HIV clinical trials are not enough

• Published in: Canadian Medical Association journal (CMAJ) 1999-03, Vol.160 (5), p.669-670
• Main Authors: Millson, M E, Rachlis, A R
• Format: Article
• Language: English
• Subjects: Acquired immune deficiency syndrome; AIDS; AIDS/HIV; Anti-HIV Agents - therapeutic use; Clinical Trials as Topic; Databases as Topic; Health care; HIV Infections - drug therapy; Humans; Nursing; R&D; Research & development; Treatment
• ISSN: 0820-3946; 1488-2329

In clinical trials of antiretroviral drugs, such as the first clinical trials of zidovudine, clinical outcomes were used as evidence of efficacy.1 It was later shown that favourable clinical outcomes were associated with viral load reductions.2,3 The new paradigm of HIV therapy is complete suppression of viral replication to preserve and restore immune function and to delay or prevent the emergence of drug resistance.4 Hence, drug efficacy is now being measured in clinical trials by the degree and duration of viral load reduction. Clinical trials designed to demonstrate favourable outcomes in terms of reduced morbidity and mortality require prolonged follow-up and large numbers of patients, and they are often not feasible because new therapeutic strategies may be introduced during the study period. This leaves the demonstration of such clinical benefits to other methodologies, including the use of long-term observational databases. To address some of these issues, which cannot be addressed in HIV clinical trials, a variety of observational databases are being used, including research cohorts (e.g., the cohort used in the Multicentre AIDS Cohort Study5), single-centre clinical care cohorts (e.g., the Southern Alberta Clinic cohort6), multicentre observational databases (e.g., the HIV Ontario Observational Database,7 the US Centers for Disease Control and Prevention Adult Spectrum of Disease database8 and the AIDS Research Consortium of Atlanta Database9), and databases set up for the purpose of administering drug treatment programs. An example of the last category is the database for the HIV/AIDS Drug Treatment Program in British Columbia, used by Dr. Robert S. Hogg and associates (page 659),10 in which demographic and clinical data are augmented voluntarily by patients and physicians. Hogg and associates have used the strongest type of data for assessing the impact of treatment in the community because the database contained information obtained from a large number of unselected providers and patients who were not influenced by research participation. Such a database can thus help in determining what actually happens in the application of treatment guidelines developed by experts in response to evidence from clinical trials. Hogg and associates have demonstrated significant population-based reductions in the rates of death and progression to AIDS among HIV-infected people treated with ERA-III antiretroviral drug regimens (2 nucleoside analogue reverse tran