Introduction

Top Abstract Introduction Experimental Procedures Results Discussion Appendix References

Cimetidine was the first H₂-receptor antagonist marketed to control gastric acid secretion (Somogyi and Gugler, 1983). It is currently sold as an over-the-counter product but continues to be remarkable in the number of literature reports documenting its role in drug interactions (Shinn, 1992; Guengerich, 1997). Cimetidine is well absorbed after oral administration but its absorption is highly variable and the appearance of a second plasma concentration maximum provides an unusual pharmacokinetic profile characteristic (Bodemar et al., 1979). Enterohepatic recirculation had been initially proposed to be responsible for the double peak (Veng Pedersen, 1981), but biliary elimination has been shown to not significantly impact drug plasma levels (Kaneniwa et al., 1986a). Gastric pH and emptying variability have been reported to influence the absorption pattern of cimetidine as a function of dose and time of administration (Oberle and Amidon, 1987). However, plasma level double peaks are also obtained in human intestinal perfusion of cimetidine (Voinchet et al., 1981). Cimetidine absorption had been shown to be lower in rat jejunum than in duodenum and ileum (Barber et al., 1979) and this absorption site dependence has been projected to account for plasma level double peaks in humans (Witcher and Boudinot, 1996). However, membrane transport contributions to this regional dependence and the mechanistic elements of cimetidine plasma level double peaking and absorption variability have not been resolved.

Most recently, cimetidine absorption in rat proximal jejunum has been shown to be exclusively via the paracellular pathway. This was based on data that indicated that intracellular cimetidine did not appreciably exit the basolateral membrane. Furthermore, intracellular cimetidine levels, as controlled by cellular uptake and elimination, appear to regulate paracellular absorption (Zhou et al., 1999). In this report, the role of jejunal S-oxidation in influencing cimetidine absorption was further explored by varying concentration and using inhibitors of cimetidine metabolism and transport in single-pass small intestinal perfusions in rats.

The current data suggest that cimetidine S-oxidation influences paracellular cimetidine absorption in rat jejunum by determining the magnitude of intracellular drug concentration. As intracellular uptake increases with lumenal cimetidine concentrations beyond the capacity of jejunal cimetidine metabolism, intracellular cimetidine levels rise, leading to the restriction of paracellular transport and a reduction of drug absorption into the systemic circulation. The reduction in paracellular transport provides a greater driving force for intracellular uptake. This, in turn, amplifies the contribution of intracellular cimetidine metabolism to jejunal elimination, resulting in elevated cimetidine S-oxide levels in the jejunal lumen. Cimetidine S-oxide is not absorbed and does not enter the rat jejunal lumen from the blood (Hui et al., 1994). Thus, intestinal elimination of cimetidine is enhanced by inhibition of paracellular transport. At high lumenal cimetidine concentrations, this form of self-inhibition depresses rat jejunal absorption. Based on cimetidine metabolite levels in human jejunum, it is proposed that this mechanism may account for double plasma level maxima. The concentration dependence and inhibition profile of this process may also contribute to the broad variability observed in cimetidine oral bioavailability.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion Appendix References

Chemicals. Cimetidine, D-glucose, 3-oxymethylglucose, D-mannitol, imipramine hydrochloride, chlorpromazine hydrochloride, erythromycin, neomycin sulfate, lincomycin, L-cysteine, L-methionine, 2(N-morpholino)ethanesulfonic acid (MES), HEPES, and polyethylene glycol 4000 (PEG 4000) were purchased from Sigma Chemical Co. (St. Louis, MO). [¹⁴C]PEG 4000 was obtained from New England Nuclear (Boston, MA). Cimetidine sulfoxide was kindly provided by SmithKline Beecham Pharmaceuticals (King of Prussia, PA). All other chemicals were of reagent grade or HPLC grade.

Rat In Situ Single-Pass Intestinal Perfusion. Male Sprague-Dawley rats (Charles River Breeding Laboratories, Wilmington, MA) weighing 250 to 300 g were fasted overnight for 16 to 22 h with free access to water and anesthetized with an intramuscular injection of 87 mg/kg ketamine and 13 mg/kg xylazine before surgery. After anesthesia, rats were placed on a heating pad under a surgical lamp to maintain body temperature. A midline longitudinal abdominal incision was made and an inlet Teflon tube (0.42 cm diameter) was inserted into the jejunum at about 10 cm distal to the ligament of Treitz. Outlet Teflon tubes were inserted 8 to 10 cm distal to the inlet cannulae. In ileal perfusions, the outlet tubing was inserted 2 cm proximal to the ileocecal junction and the inlet tube 8 to 10 cm proximal to the outlet. After cannulation, the abdomen was rinsed with isotonic saline, and covered with an isotonic saline-wetted gauze. The entire surgical area was then covered with parafilm to reduce evaporation. All inlet tubing was water-jacketed to maintain inlet perfusion solutions at 37°C. Inlet cannulae were connected to syringes that were placed in a perfusion pump (Harvard Apparatus Co., South Natick, MA). Drug solutions were perfused through the intestine at flow rates of 0.12 or 0.25 ml/min. After allowing 40 min to reach steady-state outlet concentrations, outlet perfusate samples were collected every 10 or 15 min for 90 min. When the experiment was completed, the perfused intestinal length was measured by dissecting the mesentery and blood vessels and placing the segment flat over a ruler whose surface was wetted with saline to prevent segment elongation.

Perfusion Solutions. Cimetidine solutions were made isotonic (290 ± 20 mOsm/kg) and buffered to the desired pH with MES or HEPES buffer. These perfusion solutions contained 5 mM potassium chloride, in the range of 135 mM sodium chloride for isotonicity adjustment, 10 mM MES and were adjusted to pH 5.5 or 6.5 with 10 N sodium hydroxide. In some studies, 10 mM HEPES was used in place of MES to adjust perfusion solution pH to 7.5. A nonabsorbable marker (0.01% PEG 4000 traced with radioactivity counterpart, specific activity 0.8 mCi/g) was included in the buffer to monitor water transport across the intestinal segments. Perfusion outlet drug and metabolite concentrations were corrected for water absorption or secretion as outlined in the data analysis section. All inlet drug solutions were analyzed to ensure cimetidine stability before initiating rat perfusion and then used to obtain drug water-corrected outlet-to-inlet cimetidine concentration ratios. Cimetidine sulfoxide concentrations were quantified using standard curves generated with SmithKline Beecham cimetidine sulfoxide. Jejunal perfusions were also performed with 0.4 mM cimetidine sulfoxide in four rats at pH 6.5 and a perfusion flow rate of 0.12 ml/min.

Analytical Methods. Cimetidine and cimetidine sulfoxide were analyzed by an HPLC method (Larsson et al., 1982). Samples were injected onto an HPLC system composed of a WISP autosampler model 710B, Waters 501 HPLC solvent pump (Waters Corporation, Milford, MA), and a nucleosil SA 10-µm cation exchange column, 250 × 4.6 mm (Alltech Associates, Inc., Deerfield, IL). The mobile phase, containing 20% v/v acetonitrile and 80% v/v buffer (30 mM monobasic potassium phosphate, 3.7 mM dibasic sodium phosphate, and 10 g/liter potassium chloride, adjusted to pH 4.0 with 85% phosphoric acid) was pumped at a flow rate of 1.2 ml/min. The eluting peaks were monitored at 228 nm using a variable wavelength UV detector (Spectroflow 773; Kratos Analytical Instruments, Ramsey, NJ) and the peak height together with peak area were measured with a Shimadzu integrator (Model CR 501 Chromatopac; Shimadzu Corporation, Kyoto, Japan). Retention times for cimetidine and cimetidine sulfoxide were 7.5 and 12.8 min, respectively. The minimal levels of detection for cimetidine sulfoxide and cimetidine were 50 and 80 nM, respectively. A linear response was obtained for cimetidine and cimetidine sulfoxide standard curves (r = 0.997-1.000). Quantification of metabolite was obtained as the mean value of four to five steady-state outlet perfusate samples as determined from the standard curve. The metabolite HPLC peak from perfusions was confirmed as cimetidine sulfoxide by mass spectrometry analysis in a previous study (Hui et al., 1994). The original solutions and the effluent perfusate samples were also analyzed for [¹⁴C]PEG 4000 using liquid scintillation counting. A 500-ml aliquot was mixed with 5 ml of scintillation fluid (Ecolite; ICN Pharmaceuticals, Costa Mesa, CA) and the amount of radioactivity was determined in a Beckman LS 6000 SC scintillation counter (Beckman Instruments, San Jose, CA).

Pretreatment of Rats with Antibiotics. In some selected studies, the animals were given a 7-day regimen of a neomycin and lincomycin combination (Illing, 1981; Ilett et al., 1990). Neomycin sulfate (1.5% w/w) and 0.3% w/w lincomycin were added to their food, and both drugs were also dissolved in their drinking water at concentrations of 2 and 0.5 mg/ml, respectively.

Data Analysis. Data were generated from each rat by averaging at least five perfusate collection samples at steady state. The amount of cimetidine sulfoxide detected in the perfusate is calculated as the fraction of drug metabolized (F_met).

Luminal cimetidine sulfoxide and outlet cimetidine concentrations were corrected for water transport across intestinal membrane. Intestinal water flux (J in microliters per minute per centimeter) is calculated at a given flow rate, Q (microliters per minute), per centimeter of perfused intestinal length (L) as ¹⁴C disintegration per minute using the following equation (Lu et al., 1998):
Positive values of J imply net water absorption whereas negative values indicate net water secretion. Effective cimetidine permeability (P_eff), using a modified boundary model developed by Johnson and Amidon (1988), is computed according to the following equation:
where C_outlet is the exit cimetidine concentration corrected for water transport in the perfusate and C_inlet is the input drug concentration. A water transport correction is obtained by multiplying the drug outlet-to-inlet concentration ratio by the marker inlet-to-outlet dpm ratio. R and L are intestinal radius and length of the perfused segments, respectively. The R value used in this study was estimated to be 0.23 cm (Yuasa et al., 1988). The P_eff is a measure of drug lost across the apical membrane of the intestinal epithelia including both intracellular and paracellular drug transport. Permeability defines the rate of absorption in this rat perfusion system and has been correlated with the fraction of drug absorbed from solution in human clinical studies as assessed by plasma level pharmacokinetic profiles (Amidon et al., 1988). A corrected value for the effective permeability (P_eff(cor)) was calculated using a molar summation of luminal cimetidine and cimetidine sulfoxide as C_outlet in the above equation. This corrected value is reflective of net systemic drug absorption in the presence of intestinal drug elimination. Data are presented as mean ± S.E. Statistical differences were determined by either Student's t test or ANOVA, as appropriate.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion Appendix References

Previous work had identified cimetidine S-oxide in the lumen after jejunal perfusions of cimetidine in rats and two human subjects (Hui et al., 1994). In the earlier studies in rats, significantly lower levels of the S-oxide had been observed in ileal compared with jejunal cimetidine perfusions (Hui et al., 1994). Furthermore, a maximum ratio of metabolite to initial perfusion drug concentration had been observed in the jejunum above 0.4 mM for perfusion cimetidine concentrations from 0.1 to 4 mM. Therefore, 0.4-mM drug concentrations were used in pH-dependence and metabolism-inhibition studies. Rat jejunal perfusions of 0.4 mM cimetidine S-oxide at pH 6.5 and perfusion flow rate of 0.12 ml/min resulted in water transport-corrected permeabilities of zero, confirming that the metabolite is not appreciably absorbed. Furthermore, reduction of the metabolite to the parent drug was not detectable in the outlet perfusate.

Recent studies in jejunal epithelia had determined that the cellular origin of metabolite production was the microsomal cell fraction (Lu et al., 1998). Although drug oxidation is rarely mediated by intestinal bacteria (George and Renwick, 1987), this was confirmed by comparing jejunal appearance of metabolite in antibiotic-treated versus untreated rats. Antibiotic pretreatment did increase the variability in S-oxide appearance from 0.4 mM jejunal cimetidine perfusions but metabolite production was not significantly different between treated and untreated animals (F_met = 0.66 ± 0.14 antibiotic-pretreated; F_met = 0.73 ± 0.04 untreated; n = 4 rats).

It had been shown that cimetidine S-oxide appearance was substantial in cimetidine jejunal perfusions at pH 6.5 but negligible in ileal perfusions at pH 7.5 (Hui et al., 1994). Jejunal S-oxide levels as a function of perfusion pH showed a maximum at pH 6.5, which coincided with a maximum in cimetidine jejunal permeability (Table 1) from drug perfusions at 0.4 mM. Although ileal metabolite levels were higher from perfusions at pH 6.5 versus pH 7.5, they are significantly lower than those from jejunal perfusions at pH 6.5 (Table 1). At a physiologically relevant pH of 7.5 in the ileum, ileal S-oxide appearance is negligible and ileal cimetidine permeability is equivalent to metabolite-corrected jejunal permeability from 0.4 mM cimetidine perfusions at pH 6.5 (Table 1).

Because little metabolite is produced in ileal perfusion, effective permeability calculations are not affected by metabolite production, and the amount of drug loss from the lumen reflects systemic drug absorption. The permeability of cimetidine in the ileum is constant at low concentrations (Fig. 1), indicating that the amount of drug loss is proportional to the amount available for absorption. As had been determined in a previous study, reduced permeability at higher perfusion concentrations is not the result of saturation of carrier-mediated transport at the apical membrane but rather a dose-dependent restriction of paracellular transport (Zhou et al., 1999). If jejunal permeability is calculated from drug loss in perfusate and not corrected for metabolite appearance, an unusual permeability profile is obtained (Fig. 2A). However, when lumenal drug concentration is corrected for lumenal metabolite appearance, the permeability profile and permeability magnitudes (Fig. 2B) in the jejunum are comparable to those observed in the ileum (Fig. 1). The magnitude of lumenal metabolite appearance is also observed to level off at high drug concentrations (Fig. 3). The ratio of jejunal metabolite appearance to drug loss (F_met) is proportional to perfusion concentration over the lower perfusion drug-concentration range and levels off at about 0.9 at higher perfusion concentrations (Fig. 4). Given the definition of F_met, when metabolism is saturated at higher drug concentrations, the increase in F_met indicates that cellular uptake rather than metabolism is limiting metabolite appearance. Otherwise, F_met would decrease at higher perfusion concentrations. Permeability was not influenced by perfusion flow rate at any of the concentrations studied (Figs. 1 and 2). However, at the highest drug perfusion concentration (4 mM), F_met was significantly lower at the slower perfusion rate than at the faster perfusion flow rate (Fig. 4).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effective ileal permeability (Peff) of cimetidine as a function of steady-state drug perfusion concentration at pH 7.5.  Negligible appearance of S-oxide metabolite in the ileum provides for straightforward permeability calculations as a function of lumenal drug loss. Peff at 2 and 4 mM are significantly lower than Peff at any of the lower concentrations (P < .05).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   A, effective jejunal permeability (Peff) of cimetidine as a function of steady-state perfusion concentration at pH 6.5 calculated as drug loss from the jejunal lumen; B, effective permeability corrected for drug lost to sulfoxide appearance in the lumen (Peff, corr).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Steady-state concentration of cimetidine S-oxide in the jejunal lumen, normalized for the length of the perfused jejunal segment, as a function of perfusion drug concentration. Metabolite concentrations are significantly different at all concentrations except for 2 versus 4 mM (P < .05). At 2 mM, metabolite concentration is significantly higher at the slower perfusion flow rate (P < .05). Open column, 0.123 ml/min; filled column, 0.247 ml/min.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Steady-state fraction of cimetidine metabolized (Fmet) as a function of drug perfusion concentration. Fraction metabolized is defined as the ratio of cimetidine S-oxide concentration in the jejunal lumen to the concentration of cimetidine lost from the jejunal lumen. Both concentrations are corrected for water transport and are normalized for the length of the perfused jejunal segment. Fmet are significantly different at all concentrations except for 2 versus 4 mM (P < .05). *, at 4 mM, Fmet is significantly lower at the slower perfusion flow rate (P < .05).

A 10-fold reduction in jejunal cimetidine permeability (uncorrected for S-oxide appearance) is observed when 0.4 mM cimetidine is coperfused with 5 mM methimazole (Fig. 5A). Reductions in permeability (3- to 4-fold) are generated by cimetidine coperfusion with L-cysteine, L-methionine, chlorpromazine, and imipramine. However, when steady-state cimetidine permeability is corrected for the appearance of metabolite, only methimazole and the amino acids actually reduce cimetidine absorption (Fig. 5B). Methimazole eliminates measurable appearance of cimetidine S-oxide in the jejunal lumen (Fig. 6) in parallel with the reduction in corrected permeability to near zero (Fig. 5B). Although both chlorpromazine and imipramine significantly reduce metabolite levels (Fig. 6), they do not affect the corrected cimetidine permeability (Fig. 5B). Decreases in the appearance of cimetidine S-oxide are statistically significant with coperfusion of L-methionine, whereas metabolite variability in the presence of erythromycin does not result in statistically significant changes in F_met (Fig. 6).

View larger version (30K): [in this window] [in a new window]   Fig. 5.   A, effective jejunal permeability of cimetidine from 0.4 mM steady-state drug coperfusion at pH 6.5 with 5-mM test inhibitors. Interpretation of permeability as a function of drug loss from the lumen is complicated by substantial lumenal appearance of S-oxide metabolite in the jejunal lumen. Peff is significantly lower than control with coperfusion of cysteine, methionine, chlorpromazine, imipramine, and methimazole (P < .05). B, effective jejunal permeability of cimetidine from 0.4 mM steady-state drug coperfusion at pH 6.5 with 5-mM test inhibitors in which lumenal drug loss is corrected for appearance of lumenal metabolite. Peff(cor) is significantly different from control with coperfusion of cysteine, methionine, and methimazole (P < .05).

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Steady-state fraction of cimetidine metabolized from 0.4-mM drug coperfusion at pH 6.5 with 5-mM test inhibitors. Fraction metabolized is defined as the ratio of cimetidine S-oxide concentration in the jejunal lumen to the concentration of cimetidine lost from the jejunal lumen. Both concentrations are corrected for water transport and are normalized for the length of the perfused jejunal segment. Fmet is significantly different from control with coperfusion of methionine, chlorpromazine, imipramine, and methimazole (P < .05). ND, not detected.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion Appendix References

Cimetidine plasma profiles from oral administration under fasted conditions often show a secondary maximum that is not observed when the drug is administered with a meal (Bodemar et al., 1979). The two plasma level maxima typically occur between 1 and 2 h and between 3 and 4 h after oral cimetidine administration. Given the relatively short elimination half-life of cimetidine and the fact that this double peak time frame coincides with intestinal transit time, variable absorption rate down the length of the intestine is anticipated to contribute to drug plasma level observations. Enterohepatic recirculation (Veng Pederson and Miller, 1980), intestinal bacterial reconversion of biliary metabolite (Gugler et al., 1981), variable gastric emptying (Oberle and Amidon, 1987), and region-dependent absorption (Hui et al., 1994) have all been proposed to account for these observations.

The extent of biliary secretion of cimetidine and metabolites has been shown to be insufficient to contribute to a second plasma level peak (Kaneniwa et al., 1986a). Although variable gastric emptying has been indicated to play a partial role (Oberle and Amidon, 1987), plasma level double peaks are also obtained from human intestinal perfusion of cimetidine (Voinchet et al., 1981). This study indicates that a decreased rate of absorption in the jejunum as compared with lower small intestine may contribute to plasma level double peaks and variable absorption. The first plasma level peak is generated by cimetidine absorption as the drug enters the proximal small intestine from the stomach. Then, after a slowing of the absorption rate in the jejunum coincident with systemic drug elimination, a secondary increased rate of drug absorption in the lower small intestine may generate the second plasma level peak.

Cellular uptake of cimetidine across the mucosal membrane of the small intestine includes a potential-dependent pathway and an organic cation-proton exchange mechanism (Piyapolrungroj et al., 1999). Mucosal transport by cimetidine-proton exchange favors drug secretion into the jejunal lumen as mucosal microclimate pH in the jejunum is lower than intracellular pH (Daniel et al., 1989). P-glycoprotein has also been reported to promote apical exit of cimetidine from epithelial cells (Dudley and Brown, 1996). Basolateral facilitative diffusion (Zhang et al., 1998) is proposed to favor drug transport into the cell with subsequent secretion into the intestinal lumen. Consistent with this secretory pathway, it has recently been demonstrated that intracellular cimetidine does not appreciably exit the cell across the basolateral membrane, indicating that basolateral facilitative diffusion may be predominantly unidirectional (Zhou et al., 1999). Because basolateral exit does not occur to any appreciable extent, cellular uptake and metabolism determine intracellular cimetidine levels and cimetidine intestinal absorption is predominantly paracellular.

Based on mass balance inferences, data from this study indicate that increased intracellular cimetidine concentrations reduce drug absorption by the paracellular pathway. The decrease in cimetidine permeability at concentrations above 0.4 mM in both ileum and jejunum is consistent with this mechanism of transport inhibition (Figs. 1 and 2). Whereas saturation of carrier-mediated transport more commonly accounts for permeability decreases with increasing concentration, coperfusion with methimazole supports the paracellular mechanism. Previous studies have shown that methimazole inhibits cimetidine S-oxidation by flavin monooxygenase in the small intestine (Lu et al., 1998) and does not influence cimetidine transport across rat jejunal brush-border membrane (Piyapolrungroj et al., 1999). In conjunction with jejunal perfusion data, these observations are consistent with cimetidine transport by a paracellular rather than transcellular pathway (Zhou et al., 1999). It is projected that the basolateral membrane limits transcellular cimetidine absorption and that intracellular cimetidine levels will rise with methimazole inhibition of intracellular cimetidine sulfoxidation (Fig. 6). The observed depression in cimetidine absorption rate (Fig. 5, A and B) could be projected from a reduction in drug concentration gradient driving force across cellular mucosal membranes resulting from an elevation of intracellular drug levels. High intracellular drug concentrations would be projected to promote passive permeation across the basolateral membrane. At high cimetidine perfusion concentrations, the fact that almost all drug returns to the jejunal lumen either as metabolite or unmetabolized cimetidine indicates that the basolateral membrane is the primary barrier for transcellular absorption. This is consistent with in vitro data in Caco-2 cells (Gan et al., 1998) and rat small intestinal tissue (Collett et al., 1999), showing that H₂-antagonists are preferentially transported in the basolateral to apical direction.

Cimetidine transport into jejunal epithelial cells across the lateral membrane lining the paracellular pathway may account for the surprising increase in mucosal permeability observed over the concentration range from 0.08 to 0.4 mM (Fig. 2A). To explain this concentration dependence, "systemic" permeability (corresponding to corrected-P_eff) is defined as the rate of absorption corrected for the appearance of lumenal metabolite. This determines the rate at which drug would be available to the systemic circulation from paracellular transport, because intracellular drug that is not metabolized is secreted back into the intestinal lumen. Systemic cimetidine absorption is unchanged over this concentration range, as indicated by the corrected permeability data (Fig. 2B). Defining "apical" permeability (corresponding to P_eff) as the rate of drug lost from the intestinal lumen refers to drug traversing the paracellular pathway and drug that enters epithelial cells beyond the apical membrane. The increase in apical permeability (uncorrected for metabolism) from 0.08 to 0.4 mM is the result of cellular cimetidine uptake across both apical and lateral membranes. Over this lower concentration range, intracellular drug concentrations may not be sufficiently high to saturate metabolism, and epithelial elimination increases the inward cimetidine flux across the apical and lateral membranes by maintaining sink conditions inside the cell. Essential to this interpretation, it has been shown that cimetidine sulfoxide does not compete with cimetidine for secretory transport across rat jejunal brush-border membrane vesicles (BBMV; Piyapolrungroj et al., 1999) and that the metabolite is not absorbed in jejunal perfusions. Thus, as cimetidine perfusion concentration increases, metabolism becomes saturated and intracellular cimetidine levels increase to affect a dose-dependent restriction of the paracellular transport.

Reduced paracellular absorption is coincident with increases in intracellular drug uptake and controls jejunal elimination at higher cimetidine concentrations. At low cimetidine perfusion concentrations (0.08-0.4 mM), both apical permeability (Fig. 2A) and metabolized fraction (Fig. 4) increase proportionally as intracellular cimetidine is eliminated by microsomal metabolism and metabolite export into the jejunal lumen. At high concentrations (0.4-4 mM), the capacity for metabolism and/or apical elimination is exceeded. Under these conditions, it is projected that intracellular drug concentrations increase to levels that can affect paracellular restriction of drug absorption. This is supported by the fact that both P_eff and corrected P_eff decrease as concentrations increase over the higher concentration range (Fig. 2) whereas F_met levels off between 0.8 and 0.9 (Fig. 4). The fact that the ratio of metabolite appearance to drug loss is close to 1 at high lumenal cimetidine concentrations suggests that most of the drug uptake beyond the apical membrane is not absorbed systemically but secreted back across the apical membrane as the S-oxide metabolite.

Permeability and F_met were not dependent on perfusion flow rate except at 4 mM cimetidine where a statistically lower F_met was recorded with slower versus faster perfusion flow rate (Fig. 4). Luminal sulfoxide appearance at 2 mM cimetidine was statistically greater at the slower versus the faster flow rate (Fig. 3). Because cimetidine paracellular transport is inhibited more extensively at higher lumenal cimetidine concentration, lumenal drug loss predominantly represents intracellular drug uptake with subsequent metabolite elimination into the lumen. F_met at higher cimetidine concentration is close to 1 and so a statistical difference in flow rate or concentration dependence would not be expected. The higher production of metabolite at the slower flow rate (Fig. 3) indicates that greater drug uptake rather than lower drug metabolism accounts for lower F_met values. A slower perfusion flow rate would be projected to provide a longer jejunal residence time and, therefore, a greater potential for drug uptake and lumenal metabolite appearance. A possible explanation is that a decrease in the thickness of an aqueous diffusion boundary layer at the higher flow rate reduces the aqueous resistance to cellular drug uptake. This would provide the strongest dependence of metabolite production on flow rate at a concentration approaching metabolic saturation.

Ileal cimetidine absorption demonstrates a similar self-inhibition profile (Fig. 1) in the absence of metabolite elimination into the lumen. Although the ileum might be expected to be more sensitive to paracellular restriction in the absence of cimetidine S-oxidation, the ileum is typically exposed to lower concentrations of drug because it is distal to upper intestinal absorption sites. The similarity in ileal and jejunal permeability concentration dependence (Figs. 1 and 2B) indicates that jejunal elimination and a restricted paracellular pathway promotes less cimetidine absorption in the jejunum as compared with the ileum based on exposure to higher lumenal drug concentration. Region-dependent absorption in rats has been reported, showing cimetidine plasma levels were higher from duodenal and ileal administration than from jejunal administration (Kaneniwa et al., 1986b). A similar finding has been documented in human subjects for ranitidine (Gramatte et al., 1994).

Regional differences in permeability and metabolite production are also observed with respect to perfusion pH. Whereas the uncorrected permeability reflects the sum of intracellular drug uptake and paracellular absorption into the systemic circulation, metabolite-corrected permeability defines the rate of systemic drug absorption. The fact that changes in perfusion pH significantly alter F_met and P_eff but do not alter corrected P_eff indicates that changes in pH primarily affect cimetidine intracellular uptake and subsequent lumenal secretion. This is consistent with the fact that intracellular pH and metabolism should not be influenced by changes in perfusion pH. Jejunal cimetidine P_eff was maximal at pH 6.5, whereas ileal P_eff was relatively insensitive to the lumenal pH change. Initial cimetidine uptake overshoot of equilibrium in rat jejunal BBMV was maximal at an intravesicular pH of 5.7 when extravesicular pH was 7.4 (Piyapolrungroj et al., 1999) indicative of mediated cimetidine secretion across the apical membrane. Cimetidine sulfoxide did not compete with cimetidine uptake in this vesicle preparation. The in vitro and in vivo data indicate that the relative contributions of cimetidine secretion and absorption to jejunal permeability are highly dependent on mucosal microclimate pH, which is lower in the jejunum than in the ileum (Daniel et al., 1989).

Imipramine strongly inhibited initial cimetidine uptake by rat jejunal BBMV (Piyapolrungroj et al., 1999) but proved a weak inhibitor (compared with methimazole) of cimetidine S-oxidation in jejunal microsomes (Lu et al., 1998). The perfusion data obtained in this study, indicate that cimetidine transport across the mucosal membrane is inhibited by imipramine because both uncorrected P_eff and F_met are decreased by imipramine but corrected P_eff is not. This contrast in inhibition results between methimazole and imipramine also suggests that intracellular cimetidine is controlling absorption of cimetidine by rat jejunal epithelia. Imipramine's inhibition of cimetidine uptake and methimazole's greater inhibition of metabolism favor greater intracellular cimetidine concentrations with methimazole.

Methionine inhibition of cimetidine S-oxidation in jejunal microsomes was similar in magnitude to that observed with imipramine (Lu et al., 1998). In contrast to imipramine, methionine did not inhibit cimetidine uptake by jejunal BBMV (Piyapolrungroj et al., 1999). Because methionine does not reduce cellular cimetidine uptake but does inhibit metabolism, a reduction in both P_eff and corrected P_eff is observed. The fact that no significant change is observed in F_met may be accounted for by the fact that cimetidine metabolism parallels lumenal drug loss. This is consistent with projecting greater intracellular levels of cimetidine with methionine compared with imipramine coperfusion. This data further support a role for intracellular cimetidine levels in regulating cimetidine transport across paracellular pathways.

Based on these in situ studies as well as BBMV transport and microsomal metabolism studies, a mechanism for cimetidine transport and metabolism in the jejunum is proposed (Fig. 7). Cimetidine is absorbed paracellularly and intracellular uptake is mediated by pH-dependent and potential-dependent processes in parallel with facilitative uptake across the lateral membrane lining the paracellular pathway. Paracellular transport is regulated by intracellular cimetidine concentrations as a function of intracellular uptake and metabolism. Although cimetidine elimination is substantial in the jejunum and not the ileum, self-inhibition of absorption is significant in both regions. Cimetidine elimination in the jejunum is amplified by self-inhibition of transport at higher drug concentrations. This will not usually occur in the ileum because it is distal to sites of upper intestinal absorption and therefore is exposed to lower drug concentrations. This is projected to result in region-dependent absorption underlying the clinical observation of plasma level double peaks. Based on this absorption mechanism, lumenal variables including cimetidine dose, rate of dosage form release, and meal ingestion should effect cimetidine oral bioavailability and absorption variability through their impact on lumenal cimetidine concentrations in upper versus lower small intestine.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Schematic of projected mechanism of the transport and metabolism of cimetidine in jejunal epithelia.

Although our earlier report indicated that cimetidine S-oxide appearance in the jejunal lumen was greater in rats than humans (Hui et al., 1994), surface area to volume considerations indicate that cimetidine elimination in the human jejunum is substantial (see geometric verification in the Appendix). In this regard, this mechanistic data obtained in rats should provide perspective on clinical observations with cimetidine. This data in rats suggest that intestinal elimination will contribute to a lower rate of drug delivery to the systemic circulation from the jejunum than from the ileum. Because inhibition and elimination will be greater at higher lumenal concentrations, the appearance of double plasma peaks and reductions in cimetidine bioavailability would be expected to be more prominent under fasted- than fed-state administration conditions. Meal reductions in lumenal drug concentrations should result from a slower drug delivery rate from the stomach and dilution effects in the fed-state as compared with fasted-state administration. Whereas double plasma level peaks are primarily a function of reduced jejunal permeability, variation in bioavailability is predominantly a function of jejunal elimination as determined by lumenal drug concentration. In this regard, formulation factors that influence the release pattern of a cimetidine will also influence drug bioavailability. As supported by previous clinical data, tablet formulations of cimetidine produce greater drug plasma levels than equivalent doses in solution because lumenal concentration is limited by dosage form release rate (Walkenstein et al., 1978). The fact that greater plasma level variability is observed at higher cimetidine doses is also consistent with self-inhibition of jejunal absorption (Grahnen et al., 1979; Somogyi and Gugler, 1983).