Two core assumptions of the AIDS epidemic, it turns out, could bewrong.

One rule of thumb counts on a victim's CD4-positive T cell count asthe true measure of the HIV-1 infection's status. The other construesthe long, variable time lapse between initial infection and emergenceof AIDS symptoms as proof that the virus lies low in latent form,before breaking out.

A decade-long study of 180 HIV-positive patients, reported intoday's Science, reveals that the baseline levels of virus in a patient'sblood, rather than his population of circulating CD4-positive T cells,reliably foretells progression to full-blown AIDS, and death. And itshows that the final viral onslaught does not reflect the end oflatency, but breaking the stand-off between HIV-1 replication andimmune defense.

The paper bears the title: "Prognosis in HIV-1 infection predicted bythe quantity of virus in plasma." Its principal author is microbiologistand epidemiologist Lawrence Kingsley, at the University ofPittsburgh.

"We are participants in the Multicenter AIDS Cohort Study,"Kingsley told BioWorld Today, "which is sponsored by the NationalInstitute of Allergy and Infectious Diseases [NIAID]. So in theprocess of seeing HIV-infected patients since April 1984, we'vestored their biological samples.

"In 1995, 11 years later," he continued, "a `branched-DNA signalamplification test' that measures viral RNA became available fromChiron Corp. So we went back to our freezers, thawed specimens ofblood, and used this assay, which counts RNA copies of HIV virionsper milliliter of blood, to test for the patients' viral burden.

"We measured them," Kingsley said, "from the beginning to the timewhen the individuals either developed AIDS or were last seen in ourresearch study as healthy volunteers."

The median follow-up time for patients who developed AIDS was 5.6years; for those who did not, 10.6 years.

Longest HIV Study Tells Tale

Their data allowed the Pittsburgh team to conclude that "Plasma viralload was a better predictor of progression to AIDS and death thanwas the number of CD4+ T cells."

For example:

* Only 5 percent of patients with fewer than 4,350 HIV RNA copiesdied within five years of entering the study, compared with 49percent of those with more than 36,270. Approximately 95 percent ofpatients with persistent viremia progressed to AIDS, and died within10 years.

* CD4+ T-cell counts proved not to correlate with viral burden, andof only limited prognostic value. Thus, of patients with more than500 CD4+ T cells per cubic millimeter of blood, half with viral loadsabove 10,900 died within six years, but only 5 percent with similarcell counts and viral load less than 10,900.

"Historically," Kingsley said, "CD4 count has been the best and mostwidely used test. Henceforth, we believe it should operate in tandemwith viral load assays.

"We're suggesting that we can now stage HIV disease in a fashionsimilar to the surgical staging of cancer." This assesses tumorinvasion and spread as a guide to planning and monitoring therapy.

One of the major messages from the paper in Science," Kingsleywent on, "is that viral-load testing should be done to monitor anindividual's response to any retroviral therapy. That is extremelyimportant as these new therapies become available, in determiningwhich one offers more promise more quickly."

MDs Can Order Viremia Assay

Chiron, of Emeryville, Calif., now is providing its HIV-load assay asan overnight service to primary care physicians, for $200 perspecimen, said Larry Kurtz, vice president of corporatecommunications. It also is making it available through the clinicalreference laboratories of SmithKline Beecham plc, of Philadelphia,and Corning Inc., of Corning, N.Y., he said.

Academic and industrial research laboratories may purchase the assaysystem itself for $75 to $100.

"We applied last February to the FDA for regulatory approval,"Kurtz added. (See BioWorld Today, Feb. 14, 1996, p. 3.)

Chiron biochemist John Todd is a co-author of the Pittsburgh paper,and a principal developer of the branched DNA (bDNA) assay. Heexplained how it works:

"With PCR you're taking a very few molecules _ in this case, HIVRNA _ and multiplying the number of molecules with enzymes. Soyou amplify the few that are actually in the specimen, which givesyou millions. Then you use mundane technologies to measure thosecopies.

"With bDNA, you're directly measuring the few copies that arepresent in the specimen. Consequently, the errors inherent in the PCRenzymatic multiplication step are alleviated.

"Branched DNA," Todd explained, "is a Chiron nickname for one ofthe proprietary molecules _ a synthetic piece of DNA _ that's thebasis for the whole technology."

It all takes place in a 96-well plate, and ends with light emissiongenerated by the captured HIV particles "that is directly proportionalto the amount of RNA that was there to start off with." n

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.