapine [NVP] given BID) leading to fewer opportunities to switch therapies secondary to either intolerance or lack of efficacy (M. Egger, abstract 62). However, several studies have shown that virological and immunologic responses are similar to those in industrialized countries. Tragically, most patients in resource-limited areas start ART late or very late, and mortality rate is higher, especially with low CD4 lymphocyte counts. The 4-year mortality is 15% in sub-Saharan Africa compared with 5% in North America and Europe. Tuberculosis remains the most common fatal opportunistic infection worldwide.
A review of new HIV infections in South Carolina from 2001 to 2005 showed missed opportunities for HIV testing (K. Weis et al, abstract 957). Among 4221 new infections, 73% had 1 or more health care visits before their first HIV-positive test, mostly in emergency departments. Only 20% had diagnoses that would possibly prompt an HIV test, but 42% developed AIDS within 1 year of being diagnosed. The study emphasized the importance of routine HIV screening in all health care settings, especially emergency departments, as has been recommended by the Centers for Disease Control and Prevention (CDC) (MMWR 2006;55:RR-14). A CDC study suggested that annual testing of MSM would reduce the proportion of undiagnosed HIV-positive MSM from 48% to 14% (P. Denning, abstract 956). In a New York City hospital, the use of rapid oral HIV testing found that 7% of patients tested HIV positive on a general medicine ward, 2% in the ambulatory care center, and 3% in the ED (R. Smerd et al, abstract 958). Ninety-nine percent of patients with negative rapid oral tests in the San Francisco City STD Clinic agreed to HIV RNA testing, leading to 1.1% being positive (J. Klausner et al, abstract 953).
CLINICAL TRIALS: EFFICACY/DURABILITY OF RESPONSE
Several trials looked at established agents in novel ways. ACTG A5073 (D. Mildvan et al, abstract 138) assessed once-daily versus twice-daily lopinavir/ritonavir (LPVr; softgel capsule Kaletra) plus a once-daily 2 nucleoside/nucleotide backbone in treatment-naive patients. Overall, there was no difference with regard to the probability of a sustained virological response (SVR). However, twice-daily LPVr had a statistical advantage for those with high baseline HIV RNA (≥100,000 copies/mL). In a study named GESIDA (J. Berenguer et al, abstract 504), the understudied once-daily combination of didanosine, lamivudine, and efavirenz (EFV) (given with food in this study) provided similar antiviral efficacy to that of Combivir plus EFV at 24 weeks. Abstract 503 (D. Rey et al for the DAUFIN study) reported that, in ART treatment-naive patients, the once-daily regimen of lamivudine, tenofovir (TDF), and NVP was associated with an unexpectedly high rate of early nonresponse with a high incidence of K65R and M184V mutations. Abstracts 513 (F. Pulido et al) and 514 (R. Campo et al) dealt with simplification to LPVr monotherapy; in each study, suboptimal adherence was a major risk factor for losing virological suppression. There are 2 formulations of a ritonavir tablet that have met bioequivalence criteria (Y. Cai et al, abstract 52LB); we hope to have one of these tablets available in the near future without the refrigeration requirement of the standard capsule.
Session 115 (W. Wheeler et al, abstract 648; S. Sirivichayakul et al, abstract 649; S. Eshleman et al, abstract 650; A. Low et al, abstract 651; N. Zetola et al, abstract 652; M. Para et al, abstract 653; L. Drumright et al, abstract 654; G. Somi et al, abstract 655; C. de Mendoza et al, abstract 656; and K. Nambiar et al, abstract 657) was devoted to transmitted drug resistance (TDR). The prevalence of TDR is now 11% to 15% of newly infected patients and 7% to 11% of newly diagnosed patients. In 1 cohort of 793 patients, 45% failed to reach undetectability over 6 years. Nonnucleoside reverse transcriptase inhibitor (NNRTI) TDR is associated with a higher set point compared with nucleoside reverse transcriptase inhibitor (NRTI) or protease inhibitor (PI) TDR. Unlike acquired resistant viruses, transmitted drug-resistant viruses must "back mutate" to get to wild type, and the mean time to first detection of a wild-type/drug-resistant mixture is 2 years. Transmitted drug-resistant viruses have replication capacities similar to wild-type viruses. Abstract 639 evaluated to what extent conventional genotyping underestimates the presence of transmitted resistance. Using real-time polymerase chain reaction assays for 6 common mutations (M41L, K70R, Y181C, M184V, and L90M), they found that patients with these polymerase chain reaction-detected mutations were 11 times more likely to experience treatment failure compared with those who had no resistance with conventional methods. These data further support the need for baseline resistance testing before initiation of antiretroviral therapy.
A Swiss study (V. von Wyl et al, abstract 667) compared the development of resistance with first-line boosted PI and NNRTI-based regimens. They concluded that, while boosted PI-containing regimens have similar potency as NNRTI-based regimens, boosted PI regimens lead to less resistance in cases of virological failure.
This year, 3 major classes of antiretrovirals headlined the discussion about new agents, and each drug now has a generic name (one with a trade name):
A few of the major presentations on drugs from each class that are nearest to being available for clinical use are reviewed below.
Abstract 144LB (A. Pozniak et al)-Formerly TMC278, rilpivirine is a new NNRTI that was tested at 3 different doses against EFV in ART-naive patients (with either Truvada or Combivir as the backbone). The antiviral activity was comparable at all 3 doses, and there were lower rates of central nervous system toxicity and rash in the rilpivirine arm compared with the EFV arm. It remains to be seen whether rilpivirine might challenge or even replace EFV in treatment-naive patients. Efavirenz was approved 9 years ago and has never been beaten in any clinical trial.
Abstracts 104aLB and 104bLB (M. Nelson et al and J. Lalezari et al)-Twenty-four-week data from 2 large phase 2b/3 ongoing randomized controlled trials evaluating the CCR5 inhibitor maraviroc were presented as late breakers. These studies were entitled "Efficacy and Safety of Maraviroc plus Optimized Background Therapy In Viremic, ART-Experienced Patients Infected With CCR5-Tropic HIV-1" or "MOTIVATE" 1 and 2. Triple class-experienced patients were screened with the Monogram Biosciences tropism assay (Trofile) to determine the presence or absence of the CCR5 or CXCR4 entry coreceptor. The roughly 50% with no detectable X4 (CXCR4 utilizing) virus were randomized 1:2:2 to receive an optimized background regimen alone or an optimized background regimen with maraviroc 150 mg, given either once or twice daily. The results were similar in both MOTIVATE 1 and 2, with virological and immunologic responses being superior in both arms receiving maraviroc compared with controls; in addition, there was no difference in safety between the maraviroc and control arms. Although not evident in these studies, long-term data are needed to examine the emergence of malignancies (such as occurred in the vicriviroc studies) and the concern about driving a switch from CCR5 to the potentially more pathogenic X4 virus.
Abstracts 105aLB and 105bLB (D. Cooper et al and R. Steigbigel et al)-Formerly MK-0518, raltegravir is an integrase strand transfer inhibitor. Two studies named BENCHMRK 1 and 2 were presented ("Blocking Integrase in Treatment Experienced Patients With a Novel Compound Against HIV:MeRcK"). The studies were similarly designed ongoing phase 3 randomized, double-blind, placebo-controlled trials; BENCHMRK 1 was carried out in Europe, Asia/Pacific, and Peru, whereas BENCHMRK 2 was done in North and South America. Both studies randomized patients with triple class-resistant HIV viruses in a ratio of 2:1 to optimized background plus either raltegravir 400 mg twice daily or placebo. Results were similar in both studies. Combining the data, at 16 weeks, the results showed that 61% to 62% of the raltegravir patients and 33% to 36% of the placebo patients were suppressed to less than 50 copies/mL; when raltegravir was used with either enfuvirtide or darunavir, 90% of patients were suppressed. Response was increased with lower baseline RNA, higher baseline CD4, and additional active agents in the regimen. There were no differences in adverse effects across the arms. Raltegravir failure was generally associated with 1 of 2 genetic pathways (N155H or Q148K/R/H).
METABOLIC AND CARDIOVASCULAR EFFECTS
ACTG 5142 is a randomized, open-label, prospective trial comparing an NRTI-sparing arm (EFV plus LPVr), an NNRTI-sparing arm (LPVr), and a PI-sparing arm (EFV) in treatment-naive patients; it is the first head-to-head comparison of EFV and LPVr. Earlier reported efficacy results favored EFV, but LPVr had a resistance and immune reconstitution advantage. At his meeting, Haubrich and colleagues reported on the comparative metabolic effects from the study (abstract 38). The NRTI-sparing regimen (EFV + LPVr) increased lipids significantly more than the EFV or LPV + 2 NRTI regimens. Triglyceride increases were also greater in LPVr compared with EFV + NRTI regimens, but cholesterol changes were not significantly different. Compared with EFV, LPVr had less lipoatrophy (defined as >20% reduction in extremity fat) when given with NRTI. The frequency of lipoatrophy was lowest in the NRTI-sparing and TDF-containing regimens; it was most frequent in the stavudine- and zidovudine (AZT)-containing regimens. Study 613 (D. Cameron et al, abstract 44LB) explored body composition changes with LPVr monotherapy (deintensification after 3 consecutive HIV RNA <50 copies/mL by week 24 on initial LPVr plus Combivir) versus EFV plus Combivir. The LPVr monotherapy arm was significantly associated with less lipoatrophy, greater limb fat gain, and increased triglyceride levels than the EFV arm; these findings are consistent with the observations above in ACTG 5142.
McGarth and colleagues conducted body composition assessments (metabolic and lipid parameters, lipodystrophy case scores, and imaging studies) in treatment-naive patients receiving atazanavir (ATV) with or without ritonavir (RTV) in combination with stavudine XR (a once-daily formulation that was Food and Drug Administration-approved but never marketed) and lamivudine (abstract 804). There were no clinically significant differences in body composition between the ATV and ATV/RTV regimens at 48 weeks based on computed tomography and dual-energy x-ray absorptiometry imaging and a validated questionnaire; thus, the addition of RTV to ATV does not seem to be associated with clinically significant changes in body composition at 48 weeks. Both regimens increased visceral, subcutaneous, and total adipose tissue, and neither regimen caused increases in lipoatrophy or lipodystrophy.
The SMART Study was a planned 8-year study in more than 5000 treatment-experienced patients. Patients were randomized to have CD4-guided drug conservation (DC arm with stopping ART with CD4 >350 cells/μL and resuming ART when CD4 fell to <250 cells/μL) or continuous antiretroviral therapy aimed at viral suppression (VS arm). Enrollment in the study was halted in January 2006 when the data and safety monitoring board found a significantly increased risk of disease progression (AIDS or death) in the DC arm compared with the VS arm. Abstract 41 (A. Phillips et al) reported on the retrospective analysis of cardiovascular disease (CVD) events in the SMART Study. A borderline significant excess risk of CVD was observed in DC compared with VS patients; however, increases in HIV RNA and decreases in CD4 cell counts associated with interruption of therapy were not associated with an increased risk of CVD. Results from this study suggest that ART should not be stopped because of perceived excessive risk of CVD.
The Northern California Kaiser Permanente group updated its data on HIV and coronary heart disease and myocardial infarction (MI). They estimated adjusted rates and relative risks of coronary heart disease and MI over 3 periods: pre-highly active antiretroviral therapy (HAART; 1994-1995), early HAART (1996-2000), and late HAART (2001-2006). Adjusted relative MI rates were higher among HIV-positive compared with HIV-negative subjects (D. Klein et al, abstract 807). The relative rate of MI among HIV-positive patients was highest during the early HAART period compared with the late HAART period.
Simple clinical interventions can often be critical in helping patients achieve successful adherence to antiretroviral therapy. Using simple measures of adherence and multivariate regression statistical techniques, investigators demonstrated that the complexity of antiretroviral regimens (doses per day and total daily pill burden) correlated with adherence and virological outcomes. In addition, the use of pill organizer boxes was similarly correlated with improved outcomes (A. Ammassari et al, abstract 522; and M. Petersen et al, abstract 524).
Undertaking to define patient populations at risk for self-induced drug holidays, CDC workers (J. Heffelfinger et al, abstract 529) used a multivariable logistic regression to evaluate 6129 patients treated with antiretroviral regimens during 2000 to 2004. Drug holidays were reported by 23% of 6129 patients. Multivariable logistic regression suggested that patients with illicit drug use, heavy alcohol use, MSM or injection drug use HIV acquisition risk, high pill burden, and prior poor adherence were more likely to take drug holidays.
Finally, several reports documented the excellent adherence among patients in developing countries as access to these medications increases in these settings (O. Busari et al, abstract 546; A. Gupta et al, abstract 547; and J. Awino et al, abstract 760).
While the potent antihyperlipidemia drug rosuvastatin undergoes different metabolic pathways than those usually thought to be affected by LVP/r, Colorado researchers were able to show a doubling of rosuvastatin at steady-state exposure when coadministered with LPV/r. Cautious use of this combination was advised because higher doses of rosuvastatin have been associated with increased rates of adverse effects (D. Hoody et al, abstract 564).
Complex multidirectional interactions may occur in the treatment of HIV infection and associated conditions. Histamine 2 receptor inhibitors may decrease bioavailability of ATV, which requires acidic pH for solubility and absorption, even when augmented with RTV. Conversely, TDF may affect ATV exposure by uncertain mechanisms. Varying famotidine dose and separation from an ATV/RTV/TDF regimen enabled evaluation of ATV exposure with each strategy. Famotidine 20 mg twice daily simultaneously administered with ATV, famotidine 40 mg every 24 hours separated by 12 hours, and famotidine 20 mg given 2 hours after ATV dosing resulted in ATV exposure similar to that without the H2 blocker. Decreased ATV exposure was noted when 40-mg famotidine was administered 2 hours after the ATV regimen (S. Agarwala et al, abstract 568).
Concurrent treatment of HIV and tuberculosis coinfection is frequently complicated by drug interaction, especially when rifampin is used. Two HIV treatment options were all but ruled out by 2 studies reported at this meeting. Atazanavir exposure was suboptimal even when given in doses as high as 400 mg twice a day with concurrent rifampin. Ritonavir boosting was not studied (E. Acosta et al, abstract 575). A second report involved NVP. Thirty patients on rifampin-containing tuberculosis regimens were randomized to antiretroviral regimens containing 400 mg or 600 mg of NVP daily. Nevirapine 400 mg/d was pharmacokinetically suboptimal, whereas NVP 600 mg daily was associated with a high rate of NVP hypersensitivity. Efficacy evaluation is ongoing (A. Avihingsanon et al, abstract 576).
PHARMACOKINETICS, THERAPEUTIC DRUG MONITORING
Two groups of investigators reported experience using serum concentrations to optimize the safety/toxicity of ribavirin (RBV) dosing in the treatment of hepatitis C coinfection. Marucco used a high-performance liquid chromatography method to measure steady-state trough concentrations. By logistic regression analysis, RBV trough was shown to be a predictor of early virological response and SVR for a group of patients infected with genotype 1 or 4 but was not significant for genotypes 2/3. Significant hemoglobin drop was seen more commonly in patients with C trough of greater than 2300 ng/mL (abstract 903). Similar correlations using high performance liquid chromatography with ultraviolet detection (HPLC-UV) of both serum and erythrocyte-bound RBV showed correlation with 4-week measures of efficacy and toxicity (S. Dominguez et al, abstract 893). Validation of the therapeutic drug monitoring concept in this setting and specific cutoff values for decision making await further similar research.
TDF AND RENAL TOXICITY
Associations of the antiretroviral nucleotide TDF with renal toxicity have been postulated for several years. Tenofovir is structurally similar to adefovir, which, when tested in high doses as an antiretroviral, exhibited nephrotoxic effects. Four presentations (R. Moore et al, abstract 832; H. Wai et al, abstract 833; C. Fux et al, abstract 834; and M. Goicoechea et al, abstract 835) reported efforts to better characterize possible renal effects of TDF. In retrospective cohort studies examining more than 2400 patients, both antiretroviral naive and experienced, all with more than 2 years of exposure and follow-up, 3 American and 1 Swiss team of investigators drew several conclusions. Tenofovir use over time seems to be associated with a modest but measurable risk of decline in renal function. Furthermore, concurrent use of low-dose RTV may increase this risk, despite an apparent lack of association with changes in TDF pharmacokinetics. Optimal management of various at-risk clinical scenarios awaits further investigation.
HIV AND HEPATITIS COINFECTION
Another report from the United Kingdom of acute hepatitis C virus (HCV) infection sexually transmitted among homosexuals was presented (M. Fisher et al, abstract 130). High-risk practices and concomitant sexually transmitted diseases were common among these newly HCV-infected subjects. This report adds to the growing body of evidence that HCV infection is now an emerging STD in this risk group that should lead to specific preventive interventions.
The consequences of acute HCV infection are very relevant because it often evolves into chronic infection, leading to liver disease that, in the presence of HIV coinfection, progresses faster. One of the reasons for this accelerated course of HCV-related liver disease among HIV-infected patients might be the increase in tumor necrosis factor α production (F. Yue et al, abstract 133). According to Canadian investigators, the proportion of anti-HIV- and anti-HCV-specific T cells producing this fibrogenic cytokine was significantly higher in liver biopsies of coinfected individuals. Of note, these authors did not find changes in the production of tumor necrosis factor α when HAART was also present.
Treatment of chronic HCV presents particular challenges in the HIV-HCV-coinfected patient. Another interaction between RBV and NRTIs has been recognized. In a substudy of the RIBAVIC trial, concomitant use of abacavir and RBV was found to be independently associated with decreased early virological response, defined as a decrease in the HCV RNA load of at least 2 log10 compared with baseline levels (F. Bani-Sadr et al, abstract 897). It can be hypothesized that because both compounds are guanosine derivatives, they may compete within the intracellular phosphorylation pathways. Another group of researchers reported that the use of lamivudine along with either stavudine or TDF as NRTI backbone was associated with SVR to HCV treatment (J. Mira et al, abstract 898). Further studies are needed to confirm these intriguing findings, given that multiple confounding factors could be present in these retrospective analyses.
For other reasons, AZT should be avoided when combination anti-HCV treatment is given. In a Spanish study involving 389 patients treated with pegylated interferon and 1000 to 1200 mg/d of RBV, AZT use was associated with greater decreases in hemoglobin levels at week 4 (M. Nunez et al, abstract 902). However, the doses of RBV were fairly well tolerated in this cohort, and this was reassuring of the safety of these doses in HIV-HCV-coinfected patients. Thus, severe anemia was infrequent and did not impact on SVR in this study.
Recipients of liver transplantation with recurrent HCV infection constitute an especially difficult-to-treat population. The Spanish group leading the transplantation in HIV-infected patients reported the results of pegylated interferon plus RBV treatment in 16 patients with active HCV replication after liver transplantation (J. Miro et al, abstract 890). Only 4 patients (25%) achieved SVR, and 6 others (37%) died as a result of graft loss due to the recurrence of HCV infection. Certainly, additional treatment strategies are needed to fight against this complication that has a very negative impact on the survival of HIV-HCV-infected patients receiving a liver transplant.
As for HBV-HIV coinfection, in a late-breaker presentation, researchers from Johns Hopkins claimed that entecavir, one of the Food and Drug Administration-approved drugs for the treatment of HBV, could select resistance mutations in the HIV genome (M. McMahon et al, abstract 136LB). They reported that the HIV plasma load decreased in 3 HBV/HIV-coinfected individuals exposed to entecavir in the absence of antiretroviral therapy and that the M184V mutation was present in the HIV genome of one of them. The authors also demonstrated that in vitro experiments further suggested that entecavir showed anti-HIV activity. These findings were received with surprise by the audience because this drug has been marked under the assumption that it has no activity against HIV. Further data are needed to clarify this issue because this compound is recommended for the treatment of HBV in HIV-infected individuals who do not require antiretroviral therapy by current guidelines.
The behavior strategies of HIV prevention (the ABCs of abstain, be faithful, condoms, and counseling) have not proven so successful as had been hoped. We also need biological strategies including immunization, exposure prophylaxis, microbicides, genital herpes suppression, and so on. Routine male circumcision could reduce a man's risk of HIV infection through heterosexual sex by 65%, according to the final data from the 2 National Institutes of Health-funded studies conducted in Kenya and Uganda (Lancet 2007;369:657-666). In 2 late-breaker presentations (R. Gray et al, abstract 155aLB; and M. Wawer et al, abstract 155bLB) from Rakai, Uganda, male circumcision was again shown to reduce HIV incidence and genital ulceration. There was no evidence of behavioral disinhibition, and surgical complications were comparable in HIV-positive and HIV-negative men.
A series of studies (H. Coovadia et al, abstract 13; and M. Sinkala et al, abstract 74LB) presented suggested that the benefits of breast-feeding in developing countries might outweigh the risk of vertical HIV transmission. In Africa, HIV-negative infants who were breast-fed by HIV-positive mothers from birth up to age of 6 months had increased rates of severe diarrhea that resulted in hospitalization or death compared with infants who were breast-fed for longer than 6 months. Another study found high rates of diarrhea and malnutrition among formula-fed infants in Botswana after a 2006 flood led to water contamination. The conclusion was to only consider not breast-feeding as an option in limited settings if it was acceptable, feasible, affordable, sustainable, and safe for the mother and the baby.
Cellulose sulfate (Ushercell; Polydex Pharmaceuticals) should be an excellent microbicide candidate. It is nontoxic, does not induce significant amounts of cytokines, and is active against sperm and numerous pathogens (HIV, HSV, HPV, Neisseria gonorrhoeae, Chlamydia trachomatis, etc), nontoxic to lactobacilli, nonabsorbable, and inexpensive (<$0.50 per dose). It performed well in phases 1 and 2 studies. Unfortunately, a phase 3 study in Africa and India (the CONRAD Trial-L. Van Damme et al, abstract 106LB) was halted in January 2007 because of an increased risk of HIV transmission. The search for the ideal microbicide continues.