THE MEANING OF CD4 COUNTS

By Subhuti Dharmananda, Ph.D., Director of Institute for Traditional Medicine, Portland, Oregon
July 1996.

The bone marrow is responsible for the production of the three primary types of blood cells--red
(erythrocytes), white (leukocytes), and platelets (thrombocytes). One large subgroup of the white blood
cells are the T-cells (so called because they mature in the thymus gland). Of the many types of T-cells,
there is one subgroup which has a component on its surface called CD4 (these also have a component
called CD3). The "CD4 count" is the number of these T-cells found in a microliter (less than a drop) of
blood. Only about 2% of the CD4 T-cells in the body are in the circulating blood at any time; the other
98% are tied up in lymphatic tissue. Thus, a count of 200 CD4 cells translates to about four hundred
million of those cells in circulation and a total of about two billion such cells altogether.

The number of CD4 T-cells circulating in the peripheral blood (so called, because it is sampled from
blood vessels near the skin) is strongly affected by HIV replication. HIV binds to the CD4 cell surface
component as one step towards entering the T-cell. Once it enters, it can replicate within that cell. When
it does replicate, it destroys the T-cell; further, it may set up a process (e.g. autoimmunity or cell
signaling) by which the immune system itself destroys additional CD4 T-cells. Researchers discovered a
decade ago that CD4 levels roughly correlated with disease progression in those infected by HIV. That is
why it has been so widely used as a basis for monitoring both the disease and any treatments. More
recently, it was shown that if a drug therapy could dramatically decrease the level of HIV in the blood
("viral load"), the CD4 count could rise; if the drug was then removed, the level of HIV would then
rapidly increase and the CD4 level would fall. However, the same thing happens if the drug therapy is
continued long enough, because HIV develops resistance to the drug unless two or three drugs are used
simultaneously. These short-term changes in CD4 counts are often limited by the extensive physiological
effects of prolonged HIV infection.

To count CD4 cells, it is necessary to separate the various blood cells within a sample and identify those
which have the CD4 cell surface component. This is done by machine, usually relying on a flow
cytometry device (a separation system). Unfortunately, there can be substantial variation in CD4
measurements. For this reason, one should be cautious of responding too drastically to either an increase
or a decrease from one test to another. This variability needs to be discussed in some detail before
describing interpretation of laboratory results.

If a sample of blood is divided in two and sent to two different laboratories (which use two different
pieces of equipment), the value obtained can be markedly different. For example, a reading of about 300
at one lab, could come out as 390 at another lab, or 230 at another. For those who follow CD4 levels very
closely, this variability can be unnerving. Further, if one were to take a blood sample, divide it in two,
and measure it twice at the same lab, there can be substantial variation; estimated to be as much as 20%
(thus, a reading of 300 the first time the blood is assayed could easily show up as 360 or 240 the next
time, without even hinting at any actual biological difference). The CD4 levels in the body can respond to
a number of physiological factors, including the time of day (the levels tend to be lower in the morning
and higher in the afternoon) and the body's response to infection. As one physician has said, "if you take a
walk around the block and get retested, the results will be changed." Due to the high concentration of
CD4 in the lymph tissue, a reading on one day and a reading on the following day or week can vary
substantially because of changes in lymph node conditions---not just laboratory factors. The normal range
of CD4 counts in persons without HIV infection is quoted differently by various sources, but is typically
600-1,200.

In general, two readings that are taken only a few months apart and give radically different results, are
likely to indicate either an error or a statistical fluctuation unless a clear explanation accompanies the
change (e.g. starting or ending a highly effective therapy or experiencing an opportunistic infection).
Small changes between two readings a few months apart are to be expected and are without clinical
significance. For example, a reading of 320 one time and a reading of 280 the next time may seem like a
major decline in CD4 count. In fact, the amount of CD4 T-cells in the peripheral blood may have been
about 300 both times, but measurement variability led to a slightly high reading the first time and a
slightly lower reading the second time. In a study of variation of CD4 counts (Journal of Infectious
Diseases 1994), several hundred patients who participated in a trial (at several sites) and who were in the
placebo group, were evaluated at the time of entry and eight weeks later: 6% (1 in 16) of the patients
monitored had their eight week CD4 counts either less than half or more than double the initial counts,
while the average value calculated for the group hardly changed Further, there was an observed tendency
for those who had counts that were relatively high or low within a CD4 range at the initial reading, to end
up closer to the mean value after two months, as usually occurs when the readings are influenced by
many variables.

It is not the case, as some have surmised, that CD4 levels have been relied upon by the medical
profession as an absolute and reliable indicator of health status. However, in the absence of better
indicators, the CD4 levels have been used to suggest when to initiate certain therapies (mainly
prophylaxis for PCP, to be initiated as the CD4 level drops towards 200) and to give patients an
indication (based on available statistics) of their prognosis. Today, viral load tests (PCR or branched
DNA) are used as a more reliable measure of disease status and prognosis.

The CD4 T-cells are not a very good marker of overall immune function in an individual---as
demonstrated by those individuals with low counts who maintain relatively good health. Still, the CD4
level is a reasonably accurate marker for a large group of persons to indicate susceptibility to certain
opportunistic infections and to disease progression. Based on observations of large numbers of persons
with HIV infection, the CD4 levels of 500, 200, 100, and 50 are often used to signify increasing levels of
susceptibility to infections. If a large number of people with HIV are tracked over an extended period of
time, there is a significant correlation between the average CD4 count and the frequency of various
events, including experience of first opportunistic infection, experience of specific opportunistic
infections (pneumocystis or CMV), and death.

Rapid decline in CD4 levels, which is suggestive of increased HIV activity or the body's lowered ability
to replace CD4 cells destroyed by HIV, correlates with a higher risk of experiencing a first AIDS-
defining infection. CD4 levels may drop gradually or in a step-wise fashion, with substantial decline
followed by periods of stability at the new level. Overall, an average rate of decline is about 15% per
year.

During the past five years, it has been recognized that there are other measurements which, when taken
together with CD4 level, give a much better picture of disease progression, immune status, and chances of
survival. These measures include the percentage of CD4 cells compared to all T-cells, the level of beta-2-
microglobulin, the level of cytotoxic CD8 cells (which is a subset of the total CD8 cells), and the blood
levels of HIV (viral load). However, for an individual who has all these items measured, there is still
much to contend with in making an interpretation. The uncertainty in deciding a therapeutic reponse to
the data coupled with the huge amount of data about CD4 levels, is why falling back on CD4 as a basic
measure has been attractive to most physicians. Even when the other measures are taken into account,
each person has unique circumstances that can influence health and survival.

Since drug therapies have various side-effects, they are instituted only when blood tests (or experience of
an opportunistic infection) suggest that they are needed. If the CD4 levels are relatively high, a drug
therapy is usually delayed; this may change, however, with development of new drugs. There had been an
opinion that starting antiretroviral therapy earlier might be the best way to prevent serious illness later---
which is still considered the theoretical ideal. It has been known for some time now that HIV can develop
resistance to inhibitory drugs. But it was clearly demonstrated, only recently, that the drug resistance
develops within a few days or weeks. Therefore, "monotherapy," the use of a single inhibitory drug, is no
longer deemed a useful treatment; it is now understood that the reason why early intervention with AZT
or other antiretroviral drugs did not produce increases in lifespan was largely because HIV continued its
activity during most of the time that the drug was being used.

In a long-term follow-up of AZT therapy on approximately 1,000 persons, it was shown that following an
initial increase of CD4 (from mean value of 308 to about 380) a fairly steady decline occurred at the rate
of 16% per year for three years, the same rate as occurs without the drug. The decline in the average CD4
counts was evident within six months after starting AZT therapy (the decline may have started earlier, but
monitoring was infrequent), and the CD4 value had returned to the pretreatment level in 16 months. In
this trial, AZT provided an average 16 month respite in steady decline of CD4s from the initial value, but
some decline was occurring through most of that period. As shown in other studies, much of that 16
month advantage is lost after AIDS develops, with the result of no increase in longevity. When using two
drugs together to inhibit HIV over the short term, it is shown that HIV replication decreases and CD4
counts rise day by day, over a period of about one month. Long-term effects (beyond one year) indicate
drug resistance problems. Only triple drug therapies appear immune to resistance problems during a one
year period.

In conclusion, CD4 data can be useful for various purposes as long as the uncertainties of examining any
one test in a patient are considered. An individual test can be quite misleading because of various
measurement problems, but regular testing (it is common to have tests every 3 to 6 months) can be used
to weed out some of the unusual readings. Using other measures at the same time, especially the viral
load, may give a better picture of immune status. CD4 levels can not predict what any individual will
experience, they can only be used in terms of large group statistics when examining outcomes.

REFERENCES

Hughes, MD, et.al., Within-subject variation in CD4 lymphocyte count in asymptomatic HIV
infection: implications for patient monitoring; Journal of Infectious Diseases 1994; 169: 28-36.

Boutitie, F and Pocock, SJ, Predictive value of repeated measurements of CD4 lymphocyte
counts on progression to AIDS; AIDS 1994, 8:35-41.

Vella, S, et.al., Long-term follow-up of zidovudine therapy in asymptomatic HIV infection:

results of a multicenter cohort study; Journal of Acquired Immune Deficiency Syndromes 1994;
7:31-38.

© 1996, Institute for Traditional Medicine, Portland, Oregon

EXPERIENCE COUNTS
Institute for Traditional Medicine (ITM)

Excerpts from the article Survival of patients with AIDS depends on physicians' experience treating the
disease; March 13, 1996.

"After adjusting for CD4 cell counts and severity of illness... the study showed a 46% decrease in relative
risk of death at any given time after clinical AIDS diagnosis in patients cared for by the most experienced
physicians compared with patients cared for by physicians with no previous experience [in treating
AIDS]."

"Patients of the experienced physicians were more closely monitored and treated more aggressively than
those cared for by less experienced clinicians...Early analysis shows that the more experienced physicians
used a significantly greater number of specialty consultations and had more outpatient visits for their
patients."

"Median survival of the patients cared for by the least experienced physicians was 14 months, compared
with 21 months for those treated by moderately experienced physicians and 26 months for patients
managed by the most experienced."

Good care for HIV disease includes careful and frequent monitoring, quick response to detected
problems, use of the appropriate (and sometimes, most advanced) therapies, and understanding of relative
benefits and risks of the possible interventions. The study results described above involved orthodox
medical treatments; additional benefits might be obtained through complementary approaches, such as
acupuncture, massage, chiropractic, nutritional supplements and herbs, especially as they involve frequent
monitoring and rapid response to changes in a person's condition. Practitioners with greater experience
will likely yield better results. The Institute for Traditional Medicine maintains a list of major HIV
treatment centers that provide Chinese medicine and other natural healing techniques and a Practitioner
Reference Guide to assist in finding private practitioners with a high level of experience. See the guide
"This Is What Works for HIV" for additional tips to enhancing quality of life and survival.

© 1996, Institute for Traditional Medicine, Portland, Oregon