[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
May 24/31, 2000

AIDS-Related Opportunistic Illness and Potent Antiretroviral Therapy

Author Affiliations

Phil B.FontanarosaMD, Deputy EditorIndividualAuthorStephen J.LurieMD, PhD, Contributing EditorIndividualAuthor

JAMA. 2000;283(20):2653-2654. doi:10.1001/jama.283.20.2653

To the Editor: Dr Ledergerber and colleagues1 presented data on the incidence of opportunistic infections and cancers in human immunodeficiency virus (HIV)–infected persons. They compared the incidence of specific illnesses in the 6-month period before potent antiretroviral therapy with the incidence in periods after initiation of therapy, and concluded that a decreased incidence of some (but not all) illnesses indicated an effect of therapy. However, analytic difficulties complicate interpretation of these trends for individual opportunistic illnesses.

Pretherapy incidence rates are not strictly comparable with posttherapy rates. Unlike posttherapy rates, pretherapy rates are calculated retrospectively for a group known to be alive at the end of the pretherapy period. These conditional incidence rates are too low for illnesses with a high mortality rate because some affected persons do not survive to start therapy. The effect of therapy is then underestimated if these low rates are included in trend tests. For example, diseases caused by cytomegalovirus and nontuberculous mycobacteria occur late in HIV infection. In the analysis by Ledergerber et al, mortality-induced underestimation of pretherapy incidence of these diseases may have diminished the estimated effect of therapy.

Furthermore, pretherapy rates (and, thus, trends in incidence) could have been affected by clinicians' decisions regarding when to begin antiretroviral therapy. For instance, the study found a slight effect of antiretroviral therapy on tuberculosis risk. This effect may be underestimated because rifampin is a mainstay of tuberculosis treatment, but use of rifampin prohibits concomitant use of protease inhibitors.2 Delayed initiation of antiretroviral therapy would lead to artifactually low pretherapy tuberculosis incidence rates and underestimation of the effect of therapy. Similarly, no effect of therapy on non-Hodgkin lymphoma risk was seen, but lymphoma risk in untreated patients may have been underestimated if chemotherapy precluded antiretroviral therapy. Conversely, clinicians' decisions may have caused apparent increases in pretherapy incidence if, in general, clinicians were likely to provide antiretroviral therapy to patients who had shown disease progression (ie, who had had recent opportunistic illnesses). This bias would lead to overestimation of the effect of therapy.

Given that antiretroviral therapy has dramatically reduced overall HIV-related morbidity and mortality,3 it is important to clarify effects of therapy for specific opportunistic illnesses. To this end, the authors might repeat their trend analyses, limiting their observations to the period following antiretroviral therapy. Alternatively, they could compare individuals who started therapy with appropriately matched control subjects who never received therapy. One would expect pretherapy incidence to be similar to that in control subjects and to be lower after starting therapy for diseases for which antiretroviral therapy is preventive.

Ledergerber  BEgger  MErard  V  et al.  AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss HIV Cohort Study.  JAMA. 1999;282:2220-2226.Google Scholar
Centers for Disease Control and Prevention, Prevention and treatment of tuberculosis among patients infected with human immunodeficiency virus: principles of therapy and revised recommendations.  MMWR Morb Mortal Wkly Rep. 1998;47(RR-20):1-58.Google Scholar
Palella  FJDelaney  KMMoorman  AC  et al.  Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection.  N Engl J Med. 1998;338:853-860.Google Scholar