Impact of Rabeprazole on APO-Dabigatran Exposure in Healthy Volunteers

Abstract

Background

Dabigatran is effective and safe for stroke prevention in patients with atrial fibrillation and for venous thromboembolism prevention and treatment. In Canada, APO-dabigatran, a generic formulation, has been approved based on a bioequivalence study, but its bioavailability in settings of reduced gastric acidity has not been examined.

Methods

Treatment With APO-Dabigatran Absorption (TADA) was an open-label crossover study in 46 healthy male volunteers, comparing the absorption of APO-dabigatran (150 mg) with vs without rabeprazole. The primary outcome was the 24-hour total dabigatran exposure as measured by area under the curve (AUC) and peak concentration (C_max).

Results

Compared with no rabeprazole pretreatment, the total dabigatran AUC (geometric mean [gmean] AUC_0-tz: 567.2 vs 804 ngh/mL, and gmean AUC_0-∞: 609.7 vs 804) and C_max (gmean: 64.1 vs 104.4 ng/mL) were significantly reduced with rabeprazole. The percent gmean ratios for AUC_0-tz, AUC_0-∞, and C_max (with rabeprazole vs without) were 70.5% (95% confidence interval [CI]: 51.9% to 95.7%), 71.8% (95% CI: 53.1% to 96.9%), and 61.4% (95% CI: 44.1% to 85.5%), respectively. With rabeprazole, the proportions of participants with > 50% reduction in AUC_0-tz, AUC_0-∞, and C_max were 32.6%, 30.4%, and 39.1%, respectively

Conclusions

When APO-dabigatran is administered with rabeprazole, the exposure to dabigatran is reduced by about 30%, which is similar to the level observed with Pradaxa when it was co-administered with a proton pump inhibitor. However, the finding that one-third of participants had a > 50% reduction in exposure is concerning, and it highlights the need for caution in patients who have, or are at risk of, reduced gastric acidity.

The results of multiple randomized controlled trials indicate that dabigatran etexilate, a direct oral anticoagulant (DOAC), is effective and safe for stroke prevention in patients with atrial fibrillation and for the prevention and treatment of venous thromboembolism.^1,2 These findings led to the rapid approval of dabigatran etexilate and its widespread use in over 100 countries worldwide.

The patent for dabigatran etexilate expired in Canada in 2018, and soon thereafter Health Canada approved a generic formulation marketed by Apotex (APO-dabigatran; Toronto, ON) based on the demonstration of bioequivalence to the original drug.^3,4 The original drug, Pradaxa, consists of a capsule dabigatran-coated tartaric acid pellets.^5,6 An acidic environment is essential for the dissolution of dabigatran etexilate and the absorption of dabigatran, which is a small, highly polar molecule that is otherwise poorly absorbed. This specialized dabigatran etexilate formulation ensures consistent absorption of dabigatran, despite a bioavailability of 6%-7%, including in patients taking proton pump inhibitors (PPIs).^7,8 Unlike the original drug, APO-dabigatran is formulated as a capsule containing fumaric acid. APO-dabigatran was approved after it was shown to be bioequivalent to dabigatran etexilate when tested in healthy volunteers, but it was not tested in the elderly or other patients at risk of low gastric acidity.³ This factor might be important because a reduction in bioavailability could substantially affect efficacy.

In this crossover study, we explore the possible effect of gastric acid suppression in healthy volunteers taking APO-dabigatran with vs without a PPI. We measured blood concentration of dabigatran in participants treated with a single dose of APO-dabigatran and repeated blood concentration measurement during the crossover period, following a second dose of APO-dabigatran given after 5 days of rabeprazole, a potent PPI. We chose to use a PPI to suppress gastric acid secretion, thereby simulating a reduced gut-acid environment, as found in older patients and those treated with gastric acid suppressants.

Methods

The Treatment With APO-Dabigatran Absorption (TADA) study was an open-label crossover study comparing the absorption of dabigatran (150 mg) in the presence vs absence of concomitant rabeprazole treatment (20 mg once daily for 5-7 days). The study (ClinicalTrials.gov identifier: NCT04157881) was approved by the Hamilton Integrated Ethics Board and was conducted in compliance with the protocol of the International Conference on Harmonization—Good Clinical Practice (ICH-GCP[R2]), and all applicable regulatory requirements. All participants provided written informed consent.

Participants

We recruited healthy male volunteers aged between 20 and 40 years, with a body mass index of 18-30 kg/m². We excluded subjects with a history of heart, lung, liver, kidney, gastrointestinal, genitourinary, musculoskeletal, or endocrine disease, or other systemic illness, as well as those taking regular medications or herbal supplements/remedies. We also excluded the following subjects: those with routine laboratory values outside of the reference range; those who reported smoking or excessive alcohol consumption; those with a pulse rate > 90 beats per minute, blood pressure > 140 mm Hg, systolic or > 90 mm Hg, diastolic; those who were not expected to comply with the protocol requirements or complete the trial as scheduled; and those enrolled in another investigational device or drug trial within the preceding 30 days.

Study procedures

Following a fast of at least 8 hours, participants took dabigatran 150 mg with 200 mL of water and underwent serial measurement of blood concentration of dabigatran over a 24-hour period. They subsequently underwent a 5- to 7-day washout period, during which they took rabeprazole 20 mg once daily. To ensure maximal proton pump inhibition, participants were instructed to take rabeprazole each morning and prior to APO-dabigatran administration. On the final day of rabeprazole ingestion (at least day 5), participants took a further dose of dabigatran 150 mg with 200 mL of water, after which they again underwent serial measurements of blood concentration of dabigatran over a 24-hour period.

Blood sampling and analysis

We collected blood prior to the dose of dabigatran (baseline) and then at 30, 60, and 90 minutes, and 2, 3, 4, 6, 8, and 24 hours post-dabigatran. At each timepoint, we collected blood into a 1 × 6 mL ethylenediamine tetraacetic acid tube and 1 × 4.5 mL citrate tube, as well as into 2 × 2.7 mL citrate tubes. After centrifugation at 2500 g for 10 minutes at 4°C to 8°C, we separated plasma for measurement of activated partial thromboplastin time (aPTT), dilute thrombin time (dTT), and dabigatran concentration by calibrated dTT (Hemoclot thrombin inhibitor assay; Hyphen BioMed, Neuville-sur-Oise, France), and stored additional aliquots of ethylenediamine tetraacetic acid plasma at -80°C for further testing. In addition, measurement of total dabigatran concentration in plasma samples was performed by Nuvisan GmbH, Neu Ulm, Germany, using validated high-performance liquid chromatography tandem mass spectrometry (LC-MS/MS).

The primary outcome was the 24-hour total dabigatran exposure as measured by area under the curve (AUC) and peak concentration (C_max). The secondary outcomes were as follows: proportion of participants with > 50% reduction in 24-hour total dabigatran exposure; AUC and C_max, by Hemoclot dabigatran assay; and maximum dTT and aPTT.

Statistical analysis

We assumed that a 50% reduction in C_max or AUC would be clinically relevant based on the following factors: (i) the observation that contraindicated co-medications that are potent p-glycoprotein inducers generally result in a > 50% reduction in C_max or AUC; and (ii) lesser reductions (< 50%) have not been shown to be clinically relevant with dabigatran (eg, pantoprazole co-administration did not alter the efficacy of Pradaxa).^1,7,9 Therefore, a sample size of 46 subjects was determined to be large enough to ensure that a 2-sided test with alpha = 0.05 has 90% power to detect a 50% reduction in C_max or AUC.

We analyzed pharmacokinetic and coagulation variables using a noncompartmental analysis. We calculated ratios (test vs reference treatment) for pharmacokinetic endpoints in all participants who provided pharmacokinetic measurement for both test and reference study treatments. We determined point estimates, geometric means (gmeans) and gmean ratios of natural log-transformed data, together with their 2-sided 95% confidence intervals (CIs), by repeated measures analysis of variance. We also calculated the intra-individual percent changes in total dabigatran AUC and C_max across the 2 periods, which are graphically presented in Waterfall plots. The proportions of participants (and 95% CI) with a > 50% reduction in AUC and C_max were estimated. To examine the agreement between test and reference methods (dabigatran concentration by Hemoclot and LC-MS/MS, respectively), we calculated the mean difference and 95% limit of agreement using the Bland-Altman method.

Safety

We documented all adverse events, including the frequency, severity, and causal relationships from the time of first study drug administration to 3 days after the last study drug treatment.

Results

Participant demographics

From January 2020 to August 2021, a total of 46 healthy male volunteers successfully completed the 2 treatment periods in this crossover study (Fig. 1). Baseline characteristics are presented in Table 1. The mean age of participants was 24.4 years, the mean body mass index was 23.7 kg/m², and the mean creatinine level was 81.2 μmol/L.

Figure 1.

Study flow. Apo-dabigatran, a generic formulation of dabigatran etexilate, marketed by Apotex (Toronto, ON); PK, pharmacokinetic.

Table 1.

Baseline characteristics (n = 46)

Characteristic	Value
Age, y	24.4 (4.3)
Height, cm	180.2 (6.9)
Weight, kg	77.0 (10.8)
Body mass index, kg/m2	23.7 (2.7)
Smoking, % current	21 (45.7)
Systolic BP, mm Hg	125.2 (7.2)
Diastolic BP, mm Hg	74.3 (8.1)
Heart rate, bpm	68.5 (12.0)
Creatinine, μmol/L	81.2 (13.2)
Haemoglobin, g/L	151.3 (10.3)
Platelet count, × 109/L	237.8 (48.9)
Prothrombin time, s (INR)	1.1 (0.1)
Activated partial thromboplastin time, s	32.1 (2.6)
Thrombin clotting time, s (dTT)	26.5 (1.9)
Fibrinogen, g/L	2.8 (0.6)

Values are mean (standard deviation), unless otherwise indicated.

BP, blood pressure; bpm, beats per minute; dTT, dilute thrombin time; INR, international normalized ratio.

Pharmacokinetic analyses of APO-dabigatran in the absence vs presence of rabeprazole

Figure 2 compares the plasma concentration time profile of total dabigatran (as measured by LC-MS/MS) with vs without rabeprazole pretreatment during the 24-hour sampling period. Compared with no rabeprazole pretreatment, the total dabigatran AUC (gmean AUC_0-tz [area under the concentration-time curve from dose to last timepoint with measurable drug concentration]: 567.2 vs 804 ngh/mL and gmean AUC_0-∞ [area under the concentration-time curve from dose up to infinity with extrapolation of the terminal phase]: 609.7 vs 804) and C_max (gmean: 64.1 vs 104.4 ng/mL) were significantly reduced with rabeprazole pretreatment (Table 2). The geometric mean ratios for AUC_0-tz, AUC_0-∞, and C_max (calculated as the pharmacokinetic value with rabeprazole divided by the corresponding value without rabeprazole times 100) were 70.5% (95% CI: 51.9% to 95.7%), 71.8% (95% CI: 53.1% to 96.9%) and 61.4% (95% CI: 44.1% to 85.5%), respectively. Conversely, the time from dose to maximal concentration (time to maximum concentration [t_max]; gmean: 2.4 vs 1.9 hours; ratio: 125.8%; 95% CI: 111.4% to 142.1%) and half life (t_1/2; gmean: 5.0 vs 4.3 hours; ratio: 115.2%; 95% CI: 106.9% to 124.3%) were significantly prolonged.

Figure 2.

Dabigatran plasma concentration as measured by liquid chromatography tandem mass spectrometry in the presence vs absence of rabeprazole. Data are presented as mean ± standard error; n = 46. APO-Dabi, Apo-dabigatran, a generic formulation of dabigatran etexilate, marketed by Apotex (Toronto, ON).

Table 2.

Pharmacokinetic parameters of APO-dabigatran, a generic formulation of dabigatran etexilate, marketed by Apotex (Toronto, ON) in the absence vs presence of rabeprazole

	APO-dabigatran		APO-dabigatran + rabeprazole		gmean ratio,∗ %	95% CI
	gmean	gCV, %	gmean	gCV, %		Lower	Upper
Total dabigatran by LC-MS/MS
AUC0-tz, ngh/mL	804.5	11.2	567.2	13.6	70.5	51.9	95.7
AUC0-∞, ng h/mL	849.4	11.1	609.7	13.2	71.8	53.1	96.9
Cmax, ng/mL	104.4	17.3	64.1	23.0	61.4	44.1	85.5
tmax, h	1.9	52.5	2.4	41.5	125.8	111.4	142.1
t1/2, h	4.3	14.3	5.0	15.4	115.2	106.9	124.3
Dabigatran concentration by Hemoclot assay (Hyphen BioMed)†
AUC0-tz, ng h/mL	741.3	9.3	511.5	11.5	69.0	53.6	88.9
AUC0-∞, ng h/mL	828.2	8.3	611.8	9.7	73.9	59.2	92.1
Cmax,ng/mL	93.4	16.3	54.3	23.4	58.1	41.8	80.8
tmax, h	2.0	49.5	2.4	40.4	120.6	106.9	135.9
t1/2, h	4.5	28.5	5.8	33.9	129.3	106.8	156.6

aPTT, activated partial thromboplastin time; AUC, area under the curve; AUC_0-tz, Area under the concentration-time curve from dose to last timepoint with measurable drug concentration; AUC_0-∞, Area under the concentration-time curve from dose up to infinity with extrapolation of the terminal phase; CI, confidence interval; C_max, maximum plasma concentration; dTT, dilute thrombin time; gCV, geometric coefficient of variation; gmean, geometric mean; LC-MS/MS, high-performance liquid chromatography tandem mass spectrometry; t_max, time from dose to maximal concentration; tz, time from dose to last timepoint with measurable drug concentration; t_1/2, terminal half-life.

^∗Calculated as APO-dabigatran with rabeprazole treatment gmean divided by APO-dabigatran alone gmean times 100.

^†Dabigatran concentrations are calculated from a calculated concentration curve from dilute thrombin time.

Similar pharmacokinetic differences were found when the dabigatran concentration was measured using the Hemoclot assay (Table 1). Compared with no rabeprazole, the dabigatran AUC (gmean AUC_0-tz: 511.5 vs 741.3 ngh/mL and gmean AUC_0-∞: 611.8 vs 828.2) and C_max (gmean: 54.3 vs 93.4 ng/mL) were significantly reduced in the presence of rabeprazole (Table 2). The geometric mean ratios for AUC_0-tz, AUC_0-∞, and C_max were 69.0 % (95% CI: 53.6% to 88.9%), 73.9% (95% CI: 59.2% to 92.1%), and 58.1% (95% CI: 41.8% to 80.8%), respectively. Conversely, the t_max (gmean: 2.4 vs 2.0 hours; ratio: 120.6%; 95% CI: 106.9% to 135.9%) and t_1/2 (gmean: 5.8 vs 4.5 hours; ratio: 129.3%; 95% CI: 106.8% to 156.6%) were significantly prolonged.

Pharmacodynamic analyses of APO-dabigatran in the absence vs presence of rabeprazole

Changes in clotting times are presented in Table 3. Compared with no rabeprazole, both the peak dilute thrombin time (dTT gmean: 37.5 vs 40.6 seconds; ratio: 92.3%; 95% CI: 87.2% to 97.6%) and the activated partial thromboplastin time (aPTT gmean: 43.5 vs 49.4 seconds; ratio: 88.1%; 95% CI: 82.5% to 94.0%) were shorter when APO-dabigatran was administered with rabeprazole pretreatment. The dTT t_max (gmean: 2.5 vs 2.2 hours; ratio: 117.6%; 95% CI: 102.1% to 135.4%) and the dTT t_1/2 (gmean: 53.2 vs 48.4 hours; ratio: 110.8%; 95% CI: 82.2% to 149.4%) were prolonged by rabeprazole pretreatment. Similarly, the aPTT t_max (gmean: 2.4 vs 1.9 hours; ratio: 126.8%; 95% CI: 110.6% to 145.4%) and aPTT t_1/2 (gmean: 59.2 vs 41.0 hours; ratio: 146.0%; 95% CI: 110.9% to 192.1%) were prolonged by rabeprazole pretreatment.

Table 3.

Pharmacodynamic properties of coagulation testing of APO-dabigatran, a generic formulation of etexilate dabigatran, marketed by Apotex (Toronto, ON)

	APO-dabigatran		APO-dabigatran + rabeprazole		gmean ratio,∗ %	95% CI
	gmean	gCV, %	gmean	gCV, %		Lower	Upper
Dilute thrombin time
Cmax, s	40.6	3.8	37.5	4.3	92.3	87.2	97.6
tmax, h	2.2	53.4	2.5	41.9	117.6	102.1	135.4
t1/2, h	48.4	20.3	53.2	17.8	110.8	82.2	149.4
aPTT
Cmax [sec]	49.4	5.3	43.5	4.9	88.1	82.5	94.0
tmax, h	1.9	56.6	2.4	47.5	126.8	110.6	145.4
t1/2, h	41.0	16.3	59.2	20.5	146.0	110.9	192.1

CI, confidence interval; gCV, geometric coefficient of variation; gmean, geometric mean; AUC, area under the curve; C_max, maximum plasma concentration; t_max, time from dose to maximal concentration; t_1/2, terminal half-life; aPTT, activated partial thromboplastin time; dTT, dilute thrombin time.

^∗Calculated as APO-dabigatran with rabeprazole treatment gmean divided by APO-dabigatran alone gmean times 100.

Intra-individual changes in dabigatran exposure after raboprazole

Figure 3, A and B show the percent changes in AUC_0-tz and C_max, respectively, for each participant after rabeprazole pretreatment. With rabeprazole pretreatment, the proportions of participants with a > 50% reduction in AUC_0-tz, and C_max were 32.6% (95% CI: 20.8% to 47.1%), and 39.1% (95% CI: 26.4% to 53.8%), respectively.

Figure 3.

Waterfall plots detailing percent changes in (A) dabigatran area under the curve (AUC_0-tz and (B) maximum plasma concentration (C_max) for each participant after rabeprazole pretreatment. AUC_0-tz, area under the concentration-time curve from dose to last timepoint with measurable drug concentration; ID, identification.

Correlation between dabigatran concentration by LC-MS/MS and Hemoclot

Figure 4 shows that dabigatran concentration as measured by the Hemoclot assay correlated strongly with total dabigatran concentration as measured by LC-MS/MS (correlation coefficient r = 0.970; 95% CI: 0.965 to 0.973). The mean difference between the 2 methods was 7.5 ng/ml (95% limits of agreement: -23.8 to 38.7).

Figure 4.

Correlation between the measurement of dabigatran plasma concentration using liquid chromatography tandem mass spectrometry (LC-MS/MS) and diluted thrombin time. Both are given in ng/mL concentrations. All individual data are shown. APO-Dabi, Apo-dabigatran, generic formulation of dabigatran etexilate Hemoclot thrombin inhibitor assay (Hyphen BioMed), marketed by Apotex (Toronto, ON); CI, confidence interval; L, lower; U, upper.

Correlations between total dabigatran by LC-MS/MS and clotting times

The dTT showed a strong positive correlation (r = 0.902; 95% CI: 0.889 to 0.913) with total dabigatran concentration (Fig. 5), as did the aPTT (r = 0.859; 95% CI: 0.841 to 0.875)

Figure 5.

Correlation between measurement of dabigatran plasma concentration using liquid chromatography tandem mass spectrometry (LC-MS/MS) and (A) dilute thrombin time (dTT) and (B) activated partial thromboplastin time (aPTT). All individual data are shown. APO-Dabi, Apo-dabigatran, generic formulation of dabigatran etexilate, marketed by Apotex (Toronto, ON); CI, confidence interval.

Safety

Treatment with APO-dabigatran tablet alone, or with rabeprazole pretreatment, was well tolerated. No deaths and no serious adverse events occurred. No adverse events were reported in the 8 participants who discontinued the study. Of the 46 participants who completed the study, mild adverse events were reported in 3 participants. One participant reported mild nausea, determined not to be related to study procedure or drugs, that resolved without intervention. Two participants had a mild vasovagal reaction following venipuncture, which resolved without intervention.

Discussion

APO-dabigatran has been approved in Canada, based on studies demonstrating bioequivalence, but its bioavailability has not been evaluated previously in the setting of reduced gastric acidity, a prevalent condition in older patients. Because the solubility of dabigatran etexilate is pH-dependent, the bioavailability of generic formulations may be compromised in settings of reduced acidity unless the acidifier (fumaric acid for APO-dabigatran) maintains an acidic microenvironment that optimizes the dissolution of APO-dabigatran. To investigate the bioavailability of APO-dabigatran in the setting of reduced gastric acidity, we performed a crossover study whereby healthy volunteers received APO-dabigatran with vs without pretreatment with rabeprazole. The results of our study indicate that with rabeprazole pretreatment, the bioavailability of APO-dabigatran is reduced by about 30%-40%, and that the time to maximum concentration and half-life of the APO-dabigatran are prolonged by 15%-25%. Important to note is that about one-third of participants taking APO-dabigatran had a > 50% reduction in drug exposure after rabeprazole pretreatment.

The reduction in AUC with APO-dabigatran after PPI pretreatment of about 30% is consistent with that observed in secondary analyses of clinical trials and pharmacokinetic studies of Pradaxa (Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany) when it was co-administered with pantoprazole or other PPIs.^7,8,10,11 Reassuringly, for Pradaxa, in 2 clinical outcome trials, Randomized Evaluation of Long-term Anticoagulation Therapy (RE-LY) and Randomized Evaluation of Dual Antithrombotic Therapy with Dabigatran vs Triple Therapy With Warfarin in Patients With Nonvalvular Atrial Fibrillation Undergoing Percutaneous Coronary Intervention (RE-DUAL PCI), concomitant PPI treatment did not appear to reduce its efficacy.^1,9 Although the average reduction in bioavailability with APO-dabigatran after rabeprazole was similar in magnitude to that reported for Pradaxa, the finding that about one-third of participants had a > 50% reduction in AUC after rabeprazole pretreatment is concerning because such a reduction in bioavailability might lower drug efficacy. In the absence of clinical outcome studies with APO-dabigatran, our findings highlight the need for caution in patients who have, or are at risk of, reduced gastric acidity, and they support the need for additional studies such as ours or under fed conditions when trying to establish bioequivalence of complex pharmaceuticals.³

Our results are likely to be valid because of the following factors: careful timing and completeness of drug concentration sampling; consistency in our findings when dabigatran concentration was measured by LC-MS/MS and the Hemoclot assay; consistency in both pharmacokinetic and pharmacodynamic findings; and the use of 5 days of rabeprazole to effectively reduce gastric pH.¹² Potential limitations of our study are the absence of intragastric pH sampling to confirm reduced gastric acidity with PPI treatment, and the enrollment of healthy male volunteers in whom the pharmacokinetic profile of APO-dabigatran may differ from that of older patients. Nonetheless, our study provides important data on the effect of low gastric pH on the bioavailability of APO-dabigatran.

Conclusion

Our data indicate that when APO-dabigatran is administered with rabeprazole, the exposure to dabigatran is reduced by about 30%, which is similar to that observed with Pradaxa when it was co-administered with a PPI. However, the finding that one-third of participants had a > 50% reduction in exposure is concerning and highlights the need for caution in patients who have, or are at risk of, reduced gastric acidity.