Simon Collins, HIV i-Base

Hearing loss has been associated as a complication in HIV-positive people but it is unclear whether HIV is a direct factor or whether symptoms are more strongly correlated to risk factors reported in the general population. This will be increasingly important as the HIV population ages.

Researchers from Washington DC measured cochlear function in 334 men and 178 women from two of the earliest population cohorts established to look at differences between HIV-positive and HIV-negative patients (MACS and WIHS respectively), and related to this to social factors including noise exposure and HIV and treatment history.

The mean age was 54 years for the men (46% were HIV-positive), and 45 years for the women (77% were HIV-positive). People were excluded if they had hearing-impaired clinical symptoms or recent use of orotoxic medication. Approximately 20% of people in each of the HIV-positive and HIV negative groups self-reported exposure to occupational noise.

Cochlear function was measured by distortion product otoacoustic emission (DPOAE) testing which is a non-invasive procedure using two separate tones to stimulate the cochlea. Each ear was measured twice, with a third test if results were inconsistent and the number of non-responses added as an outcome variable (0-4).

In multivariate analyses, a 10-year increase in age [OR 2.78; 95%CI 2.07, 3.73], being male [OR 5.60; 95%CI 2.98, 10.49], and being non-black [OR 2.75; 95%CI 1.57, 4.83] were significantly associated with a higher number of non-responses (all p<0.001), but not HIV status [OR 1.20; 95%CI 0.7, 2.02; p =0.52 NS]. However, neither occupational or non-occupational noise exposure was associated with reduced function (p=0.33 and p=0.93, respectively).

Age, race, and gender remained significant risk factors for increasing non-responses in the HIV-positive model. However, none of the HIV-related factors including use of monotherapy, combination therapy, HAART use, 100-cell increase in peak CD8, HIV viral load, and 100-cell increase in nadir CD4 count came near approaching statistical significance (with p-values ranging from 0.2 to 0.7).

The researchers concluded that HIV status, combination therapy, nadir CD4 count, peak CD8 count, and HIV viral load did not significantly predict decreased cochlear function in this patient group.

Reference

Torre P et al. Cochlear function among Multicenter AIDS Cohort Study (MACS) and Women’s Interagency HIV Study (WIHS) participants. 16th IAS Conference on HIV Pathogenesis, Treatment and Prevention, 17–20 July 2011, Rome. Poster abstract TUPE138. Poster. (PDF)

An analysis from the D:A:D study, presented as an oral session, reported that HIV-positive people who stop smoking can expert similar direct health benefits to HIV-negative people.

Kathy Petoumenos from the University of New South Wales, Sydney, looked at rates of cardiovascular disease before and after stopping smoking in over 27,500 HIV-positive patients from Europe, the US and Australia, who had smoking status recorded in the prospective D:A:D cohort study.

The group looked at four endpoints: fatal and non-fatal myocardial infarction (MI), a broader definition of coronary heart disease (CHD), cardiovascular disease (CVD) combining CHD and stroke, and all cause mortality. Event rates were calculated for never smokers (n=8,920), ex-smokers at D:A:D study entry (n=6,265), current smokers (n=11,951), and smokers who stopped during D:A:D follow-up (n=8,197).

Current smokers were more likely to be male (77%), white (70%), infected through IV drug use (32%) and HCV-positive (34%), but CD4, viral load, BP, lipids and ARV-exposure were broadly similar between groups.

Incidence rate ratios (IRR) were determined adjusting for age, sex, cohort, calendar year, antiretroviral treatment, family history of CVD, diabetes (men and women), and time updated lipids and blood pressure assessments. Mortality endpoint also adjusted for HCV, HBV, mode of HIV exposure, ethnicity and incidence of CVD during follow-up.

Up to February 2008, there were 432 (MI), 600 (CHD), 746 (CVD) and 1902 (mortality events) endpoints. Crude event rates were 2.85, 3.96, 4.94, and 12.45 per 1000 person years respectively.

With never-smoked as the reference, increased risks rates were 1.73 and 3.40 for previous- and current-smokers respectively. Rate ratios for patients who had stopped smoking for <1, 1-2, 2-3 and >3 years follow-up since quitting, were 3.73, 3.00, 2.62 and 2.07 respectively, compared to never-smokers. This showed significant reductions within a year of stopping that continued to reduce over subsequent years. A similar pattern was seen for CHD and CVD, and although these were not significant at all timepoints, this is likely to be related to the lower number of events in some groups and the lower number of people who stopped smoking more than two years ago. The protective trend here is clear and important (see Table 1).

Although current smokers were at 28% higher risk of mortality, with no difference between never- and former-smokers, no clear reductions were seen during follow-up since stopping, with all confidence intervals crossing 1.0, even in a sub-group at higher risk of CVD-related mortality (in patients older than 50 years).

Table 1: Incidence rate ratios by baseline smoking status and time since quitting

* Non significant (CI crossed 1.0)

An explanation for the higher rates seen in the most recent (< 1 year) quitters was explained by the likelihood that a medical incident could have been the prompt needed to stop smoking. This group would therefore be at higher risk compared to current smokers (who would have been symptomatic). This was supported by the cause of mortality being more likely to be HIV/AIDS in the never smoked group with higher rates of non-AIDS malignancies seen in the previous and stopped groups.

The study has limitations in the amount and type of data that were collected on smoking (e.g. no start/stop dates or pack-year data). However, the significant reductions on CHD with each year after stopping smoking should support cessation programme for HIV-positive people, a greater percentage of whom smoke than the general population.

comment

This is the first time that the clinical benefits of stopping smoking has been reported in HIV-positive people and these findings should not be taken for granted.

Each year, HIV-positive people are advised on the importance of modification of lifestyle for ‘healthy options’ related to the complicated etiology of cardiovascular health and any study that can show tangible benefits is important.

This is particularly important given the higher rates of lung cancers reported in other studies. Keith Sigel and colleagues reported that HIV is an independent risk factor for lung cancer after adjusting for smoking (IRR 1.80; 95%CI 1.28 2.15). [2]

Edgar Simard from the US National Cancer Institute, reported a 3-fold observed incidence of lung cancer in HIV-positive patients within 3-5 years of an AIDS diagnosis compared to the general population (and increasing cumulative incidence). [3]

Meredith Shiels and colleagues reported that lung cancer was one of the cancers that was being diagnosed at an earlier age in HIV-positive compared to HIV-negative people, and that this 3-4 year difference was statistically significant after adjustment for multiple comparisons (p<0.001). [4]

References

Unless stated otherwise, all references are to the Programme and Abstracts of the 17th Conference on Retroviruses and Opportunistic Infections. 16-10 February 2010, San Francisco. All oral abstracts are available as webcasts.

http://www.retroconference.org

1. Petoumenos K et al. Rates of cardiovascular disease following smoking cessation in patients with HIV infection: results from the D:A:D study. 17th CROI, 2010. Oral abstract 124.

http://www.retroconference.org/2010/Abstracts/38289.htm

2. Sigel K et al. HIV infection is an independent risk factor for lung cancer. 17th CROI, 2010. Oral Abstract 30.

http://www.retroconference.org/2010/Abstracts/38365.htm

3. Simard E et al. Cancer incidence and cancer-attributable mortality among persons with AIDS in the US. 17th CROI, 2010. Oral abstract 27.

http://www.retroconference.org/2010/Abstracts/38173.htm

4. Shiels M et al. Do people with AIDS develop cancer at younger ages than the general population? 17th CROI, 2010. Poster 757.

http://www.retroconference.org/2010/Abstracts/37914.htm

Keith Sigel and colleagues presented an analysis from the US Veterans Ageing Cohort Study Virtual Cohort (VC) on the relationship between HIV and lung cancer. [1]

The advantages of this database included size, smoking status data from a related health survey and a matched HIV-negative control group, although this was an almost exclusively (98%) male study. Median age was 47 years and ethnicity was approximately 40% white, 40% black. 10% Hispanic and 10% other. Approximately 30% were current smokers, 15% recently quit (< 1 year), 25-50% distantly quit (> 1 year) and 20% of HIV-positive compared to 25% of HIV-negative had never smoked.

The analysis compared over 3,700 HIV-positive and nearly 10,000 HIV-negative patients (contributing 28,500 and 76,800 person-years of follow-up respectively).

Lung cancer was defined using International Classification of Diseases (ICD-9) codes and Incidence Rate Ratios (IRR) were adjusted for age, race, smoking exposure, and Chronic Obstructive Pulmonary Disease (COPD).

The overall incidence of lung cancer per 100 person years was 0.26 compared to 0.16 in the HIV-positive vs HIV-negative groups (unadjusted IRR 1.5, 95%CI 1.2–2.0). Results from the adjusted analysis are detailed in Table 1.

Table 1: Adjusted IRR for lung cancer multivariate model including all covariates

* Reference = white race; ** Reference = never smoked.

The authors concluded that the incidence of lung cancer was significantly increased in HIV-positive men in their group, even after adjusting for smoking exposure.

comment

This study reported slightly lower rates of increased risk of lung cancers in HIV-positive individuals compared to rates that were 2- to 6-fold higher in earlier studies, that also adjusted for smoking status. [2, 3, 4]

While the approximate 2-fold increased risk associated with HIV was significant and is important, the presenter emphasised the 10-fold higher risk for current smokers (that was halved to around 5-fold for former smokers who had quit more than one year earlier).

References

Unless stated otherwise, all references are to the Programme and Abstracts of the 17th Conference on Retroviruses and Opportunistic Infections. 16-10 February 2010, San Francisco. All oral abstracts are available as webcasts.

http://www.retroconference.org

1. Sigel K et al. HIV infection is an independent risk factor for lung cancer. 17th CROI, 2010. Oral Abstract 30.

http://www.retroconference.org/2010/Abstracts/38365.htm

2. Cadranel J et al. Lung cancer in HIV infected patients: facts, questions and challenges. Thorax 2006; 61:1000-1008.

http://thorax.bmj.com/content/61/11/1000.abstract

3. Engels EA et al. Elevated incidence of lung cancer among HIV-infected individuals. J Clin Oncol 2006; 24 (9); 1383-1388. (March 2006).

http://jco.ascopubs.org/cgi/content/abstract/24/9/1383

4. Kirk GD et al. HIV infection is associated with an increased risk for lung cancer, independent of smoking. Clin Inf Dis 2007; 45:103-110.

http://www.journals.uchicago.edu/doi/abs/10.1086/518606

Advances in the treatment of HIV infection with antiretroviral therapy have led to dramatic reductions in opportunistic infections and death.  However, late presentation of HIV remains a problem and is a significant contributory cause to death in HIV-positive persons in the UK.  Furthermore, a recent UK Health Protection Agency (HPA) analysis showed that of 46,700 patients with diagnosed HIV, 19% had CD4 counts <200 cells/mm3 and therefore remain at significant risk of opportunistic infection.

These guidelines have been drawn up to help physicians investigate and manage HIV-positive patients suspected of, or having an opportunistic infection (OI).  The early chapters consider the most common presentations of OI disease such as respiratory, gastrointestinal and neurological disease.  Then follow specific organisms such as Candida spp, herpes simplex virus and varicella zoster virus. Finally, special circumstances including pregnancy, the use of the intensive care unit, fever of undetermined origin and management of imported infections, are also addressed.

Each section contains information on the background, epidemiology, presentation, treatment and prophylaxis of OIs.  Further information on the role of antiretroviral therapy is also discussed (see below).

The Guidelines Writing Group is grateful for all comments, which will be reviewed before publication.

http://www.bhiva.org

Prompted by the concern that systemic inflammation may contribute to sleep apnea, Susheel Patil and colleagues from Johns Hopkins University presented an interesting analysis from the gently named SIESTA study (Study of Immune Effects on Sleep, (HIV) Treatment and Apnea).

The study looked at obstructive sleep apnea (OSA) and the relationship with inflammation markers (TNF-alpha soluble TNF-a receptors I and II and IL-6), in three groups of men from the MACS cohort: HIV-positive and not on HAART (n=41), HIVpositive and on HAART (n=58) and HIV negative (n=60). Severity of OSI was defined by the number of events per hour detected during a nocturnal sleep study: 5-15 = mild, 15-30 = moderate, and >30 = severe. Obesity is the strongest predictor of OSI, but OSI is also independently associated with hypertension, cardiovascular disease, stroke, diabetes mellitus and reduced quality of life.

OSI >15 was higher in the HIV- negative group (57%) compared to the HAART (41%) and no-HAART (44%) groups. However, the HIV-negative group had a significantly greater mean BMI (28.6 ±7.2kg/m2) and waist circumference (98.6 +16.9cms) compared to the HAART (25.5 ±4.5kg/m2 and 93.8 +11.5cm) and no-HAART (25.4 ±4.1 kg/m2 and 91.8 ±12.8) HIV-positive groups and a trend to greater trunk weight.

When looking at participants with normal BMI (<25 kg/m2) however, the relationship indicated a trend for higher prevalence in the no-HAART group: 25% HIV-negative (n=20), 24% on HAART (n=29) and 50% in the no HAART group (n=22); (p=0.1).

Median levels of all four inflammatory markers were higher in the HIV-positive men compared to the HIV-negative men, and were higher in the no-HAART group compared to the HAART group. Within the no-HAART group, men with moderate – severe OSA had higher levels of TNF-a and IL-6 compared to men with no or less severe OSI, although this difference was not observed between men in the other groups.

The study concluded that rates of OSI were high in HIV-positive men, even when BMI was normal, and that more severe symptoms was associated with systemic inflammation suggesting a different aetiology compared to men who are HIV-negative.

Reference:

Patil SP et al. Association between systemic inflammation and obstructive sleep apnea in men with or at risk for HIV infection from the Multicenter AIDS Cohort Study (MACS). 11th Intl Workshop on Adverse Drug Reactions. 26-28 October 2009, Philadelphia. Oral abstract O-25. Antiviral therapy 2009; 14 Suppl 2: A19.

A poster by Jan Gerstoft and colleagues from Copenhagen University Hospital looked at the interaction between changes in lung function in relation to smoking and HIV treatment.

Between October 2000 and November 2001, 63 HIV-positive patients had initial lung function assessed by a panel of tests (including forced expired volume, functional vital capacity, peak flow, residual volume [RV%] and total capacity and diffusing capacity/alveolar volume [DLCO/VA%]), with follow-up assessments a median of 4.5 years later (range 3.8-4.7 years).

Most participants (87%) were already on HAART at baseline for a median of about five years (range 16-79 months) with all but two on HAART at the follow-up visit (with 85% and 89% of these patients having viral load <100 copies/mL at each time point, respectively).

Some abnormal lung function parameters were present at baseline in both smoking (n=30) and non-smoking (n=33) participants, and some were further reduced in smokers.

Specifically, DLCO/VA% was decreased in both groups, with lung function compatible with early obstructive lung disease. At follow-up these levels normalised in non-smokers and improved in smokers to the baseline levels of the non-smoking group. However, results for residual volume, which returned to normal for non-smokers, increased further in the smoking group.

The researchers concluded that this study showed that HAART was beneficial for lung status and that HIV-related changes can reverse over time in non-smokers. However, smoking masks many of these potential benefits.

COMMENT

As most participants had already been on HAART for several years at baseline, and results were not divided by HAART use and viral load, the study did not quantify the extent and timeline of the benefits due to antiretroviral therapy. Nevertheless, the suggested positive impact of HAART on lung function is important and the results reinforce the importance of smoking cessation.

Ref:
Gerstoft J et al. Changes of lung function in an optimally treated HIV population: a 4.5 year follow up study. 49th ICAAC, 12-15 September 2009, San Francisco. Poster abstract H-1561.
http://www.posters2view.com/icaac/view.php?nu=H-1561

A poster at IAS 2009 by Madhi and colleagues reported the results of a double-blind, randomised, placebo-controlled trial to examine the efficacy of influenza vaccines in HIV-positive adults. Specifically, the study tested the seasonal trivalent subunit vaccine, which protects against three H1N1 and H3N2 influenza strains, two of which were isolated in 2006 and one in 2007. [1]

This was the first community-based randomised controlled trial of the trivalent subunit influenza vaccination in HIV-positive adults. A previous randomised controlled trial in the US on HIV-positive out-patients at a military health facility found that 10/47 patients who received placebo acquired laboratory confirmed influenza versus 0/55 patients who received this type of vaccine (p<0.001). [2]

A Cochrane Review of the vaccine in healthy adults has found this type of vaccine to be 30% effective (95%CI 17%-41%) against influenza-like illness, and 80% (95% CI 56% to 91%) effective against influenza when the vaccine matched the circulating strain and circulation was high, but decreased to 50% (95%CI 27%-65%) when it did not match the circulating strains. [3]

Participants were vaccinated prior to the 2008 influenza season in South Africa. Oropharyngeal swabs were taken from patients with influenza-like symptoms or respiratory illness. Culture and PCR tests were used to identify influenza strains. Only events 14 days post-vaccination were compared.

The number of HIV-positive people enrolled was 506. Of these, 101 were ART-naive (52 received vaccine, 49 received placebo) and 405 were on ART (203 received vaccine, 202 received placebo). Median age was 36. Female to male ratio was 5 to 1 in the vaccine arm and 6 to 1 in the placebo one. Median CD4 was 372 (IQR: 254-489) and 363 (IQR: 252-517) in the two arms respectively. Nine women were pregnant in the vaccine arm and four in the placebo one.

Over 90% of patients on HAART were virally suppressed in both arms. The median time on HAART at the time of randomisation was 23 months.

The percentage of people who developed influenza on the placebo arm was 5.3% using a passive surveillance method. [4] The rate of influenza illness was 0.06 per 100 person weeks in the vaccine arm and 0.25 per 100 person weeks in the placebo one. The vaccine efficacy was 75.4% (95%CI 14-93). The protective effect against the seasonal H1N1 strain was 73.5% (95%CI 4-93). There was one case of influenza B and no cases of H3N2 or untyped A.

The authors concluded that their findings support the use of the trivalent subunit influenza vaccine in HIV-positive adults.

COMMENT

This study supports vaccinating HIV-positive adults against seasonal influenza, as is routinely recommended in the UK. Because the vaccine is developed seasonally, its efficacy will change from year to year. The seasonal vaccine does not provide protection against the current H1N1 strain of swine flu.

At present there is no data on the effect influenza vaccines would have on reducing mortality in HIV-positive people as the contribution of HIV to influenza mortality is not well understood. Nevertheless, it is plausible that the reduced influenza cases conferred by the vaccine will reduce mortality as well.

References
1. Madhi S et al. Efficacy of influenza vaccine in HIV-infected (HIV+) adults: a double-blind, placebo randomised controlled trial in South Africa. 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention. 19-22 July 2009, Cape Town. Late breaker poster LBPEB04.
http://www.ias2009.org/pag/Abstracts.aspx?AID=3805
2. Tasker SA et al. Efficacy of influenza vaccination in HIV-infected persons: a randomised, double-blind, placebo-controlled trial. Ann Intern Med. 21 September 1999, 131, 430-433.
http://www.annals.org/cgi/content/abstract/131/6/430
3. Demicheli V. Vaccines for preventing influenza in healthy adults. Cochrane Review. 2006.
http://www.cochrane.org/reviews/en/ab001269.html
4. Personal communication with S Madhi.

Acyclovir studies at IAS 2009

Connie Celum, the principal investigator on the HPTN 039 trial, together with Jai Lingappa, the medical director, and other members of the Partners in Prevention HSV-HIV Transmission Team presented findings at IAS2009 on a counterpart trial to HPTN 039. This study, in HIV serodiscordant couples, looked at whether suppressing HSV-2 in people dually infected with HSV-2 and HIV infections could reduce HIV transmissions to their HIV-negative partners. [1]

HPTN 039 was a double-blind placebo controlled trial of standard doses of acyclovir for HSV-2 suppression (400 mg twice daily) to prevent HIV acquisition among over 3200 African women and MSM in Peru and the United States who were HIV-negative and HSV-2 positive. HTPN 039 found that acyclovir suppression did not reduce HIV incidence compared to placebo. Results were reported at CROI2008 and subsequently in the Lancet. [2]

Celum presented the main results of this new double-blind placebo controlled trial, conducted at 14 sites in sub-Saharan Africa, including South Africa, Kenya, Zambia, Botswana, Tanzania, Rwanda and Uganda.

Of 6544 heterosexual HIV discordant couples screened, 3408 were enrolled. The inclusion criteria for the coinfected partner included a CD4 cell count >250/mm³. The other partner could be either HSV-2 positive or negative, but had to be HIV-negative. HIV-positive participants could not be eligible for ART at trial entry according to their country guidelines.

The HIV-positive partners were randomised to receive either 400mg acyclovir twice-daily or placebo twice-daily. Couples were followed for a maximum of 24 months. Participants were provided with ART if they became eligible for it according to country guidelines.

The primary endpoint was HIV infection in the HIV-negative partners. The secondary endpoints included plasma and genital HIV viral load in the HIV-positive partners, and HIV disease progression. The trial was set up so that if 88 ‘linked’ HIV transmissions (i.e. the virus transmitted from the enrolled partner to the seroconverting partner was determined by molecular sequencing to be linked) were observed, the trial would have high statistical power (90%) to see a 50% reduction in HIV transmissions in the acyclovir arm.

Baseline characteristics of the group included:

• 67% of HIV-positive partners were female;
• 65% of volunteers were <35 years old;
• average partnership duration was five years;
• 90% were cohabiting;
• a median of five sex acts were reported in the month prior to baseline measurements:
• 29% reported unprotected sex;
• 22% of the HIV-positive partners reported genital ulcer disease (GUD) in the prior three months;
• 4% of HIV-positive and 7% of HIV-negative partners reported outside partners respectively;
• median CD4 count was 460 cells/mm³ and median plasma HIV viral load was 4.2 log.

Monthly follow-up visits included medication provision, pill count and adherence support and individual and couple HIV risk reduction counselling. Every three months, HIV-positive partners were examined for GUD and plasma viral load and HIV-negative partners were tested for HIV and given risk reduction counselling. CD4 cell counts were taken every six months.

Retention was high. At 24 months, 92% of HIV-positive and 84% of HIV-negative participants were still in follow-up. Adherence measured by pill count was also high: 88% of all bottles were dispensed and 97% of dispensed bottle doses were taken.

No significant differences in incidence

There were 136 seroconversions at a rate of 2.8/100py (95% CI: 2.3-3.3), one after an incorrect drug kit was dispensed. Of the remaining 135, 68 occurred on the acyclovir arm and 67 on the placebo arm (HR: 0.92; 95%CI 0.60-1.41; p=0.70).

In a modified intention to treat analysis, 43 transmissions were linked by viral sequencing technology to partners on the acyclovir arm and 47 were linked to partners on the placebo arm. However, two in the acyclovir arm and four in the placebo arm were excluded from analysis because the study drug was withheld during pregnancy. Here too, there were no significant differences between the two arms. The sequencing methodology for this study was explained in a late breaker poster from Mary Campbell. [3]

Benefits of acyclovir

There were fewer GUD events in the acyclovir arm (217 vs 550; RR: 0.39; 95% CI: 0.32-0.48; p<0.001). HSV-2-positive GUD as determined by DNA PCR was also lower in the acyclovir arm (92 vs 336; RR: 0.27; 95%CI 0.2-0.36; p<0.001).

The acyclovir arm also had a 0.25 log reduction in plasma viral load (95%CI: 0.22-0.29).

A novel component of this study was evaluation of herpes suppression on HIV disease progression, an important secondary endpoint of the Partners in Prevention trial. In a separate analysis presented by Jairam Lingappa, 3,381 of the HIV-positive participants were followed up until a composite endpoint of first of CD4 cell count <200 cells/mm³, ART initiation, or death from non-trauma causes. [4]

In the acyclovir arm, 284 participants reached this endpoint versus 325 in the placebo arm (HR 0.83; 95%CI: 0.71-0.90; p=0.03). Similar reductions were found for each component of the composite endpoint analysed separately. However, Lingappa’s team further calculated that for every 43 people treated with the trial dose of acyclovir for a year, only one person would be prevented from attaining the composite endpoint. (We have previously reported findings demonstrating acyclovir and its pro-drug, valacyclovir’s effect on HIV plasma RNA levels, in the October 2006 and July/August 2008 issues of HIV Treatment Bulletin, but this is the first report documenting impact of herpes suppression on HIV disease progression.)

Among participants with CD4 counts >350 cells/mm³ at enrollment, acyclovir delayed the time to CD4 < 350 cells/mm³ (HR 0.81; 95%CI 0.71-0.93; p=0.002). Here, 20 people would need to be treated to prevent one person from progressing to a CD4 count < 350 cells/mm³.

Acyclovir effect on genital viral load

A late breaker poster by Jared Baeten et al presented the results of a substudy that examined genital HIV RNA concentrations as a surrogate marker for HIV infectivity. [5]

Endocervical and semen samples were collected from 2,521 (1,805 women and 716 men) of 3,408 HIV-positive participants. For 1,797 of these, plasma was concurrently taken. For the remainder a plasma viral load within six months was available. Since the genital samples were taken only once during the study, the genital viral load was analysed as a time-independent variable.

HIV was detected in 60% of endocervical swab samples and 57% of semen samples. The median endocervical HIV concentration was 3.2 log (IQR 2.08-3.87) overall. Genital HIV-1 concentrations were significantly lower among those randomised to acyclovir (median 2.98 vs 3.29 for endocervical swabs; p<0.001 and 2.38 vs 2.76 for semen; p=0.008). The key finding of the study was that genital HIV concentrations were higher among HIV transmitting couples, where transmission was genetically linked to the partner (3.44 vs 2.49 log copies/mL for semen, p<0.001 and 3.91 vs 3.18 log copies/swab for endocervical swabs, p<0.001). Each log increase in genital HIV-1 RNA concentration was associated with 1.85-fold increased odds of HIV transmission for semen (p<0.001) and 2.03-fold increased odds of transmission for endocervical swabs (p<0.001). The study found no significant difference in genital HIV concentration for participants whose partners acquired HIV from outside sexual partners versus those who did not transmit HIV.

However, despite a 73% reduction in GUD and 0.25 log decline in plasma HIV levels and an approximately 0.3 log decline in genital HIV levels, acyclovir conferred no reduction in
HIV transmission. The authors interpret the overall results of the trial to indicate that the plasma and genital tract HIV viral load reduction from herpes suppression with standard doses of acyclovir is too small to confer a protective effect against HIV transmission.

Future acyclovir trials

Nevertheless, given the promising effect of acyclovir on HIV viral load, Steve Reynolds described an ongoing double-blind placebo controlled trial in Rakai, Uganda. [6]

The purpose of the trial is to evaluate the effect of suppressive HSV-2 therapy among HIV-1/HSV-2 co-infected individuals on progression to AIDS, defined as CD4 count < 250 cells/mm³ or WHO stage IV disease. Volunteers with CD4 counts between 300 and 400 cells/mm³, not on ART, without WHO III/IV symptoms and no history of opportunistic infections, other than mucocutaneous Kaposi Sarcoma, candida or treated TB were eligible for inclusion. Enrollment was completed in November 2008. The trial assumes that 40% of individuals in the placebo arm will progress to CD4 counts <250 cells/mm³ or AIDS over 24 months and is powered to detect at least a 20% reduction in HIV disease progression in the intervention arm.

COMMENT

These studies show that a standard dose of acyclovir for HSV-2 suppression does not reduce HIV transmission. These are disappointing findings for an HIV prevention strategy that is already available.

A mechanism for the lack of protection has been suggested by Laurence Corey and colleagues in a recent paper in Nature Medicine. [7]

By analysing regular skin biopsies taken during acute lesions and over 20 weeks follow-up, they indentified a ‘massive localised infiltration’ of CD4 and CD8 cells, thereby increasing the targets for HIV infection. Eight weeks after lesions healed, these levels were still 8-fold higher (655 and 618 cells/mm² of skin, respectively, compared to 68 and 55 cells/mm² in unaffected skin samples).

This paper is reported in detail in the Basic Science section of this issue of HTB. [8]

It has been conventional wisdom that wider availability of acyclovir for patients with genital herpes outbreaks would reduce HIV transmissions. We now know this is incorrect, at least with the doses of acyclovir (400 mg twice daily) used in these trials. However, efforts to make acyclovir widely accessible should continue because herpes is a debilitating, unpleasant disease which acyclovir effectively treats and because HSV-2 in widely prevalent in both HIV-negative and HIV-positive people. One of the barriers to its accessibility remains its high price in many developing countries.

Despite the negative findings, this trial and its substudies have set a high standard for the testing of future HIV prevention interventions. Furthermore, modeling studies using the data from this trial provide a potential threshold of HIV plasma viral load reduction in HIV-infected persons that will be needed to impact HIV transmission.

We now need to know whether a therapeutic dose of acyclovir could delay the time until initiation of HIV treatment, and whether this would be cost effective. The trial in Rakai described by Steve Reynolds using 400 mg twice-daily dose will provide complimentary information to the Partners in Prevention trial.

Studies with higher doses of valacyclovir will evaluate whether greater reduction in plasma HIV levels is feasible compared to acyclovir 400 mg twice daily. However, this research could be overtaken by new developments in ART management, if guidelines recommend earlier treatment.

References
Unless othewise stated, all references are to the Programme and Abstracts of 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention, 19-22 July 2009, Cape Town.

1. Celum C et al. Twice-daily acyclovir to reduce HIV-1 transmission from HIV-1 / HSV-2 co-infected persons within HIV-1 serodiscordant couples: a randomized, double-blind, placebo-controlled trial. 5th IAS 2009, Cape Town. Oral abstract WELBC101.
http://www.ias2009.org/pag/Abstracts.aspx?AID=3699
2. Celum C et al. Effect of aciclovir on HIV-1 acquisition in herpes simplex virus 2 seropositive women and men who have sex with men: a randomised, double-blind, placebo-controlled trial. Lancet, 21 June 2008. 371(9630):2109-19.
http://www.ncbi.nlm.nih.gov/sites/entrez/18572080
3. Campbell M et al. Determination of transmission linkage in the Partners in Prevention Study. 5th IAS 2009, Cape Town. Poster abstract LBPEC07.
http://www.ias2009.org/pag/Abstracts.aspx?AID=3756
4. Lingappa JR et al. Daily acyclovir delays HIV-1 disease progression among HIV-1/HSV-2 dually-infected persons: a randomised trial. 5th IAS 2009, Cape Town. Oral abstract WELBC102.
http://www.ias2009.org/pag/Abstracts.aspx?AID=3714
5. Baeten J et al. Genital HIV-1 RNA concentrations and heterosexual HIV-1 transmission risk. 5th IAS 2009, Cape Town. Poster abstract LBPEA07.
http://www.ias2009.org/pag/Abstracts.aspx?AID=3770
6. Reynolds SJ. HSV-2 Suppression Trial, Rakai, Uganda. Partners in Prevention presentation reported with permission.
7. Zhu J et al. Persistence of HIV-1 receptor–positive cells after HSV-2 reactivation is a potential mechanism for increased HIV-1 acquisition. Nature Medicine. Published online: 2 August 2009 | doi:10.1038/nm.2006.
http://www.nature.com/nm/journal/vaop/ncurrent/abs/nm.2006.html
8. Jefferys R. Immune surveillance below the radar: study offers explanation for acyclovir’s failure to reduce HIV risk. HTB September/November 2009. Vol 10, No 4. (Vol 2, No 3 of HTB South).

Results of a randomised, double-blind, placebo controlled trial in Tanzania of single high-dose oral fluconazle was reported by Omar Hamza and colleagues from Muhimbili University, Tanzania.

The trial randomised 220 HIV-positive patients with clinical and mycological evidence of oropharyngeal candidiasis to receive oral fluconazole doses of either 750-mg single dose or standard dose of 150 mg once-daily for 2 weeks. Each arm included 110 patients.

Results were similar in each group and are detailed in Table 1. There were no statistically significant differences between the two groups. Approximately 95% patients were clinically cured (OR, 0.825; 95% CI, 0.244–2.789; p=0.99) and 85-75% mycologically cured (OR, 1.780; 95% CI, 0.906–3.496; p=0.129).

Table 1: Results from single dose vs 2-week fluconazole

Overall,

clinical cure was not achieved in 11 patients and for all of them, Candida species were isolated from patient specimens at baseline and on day 14. In 33 patients (15.0%), clinical cure was obtained despite persistent positive culture results on day 14 (mycological failure).

No differences were observed in relapse rates (OR, 1.073; 95% CI, 0.456–2.523; P=0.99). The average time to relapse after clinical cure was 18-20 days. Twenty-two (91.7%) of 24 patients who experienced relapse during follow-up had CD4 cell counts <200 cells/mm3, 16 (66.7%) had CD4 cell counts <100 cells/mm3, 17 (70.8%) were not receiving HAART, and 14 (58.3%) had had previous episodes of OPC.

The mean plasma fluconazole concentrations on days 1, 4 or 5, 7, and 14 in the 14-day fluconazole group were 13.35, 5.46, 1.37, and 0.32 mg/L and 4.18, 6.88, 7.94, and 7.62 mg/L, for the single-dose and 14-day groups respectively. These differences were statistically significant for days 1, 7 and 14.

Overall, adverse events were mild, and no differences in frequency of adverse events were noted between patients in the 2 treatment regimens. Because most of the study patients were in an advanced stage of HIV infection and AIDS, abnormalities in full blood count and liver function tests were common.

In this study, the mycological cure rate, with a single-dose treatment of 750 mg fluconazole, was much higher (84.5%) than the 6%–41% mycological cure rates reported from studies using 150mg single dose treatment.

The authors concluded “The use of a single high dose of fluconazole … presents the advantages of simplicity and convenience, thus improving compliance and reducing the cost of therapy. A single dose of five 150-mg tablets is less costly than fourteen 150-mg tablets taken over a 14-day course and, therefore, could be used, especially in resource-limited settings like in sub-Saharan Africa. In addition, administration of the single-dose therapy can be observed directly by medical personnel, thereby assuring patient compliance.”

Ref:
Omar J et al. Single-Dose Fluconazole versus Standard 2-Week Therapy for Oropharyngeal Candidiasis in HIV-Infected Patients: A Randomized, Double-Blind, Double-Dummy Trial. Clinical Infectious Diseases 15 November 2008, Vol. 47, No. 10: 1270-1276.