Single-Dose Pharmacokinetic Study of Lycopene Delivered in a Well-Defined Food-Based Lycopene Delivery System (Tomato Paste-Oil Mixture) in Healthy Adult Male Subjects

Epidemiological evidence suggests that high consumption of vegetables and fruits is associated with a reduced risk of cancer (1). High levels of oxidative stress may result in damage to tissue macromolecules (i.e., DNA, proteins, lipids) and the development of chronic illnesses such as cancer and cardiovascular disease (2–4). Natural antioxidants, especially the carotenoids, are present in fruits and vegetables and may mediate the protective anticancer effects suggested by the epidemiological and ecological studies. By far, β-carotene has been the most studied of all carotenoids (5). Despite overwhelming epidemiological information linking consumption of β-carotene with lung cancer protection, large intervention studies disproved a beneficial role for β-carotene in preventing human lung carcinogenesis, with some of the data suggesting a possible deleterious effect (6, 7). These unexpected results propelled the scientific inquiry toward the elucidation of the possible role of other natural antioxidants, beyond β-carotene.

Lycopene is the most prevalent carotenoid present in the human serum of Americans, accounting for roughly 50% of all plasma carotenoid content (8). Most of the dietary lycopene consumed by populations in the western world comes from tomatoes or their products (9, 10). Epidemiological studies looking at cancer protection on the basis of patterns of tomato consumption have yielded controversial results, with some but not all demonstrating a protective effect. Variability in lycopene bioavailability from different tomato products may explain these mixed results. This assertion is supported by the consistent protective effect of lycopene in epidemiological studies that have explored associations between serum lycopene concentrations and cancer risk as opposed to dietary patterns of tomato consumption (10). Published reports have confirmed that significant differences exist between different tomato products in terms of lycopene release and its gastrointestinal absorption (11–14). Of the commonly consumed tomato products, tomato paste compares favorably to fresh tomatoes and/or tomato juice (11, 15). Lycopene is a very hydrophobic molecule located within the tomato fruit matrix (16). Mechanical treatment (homogenization) and heating enhance the release of lycopene from the tomato matrix and may explain the improved bioavailability seen with consumption of processed tomato products (cooked tomatoes, tomato paste) (17). The presence of fat in the diet may also favorably affect the absorption of lycopene (18). These factors need to be taken into consideration when devising a dietary intervention for lycopene delivery or when comparing pharmacological data obtained from the use of different food-based interventions.

The putative anticancer effects of lycopene may be mediated by its extraordinary antioxidant properties (8). In fact, lycopene displays the highest singlet oxygen quenching ability among the most common carotenoids, thus providing the highest protection against oxidative damage (8, 19). Other plausible (non-antioxidant) mechanisms include the modulation of gap-junctional cell communication (19), inhibition of growth and induction of differentiation (19, 20), modulation of the insulin-like growth factor I (IGF-I)/IGFBP-3 system (21, 22) and modulation of enzymatic systems in charge of carcinogen metabolism (23, 24). Furthermore, lycopene displays strong inhibitory effects in experiments using a variety of human-derived cancer cells (21, 25). A number of short-term intervention studies using different formulations of lycopene have been reported in the literature (11, 26–30). However, with the exception of a handful bioavailability reports using short-term sampling strategies and a few short-term intervention studies with sparing sampling, limited information describing the human pharmacology of lycopene in detail have been published (31).

This report details the findings of a single-dose escalation study of lycopene delivered in a well-standardized food delivery system to healthy male volunteers. This Phase I clinical study was designed to provide detailed information (28-day sampling period) of the pharmacokinetic parameters of lycopene and its isomers after single-dose administration within the 10–120 mg dose range. We also describe the toxicity profile and report on the oral absorption of this formulation, through the evaluation of chylomicron-bound lycopene sampled at prespecified time points during the first 12 h post-administration.

This study was open to healthy male volunteers, ages 18–45, who at the time of enrollment were not using prescription medications, without history of alcohol use (72 h or more), and were not current smokers (should have quit at least 3 months before study entry). Because the most obvious future application of lycopene is for the chemoprevention of prostate carcinogenesis, we decided to focus our Phase I efforts on male subjects. Individuals were eligible if judged to be in good medical condition based on history and physical exam confirming the absence of chronic medical diseases or the use of regular prescription medications, lack of evidence for a psychiatric disorder, and performance status of 100% in the Karnofsky scale. Additional eligibility requirements included proof of normal organ function evidenced by liver and kidney function tests falling within the institutional normal values, as well as acceptable hematological function defined by WBC counts ≥4.0 K/μl, hemoglobin ≥13.5 g/dl, and platelet counts within normal institutional range.

Subjects had to be within 15% of ideal body weight based on standard weight tables and display pre-study lycopene concentrations <700 nm. Pilot single-dose studies conducted by our group in an unscreened population revealed that very little change in total blood lycopene concentrations were elicited in subjects with relatively high baseline plasma concentrations at low dosing levels (≤30 mg).⁸

Individuals unable to provide informed consent were ineligible. Additional exclusion criteria included history of gastrointestinal malabsorption (or other condition that could affect drug absorption), use of a prescription drug within the 14 days preceding study entry, and allergy to tomato-based products. Subjects were excluded if they were participating in another experimental trial where drug intake would have ended less than 4 weeks before study entry. Additionally, the presence of a health condition that in the judgment of the investigator could pose a threat to the subject's life and an abnormal electrocardiogram (EKG) made subjects ineligible.

Pre-screening telephone interviews were conducted where age, gender, smoking status, health status, alcohol use, and current medication use were ascertained. Individuals preliminarily meeting eligibility criteria were invited for a pre-study evaluation visit. Subjects were consented for participation, a blood sample was drawn to determine a baseline lycopene concentration, and a diet history was obtained. During this visit, all the evaluations necessary to determine eligibility were scheduled and research subjects reviewed and signed the informed consent. Eligible individuals were then registered onto the study.

At the time of registration, participants were assigned to a specific dose level of lycopene, which was designated in sequence depending on toxicity and last dose level filled. We proceeded with escalation to the next dose level when at most one in five patients developed grade 2 or higher toxicity. Accrual to the next dose level only began after all five patients had completed the preceding dose. The maximum tolerated dose (MTD) was predefined as the dose of lycopene below the level where at least two patients developed grade 2 toxicity or even a single patient developed grade 3 or 4 toxicity which was definitely drug related.

On the day of lycopene dosing, subjects were admitted to the General Clinical Research Center at the University of Illinois at Chicago (CRC), where a medical history, physical exam (PE) including vital signs and review of symptoms were obtained. Weight (Wt) and body surface area (BSA) were also measured. An angiocatheter was placed for IV access and blood drawing. Subjects were instructed to come to the CRC in a fasting state. After baseline blood studies were drawn, the appropriate dose of lycopene was administered p.o. with up to 250 ml of water over a maximum period of 15 min. Immediately after the appropriate dose of lycopene had been ingested, the subjects were asked to consume a negligible carotenoid-containing breakfast which provided 30% of the calories as fat and constituting 20% of the subjects energy needs for the day. Lunch was fed immediately after the 4-h draw and dinner after the 8-h draw. These meals contained negligible carotenoid content and 30% of calories as fat. Lunch provided 30% and dinner 40% of each subject's energy requirement. This energy distribution typifies the usual energy distribution in the American diet. The same meals were fed to the subjects on the second day of the study. When subjects were discharged from the CRC, they were seen by the nutritionist and instructed to follow a low carotenoid-eating plan until the completion of the study.

After lycopene administration, the subjects remained in the CRC for up to a maximum of 36 h and had hourly vital signs monitored for the initial 4 h. Thereafter, vital signs were taken once per day by the CRC nurses for the 2 days of the admission. A traditional intensive sampling strategy was used with 7 ml blood samples collected before [time zero (0)] and at 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 12, 24, 36, 48, 72, 96, 168, 240, 336, 408, 504, 576, and 672 h after lycopene administration. The intense sampling during the first 12 h assisted in the determination of the concentration-time profile of chylomicron lycopene for the assessment of lycopene absorption. The extensive blood sampling up to 672 h (28 days) was used to characterize the potential long elimination phase of lycopene. Blood collected at each sampling time point (0–672 h) was assayed for concentrations of total lycopene as well as its cis and trans isomers. In addition, samples collected during the initial 12 h were also assayed for chylomicron distribution of lycopene, as a measure of lycopene gastrointestinal absorption.

The basic formulation for the tomato juice consisted of 30 g of tomato paste incorporated into 5 ml of olive oil. This mixture was then mixed into a smooth paste and then made to 100 ml volume with purified water and blended again to give a smooth consistent juice mixture. The addition of olive oil had the purpose of improving lycopene bioavailability and of making the formula more palatable. The formula was homogenized in a blender to provide a uniform texture and to help improve lycopene bioavailability by releasing it from its matrix. Different batches of tomato paste may vary in their lycopene content. For this reason, every batch was analyzed for lycopene content and standardized to the all-trans isomer. The distribution of lycopene isomers in the formulation was 88% all-trans isomer, 8% cis isomer-1, and 4% for cis isomers-2, -3, and -4. The total volume of the formulation administered to a given research subject (within a specific dosing level) was adjusted to reflect this variability in lycopene content from batch to batch of tomato paste, so that the intended total dose of lycopene was delivered. The following dosing levels were evaluated: 10, 30, 60, 90, and 120 mg. These dosing levels corresponded to the following volumes of administration: 79 ml (10 mg), 238 ml (30 mg), 476 ml (60 mg), 769 ml (90mg), and 797 ml (120 mg). Research subjects were assigned to a treatment level, in the order in which they were identified during the enrollment period. Once a treatment level was completed, a safety assessment was undertaken and if no dose-limiting toxicity had occurred, we would open the immediate higher dosing level for accrual.

Toxicity evaluations were conducted in person at 12 h, 24 h, 48 h, 1 week, 2 weeks, 3 weeks, and 4 weeks after lycopene administration. Complete blood cell counts including platelets were obtained pre-study, 1 weeks, and 4 weeks after lycopene administration. Chemistry panel, cholesterol, triglycerides, and cholesterol fractions (high-density lipoprotein, low-density lipoprotein, and very low-density lipoprotein) were assessed during the pre-study visit and at 12 h, 1 week, and 4 weeks after lycopene administration. Adverse events were graded by a numerical score according to a defined Toxicity Grading Scale (NCI's Common Toxicity Criteria Version 2.0).

As lycopene is very lipophilic and is absorbed “packaged” into chylomicrons, the meals during the first 36 h of the study (CRC inpatient stay) were designed to be sufficiently fat-rich as to enhance lycopene absorption (see above). All meals were lycopene deplete. Subjects were instructed to avoid lycopene-containing foodstuffs throughout the duration of the study (28 days). Dietary assessments were conducted at prespecified time points throughout the study to document compliance with dietary restrictions.

During the consumption of the study formulation, participants were carefully observed so that the intended volume of the food-based delivery system was fully consumed. The following time schedule was designed to define compliance with blood draws. In this system, some leeway was allowed in obtaining blood concentrations as follows: (a) During the CRC stay (first 36 h): 15-min leeway for samples obtained from 0 to 12 h and 30-min leeway for samples obtained from 24 to 36 h. (b) During follow-up visits (48–672 h): 12-h leeway for samples from 48 to 96 h and 24-h leeway for samples from 168 to 672 h. To consider a participant evaluable for analysis, the participant should have ingested all study drug over the indicated maximum period of time (15 min) and should have at least complied with 18 of the 23 (circa 75%) scheduled blood draws for determining lycopene concentrations.

Analysis of fat-soluble vitamins and carotenoids in serum was performed following methodology described previously (32). Briefly, 200 μl of serum are mixed with 200 μl ethanol containing retinyl acetate as an internal standard, and extracted twice with 2 ml hexane (containing 0.01% BHT to prevent oxidation). The combined hexane layers are evaporated to dryness under reduced pressure (Speed-Vac centrifuge) and the residue reconstituted to the original serum volume of 200 μl with 50 μl stabilized ethyl ether and 150 μl mobile phase (methanol:acetonitrile:stabilized tetrahydrofuran, 50:45:5, v/v/v) for subsequent high-performance liquid chromatography (HPLC) isocratic separation on Novapac C18 reverse-phase column. The peaks were detected by Waters 490 Programmable Multiwavelength Detector with four channels, each analyte at its specific maximum absorbance (325 nm for vitamin A compounds, 295 nm for vitamin E compounds, 450 nm for carotenoids other than lycopene, 472 nm for lycopene). Lycopene isomers were separated using a Suplex pKb-100 C18 column with methanol:acetonitrile:isopropanol (54:44:2, v/v/v) isocratic elution according to previously published methodology (33). Extraction and analysis of fat-soluble vitamins and carotenoids in foods was undertaken according to methods previously described by our laboratory (34). The limit of detection of lycopene for this method is 0.5 μg/dl (or 10 nm). The variability of the assay is 2.7% (within-day) and 7.4% (between-day).

Chylomicrons were separated according to a method described by Borel et al. (35). Briefly, 2 ml of postprandial serum sample are carefully layered under 4 ml of 0.9% NaCl solution and centrifuged in a high-performance centrifuge (Avanti J series) at 90,000 × g for 1 h in a JA 30.5 fixed angle rotor. This method for chylomicron separation has been adapted from the original method described by Dole and Hamlin (36) and Grundy and Mok (37). After isolation, analysis for total lycopene and isomers proceeded according to analytical methodology described above.

Pharmacokinetic analysis was performed on plasma concentration-time data of 25 healthy male subjects. The plasma-concentration time data included measurements of total, trans-, and cis-lycopene concentrations from hours zero (0) until 672 h following lycopene administration. In addition, plasma concentration-time data included measurements of chylomicron concentrations of total, trans-, and cis-lycopene from hours zero (0) until 12 h following lycopene administration. The concentration-time data from each matrix was corrected by subtracting the plasma lycopene concentration at time zero (baseline) before the pharmacokinetic analysis of the data was performed. Only concentrations with a positive value were used in pharmacokinetic analysis.

Noncompartmental pharmacokinetic parameters were estimated using the microcomputer program, WinNONLIN (version 1.1, Scientific Consulting, Inc., Apex, NC). Peak plasma concentration (C_max) and the time of C_max (t_max) were determined directly from the individual observed lycopene concentration-time data. Area under the concentration-time curve from time zero (0) to 96 h (AUC_0–96) or to the last measured time point (AUC_{0–last Cp}) were determined by the log-linear trapezoidal rule. The area term was extrapolated to infinity (AUC_0–∞) using the elimination rate constant (k_el). Elimination rate constant (k_el) was obtained by nonlinear iterative least squares regression of the terminal log-linear portion of the concentration-time curve. The elimination half-life (t_1/2β) was calculated by dividing k_el into the natural logarithm of 2. The apparent clearance (CL/F) and volume of distribution (V_β/F) were calculated from the following equations:

where F is the fraction of bioavailability and assumed to be 1.

Noncompartmental pharmacokinetic parameters were also estimated using WinNONLIN. Peak plasma concentration (C_max) and the time of C_max (t_max) were determined directly from the individual observed trans- and cis-lycopene concentration-time data. Area under the concentration-time curve from time zero (0) to 96 h (AUC_0–96) or to the last measured time point (AUC_{0–last Cp}) was determined by the log-linear trapezoidal rule.

Noncompartmental pharmacokinetic parameters were estimated using WinNONLIN. Peak plasma concentration (C_max) and the time of C_max (t_max) were determined directly from the individual observed total, trans-, and cis-lycopene concentration-time data. Area under the concentration-time curve from time zero (0) to 12 h (AUC_0–12) was determined by the log-linear trapezoidal rule.

Overall differences in subject characteristics and pharmacokinetic parameters of systemic exposure (e.g., C_max and AUC) across the five different dosing regimens were evaluated by Kruskal-Wallis ANOVA. When significant differences were found, individual differences between different dosing regimens were evaluated by a Bonferroni's t test. Significance was determined at the P < 0.05 level.

Ninety-three individuals were screened by telephone. Thirty-four of these individuals underwent a pre-study evaluation and 27 were ultimately enrolled. Twenty-five subjects were evaluable due to two participants dropping out before study initiation. Minor exceptions were made by the principal investigator, which permitted the enrollment of two subjects. One subject had a BUN value (21 mg/dl) that was minimally higher than the required upper limit of normal (20 mg/dl). Another subject was also allowed to participate with a WBC count (4.7 K/μl) just under the required normal institutional value (4.8 K/μl). No other exceptions were made. One eligibility violation was committed when an individual with a baseline hemoglobin of 13.1 g/dl (13.5 g/dl required) was entered into the study. His hemoglobin concentration had risen to 13.5 g/dl 7 days into the study, which is within eligibility range. The eligibility criteria were revised twice as follows: (a) WBC count and hemoglobin concentration that allowed participation were changed from the “upper limit of normality for the institution” to ≥4.0 K/μl and ≥13.5 g/dl, respectively, and (b) the pre-study lycopene concentration that would allow participation was increased to 700 nm (from 600 nm). These changes in eligibility were implemented to facilitate and expedite accrual into the study. Table 1 summarizes the baseline characteristics for each group. No major differences were observed between the groups.

All 25 participants received the intended treatment according to the assigned dose concentration. Five subjects were assigned to each dosing level (10, 30, 60, 90 and 120 mg), and received a single dose of lycopene through a well-characterized food-based formulation. All treated subjects were included in the safety analysis. The two subjects that were enrolled in the study but withdrew were not included in the analysis as neither of them consumed lycopene. Overall the formulation was well tolerated after single oral administration. No significant toxicity (>grade 2) was observed that could be linked to lycopene ingestion. The most common adverse events that were observed included headaches (7 of 25), nausea (6 of 25), and diarrhea (4 of 25). Of these, only two episodes of nausea and two of diarrhea were felt to be related or possibly related to the study drug, and were of low grade (grade 1). Three days after ingesting lycopene, one of the participants sustained a skin laceration with a metal object in the medial aspect of his left leg, which was subsequently infected and required hospitalization for antibiotics. During the course of antibiotic treatment, he developed a skin rash and was deemed to have an allergic reaction to the antibiotics. The participant fully recovered from this unrelated event and completed all study requirements and evaluations without toxicity attributable to lycopene. No changes were seen on physical exam from baseline. Five subjects displayed “grade 1” leukopenia (defined as WBC counts between 3.0 and 3.9). It is important to mention that in most of these subjects, their baseline WBC count was close to 4.0. These abnormalities were for the most part seen at the 28-day blood draw, and could not be linked with absolute certainty to lycopene ingestion. No significant changes in hemoglobin were seen. Only one subject in the study (90 mg dose level) experienced a decrease in platelet values. This subject had a baseline platelet count of 179,000, his day 7 platelet count was 176,000, and on day 28, he displayed a minimally decreased platelet count of 143,000. The relatedness of this event to lycopene intake is unclear. Minimal changes in blood chemistry parameters were seen which were felt not related to lycopene. No significant hepatic or renal toxicity attributable to lycopene was seen. No relationships were found between lycopene concentrations and toxicities that were observed.

The mean ± SD of the plasma concentrations of total, trans-, and cis- lycopene for the five dosing levels are shown in Fig. 1. These plates illustrate the plasma concentrations between hours 0 and 672 h (inserted figure) as well as the change in baseline concentrations between hours 0 and 96 h. Table 2 summarizes the pharmacokinetic parameters for total lycopene at each dose level. After administration of single oral doses of lycopene, the mean time (t_max) to reach maximum total lycopene concentration (C_max) ranged from 15.6 to 32.6 h. Mean maximal concentration (C_max) for total lycopene ranged between 4.03 and 11.27 μg/dl (0.075–0.120 μm). Mean AUC_0–96 for total lycopene ranged from 214 to 655 μg h/dl (3.986–12.201 μmol h/l). The mean half-life for total lycopene ranged between 28.1 and 61.6 h. The ranges for the mean values of apparent CL/F and V_β/F included 98.6–286.4 ml/min and 2.12–18.54 l/kg, respectively.

Table 3 summarizes the pharmacokinetic parameters for lycopene isomers for each dose level. The ranges of mean t_max for trans and cis isomers of lycopene were 17.8–34.0 h and 22.5–56.0 h, respectively. Mean values for C_max and AUC_0–96 of the trans isomer tended to increase with ascending oral doses. The mean values of C_max for the cis isomer were of similar magnitude for dose levels between 30 and 120 mg. Similar to what was seen with total lycopene, the mean AUC_0–96 and C_max for the cis isomers were similar at dose levels of 30 and 60 mg, and at 90 and 120 mg. Isomer distribution of trans- and cis-lycopene (expressed by the trans/cis lycopene AUC_0–96 ratio) ranged from 1.47 to 1.76, and no major differences were observed between the ascending oral doses (data not shown). The trans isomer accounted for 54–65% of the mean AUC_0–96 of total lycopene, whereas cis isomer ranged from 36% to 41%.

Figure 2 displays a series of scatter plots of C_max achieved by total lycopene and isomers versus dose levels. This figure suggests that most of the gain in C_max after single-dose administration is realized in the lower range of the dosing interval (between 10 and 30 mg). Figure 3 is a series of scatter plots displaying AUCs for total lycopene and isomers versus dosing level. Similar to what was seen for C_max, the maximum increase in AUC again occurred between 10 and 30 mg. Figure 2 and 3 also reveal the high variability seen for these parameters among individuals within same dosing groups.

The noncompartmental pharmacokinetic parameters estimated from the chylomicron lycopene plasma concentrations are summarized in Table 4. The observed differences in C_max and AUC_0–12 were not significant among the five dosing groups for chylomicron-bound lycopene.

Lycopene is a carotenoid and a very potent natural antioxidant. In the western diet, it is almost exclusively consumed through the ingestion of fresh tomatoes and tomato-based products (10, 38). Epidemiological and experimental studies have suggested strong associations between lycopene and protection against a variety of epithelial cancers supporting its clinical development as a chemopreventive agent (10).

Several pharmacokinetic studies have evaluated the short-term administration of a number of different lycopene formulations. The majority of these studies have only included sparse blood sampling and offer limited pharmacological detail (11). Intensive blood sampling schemes have been applied in bioavailability studies where determinations of chylomicron-bound lycopene have been measured. However, the focus of these studies was placed on the first 8–12 h after lycopene administration and dose-response relationships were not explored (15, 39). Our study was designed to provide detailed pharmacokinetic information by using a 28-day sampling period and a broad dosing range (12-fold) of single doses of lycopene (10–120 mg) delivered in a well-standardized food-delivery system. The lowest dose of administration used in our study was based on concentrations of ingestion that in epidemiological studies have been associated with cancer protection (10). The highest dose level was defined in response to a report suggesting that substantial side effects (“lycopenemia”) may occur with protracted ingestion of very large amounts of tomato products (40). In addition, the administration of higher doses (e.g., larger than 120 mg) of our tomato paste-oil formulation would have not been feasible given the large volumes of administration that would be required.

In terms of toxicity, only two episodes of nausea and two of diarrhea were possibly related to lycopene intake and were all of mild intensity (grade 1). There was no consistent association between dose size and lycopene-related adverse events. Overall, the administration of single doses of lycopene using our formulation was feasible and safe. These results are not unexpected given the wide consumption of tomatoes in the American diet and are consistent with safety evaluations from previously published reports (41).

As lycopene is naturally present in human plasma, the pharmacokinetics of lycopene in this study were described in reference to baseline plasma concentrations. Pharmacokinetic analysis was therefore performed on the difference between the measured lycopene concentrations at sampled time points minus lycopene concentrations at baseline. The noncompartmental pharmacokinetic parameters for total lycopene by dosing group are provided in Table 2. As dietary intake of tomatoes was restricted throughout the duration of the study, in many subjects, lycopene concentrations fell below baseline value after 96 h and resulted in negative values. This is consistent with results reported by Porrini et al. (31) where a single dose of lycopene (16.5 mg) resulted in measurable increases in lycopene concentrations with a decline back to baseline at 104 h. For this reason, we have based our lycopene-exposure commentary on areas under the concentration curves generated between zero and 96 h (AUC_{0–96 h}).

Escalation of lycopene doses resulted in non-proportional increases in pharmacokinetic parameters of systemic drug exposure [e.g., C_max and AUC_{0–96 h} (Figs. 2 and 3)]. These data suggest a sigmoid (versus linear) distribution of systemic exposure parameters with increasing dosing levels of lycopene. The largest increases in systemic exposure parameters occurred during dose escalation at the lower dosing range (10–30 mg). Although the differences between systemic exposure parameters for the 10 and 30 mg groups did not reach statistical significance, the observed pattern may reflect saturability during absorption. To better characterize the pharmacokinetics of fresh lycopene absorption, we measured and analyzed the changes in chylomicron-bound lycopene concentrations during the initial 12 h after dosing. The differences observed among the five dose regimens for chylomicron-bound total, trans-, or cis- lycopene were not statistically significant. These findings may again be consistent with an absorptive process which may be saturated at very low dosing levels (<30 mg). Recent information provides indirect evidence for the presence of transporters that may mediate carotenoid transfer in mammalian tissues (42, 43). Although intestinal binding proteins have not been identified in humans or other mammals, their existence has been postulated and may be involved in facilitated and saturable absorption of carotenoids, including lycopene (44). Alternatively, the changes observed in C_max and AUCs for both total and chylomicron lycopene may reflect the high interindividual variability in plasma-lycopene responses observed and the small sample size used in our study (N = 5 per group). Other reports have described high interindividual variability in plasma responses to carotenoids including lycopene (39, 45, 46). An overall concentration ceiling is less likely as was suggested by the fact that the highest measured concentrations were observed in the highest dosing groups (90 and 120 mg). However, with only five subjects per group, a ceiling concentration cannot be properly identified with such wide variation in observed concentrations.

Given the high liposolubility of lycopene, its ability to distribute extensively in peripheral tissues does not come as a surprise and explains the large total and weight-normalized volumes of distribution that were observed (e.g., mean V_β/F ranging from 160–1320 l and 2.12–18.54 l/kg).

Time to C_max (t_max) for chylomicron-lycopene ranged between 4.2 and 8.4 h, whereas t_max for total lycopene occurred roughly after 24 h post-administration (15.6–32.6 h). This temporal pattern is consistent with that described for other carotenoids and supports the current physiological notion of lycopene absorption. Specifically, lycopene is taken up by the enterocyte, packaged into chylomicrons, and then released into the portal circulation, a process that take just a few hours (47–49). Chylomicrons are then taken up by the liver and lycopene is repackaged and released in other kinds of lipoproteins but predominantly bound to LDL, resulting in a several-hour delay for total lycopene concentrations to achieve C_max (48, 50, 51). The largest proportion of total measured lycopene in human plasma is carried in LDL (51, 52).

In our review of the literature, we were only able to identify one pharmacokinetic study of lycopene that included an intensive sampling scheme similar to the one that we used (31). Porrini et al. (31) administered a single dose of lycopene to healthy subjects and blood samples were collected for 104 h after lycopene administration. However, only one dose level (16.5 mg) was studied. In this study, peak concentrations (C_max) for total lycopene and isomers were observed earlier than in our study (6–8 h) and may reflect different experimental conditions including diet and formulation (31). Although t_1/2 was not reported in that study after single-dose administration, lycopene concentrations returned to baseline between 80 and 104 h, a similar time course to the one observed in our study (96 h). Systemic exposure parameters could not be compared as C_max were not reported and AUCs were calculated without subtracting the baseline lycopene concentration (31).

Contrary to what had been suggested by others, we were able to elicit changes in total lycopene concentrations after single-dose administration even at the lowest dosing level (10 mg). In part, this may have been possible as a consequence of our eligibility criteria which screened out individuals with high basal lycopene concentrations (>700 nm), thus providing ample opportunity to identify a concentration change if it occurred. In addition, our prolonged sampling period may have allowed us to detect a change in total lycopene which may be missed in studies using shorter sampling periods.

The elimination of lycopene has been most extensively evaluated in carotenoid-withdrawal studies where participants were asked to refrain from ingesting carotenoid-rich foodstuffs. Rock et al. (46) estimated the elimination of a common group of carotenoids in human subjects. Lycopene displayed the shortest half-life of elimination after carotenoid withdrawal (12–33 days) compared to other common carotenoids. Most of the decline occurred during the initial 2 weeks. These results were used to plan our sampling strategy which extended beyond 2 weeks (total of 4 weeks). Clearly, the results obtained through carotenoid withdrawal studies are not directly comparable to the pharmacokinetic parameters generated after single-dose administration. In our study, we observed an elimination half-life for total lycopene that ranged between 28.1 and 61.6 h and was shorter than we originally anticipated. However, these results are consistent with those reported by Porrini et al. (31) where lycopene plasma concentrations returned to baseline concentrations at 104 h.

Lycopene is predominantly present in tomatoes and tomato-derived products (such as ours) in its trans-isomeric form. At baseline conditions and at steady state, lycopene is predominantly present in plasma in its cis configuration. This observation has led to the suggestion that lycopene's cis isomers may be preferentially absorbed from the GI tract. In addition, extensive isomerization may occur during and/or after absorption. Research seems to indicate that trans-lycopene has a higher propensity to precipitate and form crystals affecting its solubility, a fact that may possibly decrease its GI absorption relative to the more soluble cis isomers (53). In our study, we measured both cis and trans isomers of total as well as chylomicron-contained lycopene. Although total plasma concentrations of the cis isomers were higher throughout the duration of the study, the changes in concentration seen relative to baseline values were higher for the trans isomer. This suggests that the trans isomer is absorbed from the GI lumen and reflects the predominance of this isomeric form in our formulation (≈88%). Interestingly, even at very early sampling times, the proportions of the different isomers in plasma and in chylomicrons did not closely resemble the actual proportions in the formulation. A higher proportion of cis isomers that would have been expected relative to their concentration in the formulation was observed. This may be a consequence of preferential absorption of cis isomers as previously suggested by other authors and/or in vivo isomerization during or immediately following absorption. Total lycopene isomeric proportions progressively returned to ratios that were closer to those observed at baseline, suggesting continuing lycopene isomerization.

Our study constitutes one of the first detailed pharmacokinetic evaluations of lycopene using a well-characterized and standardized oral food-delivery system. Our findings demonstrate that even small doses of this compound (10 mg) may induce measurable changes in plasma concentrations. Although there seems to be substantial interindividual variability in plasma concentrations after lycopene administration, progressive escalation of lycopene doses results in non-proportional increases in systemic parameters of exposure (C_max and AUC) with the largest changes seen with dose escalation at the low end of the dosing range (10–30 mg). Therefore, we anticipate that modest doses of lycopene may result during chronic administration, in substantial increases in lycopene blood concentrations, with very limited additional gain achievable with dose escalation beyond intermediate dosing levels (>30–60 mg). The shorter elimination half-life of lycopene relative to other carotenoids supports its administration on a daily basis. These assumptions require confirmation through the implementation of well-designed multiple-dose studies. Furthermore, the recommendation of a chronic dose for use in Phase II studies should be based on the desired target concentration at steady state and the projected accumulation of lycopene in the target organ. Multiple-dose studies are required to provide firm recommendations in terms of Phase II dosing. These studies would not only ascertain systemic concentrations of lycopene but would also measure the concentration in the target organ as well as the induction of modulation of relevant biomarkers. Given the large variability observed around any dose level, it is unlikely that one dose will fit all subjects in terms of inducing similar systemic exposures. It is possible then that a dose-adjusted method may be required. What seems clear at this point is that further pharmacokinetic studies are needed which should ideally include larger numbers of subjects and that would also explore the changes in systemic exposure of lycopene induced by doses that fall within the gaps that were not evaluated in this study (e.g., doses between 10 and 30 mg or between 60 and 90 mg).

Our data suggests that there seem to be three dosing regions on which differences of exposure are observed: (a) less than 30 mg; (b) between 30 and 60 mg; and (c) greater than 60 mg. On the basis of this, our group is currently conducting a 3-month study of oral lycopene administration using a similar tomato paste-oil formulation in individuals at high-risk to develop prostate cancer (e.g., increased PSA concentration) but absence of invasive malignancies in prostate biopsies. This study will explore the multiple-dose pharmacokinetics and toxicity of lycopene over a dose range of 15–78 mg/day. In addition, this study will explore the concentrations of lycopene that are achievable in prostate and oral mucosal tissue at steady state as well as the modulation of oxidative markers in the blood, mucosa, and prostate.