Estimating Prognosis in HIV-1 Infection

Progression of infection due to the human immunodeficiency virus type 1 (HIV-1) is highly variable. Cohort studies have estimated that the median time, free of serious complications after HIV-1 infection, is 7 to 11 years [1, 2]. These estimates of the incubation period are based on progression to the acquired immuno-deficiency syndrome (AIDS) as defined by the Centers for Disease Control and Prevention in 1987 [3]. With the change in the definition of AIDS [4], the median duration of AIDS-free time will probably decrease [5]. Life-threatening complications of immunosuppression develop after a shorter period of clinical latency in some HIV-1-infected persons [6]. In contrast, other persons show little, if any, loss of immune function or clinical problems during a period of 8 or more years. Differences in the infecting virus [7], the genetic makeup of the host [8], and environmental factors (including concomitant infection [9]) have been suggested as causes of the variation in the duration of clinical latency in persons not receiving antiretroviral therapy. Treatment, with inhibitors of viral reverse transcriptase such as zidovudine (Retrovir) and administration of prophylaxis for pneumonia due to Pneumocystis carinii, has increased (P > 0.05) AIDS-free time in HIV-1- infected persons [10]. Prognostic Factors Given this variable response to HIV-1 infection, what has been learned since 1983 that can help the clinician establish a prognosis? A number of clinical and laboratory factors have been identified that can aid in the estimation of prognosis in HIV-1-infected persons. The single most important measure is the enumeration of circulating T-helper or CD4+ lymphocytes. A single value establishes a statistical estimate of disease-free survival time [11], the risk for developing P. carinii pneumonia [12], and the risk for complications such as retinitis due to cytomegalovirus [13]. However, additional information can be obtained by determining the presence or absence of HIV-1-related signs and symptoms in persons within the same CD4+ lymphocyte strata. For instance, the risk for developing P. carinii pneumonia in persons with fewer than 200/L CD4+ lymphocytes is two to three times greater in symptomatic compared with asymptomatic persons [12]. Serologic assays for the core (p24) antigen of HIV-1 and antibody to p24 can define the risk for progression to serious clinical disease, but in a multivariate analysis, results of these measures did not add independently to the information provided by the CD4+ lymphocyte count [14]. Investigators at Walter Reed have suggested that delayed skin reactions to a panel of common antigens also provide useful prognostic information [15]. However, studies in the Chicago cohort of the Multicenter AIDS Cohort Study did not confirm this finding; anergy to three skin test antigens was prevalent in asymptomatic HIV-1-infected men with T-helper cell lymphocyte counts greater than 400/L, and absence of skin test reactivity did not add to information obtained from the CD4+ count alone [16]. More useful prognostic factors are serum concentrations of -2 globulin and neopterin. Both are increased in the presence of chronic inflammation or chronic infection due to activation of lymphocytes and macrophages, respectively. Fahey and colleagues [17] have shown that either measurement can be used to assess the risk for progression to AIDS in HIV-1-infected persons. These investigators also showed that combining neopterin values with CD4+ lymphocyte counts improved prognostic information [17]. CD4+ Counts and Therapeutic Decisions Currently, all therapeutic decisions are based on the CD4+ lymphocyte count and the clinical status of the patient. Zidovudine (500 mg/d) should be initiated when the CD4+ lymphocyte count is consistently fewer than 500/L [18]. The number of CD4+ lymphocytes often increases or stabilizes for a while with antiretroviral therapy. With resumption of the decline in T-helper cell counts or clinical deterioration or both, consideration should be given to changing to another reverse-transcriptase inhibitor (such as dideoxyinosine [Didanosine] or dideoxycytidine [Zalcitabine]) or combining one of these agents with zidovudine [19, 20]. A patient who has an initial CD4+ lymphocyte count fewer than 300/L, especially if symptomatic, probably should be given a combination therapy [20]. When the CD4+ count decreases to 200/L, prophylaxis for P. carinii pneumonia is started [12]. If the CD4+ lymphocyte count is fewer than 100/L, prophylaxis to prevent infection with organisms of the Mycobacterium avium complex is indicated [21]. The CD4+ count and the various serologic assays are surrogates for progression and do not measure immunologic function. This limits the physician's ability to provide an accurate prognosis and to assess the effect of therapy, as noted by Choi and colleagues [22] in this issue of Annals. Investigators, therefore, are trying to determine if more specific virologic assays can improve our ability to establish a prognosis and to provide a rational basis for intervention or modifying therapy or both. Koot and colleagues [23], in this issue of Annals, describe the association of a specific phenotype of HIV-1 (a syncytium-forming variant) and progression of infection. Other investigators [24] have suggested that quantitative measures of viral burden, such as the concentration of cell-free virus in plasma, provide prognostic information. Molecular techniques that can determine serum or cellular viral RNA, a surrogate for the level of HIV-1 replication, are also being evaluated [25-27]. Another area of research is the assessment of immunologic function [28]. As Choi and colleagues [22] describe, the benefit of zidovudine therapy for immune status is not measured completely by changes in CD4+ lymphocyte number. It is also possible that a therapeutic agent could have clinical efficacy and not increase either the CD4+ lymphocyte number or percentage. Conclusions Whether these virologic or immunologic approaches will prove useful in the management of HIV-1-infected persons or in evaluation of new therapeutic agents remains to be established. It is possible that therapy should be directed by changes in viral phenotype or genotype, especially if such changes are associated with development of HIV-1 resistance to antiviral agents (an estimate of viral burden), viral replication, or alterations in immunologic function. Such laboratory techniques will need to be accurate, reproducible, inexpensive, and readily available. The two papers in this issue of Annals raise important issues relative to our current management of HIV-1 infection and, potentially, point to future directions. Abbreviation SI: syncytium-inducing