The popularity of nonconventional therapies,
for a myriad of diseases, has increased dramatically. Most patients
use some form of alternative therapy, often concurrently with
conventional treatment and frequently without advising their
conventional health care provider. Relying on media reports, Internet
advertising, and industry marketing, many patients believe that
nonconventional therapies offer cures for literally every disease,
including cancer; that they do not interfere with other treatments;
and that they are uniformly free of toxicity at any dosage level.
Dietary antioxidants have received increasing attention from the
scientific, clinical, and nutritional foods community.[2-6] Some of
that attention has focused on the concurrent use of dietary
antioxidants, such as alphatocopherol and coenzyme Q10 with
chemotherapeutic agents, such as doxorubicin.[7-10] Claims that such
combinations represent novel cancer therapies, without mention of
potential interactions, prompt concern for their effect on long-term
outcomes. Since many patients treat themselves with oral antioxidants
during chemotherapy, clinicians need to formulate a credible position
on this subject if they are to provide their patients with timely
advice about the potential risks.
To date, no definitive human studies have demonstrated the long-term
effects of combining chemotherapeutic agents and oral antioxidants.
Fortunately, the mechanisms of action of both are understood well
enough to predict the obvious interactions and to suggest where
caution should be exercised with respect to both clinical decisions
and study interpretation.
This article will describe these potential interactions and areas of
concern, based on the available data. It will also suggest several
potential courses of action clinicians may take when patients
demonstrate an interest in alternative therapies.
Chemotherapeutic agents have many well-defined and suggested
mechanisms of actions.[12,13] Some chemotherapeutic agents, including
traditional alkylating agents and anthracycline antitumor
antibiotics, create reactive oxygen species. Reactive oxygen species
are uniformly subject to transformation to more stable compounds by
antioxidants through the simple process of electron transfer.[11,14]
Classical alkylators substitute an alkyl group for a proton in
organic tissue, as shown below:
RH + XCH2-CH2®
RCH2-Ch2 + H+ + X
This action is common at the 7 nitrogen atom of guanine, but other
sites, such as the 1 and 3 nitrogen atoms of adenine, the 3 nitrogen
atom of cytosine, and the 6 oxygen atom of guanine, may also be
alkylated. This alkylation of biologically vital macromolecules, such
as DNA, and the complex degradation reactions that follow result in
the known cytotoxic action of these drugs.
These cytotoxic actions are, however, vulnerable to interference. It
is known, for example, that greater concentrations of free thiol
groups are present in animal tumors with greater resistance to
alkylating agents. In addition, cysteine, a member of the thiol
group, can considerably reduce the antitumor effects of alkylating agents.
Thiol groups take the form, R1-S-H. Cysteine tends to lose
protons by ionization more readily than do other amino acids with
nonpolar R groups, and is often found in its oxidized form in
proteins, namely, cystine. It is likely that cysteine reduces the
cytotoxic effects of alkylating agents by interfering with the
alkylation process and subsequent reactions.
Other chemotherapeutic agents form different primary reactive oxygen
species. For example, doxorubicin, in hepatic microsomes in vivo,
increases the oxidation of nicotinamide adenine dinucleotide
phosphate (NADPH) and the transfer of electrons to molecular oxygen,
resulting in the formation of anion radicals O2.[15-17]
Antioxidant compounds have a wide variety of actions. They can
interfere selectively with free radical initiation, propagation, and
termination. They are a normal part of the human diet and have been
studied for their positive effects in the prevention or cure of some
cancers, cardiovascular disease, age-related diseases, and other disorders.[18-21]
Each antioxidant has individual actions that are often predictable.
For example, the enzyme superoxide dismutase reacts with superoxide
radicals and protons, as shown below:
2O2 + 2H ® O2 + H2O2
Copper, zinc, and manganese are essential metalloenzymes of
superoxide dismutase, which are considered antioxidants when they are
incorporated into superoxide dismutase.
The antioxidant enzyme catalase converts hydrogen peroxide to less
reactive oxygen and water, according to the following reaction:
Catalase requires iron. Yet, iron and copper are among the most
common biological oxidant catalysts, which readily transfer electrons
to oxygen and, thus, form reactive oxygen species.
Many antioxidant compounds are found in vivo. The normal antioxidant
activity derived from dietary sources is usually balanced by free
radicals generated from daily stresses, such as inflammation,
exercise, detoxification, certain chemicals, radiation, ultraviolet
light, alcohol, and fatty diets. This basal level of in vivo
antioxidant activity quenches reactive oxygen species and facilitates
repair of DNA damage, among other effects.
Dietary antioxidants can quench free radicals generated from many
sources, including chemotherapeutic agents.[23-26] This is the
principal mechanism of interaction considered herein affecting agents
that utilize reactive oxygen species for their antitumor effect. One
free radical removes electrons from stable compounds, thus creating
new reactive oxygen species, each with cytotoxic potential. The
reaction and associated cytotoxicity continue until stability is attained.
Premature stability can be reached when excess antioxidants interrupt
the chain reaction early, thus ending the cytotoxicity of that arm of
the reaction.[22,27] Such a reduction in concentration of free
radicals generated by chemotherapeutic agents has the same effect as
a reduction in dose.
Factors That May Predict Interactions
Existing data provide credible mechanisms for the interaction of
certain chemotherapeutic agents and dietary antioxidants. The six
factors discussed below are proposed as a basis for predicting
interactions between chemo-therapeutic agents and antioxidant
compounds (Table 1).
Fraction of Drug Effectiveness That Depends on Reactive Oxygen SpeciesMany
currently available chemotherapeutic agents have more than one
suggested mechanism of action, and some do not depend on oxidative
mechanisms for cytotoxicity. Alkylators, anthracyclines, and other
drug classes that utilize known oxidative mechanisms require
attention for potential free radicalantioxidant interactions,
whereas many antimetabolites, plant-derived drugs, and other agents
whose actions do not include generation of reactive oxygen species
are not likely candidates for this reaction.[11-13,16,17] Free
radicalantioxidant interactions can be highly specific.[8,22]
Nature of the Reactive Species Generated by the Chemotherapeutic AgentDifferent
chemotherapeutic agents generate different kinds of free radicals.
Protons, oxygen radicals, hydrogen peroxide, and other species have
unique interactions with tissue and antioxidant compounds. In
addition, the pharmacokinetics of the generated reactive oxygen
species warrant consideration. The distribution, metabolism, and
excretion sites of metabolites may differ from the parent compound.
The timing and half-life of reactive species can be unique as well.
Many chemotherapeutic agents, especially those that must be
metabolized to reach their active form, may vary considerably in
their effects in different individuals. For example, some patients
treated with cyclophosphamide (Cytoxan, Neosar) do not produce
carboxyphosphamide and, as a result, produce more than twice the
expected levels of highly reactive phosphoramide mustard.
Dosage and Concentration of Reactive Oxygen SpeciesThe
specific dosage and concentration of the reactive oxygen species can
determine the level of reaction and, to some extent, the clinical response.
It is well known that high-dose chemotherapy regimens can produce
better long-term outcomes even when measurable tumor response at the
lower dose is considered good. This difference may be due to the fact
that the high-dose regimen achieves a more complete cytotoxic
response, thus interfering with the ability of tumor cells to mutate
to resistant strains. It may also reflect the ability of higher in
vivo concentrations to more completely intercept and destroy micrometastases.[14,29,30]
Furthermore, the concentration of reactive oxygen species and its
relationship to maximum drug effect determines the extent to which
drug-nutrient interactions can interfere with this action. According
to the Michaelis-Menten model:
Effect = (Maximal effect × [D]) / (KD + [D])
As drug dose (D) approaches maximum effect, changes in dose have a
diminishing impact on effect. Based on this model, low-dose,
fractionated regimens would be predicted to be more vulnerable to
dosage variations resulting from drug-nutrient interactions.
Nature of the Antioxidant Since individual antioxidants
have specific actions when interacting with reactive oxygen species,
the action of the antioxidant can determine whether a reaction is
even possible. Moreover, the pharmacokinetics of the antioxidants, in
combination with those of the reactive oxygen species, can predict
whether the reactants will be in the same place at the same time.
The antioxidant mesna (Mesnex) is an excellent example of this
principle. Mesna is rapidly metabolized in vivo to mesna disulfide,
which undergoes rapid renal excretion. Its high antioxidant
concentration in the renal pathway interacts with the metabolite
acrolein, thus limiting the renal toxicity of classical alkylators,
such as cyclophosphamide and ifosfamide, without affecting their
Concentration of the AntioxidantAntioxidant activity is
a common, necessary part of many normal reactions in the body and is
provided in essentially all foods. There is always some antioxidant
activity in vivo, including during treatment with chemotherapeutic
agents of concern. Antioxidant intake at the levels present during
clinical trials should not affect outcome. The concentration of
antioxidants, together with the concentration of reactive oxygen
species, will determine the level of the reaction, likely according
to the Michaelis-Menten model.
Temporal Relationship Between the Antioxidant and ReactiveOxygen SpeciesThe
temporal relationship between antioxidant intake and the
administration of the chemotherapeutic agent of concern may be the
single best factor for predicting or ruling out interactions within
this model. There are, however, several caveats:
For drugs that undergo multiple transformations, these specific
pharmacokinetic data must be considered, since the product of the
first or second transformation may not be the one that is important.
When appropriate, a two-compartment model may be critical, since
tissue rather than plasma may be where significant drug-antioxidant
interactions are occurring.
The time course of drug effect can be affected dramatically by
impaired eliminationa not uncommon problem with this class of agents.
Markedly increasing tissue stores of antioxidants, when that is
pos-sible, may affect a drug-antioxidant interaction beyond the
period predicted by pharmacokinetics.
In vitro studies investigating the combination of chemotherapeutic
agents and antioxidants must be interpreted carefully. Human free
radical and antioxidant activity exist in a complex biochemical
framework that cannot be duplicated in a petri dish. Considering the
number of independent variables, the chances for both
false-negativeand false-positive interactions are very high.
New studies should account for the effects of interactions on
solitary tumor cells, not just on overt tumor response. Existing
models for the antitumor action of agents that generate reactive
oxygen species, the ability of antioxidants to reduce the numbers of
free radicals, and the relationship between concentration of reactive
oxygen species and its effect on micrometastases, as discussed above,
clearly suggest that a drugs ability to destroy micrometastases
may be impaired by the addition of antioxidants. Paradoxically, this
may result in an improved short-term tolerance to treatment followed
by an increased long-term chance for recurrence. For this information
to be clinically meaningful, it is critical to examine long-term
recurrence and survival rates.
1. Eisenberg DM, Kessler RC, Foster C, et al: Unconventional medicine
in the United States: Prevalence, costs, and patterns of use. N Engl
J Med 328(4):246-252, 1993.
2. National Nutritional Foods Association: Antioxidants Show Promise
in Colorectal Cancer. vol 11, p 6. Newport Beach, California,
National Nutritional Foods Association, 1997.
3. Lockwood K, Moesgaard S, Folkers K: Partial and complete
regression of breast cancer in patients in relation to dosage of
coenzyme Q10. Biochem Biophys Res Commun 199(3):1504-1508, 1994.
4. Dimirpence E, Balaguer P, Trousse F, et al: Antiestrogenic effects
of all-trans-retinoic acid and 1,25 dihydroxyvitamin D3 in breast
cancer cells occur at the estrogen response element level but through
different molecular mechanisms. Cancer Res 54:1458-1464, 1994.
5. van den Brandt PA, Goldbohm A, vant veer P, et al: Toenail
selenium levels and the risk of breast cancer. Am J Epidemiol
6. Hunter DJ, Morris JS, Stampfer MJ, et al: A prospective
study of selenium status and breast cancer risk. JAMA 264:1128-1131, 1990.
7. Shinozawa S, Kawasaki H, Gomita Y: Effect of biological membrane
stabilizing drugs (coenzyme Q1O, dextran sulfate and reduced
glutathione) on Adriamycin (doxorubicin) induced toxicity and
microsomal lipid peroxidation in mice. Gan To Kagaku Ryoho
8. Beyer RE: An analysis of the role of coenzyme Q in free radical
generation and as an antioxidant. Biochem Cell Biol 70(6):390-403, 1992.
9. Shinozawa S, Gomita Y, Araki Y: Tissue concentration of
doxorubicin in mice pretreated with alpha-tocopherol or coenzyme Q10.
Acta Med Okayama 45(3): 195-199, 1991.
10. Chinery R, Brockman J, Peeler M, et al: Antioxidants enhance the
cytotoxicity of chemotherapeutic agents in colorectal cancer: A p53
-independent induction of p21WAF/CIPI via C/EBP(beta). Nat Med
11. Chabner BA, Collins JM: Cancer Chemotherapy: Principles and
Practice, pp 276-297, 314-333. Philadelphia, JB Lippincott, 1990.
12. Black DJ, Livingston RB: Antineoplastic drugs in 1990: A review
(part I). Drugs 39:489-501, 1990.
13. Black DJ, Livingston RB: Antineoplastic drugs in 1990: A review
(part II). Drugs 39:652-673, 1990.
14. Henner WD, Peters WP, Eden JP: Pharmacokinetics and immediate
effects of high-dose carmustine in man. Cancer Treat Rep 70:877-880, 1986.
15. Connors TA: Mechanisms of clinical drug resistance. Biochem Pharm
23(suppl 2):89-100, 1974.
16. Kirkwood JM, Lotze MT, Yasko JM: Current Cancer Therapeutics, pp
1-6, 37-40. Philadelphia, Churchill Livingstone, 1996.
17. Macdonald JS, Haller DG, Mayer RJ: Manual of Oncologic
Therapeutics, pp 80-135. Philadelphia, JB Lippincott, 1995.
18. Enger SM, Longnecker MP, Shikany JM, et al: Questionnaire
assessment of intake of specific carotenoids. Cancer Epidemiol
Biomarkers Prev 4(3):201-205, 1995.
19. Jarvinen R: Carotenoids, retinoids, tocopherols, and tocotrienols
in the dietthe Finnish mobile clinic health examination survey.
Int J Vitam Nutr Res 65(1):24-30, 1995.
20. Taylor PR, Wang GQ, Sanford MD, et al: Effect of nutrition
intervention on intermediate end points in esophageal and gastric
carcinogenesis. Am J Clin Nutr 62(suppl):1420S-1423S, 1995.
21. Frommel TO, Sohrab M, Doria M, et al: Effect of beta-carotene
supplementation on indices of colonic cell proliferation. J Natl
Cancer Inst 87(23):1781-1787, 1995.
22. Garewal HS: Antioxidants and Disease Prevention, pp 19-26. New
York, CRC Press, 1997.
23. Erhola M, Kellokumpu-Lehtinen P, Metsa-Ketela T, et al: Effect of
anthracycline-based chemotherapy on total plasma antioxidant capacity
in small-cell lung cancer patients. Free Radic Biol Med
24. Koracheva S, Ribarov SR: Lipid peroxidation in lung of rats
stresses by immobilization effects of vitamin E supplementation. Lung
25. Ihara Y, Mon A, Hayabara T, et al: Free radicals, lipid
peroxides, and antioxidants in blood of patients with myotonic
dystrophy. J Neurol 242(3):119-122, 1995.
26. Abbey M, Noakes M, Nestel PJ: Dietary supplementation with orange
and carrot juice in cigarette smokers lowers oxidation products in
copper-oxidized low-density lipoproteins. J Am Diet Assoc
27. Bast A, Haenen GR, Doelman CJ: Oxidants and antioxidants: State
of the art. Am J Med 91(3c):2S-13S, 1991.
28. Hadidi AH, Coulter CE, Idle JR: Phenotypically deficient urinary
elimination of carboxyphosphamide after cyclophosphamide
administration to cancer patients. Cancer Res 48:5167-5171, 1988.
29. Culvert AH, Newell DR, Gore HE: Future directions with
carboplatin: Can therapeutic monitoring, high-dose administration,
and hematologic support with growth factors expand the spectrum
compared with cisplatin? Semin Oncol 19(suppl 2):155-163, 1992.
30. Corden BJ, Fine RL, Ozols RF, et al: Clinical pharmacology of
high-dose cisplatin. Cancer Chemother Pharmacol 14:38-41, 1985.
31. Dechant KL, Brogden RN, Pillington T, et al: Ifosfamide/mesna: A
review of antineoplastic activity, pharmacokinetic properties, and
therapeutic efficacy in cancer. Drugs 42:428-467, 1991.