Drug-Drug Interaction between Diclofenac and Gamma-Hydroxybutyric Acid

Abstract

Gamma hydroxybutyric acid (GHB) has been approved clinically to treat excessive daytime sleepiness and cataplexy in patients with narcolepsy, alcohol and opioid withdrawal, and as an anesthetic. The use of GHB clinically is limited due to its high abuse potential. The absorption, clearance and tissue uptake of GHB is mediated by proton-dependent and sodium-coupled monocarboxylate transporters (MCTs and SMCTs) and inhibition of these transporters may result in a change in GHB pharmacokinetics and pharmacodynamics. Previous studies have reported that non-steroidal anti-inflammatory drugs (NSAIDs) may inhibit these monocarboxylate transporters. Therefore, the purpose of this work was to analyze the interaction between GHB (at a dose of 600 mg/kg i.v.) and the NSAID, diclofenac, by examining the effects of these drugs on the in vivo pharmacokinetics and pharmacodynamics in rat studies. The pharmacodynamic effect evaluated was respiratory depression, a measure of toxicity observed by GHB at this dose. There was an improvement in the respiratory rate with diclofenac administration suggesting an effect of diclofenac on GHB toxicity. In vitro studies with rat blood brain endothelial cells (RBE4) that express MCT1 indicated that diclofenac can inhibit GHB transport with an IC₅₀ of 10.6 μM at pH 7.4. In vivo studies found a decrease in brain GHB concentrations and a decrease in the brain-to-plasma concentration ratio following diclofenac treatment. With this study we can conclude that diclofenac and potentially other NSAIDs can inhibit the transport of GHB into the brain, therefore decreasing GHB’s pharmacodynamic effects and toxicity.

Graphical Abstract

Diclofenac-GHB Drug-Drug Interaction

This research provides in vitro and in vivo evidence that the NSAID diclofenac reverses the toxicity (as measured by respiratory depression) of the drug of abuse γ-hydroxybutyric acid (GHB) by inhibiting the its Monocarboxylate Transporter 1-mediated uptake into the brain. These findings are consistent with clinical studies suggesting a pharmacodynamic interaction of diclofenac in patients receiving the therapeutic GHB product Xyrem®.

Introduction:

Gamma hydroxybutyric acid (GHB) has been approved for a number of clinical uses globally. In the United States, Canada and Europe GHB is approved to treat excessive daytime sleepiness and cataplexy in adult patients with narcolepsy as Xyrem® (Carter et al., 2009), and in pediatric patients in 2018 (Plazzi et al., 2018). It is approved in Germany as an anesthetic (Somsanit®), and in Austria and Italy to treat alcohol and opioid withdrawal (Alcover®) (Carter et al., 2009). Its clinical use is limited due to its high abuse potential and GHB is categorized as a Schedule I drug in the US for non-medical uses and Schedule III for its therapeutic uses (Brennan and Van Hout, 2014).

With a pKa of 4.7 (NationalCenterforBiotechnologyInformation, 2021b), GHB is ionized at physiological pH, and membrane transporters play an essential role in its tissue uptake. It has been shown that GHB is a substrate for monocarboxylate transporters, including the proton-dependent monocarboxylate transporters MCT1, MCT2, and MCT4 (Wang et al., 2006; Wang and Morris, 2007), along with sodium-dependent transporters, SMCT1, and SMCT2 (Morris and Felmlee, 2008; Cui and Morris, 2009). While MCT1, 2 and 4, as well as SMCT1 and 2, are present in the kidney and in the intestine and responsible for the absorption and renal reabsorption of GHB, MCT1 is the only monocarboxylate transporter present at the blood-brain barrier. Inhibition of these transporters has been studied extensively by our laboratory, as a possible treatment for GHB overdose (Morse et al., 2012; Morse and Morris, 2013; Vijay et al., 2015; Follman and Morris, 2019). Studies with the MCT/SMCT inhibitor L-lactate and the MCT1 inhibitors AR-C155858 and AZD-3965 have demonstrated inhibition of the renal reabsorption of GHB resulting in decreased renal and total clearance, as well inhibition of uptake into the brain, resulting in changes in both the pharmacokinetics, pharmacodynamics and toxicity of GHB (Morse and Morris, 2013; Vijay et al., 2015; Follman and Morris, 2019).

Some non-steroidal anti-inflammatory drugs (NSAIDs) have also been evaluated as inhibitors of MCT1 in cell-based studies. In our laboratory, we studied the interaction of ketoprofen and ibuprofen with MCT1 and SMCT1. It was shown that these compounds inhibit both transporters in a concentration-dependent manner (Cui and Morris, 2009). Jazz Pharmaceuticals demonstrated in a clinical trial study that diclofenac and valproic acid can alter GHB pharmacodynamics. This drug-drug interaction resulted in increased cognitive effects, including power of attention, digit vigilance accuracy and reaction times, in patients receiving Xyrem®, compared to Xyrem administration alone at therapeutic doses (Eller, 2013). Diclofenac (figure 1) is an NSAID used to treat pain, inflammatory disorders, and dysmenorrhea. It has a half-life of 1–3 hours and is highly bound to albumin (Eller, 2013). Since diclofenac may be prescribed with GHB, it is important to understand this drug-drug interaction which may affect the efficacy of GHB when used in the treatment of narcolepsy, due to changes in brain concentrations. Table 1 summarizes the physicochemical and pharmacokinetic properties of diclofenac (NationalCenterforBiotechnologyInformation, 2021a).

Figure 1.

Molecular structure of GHB (A) and diclofenac (B). Figure obtained from Pubchem (NationalCenterforBiotechnologyInformation, 2021a).

Table 1.

Physicochemical and pharmacokinetic properties of Diclofenac

	Diclofenac
Physicochemical properties
pKa	4.15
LogP	4.51
Pharmacokinetic properties
Half-life	1–3 hours
Protein Binding	99% to albumin
Elimination	Extensively metabolized by CYP2C9, UGT2B7
% of dose unchanged excreted in urine	6%

CYP2C9- cytochrome P-450 2C9: UGT2B7- uridine 5’diphosphoglucuronosyl transferase 2B7

The objective of this research was to study the interaction between NSAID, diclofenac, and GHB and the effect of this DDI on the pharmacokinetics and pharmacodynamics of GHB in rats.

Material and Methods:

Chemical and reagents

The National Institute on Drug Abuse (NIDA) provided sodium GHB salt (Na-GHB). Diclofenac was purchased from Millipore Sigma (St. Louis, MO). [3H]-GHB was purchased from American Radiolabeled Chemicals, Inc. (St. Louis, MO). Deuterated GHB (GHB-d6) was purchased from Cerilliant Corporation (Round Rock, TX). High-performance liquid chromatography grade acetonitrile was purchased from Honeywell Burdick & Jackson (Muskegon, MI).

Cell culture

The immortalized rat brain capillary endothelial (RBE4) cell line was kindly provided by Prof. P. Couraud (University Rene Descartes, Paris). Following the protocol described by Roiko et al. (Roiko et al., 2012; Roiko et al.2013), cells were cultured as monolayers on 75-cm² flasks that were coated with Type I rat-tail collagen (150 μg/ml) before plating. Cell culture media consisted of 1:1 α-minimum essentials medium (MEM)/Hams F-10 nutrient mixture supplemented with L-glutamine (2.0 mM), geneticin (300 μg/ml), human recombinant fibroblast growth factor (1 ng/ml), gentamicin (50 μg/ml), and 10% v/v FBS. Cells were incubated at 37°C in a humidified atmosphere with 5% CO₂/95% air. The culture medium was changed every 2 to 3 days, and cells were passaged using 0.25% trypsin-EDTA. For protein extraction, cells were seeded in 35×100 mm cell culture dishes at a density of 200,000 cells/well. RBE4 cells at passages 11, 13 and 22 were used for protein characterization. For uptake studies cells at passage 12 was used. Studies were conducted once cells reached a monolayer.

Cellular uptake of GHB into RBE4 cells

Once confluency was reached, uptake studies were performed with [3H]-GHB following the protocol described by Roiko et al. (Roiko et al., 2012). Briefly, cell media was removed, and cells were washed 3 times with room temperature uptake buffer. One mL of uptake buffer containing diclofenac concentrations ranging from 1–300 μM was added and allowed to incubate at 37°C for 30 minutes at pH 7.4. Optimal incubation time was determined with a time-dependent incubation study (data not shown). Once the incubation was over, the uptake buffer was removed, and 1 mL of [3H]-GHB (2.33 mM) was added to the dishes to initiate the uptake reaction at room temperature. Uptake was terminated after 1 minute by removing uptake buffer, and the cells were washed 3 times with ice-cold uptake buffer. Cells were lysed with 1N NaOH for 1 hour. Once lysis was done, the cell lysate was neutralized with 1N HCl. Radioactivity was determined using a liquid scintillation counter (1,900 CA, Tri-carb liquid scintillation analyzer; Packard Instrument Co. Downers Grove, IL). Protein concentration in the cell lysate was determined using Pierce™ BCA Protein Assay Kit (Thermo Fisher, Waltham, MA) with bovine serum albumin (BSA) as standard. Results were normalized to protein concentration. All studies were performed in triplicate.

Western Blotting

Once cells were confluent, they were lysed with lysis buffer as described by Bryniarski et. al. (Bryniarski et al., 2018). Briefly, this method includes the use of a lysis buffer containing protease inhibitor, followed by scraping with a cell scraper and then transferring the supernatant for further centrifugation for 15 minutes at 12,000 RPM at 4°C. The supernatant was then collected and protein concentrations were determined by the Bradford assay using bovine serum albumin (BSA) as the standard. Samples were diluted to achieve a concentration of 1.0 μg/μL, then 15 μg were added to the NuPAGE 4–12% Bis-Tris gel (Invitrogen, Carlsbad, CA.) and the gel run using NuPAGE MOPS SDS Running Buffer (Invitrogen). The protein was transferred to a nitrocellulose membrane utilizing NuPAGE Transfer buffer (Invitrogen) with 20% 190-proof ethanol; following transfer, the membranes were blocked with 5% (w/v) nonfat milk in Tris-buffer saline containing 10% (v/v) Tween 20 (TBST) for one hour. Primary antibodies for MCT1 (AB3540P, Millipore Sigma) were prepared in 5% (w/v) nonfat milk in TBST (1:1000) and left rocking overnight at 4°C. The following day, incubation with the secondary antibody was performed at room temperature for 1 hour, using a secondary antibody coupled to horseradish peroxidase. Then the membranes were incubated with enhanced chemiluminescence (ECL) substrates followed by visualization in ChemiDocTMXRS + system (Bio Rad, Hercules, CA). Beta-actin antibody (4970S, Cell Signaling Technology, Beverly, MA) was used as a loading control, and SeeBlue Plus 2 ladder (Thermo Fisher) was used to confirm the molecular weight of the protein.

Animals and surgery

Male Sprague-Dawley rats (Envigo, Somerset, NJ) weighing 250–330g were used for experiments. Animals were housed under controlled temperature and humidity with an artificial 12-hour light/dark cycle, and water/food was available ad libitum. All animal protocols were approved by the Institutional Animal Care and Use Committee at the University at Buffalo. Animals were allowed to acclimate to their environment one week before any procedure. Surgical implantation of the jugular cannula was implanted under anesthesia with ketamine/xylazine solution. After surgery, cannulas were flushed daily with 40 IU/mL heparinized saline to maintain patency. Animals were allowed to recover a minimum of three days after surgery before any experiment was conducted.

Pharmacokinetic/Pharmacodynamic (PK/PD) Studies

The effects of diclofenac on the PK/PD studies of GHB were assessed utilizing whole-body plethysmography and serial blood and urine sampling (model PLY4213; Buxco Research Systems, Wilmington, NC) as previously used in our laboratory (Morse et al., 2012; Follman and Morris, 2019). Studies consisted of administration of diclofenac five minutes before the administration of GHB. Diclofenac was administered as 50 and 75 mg/kg intravenously as a 50mg/mL solution in 20 % cyclodextrin/sterile double distilled water (w/v). GHB was administered at a dose of 600mg/kg as a 300mg/mL solution in sterile double distilled water. Solutions were administered intravenously via the jugular vein cannula. On the day of the study, rats were placed in the plethysmography chambers one hour before the study and allowed to acclimate for 45 minutes before baseline readings were recorded. GHB administration was considered time 0 and respiration measures were recorded at 2.5, 5, 7.5, 10, 15, 20, 25, 30, and every 15 minutes after that for a duration of 8 hours. To study the toxicodynamic effects, respiratory parameters measured were breathing frequency (rate), tidal volume and minute volume (rate*tidal volume). To study the drug effects on GHB toxicokinetics, blood samples were collected after GHB administration at 3, 11, 21, 31, 61, 121, 181, 241, 301, 331, 361, and 481 min. Urine was collected in intervals from 0–1, 1–2, 2–4, 4–6, and 6–8 h.

Plasma/Brain concentrations over time study.

The brain to plasma partitioning of GHB was analyzed after administration of 600 mg/kg GHB administered i.v., as previously performed in our laboratory (Morse and Morris, 2013; Follman and Morris, 2019). Animals were administered 600 mg/kg GHB i.v. alone or with diclofenac, administered 5 minutes before administration of GHB as a 75 mg/kg i.v. dose. Animals were sacrificed at 60 minutes after GHB administration, corresponding to 65 min after diclofenac administration. Terminal plasma and whole brain samples were collected at the time of sacrifice. Brain samples were snap-frozen in liquid nitrogen and stored at −80°C until analysis. Blood samples were centrifuged, and plasma was stored at −80°C until analysis.

Sample Analysis

The plasma and urine concentrations of GHB were determined using a previously validated liquid chromatography coupled to mass spectrometry assay (Felmlee et al., 2010). Briefly, plasma samples collected before the 241 min period were prepared by diluting 5 μL of the sample with 50 μL of blank plasma; for plasma samples, after this time point, 50 μL of the sample was utilized. The standard curve concentration ranged from 1 to 500 μg/mL. GHB standards were prepared by adding 5 μL of stock solution to 50 μL of blank plasma. GHB-d6 (5 μL), internal standard, was added to all samples. Acetonitrile (800 μL) was added to precipitate proteins; samples were centrifugated for 20 minutes at 10,000 rpm at 4°C. Supernatant was dried under a stream of nitrogen and reconstituted in 250 μL of the mobile phase. Urine samples were diluted 100x with blank urine and 5 μL of the internal standard. The standard curve was prepared by adding 5 μL of stock solutions to 25 μL of blank urine and 5 μL of the internal standard. Methanol (1 mL) was added, and double-distilled water (470 μL for samples, 465 μL for samples). Samples were centrifugated for 20 minutes at 10,000 rpm at 4°C. Supernatant was transferred to a clean vial for analysis.

Data/Statistical Analysis

PK/PD and cellular uptake data were graphed utilizing GraphPad Prism 7 (GraphPad Software, La Jolla, CA). The pharmacokinetic parameters were determined using Excel add-ins PK solver (Zhang et al., 2010). IC₅₀ parameter fitting was performed utilizing GraphPad Prism 7. The area under the plasma concentration-time curve (AUC) was determined using the trapezoidal method. The total clearance (CL) was determined as dose/AUC, the renal clearance (CL_R) was determined as A_e,∞/AUC, where A_e,∞ is the total GHB excreted unchanged in the urine, and the non-renal clearance (CL_NR) was determined as CL-CL_R. The pharmacodynamic parameters analyzed were the area below the effect curve (ABEC) and maximum effect (E_max). One-way analysis of variance followed by Dunnett’s post hoc and student t-test was performed on the PK/PD parameters obtained for all studies to determine statistical significance using Graph Pad Prism 7. The difference resulting in p<0.05 were considered significant.

Control data, 600 mg/kg GHB i.v., was obtained from previous publications performed by our laboratory (Morse and Morris, 2013; Vijay et al., 2015). The historical data was used to reduce the number of animals utilized in these studies. All experiments were conducted utilizing the same equipment and protocols.

Results:

Protein Expression of Monocarboxylate Transporters in RBE4 cells.

The protein expression of MCT1 was confirmed to be present in the cells. The MCT1 expression normalized by β-actin were 2.23 ± 0.26, 1.51 ± 0.35 and 1.73 ± 0.40 for passages 11, 13 and 22 (p > 0.05), suggesting unchanged MCT1 expression with the passage numbers examined (figure 2).

Figure 2.

Protein characterization of RBE4 cells by western blotting. Cells passages 11, 13 and 22 were analyzed.

Effect of Diclofenac on GHB cellular uptake.

To study the inhibitory effect of diclofenac on GHB uptake in RBE4 cells were pre-incubated with diclofenac (1–100 μM) under sodium-containing buffer conditions at pH 7.4 (figure 3). The inhibition constant (IC₅₀) of 10.6 μM was determined by fitting the data using GraphPad Prism 7.

Figure 3.

Effect of diclofenac on GHB uptake into RBE4 cells. Cells were incubated with diclofenac for 30 minutes before the uptake of GHB was completed. The uptake reaction was carried out for 1 minute at a pH 7.4.

Effect of Diclofenac on GHB pharmacokinetics.

The administration of diclofenac (50 and 75 mg/kg i.v.) 5 minutes before administration of GHB (600 mg/kg i.v.) did not affect the GHB renal or total clearances. The area under the plasma concentration-time curve (AUC) and the urinary excretion of GHB (figure 4) were similar to that with the GHB control group. There were no differences in pharmacokinetic parameters between the GHB control group or pre-treatment with diclofenac (table 2). Higher doses of diclofenac could not be evaluated due to toxicity.

Figure 4.

GHB plasma (A) and urine (B) concentrations after administration of 600mg/kg GHB i.v. alone and with pre-treatment of diclofenac (50 or 75 mg/kg i.v.). Data presented as mean ± SD (n=3–5).

Table 2.

Effect of diclofenac pre-treatment on GHB pharmacokinetics

Parameter	GHB (n=5)	GHB + Diclofenac 50 mg/kg (n=3)	GHB + Diclofenac 75 mg/kg (n=4)
AUC (mg*min/mL)	102 ± 12	95.4 ± 6.8	101 ± 12
CL (mL/min/kg)	6.00 ± 0.74	6.31 ± 0.44	5.95 ± 0.67
CLR (mL/min/kg)	1.68 ± 0.75	1.74 ± 0.61	1.77 ± 0.51
CLNR (mL/min/kg)	4.31 ± 0.34	4.79 ± 0.26	4.18 ± 0.47

Data presented as mean ± SD (n=3–5). Diclofenac was administered at a dose of 50 and 75mg/kg i.v. 5 minutes before GHB (600 mg/kg i.v.).

Effect of Diclofenac on GHB pharmacodynamics.

Diclofenac administration (50 and 75 mg/kg i.v.) 5 minutes before GHB (600 mg/kg i.v.) improved the respiratory depression produced by GHB. Respiratory parameters of breathing frequency, tidal volume, and minute volume are presented in figure 5. There was a statistically significant decrease in the ABEC for breathing frequency and also a significant increase in frequency E_max for both doses of diclofenac (table 3). The tidal volume ABEC is significantly lower for the higher dose of diclofenac (table 3).

Figure 5.

Effect of diclofenac pre-treatment on respiratory parameters. Diclofenac (50 or 75 mg/kg) was administered intravenously 5 min before GHB (600mg/kg i.v.) administration. Time zero represents the administration of GHB. Data presented as mean ±S.D. (n=3–5).

Table 3.

Effect of diclofenac pre-treatment on GHB-induced respiratory depression

Toxicodynamic Parameter	GHB (n=5)	GHB + Diclofenac 50 mg/kg (n=4)	GHB + Diclofenac 75 mg/kg (n=4)
Frequency ABEC (breaths)	7161 ± 204	4340 ± 1664*	4337 ± 1324*
Frequency Emax (breaths/min)	30.7 ± 5.0	44.9 ± 12.0*	46.5 ± 6*
Tidal Volume ABEC (mL/breath*min)	207 ± 51	128 ± 77	71.8 ± 34.3*
Tidal volume Emax (mL)	3.63 ± 0.69	2.96 ± 1.37	2.13 ± 0.44
Minute Volume Emax (mL/min)	81 ± 16	64.6 ± 12.9	63.5 ± 21.9

Data presented as mean ± SD (n=3–5). Diclofenac was administered at a dose of 50 and 75 mg/kg i.v. 5 minutes before GHB (600 mg/kg i.v.).

^*Significantly different from GHB alone (p < 0.02) determined by One-way ANOVA with Dunnett’s post hoc test.

Effect of diclofenac on GHB brain to plasma partitioning.

As shown in figure 6 and table 4, coadministration of 75 mg/kg diclofenac significantly decreased the plasma and brain concentrations and the brain-to-plasma concentration ratio observed with GHB alone, when determined at 60 minutes after GHB administration.

Figure 6.

Effect of diclofenac on GHB plasma (A), brain (B), and brain to plasma ratio (C) concentrations at 60 minutes post intravenous GHB dose. Diclofenac (75 mg/kg i.v.) was administered 5 minutes before GHB. GHB (600 mg/kg i.v.) was administered at time zero. Animals were euthanized at 60 minutes post GHB dose. Student t-test was used to determine statistically significant differences. Data are presented as the mean ± S.D. (n = 3–4). *, p< 0.05

Table 4.

Effect of diclofenac on the brain to plasma partitioning of GHB

Treatment	Cplasma (μg/mL)	Cbrain (μg/mL)	Brain/Plasma ratio
GHB (n=3)	890 ± 20	243 ± 10	0.273 ± 0.006
GHB + diclofenac (n=4)	718 ± 62*	131 ± 58*	0.180 ± 0.070*

Data presented as mean ± SD (n=3–5). Diclofenac was administered at a dose 75 mg/kg i.v. 5 minutes before GHB (600 mg/kg i.v.).

^*Significantly different from GHB alone (p < 0.05).

Discussion:

The interaction between the NSAIDs ibuprofen and ketoprofen and GHB have been studied previously in our laboratory. It was found that ketoprofen and ibuprofen inhibit the uptake of GHB in vitro in cell studies, by inhibiting both MCT1- and SMCT1- mediated transport (Cui and Morris, 2009). Interestingly, ibuprofen has been demonstrated to be an inhibitor but not a substrate for SMCT1 (Itagaki et al., 2006). When we examined the effect of diclofenac on the in vitro uptake of GHB in rat brain endothelial cells that express MCT1, the only isoform of MCT present at the BBB, we found inhibition with a IC₅₀ of 10.6 μM.

Treatment with diclofenac did not alter GHB pharmacokinetics, based on plasma concentration-time profiles. There was no change in AUC, total clearance, renal clearance and non-renal clearance. This was not unexpected, since the major effect on drug clearance of MCT inhibitors is through inhibition of the renal reabsorption of GHB resulting in increased renal clearance. However, diclofenac is extensively metabolized and eliminated only to a minor extent as unchanged drug in the urine (Riess et al., 1978), so would not be expected to be present in the renal tubular fluid at high enough concentrations to inhibit GHB reabsorption. On the other hand, when we examined respiration parameters, we observed a dose-dependent improvement in respiratory rate. Respiratory rate is used as our pharmacodynamic endpoint since it represents one of the adverse effects of GHB and is the main reason for toxicity and death following GHB overdoses. There was a significant decrease in frequency ABEC and on tidal volume ABEC for the higher dose (75 mg/kg) of diclofenac when compared to GHB alone administration. Of note, the administration of diclofenac alone did not produce a change in respiration parameters (data not shown). Significantly, brain concentrations and the brain-to-plasma concentration ratio was significantly decreased when diclofenac was administered with GHB, suggesting MCT1-mediated inhibition of the brain uptake of GHB occurs in vivo, resulting in the potential for decreased pharmacodynamic effects.

Jazz Pharmaceuticals (Eller, 2013) reported significant effects of diclofenac on GHB (Xyrem®) pharmacodynamics in clinical studies. They observed an improvement in cognitive effect and attention in the presence of diclofenac, suggesting that decreased brain concentrations of GHB may be responsible for these effects. It is crucial to understand these drug-drug interactions since it will affect GHB efficacy for the therapeutic use of GHB in narcolepsy or for other therapeutic uses. Our study results indicate significant effects of diclofenac on brain concentrations of GHB in rats. It is possible that the diclofenac metabolites formed in vivo, namely diclofenac acyl glucuronide or 4-hydroxydiclofenac may contribute to the inhibition of MCT1 (Tang, 2003). Dose adjustment may be required for patients that take GHB in conjunction with diclofenac or other NSAIDs, based on this preclinical data.

In conclusion, our preclinical studies have demonstrated that the co-administration of the diclofenac with GHB can result in significant drug-drug interactions mediated by monocarboxylate transporters. Diclofenac altered the pharmacodynamics of GHB due to decreased uptake of GHB in the brain. Further clinical studies are necessary to evaluate the effects of NSAIDs on the therapeutic effects and toxicity of GHB.