Treatment access: more optimistic results from scale up programmes
Simon Collins, HIV i-Base
There were hundreds of posters relating to treatment access that were all very similar in describing successful results from roll-out programmes. All these studies are important. They document treatment success of local, regional or national significance, but contain few medical surprises. And as James McIntyre remarked in his rapporteurs report: In addition to large scale high impact access programmes thousands of smaller programmes are treating hundreds of thousands of people successfully at a community level. What he called the long tail of ART access. 
In summary, treatment works well for both adults and children in all settings. We just need a lot more of it and successful treatment is often limited by late diagnosis, poor nutrition, TB coinfection (though ARVs reduce the incidence of TB in HIV-infected patients), fees at point of access, and, increasingly, access to second-line treatment.
Adherence rates are often significantly higher than reported in Western studies, characterised in a meta-analysis presented by Mills and colleagues, comparing African and Western studies. 
They included 30 studies from North America (2 abstracts) and 22 studies (15 abstracts) from Africa (from 11 Sub-Saharan countries). All African studies were published after 2002. Patient self-report was used to assess adherence in 70% of US and 82% of African studies and similar thresholds for measuring appropriate adherence (eg. 100%, >95%, >90%, >80%) were used. In their pooled analysis, African patients had significantly greater levels of adherence: 77.1% [95%CI 67.3, 85.6] than North American patients: 54.7% (95% CI, 48.0, 61.3], comparison odds ratio 2.5, 95% CI, 2.2 to 2.8, P<0.0001.
Several of the earliest programmes also reported data on durability of treatment. For example, Goemaere and colleagues reported 5-year follow-up data from the Khayelitsha programme, which has been providing free treatment now for over five years. 
In this prospective cohort from three clinics outside Cape Town, a preliminary analysis from 1729 adults who started ARVs by the end of 2004, 76% of patients remain in care after four years, of whom 84% are still on their first-line regimen.
Median baseline CD4 count for new patients almost doubled from 46 cells/mm3 in 2001 to 85 cells/mm3 in 2004, and mortality over the first 6-months treatment fell from 13% to 7%. Two-thirds of the deaths in the programme occur in first six months. The proportion of patients with viral loads < 400 copies/mL at 6 months remained stable at between 88% and 91%. Toxicity-related treatment changes were reported as 16.7% and 8.3% cumulatively by 36 months, for patients starting with d4T or AZT respectively.
A second study from MSF presented by Sauvageot and colleagues reported >3-year follow-up from over 1,100 adults (>13 years old, 49% women) from 6 MSF programs in 5 countries (Cambodia, Cameroon, Kenya, Malawi, Guatemala). 
At baseline, 90% of patients were ART naive, 89% in WHO stage III/IV, with a median age of 34 years (IQR: 29-41) and a median CD4 count of 62 cells/mm3 (IQR: 18-136). Baseline median BMI was 20 kg/m2 (IQR: 18-22). 94% received the 2003 WHO recommended first line regimens (94%).
At 3 years, 768 patients (68%) were still on treatment, 235 (21%) had died, 102 (9%) were lost to follow-up (LTF) and 31 (3%) had transferred care to a different centre. For patients still on treatment, median CD4 level had increased to 326 cells/mm3 (IQR: 229-463) and BMI to 22 kg/m2 (IQR: 20-25).
The probability of still being followed at 3 years was 0.71% [95%CI, 0.68-0.73] using death and loss-to-follow-up as endpoints, and 0.87 [95%CI, 0.84-0.89] among patients still on treatment at 1 year. The probability of not developing a new WHO stage IV or III condition was 0.72 [95%CI, 0.70-0.75] and 0.64 [95%CI, 0.61-0.67], respectively. The probability of changing a single ARV drug because of intolerance or switch to a second line regimen was 0.29 [95%CI, 0.26-0.32] and 0.05 (95%CI, 0.04-0.06), respectively.
Just prior to the conference the WHO released new guidelines for treatment of HIV that included for the first time a caution against using d4T in first line treatment, because of the higher risk of side effects (see later in this issue of HTB).  While the virological durability reported in the MSF studies above is important and impressive, an indication of the impact of d4T on lipodystrophy in Africa was given in an MSF study from Rwanda. 
All patients (n=226) attending two MSF clinics in Kigali from November 2005-February 2006, who had been on WHO-recommended first line regimens for over one year were assessed for symptoms of lipodystrophy, using a Lipodystrophy Case Definition Study questionnaire and clinical examination
Of the 226 patients assessed, >90% used d4T/3TC-based regimens (187 with nevirapine and 20 with efavirenz). Only 19 patients used AZT/3TC-based regimens (n=13 with nevirapine and 6 with efavirenz).
Mean (SD) age/time on HAART was 38 (+/-8) years and 18.1 (+/-4.5) months; 77% were women. Body fat changes were reported by 65 patients and clinically confirmed in 56 cases, resulting in an overall prevalence of 24.8%. Fat loss was observed in 11.8% (n=24); fat accumulation in 4.5% (n=9), and mixed patterns in 10.7%(n=23);
Women were more likely to have lipoatrophy (fat loss) compared to men (p=0.01). Use of d4T was significantly associated with lipodystrophy in general (p=0.04) and fat loss in particular (p=0.02). Baseline and maximal weight on HAART were significantly higher in patients with lipoatrophy (7+/-2 and 8+/-2 kg difference, respectively (p<0.01)). Lipoatrophy was associated with recent onset weight loss (-5.1+/-2.9 kg vs -1.0 +/-0.9 kg p<0.001), occurring at a faster rate (0.52 +/-0.07 kg/wk vs 0.18 +/-0.02 kg/wk, p<0.001). No association of lipodystrophy/lipoatrophy was seen with clinical/immunological parameters, age of patients and time on HAART.
In Kevin de Cocks plenary from 3 by 5 to universal access he showed that the scale of treatment of women has run in parallel and sometimes exceeded the feminisation of the epidemic in high prevalence countries; and that 60% -70% of adults receiving treatment are women. 
TB is the leading cause of death in Africa and in many places the leading cause of death for people with HIV.
A very concerning study of extensively drug resistant (XDR) TB reported by Neel Gandhi from KwaZulu South Africa reported a high prevalence acquired both nosocomially and in the community with a very high mortality rate. 
A total of 10% of all positive TB isolates were not only resistant to INH and rifampicin (MDR TB), but also XDR, resistant to all first and second line TB drugs (INH, rifampicin, ethambutol, streptomycin, ciprofloxacin, kanamycin). Most people had no prior history of TB therapy and 36% had no prior hospitalisation. Two health care workers (and possibly four others were infected with XDR TB and died). Of all people with XDR TB, 98% died.
Unless stated otherwise, all references are to the Programme and Abstracts of the XVI International AIDS Conference, Toronto, Canada. 13-18 August 2006.
- McIntyre J. Rapporteur Report, Track B. FRPL0102.
- Mills E, Nachega J, Buchan I et al. Adherence to antiretroviral therapy in Africa versus North America: a meta-regression analysis. TUPDB03.
- Goemaere E, Darder M, Hilderbrand K et al. Five years of free ART in resource-limited settings: the experience in Khayelitsha. Abstract WEPE0072.
- Sauvageot D, Ferradini L. Clinical and immunological long-term outcomes on adults after 3 years of ART in Medecins sans Frontiers (MSF) programs: a multicentric analysis. Abstract WEPE0074.
- Antiretroviral therapy for HIV infection in adults and adolescents in resource-limited settings: towards universal access. August 2006 revision.
- van Griensven J, De Naeyer L, Mushi T et al. How common is lipodystrophy after at least 1 year of WHO first line antiretroviral treatment in Kigali, Rwanda? Poster abstract WEPE0140.
- de Cock K. From 3 by 5 to universal access. Plenary session. Oral abstract WEPL01.
- Gandhi NR, Moll A, Pawinski R et al. High prevalence and mortality from extensively drug-resistant (XDR) TB in TB/HIV coinfected patients in rural South Africa. Late breaker abstract THLB0210.