Impact of antioxidant supplementation on chemotherapeutic toxicity: A systematic review of the evidence from randomized controlled trials
Version of Record online: 11 JUL 2008
Copyright © 2008 Wiley-Liss, Inc.
International Journal of Cancer
Volume 123, Issue 6, pages 1227–1239, 15 September 2008
How to Cite
Block, K. I., Koch, A. C., Mead, M. N., Tothy, P. K., Newman, R. A. and Gyllenhaal, C. (2008), Impact of antioxidant supplementation on chemotherapeutic toxicity: A systematic review of the evidence from randomized controlled trials. Int. J. Cancer, 123: 1227–1239. doi: 10.1002/ijc.23754
- Issue online: 16 JUL 2008
- Version of Record online: 11 JUL 2008
- Manuscript Accepted: 8 MAY 2008
- Manuscript Received: 17 JAN 2008
- alternative medicine;
- complementary medicine;
Much debate has focused on whether antioxidants interfere with the efficacy of cancer chemotherapy. The objective of this study is to systematically review the randomized, controlled clinical trial evidence evaluating the effects of concurrent use of antioxidants with chemotherapy on toxic side effects. We performed a search of literature from 1966-October 2007 using MEDLINE, Cochrane, CinAhl, AMED, AltHealthWatch and EMBASE databases. Randomized, controlled clinical trials reporting antioxidant-based mitigation of chemotherapy toxicity were included in the final tally. Searches were performed following a standardized protocol for systematic reviews. Only 33 of 965 articles considered, including 2,446 subjects, met the inclusion criteria. Antioxidants evaluated were: glutathione (11), melatonin (7), vitamin A (1), an antioxidant mixture (2), N-acetylcysteine (2), vitamin E (5), selenium (2), L-carnitine (1), Co-Q10 (1) and ellagic acid (1). The majority (24) of the 33 studies included reported evidence of decreased toxicities from the concurrent use of antioxidants with chemotherapy. Nine studies reported no difference in toxicities between the 2 groups. Only 1 study (vitamin A) reported a significant increase in toxicity in the antioxidant group. Five studies reported the antioxidant group completed more full doses of chemotherapy or had less-dose reduction than control groups. Statistical power and poor study quality were concerns with some studies. This review provides the first systematically reviewed evidence that antioxidant supplementation during chemotherapy holds potential for reducing dose-limiting toxicities. However, well-designed studies evaluating larger populations of patients given specific antioxidants defined by dose and schedule relative to chemotherapy are warranted. © 2008 Wiley-Liss, Inc.
The published literature suggests that antioxidant supplements are taken by 13–87% of patients with cancer.1–10 Such a wide range of percentages might result from the variability of definitions of complementary and alternative (CAM) medicine used in the different studies, and to differences in the cancer types, age, education, economic status, and ethnicity of the groups assessed.11 Because cancer patients may take antioxidant supplements to help alleviate side effects from toxic chemotherapies, we systematically reviewed studies that evaluated the effects of antioxidants on chemotherapy-related toxicities. Much debate has focused on the use of antioxidant supplements by patients undergoing chemotherapy due to concerns that the antioxidants may interfere with the mechanism of action of the therapeutic agent and subsequently decrease its efficacy.12 However, others argue that antioxidant supplements are beneficial to patients undergoing chemotherapy because they enhance the efficacy of the chemotherapy, as well as alleviate toxic side effects, allowing patients to tolerate chemotherapy for the full course of treatment and lessen the need for dose reduction.13 We have previously reviewed randomized controlled trials in which antioxidants were given with chemotherapy and survival and tumor response outcomes were measured.14 No indication was found that antioxidants were associated with decreased survival or tumor response, although further research with very large sample sizes would be required to definitively reject the hypothesis of interference. Our analysis suggested, in fact, that concurrent use of supplements and chemotherapy treatments might produce better tumor response rates and increased chances of survival, although small sample sizes and low quality of studies precluded firm conclusions.
A primary mechanism of many chemotherapy drugs against cancer cells is the formation of reactive oxygen species (ROS), or free radicals. Drugs that form ROS include but are not limited to alkylating agents (e.g., melphalan, cyclophosphamide), anthracyclines (e.g., doxorubicin, epirubicin), podophyllin derivatives (e.g., etoposide), platinum coordination complexes (e.g., cisplatin, carboplatin) and camptothecins (e.g., topotecan, irinotecan). Unfortunately, these free radicals are often the source of serious side effects, as well. For example, nephrotoxicity, ototoxicity and peripheral neuropathy are often produced by cisplatin and other platinum-based chemotherapies.15, 16 Anthracyclines such as doxorubicin often cause cardiotoxicity.17 Drugs such as the taxanes (e.g., paclitaxel, docetaxel), vinca alkaloids (e.g., vincristine, vinblastine), antimetabolites (e.g., methotrexate, fluorouracil, cytarabine) generate lower levels of oxidative stress, and free radical damage is not considered their primary mechanism of action.13, 18 Asparaginase and dactinomycin are examples of chemotherapy drugs that use mechanisms other than oxidation for their anticancer effects. Clearly, interactions between chemotherapeutic compounds and antioxidants are complex and other factors affect the production of free radicals such as dose, localization and metabolism of the drug. In addition, some antioxidants have the potential to also act as oxidative molecules, depending on their use and/or relative concentration.
Of the supplements included in this review, antioxidant mechanisms range from free radical scavengers that act as reducers or that break lipid chains (melatonin, NAC, Vitamin E, GSH, beta carotene and vitamin C) to antioxidant enzymes formed by combining with a protein to form selenoproteins (selenium, GSH). Other mechanisms include metal chelators (Vitamin C, EGCG) or cellular protectors from free radical attack (vitamins A, C, E and melatonin) while some target and repair DNA aberrations (EGCG). Thus, understanding antioxidant-chemotherapy interactions is difficult enough in simple in vitro cell systems but exceedingly more difficult to define when using more complex animal tumor models. Further, the pharmacokinetics or pharmacodynamics of chemotherapy agents may be affected by certain antioxidants. These factors underline the need for an examination of the role of antioxidants in well-designed randomized clinical trials.
Cancer patients often have low antioxidant levels prior to chemotherapy treatment,19 and higher levels of oxidative stress have been linked with more aggressive cancers.20–22 Therefore, administration of the aforementioned drugs exacerbates oxidative stress in cancer patients, as shown by DNA oxidation and lipid peroxidation levels during and after cancer therapy.23, 24 Theories suggest that antioxidant supplementation during the administration of these chemotherapies either hinders the cytotoxic mechanism(s) of chemotherapy by quenching ROS produced by the drug, or helps protect healthy cells from additional oxidative stress and toxicity from treatment. Clearly, the heart of the dilemma for patients with cancer lies in trying to understand whether antioxidant therapy will increase their quality of life through protection of normal tissues and possibly slow disease progression by lowering oxidative levels or interfere with the eventual clinical outcome of their disease.
Alternatively, patient outcomes may be improved by antioxidants through improving the therapeutic index of coadministered chemotherapy drugs, i.e., increasing a patient's ability to tolerate full doses of antineoplastics with uninterrupted treatment schedules. The toxic side effects of chemotherapy often lead to dose reductions, interruptions and delays in chemotherapy treatment, and incomplete courses of treatment. A reduction in these side effects might result in an improved quality of life for the patient, and possibly better survival rates. In a recent study of colon cancer patients over age 65, those who received a full 5–7 months of chemotherapy had higher survival rates than those who only received 1–4 months of treatment. Furthermore, mortality rate among the 30% of patients who dropped out of chemotherapy treatment early was twice that of the group who completed therapy.25 The reduction of toxic side effects from chemotherapy has clinical relevance, and many studies have evaluated the potential for antioxidants to contribute to this reduction. This review evaluates randomized, controlled trials in which studies measured side effects in patients given antioxidants concurrently with chemotherapy to determine if the antioxidants increased or decreased the side effects of the chemotherapy.
Material and methods
Electronic databases (MEDLINE, CENTRAL (The Cochrane Library), CinAhl, AMED/AltHealthWatch (combined) and EMBASE) were searched from inception through the last week of October, 2007. Database selection was based on inclusion of peer-reviewed alternative and complementary medicine articles. An identical search string was used in all databases with the exception of Medline which required the string to be altered to fit the database's particular terminology. Three categories were combined in attempts to cover as many variations as possible for (i) cancer (5), (ii) chemotherapy (24), and (iii) antioxidants (32). Authors will provide the detailed search string upon request. Searches were not restricted by language and were performed in duplicate (AK and MM). In the case of a disagreement, disputes were resolved through discussion among the authors until consensus was reached. Non-English abstracts were translated and if they appeared to meet inclusion criteria, the entire article was translated into English. Key articles and review papers were hand searched for additional references.
Searches yielded the following results: MEDLINE (368), AMED/AltHealthWatch (254), CENTRAL (284), CinAhl (90) and EMBASE (85). Abstracts were read as an initial screening. Full text was obtained for pertinent articles. In certain cases, further clarification or data was needed and authors were contacted. The resulting articles were screened for inclusion according to the criteria mentioned later.
Type of study
Only randomized, controlled trials that provided data on side effects were included in the review. Outcome data regarding side effects must be clinically-relevant and quantifiable according to the National Cancer Institute Common Toxicity Criteria or other standardized methods for assessing toxicities, e.g., quality of life.
Cancer patients who were currently undergoing only chemotherapy (not radiation) were included. All cancer types were included.
Type of therapy
Antioxidants were given (orally or intravenously) to patients concurrently with chemotherapy. Only trials administering chemotherapies that utilized ROS-producing mechanisms were included. [ROS-generating chemotherapy (doxorubicin, epirubicin, daunorubicin, idarubicin, cisplatin, carboplatin, oxaliplatin, bleomycin, carmustine, cyclophosphamide, melphalan, etoposide, mitomycin, vinblastine, vinorelbine, paclitaxel, docetaxel) together with an antioxidant compound (vitamin C, vitamin E, vitamin A, melatonin, glutathione, N-acetylcysteine, polyphenols, green tea catechins, carotenoids, carnitine, selenium, ellagic acid, curcumin, coenzyme Q10, lycopene, flavonoids, and isoflavones, including chemical names and synonyms of vitamin names]. Only antioxidant phytochemical extracts were included (not whole herbs or multicomponent herbal mixtures that contained phytochemical antioxidants) because of the potential for confounding of results by nonantioxidant activities of complex herbs and mixtures.
All data were obtained from published peer-reviewed reports for each trial. The Jadad scoring method was used by AK and CG to assess the quality of the included articles. This validated scale allows assessment of the methodological quality of each trial by analyzing the randomization, blinding methods, and description of patient dropouts/withdrawals of each study. This results in a score between 0 and 5, 0 indicating a weak study design and 5 indicating a strong study design.26
The screening and subsequent quality assessments of included articles were limited by what was available in the written report alone. Authors were contacted in certain cases for verification of randomization. The authors of this article have attempted to avoid publication bias by only including randomized, controlled trials that inherently reduce bias. However, the possibility of publication bias (preferential publication of positive trials) cannot be excluded.
Of 965 references screened, 33 met the inclusion criteria, for a total of 2,446 patients evaluated. A flow chart shows the number of articles excluded for each factor (Fig. 1). Over half of the initial articles were excluded because they were not randomized, controlled trials (486). Many others were excluded because the antioxidant was not administered concurrently with chemotherapy (245). The remainder were excluded because the antioxidant was synthetic (44), the study included radiation (7), or the study was a preliminary report of an included study (6). Synthetic antioxidants have been included in previous reviews and therefore, were not included in this study.27, 28 All included papers were published in English; the papers in other languages were among the excluded classifications. The Jadad scoring method was used to evaluate all included studies. Jadad method and scores nearly spanned the entire range of quality from five to one. Jadad scores and the supplements reported in the trials are found in Tables I–IV.
|Reference||Tumor type(s)||No. of pts||GSH protocol||Chemotherapy regimen||Toxicity mitigation in GSH group vs. Control group||Responses in GSH group vs. Control group||Conclusion||Jadad score|
|Cascinu et al., 200229||Advanced colorectal cancer||n = 52, 26 chemo + GSH, 26 chemo + placebo||1,500 mg/m2 given IV over 15 min, immediately before chemo||Oxaliplatin 100 mg/m2 as IV infusion, followed by 5FU, 1500 mg/m2 as IV 24-hr infusion with leucovorin, 150 mg/m2 as infusion||30% vs. 100% in GSH vs. control groups experienced grade 2-4 neurotoxicity (p = 0.004); incidence and severity of other toxicities were similar between the groups||CR + PR rates were 27% vs. 23% in GSH vs. control groups; neither group reported a CR; Median survival: 16 vs. 17 months||GSH group experienced significantly reduced neuropathy vs. control group||5|
|Schmidinger et al., 200030||Advanced NSCLC and HNC||n = 20, 6 with NSCLC, 14 HNC; 11 chemo + GSH, 9 chemo + placebo||5,000 mg/m2 given IV over 15 min, immediately before chemo||CDDP 80 mg/m2 as IV infusion. HNC pts received 450 mg/m2 5FU by IV bolus, NSCLC received 120 mg/m2 IV VP-16, with cycles every 4 wks||Significant decrease in hemoglobin (p = 0.04), platelet counts (p = .03), and white blood cell counts (p = 0.004) in placebo vs. GSH groups; neither group experienced neurotoxicity||CR + PR rates were 55% vs. 50%; CR rates were 9% and 0%; Median survival: 13.1 months vs. 10.5 months||GSH group had significantly reduced hematological toxicities vs. the control group||2|
|Smyth et al., 199731||Ovarian cancer (Stages I-IV)||n = 151, 74 chemo + GSH, 77 chemo + placebo||3,000 mg/m2 given IV over 20 min, immediately before chemo||CDDP 100 mg/m2 as IV infusion every 3 weeks for six courses||58% vs. 39% were able to receive 6 cycles of CDDP (p = 0.04); 39% vs. 49% experienced neurotoxicity (p = 0.22)||CR + PR rates were 73% vs. 62% (p = 0.25); CR rates were 46% vs. 9%; survival rates were similar (stated in article)||GSH group had improved QoL scores, weight gain, neuroprotection, and nephroprotection and nonsignificantly higher tumor response rates vs. the control group||5|
|Bogliun et al., 199632||Advanced ovarian cancer||n = 54, 27 chemo + GSH, 27 chemo alone||2,500 mg/m2 given IV over 15 min, immediately before chemo||CDDP 50 mg/m2 as IV infusion in 26 pts; CDDP 75 mg/m2 given IV in 28 pts||26% vs. 50% experienced neurotoxicity; 37% vs. 78% experienced oliguria||CR + PR rates were 70% vs. 59%; CR rates were 22% and 11%; no survival rates reported; no statistical analysis due to small sample size||GSH group had less neurotoxicity and oliguria, and higher tumor response rates than the control group||1|
|Cascinu et al., 199533||Advanced gastric carcinoma||n = 50, 25 chemo + GSH, 25 chemo + placebo||1,500 mg/m2, given IV over 15 min, immediately before chemo||CDDP 40 mg/m2 and 5FU 500 mg/m2 given IV, 4-epidoxorubicin, IV bolus, 9 weekly treatments||17% vs. 89% experienced neurotoxicity (p = .0001); other toxicities were similar between the two groups||CR + PR rates were 76% vs. 52%; CR rates were 20% and 12%; survival rates were 14 vs. 10 months||GSH group had significantly less neurotoxicity and higher tumor response rates than the control group||5|
|Colombo et al., 199534||Relapsed, advanced ovarian cancer||n = 33, 16 chemo + GSH, 17 chemo alone||2,500 mg/m2 given IV over 15 min, immediately before chemo||CDDP 50 mg/m2, 9 weeks as IV infusion||13% vs. 27% experienced neuropathy; other toxicities were similar between the two groups||CR + PR rates were 75% vs. 60%; CR rates were 44% and 27%; survival rates were 21 vs. 15.9 months||GSH group had less neuropathy and higher tumor response rates than the control group||2|
|Parnis et al., 199535||Advanced ovarian cancer||n = 24||1,500 mg/m2 given IV 15 min prior to chemo||CDDP 40 mg/m2 over 2-h for either 2,3 or 4 days every 4 weeks||No statistical difference in toxicities between antioxidant and control groups.||n/a||GSH failed to reduce toxicities based on short ½ life.||2|
|Catalano et al., 200136||Colorectal Cancer||n = 52||1,500 mg/m2 given IV 15 min prior to chemo OR saline placebo||Oxaliplatin 100 mg/m2 2-h infusion d1, leucovorin 250 mg/m2 2-h infusion followed by 5-FU 1500 mg/m2 24-h infusion d1-2, q 2 wks.||No difference after 4 cycles. After 6–8 cycles, 63% exp. Grade 2–4 neurotox in control vs. 9.5% exp. Grade 2–4 neurotox in GSH group (p = 0.001).||n/a||GSH reduced incidence of moderate to severe neurotoxicity.||2|
|Fujimoto et al., 198337||Gastric Cancer||n = 207||30 mg/kg given IV every day from start of chemo to discharge||5-FU prodrug (FT-207) 16 mg/kg/day IV until discharge, then 12 mg/kg/day orally for 24-36 mos.||No significant difference in GI toxicities, higher serum 5-FU levels in GSH group||Similar survival rates||GSH group had no differences in toxicity, but significantly higher survival rates for stage III patients.||1|
|Choi et al., 200738||Advanced or metastatic cancers||n = 51||30 g/day given orally 3 days prior to chemo and 15 days after||LV 100mg/m2 IV over 30 min then FU 500 mg/m2 cont. infusion for 5 days||9% GSH vs. 38% control exp. Grade 2-4 mucositis/stomatitis (p < 0.001)||n/a||GSH prevented severe mucositis/stomatitis.||2|
|Wang et al., 200739||Colorectal Cancer||n = 86||30 g/day given orally for 7 days every 2 weeks on first day of chemo||Oxaliplatin 85 mg/m2 IV days 1 and 15 plus folinic acid 20 mg/m2 over 10-20 min, followed by 500 mg/m2 bolus 5-FU days 1,8, 15.||After 6 cycles, GSH group had less neuropathy, particularly grade 3–4 11.9% vs. 31.8% (p = 0.04) and fewer GSH patients (7.1%) needed dose reductions than control (27.3%) (p = 0.02)||Similar survival rates, CR rates 52% GSH vs. 48% control||GSH significantly reduced incidence and severity of neuropathy as well as the need for dose reduction of oxaliplatin.||1|
|Reference||Tumor type(s)||No. of pts||MLT protocol||Chemotherapy regimen||Toxicity mitigation in MLT group vs. Control group||Responses in MLT group vs. Control group||Conclusion||Jadad Score|
|Lissoni et al., 200340||Advanced NSCLC||n = 100, 50 chemo + MLT, 50 chemo alone||20 mg orally in evening||CDDP 20 mg/m2 as IV infusion for 3 days; etoposide 100 mg/m2/day IV for 3 days||18% vs. 41% experienced neurotoxicity (p < 0.01); 14% vs. 20% experienced thrombocytopenia (p < 0.01); 6% vs. 41% experienced weight loss > 10% (p < 0.001); 8% vs. 35% experienced asthenia (p < 0.005)||CR + PR rates were 35% vs. 18% (p < 0.05); CR rates were 4% vs. 0%); no control pts alive after 2 yrs, while 6% in the MLT group were alive after 5 years (p < 0.001)||MLT group had significantly reduced toxicities and improved tumor response and survival rates and vs. control group||1|
|Cerea et al., 200341||Met. colorectal cancer||n = 30, 14 chemo + MLT, 16 chemo alone||20 mg orally in evening||CPT-11 given IV at 125 mg/m2 per wk for 9 consecutive wks||29% vs. 38% experienced diarrhea grade 3–4 (associated with 50% dose reduction) (NS)||CR + PR rates were 36% vs. 13% (NS); neither group reported a CR; survival rates not reported||Toxicities were reduced in MLT group, but not statistically significant; significant disease control in MLT vs. control group (86% vs. 44%, p < 0.05||2|
|Lissoni et al., 199942||Advanced NSCLC, Breast cancer, GI tumors, HNC||n = 250, 104 NSCLC; 77 breast ca; 42 GI tract cancer; 27 HNC||20 mg orally in evening||NSCLC: CDDP + VP-16 or GEM alone, Breast cancer: DOX or mitoxantrone or paclitaxel alone, GI tumors: 5FU, FA, HNC: 5FU + CDDP. No doses given.||20% vs. 43% experienced myelosupression (p < 0.001); 2% vs. 13% experienced neurotoxicity (p < 0.05); 2% vs. 10% experienced cardiotoxicity (p < 0.02); 10% vs. 30% experienced stomatitis (p < 0.02); 27% vs. 63% experienced asthenia (p < 0.001)||CR + PR rates were 34% vs. 15% (p < 0.001); CR rates were 5% and 0% (p < 0.02); 1-yr survival rates were 51% vs. 23% (p < 0.001)||MLT group had significantly reduced toxicities vs. control group and significantly improved tumor response and survival rates vs. control group.||3|
|Lissoni et al., 199743||Advanced NSCLC||n = 70, 34 chemo + MLT, 36 chemo alone||20 mg orally in evening||CDDP 20 mg/m2 as IV infusion for 3 days; VP-16 100 mg/m2/day IV for 3 days||12% vs. 36% experienced myelosuppression (p < 0.05); 0% vs. 14% experienced neuropathy (p < 0.05); 9% vs. 33% experienced asthenia (p < 0.01); and 0% vs. 44% experienced weight loss > 10% (p < 0.001)||CR + PR rates were 32% vs. 17% (NS); CR rates were 3% vs. 0%); 1-yr survival rates were 44% vs. 19% (p < 0.05)||MLT group had significantly reduced toxicities vs. control group and significantly higher 1-yr survival rates vs. control group;||1|
|Lissoni et al., 199744||Met. solid tumors||n = 80, lung = 35; BC = 31; GI = 14||20 mg/day orally in evening||Lung (CDDP/VP-16); BC (mitoxantrone); GI (5-FU plus folates). No doses given.||MLT group sign less malaise, asthenia and thrombocytopenia. Stomatitis/neuropathy less (ns). Alopecia/vomiting not influenced by MLT.||n/a||MLT may prevent myelosuppresion and neuropathy.||1|
|Ghielmini et al., 199945||Prev. untreated, inoperable Lung cancer pts||n = 20||40 mg orally in evening for 21 days, starting 2 days b/f chemo||Carboplatin (5 areas under the curve) for two cycles on day 1, VP-16 (150 mg/m2 IV) days 1–3 every 4 weeks.||No significant difference in hematological parameters of toxicity.||n/a||MLT not protective against myelotoxic effects of carboplatin and etoposide.||4|
|Lissoni 200746||Met. NSCLC or GI cancer||n = 370||20 mg/day orally in evening||NSCLC: (CDDP/VP-16 or CDDP/GEM); CRC: (OXA/5FU and folates); GI: PELF regimen or 5-FU and FA. No doses given||MLT group had less thrombocytopenia 4% vs. 22% (p < 0.01), neurotoxicity 5% vs. 12% (p < 0.05), asthenia 27% vs. 52% (p < 0.01), cachexia 5% vs. 20% (p < 0.005)||CR + PR rates were 36% vs. 20% (p < 0.001); 2-yr survival rates 25% vs. 13% (p < 0.05)||MLT group experienced significantly lower levels of toxicity vs. control.||1|
|Reference||Tumor type(s)||No. of pts||Antioxidant protocol||Chemotherapy regimen||Toxicity mitigation in Antioxidant group vs. Control||Responses in Antioxidant vs. Control group||Conclusion||Jadad Score|
|Pace et al., 200347||Various malignant tumors: lung (15), HNC (5), ovarian (3), urethral (2), gastric (1), testicular (1)||n = 27, 13 chemo + vitamin E vs. 14 chemo alone||300 mg/d, alpha-tocopherol orally before chemo; then cont'd. for 3 months after treatment||CDDP administered in varying doses and schedules based on specific tumor site, e.g., for lung cancer, 75 mg/m2 IV on day 1 and GEM 1000 mg/m2IV on day 1 and day 8 every 3 weeks||30.7% vs. 85.7% experienced neurotoxicity (p < .01); other toxicities were similar between the two groups||CR + PR rates were 62% vs. 73% (NS); CR rates and survival rates were not reported||Vitamin E group had a significant reduction in severity and incidence of neurotoxicity. Control group had higher tumor response rate than vitamin E group.||2|
|Wadleigh et al., 199248||17 solid tumors, 1 acute leukemia||n = 18||400 mg/ml topical when lesions were observed||Receiving various chemos||6/9 Vit E pts had complete resolution of oral lesions vs. 1/9 control (p = 0.025)||n/a||Vit E may be effective in treatment of chemo-induced mucositis.||2|
|Whittaker and Al-Ismail, 198449||Acute myeloid leukemia||n = 63||200 mg Vit E (alpha tocopherol) per day, thrice orally, average 25mg/day orally digoxin or nothing.||Doxorubicin 30 mg/m2 IV over 5 mins on 5th and 6th days, then maintained at 45 mg/m2 IV monthly for at least one year||Systolic time interval measurements suggested protective role for digoxin over control (p < 0.05).||n/a||Alpha-tocopherol showed no cardiac protective effect.||1|
|Argyriou et al., 200650||Solid or non-myeloid malignancy||n = 30 (completed; 40 enrolled)||600 mg/day orally during chemo and for 3 months after.||Cisplatin-based therapy||Neurotoxicity experienced in 3/14 (21.4%) Vit E pts vs. 11/16 (68.5%) control. p = 0.026.||n/a||Vit E may have important neuroprotective effects.||3|
|Argyriou et al., 200651||Solid or non-myeloid malignancy||n = 32||300 mg/2x per day orally||Either 175 mg/m2 IV paclitaxel plus carboplatin at an AUC of 6 on day 1, or 175 mg/m2 IV paclitaxel plus 80 mg/m2 epirubicin on day 1.||Neurotoxicity exp. In 3/16 (18.7%) vit E pts vs. 10/16 (62.5%) in control (p = 0.03). PNP score (neurotoxicity measurement) vit E 2.25 vs. control 11 (p = 0.01).||n/a||Vit E protects from peripheral nerve damage.||2|
|Reference||Tumor type(s)||No. of pts||Antioxidant protocol||Chemotherapy regimen||Toxicity mitigation in Antioxidant group vs. Control||Responses in Antioxidant vs. Control group||Conclusion||Jadad Score|
|Pathak et al., 200552||Advanced NSCLC (stages IIIb and IV)||n = 136, 64 chemo + mixed antioxidants vs. 72 chemo alone||Oral ascorbic acid (6100 mg/day), vitamin E (1050 mg/day) and beta-carotene (60 mg/day)||Paclitaxel (225 mg/m2 IV as 3-hour infusion on day 1) and carboplatin (dosage based on most recent creatinine clearance value before each chemo cycle)||No statistical difference in toxicities between antioxidant and control groups||CR + PR rates were 37% vs. 33% (p = .28); CR rates were 3% vs. 0%; 1 yr survival rates were 39% vs. 33%; 2 yr survival rates were 16% vs. 11% (p = .20)||Antioxidants did not reduce toxicities. No statistically significant difference in response or survival rates between groups, however, antioxidant group had non-significant advantage in both.||2|
|Falsaperla et al., 200553||Hormone-refractory prostate cancer (chemo naïve)||n = 48 consecutive pts, 24 chemo + Ellagic acid vs. 24 chemo alone||Ellagic acid, 180 mg (60 mg every 8 h) taken orally before meals during & between chemo cycles||Vinorelbine (25 mg/ mq IV, weekly for 6 wks) and estramustine (280 mg, 3x/day, for 42 days)||33% vs. 75% experienced neutropenia (p < .05). Data also showed non-statistically significant decrease in anemia, nausea, anorexia, diarrhea, and neuropathy in antioxidant group||CR + PR rates were 58% vs. 25%; CR rates were 25% vs. 0% (NS); 2-year survival rates were 75% vs. 58% (NS)||Ellagic acid group had significantly decreased neutropenia as well as non-statistically significant reductions in other toxicities; also had higher tumor response and 2-yr survival rates.||2|
|Weijl et al., 200454||Various malignant tumors: testicular (16), osteo-sarcoma (13), GI (6), urogenital (5), H&N (5), melanoma (3)||n = 48 pts, 25 chemo + antioxidants vs. 23 chemo + placebo||Oral vitamin C (1 g, L-ascorbic acid), vitamin E (400 mg, as dl-alpha-tocopherol-acetate) and selenium (100 μg), all dissolved in milky white beverage||CDDP by IV in varying dose intensities (highest planned dose: 100 mg/m2) Each cycle 1-5 days of cytostatic drug infusions repeated every 21 days.||No significant reduction in nephrotoxicity and ototoxicity, except in correlation analysis with respect to plasma antioxidant levels; also, more pts in antioxidant group received highest planned CDDP dosages.||CR + PR rates were 44% vs. 48%; CR rates were 36% vs. 26%; survival rates were not reported||More pts in antioxidant arm were able to receive optimal doses of CDDP. Response rates were similar between the two groups, however, CR rates were higher in antioxidant group than control group; Authors report poor pt adherence (46% of all pts did not drink the antioxidant beverage during the whole study period).||4|
|Meyskens et al., 199555||CML in chronic phase (persistent leukocytosis of at least 30,000 mm3 found on at least 2 occasions)||n = 124, 57 chemo + vitamin A vs. 67 chemo alone||Oral vitamin A (50,000 IU/day, as retinol)||Intermittent oral pulse busulfan: 8 mg/m2 for 4 days every 4 weeks until chronic stable phase was reached in terms of leukocyte counts (<50,000 mm3 and >6000 mm3; chemo restarted when counts reached 50,000 mm3)||23% vs. 4% experienced more grade 2+ toxicities (p = .002)||No tumor response rates were reported; 5-yr survival: 48% vs. 30%; after adjustment for survival-related factors||Only study where antioxidant group experienced significantly more toxicities than control group. Significantly greater risk of disease progression (53%; p = .022) and death (60%; P = .014) in chemo alone group vs. vitamin A supplemented patients; Vitamin A had higher 5-yr survival rates vs. control group.||2|
|Myers et al., 198356||Various malignant tumors: breast, lymphoma, soft-tissue sarcoma||n = 24, 12 chemo + NAC vs. 12 chemo alone||NAC, oral, 5.5 gm/m2 prior to each chemo treatment||Doxorubicin, 75 mg/m2 IV every 4 wks||NAC group experienced slightly more toxicities (nausea, alopecia, diarrhea, leucopenia) vs. control group||CR + PR rates were 17% vs. 7%; CR rates were 4% vs. 0%; no survival rates were reported; no statistical analysis was conducted due to diversity of tumor types||NAC group experienced slightly more toxicities vs. control group and had higher tumor response rates vs. control group||2|
|Lin et al., 200657||Stage III Colon Cancer||n = 14, NAC group = 5, placebo = 9||NAC, oral, 1200 mg||85 mg/m2 IV oxaliplatin biweekly, and weekly IV bolus of 425 mg/m2 IV 5-FU and 20 mg/m2 LV||After 12 cycles, NAC pts 1/5 grade 2-4 neuropathy vs. control pts 8/9 grade 2–4 neuropathy (p < 0.05)||n/a||NAC reduces incidence of neuropathy.||1|
|Waldner et al., 200658||Non-Hodgkins Lymphoma||n = 40||L-carnitine, 3g/day IV prior to chemo, then 1g/day oral for 21 days after.||CHOP (day 1: 750 mg/m2 IV cyclophospha-mide, 1.4 mg/m2 IV vincristine and 50 mg/m2 IV doxorubicin, days 2–5: 100 mg oral prednisolone)||No cardiotoxicity detected in carnitine or placebo group.||n/a||Carnitine improved oxidative metabolism however, not a clinically relevant outcome.||1|
|Iarussi et al., 199459||Children with leukemia or non-Hodgkins lymphoma||n = 20||Co-Q10, 100 mg orally twice daily.||Anthracyclines (cumulative dose fixed at 240 mg/m2=120 mg/m2 IV daunorubicin and 120 mg/m2 IV doxorubicin)||Co-Q10 group had left ventricular fraction shorter than control.||n/a||Protective effect of Co-Q10 on cardiac function with anthracyclines.||2|
|Sieja and Talerczyk, 200460||Ovarian Cancer||n = 62||Se for study group in addition to orally taken mix of β-carotene, vit c, vit E, vit B2, vit B3, for both.||100 mg/m2 IV CDDP and 600 mg/m2 IV cyclophosphamide||Significant increase in WBC in Se group (p < 0.001); Sig. decrease in all side effects, except diarrhea.||n/a||Beneficial effects of Se found when taken w/chemo.||2|
|Federico et al., 200161||Cancer of digestive tract||n = 60||Oral Se (200 ug/day) plus Zinc (21 mg/day) for 50 days||500 mg/m2 IV MTX day 1, 250 mg/m2 IV 5-FU day 2, and 600 mg/m2 IV L-folinic acid day 2.||All pts malnourished at baseline. 21/30 (70%) Sel. Group did not exp. Further decline, but had increase in appetite. 24/30 (80%) control had sig decline of all parameters (body wt, etc.) (p > 0.01)||n/a||Se plus Zinc may improve general clinical course.||1|
The majority of the articles reported the use of glutathione (GSH) with chemotherapy (11) (see Table I). Melatonin (MLT) was dispensed in 7 of the studies, 6 of which were from the same group in Italy (see Table II). Five studies investigated vitamin E (see Table III). Two studies included selenium, two looked at a mix of antioxidants (vitamins C and E and selenium) and (vitamins C and E and beta-carotene) and two looked at N-acetylcysteine (see Table IV). Only 1 study was included for each of the following: vitamin A, ellagic acid, CoQ10 and L-carnitine (see Table IV).
Because the studies evaluated a variety of antioxidants in patients of several cancer types, meta-analysis was not considered advisable, and systematic review was chosen to summarize results of the studies.
Summary of studies
Of 9 GSH studies that evaluated neurotoxicity, 6 reported less neurotoxicity in the GSH group than the control group.29, 32–34, 36, 39 Four of these studies found the decrease to be statistically significant.29, 33, 36, 39 Myelosuppression was generally reported as similar between GSH and control groups with only 1 analysis where the control group experienced more anemia than the GSH group.30 The remaining GSH studies reported similar general toxicities between the GSH and control groups. No GSH studies found a higher incidence of toxicities in the GSH group than the control group.
Seven studies reported a range of decreased toxicities from the antioxidant, many significantly, over control groups. Five studies, all from Lissoni et al., reported neurotoxicity was decreased in the MLT groups, with four of those statistically significant.40, 42–44, 46 Myelosuppression was also consistently lowered in the MLT groups. Five studies, again by Lissoni et al., reported statistically significant decreases in myelosuppression.40, 42–44, 46 Ghielmini et al. reported no difference between myelosuppression between the 2 groups.45 All reports on cachexia (3)40, 43, 46 asthenia (5)40, 42–44, 46 stated a significantly significant decrease in the toxicities of the MLT group. The study by Cerea et al. found the MLT group experienced less grade 3–4 diarrhea requiring a 50% dose reduction than the control group.41 Of note in the MLT studies is the increase in tumor response and survival times of the MLT groups over the control groups. However, the advanced disease stage of subjects in these trials, and the paucity of studies from research groups other than Lissoni et al. limit the generalizability of these results.
Two studies reported on mixtures of antioxidant supplements. Both of these studies reported side effects to be similar between the antioxidant group and the control. Specifically, Pathak et al. reported alopecia, myelosuppression, diarrhea and neuropathy to be the same.52 Weijl et al. evaluated nephrotoxicity and ototoxicity and found similar results between the 2 groups.54
Ellagic acid (EA)
Falsaperla et al. investigated patients with hormone refractory prostate cancer and found no difference between the ellagic acid group and the control in terms of diarrhea, however, 33% of the antioxidant group versus 75% of the control experienced neutropenia (p < 0.05).53
A study by Meyskens et al., found no difference between myelosuppression between the 2 groups. However, the study reported 23% of the vitamin A group experienced more grade 2+ toxicities versus 4% grade 2+ toxicities in the control group (p = 0.002).56 This was the only statistically significant increase in toxicities in the antioxidant group of all the studies, although the vitamin A-specific toxicities were atypical of toxicities reported by other studies included in this review i.e., dry skin, personality changes, dry mouth, anxiety. While the control group experienced less Grade 2+ toxicity, they also had a significantly increased risk of disease progression and death compared to the vitamin A group.
Two studies evaluated NAC, however, one solely reported on the effect of the supplement on neuropathy. In a study by Lin et al., the NAC group experienced a marked decrease in neuropathy in comparison to the control group. Only 20% of the NAC group experienced grade 2+ neuropathy versus 89% in the control (p < 0.001).57 In comparison, a study of NAC by Myers et al. found nonsignificantly increased toxicities in the NAC group over the control. These increased toxicities were alopecia and diarrhea, of 8 total toxicities measured. Other toxicities measured were similar between the 2 groups.56
A study by Waldner et al. evaluated L-carnitine for potential protection against cardiotoxicity.58 However, neither the L-carnitine or control group experienced cardiotoxicity.
Coenzyme Q10 (CoQ10)
Iarussi et al. also evaluated the potential for an antioxidant, CoQ10, to protect from cardiotoxicity.59 By evaluating the left ventricular fraction time of each group of patients, Iarussi determined the CoQ10 group had a lower ejection fraction than the control group and therefore was protected from cardiotoxicity.
Two studies of selenium found significant decreases in toxicities in the antioxidant groups. Sieja et al. found a statistically significant decrease in alopecia, myelosuppression and asthenia.60 Federico et al. also reported a statistically significant decrease in asthenia in the selenium group versus the control.61
The majority of evidence from studies included in this review suggests that antioxidant supplementation may reduce the toxic effects of ROS-generating chemotherapies. Of a total of 86 separate reports on toxicities, 53% in (46) showed the antioxidant group experienced less toxicity than the control group. Of that 53, 82% of those studies reported a statistically significant difference in toxicity. About 43% in (37) of the reports stated no difference in toxicities between the antioxidant group and the control. About 4% of the reports showed an increase in toxicities in the antioxidant group over the control group. Only one of the reports of increased toxicities was statistically significant. This report stated the antioxidant group experienced more general toxicities of grade 2+ than the control, although these results were not surprising due to the well documented toxicities of high-dose vitamin A.55 In addition, the general toxicities named in this study were not reported by any other studies reviewed. In another study (NAC) that reported higher toxicities in the antioxidant group than the control group, 2 of 8 side effects measured were nonsignificantly higher in the antioxidant group. One of the 2 reported increased toxicities was diarrhea, a known side effect of oral NAC supplementation.56 In the cases of myelosuppression,30, 40, 42–44, 46, 53, 60 asthenia,40, 42–44, 46, 60, 61 weight loss,31, 40, 43, 46 cardiotoxicity42, 59 and nephrotoxicity,42 decreased toxicities were statistically significant in every report.
The most frequently investigated toxicity was neurotoxicity in (19), a major side effect of platinum-based chemotherapies. Of 19 total assessments of neuropathy, 15 reported decreased neurotoxicity in the antioxidant group.29, 32–34, 36, 39, 40, 42–44, 46, 47, 50, 51, 57 Of these 15 reports of decreased neurotoxicity, 12 were statistically significant.29, 33, 36, 39, 40, 42, 43, 46, 47, 50, 51, 57 No studies reported increased neurotoxicities from antioxidant supplementation.
Myelosuppression was the second most frequently investigated toxicity with 17 assessments, 8 of which reported significantly decreased toxicities in the antioxidant group.30, 40, 42–44, 46, 53, 60 However, 9 studies reported no difference between the incidence of myelosuppression in the antioxidant group and the control group.31, 33, 34, 37, 45, 48, 52, 55, 56 Other toxicities reported to be the same between the 2 groups were alopecia,33, 40, 43, 44, 50–52 diarrhea,43, 52, 53 nausea/vomiting,44, 50, 51, 56 nephrotoxicity43, 54 and ototoxicity.34, 35, 54
Given these findings, it appears that some antioxidant nutrients may be more effective than others at preventing or alleviating treatment-related toxicities. Vitamin A and L-carnitine, for example, are considered weak antioxidants, perhaps accounting for their lack of toxicity-mitigating effects. Additionally, the potentially cytoprotective effects of antioxidants may be dose-dependent or possibly require appropriate synergists to amplify cytoprotective effects. The positive findings of the glutathione studies may result in part from intravenous administration of the antioxidant, allowing for high doses of the antioxidant. As a result of some investigators' concerns that antioxidants may interfere with the efficacy of chemotherapy, doses of antioxidants may have been administered at suboptimal levels. Future studies should consider a range of doses for antioxidant supplements given concurrently with chemotherapy.
While this review only included studies with chemotherapy regimens of at least one drug thought to produce higher levels of oxidative stress (e.g., anthracyclines, platinum coordination complexes), there is some indication that these free-radical producing drugs may have additional mechanisms of action.62 Thus, while antioxidants may have reduced free-radical damage to normal tissues and potentially diminished side effects, other non-oxidative mechanisms of action of the chemotherapy may still produce toxicities unaffected by antioxidant supplementation.
Tumor response rates were not the focus of this review, however it is noteworthy that all but one47 of the antioxidant supplemented groups in studies reporting tumor response experienced the same or better response than the control group. No studies reported significantly worse survival or response in the antioxidant supplement group, as reported in our previous publication, which reviewed studies that reported tumor response or survival.14 Overall, MLT, vitamin E and GSH showed consistent and promising reductions in toxicities, in particular, neurotoxicity. Of the MLT and vitamin E studies that reported neurotoxicity data, all but one44 reported a statistically significant reduction in neurotoxicities in the antioxidant supplemented group over the control group. Additionally, the MLT studies, while of different tumor types and a range of sample sizes (20–370), showed a consistent reduction in myelosuppression. Every MLT study was conducted with 20 mg of MLT taken orally in the evening, beginning 7 days prior to chemotherapy,40–44, 46 except the Ghielmini study45 which gave 40 mg orally in the evening, beginning 2 days prior to chemotherapy. The Ghielmini study was the smallest MLT study with only 20 patients and was the only MLT study that did not show a reduction of myelosuppression. All but one of the other studies showed a statistically significant reduction in myelosuppression. While the studies that reported a reduction in myelosuppression did not receive high Jadad scores, the sample sizes were some of the largest of all the studies included in this review. Two important factors to be determined by future studies are the optimal dose of MLT as well as the optimal start date of initial MLT treatment from the outset of chemotherapy. Furthermore, while these MLT studies are consistent in their larger sample sizes, dosing and results, all but one were conducted by a single research group. Replication of the findings by other investigators is urgently needed.
Four of the 5 vitamin E studies showed statistically significant reductions in toxicities, primarily neurotoxicities. Doses ranged from 300–600 mg of vitamin E across various cancer types. The quality of the studies spanned from Jadad scores of 3 to one study which scored a one. Of note, the study of the lowest quality was the only study to not report a statistically significant decrease in toxicity. This study by Whittaker recruited 63 patients, but only 25 (7 on vitamin E) survived to be included in statistical analysis for “cardiac damage.”49 Because none of the vitamin E studies had high Jadad scores or large sample sizes, statistically significant reductions in toxicities reported by the other 4 studies cannot be generalized for clinical use. Future studies should employ larger samples and better methodology.
Finally, while the GSH studies spanned from some of the highest quality studies within the review to some of the lowest, GSH appears to be one of the most promising antioxidants for the reduction of neurotoxicity from platinum-based chemotherapy. Unfortunately, there was little consistency among the studies with various tumor types, sample sizes, doses and routes of administration (IV and oral, 1,500 mg/m2 to 30 g/day p.o.). However, those studies of the highest quality reported a reduction in neurotoxicity, 2 of which were statistically significant.29, 33 GSH should be further studied for its potential to reduce neurotoxicities for patients on platinum-based chemotherapy, in particular, for colorectal and gastric cancers. However, future research on concurrent use of antioxidants and chemotherapy should employ larger sample sizes, single cancer types and better research designs.
This summary of 33 RCTs encompassed cancer patients of diverse tumor and treatment type. All but one of the studies included less than 300 patients and therefore should generally be regarded as Phase II studies. Larger sample sizes are necessary in order to reliably assess modest yet clinically important treatment effects. Some studies identified in this review may have been designed and powered to detect differences in treatment response or survival rather than toxicity. Lack of adequate statistical power to detect differences in toxicities would render those studies showing similar or non-significantly better results in the antioxidant arms difficult to interpret. In the absence of statistical power calculations (either a priori or post hoc), a common problem in older randomized trials, other clinically important effects may have been missed in the smaller trials.63 The limited number and power of the studies included make it impractical to conclusively rule out the possibility that antioxidants have no effect or even a negative effect on chemotherapy toxicities. To exclude this possibility, very large and well-designed trials would have to be carried out.
Other limitations include the predominately advanced stage of the patients participating in these studies and compliance issues. Because the majority of the subjects in the included studies had advanced or relapsed disease; the applicability of these results in patients with earlier, more chemosensitive disease is not addressed by these studies. Noncompliance was an issue in only 1 study, resulting in a 46% noncompliance rate among the antioxidant group.54
While quality of the studies was assessed using the Jadad method, this method is not without limitations. It is useful in its assessment of a randomized trial's methodological quality, yet, it fails to take into account limitations of studies such as sample size and statistical power. In this review, the Jadad method was particularly pertinent due to the lack of blinding or placebo controls in many of the studies; less than one-third of the studies (10) included double-blinded, placebo-controlled methods.29, 31, 33, 35, 36, 45, 48, 50, 54, 60 Of the studies that did use double-blinding, 6 evaluated neurotoxicity, 4 of which reported a statistically significant reduction of neurotoxicity in the antioxidant supplemented groups.29, 33, 36, 50 In addition, the response rates were similar to or non-significantly greater in antioxidant groups than those of control groups in all 4 studies.
Lastly, while extensive efforts were made to comprehensively review the literature through searching multiple databases, hand searching reference lists, personal communications, etc., extending the search to include hand searching of conference proceedings, dissertations and theses and additional clinical trial registries may have further reduced bias and possibly produced more negative studies.
Implications for clinical practice
This systematic review provides preliminary evidence, limited by quality and sample size of the reviewed studies, suggesting that certain antioxidant supplements may reduce adverse reactions including neurotoxicity, asthenia, stomatitis/mucositis, and weight loss. Significant reductions in toxicity may alleviate dose-limiting toxicities so that more patients are able to successfully complete prescribed chemotherapy regimens, suggesting an improved therapeutic index. The aforementioned study by Neugut et al.25 demonstrated the relationship between more doses of chemotherapy and higher survival rate. In this systematic review, 5 included studies reported the antioxidant group (3 GSH, MLT and a mixture) experienced better treatment tolerance in terms of either less dose-reduction34, 39, 41, 54 or higher rates of completing full chemotherapy regimens than control groups (GSH).31 None of the included studies reported that the control group had higher doses or more full cycles of chemotherapy than the antioxidant group. Because of the potential for the relationship between the reduction of dose-limiting toxicities allowing for full chemotherapy cycles and the subsequent potential for increased tumor response and/or survival, it is critical that future antioxidant/chemotherapy studies employ proper sample sizes and methodologies so that the results are of clear clinical relevance.
Our previous review found no evidence of antioxidant interference with chemotherapy mechanisms, with a possibility that antioxidants may even improve tumor response or patient survival.14 Combining this result with the potential for improvement of toxic side effects by antioxidants reported in the present review, additional strategies for further research on antioxidants and chemotherapy are now warranted. First, toxicities should be chosen that occur regularly in high percentages of patients, are clinically significant in terms of requiring dose reductions or impairing quality of life, and are poorly controlled by currently available means. Accurate means for measuring the toxicities of interest should be available. Neurotoxicity from platinum-based drugs (which have strong ROS mechanisms of action) may be one appropriate model to research. Animal studies might be employed to select antioxidants and modes of administration that have the highest probabilities of clinical success. Subsequently, clinical trials progressing through Phase I, II and III models could be undertaken. This would advance the present state of knowledge regarding antioxidants and chemotherapy from the existing series of Phase II-type studies towards a more thorough assessment. Finally, most studies to date have been performed in patients with advanced or relapsed disease. For these patients, an improved therapeutic index is of special relevance to allow continuation of chemotherapy, since lengthy chemotherapy regimens may be applied to retard disease progression even in the absence of complete remission. Thus, regimens for advanced disease, or second-line and later regimens may be of particular interest to evaluate in future research.
- 4Use of complementary/integrative nutritional therapies during cancer treatment: implications in clinical practice. Cancer Control 2002; 9: 236–43., , , , , .
- 11Reasons for and characteristics associated with complementary and alternative medicine use among adult cancer patients: a systematic review. Integr Cancer Ther 2004; 3: 1–12., , , .
- 18Goodman and Gilman's “the pharmacological basis of therapeutics,” 9th edn. New York: McGraw Hill, 1996., , , , .
- 27Non-small cell lung cancer: the role of cytoprotection in treating therapy-related toxicity. J Support Oncol 2004; 2: 13–7..
- 36Neuroprotective effect of glutathione (GSH) on oxaliplatin (L-OHP) based chemotherapy in advanced colorectal cancer patients (pts): a randomized double-blind placebo-controlled trial. Eur J Cancer 2001; 37: S280–S281., , , , , , , , , .
- 41Biomodulation of cancer chemotherapy for metastatic colorectal cancer: a randomized study of weekly low-dose irinotecan alone versus irinotecan plus the oncostatic pineal hormone melatonin in metastatic colorectal cancer patients progressing on 5-fluorouracil-containing combinations. Anticancer Res 2003; 23: 1951–4., , , , , , , , .
- 43A randomized study of chemotherapy with cisplatin plus etoposide versus chemoendocrine therapy with cisplatin, etoposide and the pineal hormone melatonin as a first-line treatment of advanced non-small cell lung cancer patients in a poor clinical state. J Pineal Res 1997; 23: 15–9., , , , , , , , .
- 59Protective effect of coenzyme Q10 on anthracyclines cardiotoxicity: control study in children with acute lymphoblastic leukemia and non-Hodgkin lymphoma. Mol Aspects Med 1994; 15S: S207–S12., , , , , , , .