Use of dabigatran and rivaroxaban in non-valvular atrial fibrillation: one-year follow-up experience in an Italian centre

Abstract

Background

Direct oral anticoagulants (DOAC) have been shown to be non-inferior to traditional vitamin K antagonists in preventing stroke and arterial thromboembolism in patients with non-valvular atrial fibrillation. Nevertheless, it is mandatory to record side effects and individual adherence to DOAC treatment.

Materials and methods

In this single-centre experience, patients with non-valvular atrial fibrillation were prospectively observed after switching from a vitamin K antagonist to dabigatran or rivaroxaban. The efficacy, safety, and tolerability of the novel treatment, and adherence to it, were evaluated over a period of 1 year. Clinical data were integrated with records of haemorrhagic and non-haemorrhagic complications. All the subjects were given an anonymous self-report questionnaire on the degree of their adherence/satisfaction with the treatment.

Results

Of 196 patients with non-valvular atrial fibrillation (median age, 78.5 years) who switched from a vitamin K antagonist to DOAC, 178 completed the 1-year follow up, of whom 87 were given dabigatran and 91 rivaroxaban. The efficacy of the two DOAC was similar. Patients given dabigatran had a higher frequency (n=32) of non-haemorrhagic complications (OR: 3.3; 95% CI: 1.7–7.8), which occurred earlier (HR: 6.1; 95% CI: 3.0–12.6) than those (n=7) recorded in subjects on rivaroxaban. The degree of satisfaction with therapy was higher among patients on rivaroxaban (mean score 9.1, SD 1.0) than among those on dabigatran (mean score 8.7; SD 0.9; p=0.01).

Discussion

Overall, in this experience, DOAC were shown to be effective, safe alternatives to vitamin K antagonists. Nevertheless, compared with rivaroxaban, dabigatran resulted in a higher rate and earlier occurrence of non-haemorrhagic events, and a lower satisfaction score.

Introduction

The prevalence of non-valvular atrial fibrillation (NVAF) is increasing worldwide. The incidence of stroke in patients with NVAF is calculated to be around 3–4%¹, although the percentage increases significantly (20%) in patients over 80 years old². Anticoagulant therapy represents the cornerstone of prophylaxis, being recommended for patients at high risk of thromboembolic events, in order to reduce stroke and/or systemic arterial embolism. However, anticoagulant treatment has several drawbacks in elderly people and the beneficial effect of vitamin K antagonists (VKA) must be weighed against the risk of major bleeding³. Moreover, physicians are often afraid to use oral anticoagulation in elderly people and prefer antiplatelet agents when patients cannot be monitored or controls are not feasible. These issues encouraged the development, over the last 10 years, of novel direct oral anticoagulants (DOAC), which may not require laboratory monitoring. DOAC have been shown to have non-inferior or superior efficacy compared to VKA, and be associated with a lower incidence of major bleeding, particularly intracranial haemorrhage, as demonstrated in well-known controlled clinical trials^4–7. The ability to inactivate bound serine proteases makes DOAC stronger than warfarin and heparins. In addition, compared to warfarin, DOAC have a favourable risk-benefit balance for preventing stroke and systemic arterial thromboembolism. Caution is recommended with prescribing dabigatran to elderly patients because this DOAC, at a dose of 150 mg twice daily, has been associated with a higher risk of major bleeding. Coagulation monitoring is not required, but patients should be followed up periodically with the aim of detecting new health states that could change the expected effectiveness or safety of the drugs⁸.

The Italian Medicines Agency (Agenzia Italiana del Farmaco - AIFA) has so far licensed three DOAC, which have been commercially available since 2013–2014: dabigatran etexilate (Pradaxa^® [Boehringer Ingelheim, Rhein, Germany] 150/110 mg bid), rivaroxaban (Xarelto^® [Bayer Pharma AG, Berlin, Germany] 20/15 mg od) and apixaban (Eliquis^® [Bristol-Myers Squibb, New York, USA] 5/2.5 mg bid). Prescription of these drugs was allowed only by some specialists, specifically cardiologists, internists, neurologists, geriatricians and haematologists. To determine eligibility for prescription of DOAC, the AIFA requires answers (yes/no) to a questionnaire collecting some personal information and data on a potential patient’s health status in order to evaluate thromboembolic and haemorrhagic risk profiles on the basis of the CHA2DS2-VASc and HAS-BLED scores, respectively (age, gender, congestive cardiac failure, hypertension, diabetes mellitus, abnormal liver and renal function, prior stroke/transient ischaemic attack or peripheral arterial thromboembolism, bleeding, labile time-in-therapeutic range with VKA in the preceding 6 months [≤70% for dabigatran and ≤60% for rivaroxaban]) and the presence of conditions making it objectively difficult for the patients to undergo laboratory controls (no detectable venous access, great distance from the surveillance centre, chronic confinement to bed and the impossibility of patients performing periodic, long-term laboratory monitoring). A positive response to this last item mandatorily makes the patient eligible. The patient must have a CHA2DS2-VASc score ≥1 to be prescribed dabigatran, whereas a score >3 is required for rivaroxaban; a HAS-BLED score >3 is necessary for the prescription of either DOAC. No specific indications on clinical and laboratory monitoring are mentioned.

In consideration of the importance of therapy with the new oral anticoagulants, post-marketing surveillance of the clinical benefits and disadvantages of DOAC is desirable in order to assess the clinical impact of these drugs in “real life”. Indeed, despite the availability of antidotes to DOAC and growing experience with the management of bleeding, this is the most feared complication and the one that is hardest to deal with¹⁰.

Here we report data from a 1-year follow up in a clinical setting of a cohort of 196 NVAF patients switched from treatment with a VKA (warfarin or acenocoumarin) to a DOAC (dabigatran or rivaroxaban) in an Italian centre for monitoring anticoagulant therapy.

Materials and methods

Patients

A cohort of 196 Caucasian outpatients (median age: 78.5 years, range: 50–94), consisting of 112 women (57.1%; median age 80 years, range 52–94) and 84 men (42.8%; median age 77 years, range 50–91) with different types of NVAF and on treatment with VKA were proposed for switching from VKA to a DOAC and followed up prospectively from June 2013 to December 2014. Because of the later authorisation for marketing in Italy of apixaban, patients who were prescribed this DOAC were not considered in this study.

After approval by the local Ethics Committee, the study was carried out according to the principles of the Declaration of Helsinki; informed consent was obtained from all participants.

Patients were enlisted according to AIFA eligibility criteria for prescription of a DOAC (www.agenziafarmaco.com). The majority of patients accepted a switch to a DOAC because of an unsatisfactory time-in-therapeutic-range (<60%). A complete clinical summary and information on the patients’ habits, diet, co-medication, illnesses, bleeding complications and scheduled surgical or invasive procedures were obtained from all participants by specially trained staff.

Patients receiving dabigatran

Dabigatran was given to 93/196 (47.4%) patients: 49 women (52.6%; median age 80 years, range 58–87) and 44 men (47.3%; median age 76 years, range 56–89). The majority of subjects received dabigatran 110 mg bid (79/93: 84.8%). The higher dose of dabigatran (150 mg bid) was reserved to younger patients (14/93: 15.05%, aged <75 years) with a lower risk of bleeding and normal renal function (as assessed using the Cockcroft-Gault equation).

Patients receiving rivaroxaban

Of the 196 patients who entered the study, 103 (52.5%) were given rivaroxaban: 63 women (61.1%; median age 79 years, range 52–94) and 40 men (38.8%; median age 79 years, range 50–91). Rivaroxaban 20 mg od was given to 92/103 patients (89.4%) while only 11/103 individuals (10.6%) took a dose of 15 mg od because of a calculated creatinine clearance <30 mL/min.

Definition of bleeding events

All bleeding events were classified as major, non-major clinically relevant (NMCRB), or minor using the International Society on Thrombosis and Haemostasis (ISTH) definitions¹¹. Major bleeding was defined as overt bleeding with any of the following: documented transfusion of at least 2 units of red blood cells; a drop in haemoglobin of 2 g/L or more; surgical revision because of bleeding; bleeding into a critical site (intracranial, intra-ocular, intra-articular, retroperitoneal, overt gastrointestinal bleeding) or fatal bleeding.

NMCRB was defined as non-major bleeding compromising haemodynamics; any bleeding leading to hospitalisation; a subcutaneous haematoma larger than 25 cm², or 100 cm² if there was a traumatic cause; an intramuscular haematoma documented by ultrasonography; epistaxis that lasted for more than 5 minutes, was repetitive (i.e. two or more episodes of bleeding more extensive than spots and a handkerchief within 24 hours), or led to an intervention (i.e. packing or electrocoagulation); gingival bleeding occurring spontaneously (i.e. unrelated to eating or tooth brushing) or lasting for more than 5 minutes; haematuria that was macroscopic and was spontaneous or lasted for more than 24 hours after instrumentation (i.e. catheter placement or surgery) of the urogenital tract; macroscopic gastrointestinal haemorrhage, including at least one episode of rectal blood loss, if more than a few spots on toilet paper; haemoptysis, if more than a few speckles in the sputum and not occurring within the context of pulmonary embolism; or any other type of bleeding considered to have clinical consequences for a patient such as medical intervention, the need for unscheduled contact (visit or telephone call) with a physician, temporary cessation of a study drug, association with pain or impairment of activities of daily life.

Minor bleeding was defined as any overt bleeding event that did not fulfil the criteria for major bleeding or NMCRB.

Clinical follow up and testing

All patients enrolled underwent clinical examination and laboratory evaluation of haemoglobin concentration, haematocrit, hepatic and renal function. Creatinine clearance was evaluated using the Cockroft-Gault formula. After enrolment, patients had study visits scheduled at 1 month and at 3-monthly intervals thereafter. The clinical examination was integrated by records of any major bleeding, NMCRB or minor haemorrhagic complications or side effects. In addition, data on temporary discontinuation of anticoagulant therapy were requested. Each subject was given an anonymous self-report questionnaire on their degree of adherence/satisfaction with the treatment, determined on a visual analogue scale (VAS) from 0 to 10 (Figure 1)¹².

Figure 1.

VAS questionnaire on adherence/satisfaction with DOAC therapy.

VAS: visual analogue scale; DOAC:Direct oral anticoagulants.

Statistics

All the analyses were performed using the Statistical Package for Social Sciences (SPSS 11.0 for Windows). The significance of any differences in means was evaluated by a non-parametric test, whereas the significance of any differences in proportions was tested by chi-square statistics. Odds ratio (OR) and 95% confidence intervals (CI) were calculated. The 95% CI of a proportion was calculated using the modified Wald or exact method as required. Adjusted OR and their 95% CI were calculated by logistic-regression models that controlled for age at initiation of oral anticoagulation, sex, and co-medication. Event-free survival curves were calculated using the Kaplan-Meier method and comparison of survival functions was based on log-rank testing. Adjusted hazards ratios (HR) and their 95% CI were calculated by Cox proportional hazards models that controlled for age at initiation of oral anticoagulation, sex, and co-medication. Overall incidence rates were measured from the date of enrolment and patients’ survival was censored at the occurrence of adverse events or when lost to follow-up. In all cases, p values <0.05 are considered statistically significant.

Results

Patients on dabigatran

Of 93 patients who were switched to dabigatran, 87 (93.6%) (41 men and 46 women) completed the 1-year follow up. Of these, 77 (88.5%) were given dabigatran 110 mg bid and 10 (11.4%) the higher dose (150 mg bid).

Six patients (6.4%) did not complete the follow up and discontinued therapy. Major bleeding (gastrointestinal) occurred in a 78-year old woman on dabigatran 110 mg bid. A NMCRB (abundant, recurrent epistaxis that required hospitalisation with anterior and posterior nasal packing) associated with severe thrombocytopenia (platelet count: 8×10⁹/L) was observed in a 76-year old man on dabigatran 110 mg bid. The remaining four patients (1 patient on 150 mg and 3 on 110 mg bid) discontinued anticoagulant therapy with dabigatran because of persistent stomach pain and heartburn. All the patients asked to return to their previous oral anticoagulant therapy with a VKA either because of burdensome symptoms or the fear of further complications due to the novel therapy, despite the contrary opinion of clinicians.

Among the patients who completed the 1-year follow up, 21 (29.6%) had milder side effects (n=16; 76.2%) and NMCRB or minor bleeding (n=10; 47.6%), which did not need discontinuation of therapy (Table I). Of these 21 patients, 15 (71.4%) complained of gastrocolic dyspepsia soon after switching to dabigatran; two were taking 150 mg (both men) and 13 were taking 110 mg (7 women and 6 men). These patients did not stop taking the drug and their symptoms resolved spontaneously after about 3–4 weeks. Six patients suffered from NMCRB, one man on 150 mg and five people on 110 mg bid (3 women and 2 men), whereas four had minor bleeding, one woman on 150 mg and three people on 110 mg (2 women and 1 man). No transient ischaemic attacks, or cerebral or peripheral thromboembolic accidents were observed.

Table I.

Adverse events and bleeding complications in patients on dabigatran and rivaroxaban during a 1-year follow up.

	DOAC
	Dabigatran	Total	Rivaroxaban	Total
Adverse event (n)	Severe thrombocytopenia (8×109/L)* (1)	21	Elevated transaminases (1)	1
	Epigastric pain and heartburn (4)
	Transient elevated gammaglutamyl ranspeptidase (1)
	Gastro-colic dyspepsia (15)
Major bleeding (n)	Gastrointestinal bleeding (1)	1	Metrorrhagia (1)	2
			Rectal bleeding (1)
Non-major clinically relevant bleeding (n)	Haemoptysis (2)	7	Haemoptysis (1)	2
	Anal bleeding (1)		Macroscopic haematuria (1)
	Recurrent epistaxis (1)*
	Macroscopic haematuria (2)
	Gengivorrhagia (1)
Minor bleeding (n)	Mild epistaxis (2)	4	Mild epistaxis (1)	2
	Small subcutaneous haematomas (1)		Mild thrombocytopenia (80×109/L) (1)
	Subconjunctival haemorrhage (1)
Total (n)		32		7

^(*)Episodes occurred in the same patient;

DOAC:Direct oral anticoagulants.

Patients on rivaroxaban

Of the 103 patients who were prescribed rivaroxaban, 91 (88.4%; 56 women and 35 men) completed the 1-year follow up. Of these 91 patients, 67 (73.6%) received 20 mg od and 24 (26.4%) were given 15 mg od.

Of the initial 103 patients, three (2.9%), all on rivaroxaban 20 mg od, discontinued treatment. They withdrew from the study because of major bleeding (metrorrhagia in an 85-year old woman and rectal haemorrhage in a 75-year old man), and a more than 10-fold increase of transaminases in a 75-year old man. All the patients asked to return to their previous oral anticoagulant therapy with warfarin.

Of the 91 patients who completed the 1-year follow up, four subjects (5.4%), all on 20 mg od, had NMCRB (2 women) or minor bleeding (1 man and 1 woman), which did not require discontinuation of therapy (Table II). No cerebral or peripheral arterial thromboembolic events were observed.

Table II.

Main characteristics of the patients anticoagulated with DOAC.

	DOAC
	Dabigatran (n=93)	Rivaroxaban (n=103)	All (n=196)
Age median (range)	77 (56–89)	79 (50–94)	78.5 (50–94)
Women, n (%)	49 (52.6)	63 (61.1)	112 (57.1)
Dosage, n (%)	110 mg bid: 79 (84.9)	15 mg od: 92 (89.3)
	150 mg bid: 14 (14.1)	20 mg od: 11 (10.7)
Missed patients, n (%)	0	9 (8.7)	9 (4.5)
Major bleeding, n (%)	1 (1.1)*	2 (1.9)*	3 (1.5)*
Non-major clinically relevant bleeding, n (%)	11 (11.8)**	4 (3.9)	15 (7.7)
Minor bleeding, n (%)	4 (4.3)	2 (1.9)	6 (3.0)
Non-bleeding side-effects, n (%)	20 (21.5)°	1 (1.0)§	21 (10.7)
Complications
without discontinuation of treatment, n (%)	26 (28.0)	4 (3.9)	30 (15.3)
with discontinuation of treatment, n (%)	6 (6.4)	3 (2.9)
Patients who completed the follow up, n (%)	87 (93.5)	91 (88.3)	178 (90.8)

^*All discontinued therapy;

^**1 patient discontinued dabigatran;

^°4 patients discontinued dabigatran;

^§1 patient discontinued rivaroxaban;

DOAC:Direct oral anticoagulants.

Multivariate analysis

The possibility that the occurrence of adverse effects might be different between groups was investigated in a binary logistic regression model, adjusted for age at initiation of oral anticoagulation, and sex. A significant difference was found (p=0.001) between patients taking dabigatran, who showed higher frequencies of major bleeding and adverse events (n=32) than those (n=7) in subjects who received rivaroxaban (OR: 3.3; 95% CI: 1.7–7.8). Kaplan-Meier analysis showed that non-haemorrhagic complications occurred earlier in patients on dabigatran (log-rank test: p<0.001) (Figure 2). A Cox proportional hazards model that controlled the same variables was performed to address the possibility of a different time-to-event between the groups of patients. The analysis confirmed that events occurred earlier among patients anticoagulated with dabigatran (HR: 6.1; 95% CI: 3.0–12.6) (p=0.05).

Figure 2.

Cumulative event-free survival in DOAC anticoagulated patients.

Kaplan-Meier curve, showing cumulative event-free survival in DOAC anticoagulated patients taking dabigatran (solid line) or rivaroxaban (dotted line).

DOAC: direct oral anticoagulants.

Missing follow-up data

Overall nine of the 196 patients (4.5%) missed a periodic 3-month follow-up. These patients were part of the group of 103 (8.7%) patients taking rivaroxaban. None of the patients missed a follow-up because of clinical problems. Most of them preferred to be followed up by their own general practitioners. The main characteristics of the patients anticoagulated with the two DOAC are summarised in Table II.

Adherence/satisfaction with treatment

A total of 133 out of 178 patients who completed the 1-year follow up (74.4%, median age 78 years, range 50–94) answered the VAS questionnaire: 65 of 87 (74.7%) patients on dabigatran (median age 76 years, range 56–93) and 68 of 91 (74.7%) on rivaroxaban (median age 79 years, range 50–94). All patients showed good adherence to the new oral therapy and scored from a minimum of 7 to a maximum of 10 (Figure 3). Patients on rivaroxaban assigned a higher score (mean 9.1, SD 1.0) than patients on dabigatran (mean 8.7; SD 0.9; Mann-Whitney U-test p=0.01).

Figure 3.

VAS adherence/satisfaction scores in patients on dabigatran or rivaroxaban.

The numbers of patients with each score are reported above the columns. VAS: visual analogue scale.

Discussion

DOAC represent an undisputed innovation as long-term oral anticoagulant therapy in patients with NVAF^13,14. They offer many advantages over VKA, making anticoagulation a more acceptable treatment, particularly in elderly people. In the “real life” setting, the most practical advantage is the use of a fixed dose that does not need laboratory monitoring. Nevertheless, other advantages must be taken in account, such as minimal drug–drug and food interactions, a predictable pharmacokinetic profile, the wide therapeutic windows, and a lower risk of intracranial haemorrhage. Attention must be maintained when using the highest dose of dabigatran (150 mg bid) since this has been associated with a similar risk of intracranial haemorrhage as that seen with VKA. These considerations make the use of DOAC in “real life” somewhat different from that in clinical trials.

Our report, concerning 196 patients with NVAF switched from a VKA to a DOAC (dabigatran or rivaroxaban), presents the 1-year experience of post-marketing surveillance of adherence and satisfaction with DOAC in an Italian centre, as well as an evaluation of the drugs’ efficacy and safety. Both dabigatran and rivaroxaban were found to be effective and safe in the majority of patients treated, including elderly people aged 80 years and over.

In accordance with published data, various adverse events were recorded¹⁵. More cases of NMCRB were reported in patients on dabigatran than in patients on rivaroxaban; the difference was not, however, statistically significant. In addition, although almost all the patients on dabigatran were taking the lower dosage of the drug, non-haemorrhagic clinical complications occurred significantly more frequently and earlier among patients taking this DOAC than among those taking rivaroxaban. The most common side effects were gastrointestinal disorders, which needed discontinuation of therapy in four patients.

A good level of adherence/satisfaction with therapy was observed for both drugs, although the score was higher among patients taking rivaroxaban. One case of severe thrombocytopenia that developed after the introduction of dabigatran and disappeared after discontinuation of the drug was considered dabigatran-induced. As reported in the literature, thrombocytopenia is a rare complication of DOAC use and is associated with an increased risk of life-threatening bleeding, so attention must be maintained, particularly in elderly people¹⁶. Of 103 patients to whom rivaroxaban was prescribed, nine (8.7%) missed a periodic followup without a detectable cause. The relatively high percentage of patients missing follow-ups highlights the risk of “self-treated patients” and the need to interact closely with all patients and their general practitioners in order to pick up or prevent clinical complications.

Conclusions

The recent introduction of DOAC in the anticoagulant treatment of NVAF provides a useful alternative to the traditional standard therapy with VKA. We assessed the efficacy and safety of switching from warfarin to DOAC in “real life”, based on data from a 1-year follow-up (2013–2014) of patients affected by NVAF who switched from a traditional VKA to dabigatran or rivaroxaban. New oral anticoagulants have demonstrated their efficacy as alternatives to VKA. Although these agents do not require routine coagulation monitoring, assays to assess the level of anticoagulation may be of assistance in certain circumstances, such as in the case of overdose, as may be measurements of creatinine clearance in elderly patients in order to avoid haemorrhagic or thromboembolic events during treatment.