Evaluation of oral anticoagulants in atrial fibrillation patients over 80 years of age with nonsevere frailty

Abstract Background The safety and efficacy of an oral anticoagulant (OAC) treatment and the difference between direct OACs (DOACs) and warfarin in nonsevere frail elderly patients with AF are unclear. Methods This was a retrospective and observational study of 354 patients over 80 years of age with nonsevere frailty who were diagnosed with AF and treated with OACs. Nonsevere frailty was defined as a clinical frailty scale score of <7. Bleeding and thromboembolic events during the OAC treatment were followed up. Results Of 354 patients enrolled, 273 (77.1%) received DOACs and 81 (22.9%) received warfarin. Of 273 patients receiving DOACs, there were 210 (76.9%) prescribed with appropriate doses of DOACs. Of 81 warfarin‐treated patients, 53 (65.4%) were prescribed an appropriate dose of warfarin. During a follow‐up of 33.1 (14.0‐51.0) months, 15 patients (1.5/100 person‐years) had bleeding events and 10 (1.0/100 person‐years) had thromboembolic events while on an OAC treatment. The incidence ratio of bleeding events in patients receiving DOACs was lower than that in those receiving warfarin (1.0/100 person‐years vs 2.9/100 person‐years, hazard ratio [HR]: 0.26, 95% confidence interval [CI]: 0.07‐0.91, P = .036). There was no significant difference in the incidence of thromboembolic events between the DOAC and warfarin treatment groups (0.88/100 person‐years vs 1.4/100 person‐years, HR: 0.63, 95% CI: 0.16‐2.57, P = .52). Conclusions OACs are substantially safe and effective for preventing thromboembolic events in nonsevere frail patients over 80 years of age. Particularly, DOACs can be used more safely than warfarin.


| INTRODUC TI ON
Atrial fibrillation (AF) is the most common type of arrhythmia in clinical practice. It is estimated that AF is associated with approximately 30% of ischemic strokes in patients over 80 years of age. 1 The elderly population is growing worldwide, and is expected to result in a more than four-fold burden of AF among patients over 80 years of age by the year 2050, representing more than half of all cases of AF. 2 The evaluation of frailty is important to help assess the suitability of an oral anticoagulant (OAC) treatment for preventing thromboembolic events in AF patients. Frailty has been found to be associated with poor clinical outcomes related to medical management. 3 Following this, in severe frail elderly patients with AF, it may be appropriate to use no anticoagulation to avoid bleeding events. 4 A previous study reported that patients classified as being nonsevere frail patients were 3.5 times more likely to receive OACs than the severe frail patients. 5 However, there are limited data about the safety and efficacy of an OAC treatment in octogenarians whose frailty is not so severe.
The objective of this study was to evaluate the safety and efficacy of an OAC treatment and the difference between direct OACs (DOACs) and warfarin for the management of AF among nonsevere frail patients over 80 years of age.

| Study population
This was a retrospective and observational study of patients over 80 years of age with nonsevere frailty who were diagnosed with AF under regular outpatient care or under admission and were newly treated with the DOACs (dabigatran, rivaroxaban, apixaban or edoxaban) or warfarin to prevent thromboembolic events between January 2011 and August 2017. All patients were treated at our institution. The patient observation began at the start of the use of the OACs. The start date of the OACs was considered to be the date the prescription was dispensed. Patients were categorized into DOAC or warfarin groups based on their administered OACs. We also included the patients treated with antiplatelet drugs in combination with OACs.

| Assessment of frailty
Frailty was assessed with the Clinical Frailty Scale (CFS) of the Canadian Study on Health & Aging, which has been verified as a useful rapid assessment tool of frailty. 6,7 The CFS is a measure of frailty based on a clinical judgement that takes into account cognition, mobility, function, and co-morbidities, with scores ranging from 1 (very fit) to 9 (terminally ill). 8 Each patient was attributed a CFS score by each physician when AF was diagnosed.
Nonsevere frailty was defined as a CFS score of <7 in the present study.

| Study outcomes
The primary safety outcome of the study was represented by the incidence rate of bleeding events composed of major bleeding (MB) and clinically relevant non-major bleeding (CRNMB). The primary efficacy outcome included the incidence of thromboembolic events composed of ischemic strokes and systemic embolisms (SEs). Because of the retrospective characteristic of this study, the incidence ratio of bleeding or thromboembolic events did not directly reflect the safety or efficacy of each OAC alone. To monitor the adverse effects due to OACs, interviewing and examining the patients or obtaining usual blood tests, were performed every 1-3 months during the OAC treatment. Additional testing was performed as necessary when a bleeding or thromboembolic event was suspected. MB was defined as bleeding requiring hospitalization, bleeding requiring a transfusion of at least 2 units of packed red cells, or bleeding occurring at a critical site during the use of OACs.
CRNMB was defined as bleeding not meeting the criteria for major bleeding, but requiring medical intervention, unscheduled contact with a physician, or temporary cessation of the OAC treatment.
Ischemic strokes were defined as a loss of neurological function of a sudden onset lasting ≥24 hours. SEs were defined as thromboembolisms outside the brain. Patients were followed until their first bleeding or thromboembolic event, discontinuation of the treatment, a treatment switch to a different OAC, patient death, or the end of the study period (August, 2018).

| Statistical analysis
Data analyses were performed using EZR on R-commander version 1.24 software (Saitama Medical Center, Jichi Medical University). All continuous variables were tested for the normality of the distribution using the Kolmogorov-Smirnov test.
Continuous variables with a normal distribution were described as the mean ± standard deviation (SD), continuous variables with a skewed distribution were described as the median (quartile: 25%-75%), and categorical variables were described as frequencies and percentages. Comparisons between groups were analyzed by univariate logistic analysis (Fisher's exact test, Unpaired t test, or Mann-Whitney test) and multivariate analysis using a logistic regression analysis. The incidence ratio was calculated using the person-year method (events per 100 person-years).
The relationship of the OAC treatment and the incidence ratio of bleeding or thromboembolic events was analyzed using the Kaplan-Meier method, and the curves were compared using a log-rank test. A multivariate analysis using a Cox proportional hazard model was constructed to assess the development of bleeding and thromboembolic events. These models were adjusted by the age, gender, HAS-BLED score, BW, CrCl, usage of antiplatelet drugs, and dosages of OACs for bleeding events; and adjusted by the age, hypertension, diabetes mellitus, past history of an ischemic stroke, and CHADS 2 score for thromboembolic events. In all tests, a P-value of .05 was considered as the cut-off for statistical significance.

| Ethical consideration
This study was performed in accordance with the Code of Federal Regulations and the Declaration of Helsinki. The present study was approved by the Toho University Omori Medical Center Ethical Committee (number: M17139), and informed consent was obtained from each patient before the study.

| Patient characteristics
Among 440 consecutive patients aged 80 years and over with nonsevere frailty who were diagnosed with AF, we retrospectively analyzed 354 patients who were treated with DOACs (dabigatran, rivaroxaban, apixaban or edoxaban) or warfarin treatment at our institution between January 2011 and August 2017. The reasons why 86 patients were not treated with OACs were no justification provided (23.3%), a history of bleeding (16.3%), patient refusal (16.3%), severe renal dysfunction (14.0%), active bleeding (11.6%), or a poor patient condition (18.6%). The study flow chart is shown in Figure 1. The mean age was 83.8 ± 3.6 years, and 48.0% were male. The mean BW was 52.1 ± 11.0 kg and mean body mass index (BMI) 21.6 ± 3.9 kg/  Table 1.    Table 3. person-years) experienced thromboembolic events. The incidence ratio of thromboembolic events did not differ significantly among each of the four DOACs (P = .053). Table 4 shows the incidence of thromboembolic events in each OAC subgroup, that is, according to the prescribed OACs and dosages.  Note: Data are expressed as the mean ± SD, or number (%). Abbreviations: BW, body weight; CrCl, creatinine clearance. P < .05 was considered as significant.

| DOAC treatment
The patients receiving DOACs were divided into three groups, ap- did not differ among the three treatment groups, respectively.

| Warfarin treatment
The patients receiving warfarin were divided into two groups, an appropriate dose of warfarin group (n = 53) and an inappropriate

| Main findings
Our data suggested that OACs are substantially safe and effective for reducing the risk of AF-associated thromboembolic events in nonsevere frail patients over 80 years of age. Particularly, DOACs can be used more safely than warfarin.

| Assessment of the safety and efficacy outcome
During the follow-up period, there were few bleeding events and rare thromboembolic events. The present study had a non-inferior safety and efficacy outcome while on the OAC treatment compared to the real-world studies that included patients older than 65 years. 11,12 Our data revealed that the OAC treatment was safe and effective for the management of AF in octogenarians if their frailty was not so severe.

| Comparison of the safety and efficacy outcome between DOACs and warfarin
While it is likely that the DOACs will eventually replace warfarin for the management of AF in clinical practice, the role of DOACs in elderly patients remains to be fully defined. Phase III trials revealed that DOACs were at least as safe and effective as dose-adjusted warfarin, 13 however, elderly and frail patients were represented to a lesser extent. Recently, several real-world studies have confirmed the safety and efficacy of DOACs in the elderly, although the frailty was not likely to be mentioned. 14,15 In the present study, bleeding events were significantly less frequent in the DOAC group than with the warfarin group, and the frequency of thromboembolic events did not differ significantly between the two groups. When comparing the DOAC and warfarin groups, the CrCl was significantly lower in the warfarin group than the DOAC group (37.2 ± 15.6 vs 46.9 ± 15.7, P < .01), and the usage of antiplatelet drugs was significantly more frequent in the warfarin group than the DOAC group

| Antiplatelet therapy in combination with OACs
Biologic plausibility suggests that antiplatelet therapy in combination with OACs could lead to worse bleeding outcomes. However, a multi- interval; DOACs, direct oral anticoagulants. *Adjusted by the age, hypertension, diabetes mellitus, past history of an ischemic stroke, and CHADS 2 score. The rate did not differ significantly between the two groups during antiplatelet therapy were made by the each attending physician. Antiplatelet drug administration tended to be more likely to be stopped early because of the fear of bleeding. As a result, the antiplatelet therapy may not have been associated with the development of the bleeding. Further research is necessary to seek the optimal antiplatelet therapy in combination with OACs in this population.

| Assessment of the prescription of inappropriate doses of OACs
Whether the dose recommendations of DOACs are adhered to in clinical practice remains a major concern, especially among elderly patients. In this study, approximately 30% of patients received inappropriate doses of DOACs according to the approved Japanese recommendations. As shown in a previous study, in which the majority of the patients on inappropriate doses of DOACs were found to be on a lower dose than recommended, 17,18 we also found that under-treatment was more frequent than an over-treatment in this setting. This is probably due to the fear of bleeding complications.
Moreover, in the group with DOACs alone, given that an antidote was available only for dabigatran, 19 prescribing lower doses makes clinicians more confident of a safer prescription. On the one hand, an inappropriate dose of warfarin was also often prescribed in the clinical practice even though an antidote was available for warfarin.

| Safety and efficacy outcome according to the prescribed OACs and dosages
The incidence of bleeding and thromboembolic events did not dif-

| Limitations
This study had some potential limitations. First, a single measure of frailty was used, so patients may have been misclassified based on other definitions of frailty. Second, this study was a retrospective and observational study conducted at a single center. The selections of DOACs or warfarin were not randomized as the decisions were made by the attending physicians. There were variations in the indication for DOACs or warfarin. The observation period also differed according to the OAC treatment regimen. Third, this study had a small number of patients, which might have resulted in a statistical bias. Especially, there were only 28 patients taking edoxaban when our survey was conducted (January 2011 to August 2017), because of the limited spread within the market during that early period after it obtained approval. Fourth, the follow-up duration was relatively short. Further research is necessary with more patients and a long-term follow-up.

| CON CLUS IONS
OACs are substantially safe and effective for reducing the risk of AF-associated thromboembolic events in nonsevere frail patients over 80 years of age. Particularly, DOACs can be used more safely than warfarin.

ACK N OWLED G EM ENTS
This manuscript was supported in part by Grants-in-Aid (24591074 and 15K09103 to TI) for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.