The Oncologist, Vol. 12, No. 7, 808-815, July 2007; doi:10.1634/theoncologist.12-7-808 © 2007 AlphaMed Press
Commentary: A Case for Minimizing Folate Supplementation in Clinical Regimens with Pemetrexed Based on the Marked Sensitivity of the Drug to Folate AvailabilityDepartments of Medicine and Molecular Pharmacology, The Albert Einstein College of Medicine Cancer Center, Bronx, New York, New York Key Words. Antifolates • Folate pools • Folic acid • Pemetrexed Correspondence: I. David Goldman, M.D., Departments of Medicine and Molecular Pharmacology, The Albert Einstein College of Medicine Cancer Center, Bronx, New York 10461, USA. Telephone: 718-430-2302; Fax: 718-430-8550; e-mail: igoldman{at}aecom.yu.edu Received November 27, 2006; accepted for publication April 23, 2007.
Pemetrexed is a novel antifolate recently approved for the treatment of pleural mesothelioma and non-small cell lung cancer. In clinical regimens, pemetrexed is administered in conjunction with folic acid to minimize toxicity. However, excessive folate supplementation may also diminish the activity of this agent. The current study demonstrates, in several human solid tumor cell lines, that when extracellular 5-formyltetrahydrofolate levels are increased in vitro, within the range of normal human blood levels, there is a substantial decrease in pemetrexed activity upon continuous exposure to the drug. This was accompanied by a comparable lower level of trimetrexate activity consistent with an expansion of tumor cell folate pools. Likewise, when cells were exposed to pemetrexed with a schedule that simulates in vivo pharmacokinetics, there was markedly less cell killing with higher extracellular folate levels. Data are provided to indicate that 5-formyltetrahydrofolate is an acceptable surrogate for 5-methyltetrahydrofolate, the major blood folate, for this type of in vitro study. These observations and other reports suggest that, in view of the rise in serum folate and fall in serum homocysteine that has accompanied folic acid supplementation of food in the U.S., the addition of folic acid to regimens with pemetrexed should be limited to the lowest recommended level that provides optimal protection from pemetrexed toxicity. Disclosure of potential conflicts of interest is found at the end of this article.
Pemetrexed (Alimta®; Eli Lilly and Company, Indianapolis, IN) is a new-generation antifolate recently approved for the treatment of mesothelioma and non-small cell lung cancer based on the results of two phase III trials [1, 2]. Pemetrexed (PMX), as administered, is a monoglutamate with only a modest level of inhibitory activity at its target enzymes. However, following transport into susceptible tumor cells, potent polyglutamate derivatives are formed [3, 4]. The pentaglutamate of PMX (Ki = 1.3 nM) is an approximately 100-fold better inhibitor of thymidylate synthase (TS) than the monoglutamate (Ki = 109 nM) [5, 6]. Similarly, the pentaglutamate is a far more potent inhibitor of PMX's secondary target, glycinamide ribonucleotide transformylase (GARFT) (Ki = 65 nM), than the monoglutamate (Ki = 9,300 nM) [5, 6]. Because formation of polyglutamate derivatives is a key determinant of the activity of PMX, factors that influence this process are critical to drug action. One key element is the level of physiological folate cofactors within cells. As is the case for methotrexate [7], cellular folates inhibit PMX polyglutamation at the level of folypolyglutamate synthase (FPGS) [8]. Physiological folates also compete with PMX at its target enzymes, TS and GARFT [9]. This was well documented in murine leukemia cells where, as the level of cellular folates was increased, by increasing extracellular folate, there was a marked fall in the activity of a variety of antifolates [8]. PMX was among those antifolates that was most influenced by a higher intracellular folate content. The converse effect, greater PMX activity, is seen with folate depletion [10, 11]. This effect of physiological folates on PMX activity must be considered within the context of current clinical regimens that mandate folic acid and B12 administration prior to, and concurrent with, PMX. This was based upon the observations that (a) PMX toxicity correlates with an increasing plasma homocysteine level, which is the most sensitive indicator of folate and/or vitamin B12 deficiency [12]; (b) added folic acid markedly reduces drug toxicity and allows the administration of substantially more cycles of PMX [1, 13, 14]; and (c) there is a greater therapeutic window for PMX in mice fed a folic acid–replete, as compared with a folic acid–deficient, diet [15]. This use of folic acid supplementation followed earlier reports by Grindey and coworkers demonstrating the adverse impact of both folate depletion and folate excess on the activity of dideazatetrahydrofolate. This led to regimens in which folic acid was coadministered with, and substantially reduced the toxicity of, this agent [16, 17].
Initial clinical studies with PMX included patients from countries where cereal-grain products were not supplemented with folic acid and were initiated before foods in the U.S. were fortified with folic acid. This fortification has increased blood folate levels in the U.S. more than twofold with a corresponding twofold decrease in the percentage of the population with elevated homocysteine levels [18]. Further, current guidelines leave the folic acid dose up to the administering physician over a threefold range of 350 µg to 1 mg/day. Population-based studies have identified a linear relationship between folate intake and blood folate level [19, 20]; a meta-analysis indicated a 2.5 ng/ml ( The current study evaluates the impact of extracellular folate on PMX cytotoxicity in human solid tumor cell lines in vitro. Because the physiological blood folate 5-methyltetrahydrofolate (5-CH3-THF) undergoes oxidative degradation in vitro, 5-formyltetrahydrofolate (5-CHO-THF), which shares many similar characteristics, was used as a surrogate. The extracellular 5-CHO-THF concentrations encompassed the current normal range of plasma folate levels in the U.S. (13.1–74.3 nM; mean, 30.5 nM) [23].
Chemicals and Cell Lines PMX was provided by Eli Lilly and Company (Indianapolis, IN); (6S) 5-CHO-THF was purchased from Schircks Laboratories (Jona, Switzerland). (6 R,S) 5-CH3-THF, ß-mercaptoethanol (BME), and cyanocobalamin (vitamin B12) were from Sigma (St. Louis, MO). Trimetrexate (TMQ) was obtained from Parke-Davis (Ann Arbor, MI). HCT-15 colon cancer and A549 and H1299 lung cancer cell lines were obtained from American Type Tissue Culture Collection (Manassas, VA). The human mesothelioma cell lines NCI-2052 and NCI-2373 were obtained from the National Cancer Institute (Bethesda, MD).
Cell Culture
Growth Inhibition Studies
Pulse Exposure to PMX
Statistical Analyses
Effect of Extracellular 5-CHO-THF or 5-CH3-THF Level on PMX and TMQ Growth Inhibition To examine the equivalence of 5-CHO-THF and 5-CH3-THF as folate substrates under these in vitro conditions, HCT-15 colon cancer cells were adapted for at least 1 week to either 25 nM 5-CHO-THF or 80 nM 5-CH3-THF. The half-life of 5-CH3-THF in culture medium was previously reported to be approximately 24 hours in the presence of BME [24]. Therefore, to maintain levels of 5-CH3-THF, 0.1 mM BME was added to the medium and half the concentration of 5-CH3-THF (40 nM) was replenished in the medium once daily. Following this, growth inhibition by PMX was assessed. Differences in growth inhibition by TMQ were also assessed as a sensitive measure of change in cellular folate cofactor pools. The TMQ IC50 was virtually identical whether the cells were grown in 5-CH3-THF or 5-CHO-THF (Table 1). The addition of 0.1 mM BME to the medium with 5-CHO-THF had no effect on TMQ sensitivity (data not shown). Use of 5-CH3-THF was not limited by levels of cyanocobalamin (vitamin B12), because the addition of 50 ng/ml of this vitamin to the medium (a tenfold excess) did not affect TMQ sensitivity (data not shown). This lack of change in sensitivity to TMQ indicates that cellular folate pools were comparable whether cells were grown in 5-CHO-THF or 5-CH3-THF. Similarly, the sensitivity of HCT-15 cells to PMX was unchanged whether cells were grown in 5-CHO-THF or 5-CH3-THF (Table 1).
Impact of Extracellular Folate Level on the Sensitivity of Human Solid Tumor Cell Lines to PMX or TMQ Concentrations of 5-CHO-THF were chosen to encompass the physiological range of 5-CH3-THF blood levels [23] and those observed with folic acid supplementation [20]. After 1 week of adaptation to these folate concentrations, cells were exposed to either PMX or TMQ. As depicted in Table 2, sensitivity of HCT-15 cells to TMQ was markedly decreased with higher concentrations of extracellular 5-CHO-THF, consistent with substantially greater cell folate pools. This was accompanied by a comparable, marked increase in PMX IC50 of 9.7 ± 0.9 nM to 183.3 ± 16.7 nM (Table 2). Over an extracellular 5-CHO-THF range of 10–62.5 nM there was a sixfold increase in the PMX IC50 in HCT-15 cells. Large increases ( 3.6-fold) in PMX and TMQ IC50s were also observed for A549 lung cancer, H2052 (Fig. 1 and Table 2), and H2373 human mesothelioma cell lines over this range of extracellular 5-CHO-THF levels with IC50s indicated in Table 2. Hence, higher extracellular folate levels correlate inversely with PMX activity in vitro. This is illustrated further in Figure 2, where IC50 is plotted as a function of the log of the extracellular 5-CHO-THF level. There were larger differences in the impact of extracellular folate on TMQ activities among the different cell lines. Hence, there is variability in the extent to which tumors respond to changes in the availability of folates.
Effect of Extracellular Folates on Pulse Exposure to PMX That Simulates In Vivo Pharmacokinetics in Clinical Regimens Because PMX is administered clinically as an i.v. bolus, tumor cells in vivo are exposed to high concentrations of the drug for brief periods after which extracellular PMX levels decline rapidly as the drug is cleared from plasma. To reproduce these conditions, an exposure schedule was formulated based on published PMX pharmacokinetics [29]. This exposed cells, growing under different levels of extracellular 5-CHO-THF, to PMX concentrations that were lowered over three steps, beginning 1 day postseeding and ending day 3 postseeding (Fig. 3). Viable cells were then counted on day 4, 5, or 7 and expressed as a percentage of untreated cells. As depicted in Figure 3, when H1299 lung cancer cells were treated with all three steps of PMX exposure, there were only 4%–7% viable cells in the groups growing in 4–25 nM 5-CHO-THF but approximately 28% of cells growing in 62.5 nM 5-CHO-THF were viable. Further, while cells growing at higher folate conditions of 25 or 62.5 nM 5-CHO-THF were able to recover and nearly triple in number in the 3 days following PMX exposure, cells growing under lower folate conditions of 4 nM or 10 nM 5-CHO-THF had lower growth rates, implying that most of these cells remained growth-arrested. When cells were exposed to only the first step, or the first two steps, of PMX exposure, there was an incrementally higher percentage of viable cells with higher extracellular 5-CHO-THF levels.
The current study demonstrates the marked decrease in PMX activity that accompanies an increase in the availability of folates. These observations are similar to findings in murine leukemia cells [8] and are consistent with reports by other laboratories [10, 11], all indicating the adverse impact of increasing physiological folates on PMX activity. What is particularly striking in the current study is the degree to which PMX activity, measured by growth inhibition or percentage of cell kill, decreases as the extracellular folate level is increased over the physiological range of folate blood levels [23]. It is, however, important to point out that the impact of these in vitro folate concentrations on PMX activity does not necessarily extrapolate quantitatively to the in vivo setting and may not reflect the alterations in PMX activity that might occur at similar actual blood levels. The point here is the impact of the trend—higher extracellular folate levels are consistent with lower PMX activity. These folate concentrations are relevant to blood folate levels in the U.S. following implementation of folic acid supplementation [20]. In countries like The Netherlands, folic acid is not added to food and blood folate levels are lower [30]; however, serum folate levels increased from 14 nM to as high as 75 nM with folic acid supplements of up to 800 µg daily [22]. The lower PMX activity occurred whether exposure to PMX was continuous or in a graded-pulse protocol that simulates in vivo pharmacokinetics. Indeed, folic acid supplementation in current PMX regimens is likely to increase blood folate levels beyond the highest normal values reported in the most recent Centers for Disease Control survey in the U.S. and used in the current study [23]. Current clinical regimens require the routine administration of folic acid and vitamin B12 beginning at least 1 week prior to, and concurrent with, PMX chemotherapy. This vitamin supplementation reduces toxicity and improves outcome in patients treated with PMX [1, 13, 14]. While the benefits of vitamin supplementation are clear, the optimal level of folate supplementation required is not. This is of particular importance now that food in the U.S. is supplemented with folic acid and because many patients take multivitamins and other supplements that contain folic acid. Also of concern is the wide range of folic acid supplementation (350–1,000 µg/day) to be started 1–3 weeks prior to the first PMX dose [31] or, as recommended by the manufacturers, at least five daily doses beginning 1 week prior to starting PMX [32], in all cases left to the discretion of the administering physician. If 350 µg/day of folic acid, with five doses 1 week prior to PMX, is considered to be sufficiently protective, it makes little sense to allow a nearly threefold higher dose of 1 mg/day, for up to 3 weeks prior to PMX, which has the potential for diminishing the activity of the drug.
PMX hematological and gastrointestinal (GI) toxicities correlate closely with increasing plasma homocysteine levels [12]. However, homocysteine reduction approaches a nadir as the folic acid dose is increased; 90% of the maximum reduction in blood homocysteine can be achieved with 400 µg folic acid per day [21, 22]. In the U.S., mean plasma folate levels have increased with folic acid enrichment of grain products (effective since 1998) from approximately 11 nM/l to 23–30 nM/l [18, 23, 33] and nearly 80% of the population in the U.S. now has homocysteine levels <9.0 µM [23]. Hence, it is unclear as to when the level of folic acid intake from all sources produces a maximum decrease in toxicity beyond which higher doses may decrease the activity of PMX and how this relates to homocysteine blood levels. While the risk for developing PMX toxicity increased threefold in patients with homocysteine blood levels >11.5 µM [12], there was only about 30% less PMX toxicity in patients who had homocysteine levels of <7.5 µM as compared with patients with homocysteine levels within the normal range (7.5–11.5 µM), and the latter difference was not statistically significant. Further, studies on the effect of vitamin supplementation on PMX response rates and survival suggest improvement in patients with gastric cancer and mesothelioma but lower response rates in patients with breast cancer [34]. Interestingly, patients with breast cancer in that report had lower mean levels of homocysteine ( Based on all available information, excessive folate supplementation can only have an adverse impact on PMX activity and should be avoided. Hence, a reasonable guideline at this time should be a recommended dose of folic acid supplementation at 350 µg/day or, at most, the 400 µg/day that tends to be standard in multiple vitamins, with five daily doses within 1 week before the start of PMX. Because patients often wish to take a multiple vitamin, folic acid may be administered as a constituent of a multiple vitamin (350–400 µg/day) and patients should be instructed specifically to take no other vitamin or food supplement that might contain folic acid. While this paper focuses on the folic acid component of supplementation, vitamin B12 (1,000 µg i.m. the week preceding PMX then every three cycles thereafter) is an indispensable component of this regimen because pretreatment methylmalonic acid levels independently predict GI toxicity (diarrhea and mucositis) [12]. There is no evidence that excess vitamin B12 has any negative effect on PMX activity.
The authors indicate no potential conflicts of interest.
This work was supported by grants from the National Institutes of Health, CA-82621, and the Mesothelioma Applied Research Foundation Alvin Rehbeck Memorial Grant.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||