Drugs that interact with grapefruit and/or grapefruit juice (GJ) undergo cytochrome p450 oxidative metabolism in the intestinal wall or liver. GJ contains various furanocoumarins which have been demonstrated to affect the cytochrome p450(CYP) system (especially at isoenzyme CYP3A4) by binding to the isoenzyme as a substrate and impairing first-pass metabolism, by direct inactivation or inhibition of the enzyme (mechanism-based inhibitors). The net effect on the CYP enzymes from this inhibition seems to be a selective down-regulation of CYP3A4 in the small intestine.²⁶ For certain drugs which are known to be CYP 3A4 metabolized, less drug is metabolized prior to absorption, and greater amounts of these drugs reach the systemic circulation, leading to higher blood levels, and potentially to increases in therapeutic and/or toxic effects.

Naringin

Naringin is the main bioflavonoid in GJ.  Naringin is not a potent CYP inhibitor, but is partially metabolized by enteral bacteria to naringenin, which is a potent inhibitor of p450 enzymes, and early research into GJ interactions proposed that naringin was the component of GJ responsible for the interactions,^4-6  although it was thought possible that another unidentified component in grapefruit may also have been responsible, since giving naringin alone does not seem to cause the same degree of inhibition as GJ.^4-5,21 Recent evidence clearly indicate that furanocoumarins, especially dihydroxybergamottin are the chemicals responsible for GJ interactions.

Furanocoumarins/Dihydroxybergamottin

Researchers have isolated a group of compounds from GJ called furanocoumarins, which appear to be specific CYP3A4 inhibitors.^27-29 A study on extracts of GJ interacting with rat and human p450 found that naringin accounted for only 10% of the inhibition of CYP activity seen with GJ.²⁸ In vitro results show that a compound known as 6',7'-dihydroxybergamottin may be the chemical which accounts for the difference in effects on CYP3A substrates caused by GJ versus naringenin.²⁷ A separate study with in vitro data determined that several compounds found in GJ inhibit CYP3A4 enzymes.  Specifically, these were nootkanone (a sesquiterpene), and 4 derivatives of coumarin, geranyloxycoumarin, bergamottin, and 2 chemical with very long technical names, denoted as GF-I-1 and 4.²⁹ Additional CYP3A inhibitory chemicals GF-I-5, GF-I-6 and bergapten have also been identified.¹³² The concentrations of nootkanone and bergamottin required to inhibit CYP3A4 enzymes however, was found to be higher than the naturally occurring concentrations found in GJ.

Results of confirmatory in vivo testing of CYP3A4 inhibition with externally administered  GF-I-1, GF-I-4, GF-I-5 and GF-I-6 have not been published at present. It should also be noted that wide inter-individual variability in response to these interactions have been documented in studies.^2,11

A recent study determined that the chemical moiety 5-geranyloxyfurocoumarine is essential for CYP3A4 inhibitory activity. This moiety is found in dihydroxybergamottin, as well as GF-I-1 and GF-I-4.¹²⁷ A total of 6 furanocoumarin derivatives are suggested to be clinically active constituents of GJ.¹²⁹

Commercial manufacturing of GJ is a multi-step process that includes washing of unpeeled fruit prior to squeezing whole fruits to juice. After squeezing, juice is heated for pasteurization, and volatile components may evaporate. Grapefruit oil, an important constituent of grapefruit peel, is sometimes added to the concentrate as a flavor enhancer.

A study found that epoxybergamottin, a furanocoumarin extracted from grapefruit peel is also an inhibitor of CYP 3A4. This compound has previously been found in only minor quantities in GJ, compared to 6'7'-dihydroxybergamottin. The authors note that it is possible that this compound could be distributed into the juice, during the commercial juice manufacturing process. Epoxybergamottin is not chemically stable in acidic medium, and may become hydrolyzed to 6'7'-dihydroxybergamottin.¹³⁴

P-glycoprotein

A study performed in cellular models indicates that GJ significantly activates p-glycoprotein mediated reduction in bioavailability, partially counteracting the CYP3A4 inhibitory effects of GJ.⁴² This experiment was performed in laboratory (CaCo-2) cell cultures, and was not a human trial. The results of this experiment were later found to be due to a laboratory artifact and were retracted¹³², though the retraction was not published by the original author. Subsequent results indicate that GJ likely acts as a weak p-glycoprotein inhibitor.

A recent study using GJ and digoxin shed further light on p-glycoprotein effects in humans. In a randomized crossover fashion, 12 healthy volunteers (7 male, 5 female) were given a single oral dose of digoxin 0.5 mg with either water or 250 mL of SSGJ 30 minutes prior to the dose, and 3.5, 7.5 and 11.5 hours after the dose. A two-week washout between study periods was employed.

Compared to the water group, mean peak blood concentrations (Cmax) rose 21%, time to peak level (Tmax) rose 8%, and total drug exposure at 48 hours (AUC48) rose 9%. None of these changes were statistically significant.  AUC at 4 and 24 hours also rose 9% and this was considered statistically significant. 

The authors consider GJ a weak inhibitor of P-glycoprotein, since studies with other known P-glycoprotein inhibitors show larger increases in digoxin AUC. They concluded that the study did not support the role of GJ as an important P-glycoprotein inhibitor, and that the modest increases in digoxin concentrations observed did not require any recommendations to alter digoxin dosing or titration when taken with GJ.¹¹⁴

In a similar study, 7 healthy volunteers (4 male, 3 female) received a single oral dose of digoxin 1 mg with with either water or GJ in a randomized crossover design. A washout of two weeks between study periods was employed.

During the GJ phase, 240 mL of SSGJ was consumed 3 times daily for 5 days prior to digoxin administration, and 6 days after digoxin administration, in an attempt to maximize the effect on p-glycoprotein.

Compared to water, administration of digoxin with GJ showed peak blood levels (Cmax) decreased by a mean of 16%, total drug exposure (AUC) increased 3%, Time to peak concentration (Tmax) increased 25% and half-life was essentially unchanged (decreased by 3%). None of these changes were determined to be statistically significant, although significant interindividual variability was observed.

The authors concluded that when results are taken with the previous study, GJ ingestion does not have a significant effect on intestinal P-glycoprotein activity, and that P-glycoprotein inhibition does not play an important role in grapefruit interactions.¹¹⁵

Seville Orange Juice

A trial of administration of cyclosporine with GJ and Seville orange juice, both of which contain 6',7'-dihydroxybergamottin, was conducted recently.  7 healthy volunteers were given either water, single-strength GJ, or Seville orange juice 30 minutes prior to a 7.5 mg/kg dose of cyclosporine (Sandimmune^® brand) in a randomized crossover design.  The effect of Seville orange juice on CYP 3A4 concentrations in 2 individuals (via duodenal biopsy) and the effect of 6',7'-dighydroxybergamottin on p-glycoprotein activity in vitro were also assessed.  The AUC for cyclosporine was increased 55% with GJ, whereas Seville orange juice did no significantly affect the AUC of cyclosporine. The two individuals who had duodenal biopsy showed clearly decreased enterocyte concentrations of CYP 3A4, suggesting that 6',7'-dihydroxybergamottin is not solely responsible for the increased cyclosporine levels when given with GJ.  The in vitro studies confirmed that 6',7'-dihydroxybergamottin had no effect on p-glycoprotein.

So, if the two juices had similar concentrations of 6',7'-dihydroxybergamottin, why did cyclosporine blood levels increase with GJ, but not with Seville orange juice ?  P-glycoprotein is known to play a significant role in cyclosporine availability, and based on this study, and in agreement with the study listed above, it is reasonable to conclude that GJ contains a compound or compounds that inhibit P-glycoprotein activity, which are not found in Seville oranges.  Further studies are needed to identify inhibitors of p-glycoprotein in GJ and to evaluate the relative contribution of reduced p-glycoprotein and CYP 3A4 activity to the increased oral bioavailability of other drugs.⁵¹ Seville oranges may selectively "knock out" CYP 3A4 activity, while the inhibitor(s) of p-glycoprotein in GJ appear to be different from those compounds identified as inactivating CYP 3A4.¹³²

Eleven healthy volunteers (6 male, 5 female) received 30mg dextromethorphan daily for 5 days with either 200 mL single-strength GJ (study day 2), 200 mL seville orange juice (SOJ) (study day 4), or water (study day 1, 3, 5) in a linear, non-crossover fashion. Immediately, the non-crossover design raises a concern, since the effect of GJ on the enzymes is known to persist for 72 hours or more. A 3 -day washout was given after study day 2, and at least 7-days passed between study day 2 and 4. Blood levels were not collected in this study. Urine point assays, and 8-hour total urine concentrations were used to determine dextromethorphan availability in a complicated series of analyses, based partially on animal data, and some assumptions.

The fraction of the administered dose of dextromethorphan that escaped first pass metabolism were found to increase significantly and similarly when GJ or SOJ were taken with dextromethorphan on study days 2 and 4, although lack of an adequate washout period after GJ intake on day 2 was provided, the administration of SOJ with dextromethorphan produced an identical effect on the dextromethorphan pharmacokinetic profile as was observed with GJ. Two volunteers experienced drowsiness, but were found to be CYP 2D6 "poor metabolizers" (the alternate metabolic pathway for dextromethorphan and its metabolites).  It also appears that both GJ and SOJ affected p-glycoprotein transport protein activity.⁹⁴

Ten healthy volunteers (5 male, 5 female) received a dose of felodipine 10mg with 240 mL of either fresh squeezed Seville orange juice, commercial orange juice, or grapefruit juice (diluted to contain the same total molar concentration of bergamottin and 6'7'-dihydroxybergamottin as the Seville orange juice). This study was conducted in randomized crossover design with at least a 7 day washout between study periods.

Seville orange juice increased felodipine AUC 76% and Cmax was increased 61%, compared to commercial orange juice (control). As expected, grapefruit juice increased felodipine AUC by 83% with Cmax increased 88%

An additional CYP 3A4 inhibitor, bergapten was found in seville orange juice, but not in grapefruit juice. Bergapten appears to have 1/3 the potency of 6'7'-dihydroxybergamottin when tested with midazolam in intestinal cell concentrations.¹³²

Pomelo & Other Related Citrus Fruits

A case report exists of an interaction between tacrolimus and pomelo (also known as pummelo), a grapefruit-related citrus fruit. A patient taking tacrolimus post-renal transplant was stabilized on a tacrolimus dosage of 6 mg/day with tacrolimus blood levels stable in the therapeutic range of 8-10 ng/mL.  A subsequent check of the tacrolimus level was increased at 25.2 ng/mL. The patient had no subjective symptoms. Upon further questioning, it was revealed that he had consumed approximately 100g of pomelo from his garden just prior to the tacrolimus on the day before blood sampling for tacrolimus level determination.¹¹⁷

Follow up testing by the above authors confirmed that some forms (1 of 3 tested) of pomelo juice extract were as potent inhibitors of CYP 3A4 as grapefruit juice extract in vitro. The pomelo extract had little effect on p-glycoprotein in a cellular model.¹²⁴

In 2004, Japanese researchers developed an enzyme-linked immunosorbent assay specific for coumarin derivatives that contain the geranyloxy- side chain, known to be the chemical moiety in grapefruit responsible for CYP3A4 inhibition. This allowed other fruits to be screened for containing 6'7'-dihydroxybergamottin.  

Of 15 fruits screened, white grapefruit (control), red pummelo (pomelo), sweetie (oro blanco), melogold, banpeiyu pummelo, hassaku orange, sour (seville) orange, lime and natsudaidai showed significant immunoreactivity, indicating the presence of furanocoumarin derivatives. Navel orange, sweet orange and yuzu showed slight immunoreactivity, while iyokan orange, satsuma mandarin, ponkan mandarin and dekopon mandarin showed minimal immunoreactivity.

The researchers noted that besides the Rutaceae family (grapefruit-like) many plants of other families such as Umbelliferae, Leguminosae and Moraceae also contain furanocoumarin derivatives. Many plants of these families are used as common vegetables or traditional medicines, and it is possible that furanocoumarin derivatives contained in these plants could change the pharmacokinetics of certain drugs.¹²⁹

In a related study, Japanese researchers performed in vitro testing of local citrus fruits by incubating citrus fruit residue with human liver extract, midazolam, and measuring residual CYP 3A activity. Fruit juice residues prepared from banpeiyu pummelo, hassaku orange, takaoka-(suisho) buntan pummelo and kinkan (Tamatama) inhibited microsomal CYP 3A activity. Banpeiyu pummelo inhibition was strongest, though weaker than the white grapefruit control. No significant CYP 3A inhibition was noted with ama-natsu orange, dekopon mandarin, hyuga-natsu orange, unshu-mikan (satsuma mandarin) or navel orange. These findings are in agreement with the above study which looked at enzyme-linked immunoreactivity testing.¹³⁰

In another related study by Japanese researchers, in vitro testing of tropical fruits was performed by incubating citrus fruit residue with human liver extract, midazolam, and measuring residual CYP 3A activity. Residues prepared from star fruit, pomegranate, and common papaw significantly inhibited CYP 3A activity. In all three cases, inhibition was stronger than the white grapefruit control. No significant inhibition was noted with valencia orange, mango, rambutan, kiwi fruit, dragon fruit or passion fruit.¹³¹

Time course of GJ-drug Interactions

A research group conducted a study with simvastatin to characterize the duration of the GJ induced CYP 3A4 inhibition. Ten healthy volunteers (9 male, 1 female) received simvastatin 40 mg with either water as a control, and either 0, 1, 3 or 7 days after drinking high-dose GJ (200 mL double-strength GJ) three times daily for 3 days in a non-randomized crossover fashion.  As seen with the previous study, significant increases in simvastatin levels after GJ intake were observed on the day 0 study when compared with water. Simvastatin AUC was increased 1250%, with Cmax increased 1104%. Time to peak concentration (Tmax) was also prolonged from 2 hours to 4 hours (100% increase).  This effect was significantly reduced if 24 hours elapsed between the last GJ intake and simvastatin dosing. At this time, simvastatin AUC was increased 105%, and Cmax was increased 136%.  The authors noted that the effect of even high-dose GJ 24 hours after ingestion is only about 10% of that seen with concurrent GJ and simvastatin intake. AUC and Cmax of simvastatin when taken on day 3 and day 7 after last GJ intake were not significantly elevated compared to control, indicating that the interaction potential of even high amounts of GJ intake dissipates within 3-7 days after last GJ ingestion. 

This is useful for characterizing the time course of GJ-drug interactions, and fits with prior expectations that the GJ effect can last up to 3 days after last GJ ingestion.⁹⁰

In a trial to demonstrate the time course of recovery of CYP 3A4 function after grapefruit intake, midazolam was chosen as the CYP 3A4 substrate.

25 healthy volunteers (21 male, 4 female) received midazolam 6 mg orally as a control prior to grapefruit exposure. Then, midazolam 6 mg was given orally 2 hours following 280 mL of single-strength grapefruit juice. Then, subjects were randomly assigned to receive a third dose of midazolam 6 mg orally either 26 hours, 50 hours or 74 hours post grapefruit juice ingestion (i.e. 1-3 days post grapefruit intake). No additional grapefruit juice (GJ) was administered at any time during the study.

Increases in midazolam Cmax (peak blood level) and AUC (total drug exposure) relative to the control (no grapefruit intake) are shown below.

	2h post-GJ	26h post-GJ	50h post-GJ	74h post-GJ
Cmax	54%	25%	5%	no change
AUC	65%	21%	21%	6%

Plotting an AUC ratio vs. time curve allowed estimation of a half-life for CYP 3A4 enzyme recovery of 23 hours. Based on this study, the authors concluded that the recovery of enteric CYP 3A4 function should be close to complete within 3 days of a single exposure to grapefruit juice.¹²⁶

Effect of Grapefruit Dose

Large dose of grapefruit may affect CYP 3A4 in the liver, in addition to intestinal CYP 3A4.

In phase I of this study, 8 healthy male volunteers received 240 mL of double strength GJ or water three times daily for 2 days, followed by 240 mL of double strength GJ 90, 60 and 30 minutes prior to a single 2 mg oral midazolam dose. The study was conducted in randomized fashion, with a 7 day washout period between treatment arms. Radiolabelled erythromycin was given intravenously at the same time, to conduct an erythromycin breath test (EBT), a relatively specific test of hepatic CYP 3A4 function.

In phase II, 16 healthy male volunteers received either a single dose of 240 mL of single-strength GJ, double-strength GJ or water 60 minutes prior to a single 2 mg oral midazolam dose. The study was conducted in randomized fashion, with a 7 day washout period between treatment arms. Again, radiolabelled erythromycin was given intravenously at the same time, to conduct EBT. 

In the single dose/single strength GJ group, AUC increased by 100%, Cmax increased by 115%, and half-life increased by 32% compared to placebo. In the single dose/double strength GJ group, AUC increased by 114%, Cmax increased by 99%, and half-life increased by 15% compared to placebo.

In the multiple dose/double strength GJ group, AUC increased by 495%, Cmax increased 168%, and half-life increased 110% compared to placebo.

In the single dose GJ groups, EBT excretion was essentially unchanged from placebo, while in the multiple dose GJ group, EBT excretion decreased by 18%.

EBT excretion and half-life are measures of systemic CYP 3A4 activity. In most other low-dose GJ studies half-life is usually, but not always, similar in the GJ group to placebo. The magnitude of effect on EBT in the multiple dosing group was similar to that of other known CYP 3A4 inhibitors, such as mibefradil, but less that ketoconazole, one of the most potent CYP 3A4 inhibitors. The authors conclude that large amounts of GJ could inhibit hepatic as well as intestinal CYP 3A4, and that patient who take drugs metabolized by CYP 3A4 should avoid consuming large amounts of GJ to eliminate the possibility of adverse drug interactions.¹³³

Confirmation of effect of raw grapefruit and extract

To confirm that the activity of grapefruit segments and an extract of the peel and rind had similar drug interaction potential to the juice, 12 healthy volunteers (10 male, 2 female) received a single dose of felodipine extended-release 10 mg with either 250 mL single strength GJ, peeled grapefruit segments of one grapefruit equivalent to 250 mL, the ethanol-based extract of a single grapefruit diluted to 250 mL with water, or water in a four-way randomized crossover fashion. 

All three forms of grapefruit significantly increased the bioavailability compared with water alone. Time to peak concentration (Tmax) was shortened significantly with GJ and the extract, while none of the 3 grapefruit preparations affected the half-life of felodipine. Clinical effects on blood pressure and heart rate were not reported.

The authors concluded that based on current knowledge of which drugs interact, if there is a concern for a drug interaction with GJ, it seems logical to avoid consumption of grapefruit segments as well during pharmacotherapy with the affected drug(s). Confectioneries, like marmalades, made from grapefruit peel may also cause a drug interaction.¹⁰⁶

Organic Anion Transporting Polypeptide (OATP) effects

Recently, new information has become available that may change the way we think about Grapefruit-drug interactions, and explains why sometimes a decrease in bioavailability is seen with GJ and some drugs.

Twenty-four healthy volunteers received a single 60 mg dose of fexofenadine following 2 days of either 240 mL double-strength GJ or water in a randomized crossover fashion.  AUC was decreased a mean of 33% and Cmax was decreased a mean of 36% in the GJ group. No changes in QTc intervals on the EKG were noted in any patient receiving fexofenadine and GJ.¹⁰⁴ This is unexpected since fexofenadine does not undergo significant biotransformation by CYP enzymes. However, co-administration of fexofenadine with ketoconazole and erythromycin, which as known CYP 3A4 inhibitors, resulted in significant increases in the extent of absorption of fexofenadine.¹⁰⁵ (Ketoconazole and erythromycin have since been shown to inhibit p-glycoprotein.^{ref1, ref2}) This builds on a recent study detailed below, which suggested that fexofenadine may be a substrate for organic anion transporting polypeptide (OATP), another transporter system, and that grapefruit, and possibly orange, apple, and grape juices may also affect disposition of fexofenadine through effects on OATP.

Ten healthy volunteers (6 male, 4 female) received a single dose of fexofenadine 120mg with 1200 mL of single-strength GJ, 25% strength GJ, plain orange juice, or apple juice in a 5-way randomized crossover design. Compared to the water group, all four fruit juice groups had significantly decreased AUC Cmax, as shown in the table.
 

 
Time to peak concentration (Tmax) and half-life were not significantly affected by any of the juices, consistent with reduced bioavailability as opposed to increased systemic elimination. The authors concluded that the results of this study provide support for a new model of food-drug interaction, that includes the role of OATP, as well as CYP 3A4, and p-glycoprotein.⁸³  In a non-clinical experiment, this study also looked at the effects of varying concentrations of GJ, orange juice, apple juice and grape juice on p-glycoprotein and OATP activity.

In CaCo-2 cell monolayers, 5% strength GJ, homogenized grapefruit segments, and apple juice did not significantly affect transport of p-glycoprotein substrates digoxin and vinblastine. However, 5% strength GJ, orange juice and apple juice significantly inhibited the uptake of fexofenadine by OATP. (5% strength juices were chosen to prevent damage to cell integrity from stronger concentrations).