Review of high-dose intravenous vitamin C as an anticancer agent
Abstract
In the 1970s, Pauling and Cameron reported increased survival of patients with advanced cancer treated with high-dose intravenous (IV) vitamin C (L-ascorbate, ascorbic acid). These studies were criticized for their retrospective nature and lack of standardization of key prognostic factors including performance status. Subsequently, several well-designed randomized controlled trials failed to demonstrate a significant survival benefit, although these trials used high-dose oral vitamin C. Marked differences are now recognized in the pharmacokinetics of vitamin C with oral and IV administration, opening the issue of therapeutic efficacy to question. In vitro evidence suggests that vitamin C functions at low concentrations as an antioxidant but may have pro-oxidant activity at high concentrations. The mechanism of its pro-oxidant action is not fully understood, and both intra- and extracellular mechanisms that generate hydrogen peroxide have been proposed. It remains to be proven whether vitamin C-induced reactive oxygen species occur in vivo and, if so, whether this will translate to a clinical benefit. Current clinical evidence for a therapeutic effect of high-dose IV vitamin C is ambiguous, being based on case series. The interpretation and validation of these studies is hindered by limited correlation of plasma vitamin C concentrations with response. The methodology exists to determine if there is a role for high-dose IV vitamin C in the treatment of cancer, but the limited understanding of its pharmacodynamic properties makes this challenging. Currently, the use of high-dose IV vitamin C cannot be recommended outside of a clinical trial.
Introduction
Vitamin C is an essential micronutrient for humans, who lack the enzyme required for its synthesis. Vitamin C is well known for its antioxidant activity although it is only one of a large variety of dietary antioxidants. In the 1950s, vitamin C was originally hypothesized to be protective against cancer,1, 2 but in the 1970s, Ewan Cameron and Linus Pauling suggested that it also had a therapeutic effect, reporting increased survival of patients with advanced cancer following high-dose IV vitamin C treatment.3 In contrast, several subsequent randomized controlled trials (RCTs) of high-dose oral vitamin C failed to demonstrate a similar benefits,4-6 opening the issue of therapeutic effectiveness to controversy.
More recent studies showed that high plasma concentrations of vitamin C can only be achieved if vitamin C is administered intravenously or intraperitoneally as the rate of absorption from the gut is limited with oral administration.7, 8 Plasma concentrations in humans following high-dose IV vitamin C are approximately 200 times higher than those achieved following oral administration.8 High-dose oral and high-dose IV vitamin C treatment therefore have to be considered as distinct therapeutic approaches.
The purpose of this review is to survey the literature to analyze the antitumor effects of vitamin C in both human and animal studies in terms of the dose and achieved plasma vitamin C concentrations. In analyzing the effects of vitamin C in rodent models, it must be noted that rodents synthesize their own vitamin C and therapeutic effects are limited to “high plasma vitamin C.” The five questions this review will address are therefore (i) Does vitamin C (at high plasma concentrations) have antitumor activity in rodent systems? (ii) If so, how does vitamin C potentially exert its antitumor activity? (iii) Does vitamin C (at high plasma concentrations) have clinical antitumor activity? (iv) Is vitamin C safe alone and in conjunction with chemotherapy? and (v) If the answer is not clear, how can the issue of therapeutic efficacy be addressed in the future?
Does Vitamin C (at High Plasma Concentrations) Have Antitumor Activity in Rodent Systems?
Despite a large amount of preclinical in vitro data on the effects of vitamin C on tumor cells, there are few reports on the antitumor activity of high-dose vitamin C in xenografts of human tumors in immunodeficient mice, a well-characterized method of predicting potential antitumor activity in humans (Appendix I). Chen et al. demonstrated that vitamin C (4 g/kg intraperitoneal [IP] once or twice daily) reduced tumor growth and weight by 41–53% using ovarian (Ovcar5), pancreatic (PAN01) and glioblastoma (9 L) tumor cell lines in mice.9 The effective plasma concentration that decreased survival by 50% (EC50) was less than 10 mM in 75% of the tumor cells tested, but in contrast cytotoxicity was not evident in normal cells at ascorbate concentrations exceeding 20 mM. The lack of any complete response led them to propose that the role for vitamin C may be as an adjunct alongside chemotherapy. Similar results have been shown in mesothelioma cell lines.10 In keeping with this premise, Espey et al. demonstrated that vitamin C (4 g/kg) inhibited the growth rate of PAN02 pancreatic tumors by approximately 40%, and also augmented the effect of concurrently administered gemcitabine (30 mg/kg).11 This dose of vitamin C in rodents is equivalent to 1.5 g/kg in humans.12
How Does Vitamin C Potentially Exert Its Antitumor Activity?
One of the issues with the use of high-dose IV vitamin C is the lack of understanding of its potential mechanism of action. Although initial theories centered on modification of biological responses, more recent research has concentrated on the importance of both extracellular and intracellular effects of vitamin C. McCormick postulated that vitamin C protected against cancer by increasing collagen synthesis.1 Cameron and Pauling later hypothesized that the association between vitamin C and hyaluronidase activity was key. They speculated that a high intake of vitamin C increased the biosynthesis of a physiological inhibitor of hyaluronidase (PHI) and subsequently reduced the invasiveness of proliferative disease.13 Experimental studies, while demonstrating that scorbutic guinea pigs have higher levels of PHI compared with those receiving oral vitamin C, failed to show a link between PHI levels and the amount of vitamin C given.14
High-dose oral vitamin C has been shown to be a potent immunomodulator, enhancing the activity of natural killer (NK) cells in vivo.15 NK cells are thought to be important in immune surveillance, preventing growth and dissemination of tumor cells.16 Heuser and Vojdani demonstrated that vitamin C caused an increase not only in NK activity but also B- and T-cell activity in patients previously exposed to a toxic chemical.17 Both these studies used oral vitamin C. Other studies have found evidence to the contrary so while it is a potential mechanism, this remains unproven.18
Extracellular mechanisms for vitamin C action
Studies by Chen et al. established that high doses of vitamin C in mice (IV or IP) generated significant extracellular concentrations of hydrogen peroxide (H202).19 H202 is central to a diverse range of physiological responses pivotal to disease progression, including angiogenesis, and oxidative stress.20 They hypothesized that the presence of a catalyst such as ferric ion within the extracellular matrix of tumors could oxidize vitamin C to ascorbate radicals, which then could then donate electrons to oxygen to form superoxide radicals (O2−) (Fig. 1). This would be converted by superoxide dismutase to the potentially tumoricidal peroxide ion.19 Because peroxide is rapidly converted to oxygen in blood, the accumulation of peroxide would occur only in tissues. The identity of the catalyst was not elucidated but a later study suggested that it could be the metalloprotein ferritin, which is secreted by some tumor cells.22 Extracellular peroxide, depending on concentration, can have a cytotoxic effect.
Extracellular mechanism of action of high-dose intravenous vitamin C. Legend: Two possible mechanisms for the generation of extracellular hydrogen peroxide (H2O2) in response to ascorbate. In the first, as shown on the left-hand side, molecular oxygen is reduced to superoxide by a molecular complex of ascorbate and an as yet uncharacterized metalloprotein catalyst (such as ferritin).9 In the second, as shown on the right-hand side, ascorbate is first taken up by a cell such as a neutrophil, either directly by an ascorbate transporter or indirectly as dehydroascorbate by the glucose transporter. Here it stabilizes tetrahydrobiopterin (BH4), preventing its degradation and leading to activation of the enzyme nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, which reduces molecular oxygen to superoxide.21 In each case, peroxide is generated subsequently by the enzyme superoxide dismutase.
Chen et al. demonstrated that high (pharmacologic) but not low (physiological) concentrations of vitamin C killed cancer but not normal cells, with cell death dependent on extracellular vitamin C concentrations.9, 19, 23 H202 generation displayed a linear relationship with the formation of the vitamin C radical.23 The pattern of cell death changed from apoptosis to pyknosis/necrosis as vitamin C concentrations increased, in keeping with H202-mediated cell death.23 When H202 scavengers were employed they were protective against cell death.23 H202 generated by vitamin C oxidation and exogenously added H202 produced cell death curves that were indistinguishable.23
Chen et al. did not demonstrate a lower level of antioxidant enzymes (catalase, superoxide dismutase and glutathione peroxidase) in malignant cells to explain the selective death of cancer cells.23 Instead they theorized that H202 diffuses into sensitive cancer cells and causes toxicity by adenosine triphosphate (ATP) depletion.19 The selective cancer cell death may be explained by the different mechanisms of ATP generation with cancer cells primarily using anaerobic generation and normal cells aerobic generation. It remains uncertain whether this in vitro and in vivo data will translate into clinical benefit.
Intracellular mechanisms for vitamin C action
Vitamin C is taken into cells by sodium-dependent vitamin C transporter 1 (SVCT1) and SVCT2 sodium-dependent transporters that are members of the SLC23 family.24 The oxidized form of vitamin C, dehydroascorbate, can also be taken up by the glucose transporter (GLUT) and reduced in the cell to ascorbate.25 Macrophages and vascular endothelial cells can express high levels of SCVT2 and thus concentrate vitamin C to a high millimolar degree.26, 27
High intracellular vitamin C concentrations are proposed to inhibit hypoxia-inducible factor (HIF)-1α activation. HIF plays an important role in determining patterns of gene expression in cancer and is another potential target of vitamin C action. HIF-1α is broken down by hydroxylases, which require iron and ascorbic acid as cofactors.28 Vitamin C deficiency has been shown in vitro to compromise hydroxylation of HIF and upregulate HIF-1α.29, 30 HIF-1α overexpression promotes tumor progression through angiogenesis, confers resistance to chemotherapy and radiotherapy, and carries a poor prognosis.29, 31, 32
Vitamin C levels have been studied in patients with endometrial cancer, demonstrating significantly lower levels of vitamin C in high-grade tumors compared with paired normal tissue.33 Markers of HIF-1 activation (HIF-1α protein, GLUT-I and BCL 2/adenovirus E1B) were elevated in samples with low vitamin C levels, in keeping with the theory that low tissue vitamin C concentrations upregulate the HIF-1 pathway. There was also an inverse correlation between vascular endothelial growth factor levels and vitamin C concentrations. Similar findings have been demonstrated in gliomas.32
Genetic approaches and small-molecule inhibitors targeting HIF-1 have proven effective at decreasing resistance to chemotherapeutics in a number of different cancers.34, 35 A study using vitamin C with a recombinant adenovirus-associated virus (rAAV) vector bearing small-interfering RNA targeting HIF-1α (rAAV-siHIF) in pancreatic tumors in athymic mice found that vitamin C could inhibit expression of HIF-1α protein but not messenger RNA expression.36 It inhibited the growth in early and middle stages of disease but not advanced stages. The lack of blood supply in advanced stages is thought to compromise the delivery of vitamin C to the tumor.36
Does Vitamin C (at High Plasma Concentrations) Have Clinical Antitumor Activity?
In 1974 Cameron and Campbell published the first clinical trial suggesting the therapeutic role for vitamin C in cancer.37 Fifty patients with no further conventional treatment options were treated with IV and oral vitamin C (20% receiving oral only). Of five tumor regressions described, one occurred in a patient with ovarian cancer who had extensive pelvic disease at initial diagnostic laparotomy, which was not present at autopsy. However, this did not seem to prolong survival with the patient dying on day 33. Two other patients reported to have regression had no histological diagnoses of incurable disease. One of these patients diagnosed with advanced pancreatic cancer at laparotomy was found to have chronic pancreatitis on autopsy. A further patient with lymphoma had evidence of remission while on vitamin C, which recurred once treatment stopped. Remission was again achieved on restarting vitamin C.38 Another 48% of patients reported a subjective improvement quantified by a reduction in analgesic use and need for paracentesis.37 However, with no control group, a placebo effect for these symptomatic improvements cannot be excluded.
Cameron and Pauling then published two historically controlled trials, each comparing 100 patients treated with high-dose vitamin C, with 1000 controls matched by age, sex, tumor site and histological features (Fig. 2).3, 39 Both trials found a significant prolongation of mean survival (210 vs 50 days and 293 vs 39 days, respectively).3, 39 However, neither of these trials was standardized by two critical prognostic factors: performance status and stage.4, 40 The consistency of determination of “untreatability” is also controversial in the initial trial: 20% of the control group died within a few days of being deemed untreatable compared with none in the treatment group. The latter trial retrospectively analyzed the time from first hospital admission to date of untreatability (>1 year in 27% of patients in the treatment group and 23% in control group – not statistically significant) to address this concern.
Clinical studies of high-dose vitamin C in cancer. Summary of survival results from published studies on high-dose vitamin C administration to patients with cancer. Control patients are shown in yellow and patients treated with vitamin C in blue. The bars on the left-hand side represent trials with intravenous administration and those on the right with oral administration. +Reported mean survival. ++Reported median survival. ∧Randomized controlled trials. *Compared nil versus low versus high-dose vitamin C using combination of oral and intravenous (IV) dosing. **Compared low versus high-dose vitamin C using combination of oral and IV dosing.
The Mayo Clinic conducted three RCTs to examine the efficacy of vitamin C (Fig. 2), none of which showed a definitive benefit in terms of survival or quality of life.4-6 They compared 10 g of vitamin C administered orally versus placebo. The initial trial used patients unsuitable for further systemic therapy either because of progression during treatment or because their general condition precluded further treatment.4 These negative results were refuted due to concerns that, if vitamin C acts by improving host resistance, prior treatment would obscure any benefit.41, 42 All but nine patients had had previous treatment compared with only 4 of the 100 patients in the initial Cameron and Pauling trial.42
The second RCT was conducted in 100 cytotoxic-naive patients with colorectal cancer.5 None of the 38 patients with measurable disease demonstrated disease response.5 This trial was criticized as it was limited to patients with colorectal cancer and it was questioned whether negative results in this tumor group were transferable to other primary sites.5, 40 However, 20% of the patients in the initial Cameron and Pauling trial had colorectal cancer and they demonstrated similar survival benefit to other tumor subtypes.3, 39 Subsequently, 144 patients with predominantly lung and colorectal primaries were studied. They described an initial benefit in overall well-being but this was lost by 6 weeks.6
The results from these RCTs led to the current opinion among oncologists that high-dose vitamin C is ineffective. However, it is now recognized that vitamin C pharmacokinetics differ significantly with oral and IV dosing.8 Plasma vitamin C concentrations were not measured in any of these trials.
Around this time two further trials published similar results to the historical trials, demonstrating prolongation of survival times and improvement in quality of life (Fig. 2).43, 44 Cameron and Campbell published the only trial that measured plasma levels in association with survival time and demonstrated a linear relationship between dose and IV plasma vitamin C levels.43 Levels above 3 mg/dL (0.17 mM) were reported as desirable but this concentration based on recent literature appears too low to exert a pro-oxidant effect.43 Treatment was not randomized and was dependent on clinician preference, creating the potential for selection bias. Again there was no stratification by performance status.
Since this time case series and reports have continued to raise interest in a therapeutic role for high-dose Vitamin C, but there has been limited correlation with plasma vitamin C concentrations.38, 45-49 Riordan et al. and Padayatty et al. published the largest of these series with seven45 and three patients, respectively.46 These series cover a spectrum of malignancies but there is significant overlap, with many of these cases published repeatedly.37, 38, 45-47 Padayatty et al. reported on three patients with renal cell cancer (RCC), non-Hodgkin lymphoma (NHL) and bladder cancer, in keeping with the guidelines of the US National Cancer Institute Best Case Series Program.46 Two of these patients were also included in the Riordan series et al.45
All of these patients used IV vitamin C either with standard therapy or alongside other alternative therapies, making it impossible to definitively assign clinical benefit to vitamin C. They described positive results with either improved health status or slowed disease progression but with no control group this is inconclusive. RCC is a malignancy that has a variable natural history and one that rarely can undergo spontaneous regression (although usually following nephrectomy, which was not the case in this report).50 The latter two cases reported received standard therapy with radiation and surgery, respectively. All of these cases had the slim potential for long-term remission with these therapies.51 Interestingly, the case of NHL did have nodal relapse confirmed histologically that appeared to regress with vitamin C. There were no plasma vitamin C concentrations measured to help establish a dose–response relationship. There was often a lack of histological diagnosis in the case of recurrent or metastatic disease. In view of these factors, they do not provide definitive evidence of a beneficial or detrimental role for IV vitamin C.
Drisko et al. published a case series of two patients with advanced stage IIIc ovarian cancer treated with chemotherapy and IV vitamin C (60 g twice weekly).49 Both these patients had optimal surgical debulking, a key prognostic determinant of patient outcome.52 These patients were reported to demonstrate prolonged survival with both patients alive over 3 years out from diagnosis. Treated stage IIIc ovarian cancer has a 5-year survival rate of around 30%.53 The survival for these patients consequently may be explained by optimal conventional therapy. However, neither of these patients had subsequent chemotherapy over this time.
A recent retrospective multicenter epidemiological cohort study examined the effect of IV vitamin C (7.5 g weekly) on quality of life during adjuvant chemotherapy and radiotherapy and aftercare in patients with breast cancer.54 Mean intensity scores in patients treated with vitamin C during adjuvant therapy were improved (0.25 vs 0.4, P = 0.013), but the absolute difference was small and unlikely to be of clinical significance (score of 0 representing no symptoms and 1 representing mild complaints). They showed a mean Eastern Cooperative Oncology Group performance status during adjuvant therapy of 1.596 in the study group and 2.067 in the control (P = 0.002). Tumor status was reported to be stable at 6 and 12 months; however, longer follow-up is necessary to evaluate the effect on survival and relapse, two critical outcomes of adjuvant therapy.
In palliative patients, improvement in quality of life is an important component of care and a key end point.55 Yeom et al. investigated the effect of IV vitamin C (10 g twice a day for 3 days) on quality of life in 39 palliative patients.56 They demonstrated a significant improvement in quality of life with higher scores for physical, emotional and cognitive function and lower scores for fatigue, nausea and vomiting, pain and appetite loss (both P < 0.005) in a single assessment 1 week posttreatment.56 Although these results are suggestive of a benefit, with no control group a placebo effect cannot be excluded. The duration of benefit was also not assessed.
Cameron hypothesized palliative patients experience a strong “reverse placebo effect” because of previous treatment failures that counterbalance “placebo and anticipation” effects.3, 39 This has not been described in the literature elsewhere.
Clinical pharmacokinetics
Vitamin C has different functions at physiological and pharmacological plasma concentrations. Oral vitamin C administration is associated with tightly controlled plasma concentrations regulated by the pharmacokinetic principles of bioavailability and clearance.7 Saturation of bioavailability mechanisms occurs at oral doses of 400 mg daily equating to blood levels of 60–100 μM.8 IV dosing bypasses this tight control, achieving plasma concentrations up to 20 mM.7, 8
In vitro evidence suggests plasma concentrations of 10 mM are necessary for an antitumor effect and this appears achievable clinically only with IV administration.8 Padayatty et al. postulated that the vitamin C-free radical species, ascorbyl radical, forms only when human plasma concentrations are greater than 10 mM and that it is this radical or its unpaired electron that induces oxidative damage in cancer cells.8
Casciari et al. published a trial to determine the dose necessary to achieve this level in humans.57 In a single patient with colon cancer, they found that a dose of 60 g but not 30 g was effective at attaining this. A recent phase I trial of 24 patients demonstrated that IV vitamin C to a dose level of 1.5 g/kg three times per week was safe and achieved plasma concentrations >10 mM for several hours.58 While average follow-up was only 10 weeks, this trial did not demonstrate objective tumor response at these levels.58 Riordan et al. published a series of 24 patients treated with IV vitamin C 150 mg/kg/day and 710 mg/kg/day.59 The mean plasma level was 1.1 mM (below the expected therapeutic target). They did not demonstrate a correlation between dose and plasma concentration. The reason for this is unclear. Other factors such as critical illness, renal function and chemotherapy regimens may alter plasma vitamin C concentrations and affect the plasma concentration necessary for cytotoxicity.58
Is Vitamin C Safe alone and in Conjunction with Chemotherapy and Radiotherapy?
Vitamin C is generally regarded as an innocuous compound with a favorable therapeutic index (Table 1). While many centers have used high doses, there is limited published data around the safety at these doses. Cases of acute hemolysis in patients with underlying glucose-6-phosphate dehydrogenase (G6PD) deficiency have been reported in patients treated with high-dose vitamin C with at least one fatality described.62, 63 All patients should be screened for G6PD deficiency prior to starting vitamin C therapy.62
| Major side effects |
| Glucose-6-phosphate deficiency† |
| Renal stones – particularly oxalate stones† |
| Tumor acceleration |
| Minor side effects |
| Dyspepsia, nausea and altered bowel habit |
| Increase iron absorption‡ |
| Raise urinary uric acid levels and excretion of calcium and iron60 |
| Fluid overload – caution in patients with ascites, heart failure |
| Interfere with routine laboratory parameters – B12, glucose and fecal occult blood |
| Side effects falsely attributed to vitamin C60, 61 |
| Mutagenicity |
| Rebound scurvy |
| Infertility |
| Hypoglycemia |
| Destruction of vitamin B12 |
- †Absolute contraindication to intravenous vitamin C. ‡Important in patients with hemochromatosis.
Caution should also be taken in patients with a history of renal stones.60 Acute obstructive renal failure secondary to oxalate stones has been reported in a patient with underlying renal impairment.64 The effect of vitamin C on oxalate excretion is controversial with some believing excessive ingestion of AA increases the formation of oxalate stones.60
In patients with widespread and rapidly proliferating tumors, vitamin C has been reported to cause tumor acceleration and precipitate tumor hemorrhage and necrosis.37, 65 The initial Cameron and Campbell trial described four patients in this category.37 Potentially, this could also be explained by the natural history of the underlying cancer.
Dyspepsia, nausea and altered bowel habit are the most frequently reported side effects, particularly following oral administration.59, 60 High doses of oral vitamin C have been shown to affect iron absorption and interfere with many routine laboratory parameters.59, 60 In patients with congestive heart failure and ascites, the high fluid intake associated with administration may exacerbate their condition.60
Role of vitamin C in combination with chemotherapy and radiation
The literature reports that 30–95% of patients with cancer try unconventional therapies, with the majority using these as adjuncts to their standard care with the intention to improve their quality of life and symptom control.66-68 Despite this wide use, it remains unclear whether the concurrent use of antioxidants with chemotherapy and radiotherapy is beneficial or detrimental.69 Because of the paucity of clinical trial evaluation, the evidence to date is mostly derived from in vitro and in vivo data, and observational records. There are no published RCTs examining high-dose IV vitamin C in conjunction with chemotherapy or radiotherapy, making it difficult to definitively assess safety and efficacy.
Vitamin C has been studied in combination with a number of cytotoxic agents in vitro and in vivo with conflicting outcomes on efficacy11, 70-72 (see Appendix II). It is theorized that vitamin C may sensitize refractory cancers to radiotherapy and chemotherapy.11, 70-73 Koch and Biaglow studied dehydroascorbic acid alongside radiation in hypoxic Ehrlich cells in ascites in vivo, demonstrating increased inhibition of cell growth with half the radiation dose.74 Similar findings were found in neuroblastoma and glioma cell lines treated with 5-fluorouracil (5FU), sodium d-ascorbate and radiation.75 Contrary to this, Witenberg et al. demonstrated a reduction in apoptosis from ionizing radiation in myeloid leukemia cells treated with dehydroascorbic acid.76
This sensitizing effect is postulated to be due to the increased H202 generation secondary to vitamin C administration. However, vitamin C may also result in a reduction in HIF-1, which in xenograft models has been shown to be associated with heightened radiation sensitivity.77 In keeping with this theory, putative small-molecule inhibitors of HIF-1 have demonstrated enhanced tumor responsiveness to radiation in vitro, supporting the use of this as a target in association with conventional therapies.78
High-dose vitamin C is also postulated to reduce the toxicity of chemotherapy because of restoration of plasma vitamin C concentrations and thus antioxidant capacity.54 In vitro evidence has also suggested that vitamin C may reduce the cardiac toxicity associated with doxorubicin without compromising efficacy, postulated to be related to peroxidation of cardiac lipids.79
How Can the Issue of Therapeutic Efficacy Be Addressed in the Future?
There are a number of trials underway in both the phase I and II setting (Table 2). Although the methodology exists for investigating the role of high-dose vitamin C in cancer therapy, it is hindered by uncertainties including the target population and markers and predictors of response. The inconsistency and current level of evidence to support a clear scientific rationale also makes the likelihood of funding for the necessary research problematic.
| Investigator | Phase | Title | Dose of vitamin C | Objectives |
|---|---|---|---|---|
Levin NCT00441207 |
I | A phase I study of high-dose IV vitamin C treatment in patients with solid tumors | Not specified | Evaluate safety and tolerability of IV vitamin C. Observe evidence of tumor response to vitamin C and compare the level of fatigue, pain control, quality of life before and after vitamin C. |
Hoffer et al. NCT01050621 |
I–II | Phases I–II clinical trial of combination conventional cytotoxic chemotherapy and IV vitamin C in patients with advanced cancer or hematological malignancy for whom cytotoxic chemotherapy alone is only marginally effective | Dose escalation beginning with 0.9 g/kg escalating to 1.5 g/kg IV two to three times per week, bracketing chemotherapy | Evaluate safety and tolerability of IV vitamin C when administered alongside cytotoxic therapies. To assess tumor response. To assess effect on quality of life. Determine the effect of chemotherapy on pharmacokinetics of IV vitamin C. |
Mikines NCT01080352 |
II | Evaluation of cytotoxicity and genetic changes of high-dose IV vitamin C infusions in castration-resistant metastatic human prostate cancer | 20 g IV vitamin C weekly | PSA change after 12–20 weekly treatments. Changes in bone metastases and markers of bone activity (bone-specific ALP, PINP, NTX). Pharmacokinetics in elderly cancer patients. |
Monti NCT00626444 |
II | Pilot trial of IV vitamin C in refractory non-Hodgkin lymphoma (NHL) | IV vitamin C to achieve plasma level of 300–350 mg/dL given three times per week | Evaluate safety and tolerability of IV vitamin C and tumor shrinkage. |
Edman NCT00954525 |
I | IV vitamin C in combination with standard chemotherapy for pancreatic cancer | 50, 75 or 100 g IV vitamin C three times per week | Evaluate safety and tolerability of IV vitamin C and tumor shrinkage. |
| Drisko | II | Safety of oral antioxidants and IV vitamin C during gyn cancer care | Oral and IV vitamin C two to three times per week (individual doses not specified) | Evaluate safety of adding high-dose Antioxidants to chemotherapy in the treatment of gynecologic malignancies (uterine, cervical or epithelial ovarian). Evaluate tumor response rates in patients with gynecologic malignancies treated with antioxidants to include IV and oral ascorbic acid, IV glutathione, oral mixed carotenoids, mixed tocopherols and vitamin A. |
- ALP, alkaline phosphatase; IV, intravenous; NTX, N-terminal telopeptide; PINP, procollagen type I N-terminal propeptide; PSA, prostate-specific antigen.
The ready availability and accessibility of IV vitamin C to patients makes a true placebo-controlled trial difficult, with crossover likely to be a major confounding factor. Comparison of high-dose oral versus high-dose IV vitamin C (at least 1.5 g/kg three times per week as used in phase I trials) may address this, that is, low plasma concentration versus high (Table 3). A third placebo-controlled arm could be added.
| Factors to be considered | Recommendations |
|---|---|
| Population | 1. Palliative patients with no further chemotherapeutic options |
| 2. Alongside chemotherapy in refractory disease – ideal to use alongside one specific agent | |
| 3. One tumor subtype on observation alone, for example, good prognosis renal cell cancer | |
| 4. Across tumor subtypes based on initial data from Cameron trial | |
| Trial arms | 1. Comparison of oral high-dose (low plasma concentration) vitamin C versus IV high-dose (high plasma concentration) vitamin C |
| 2. Three arm trial with placebo versus oral high-dose vitamin C versus IV high-dose vitamin C | |
| 3. Placebo versus high-dose IV vitamin C | |
| 4. Similar arms as described above alongside chemotherapy or radiotherapya | |
| Pharmacokinetics | 1. Baseline and on-treatment assessment of plasma vitamin C levels |
| 2. If alongside chemotherapy, pharmacokinetic studies of the chemotherapeutic agent | |
| Markers of response | 1. Pre- and posttreatment biopsy if easily obtainable tissue |
| 2. Catalase genotype evaluation as subgroup analysis to determine if there is a potential target population with increased efficacy | |
| 3. CT/MRI to assess response – timing of imaging remains controversial | |
| 4. PET-CT scan to assess tumor metabolic activity pre- and posttreatment | |
| End points | 1. Assessment of tumor response |
| 2. Progression-free and overall survival | |
| 3. Quality of life assessment |
- a Addition of vitamin C alongside chemotherapy or radiation treatment has a number of ethical concerns because of the potential for both a beneficial or detrimental interaction. Vitamin C could be studied in a population who have developed chemoresistance and have no further treatment options. CT, computed tomography; MRI, magnetic resonance imaging; PET, positron emission tomography.
The target population remains unclear. There is no clear cancer type or phase of care defined for the role of vitamin C. In view of this, vitamin C could be trialed as an isolated treatment in patients who have exhausted conventional treatments. An alternative approach would be to use vitamin C in patients on observation such as those with asymptomatic biochemical progression of ovarian or prostate cancer, asymptomatic pulmonary metastases in good prognosis RCC or indolent low-grade NHL. Regardless of the population used, patient safety remains paramount. All patients should have a G6PD screen and caution should be exercised in patients with a history of renal stones.
Baseline and on-treatment vitamin C levels should be assessed and used to titrate dose to achieve a concentration of at least 10 mM based on in vitro data.8 This would help improve knowledge on the pharmacokinetic properties of vitamin C and help establish a dose–response relationship.
Based on experience gained from chemotherapeutic agents, further delineation of the mechanism of action may identify a target population with higher response rates. If vitamin C acts via H2O2 formation, population-based differences in the genotype and phenotype of catalase expression and activity may be relevant.80 Those with low levels of catalase activity may be more sensitive to the effect and toxicity of high-dose vitamin C. Assessment of this in a clinical trial would help determine if this theoretical sensitivity translates to clinical response.
It remains uncertain whether the use of vitamin C in conjunction with chemotherapy and radiotherapy has a beneficial or detrimental interaction. Trials using vitamin C alongside chemotherapy need to include analyses of the pharmacokinetic properties of both vitamin C and the chemotherapeutic agent. In vitro and in vivo data suggest vitamin C may overcome chemoresistance and improve chemosensitivity.11, 70-73 This has been demonstrated in vivo in pancreatic cancer cell lines in combination with gemcitabine, making this a potential population to start with.11
Biopsies pre- and post treatment may help identify predictors and markers of response. This has not been performed in the trials to date. Method and timing of assessment of tumor response is hindered by our limited understanding of the mechanism of action and lack of specific biomarkers. Improvement of our translational knowledge is critical for future research. Fluorodeoxyglucose-positron emission tomography could be used as a surrogate marker of response by assessing the effect of high-dose vitamin C on tumor metabolism. There is evidence to support the use of PET to assess response to other modalities such as chemotherapy and chemoradiation.81
Conclusion
Although the rates of utilization of vitamin C therapy remain uncertain, its popularity has increased over the years since the first suggestion of its chemotherapeutic activity in the 1970s. Although there is currently no definitive evidence that IV vitamin C improves quality of life, progression-free or overall survival, the published RCT data do not negate potential benefit based on an improved understanding of vitamin C pharmacokinetics.
The pharmacokinetic properties of IV and oral vitamin C are critical in interpreting the data to date. Based on in vitro data, it is now recognized that the plasma levels necessary for cytotoxicity requires IV dosing. However, a phase I trial of 24 patients did not demonstrate objective responses despite using IV doses of up to 1.5 g/kg.58
Vitamin C is well known for its antioxidant activity, but it is the proposed pro-oxidant activity at high concentrations that remains controversial. This situation is perpetuated by the lack of a defined mechanism of action. While intracellular and extracellular generation of H2O2 is the most common theory, clarification of this and determination whether this will translate to a clinical benefit is critical in future research.
Although high-dose IV vitamin C appears relatively innocuous given alone, it does have the potential to cause harm in patients with G6PD deficiency and previous renal stones. It remains uncertain whether vitamin C is clinically safe when given alongside chemotherapy and radiotherapy.
Despite 40 years of research since the initial reports on high-dose IV vitamin C, its use remains controversial. The methodology exists to determine if high-dose IV vitamin C does have an anticancer effect, but the ability to design and conduct studies is impaired by the lack of a consistent scientific rationale, the ready availability and the use of this agent by practitioners already convinced by current evidence. This makes a placebo-controlled trial difficult.
There are highly polarized views on the use of high-dose vitamin C for cancer treatment, with passionate advocates balanced by passionate critics. This is a key reason for why carefully controlled clinical trials, rather than a review of the literature, are needed to obtain a clear view of this field.
Acknowledgment
No funding for this project.
Appendix I
Data in Cell Lines
| Cell lines | Author | Type – human/animal | Dose or target ascorbate concentration | Results | Proposed mechanism of action |
|---|---|---|---|---|---|
| Ehrlich ascites carcinoma and leukemia cells | Poydock82 | Mouse | Vitamin C 0.4 g/kg and B12 | Inhibits cell growth but no effect on normal fibroblasts | |
| Hepatocellular carcinoma, bladder, breast, prostate and cervix | Verrax and Calderon83 | Mouse and human | Ascorbate concentration 50 μM to 33 mM | Decrease tumor growth rate | H202-mediated cytotoxicity |
| Hormone-refractory prostate cancer | Pollard et al.84 | Mouse | Ascorbate 4 g/kg | Inhibit tumor growth by approximately 50% | Not discussed |
| Human leukemic, preleukemic and myeloma | Park and Kimler85 | Human | L-ascorbic acid – minimum effective dose of 0.03 mM | Both enhancement and suppression of growth. Consistent effect within individual cell lines | Multistep model with oxidation reduction and/or electron transfers that may finally lead to free radical scavenging and cell protection or to generation of free radicals and cell death |
| Leukemic cells | Park et al.86 | Human | L-ascorbic acid 0.3 mM | Suppress in vitro growth | |
| Lymphoid neoplasm, lymphocytic leukemia, acute lymphoblastic leukemia, epidermoid carcinoma, fibrosarcoma | Leung et al.87 | Mouse and human | Sodium ascorbate, d-isoascorbic acid | Cytotoxic to malignant cell lines but not nonmalignant cells | Complex – likely deplete amino acids essential for cell growth; induce chromosomal aberrations; production of H202 |
| Melanoma | Bram et al.88 | Mouse and human | L-ascorbic acid 0.2–0.5 mM | Twofold to 10-fold more sensitive to ascorbate | Vitamin C essential for melanin biosynthesis and preferentially incorporated into melanoma tissue. H202-mediated cytotoxicity |
| Mesothelioma | Takemura et al.10 | Human | Ascorbate concentration 50–1000 μM | Dose-dependent reduction in cell viability | Production of reactive oxygen species particularly H202 accompanied by disruption of mitochondria structure |
Neuroectodermal cell lines Neuroblastoma and melanoma |
De Laurenzi et al.89 | Human | Ascorbate concentration 10 nM to 1 mM | Slower growth at low concentrations in NB cells (10–100 nM). Increased cell death at 1 mM in both melanoma and NB cells | Induces programmed cell death with DNA fragmentation likely related to H202 formation |
| Ovarian, pancreatic and glioblastoma | Chen et al.9 | Human | Ascorbate 4 g/kg | Inhibit cell growth | H202-mediated cytotoxicity |
| Pancreatic | Du et al.90 | Human | Ascorbate concentration 0–20 mM | Inhibited tumor growth and prolonged survival | H202-mediated cytotoxicity |
| Pancreatic | Espey et al.11 | Mouse and human | Ascorbate 4 g/kg | Inhibited tumor growth | H202-mediated cytotoxicity |
| Promyelocytic | Sestili et al.91 | Human | Ascorbate concentration 300 μM | Potentiate cell death | Similar mechanism to H202 as reversed with catalase |
- Neuroblastoma (NB) doses of 4 g/kg in rodents are proposed to be equivalent to 1.5 g/kg in humans.
Appendix II
Studies of Chemotherapy with Vitamin C
| Type of cancer | Author | Setting | Dose or concentration of vitamin C | Chemotherapy | Results | |
|---|---|---|---|---|---|---|
| Preclinical in vitro – negative results | ||||||
Leukemia Lymphoma |
Heaney et al.92 | Cell lines and murine xenografts | 0 and 18 mM | Doxorubicin Cisplatin Methotrexate Vincristine Imatinib |
Detrimental results | Used dehydroascorbic acid |
| Preclinical in vitro – positive results | ||||||
Breast carcinoma Endometrial adenocarcinoma Oral epidermoid carcinoma |
Noto et al.93 | Human cell lines | Vitamin C 1 μM to 10 mM Vitamin K3 10–105 nM |
Effective in combination but higher doses needed for less effect for individual vitamins | Mediated through H202 production | |
Breast cancer Colon fibroblast Melanoma Pancreatic cancer Skin fibroblast |
Casciari et al.57 | Human cell lines | Ascorbic acid 10 mM for 2 days | Doxorubicin | Potentiate cell death at high levels. Additive effect at high levels with doxorubicin but at levels that may not be clinically feasible. At low doses reduced benefit of doxorubicin. No clear synergistic role with vitamin K3 |
|
| Breast | Kurbacher et al.71 | Human cell lines | 1 μM or 100 μM | Doxorubicin Cisplatin Paclitaxel |
Synergism with doxorubicin Partly synergistic with cisplatin and paclitaxel |
Formation of oxyradicals in use with doxorubicin and cisplatin. Possible action mediated via p-glycoprotein. Dose-related phenomenon – low doses with doxorubicin, with the highest concentrations necessary in conjunction with cisplatin |
| Cervical | Reddy et al.73 | Human cell lines | 0.1 μM to 10 mM | Cisplatin Etoposide Adriamycin Bleomycin |
Potentiates | Down regulates AP-1 and stabilizes p53 |
| Lymphoma | Dai et al.94 | Mouse cell lines | 500 μM | Arsenic | Potentiate | Deplete GSH and increase H202 death |
| Multiple myeloma | Grad et al.70 | Human cell lines | 0 and 100 μM | Arsenic | Potentiate benefit | Deplete GSH and increase H202 death |
| Nonsmall-cell lung cancer | Song et al.72 | Human cell lines | 25 μg/mL | Vincristine | Potentiate by reducing resistance | Speculated to be due to modification of cell membrane |
| Neuroblastoma | Prasad et al.95 | Mouse cell lines | L-ascorbic acid at 500 μg/mL (2.53 mM) and sodium d-isoascorbate | 5-Fluorouracil (5FU) X-irradiation Bleomycin Prostaglandin E1 Sodium butyrate Vincristine 6-thioguanine CCNU Methotrexate DTIC |
Potentiate 5FU, irradiation, Bleomycin, Prostaglandin E1, Sodium butyrate. No effect on vincristine, 6-thioguanine, CCNU. Reduced cytotoxicity of DTIC and methotrexate |
Multiple mechanisms of action: inhibition of catalase leading to accumulation of H202 and subsequent cell death |
| Esophageal | Abdel-Latif et al.96 | Human cell lines | 20 mM | Cisplatin 5FU |
Potentiate | Inhibit translocation of NF-κB and AP-1 |
| Preclinical in vivo – positive results | ||||||
Bladder Breast Cervix Hepatocarcinoma, Prostate |
Verrax and Calderon83 | Mouse model | Ascorbate concentration 50 μM to 33 mM | Etoposide Cisplatin 5FU Doxorubicin Paclitaxel |
Potentiate in combination | H202-mediated cytotoxicity |
| Liver ascites | Taper and Roberfroid97 | Mouse model | Vitamin C 1 g/kg Vitamin K3 10 mg/kg |
Vincristine | Potentiates in combination but not alone | Through H202 production |
| Pancreatic | Espey et al.11 | Mouse model | Vitamin C 4 g/kg | Gemcitabine | Potentiates in combination with activity alone | Through H202 production |
| Urothelial | Kassouf et al.98 | Mouse model | Vitamin C 23 g/L Vitamin K3 0.23 g/L |
Gemcitabine | Potentiates in combination | Increased apoptosis |
| Clinical trials | ||||||
| Multiple myeloma | Qazilbash et al.99 | Phase II (n = 48) | Vitamin C 1000 mg IV for 5 days | Melphalan ± arsenic and ascorbic acid | Neutral – longer follow-up necessary |
- AP-1, activator protein 1; CCNU, lomustine; DTIC, dacarbazine; GSH, glutathione; IV, intravenous; NF-κB, nuclear factor κB.
