The use of highly active antiretroviral therapy (HAART) has dramatically reduced the appearance of many AIDS-related infections and deaths in people with HIV/AIDS (PHAs) who are able to tolerate and adhere to such therapy. HAART, however, is not a cure, and the immune systems of PHAs who use HAART do not fully recover from HIV infection. Incomplete recovery may be another reason why some PHAs develop certain infections despite the use of HAART.

Doctors in Rome, Italy, have been studying the health of some PHAs using HAART. In one report, researchers analysed the information in their database collected in 1999 on nearly 1,400 PHAs who took and did not take HAART. As well, the researchers also analysed their data based on the PHAs' response to HAART. They found that those PHAs who "responded" developed different infections from those PHAs who did not respond to therapy. By "responded" the doctors meant that these PHAs' viral load fell below the 500 copy mark and their CD4+ counts rose to at least 200 cells. "Non-responders" did not develop such improvements.

The researchers recruited 1,389 HIV positive subjects, of whom only 34% (474) were receiving anti-HIV therapy that consisted of at least three drugs. The researchers tracked which subjects developed particular AIDS-related infections and/or cancers. They found the following trends:

• Untreated subjects were twice as likely to develop the life-threatening pneumonia PCP (Pneumocystis carinii pneumonia) than subjects taking HAART. This difference was statistically significant; that is, not likely due to chance alone.

• Among HAART-users, about 25% of the "responders" developed a form of cancer called lymphoma. In "non-responding" users of HAART, the proportion who developed lymphoma was much lower, about 6%. Similarly, among people who remained off treatment, the figure was similar, about 5%. This difference in lymphoma between responders and non-responders was statistically significant.

• The proportion of women who developed invasive cervical cancer was about 1% in untreated women, 4% in non-responders who used HAART and nearly 8% in women who responded well to HAART. This difference in the development of cervical cancer between the three groups of subjects was not statistically significant, but it certainly is troubling.

The Italian researchers note that even in subjects with a "good" response to HAART, serious AIDS-related infections/cancers can occur. We hope that other research teams who have large databases with health-related information on PHAs will try to confirm these findings.

1. JAIDS 2001;26(4):394-395.