Gay Men's Health Crisis "Treatment Issues", Vol 8, No. 10, November, 1994
Dave Gilden

PCR and bDNA tests do not count actual HIV particles directly, but assess the amount of HIV genetic material present. Results are recorded as the number of viral "copies" or "equivalents" per unit measure. The tests for HIV RNA discussed here measure free virus particles in blood plasma. The production of new HIV reflects current viral replicative activity, which therapy can rapidly inhibit.

Similar assays for HIV DNA measure the number of infected cells with HIV incorporated in their genes. Since present and most upcoming treatments do not directly eliminate infected cells or the HIV therein, HIV DNA is a more stable quantity less sensitive to changes in therapy. Infected cells may be actively producing new virus particles, or, much more commonly, they possess quiescent or defective HIV genes. The latter cells may live for many months.

Utilizing Viral Load Assays

HIV load analyses could be applied in two distinct manners:

* In medical practice, they could be used to make individual treatment decisions. Rising HIV levels in a patient's blood could be taken as a sign of increasing failure on the part of the immune system or current drug treatment to control the virus. A physician could then recommend either beginning anti-HIV therapy or altering the present regimen to stave off further disease progression. Changing treatment in this manner is called a medical strategy approach and involves not just prescription of successive single-agent therapies against HIV, but juggling any of the proposed multidrug combinations as soon as their components are available.

* In clinical trials, a treatment's antiviral activity could be measured in terms of the reduction in the amount of HIV in trial participants' blood. Single compounds or multidrug combinations that cause the greatest decreases would be candidates for further testing and possibly accelerated approval by the Food and Drug Administration. The bDNA and PCR techniques have an intuitive appeal. The obvious goal of antiviral therapy is to inhibit HIV and thus allow an infected individual to maintain or recover health. The tests can monitor drugs' effect on HIV in an immediate, "real-time" fashion that can never be duplicated by CD4 counts, which vary more slowly in response to therapy while fluctuating sharply for extraneous reasons.

The Uncertainties

But serious questions remain concerning the significance of these tests. It may not be so difficult to establish that rising HIV levels correlate with and even predict declining immune function and worsening physical symptoms. (Several reports1,2 already suggest that this is true.) It is also possible to chart the decline in HIV load that occurs immediately after someone begins a new antiviral therapy.3 Whether this artificially induced drop in virus produces better health is a more knotty issue, however. Consistently low CD4 cell counts are considered a very accurate predictor of future physical symptoms. But it turns out that raising CD4 count over the short term through antiviral therapy has had disappointing results. Physical condition has not improved correspondingly during clinical trials.4 (In one study of people with advanced disease, though, trial participants experienced an increase in survival greater than their CD4 count rises would indicate.5)

Resolving this issue runs up against the modest, transient effects that current anti-HIV medications have. Stronger, more sustained increases in CD4 counts, for example, might have an incontestably positive effect on physical health. Also, HIV infection focuses on the center of immune activity in the lymph nodes and other lymphoid tissue, not in the blood. Detecting a lowering of HIV in the bloodstream may not reflect what is occurring in the lymph system. One study presented at the Tenth International Conference6 found that in people with high CD4 counts (average of 654), initiating AZT therapy did not alter the production of new HIV particles within lymph nodes or the number of infected cells. Adding ddI to prior AZT therapy did decrease new HIV production in the lymph nodes in four of six patients, but this decrease was not statistically significant.

An underlying problem is that HIV is a subtle infection because there are no outward symptoms of immune deficiency per se. Directly monitoring a treatment's effect requires some way of precisely describing immune function. CD4 cell count by itself is not sufficient because it does not indicate how well those cells or the immune system as a whole are functioning. Even if therapy-induced reductions in the amount of blood-borne HIV do trigger a rise in CD4 counts, this does not guarantee that the immune system will recover.

Surrogate Marker Conference Reports

In an effort to find a way out of this labyrinth, many of the chief contributors to recent developments in the surrogate marker field -- including various measurements of HIV activity and immune function -- attended a three-day conference in Arlington, Virginia from October 12 to 14, 1994.

The presentations at the conference comprised a vast review of the subject, although many of the data had been released previously, at the Tenth International Conference and elsewhere. [The studies mentioned below were reported at both conferences, although in greater detail at the October surrogate marker event. Footnote references will refer to the International Conference Abstract Book, volumes I and II, because this is a published, publicly available source.]

Walter Reed Army data: One of the chief reports supporting the utility of viral load assays comes from a. U.S. Army study7 investigating the relationship between HIV level and the emergence of AZT resistance in those on long-term AZT therapy. The 100 trial participants had a mean CD4 count of 252 and were followed for two to three years. As reported by Douglas Mayers, M.D., of the Walter Reed Army Institute of Research, researchers found that emergence of AZT resistance was associated with HIV levels of more than 100,000 copies of HIV RNA per milliliter of blood plasma.

The strongest predictors of death in the group turned out to be low CD4 count and even more significantly, HIV levels of more than one million HIV RNA copies per milliliter of plasma. Another Army study8 presented by Dr. Mayers followed 32 individuals over two to three years after they had switched from AZT to ddI monotherapy. In this group, only low CD4 count (less than 100) was a statistically significant predictor of death, with high plasma HIV load (an RNA copy number of greater than 300,000 per milliliter) being not quite significant.

Veteran Affairs review: A report9 suggesting that reductions in HIV level brought on by therapy do indeed lead to clinical improvement has been prepared by William O'Brien, M.D., of the Veteran Affairs Medical Center in Los Angeles, and his colleagues in a nationwide trial known as VA CSP 298. The group analyzed old blood samples from this trial, whose original results were published two years ago.10 (VA CSP 298 was the first major study to conclude that AZT therapy early in HIV infection, before CD4 count drops below 200, delayed progression to AIDS but did not prolong survival compared to later treatment, starting when CD4 count falls below 200.)

Using a statistical model, investigators found that increases in CD4 count in response to treatment could account for only 37 percent of the observed delay in progression to AIDS from early AZT therapy. By comparison, meaningful reduction in virus levels seemed to account for 67 percent of the clinical benefit. But the study could say nothing about the association between reducing HIV load and prolonging survival. AZT's inhibition of HIV is too temporary to make a judgment on this score, according to Dr. O'Brien. "It's not just how low virus load goes, but for how long," he notes.

(Trial participants in the early treatment arm started AZT at the beginning of the trial and experienced a 70 percent average decrease in HIV level by the second month. This level returned to baseline by the eighth month on average.) ACTG look-back: Another review of an older study came to similar conclusions as the VA CSP 298 review. At the October surrogate markers meeting, Seth Wells, Ph.D., Sc.D., a senior statistician at the Harvard University AIDS Clinical Trail Group Analysis Center, described findings gleaned from blood samples taken in the course of the ACTG trial 116B/117. This trial tracked the effect of switching to ddI after at least four months of AZT therapy. High HIV levels at study entry were associated with more advanced disease and quicker disease progression, and those whose HIV levels increased had a greater risk of further progression regardless of their initial viral load.

Trial participants who continued to receive AZT experienced on average a persistent increase in HIV load whereas those who were changed to ddI had a modest reduction (of one-half logarithm or about three-fold) that bottomed out at week eight. Interestingly, this decrease was not reflected in any appreciable CD4 cell augmentation. This result differed from the results in ACTG 116A and VA CSP 298, which included people who had not yet started therapy and whose immune deficiency was less advanced.

In an interview, Dr. Wells observed, "I'm encouraged by the data, but the patients in the 116B/117 trial were a very select population [with advanced disease] and you can't generalize from that data to other groups, including the people in 116A."

Complicating Factors

An indicator that works on a generalized population level might not work when applied mechanically to individuals because each personal response to HIV infection differs from the composite average. One of the co-investigators in ACTG 116-117, Robert Coombs, M.D., Ph.D., of the University of Washington, described some of the interacting factors that might explain different individual and subgroup outcomes to changing HIV burden.

The first factor is initial virus load. In people with low levels of HIV particles in their blood -- RNA copy numbers of less than 10,000 per milliliter -- there is little detectable treatment-induced reduction, and changes in CD4 count are related to changes in CD4 proliferation rates and the virulence of a patient's particular viral strain. (The appearance of a so-called syncytia-inducing, or SI, strain of HIV is one major sign of increased virulence.) In ACTG 116B/117 (remember, these were people who were previously taking AZT and had advanced immune deficiency), many of the patients who were switched to ddI experienced some reduction in HIV load and an accompanying increase in CD4 count. The trial participants kept on AZT largely saw just the reverse: a continuing increase in viral load and a decrease in CD4 count. This is a sign of AZT's waning benefit over time.

There were some participants in 116B/117, though, who perversely had an increase in both viral load and CD4 count. Dr. Coombs thought this phenomenon was due to higher CD4 cell replication rates in such people. Another subgroup witnessed a simultaneous decline in both viral load and CD4 cell count. Most of these people possessed a virulent SI strain of HIV, Dr. Coombs noted.

All these interacting elements make it hard to determine the role viral load can play in clinical trials, and harder still in individual treatment decisions. Dr. Wells' group at Harvard is formulating ways to validate viral load as a surrogate marker. He says, "Viral load seems like a great marker for phase I and II testing [for screening new drugs for anti-HIV activity]. But we will need to follow extraordinary numbers of people to assess how well viral load works as a surrogate for [clinical] outcomes. The size of the studies has to do with the magnitude of the treatment effects we're observing."

Dr. O'Brien concurs in this assessment. "There is not enough data to change our whole policy," he says. Before viral load can enter into wide-scale use as a predictor of increasing immune deficiency in HIV infection, its relationship to physical symptoms will have to be convincingly demonstrated. The rate of opportunistic conditions has long been considered the "gold standard" for gauging therapies' benefits. Given the lack of a more direct measure of immune function, it is considered a real measurement of disease progress and ranked as a "clinical" endpoint rather than a surrogate.

Difficulties with Clinical Endpoints

Frank Rhame, M.D., an AIDS researcher who heads the HIV clinic at the University of Minnesota, goes a step further and argues, "You have to ground surrogates in studies showing a survival difference not in those counting up AIDS-defining events, which is itself a surrogate marker." At the October surrogate marker conference, Dr. Rhame explained in great detail the practical problems that have arisen when relying on the incidence of new opportunistic conditions to decide the outcome of clinical trials.

One of the major difficulties encountered when interpreting data on opportunistic conditions is that there is no ranking system for judging the seriousness of illnesses, which can range from comparatively minor candida infections to life- threatening conditions. Also, trials usually count only the first event during follow-up, ignoring whether a given individual experiences a whole string of infections or not. Finally, trial sites may misdiagnose illnesses or miss them entirely when trial participants fail to return for follow-up -- frequently because they are too sick.

Unfortunately, survival has its own drawbacks for evaluating the effect of antiretroviral drugs. It is time-consuming and unethical to wait for a sufficient number of deaths before judging the utility of a therapeutic technique. In addition, death depends on the history of opportunistic conditions, not just on HIV, and this history is affected by therapies unrelated to antiviral medications and their effect on surrogate markers for immune decline. Another concern is that focusing on survival really focuses on the rapid progressors who provide a disproportionate number of a trial's fatal events. Rapid progressors' performance in a trial may not be representative of what happens in other subgroups.

Many Markers, Few Treatments

The surrogate marker conference heard presentations describing the utility of at least seventeen candidate surrogate markers, including various measurements of HIV level and virulence, lymph node condition and immune status. Different kinds of therapy affect the AIDS disease process in different ways, so probably no one marker will prove universal enough to be used alone. Viral load and CD4 count for now promise to be the most applicable and informative combination, but it will take considerable effort to sort out the contribution all these different lab values can make, both for testing new drugs and determining patients' individual treatment strategies.

Thomas Merigan, M.D., a co-chair of the surrogate marker conference, said, "We're trying to use surrogate markers to license drugs with slight effect. How can we be definite?" Evaluating strong treatments can be accomplished in a clear- cut manner. The lack of such treatments creates both the need for surrogate markers and the uncertainties concerning their application.

1 Saksela K et al. Proceedings of the national academy of science USA. Feb 1 1994; 91(3):1104-8.

2 Mellors J et al. First National conference on human retrovirus infections. Dec 1993; (abstract 274):103.

3 Dewar RL et al. Journal of infectious diseases. Nov 1994; 170(5)1172-9.

4 Fleming T. Statistics in medicine. Jul 30 1994; 14(14):1423-35.

5 De Gruttola V et al. Journal of acquired immune deficiency syndromes. Apr 1993; 6(4):359-65.

6 Cohen O et al. Tenth international conference on AIDS abstract book. Aug 7-12 1994; I(abstract 001B):7.

7 Vahey MT et al. Tenth international conference on AIDS abstract book. Aug 7-12 1994; I(abstract 253B):75.

8 Mayers DL et al. Tenth international conference on AIDS abstract book. Aug 7-12 1994; II(abstract 359B):17.

9 O'Brien WA et al. Tenth international conference on AIDS abstract book. Aug 7-12 1994; I(abstract 254B):75.

10 Hartigan PM et al. New England Journal of Medicine. Apr 16 1992; 326(16): 1037-42.

Copyright (c) 1994 - Gay Men's Health Crisis. Noncommercial reproduction encouraged. DISTRIBUTED BY GENA/aegis * 714.248.2836 * 8N1/Full Duplex * v.34.
9411
GM081002

Copyright © 1994 - Treatment Issues. Reproduced with permission. Treatment Issues is published twelve times yearly by GMHC, Inc. All rights reserved. Noncommercial reproduction is encouraged. Subscription lists are kept confidential. GMHC Treatment Issues, The Tisch Building, 119 West 24th Street, New York, NY 10011  fredg@gmhc.org  http://www.gmhc.org

AEGiS is made possible through unrestricted grants from Boehringer Ingelheim, Elton John AIDS Foundation, iMetrikus, Inc., John M. Lloyd Foundation, the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2003. This material is designed to support, not replace, the relationship that exists between you and your doctor.

AEGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.

Copyright ©1980, 2003. AEGiS. All materials appearing on AEGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of AEGiS, or the party credited as the provider of the content. .