perspektiv på hiv Perspektiv på hiv



Summary from the 18th Conference
on Retroviruses and Opportunistic Infections (CROI), Boston, 2011

Per Björkman
Dept. of Infectious Diseases, Malmö,
Skåne University Hospital,
 Lund University, Sweden

Per Björkman är docent i infektionssjukdomar och överläkare på infektionskliniken i Malmö. Han har i många år arbetat med bland annat hiv, hepatiter, tuberkulos och andra infektionssjukdomar av olika slag. Han har också engagerat sig i internationellt arbete ibland annat Afrika och Asien. Han är ledamot i styrelsen för Stiftelsen Läkare mot AIDS Forskningsfond. Den text som här publiceras är av intresse för såväl läkare som behandlar hivpatienter som för en intresserad allmänhet även om allt knappast är av intresse för bägge dessa grupper. Så föreställer jag mig exempelvis att de två följande ämnesorådena, tuberkulos och cryptococcos, som bägge är av stort intresse för läkare med ansvar för hivpatienter är av mindre intresse för icke medicinsk personal, medan däremot det material som följer därpå och som handlar om ”nya strategier för hivprevention” efter vad jag föreställer mig borde vara av stort intresse för både läkare med spetskompetens inom det aktuella området och för en intresserad allmänhet – för mig det haft stort intresse både i min egenskap av hivläkare och intresserad allmän, eller mer specifikt i min egenskap av homosexuell man. Sak samma torde gälla andra. Det är inte särskilt svårt att snart nog upptäcka vad som intresserar en själv i denna allt sammantaget högintressanta och högkoncentrerade sammanställning av data från en högintressant konferens.
Lars Moberg



In the updated WHO guidelines from 2010, initiation of antiretroviral therapy (ART) is proposed for all HIV-infected patients with tuberculosis (TB); however, the timing of ART has not been clarified. Data from two studies were presented at the conference, comparing immediate ART (started within 2 weeks of TB treatment initiation) and early ART (started after 2 months of TB treatment; ie. after completion of the intensive phase). In the STRIDE study, 806 patients with confirmed (46% of patients) and clinically suspected TB, and with CD4 cell counts <250 cells/mm3 were included, with a median CD4 cell count of 77 cells/mm3. Patients initiated ART within a median of 10 vs. 70 days after starting TB treatment, respectively. Overall mortality rates were not significantly different in the two arms; however, in the subgroup of patients with CD4 cell counts <50 cells/mm3, the risk of death was significantly reduced in those starting ART immediately (15.5% vs. 26.6%). Irrespective of CD4 cell count stratification at baseline, patients in the immediate ART arm had significantly higher incidence of immune reconstitution inflammatory syndrome (IRIS; 11% vs. 5%). There were no significant differences between the arms with regard to confirmed or suspected TB diagnosis, apart from a greater incidence of IRIS in those with microbiologically confirmed TB.

Similar findings were reported in the SAPiT study, conducted in South Africa in 642 ambulatory patients with smear-positive pulmonary TB with CD4 cells <500 cells/mm3 (median 150). Although overall mortality was similar in both arms, patients with CD4 cells <50 cells/mm3 who started ART within 4 weeks of TB treatment had a 68% reduction in mortality, which was a significant improvement compared to later start of ART. Overall rates of IRIS were much higher in the immediate ART arm (46.8% vs. 9.9%), as well as drug switches.

Following these results, it seems clear that immediate ART should be prioritized for co-infected persons with the most advanced immunosuppression. This is also in agreement with the CAMELIA study conducted in Cambodia. In this cohort, an overall benefit of immediate ART was observed, with a reduction of mortality of 34%; however, participants in this study had extremely advanced HIV infection, with a median CD4 cell count at inclusion of 25 cells/mm3. Early identification of patients with CD4 cells <50 cells/mm3 is critical; in settings with limited access to CD4 laboratory technology, different scoring systems may be useful. In addition, parameters other than CD4 cell counts may be important to assess outcomes; it is likely that factors other than the CD4 cell threshold of 50 cells/mm3 should be taken into account when making the important decision on when to start ART in this group of severely sick patients.

Two studies suggest that active TB negatively impacts CD4 cell recovery during ART. In 4 large Italian cohorts, patients with TB at baseline had a lower likelihood of achieving CD4 cell counts greater than 300 and 500 cells/mm3, respectively, compared to patients with other AIDS-defining diagnoses at ART initiation. Among patients receiving ART in a referral centre in Kampala, Uganda, the CD4 cell recovery was significantly slower in patients who developed TB after starting ART, whereas subjects with prevalent TB at ART initiation had similar CD4 cell evolution as compared to patients without TB. The mechanism underlying this phenomenon has not been elucidated; both an effect of TB-related immune activation and a higher risk of ART failure could be responsible.

Although screening for TB is recommended in HIV-infected persons, the diagnosis of TB is difficult in immunosuppressed hosts. The most promising new diagnostic method is probably the rapid PCR machine GenXpert, which can provide a microbiological result within less than two hours, in addition to identification of common rifampicin resistance mutations (which may be used as a marker for multi-drug resistant TB). The lower limit of detection using this technique is 131 colony-forming units/mL, compared to >10 000 colony-forming units/mL for smear microscopy; the only microbiological diagnostic method available in most settings. Efforts are under way to increase access to better diagnostics in high-burden countries through the FIND initiative; however, costs and infrastructural constraints are likely to pose great problems. GenXpert has been validated for use on sputum samples, but seems to have good performance for lymph node aspirates as well, according to a South African study. The technique will also be tested on stool samples, which might be an alternative to gastric aspirates, especially in children.

Lipoarabinomannan (LAM) is a lipopolysaccharide component of the mycobacterial cell wall which is released from metabolically active or degrading bacteria, and can be detected in urine with an ELISA technique. For screening purposes in HIV-positive patients, LAM seems to perform poorly, but the sensitivity for detecting active TB increases with lower CD4 cell count. It is possible that LAM might be useful for diagnosing extrapulmonary TB in such patients; however, determination of specificity might be problematic in such an evaluation since detection of mycobacteria in these conditions may be extremely difficult and require extensive investigations. 



After TB, cryptococcosis is the most common serious opportunistic infection worldwide, and is the leading cause of adult meningitis in many regions of Sub-Saharan Africa. Rapid reduction of fungal burden in the cerebrospinal fluid (CSF), as well as intensive management of raised intracranial pressure by CSF drainage, has been associated with improved survival. Starting ART is obviously important, but the timing of ART in patients with cryptococcal meningitis remains uncertain, and currently there are no evidence-based recommendations on when to start ART. Severe and fatal variants of IRIS are not rare in this condition.

Combination antifungal therapy with amphotericin B and flucytosine have previously been shown to provide the best antifungal activity, but flucytosine is rarely available in resource-limited settings. A study performed in South Africa showed comparable results with regard to mortality, side effects and CSF antifungal activity, with amphotericin B given for 2 weeks combined with either flucytosine or high-dose fluconazole (800 mg or 1200 mg once daily).  From Uganda, a study of 30 patients was presented in which amphotericin B was given for 5 days. Shortening amphotericin therapy could facilitate treatment and reduce the need for laboratory monitoring of toxicity. Preliminary results showed adequate antifungal activity and no serious adverse effects related to amphotericin. Another approach to optimize treatment of cryptococcal meningitis is the addition of interferon-gamma, which was shown to be well tolerated and to increase rates of fungal clearance, even if only two doses were given on treatment days 1 and 3. 


HIV prevention – new strategies

In a plenary lecture, Jonathan Mermin (CDC, Atlanta) summarized the status of HIV prevention in the USA. Among 56 000 new infections annually, 53% occur in men who have sex with men (MSM) and 45% in black persons. African Americans have an 8-fold higher risk of being HIV-positive than white persons. It is estimated that more than 50% of new infections are transmitted from persons with undiagnosed HIV (considered to constitute 21% of the total number of persons living with HIV in the country). This has led to introduction of opt-out testing in settings such as emergency rooms and primary health centres in some regions. Furthermore, linkage to care for those with diagnosed HIV infection remains unsatisfactory; only 69% of subjects were estimated to have attended a clinic for HIV care within one year of diagnosis.

New strategies for HIV prevention that recently have been shown to reduce rates of transmission in randomized studies are vaginal microbicide gel and pre-exposure oral ARV prophylaxis (PreP). Both these interventions are based on tenofovir (TDF; with or without [1] emtricitabine [FTC]). In the iPrEx study, 2 500 high-risk MSM (median 18 partners within the preceding 12 weeks) were recruited in the Americas, South Africa and Thailand. A 44% reduction in HIV incidence was found in men receiving daily FTC/TDF compared to placebo (48 seroconversions in the FTC/TDF arm vs. 83 in the placebo arm; p=0.02). The risk of HIV acquisition was strongly associated with drug levels; only 9% of seroconverters in the active arm had detectable TDF plasma concentrations, as compared to 51% of those who remained HIV-seronegative. No participant with regular dosing and adequate plasma drug levels seroconverted. Persons with the highest risk behaviour (unprotected receptive anal intercourse) showed the highest benefit from the intervention. Interestingly, adherence studies using self-reports, pill counts and pharmacy refills indicated good adherence levels, but this was obviously not the case when drug concentrations were taken into account (only 62% of men with reported perfect adherence had detectable TDF in plasma).

Despite the proven efficacy of PreP in this study, several concerns remain with regard to its use in routine settings; including potentially increased risk behaviour, selection of FTC/TDF resistant variants, long-term side effects and cost-effectiveness.

In the iPrEx cohort, risk behaviour was reduced during the study period, which was attributed to a package of interventions apart from providing drugs – such as counselling, repeated HIV testing, provision of free condoms and STI screening. Side effects were rare and not serious. Special attention was given to a potential risk of bone loss. Low bone-mineral density was detected at base-line, associated with the use of amphetamine and poppers, but there was only a slight decrease of bone mass after PreP was started.

The use of PreP will require accurate identification of primary HIV infections (which were detected at baseline in 10 iPrEx participants), and regular HIV testing with linking to adequate care in persons found to be HIV-infected. Although very few resistant strains were observed in the iPrEx cohort, the risk of selection of resistance cannot be reliably estimated based on the limited data available. The real effectiveness (as well as cost-effectiveness – the annual cost is estimated at 12 000 USD) of PreP remains uncertain, and there are several unresolved practical issues, not least the logistical implementation of this intervention.   


New drugs for hepatitis C co-infection [2]

New directly acting antiviral molecules, such as inhibitors of the NS3/4A protease and the NS5 polymerase, are expected to improve treatment results for hepatitis C dramatically. Their use may be especially important in HIV co-infected persons, who show worse treatment outcomes with standard combination therapy than HCV-monoinfected patients.

Interim results of telaprevir or placebo given in combination with pegylated interferon-alfa 2a and ribavirin were presented in a late-breaker session. Among 79 participants, 59 were on stable ART (based on either efavirenz [EFV] or atazanavir [ATV/r]), and 20 were not taking ART (CD4>500 cells/mm3, HIV RNA< [3] 5 log). After 4 and 12 weeks, around 65% had undetectable HCV RNA in the telaprevir arm, compared to around 15% in those receiving placebo. Telaprevir recipients were more likely to develop moderate rash (14% vs. 1%), but no cases of severe rash have been noted so far. Median telaprevir concentrations were comparable in patients with and without concomitant ART, and no effect of telaprevir on the drug concentrations of EFV or ATV/r was observed.

Since the protease inhibitors (telaprevir as well as boceprevir) are both metabolized by the CYP 3A pathway, drug-drug interactions are expected, and have not yet been extensively studied with regard to ARV drugs. For telaprevir, different interactions with boosted protease inhibitors have been found; decreased concentration of darunavir and decreased elimination of atazanavir, whereas the concentration of lopinavir was unaffected. The exposure to tenofovir was increased by 30% in subjects receiving telaprevir. EFV increases the elimination of telaprevir by around 20%, and for this reason higher dosing of telaprevir with EFV is recommended (1125 mg every 8 hours). For boceprevir, no ARV data were presented.


New data on antiretroviral treatment

Results from a prospective trial, mostly conducted in low-income countries, comparing EFV given in combination with either co-formulated 3TC/AZT or FTC/TDF to 1 045 patients (46% women) with CD4 cell counts <300 cells/mm3, showed similar rates of death, disease progression, immune reconstitution and virological suppression for the two treatment arms. Side effects occurred more often in participants receiving 3TC/AZT – mostly anemia and neutropenia, but the rate of serious metabolic events was also higher in the 3TC/AZT arm (4% vs. 1%). An interaction with gender was noted, with a higher risk of adverse events in women receiving 3TC/AZT compared to FTC/TDF. Overall, the treatment outcomes in this study were excellent, with fewer events registered than estimated.

The efficacy of once-daily dosing with raltegravir (800 mg) has been compared to standard dosing (400 mg bid), in combination with FTC/TDF. Although 83% of subjects receiving once-daily raltegravir achieved HIV RNA<50 copies/mL, this was significantly inferior to standard dosing (89% suppression). The difference was greatest in patients with HIV RNA>5 log, but once-daily dosing of raltegravir cannot be recommended even to subjects with lower viral load based on these data.

Intensification of ART using raltegravir and the CCR5 inhibitor maraviroc has been investigated in acutely infected subjects, with the aim of achieving optimized control of viral replication, hypothesized to be associated with improved immunological responses and decreased immune activation. HIV replication was assessed by an ultra-sensitive single genome copy assay. Although the initial viral load decline was faster in the intensification arm, no advantage of the intensified regimen was demonstrated compared to patients who received a standard 3 drug ARV regimen; 3/11 subjects in the standard arm vs. 9/21 in the intensification arm had HIV RNA<1 copy/mL. Interestingly, 3 patients in the intensification arm failed to reach HIV RNA<50 copies/mL at week 48, despite showing adequate adherence and drug concentrations.

New treatment strategies have been inspired by the apparent eradication of chronic HIV infection in a patient who received an allogenic bone marrow graft for acute leukemia from a donor homozygous for the 32 base-pair deletion of the CCR5 receptor. The recipient remains alive off ART and without detectable HIV RNA in all sampled compartments nearly three years post-transplantation. Researchers have used CCR5-specific zinc finger nucleases to disrupt the CCR5 receptor on CD4 cells, and have managed to render these cells resistant to HIV infection in vitro. Similar in vitro experiments were reported to disrupt the CXCR4 receptor which is used for cell entry by X4 tropic viruses. In a clinical experiment, 6 patients on suppressive ART with CD4 cell counts between 200 and 500 cells/mm3 received infusions with CCR5 knock-out CD4 cells at different dosages. Engraftment of these cells was demonstrated by detectable proportions in peripheral blood and rectal mucosa of CD4 cells with CCR5 disruption during a follow-up period of 11 months. Furthermore, significant increases in CD4 cell counts were found in all patients, and no significant adverse effects have been observed hitherto.   


The role of chronic immune activation - caused by bacterial translocation?

The blocking of the CCR5 receptor has been hypothesized to have immunomodulatory properties which could lead to reduced immune activation, which is now recognized to be the most important pathogenetic factor in HIV disease progression. Addition of maraviroc or placebo for 24 weeks was studied in patients showing blunted immunological response to ART (persistent CD4 cell counts <350 cells/mm3) despite having virological suppression for over one year. No effect on CD4 cell counts was noted in subjects in the maraviroc intensification arm. Markers of immune activation showed a discordant pattern, with decreased lipopolysaccharide concentration but increased concentration of soluble CD14 receptor in plasma. In 15 sampled patients, rectal biopsies showed a nearly 2-fold increase in the percentage of activated CD4 and CD8 cells in subject receiving maraviroc. The mechanism for this unexpected finding could be that CCR5 chemokine ligands increase in compensation to the blocking of the CCR5 receptor, and that these ligands can use alternate T cell and macrophage receptors for signaling.

Other researchers presented data to support the association between chronic immune activation and failure to restore peripheral CD4 cell counts on suppressive ART. Patients with such immuno-discordant responses were found to have higher levels of IL-6 as a marker of immune activation, as well as higher levels of fibrosis in gut-associated lymphoid tissue from rectal biopsies. A proposed mechanism is persistent bacterial translocation from the gut, occurring as a consequence of the dramatic depletion of gut-associated CD4 cells early in the course of HIV infection and disruption of the intestinal barrier. The composition of the intestinal microbial flora could also be involved; one study demonstrated a shift in the gut microbiota in HIV-infected patients with higher proportions of Enterobacteriacae as compared to Lactobacilli and Clostridiales in controls. In vitro experiments of antigen-presenting cells from HIV-infected subjects have also shown hyper-responsiveness after bacterial stimulation, with greater production of pro-inflammatory cytokines compared to HIV-negative controls. The in vitro pro-inflammatory response was increased both to pathogenic (Salmonella typhimurium) and to non-pathogenic bacteria (Lactobacillus plantarum).

Chronic liver disease may contribute to bacterial translocation in HIV infection. Markers of bacterial translocation were found to be increased in patients co-infected with either hepatitis B or C virus, as compared to patients with HIV mono-infection. In addition, markers of bacterial translocation could also be informative of the risk of progression of liver disease, but the association between such markers and the stage of liver fibrosis is not evident from available data.


Mother-to-child transmission and pediatric ART

A Thai study showed vertical transmission of CMV to be independently associated with MTCT of HIV in formula-fed infants who received AZT for prevention of MTCT. Overall, 84% of HIV-infected infants had CMV infection diagnosed at 18 months of age, compared to 63% of those who remained HIV-negative. The increased risk of HIV transmission was found both in infants infected with CMV in utero as well as in those infected during delivery. 

Extended nevirapine (NVP) prophylaxis to breast-feeding infants born to HIV-infected South-African women was shown to be associated with a decreased risk of MTCT. Among children to mothers not on ART for maternal health, 1.4% of those receiving NVP prophylaxis for 6 months had acquired HIV infection at 6 months, compared to 3.4% of infants treated for 6 weeks. The protective effect was greatest in infants to mothers who were on ART for maternal health (29% at randomization); among these, only 0.2% were infected at 6 months. Over 90% of infants stopped breastfeeding between 6 and 9 months of age. This finding supports increased efforts to start triple ART for women during pregnancy, to be continued after delivery, at least for the duration of breast-feeding.

Previous studies have demonstrated better outcomes of ART based on boosted lopinavir (LPV/r) compared to NVP in verticallly infected children exposed to single-dose NVP for PMTCT. In a study of 288 children unexposed to NVP enrolled in 10 sites in Sub-Saharan Africa and India, LPV/r was also shown to be superior to NVP, with a significantly lower risk of treatment failure (31 vs. 11 cases) and death (5 vs. 1) during the 24 week follow-up period. 

(1) Se Aktuellt om hiv – nyhetsbrev från Noaks Ark av 20111014 som redovisar att amerikanska National Institutes of Health stänger den arm i studien ”VOICE” – HIV prevention study in women som endast innehåller tenofovir=Viread® till skillnad från armen med Truvada® (tenofovir och emtricitabin) och ännu en arm med tenofovir i i form av en microbicid gel-beredning av substansen. Orsaken till att armen med tenofovir per oralt intaget stängs är att någon signifikant skillnad från placebo kommer inte att kunna uppvisas. Exakta siffror i antalet smittade anges inte utan det kommer senare när studien avslutas. Lars Moberg
2) Se också artikel av Jenny Stenkvist LM
3) < =less than, mindre än  LM



© Perspektiv på hiv | Redaktör: Dr. Lars Moberg   . TOP