Abstract
Selenium is an essential trace element involved in many biological processes that are mediated through, at least, 25 selenoproteins expressed in humans. Extensive study of selenium compounds has demonstrated growth inhibition of malignant cells in a vast array of experimental models. Moreover combining selenium with conventional cancer therapy has yielded promising results in both preclinical studies and a cohort of human trials. The aim of this review is to highlight the current research evaluating the role of selenium compounds in combination with chemotherapy and radiation. Pharmacodymanic mechanisms responsible for the differential effects of the commonly studied compounds on healthy and malignant cells are presented and the pertinent in vitro and in vivo data summarised. The clinical utility of this approach is discussed both in terms of anti-tumour efficacy and toxicity prevention. Finally a case is made for novel trial designs to facilitate rapid progression into pivotal studies.
Selenium (Se) is one of the most extensively studied trace elements, especially with regard to cancer. Most of this research relates to nutritional doses of Se in healthy people as primary prevention of cancer, with strong evidence that populations with low Se intakes are at higher risk of cancer (1-3); in these groups, supplementation may reduce cancer incidence and cancer-specific mortality (4). However it appears that there can be “too much of a good thing” with Se; not only does supplementing with supranutritional doses of Se in Se-replete populations not confer protection against cancer (5), laboratory and clinical studies suggest this may even increase the risk of cancer (6, 7).
Many patients extrapolate the prevention data and take Se (often at supranutritional doses) to try and control their existing cancer (8, 9). There is very little clinical research relevant to this practice but one study on prostate cancer showed no restraint of prostate-specific antigen (PSA) velocity and, of greater concern, in those with the highest baseline serum Se, supplementation with 800 μg daily of selenized yeast actually significantly increased PSA velocity (10).
While this generates concern about patients using Se as a sole treatment for cancer, especially in those who have adequate Se intakes, there is substantial preclinical research and some early clinical trials suggesting that Se compounds may significantly protect against the normal tissue toxicities of cancer therapies without compromising their anticancer efficacy (or even enhancing it) (11-13). The mechanisms that mediate this are being increasingly explored and understood and, in contrast to primary prevention, the beneficial interactions of Se with cancer treatments appear to be dose-dependent, with maximum efficacy at doses much greater than nutritional requirements (13).
While this is encouraging, considerable uncertainties slow the progress of research in this area. It is not clear what chemical form of Se is most effective, nor at what dose (which has varied from <100 to 90,000 μg/day). While the Se dose has sometimes been guided by preclinical pharmacokinetic (PK) studies, there has not been a clear evaluation of the pharmacokinetic-pharmacodynamic (PK-PD) relationship for various Se compounds, nor comparison of their differential effects on normal and malignant cells at different concentrations for each compound. Furthermore the toxicities of supranutritional doses of the different Se compounds in clinical use have not been systematically evaluated. Of particular concern is the genotoxicity of some forms of Se, particularly the inorganic salts (14), and the possibility that using these in conjunction with other genotoxic therapies (such as radiation and alkylating agents) could increase the risk of second malignancies and myelodysplasia. No clinical trials to date have reported on such end-points.
This review focuses on the preclinical and clinical evaluation of the interaction of Se compounds with chemotherapy and radiation, the PK-PD relationship of these compounds and differential effects and thresholds in normal and malignant cells. Ultimately we wish to use this information to inform the rational development of further clinical trials that can definitively assess whether these Se compounds can meaningfully reduce the burden of toxicities of cancer treatments for patients while maintaining or improving their efficacy, as suggested by preclinical studies.
Selenium – Role in Human Health
It is well known that Se intake varies hugely worldwide due to differences in diet and soil content of Se in different geographical locations. Although a recommended dietary intake of Se has been established, it is difficult to give a broad recommendation for a specific dose (6). ‘Nutritional’ doses ranging from 50-200 μg/day have been used mainly for primary prevention (15), and Se supplementation at these doses for Se-deplete individuals has been associated with lower overall mortality and incidence of certain cancer types (6). However, as alluded to earlier, it is well known that taking Se at supranutritional doses over a long duration can cause adverse effects, a reminder that Se is also recognised as a toxic element (6).
The role of Se in maintaining health is complex and it has many important biological functions including in redox signalling, thyroid metabolism, immune function, detoxification and antioxidant activity (6). These effects are primarily mediated by over 25 specific encoded selenoproteins, in which selenocysteine replaces cysteine. Among the more widely studied selenoproteins are the glutathione peroxidases (involved in antioxidant activity and sperm motility) and thioredoxin reductases (involved in intracellular redox regulation and signalling). The pronounced immunostimulant effects of Se can cause a wide array of effects, from increased proliferation of activated T-cells and natural killer cell activity to enhanced lymphocyte-mediated tumour cytotoxicity (6).
Selenium Compounds – Metabolism, Pharmacology and Toxicity
Se exists in different chemical forms and these compounds can be broadly categorized into organic and inorganic forms. The metabolic pathway of dietary Se is shown in Figure 1.
There is good evidence that the increased Se status attained after supplementation with organic forms of Se such as Se-methyl-selenocysteine (MSC) and L-selenomethionine (SLM), is maintained for a longer period after its discontinuation compared to inorganic forms such as sodium selenate reported whole-body half-lives of SLM and sodium selenate in humans were 252 and 102 days respectively (16). Accordingly, foods or supplements containing SLM can maintain the activities of selenoenzymes during periods of Se depletion for longer than those containing inorganic Se, owing to the recycling of SLM catabolised from protein stores.
Animal laboratory studies have shown that the organic forms of Se are both more effective and safer than the commonly-used inorganic forms such as sodium selenite (SS), which are more genotoxic (14). From 50% lethal dose (LD50) determinations, SS was found to be fourfold more toxic than SLM when administered to mice intravenously and three-fold more toxic than Se-yeast when given orally to rats (17).
In humans safety data is more limited. The No Observed Adverse Effect Level (NOAEL) for Se in humans is variably reported as 400-850 μg/day (18,19). Chronic Se toxicity would be expected after long-term consumption (over months to years) of more than 2,400-3,000 μg/day but is reversible (20). Based on the published literature, likely symptoms of toxicity (due to environmental exposure) include brittle hair and nails and their loss, gastrointestinal disturbances, skin rash, garlic breath odour (caused by volatile selenium compounds) fatigue, irritability, and nervous system abnormalities. Impaired natural killer cell activity and endocrine disturbances may also occur (20).
Three major forms of Se (MSC, SLM and SS) are widely used in supplementation in both preclinical and clinical studies. The organic Se compound MSC is a water-soluble amino acid that is absorbed in mammals from the gastrointestinal tract and is well-tolerated (21). It is readily converted to the active moiety methylselenol through one-stage β-lyase conversion and remains more bioavailable for anticancer effects compared with the other forms (22). Methylselenol can be demethylated to yield selenide or methylated to yield dimethyl selenide, then released in the breath; methylated again, it yields trimethyl selenonium, which is excreted in urine (23). A phase 1, single-dose, dose-escalation study in 15 human subjects evaluated the toxicity and PK profile of MSC, administering 400, 800 and 1,200 μg of MSC orally and measuring Se plasma and urine levels. There was no significant clinical or laboratory toxicity and little difference was seen in PK parameters at the 400 and 800 μg dose levels (23).
The organic Se compound SLM is well-tolerated in humans at a dose of 7,200 μg (or comparable dosing at 3,600 μg/m2) twice daily for 7 days followed by a once-daily maintenance dose for 10 weeks (24, 25). The only toxicity attribute to SLM was garlic-like odour in breath and urine that was seen more commonly during the induction SLM week and was found to disappear with prolonged treatment. All patients given 4,800 μg SLM twice daily achieved plasma Se levels >15 μM, the Se concentration required for reduced chemotherapy-induced toxicity and enhanced antitumor efficacy of chemotherapeutic drugs in preclinical animal models (13). The plasma Se dose levels exceeded 30 μM by day 28, thereby confirming the feasibility of using high-dose SLM in the clinic in this setting. In another study, selenized yeast (which predominantly contains SLM) was given in doses of 3,200 μg daily for up to 24 months in 16 patients with prostate cancer on ‘watchful waiting’, with five patients reporting minor toxicities, including garlic breath, brittle nails or hair, stomach upset or dizziness, without any correlation with plasma Se levels (18). None of these patients reported symptoms of peripheral neuropathy.
Metabolic pathway of dietary Se in humans, from Rayman et al. (93). SeMet, Selenomethionine; SeCys, selenocysteine; GSSeSG, selenodiglutathione; γ-glutamyl-CH3SeCys, γ-glutamyl-Se-methyl-selenocysteine; H2Se, hydrogen selenide; HSePO3, selenophosphate; CH3SeCys, Se-methylselenocysteine;CH3SeH, methylselenol; (CH3)2Se, dimethyl selenide; SeO2, selenium dioxide; (CH3)3Se, trimethylselenonium ion.
The inorganic compound sodium selenate (which is metabolised to selenite) was evaluated in a recent phase I trial, which established a maximum tolerated dose (MTD) of 60,000 μg daily orally in patients with metastatic castration-resistant prostate cancer (26). Dose-limiting toxicities of fatigue and diarrhoea were seen at 90,000 μg, with adverse events including nail disorders, muscle spasms, alopecia and nausea. Some of these were attributed to the accumulation of selenite, the more potent metabolite of selenate. In the same trial, slowing of PSA velocity was also observed, and one patient had >50% reduction in serum PSA. Another clinical trial adding SS 200 μg/kg/day (equivalent to 14,000 μg/day in a 70-kg adult) to chemotherapy in lymphoma patients reported less toxicity with combination therapy than with chemotherapy alone, including infections and reduction in cardiac function (27).
Pharmacodynamics of Selenium Compounds in Normal and Malignant Cells
Se affects varied cellular processes and molecular pathways that may be involved in the anti-cancer effect of Se [reviewed in (3)] and include:
Seleno-enzymes involving the reduction of DNA damage, oxidative stress and inflammation;
Induction of phase II conjugating enzymes involving detoxifying carcinogens and reducing DNA adduct formation;
Enhancement of immune response including cytotoxic lymphocyte and natural killer cell activity;
Increase in tumour-suppressor protein p53, which inhibits proliferation, stimulates DNA repair and promoting apoptotic death by acting as a transcription factor for several genes, including the growth arrest and DNA damage-inducible (GADD) genes;
Inactivation of protein kinase C, a signalling receptor that plays a crucial role in tumour promotion by oxidants;
Alteration in DNA methylation, as abnormal methylation patterns are associated with neoplasia and inactivation of tumour-suppressor genes;
Perturbation of the cell cycle, resulting in growth inhibition and may allow DNA repair to take place;
Induction of apoptosis of cancer cells, which generally involves the sequential activation of the caspases;
Inhibition of angiogenesis required for the growth and metastasis of tumours.
Reassuringly, there is a significant body of published work demonstrating differential effects of Se compounds in normal versus malignant cells or tissues that would be favourable clinically, as summarised in Table I. However, while there is clear evidence from preclinical work that the efficacy and toxicity of Se compounds varies greatly (11,14), European clinical trials mostly use SS and those in the US use SLM, with no direct comparison of Se compounds or their PK-PD dose relationship.
Selenium with Chemotherapy and Radiotherapy
The challenge in using Se compounds in cancer patients in conjunction with chemotherapy and/or radiotherapy lies in being able to reduce normal tissue toxicities of these treatments without compromising their antitumour effects (or preferably enhancing those). However, despite numerous studies on the PD of Se in normal and malignant cells, it is still not known which form and dose of Se can be safely used and has the most favourable differential effect in normal and malignant tissues, especially in conjunction with chemotherapy and radiation. Current doses of Se are empirical or guided by PK, although the PK-PD relationship has not been established in humans, hence the optimal form and dose of Se to be used with chemotherapy or radiotherapy remains unclear. However Joel et al. have demonstrated that PD biomarkers of Se effects can be measured in human white blood cells (WBC) in vitro and in vivo (28), thus enabling the relationship between Se PD and PK to be evaluated in clinical studies and determine the optimal Se compound and dose to be incorporated into potentially pivotal trials.
Antitumour Efficacy
While Se is selectively cytotoxic to cancer cells at higher doses (summarised in Table I), it commonly augments the anticancer efficacy of chemotherapy and radiation in cell culture and tumour xenograft models (summarised in Table II). Therapeutic synergy has been demonstrated between supranutritional doses of Se compounds and chemotherapy drugs, including cisplatin, carboplatin, oxaliplatin, irinotecan, docetaxel, fluorouracil and doxorubicin in human tumour xenografts of colorectal, ovarian, prostate and small cell and non-small cell lung carcinoma, head and neck squamous cell carcinoma and leukaemia (11,13, 29-32).
However the strength of the interaction varied greatly, with differing xenograft models showing up to a 20-fold difference in Se dose potency) (11). Furthermore, the xenograft model and Se dose also influenced the apparent efficacy of the Se compounds. For example, at their MTD of 200 μg/day orally (approximately 8,000 μg/kg/day), MSC and SLM were both superior to SS in combination with irinotecan in head and neck squamous cell carcinoma xenografts (Figure 2) (11) whereas SS and SLM (1,000-1,500 μg/kg/day i.p.) were equally dose-potent and effective when combined with cisplatin in an ovarian cancer xenograft model (32).
Attention is also drawn to the schedule-dependency of Se in combination with other cancer treatments: the effects of MSC or SLM were maximal after 7 days pre-treatment, with little or no benefit seen when co-administered without pre-incubation with Se in human colorectal and head and neck squamous cell carcinoma xenograft models in mice (11, 33).
Few clinical studies have been conducted that evaluated the impact of Se supplementation during chemotherapy or radiation on treatment efficacy. Muecke et al. randomised 81 Se-deficient women to oral SS (500 μg daily with radiotherapy and 300 μg on non-radiotherapy days) or radiotherapy alone following surgery for gynaecological malignancies (34). At a median follow-up of 67 months, disease-free survival was not significantly different (log-rank p=0.65), though there was there a trend for improved overall survival in the selenium-treated group (log-rank p=0.09) (35).
Mix et al. conducted a randomised, placebo-controlled phase 2 pilot study in 18 patients with locally-advanced head and neck squamous cell carcinoma receiving cisplatin concurrently with radiation, administering placebo or 3,600 μg/m2 SLM twice daily orally for 7 days prior to treatment then once daily until 3 weeks after chemoradiation completion. In this small study, there was no difference observed in disease-free or overall survival (log-rank p=0.39 and p=0.39 respectively) (24).
Asfour et al. reported a trial randomising 50 patients with non-Hodgkin lymphoma to either very high doses of SS (200 μg/kg/day) with cyclophosphamide, vincristine, doxorubicin and prednisone chemotherapy or chemotherapy alone (27). A significantly greater tumour response rate was observed in the Se group than controls (60% vs. 40% respectively), with correspondingly lower levels of the oncogene B-cell lymphoma 2 (BCL2) in bone marrow aspirates in Se-treated patients compared to controls.
Pharmacodynamic mechanisms of selenium supplementation in normal and malignant cells.
Tumour cure rates in nude mice with head and neck squamous cell carcinoma xenografts (5 mice per group). Se compounds differ in their ability to improve antitumour activity of irinotecan in xenografts. MSC: Se-Methyl-selenocysteine; SLM: seleno-methionine; MSA: methyl-seleninic acid; SS: sodium selenite. (group means reported) Courtesy of Dr Y Rustum.
Reduction in Toxicity
Table II summarises extensive preclinical work that demonstrated the protective effects of various Se compounds against the toxicities of radiation (12, 36) and organ-specific toxicities of many cytotoxic drugs, including myelo-suppression, mucositis, diarrhoea, alopecia, cystitis and nephrotoxicity (11, 13, 30, 37). Specific toxicities of cisplatin in kidneys, bone marrow and intestine were improved without affecting antitumour efficacy (30, 38-41) or pharmacokinetics (42). Impressive protection from lethal effects of six cytotoxic drugs with 200 μg/day MSC was demonstrated in nude mice (13).
Similar protection by Se compounds from the normal tissue toxicities of chemotherapy has been reported in several clinical trials (Table III). In the setting of high-dose chemotherapy and allogeneic haematopoietic stem cell transplantation, a double-blind randomised placebo-controlled study of oral selenized yeast in 77 patients reported a reduction in the incidence and duration of severe oral mucositis (43). In a trial randomising 41 patients to 4,000 μg Se (as seleno-kappacarrageenan) orally daily for 4 days before and after the first dose of cisplatin-based chemotherapy or to chemotherapy alone, significantly less myelosuppression and nephrotoxicity was seen with Se (44). A double-blind trial involving 62 women receiving cisplatin-based chemotherapy for ovarian cancer randomized patients to Se as selenized yeast, 200 μg per day for 3 months, starting concurrently with chemotherapy (45). Those patients randomized to Se experienced significantly less toxicity at 3 months of treatment, including gastrointestinal, alopecia, weakness and neutropenia. The previously referenced four studies using high-dose SS in patients with non-Hodgkin lymphoma reported the Se group had significantly fewer toxicities (including infection and fall in cardiac left ventricular ejection fraction) and reduced apoptosis of neutrophils (46).
Hearing loss following chemoradiation in selenium-, and placebo-treated groups. (n=6, 2 Selenium treated, 4 placebo, mean±SEM) Courtesy of M. Jameson.
Data from the aforementioned study conducted by Muecke et al. reported significantly increased blood Se with SS supplementation (the primary endpoint) but analysis of secondary endpoints showed a significant reduction in the actuarial incidence of grade 2 or more diarrhoea (from 46.6% to 21.0%, p=0.039), without any Se-related side-effects (12). SS has also been assessed as a potential modulator of radiation-related toxicities in 39 head and neck cancer patients in a randomised controlled trial (36). Using the same dosing schedule of SS as the study of Muecke et al. (12), there was no statistically significant difference in the incidence of severe toxicities.
More recently SLM has been trialled alongside cisplatin-based chemoradiation for both head and neck squamous cell cancer and non-small cell lung cancer (NSCLC), with treatment toxicity as the primary outcome (24, 47). In the head and neck trial 18 patients were randomised to pre-treatment with 3,600 μg/m2 SLM twice daily or placebo before radiotherapy and then daily thereafter. While no significant difference in grade 3 toxicities was observed, 50% of Se-treated patients had grade 0-1 mucositis compared to 25% in the placebo group (24). In addition, a trend for reduced high-frequency hearing loss in Se-treated patients was seen (Figure 3; unpublished data, M. Jameson). The single-arm NSCLC study in 16 patients, administering 4,800 μg/m2 SLM twice daily for 7 days pre-radiotherapy then daily during chemoradiation, reported no Se-related toxicity but a lower than expected rate of grade III or higher toxicities was observed, particularly for pneumonitis and anaemia (47).
Combination of selenium with systemic therapies and radiation.
Conclusion
While there have been major advances in cancer therapy in recent years, especially with the development of targeted-therapies, cytotoxic chemotherapy and radiation still remain a mainstay of treatment for many malignancies. The considerable toxicity of these treatments remains a major challenge in cancer management, one that has not been overcome despite extensive research into protective strategies (48).
In this context, Se appears to be unique in terms of its ability to protect against carcinogenesis as well as selectively target existing cancer cells and synergise with other cancer therapies, while protecting normal tissues from the cytotoxic effects of those treatments. These distinctive features enable, at least in preclinical models, administration of higher doses of cytotoxic agents than would be feasible otherwise, with improved cancer outcomes. If this widening of the notoriously narrow therapeutic index of cytotoxic treatments with Se is replicated in the clinical setting, this would represent a very significant advance in cancer management, and further clinical evaluation is clearly justified.
However, before large-scale clinical trials are undertaken to evaluate the ability of Se to modulate the efficacy and toxicity of anticancer therapies, more research is needed to determine which Se compound, and at what dose, can be most safely and effectively used in such trials in patients with cancer. The preclinical data suggest that organic Se compounds (such as MSC and SLM) are safer and likely more effective than inorganic compounds such as SS, particularly at the higher doses that appear to be optimal for anticancer efficacy and beneficial interactions with chemotherapy and radiation.
Clinical trials evaluating outcomes when selenium compounds were given with chemotherapy and/or radiotherapy.
More rigorous evaluation in clinical trials is needed to characterise the relationship between PK and PD endpoints and comparative safety (including genotoxicity) for the Se compounds in clinical use. Innovative trial designs could include patients selected with malignancies that enable the evaluation of differential effects of Se compounds in normal cells and malignant cells. These trials are critical to inform the selection of the optimal Se compound and dose for future clinical trials and equally to avoid less effective or more toxic forms and doses of Se.
Footnotes
↵* These Authors contributed equally to this study.
This Article is freely accessible online.
- Received September 6, 2017.
- Revision received October 3, 2017.
- Accepted October 12, 2017.
- Copyright© 2017, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
