MARIJUANA AND IMMUNITY
Journal of Psychoactive Drugs Vol. 20 (1), Jan--Mar 1988, 3-8.
Leo E. Hollister, M.D., Professor of Psychiatry and Pharmacology,
University of Texas Medical School, Harris County Psychiatric Center, P.O.
Box 20249, Houston, Texas 77225
Few areas of scientific research have been as controversial as the
effect of marijuana on immune defenses. The effects of marijuana on health
in general have been marked by polarities of belief or interpretation of
evidence often due to the particular prejudices of investigators. In
addition, evidence of altered immune functions is derived mainly from in
vitro tests or ex vivo experiments, which employed doses of cannabinoids
far in excess of those that prevail during social use of marijuana.
Finally, the clinical significance of the experimental observations is
difficult to assess.
The present review will attempt to objectively assess the evidence.
Other recent reviews of the subject have also appeared, with varying
degrees of intensity of coverage (Hollister 1986; Maykut 1985; Munson &
Fehr 1983; Rosenkrantz 1976). For purposes of a more systematic
discussion, immunity will be considered as several separate topics: (1)
cell-mediated immunity, (2) humoral mechanisms, (3) cellular defenses,
and (4) immunogenicity of marijuana or cannabinoids.
Lymphocytes exposed to several mitogens divide rapidly, increase
protein and nucleic acid synthesis, and show morphological changes
resembling blasts. This test of the ability of T-lymphocytes to transform
themselves measures one potential aspect of cell-mediated immunity. A
direct way to measure the activation of lymphocytes is to measure the rate
of incorporation of a nucleic acid, such as 3H-thymidine, into the cells
following addition of the mitogen to the culture. All studies are
conducted in vitro.
An early study (Nahas et al. 1976) measured 3H-thymidine uptake
in normal human lymphocytes stimulated by both phytohemagglutinin (PHA)
and allogenic cell mixed lymphocyte culture (MLC). The incorporation of
3H-thymidine was equally inhibited by 10 -5M to 10 -6M concentrations of
delta-9-tetrahydrocannabinol (THC), delta-9-tetrahydrocannabinol
(delta-8-THC), their corresponding 11-OH metabolites, a variety of other
inactive cannabinoids, and olivetol. THC in similar concentrations also
depressed 3H- leucine and 3H-uridine uptake, indicating an effect of
protein and RNA synthesis as well. These concentrations of THC were 10 to
20 times greater than those reported earlier by the same group (Nahas et
al. 1974) as having similar effects. In this study, cell-mediated
immunity was evaluated in 51 young, chronic marijuana smokers whose
lymphocytes were stimulated in vitro by PHA and MLC. As compared with
normal controls, 3H-thymidine uptake was reduced.
Klein and colleagues (1985) added the predominantly T-cell
mitogens, PHA and concanavalin A (Con A), and the B-cell mitogen, E. coli
lipopolysaccharide (LPS), to mice spleen cells treated with varying
concentrations of THC and its active metabolite 11-OH-THC. Both
T-lymphocyte and B-lymphocyte proliferation in response to mitogens were
suppressed by THC, but considerably less by 11-OH-THC. Proliferation of
both types of lymphocytes was completely inhibited by concentrations of
(10 mcg / ml) that were not directly lytic to the cells. Lower
concentrations of THC were found to inhibit B-lymphocytes than those
required for T-lymphocytes, suggesting that humoral immunity was more
impaired than cell-mediated immunity in this system.
By no means have all studies of cell-mediated immunity in marijuana
smokers or in vitro exposure of T cells to cannabinoids---often conducted
in exactly the same way--- shown evidence of immunosuppression. Indeed,
the inconsistency of study findings has led to the present state of
White, Brin and Janicki (1975) obtained peripheral blood
lymphocytes from 12 healthy long-term marijuana smokers. The blastogenic
response to PHA and pokeweed mitogen were measured in vitro by 3H-thymidine
uptake. The responses of lymphocytes from the marijuana smokers were not
significantly different from those who did not smoke the drug.
Lau and colleagues (1976) observed eitht chronic smokers of
marijuana in a hospital setting over a 30-day period. Each subject
received a placebo during the first six days, followed by THC in oral doses
up to 210 mg / day for the next 18 days, and then a placebo for the last
sic days. The response of their lymphocytes to PHA stimulation, as
measured by 3H-thymidine uptake, was no different in either of the three
Rachelfsky and Opedz (1977) stimulated normal human lymphocytes
with PHA and with MLC, and 3H-thymidine uptake was measured. The uptake
thymidine was unchanged in lymphocytes exposed to 1.9 X 10 -4M or 12.0 X
-4M concentrations of THC. Higher concentrations of THC precipitated in
the medium. Changes were comparable in cells exposed to THC and in those
not so exposed.
Kaklamani and colleagues (1978) obtained peripheral lymphocytes
from 12 chronic users of marijuana and 15 nonusing control subjects.
Lymphocytes from the experienced marijuana users were obtained both before
and after smoking hashish. Incorporation of 14C-thymidine, after PHA
stimulation and formation of rosettes of sheep erythrocytes, was no
different between the normal controls an the marijuana users either before
or after the latter had smoked hashish.
Whatever immunosuppressive effects marijuana may have, they are not
dependent on psychoactive components. A variety of cannabinoids, which
have no apparent central nervous system activity, share an apparent
immunosuppressive action (Smith et al. 1978).
T-Lymphocyte Rosette Formation
Another commonly used measure of cell-mediated immunity is the
ability of T-lymphocytes to form in vitro rosettes of sheep erythrocytes
surrounding T cells. A dose related decrease in rosette formation was
found in sensitized T cells exposed in vitro to various concentrations of
THC in this medium (Cushman, Khurana & Hashim 1976). Cushman and Khurana
(1977) tested 10 subjects during a four-week cycle of marijuana smoking,
so that the subjects were exposed chronically rather than acutely, and the
results showed a decrease in early T-cell rosette formation, but no change
in either late T-cell or B-cell rosettes. When Gupta, Grieco and Cushman
(1974) compared 23 chronic marijuana smokers with 23 nonsmokers, T-cell
rosettes were decreased in the users as compared with the nonusers, but
rosettes associated with B-lymphlcytes were not different, suggesting a
selective effect on cell-mediated immunity.
Petersen, Grahan and Lemberger (1976) tested three subjects who
smoked "street" marijuana for rosette formation and blastogenesis.
of the three showed decreased rosette formation and impaired blastogenesis
following stimulation of their lymphocytes with PHA. In another trial,
rosette formation was measured in six persons three to six hours after they
smoked a marijuana cigarette containing 10 mg of THC. Rosette formation
was impaired in five of the subjects, and became normal 24 hours later in
all but one subject. Thus it appeared that the effects of marijuana on
T-lymphocytes are variable and reversible, suggesting that factors other
than exposure to marijuana itself may be involved.
Mice immunized with sheep erythrocytes were treated with
intraperitoneal doses of 10, 25, and 40 mg / kg / day dose (Lefkowitz et
al. 1978). Monkeys were exposed to three levels of marijuana smoke over
six-month period. Plasma immunoglobulins (IgG and IgM) were decreased in
those monkeys exposed to medium and high concentrations of smoke. In vitro
tests by Dual and Heath (1975) of the response of lymphocytes to Con A
were decreased. Thus, both humoral and cell-mediated immunity appeared to
be affected. However, the authors asserted that it is impossible to assess
the in vivo implications from tests of this sort.
Cushman and Khurana (1977) tested 10 subjects during a four-week
cycle of marijuana smoking, so that the subjects were exposed chronically
rather than acutely. The results showed a decrease in early T-cell rosette
formation, but no change in either late T-cell or B-cell rosettes.
These studies also indicated that T-lymphocyte function, as
measured by rosette formation, was decreased when the cells are exposed
cannabinoids either in vitro or ex vivo. However, these impairments were
Other Measurements of Cell-mediated Immunity
A number of other measurements of cell-mediated immunity have
pointed in the same direction. Although both impaired allograft rejection
and decreased hemagglutinin titers were found in animals treated with
cannabinoids, the effect on allograft rejection as a measure of
cell-mediated immunity was greater (Munson et al. 1976). Susceptibility
to infection with herpes simplex virus type 2 applied directly to the
vagina was increased in mice that had received doses of mg / kg / day of
THC (Mishkin & Cabral 1985). A similar increased susceptibility was
found in guinea pigs treated with doses of 4.0 and 10 mg / kg / day (Cabral
et al. 1986).
Moorhan and colleagues assessed the LD-50 dose of Listeria
monocytogenes in mice treated with THC in doses of 38, 75 and 150 mg / kg.
The LD-50 was decreased 10-, 17- and 657-fold by each dose, respectively.
Marijuana extract was less active. A similar challenge with herpes simplex
type 2 virus showed a 96-fold decrease following administration of THC,
with no decreased resistance following administration of marijuana extract.
These situations are not at all comparable to human exposure.
The vast majority of people can be made sensitive to
dinitrochlorobenzene (DNCB), a powerful skin sensitizer. DNCB is often
used with "recall" antigens (e.g., tuberculin and mumps) to test
patients for anergy. Sensitivity to DNCB was found in all 34 chronic
marijuana smokers who were tested as compared with 96 percent of 279
healthy nonsmokers. On the other hand, 384 patients with cancer, whose
cell-mediated immunity is sometimes decreased, showed a positive reaction
in only 70 percent of those tested (Silverstein & Lessin 1974). Such
evidence raises questions about the clinical significance of experiments
that have shown evidence of cell-mediated immunity from cannabinoids.
It has been hypothesized that the membrane-disordering effects of
THC may affect the binding of antigens to cellular receptors, accounting
for a decrease in cell-mediated immunity. On the other hand, the
combination of increased membrane disorder and inhibition of
acyltransferase activity in B cells and T cells could impair the transfer
of cellular constituents (Baczynsky & Zimmerman 1983b). Regardless of
whether the action is a nonspecific one at the cell membrane or at a more
primary site, impaired immunity remains precisely that. However, a cell
membrane site of action could explain the apparent transitory nature of
observed alterations in cell-mediated immunity, as well as the requirement
for much larger concentrations of cannabinoids than those usually
encountered during social use of marijuana.
Summary of Effects of Cell-mediated Immunity
In summary, the effects of cannabinoids on cell-mediated immunity
are contradictory. Such evidence as has been obtained to support such an
effect has usually involved doses and concentrations that are orders of
magnitude greater than those obtained when marijuana is used by human
subjects. Clinically, one might assume that sustained impairment of
cell-mediated immunity might lead to an increased prevalence of malignancy,
as seen in the current epidemic of acquired immune deficiency syndrome
(AIDS). No such clinical evidence has been discovered or has any direct
epidemiological data incriminated marijuana use with the acquisition of
human immunodeficiency virus infection or the clinical development of AIDS.
Even though some degree of impairment of immune responses were to occur,
the remaining immune function may be adequate, especially in the young
persons who are the major users of cannabis.
Transformation of B-Lymphocytes
Transformation of B cells stimulated by the mitogen LPS was
inhibited more than were T cells stimulated by PHA following the same doses
of THC in mice (Klein et al. 1985). This evidence of diminished B-cell
reactivity following the administration of THC was confirmed in another
study (Munson et al. 1976) that showed a dose-dependent suppression from
doses of 50, 100, and 200 mg / kg of THC in mice. These doses are
enormous, of course.
A frequently used measure of humoral immunity is the ability of
splenic lymphocytes from mice that are immunized against sheep erythrocytes
to form plaques when exposed in vitro to sheep erythrocytes. Levy and
Heppner (1981) found that both THC and haloperidol produced
dose-dependent reductions in hemolytic plaque-forming cell (PFC) numbers
at the time of peak reactivity (day 4) in control mice. Treatment with
THC and haloperidol only delayed the time of peak PFC formation by 24 to
hours (doses were high enough to produce signs of gross behavioral
toxicity)> Neither THC nor other cannabinoids had any effect on the titer
of serum hemagglutinating antibody measured seven days after immunization.
Baczynsky and Zimmerman (1983a) immunized mice with sheep
erythrocytes on Day 1 (primary immune response) and on Days 1 and 28
(secondary immune response) and measured hemagglutinin titers. Mice
treated with 10 mg or 15 mg of THC during the primary immunization period
exhibited a suppression of the primary humoral immune response. These
doses also suppressed the secondary immune response, even when given during
the period of primary immunization. Mice treated with THC during the
secondary immunization period showed no measurable response. Other
cannabinoids had no effect.
Immature mice immunized with sheep erythrocytes also showed
suppression of the immune response when treated with THC in doses of 1.0,
5.0, and 10.0 mg / kg. Splenic weight was reduced and PFC as well as
hemagglutinin titers were lower than controls. Suppression was specific
for THC and was not observed with cannabidiol or cannabinol, even at doses
of 25 mg / kg (Zimmerman et al. 1977). Some evidence of tolerance or
hyporesponsiveness to this humoral antibody suppression by THC was found
when m;ice were treated with THC for five days prior to immunization as
well as afterward (Loveless, Harris & Munson 1981-1982).
Rosenkrantz, Miller and Esber (1975) immunized rats with a single
intraperitoneal dose of sheep erythrocytes during, before and after
administration of THC in order to determine its effect on the inductive
productive phases of the primary immune response. Following a dose of 10
mg / kg, THC decreased the primary immune response by 33 to 40 percent;
inductive phase was decreased by 48 to 78 percent by all doses of THC and
the productive phase was decreased by 26 to 59 percent by the higher doses.
The same group (Luthra et al. 1980) tested the primary immune
response of rats to intraperitoneal ad;ministration of sheep erythrocytes
after five to 26 days following pretreatment with THC in order to determine
if tolerance developed to the immunosuppressant effects. As measured by
splenic antibodly-forming cells and hemagglutinin/hemolysin titers, no
evidence of tolerance was found.
Summary of Effects of Humoral Immunity
In summary, humoral immunity, as tested by a number of in vitro
procedures, seems also to be impaired by cannabinoids, but this effect was
most evident for THC. The clinical significance of such changes is
questionable due to the great concentrations of cannabinoids used and the
lack of any epidemiological evidence of increased bacterial infections in
chronic users of marijuana.
Leukocytes and Lymphocytes
When 10 subjects were followed through a four-week cycle of
marijuana smoking, no change was observed either in peripheral leukocyte
absolute lymphocyte counts (Cushman & Khurana 1977). Leukocytes from
chronic marijuana smokers were compared with those from five nonusers of
the drug for their ability to migrate after exposure to THC or marijuana
extract. Both treatments inhibited leukocyte migration without killing the
cells, both in cells from users and nonusers of marijuana. The prevailing
THC concentration needed to accomplish this was 2.0 mcg / ml, a couple of
orders of magnitude greater than any THC plasma concentrations usually
found clinically (Schwartzfarb, Needle & Chavez-Chase 1974).
Natural Killer-cell Activity
Natural killer-cell activity in rats was decreased by subchronic
treatment (25 days) with THC, but not after acute treatment (one day).
This effect was not found in rats treated simultaneously with naloxone,
suggesting possible involvement of the opiate system (Patel et al. 1985).
When injected into mice, both THC and its active metabolite 11-OH-THC
suppressed splenic natural killer-cell activity in vitro. The tissue
concentrations of the cannabinoids were reported as being 5.0-10.0 mcg /
ml, about two orders of magnitude greater than those that might be
experienced during the social use of marijuana (Klein, Newton & Friedman
Macrophages work closely with T cells as part of the immunological
defense system. On glass surfaces, macrophage cultures normally show
spreading, which is an indication of their mobility. The addition of THC
to the medium inhibited the degree of spreading. It also inhibited the
phagocytosis of yeast particles (Lopez-Cepero et al. 1986). However,
another experiment (Munson et al. 1976) using intact mice that were
treated with a single dose or multiple doses of THC could not demonstrate
impairment in reticuloendothelial activity, as measured by the
intravascular clearance of colloidal carbon.
Summary of Effects on Cellular Defenses
It is somewhat surprising that newer techniques of cell sorting,
which permit determination of absolute counts of T- and B-lymphocytes as
well as subsets of T-lymphocytes, have not been utilized. The evidence
from the in vitro studies is weakened by the high concentrations of drug
that were used. Clinically, evidence for impairment of cellular defenses
has not been forthcoming.
IMMUNOGENICITY OF CANNABINOIDS
It has been hypothesized that THC, a relatively simple chemical,
can act as a hapten and become an immunogen. If such were the case,
tolerance to THC might be explained on an immunological basis as well as
the rare reports of allergic reactions. Azathioprine, an immunosupressant,
had a modest effect in mitigating the hypotensive effects of THC in
spontaneously hypertensive rats. Spleen cells from mice treated with THC
showed slightly more blast transformation in culture than untreated spleen
cells, either with or without THC being added to the culture medium.
However, the degree of blast transformation was far less than that produced
by PHA. This somewhat weak evidence for an immunogenic action of THC came
from a laboratory that later stressed the immunosuppresant effects of
marijuana (Nahas, Zugury & Schwartz 1973).
Watson, Murphy & Turner (1983) employed a technique used to test
compounds for their potential for producing allergic contact dermatitis
that also maximizes the degree of skin sensitization of guinea pigs.
Sensitivity was greatly increased by THC and cannabinol, but less so with
In a clinical study conducted in the southwestern United States
(Freeman 1983), skin tests were applied to 90 patients with various forms
of atopy. The test was positive for 63 patients for marijuana pollen as
compared to only 18 who had reacted to tobacco leaf. However, it is
unlikely that marijuana pollen contains many cannabinoids, but rather
contains proteins that may be sensitizing;
A series of 28 marijuana smokers showed predipitins for Aspergillus
antigens: 13 were positive as compared to one of 10 controls. Lymphocytes
showed significant blastogenesis in three of those subjects who tested
positive. Seven of these 23 subjects reported bronchospasm following the
smoking of marijuana, and one patient had evidence of systemic
aspergilliosis (Kagen et al. 1983). As it is well known that marijuana
contains contaminants, including molds and fungi, it is not surprising that
these should cause allergic reactions in some users. The study does not
indicate that cannabinoids themselves are immunogens.
Skin testing with cannabinoids seems to be useless for determining
the rare patient with sensitivity to marijuana. A variety of intradermal
tests with various cannabinoids and common allergens was applied to 63
marijuana users by Lewis and Slavin (1975). Two users, who were clearly
atopic with a past history of bronchial asthma, also reported experiencing
asthma after some exposures to marijuana. A third subject, with a history
of allergic rhinitis, also experienced similar symptoms following marijuana
use. All three of these subjects had negative skin tests to cannabinoids.
On the other hand, seven subjects who tested positive for hemp and one who
tested positive to delta-9-THC had no clinical manifestation of marijuana
A 29-year-old woman (known to be allergic to ragweed) experienced
symptoms of an anaphylactoid reaction that lasted 20 to 30 minutes
immediately after smoking marijuana. Skin tests with THC showed a 2+
reaction, and with cannabidiol a 1+ reaction (Liskow, Liss & Parker
1971). The weakly positive skin tests do not necessarily indicate that the
reaction was due specifically to the cannabinoids.
Summary of Immunogenicity
While it is possible that a few persons may become truly allergic
to cannabinoids, it is far more likely that allergic reactions, which have
been exceedingly rare following the use of marijuana, are due to
contaminants. Marijuana is grown in the field and harvested along with
everything else (e.g., bacteria, fungi, molds, parasites, worms,
chemicals) that may be found in such field plants. That such impure
material, when smoked and inhaled into the lungs, causes so little trouble
is really a marvel.
SUMMARY AND CONCLUSIONS
Despite the fairly large literature that developed during the past
15 years or so, the effect of cannabinoids on the immune system is still
unsettled. The evidence has been contradictory and is more supportive of
some degree of immunosuppression only when one considers in vitro studies.
These have been seriously flawed by the very high concentrations of drug
used to produce immunosuppression and by the lack of comparisons with other
membrane-active drugs. The closer that experimental studies have been to
actual clinical situations, the less compelling has been the evidence.
Although the topic was of great interest during the 1970's, as
indicated by the preponderance of the references from that period, interest
has waned during the present decade. This waning of interest suggests that
perhaps most investigators feel that this line of inquiry will not be
rewarding. The AIDS epidemic has also diverted the attention of
immunologists to the far more serious problem of the truly devastating
effects a retrovirus can have on a portion of the immune system.
The relationship between the use of social drugs and the
development of clinical manifestations of AIDS has been of some interest,
however. Persons infected with the virus but not diagnosed as AIDS have
been told to avoid the use of marijuana and / or alcohol. This advice may
be reasonable as a general health measure, but direct evidence that heeding
this warning will prevent the ultimate damage to the immune system is
The author wishes to thank Matthew Edlund, M.D., who provided
useful critical comments during the preparation of this article.