Such treatment improved CD4 counts in Scottish patients

EDITOR—Egger et al describe the positive impact of antiretroviral combination treatment of people with HIV infection in Switzerland.1 We analysed data from Scotland's national CD4 lymphocyte monitoring scheme to describe the effect of antiretroviral combination therapies on progression of HIV disease among infected people in Scotland.2

We divided people undergoing monitoring of CD4 counts over two consecutive years in 1993-7 into four cohorts. Altogether 770 patients underwent CD4 cell count monitoring during 1993 and 1994, 731 during 1994 and 1995, 706 during 1995 and 1996, and 708 during 1996 and 1997. Median differences in CD4 cell counts were calculated by comparing the patients' first CD4 cell counts in years one and two. In each cohort the median first CD4 cell count in year one (baseline) was similar, ranging between 247 and 290 × 10⁶ cells/l. For each of the three cohorts spanning 1993-6, the median loss of CD4 lymphocytes over consecutive years ranged between 24 and 32. For the 1996-7 cohort, however, there was a median gain of six CD4 cells (95% confidence interval 0 to 12) (figure).

No of new cases of AIDS and death from AIDS, with median changes (and 95% confidence intervals) in CD4 cell counts in Scotland, by year

• Download figure
• Open in new tab
• Download powerpoint

It is well recognised that progression of HIV disease is associated with the loss of CD4 lymphocytes in the peripheral circulation. The dramatic change from median losses of around 30 cells per year during 1993-6 to a median gain of 6 cells between 1996 and 1997 suggests that giving combination therapy regimens during this time has had a major impact on preventing CD4 count depletion and has possibly contributed to cell gain.

The extent of the change in therapeutic practice in Scotland is difficult to gauge because surveillance data from the Scottish Centre for Infection and Environmental Health on treatment are incomplete. A minimum of 34% of cases in the 1996-7 cohort, however, were taking dual and a further 23% triple regimens at some stage during 1997, compared with 17% and 6% during 1996. Furthermore, most clinicians who manage patients with HIV infection in Scotland indicated that they have been giving combination therapies where possible since the latter part of 1996. In keeping with these findings, the figure shows that the annual numbers of cases of diagnosed AIDS and deaths from AIDS have decreased since 1995.

The local difficulties in funding treatment described in England3 are equally applicable in Scotland. Despite these difficulties it seems that the benefits of combination therapy that were observed in randomised controlled trials are now being successfully translated into clinical practice.

Long term follow up of patients under triple therapy is necessary

EDITOR—Egger et al reported reduced progression of and mortality from HIV disease with the new antiretroviral combination therapies in a Swiss HIV cohort during 1988-96.1 They were not, however, able to assess the contribution of triple therapy with protease inhibitors, which is an important recent development in combination therapy.2 We attempted to do so in the Aquitaine cohort in France.3

At enrolment our cohort (n=3550) was comparable with the Swiss cohort in age and HIV transmission group after stratification for the inclusion calendar period. There was an increasing proportion of men in the Aquitaine cohort in 1995-6. Clinical stage differed in each inclusion period, the Aquitaine cohort having consistently more patients with C stage disease at inclusion. The follow up per period was 20% longer in our group on average.

We distinguished five groups of antiretroviral treatment: monotherapy alone with a nucleoside analogue; dual therapy alone with two nucleoside analogues; triple therapy alone including one protease inhibitor; monotherapy followed by dual therapy; and monotherapy or dual therapy followed by triple therapy. The few patients who were treated with protease inhibitors but were not having triple therapy were excluded from the analysis.

A multivariate analysis of the risk of progression to a first diagnosis of AIDS or death was performed with a Cox proportional hazards regression model; the five groups were compared with patients who had never been treated, on the basis of intention to treat. The model was adjusted for CD4 cell count, age, sex, disease stage, history of intravenous drug use, use of prophylaxis against opportunistic infections, and period of enrolment (table). Time was measured from the date of first CD4 cell count under 200 × 10⁶ cells/l.

Dual therapy with (relative hazard 0.19 (95% confidence interval 0.14 to 0.25)) or without (0.29 (0.17 to 0.49)) previous antiretroviral treatment and triple therapy with (0.04 (0.03 to 0.06)) or without (0.07 (0.02 to 0.29)) previous antiretroviral treatment were independent protective factors of the risk of death compared with the absence of the antiretroviral treatment. Monotherapy alone did not change the vital prognosis (relative hazard 0.91 (0.75 to 1.1)). Triple therapy was more efficient than dual therapy, with a stronger protective effect regardless of history of previous treatment. Those results were comparable for the progression to AIDS except for monotherapy, which increased the risk of reaching the AIDS stage (relative hazard 2.00 (1.41 to 2.82)).

Like Egger et al's our results confirm the reduction in disease progression and mortality with introduction of antiretroviral combination therapies. We detailed the role of protease inhibitors in this risk reduction, confirming through an observational cohort the results of clinical trials.2 A possible bias in our findings was that the first patients treated by triple therapy without previous treatment were more likely to be patients with C clinical stage disease. This implies that the protective effect of this new class of drugs in combination therapy is likely to be more important than observed so far. Long term follow up of patients receiving triple therapy such as in the Swiss or Aquitaine cohorts is necessary to confirm the efficacy of protease inhibitors under routine clinical circumstances.