Burden of upper gastrointestinal symptoms in patients prescribed dabigatran for stroke prevention

Abstract

Background:

Dabigatran, a non-vitamin K antagonist oral anticoagulant, has been shown to prevent stroke in patients with non-valvular atrial fibrillation. Nonetheless, studies show that 10%–30% of those prescribed dabigatran experience dyspepsia that may eventually lead to discontinuation of therapy and loss of clinical benefit.

Aim:

To evaluate the gastrointestinal tolerability of dabigatran utilizing a validated questionnaire, as well as determining subsequent non-compliance and drug discontinuation.

Method:

This is an observational study. All patients were assessed by a validated questionnaire, Hong Kong dyspepsia index, prior to drug prescription and again 4 weeks later.

Results:

In this study, 115 patients with non-valvular atrial fibrillation (mean age: 74.6 ± 11.4 years; mean CHA₂DS₂-VASc score was 3.39 ± 1.59) were prescribed dabigatran. At baseline, the mean Hong Kong dyspepsia index was 12.9 ± 1.6 and nine patients had significant dyspepsia (Hong Kong dyspepsia index ⩾ 16). After 4 weeks, the mean Hong Kong dyspepsia index was similar at 12.6 ± 1.9 (p = 0.23). There was no change in Hong Kong dyspepsia index after initiation of dabigatran in 59 (51.3%) patients, and improvement in 37 (32.2%). Only 19 (16.5%) patients had worsening of Hong Kong dyspepsia index, and among these 19 patients, only 1 patient (0.9%) discontinued dabigatran due to significant dyspepsia.

Conclusion:

Worsening of dyspepsia with dabigatran 110 mg twice daily was uncommon with correct drug administration and clear instructions provided. Systematic assessment of dyspeptic symptoms using a validated questionnaire (i.e. Hong Kong dyspepsia index) before and after treatment initiation allows a more objective comparison of dyspeptic symptoms.

Introduction

Atrial fibrillation (AF) confers a five-fold increased risk of ischemic stroke.^1–4 A vitamin K antagonist (VKA) oral anticoagulant (OAC) has been shown to be effective in preventing ischemic stroke in patients with AF.⁵ International guidelines recommend the use of oral anticoagulation to prevent stroke in such patients, except in those with a CHA₂DS₂-VASc score = 0.^6,7 Nevertheless, studies have shown that VKA-OACs are overtly underused in clinical practice due to the difficulty in maintaining good anticoagulation control, as well as the fear of life-threatening bleeding complications.^8–10

Dabigatran, the first commercially available non-vitamin K antagonist oral anticoagulant (NOAC), has revolutionized the management of stroke prevention in AF. In the RE-LY (Randomized Evaluation of Long-term anticoagulation therapY) study, dabigatran 110 mg twice daily was shown to be comparable to an adjusted dose of VKA in reducing stroke and systemic embolism, but associated with a lower risk of major bleeding, in particular intracranial hemorrhage.^11,12 Dabigatran 150 mg twice daily has been shown to be superior to VKA in reducing thromboembolic events, without an increased risk of major bleeding.^11,13,14 Nonetheless, in the RE-LY study, 1 in 10 patients prescribed dabigatran developed dyspepsia that was twice as prevalent compared with those prescribed VKA and led to non-compliance and drug discontinuation.¹¹

It has been reported that dyspepsia accounted for 30% of all dabigatran discontinuation in a “real-world” clinical setting,¹⁵ but there is a paucity of prospective data concerning the occurrence of upper gastrointestinal symptoms. Appropriate advice and pre-emptive measures such as advising patients to take the drug with meals may reduce the incidence of upper gastrointestinal symptoms and consequent discontinuation rate. In this study, we sought to assess the gastrointestinal tolerability of dabigatran, as well as the consequent non-compliance and drug termination in order to identify clinical parameters that could predict dabigatran intolerance.

Methods

Patients

Dabigatran-naïve patients who were ⩾18 years old, had documented non-valvular AF, a CHA₂DS₂-VAS_c score ⩾1, and preferred dabigatran to other available OACs were recruited to this study. Those who had significant valvular disease, previous valvular replacement, life expectancy <1 year, creatinine clearance or estimated glomerular filtration rate <30 mL/min, pregnancy or pregnancy plan during the trial period, or any other medical condition that rendered the patient unsuitable for dabigatran were excluded.

Study design

This was a single-center, observational study conducted at Queen Mary Hospital, Hong Kong, between 1 January 2014 and 31 December 2014 that evaluated the upper gastrointestinal symptoms upon initiation of dabigatran in patients with AF. The study complied with the Declaration of Helsinki and was approved by the locally appointed ethics committee. After enrollment, all patients were prescribed dabigatran 110 mg twice per day. They were instructed to take doses 12 h apart, with a meal, and then remain upright for at least 30 min. Demographic data including cardiovascular risk factors, medications, and prior history of gastrointestinal pathologies were recorded at baseline. Individual ischemic stroke risk was calculated at baseline using the CHA₂DS₂-VASc score (C: congestive heart failure (1 point); H: hypertension (1 point); A₂: age 65–74 years (1 point) and age ⩾ 75 years (2 points); D: diabetes mellitus (1 point); S: prior stroke or transient ischemic attack (2 points); VA: vascular disease (1 point); and Sc: sex category - female (1 point) scores as described in recent guidelines).^6,7 Symptoms of dyspepsia were quantified before and 4 weeks after initiation of dabigatran using a previously validated questionnaire, the Hong Kong dyspepsia index (HKDI).^16,17 Briefly, the index comprises questions about 12 symptoms (epigastric pain, upper abdominal bloating, upper abdominal dull ache, epigastric pain before meals, epigastric pain when anxious, vomiting, nausea, belching, acid regurgitation, heartburn, feeling of acidity in the stomach, and loss of appetite) that are graded on a five-point Likert scale: 1 (none), no symptoms; 2 (mild), symptoms can be easily ignored; 3 (moderate), awareness of symptoms but easily tolerated; 4 (severe), symptoms sufficient to cause interference with normal activity; and 5 (incapacitating), incapacitating symptoms with an inability to perform daily activities and/or require days off work. The index has been previously demonstrated to be discriminative between dyspeptic patients and controls (HKDI ⩾ 16).¹⁶ The serial change in HKDI correlated well with improvement and worsening of dyspepsia symptoms after therapeutic intervention (Kendall’s τ = 0.21, p = 0.02). The primary endpoint of this study was the change in HKDI before and 4 weeks after initiation of dabigatran. The secondary endpoint was dabigatran discontinuation within 4 weeks.

Statistical analysis

Continuous variables are expressed as mean ± standard deviation. Categorical variables are presented in frequency tables. Statistical comparisons were performed using Student’s t-test or Fisher’s exact test, as appropriate. The change in HKDI scores from immediately before to 4 weeks after initiation of dabigatran was assessed by paired-sample t-test. Calculations were performed using SPSS software (version 21.0). A p-value <0.05 was considered statistically significant.

Result

A total of 115 patients with non-valvular AF were recruited in this study. Table 1 summarizes their clinical characteristics. The mean age was 74.6 ± 11.4 years, and more than 50% of patients were ⩾ 75 years. The mean CHA₂DS₂-VASc score was 3.39 ± 1.59. Hypertension was present in 80 (69.6%) patients, diabetes mellitus in 30 patients (26.1%), and history of ischemic stroke or transient ischemic attack in 26 (14.8%). Approximately one-third of patients were prescribed warfarin and another third were on aspirin prior to initiation of dabigatran. In addition, five patients had a past history of peptic ulcer (4.3%), four patients had gastritis (3.5%), one had a gastric polyp (0.9%), and one had a gastrointestinal stromal tumor (0.9%).

Table 1.

Baseline characteristics of the entire study population.

	All (n = 115)
Age (years)
Mean age	74.6 ± 11.4
65–74	34 (29.6)
⩾ 75	62 (53.9)
Female, n (%)	55 (47.8)
Hypertension, n (%)	80 (69.6)
Diabetes mellitus, n (%)	30 (26.1)
Coronary artery disease, n (%)	13 (11.3)
Prior myocardial infarction, n (%)	6 (5.2)
Hypertrophic cardiomyopathy, n (%)	5 (4.3)
Heart failure, n (%)	26 (22.6)
Peripheral artery disease, n (%)	1 (0.9)
Stroke/transient ischemic attack (TIA), n (%)	17 (14.8)
Prior gastrointestinal bleeding, n (%)	2 (1.7)
Mean CHA2DS2-VASc	3.39 ± 1.59
Estimated glomerular filtration rate, n (%)	66.3 ± 15.2 mL/min
Previous warfarin use, n (%)	39 (33.9)
Previous aspirin use, n (%)	37 (32.2)

At baseline, the mean HKDI was 12.9 ± 1.6 with a high level ⩾16 present in nine patients (7.8%), that is, significant dyspepsia. A concomitant proton pump inhibitor was prescribed for 47 (40.9%) patients and a histamine-2 (H2) antagonist for 28 (24.3%) patients. No patient was newly commenced on either proton pump inhibitor or H2 antagonist in between these two HKDI questionnaires. After 4 weeks of dabigatran therapy, the mean HKDI was 12.6 ± 1.9, not significantly different to the baseline (p = 0.23, Figure 1). There was no change in HKDI in 59 (51.3%) patients, improvement in 37 (32.2%), and worsening of HKDI in only 19 (16.5%) patients. Of note, the number of patients with significant dyspepsia, that is, HKDI ⩾ 16, reduced from 9 (7.8%) to 4 (3.5%) after initiation of dabigatran. Only 1 patient (0.9%) who had a past history of duodenal ulcer discontinued dabigatran due to significant dyspepsia despite concomitant use of a proton pump inhibitor. During the 4-week study period, there was no gastrointestinal bleeding or other bleeding episode in our study population.

Figure 1.

Scatter plot of Hong Kong dyspepsia index at baseline and 4 weeks after initiation of dabigatran. The red dashed line dented the cutoff (⩾16) for significant dyspepsia.¹⁶

Table 2 compares the baseline characteristics between patients with and without worsening of HKDI after initiation of dabigatran. Compared with those without worsening of HKDI, patients with worsening HKDI score were more likely to be female (43.8% vs 68.4%, p = 0.05). There was no significant difference in the age, renal function, co-morbidities, baseline HKDI, past history of gastrointestinal disease, or concomitant prescription of a proton pump inhibitor or H₂ blocker between the two groups of patients (all with p ⩾ 0.05).

Table 2.

Baseline characteristics of patients with and without worsening HKDI 4 weeks after initiation of dabigatran.

	Patients without worsening HKDI (n = 96)	Patients with worsening HKDI (n = 19)	p-value
Mean age, (years)	74.1 ± 11.8	76.9 ± 8.8	0.33
Female, n (%)	42 (43.8)	13 (68.4)	0.05*
Hypertension, n (%)	63 (65.6)	17 (89.5)	0.05
Diabetes mellitus, n (%)	24 (25.0)	6 (31.6)	0.55
Prior myocardial infarction, n (%)	5 (5.2)	1 (5.3)	1.00
Hypertrophic cardiomyopathy, n (%)	4 (4.2)	1 (5.3)	1.00
Heart failure, n (%)	21 (21.9)	5 (26.3)	0.67
Stroke/transient ischemic attack (TIA), n (%)	15 (15.6)	2 (10.5)	0.73
Prior gastrointestinal bleeding, n (%)	1 (1.0)	1 (5.3)	0.30
Prior upper gastrointestinal pathology, n (%)	9 (9.4)	3 (15.8)	0.42
Baseline HKDI⩾ 16	9 (9.4)	0 (0)	0.35
Mean CHA2DS2-VASc	3.28 ± 1.63	3.95 ± 1.22	0.10
Estimated glomerular filtration rate (mL/min)	66.8 ± 15.0	63.3 ± 16.0	0.37
Previous warfarin use, n (%)	34 (35.4)	5 (26.3)	0.44
Previous aspirin use, n (%)	30 (31.3)	7 (36.8)	0.63
Concomitant proton pump inhibitor, n (%)	37 (38.5)	10 (52.6)	0.25
Concomitant H2 blocker, n (%)	24 (25.0)	4 (21.1)	0.71

Discussion

In this study, we evaluated the gastrointestinal tolerance of dabigatran in patients who were naïve to dabigatran. Most patients were prescribed either a proton pump inhibitor (40.9%) or a H2 antagonist (24.3%). The HKDI was administered to enable an objective evaluation of dyspeptic symptoms prior to and 4 weeks after the use of dabigatran. Our study showed that when systematically assessed by the HKDI, worsening of dyspepsia with dabigatran 110 mg twice daily was not common, and drug discontinuation due to gastrointestinal symptoms was infrequently encountered. Although previous studies including the RE-LY trial and real-world cohorts showed a high incidence of “dyspepsia,”^11,15,18 this phenomenon was not observed in this study. Of note, the definition of “dyspepsia” employed in these previous studies was less clear and none systematically assessed the occurrence and severity of dyspeptic symptoms before and after initiation of dabigatran therapy.^11,15,18 Patients are generally referred with the description of chronic or recurrent pain or discomfort centered in the upper abdomen,¹⁹ thus assessment of dyspepsia is largely based on patient reports, rather than physiological or laboratory parameters.¹⁶ In this study, we employed a validated questionnaire that comprised the ratings of 12 relevant dyspeptic symptoms to quantify the presence and severity of dyspepsia before and after initiation of dabigatran therapy.¹⁶ Interestingly, with the use of a validated questionnaire, we found that fewer patients reported a worsening of dyspeptic symptoms after initiation of dabigatran. This is in line with the results of a recently published Danish cohort that showed that dabigatran was not associated with increased hospitalization for dyspepsia-like diagnosis, an objective outcome measure.²⁰ Of note, only one patient (0.9%) in this study discontinued dabigatran due to significant dyspepsia despite concomitant use of a proton pump inhibitor. The previous findings of a high incidence of dyspepsia associated with the use of dabigatran may be related to the lack of systematic symptom assessment. Pre-existing dyspepsia was not recognized or documented with consequent injudicious attribution of dyspeptic symptoms to the use of dabigatran, rather than a true increase in the incidence or severity of dyspepsia with dabigatran itself.

Another explanation of our results is the focused education of patients about correct administration of dabigatran. Although there is no evidence regarding clinical management of upper gastrointestinal symptoms related to dabigatran, data from the RE-LY study have shown that taking dabigatran with two glasses of water or with food and remaining upright for at least 30 min afterwards are reasonable measures to reduce drug-related dyspepsia.²¹ The duration of our study lasted only 4 weeks. Although data from previous studies have suggested that dyspepsia is more likely to occur early after initiation of dabigatran,^20,22 it remains possible that more patients will develop worsening of dyspepsia with time. In addition, the impact of routine use of a proton pump inhibitor or H2 antagonist for patients prescribed dabigatran could not be assessed from this study, and since such practice is becoming more common, it is possible that fewer patients will experience dyspepsia. Finally, we only studied dabigatran at a lower dose of 110 mg twice daily. Previous studies have reported no increase in non-bleeding gastrointestinal adverse events when dabigatran dose was increased to 150 mg twice daily.^20,21

Limitations

First, this was a single-center study in an Asian population. Due to differences in genetic and environmental factors that may contribute to dyspeptic symptoms, our results may not be generalized to patients of the other ethnicities. Second, we studied a single dose of dabigatran, 110 mg twice daily. The gastrointestinal effects associated with dabigatran 150 mg twice daily remain uncertain. Third, our study lasted for 4 weeks. Longer follow up will be needed to identify those who develop worsening of dyspeptic symptoms late after initiation of dabigatran therapy. Finally, the prescription of a proton pump inhibitor or H2 antagonist was the individual decision of the physician and not randomized.

Conclusion

Our study showed that with correct administration, worsening of dyspepsia with dabigatran 110 mg twice daily was uncommon. Patients should be given instructions regarding administration of dabigatran at the time of treatment initiation and with regular reinforcement thereafter. Systematic assessment of dyspeptic symptoms using a validated questionnaire before and after treatment initiation enables a more objective comparison of symptoms and may prevent unnecessary drug discontinuation.