Research Article

Surrogate markers for survival in patients with AIDS and AIDS related complex treated with zidovudine.

BMJ 1991; 302 doi: https://doi.org/10.1136/bmj.302.6768.73 (Published 12 January 1991) Cite this as: BMJ 1991;302:73
  1. M. A. Jacobson,
  2. P. Bacchetti,
  3. A. Kolokathis,
  4. R. E. Chaisson,
  5. S. Szabo,
  6. B. Polsky,
  7. G. T. Valainis,
  8. D. Mildvan,
  9. D. Abrams,
  10. J. Wilber
  1. Department of Medicine, University of California, San Francisco.

    Abstract

    OBJECTIVE--To determine whether early effects of zidovudine treatment on CD4+ lymphocyte count and concentrations of beta 2 microglobulin, neopterin, or HIV p24 antigen or antibody are correlated with survival in patients with AIDS or AIDS related complex. DESIGN--Retrospective study of changes in laboratory markers and survival. SETTING--Multicentre trial at university hospital clinics. SUBJECTS--90 Patients with AIDS or AIDS related complex. INTERVENTION--Patients started zidovudine 200 mg orally every four hours. Fifty six of the patients died a median 17 months after starting zidovudine; the remaining 34 patients were followed up for a median 25.5 months. MAIN OUTCOME MEASURES--Changes in CD4+ lymphocyte count and serum concentrations of p24 antigen and antibody, beta 2 microglobulin, and neopterin; survival of the patient. RESULTS--The pretreatment characteristics that independently predicted poor survival were determined using a multivariate proportional hazards model: a diagnosis of AIDS (v AIDS related complex), age over 45 years, and the logarithm of serum neopterin concentration. When these baseline characteristics were controlled for the logarithm of CD4+ lymphocyte count at weeks 8-12 of treatment (p = 0.007) and an increase in serum beta 2 microglobulin concentration at weeks 8-12 (p = 0.05) also independently correlated with survival. In the 38 patients with a better pretreatment prognosis, 24 month survival estimated by the product-limit method was 88% for those with a good response on both surrogate markers during early treatment compared with only 50% for those with a poor response on either marker. In the 38 with a worse pretreatment prognosis, 24 month survival was estimated to be 49% for those with a good response on both surrogate markers compared with only 18% for those with a poor response on either. CONCLUSION--These data suggest that CD4+ lymphocyte count at 8-12 weeks and, perhaps, change in serum beta 2 microglobulin concentration could be surrogate end points for clinical outcome in trials of antiretroviral drugs for patients with HIV disease.