Association of Apixaban Therapy and Prothrombin Time in Patients With Atrial Fibrillation

– Single Center Cohort Study –

Abstract

Background: This study evaluated whether measuring prothrombin time (PT) using particular reagents of interest predicted apixaban-associated anticoagulant activity in Japanese patients with non-valvular atrial fibrillation (NVAF).

Methods and Results: Two reagents, Shinplastin Excel S and Coagpia PT-N, were used to evaluate PT under apixaban therapy. From June 2013 to February 2014, 103 NVAF patients were recruited, and PT was measured at 3 time points: (1) anytime in the outpatient clinic, (2) at peak, and (3) at trough. In spike-in experiments using pooled citrated normal human platelet-poor plasma with these PT reagents, apixaban prolonged PT values in a concentration-dependent manner. PT values significantly correlated between both reagents (r=0.97) in outpatients. PT values in outpatients taking 5-mg apixaban bid were significantly prolonged and had wide inter- and intraindividual variability. Peak values were significantly higher than trough values, with both values higher than normal. The dose change of apixaban from 5 mg bid to 2.5 mg bid in outpatients halved the degree of PT prolongation in each NVAF patient.

Conclusions: The PT value measured by these specific reagents can predict apixaban-associated anticoagulant activity, although there is significant interpatient variability. (Circ J 2014; 78: 2651–2656)

Apixaban, a non-vitamin K antagonist oral anticoagulant (NOAC), was approved in Japan in March 2013 for the prevention of stroke in patients with non-valvular atrial fibrillation (NVAF).^1,2 Apixaban does not require frequent monitoring of coagulation activity or dose adjustment because its pharmacokinetics and pharmacodynamics are predictable.^3,4 In addition, it has been reported that most conventional coagulation assays, such as prothrombin time (PT) or activated partial thromboplastin time (aPTT), are not sensitive enough to monitor apixaban plasma concentrations because of several limitations.^5,6 However, in the clinical setting, including the recurrence of stroke, bleeding, and urgent surgery, it would be important to understand apixaban-associated anticoagulant activity. One study reported that the measurement of chromogenic Xa activity is useful for estimating apixaban plasma concentration.⁶ However, because that is a technically demanding and time-consuming procedure, it would be convenient to have a simpler way to estimate apixaban’s plasma concentration in daily clinical practice.

In this regard, a series of studies showed that PT was prolonged by the use of rivaroxaban, another Xa inhibitor available in Japan, and that the absolute PT value was associated with plasma Xa activity and could be useful for approximately predicting the anticoagulant activity of rivaroxaban.^7–9 In contrast, Suzuki et al reported that the PT value in patients taking rivaroxaban had wide interindividual and peak-trough variability, depending on the sampling time point after intake.¹⁰

Apixaban is administered as a 5-mg tablet twice daily.¹ The recommended dose of apixaban is 2.5 mg bid in NVAF patients with at least 2 of the following characteristics: >80 years of age, body weight <60 kg, or serum creatinine level >1.5 mg/dl.

Although taking apixaban 5 mg bid has been reported to prolong PT when using commercially available PT reagents and that the degree of prolongation is concentration-dependent,⁶ PT values measured using most PT reagents were within the normal range, and it was concluded that these reagents were not sensitive enough to predict apixaban-associated Xa activity. However, one of the tested reagent, Triniclot PT Excel S^® (trade name in Japan: Shinplastin Excel S^® [Kyowa Medex Co, Tokyo, Japan]), showed PT prolongation by 1.1–2-fold with the use of apixaban 50–500 ng/ml in the normal pooled-plasma preparation.⁶

As the important issue during daily clinical practice is to avoid adverse hemorrhagic events with the use of NOACs, including apixaban, in Japanese senile and frail patients with NVAF, measuring PT values using the aforementioned reagent might be viable for monitoring NOAC-associated anticoagulant activity. In this regard, we recently found in a preliminary study that another PT reagent, Coagpia PT-N^®, which has been used in Yamaguchi Grand Medical Center, also detects significant PT prolongation with the use of apixaban 5 mg bid.

Therefore, this study comprised 2 investigations: a comparison of the effect of apixaban on PT prolongation using Triniclot PT Excel S^® and Coagpia PT-N^®, and second, an evaluation of the distribution of PT and determination of the clinical usefulness of using a specific reagent of interest to measure PT in Japanese patients with NVAF under apixaban therapy.

Methods

Reagents for the Measurement of PT

Two coagulation assay reagents for the measurement of PT were used in this study, one of which was previously reported to be relatively sensitive for apixaban plasma concentration compared with other reagents (Shinplastin Excel S; normal range: 12.6–15.7 s, Kyowa Medex Co, Tokyo, Japan),⁶ and the other has been used by use in daily clinical practice (reagent: Coagpia PT-N; analyzer: Coapresta 2000^®, normal range: 10.0–13.0 s, Sekisui Medical Co, Tokyo, Japan). In a preliminary study, apixaban 5 mg bid prolonged the PT value by 1.5–1.8-fold using both reagents, therefore, we first compared the effects of apixaban on PT prolongation using the 2 reagents in an in vitro citrated normal pooled-plasma spike-in experiment and in 50 samples obtained from outpatients. As both reagents could similarly detect PT prolongation above the normal range with the use of apixaban, we mainly used Coagpia PT-N^® reagent for the rest of the PT measurement analysis under apixaban therapy.

Study Population

From June 2013 to February 2014, 103 NVAF patients were recruited to this study after giving written informed consent to participate. We recorded their clinical characteristics and the distribution of PT using PT reagents.

Coagulation Assay

(1) To obtain the standard curve between apixaban plasma concentration and the PT value, PT was spiked at increasing concentrations of apixaban (0, 50, 100, 200, 300, 400, 500 ng/ml) in pooled citrated normal human platelet-poor plasma. The standard curve was drawn twice.

(2) PT values of blood samples taken from patients at the outpatient clinic obtained using 2 reagents (Shinplastin Excel S and Coagpia PT-N) were compared (5 mg bid: n=46, 2.5 mg bid: n=4).

(3) (a) PT values were measured by Coagpia PT-N in the blood collected anytime at the outpatient clinic, approximately 2–3 h after apixaban intake (5 mg bid: n=40, 2.5 mg bid: n=20).

　　(b) Peak and trough PT values were measured 3 h after drug intake and just before drug intake during steady hospitalization, respectively (5 mg bid: n=17).

　　(c) In the outpatient clinic, PT was sequentially measured in NVAF patients who met at least 2 of the following clinical characteristics: >80 years of age, body weight <60 kg, serum creatinine level >1.5 mg/dl; these patients started apixaban therapy at 5 mg bid and were later changed to 2.5 mg bid, according to the manufacturers’ recommendation (n=5).

Exclusion Criteria

Patients who used anticoagulation therapy (vitamin K antagonist, dabigatran, and rivaroxaban) within 30 days were excluded, as were patients with creatinine clearance (CCr) <25 ml/min. Estimated CCr was calculated using the Cockcroft-Gault equation.^11,12

Statistical Analysis

Statistical analysis was performed using the STATVIEW program (Abacus Concepts, Berkeley, CA, USA). Values are presented as the mean ± standard deviation. A chi-square test was used to compare the characteristics of patients taking 2 different doses of apixaban. Pearson product-moment correlation coefficient was used to correlate the distribution of PT values between the PT reagents. Overall statistical significance among the serial measurement of PT in AF patients was analyzed by repeated measures ANOVA, followed by an individual comparison by paired t-test as appropriate. P<0.05 was considered statistically significant.

Results

Patients Characteristics

Patient characteristics are listed in Table. A standard dose of apixaban (5 mg bid) was administered to 75 patients, and reduced dose (2.5 mg bid) was administered to 28 patients. In 5 patients, the dose was reduced from 5 mg bid to 2.5 mg bid because they met at least 2 of the clinical characteristics for the reduced dose protocol during the follow-up period. There were no episodes of cerebral or gastrointestinal bleeding during the follow-up period under apixaban therapy.

Table. Clinical Characteristics of the Patients With Regard to Apixaban

	Total (n=103)	5 mg twice daily (n=75)	2.5 mg twice daily (n=28)	P value
Clinical characteristics
Age, years	73.8±9.9	70.0±8.3	83.7±6.2	<0.001
Female sex, n (%)	49 (48)	33 (44)	16 (57)	0.234
Weight, kg	56.2±10.6	60.0±10.2	49.1±7.5	0.003
Creatinine clearance, n (%)
Normal, >80 ml/min	23 (22)	22 (29)	1 (4)	0.005
Mild impairment, >50–80 ml/min	47 (46)	40 (53)	7 (61)	0.010
Moderate impairment (>30–50 ml/min)	30 (29)	13 (17)	17 (61)	<0.001
Severe impairment (≤30 ml/min)	3 (3)	0	3 (11)	0.004
Prior myocardial infarction, n (%)	7 (7)	5 (6)	2 (7)	0.932
Prior clinically relevant or spontaneous bleeding, n (%)	4 (4)	2 (3)	2 (7)	0.295
Qualifying risk factors, n (%)
Age ≥75 years	47 (46)	21 (28)	28 (100)	<0.001
Prior stroke, TIA, or systemic embolism	12 (12)	9 (12)	3 (11)	0.856
Heart failure or reduced left ventricular ejection fraction	28 (27)	18 (24)	10 (34)	0.234
Diabetes	21 (20)	17 (23)	4 (14)	0.348
Hypertension requiring treatment	59 (57)	42 (56)	17 (61)	0.667
CHADS2 score
Mean	1.75±1.18	1.55±1.17	2.25±1.06	0.025
Distribution, n (%)
≥2	56 (54)	34 (45)	22 (79)	0.003
1	33 (32)	27 (36)	6 (21)	0.159
0	14 (14)	14 (19)	0 (0)	0.003
Use of PPI or H2-blocker	38 (37)	24 (32)	14 (50)	0.092
Criteria for reduced dose (2.5 mg twice daily), n (%)
Age ≥80 years	–	–	28 (100)	–
Weight ≤60 kg	–	–	28 (100)	–
Serum creatinine ≥1.5 mg/dl	–	–	0	–
Thromboembolic events, n (%)
Stroke, TIA, or systemic embolism	0	0	0	–
Side effects, n (%)
Hemorrhagic stroke	0	0	0	–
Epigastric symptoms	3 (3)	2 (3)	1 (4)	0.808
Major bleeding (extracranial)	0	0	0	–
Minor bleeding
Epistaxis	6 (6)	4 (5)	2 (7)	0.727
Subcutaneous hemorrhage	1 (1)	1 (1)	0	0.539
Follow-up, days	153±122	175±121	93±106	<0.001

PPI, proton-pump inhibitor; TIA, transient ischemic attack.

Concentration-Dependent Prolongation of PT

Apixaban prolonged the PT value in a concentration-dependent manner (Figure 1). In addition, the slopes of the PT prolongation curve with both reagents were almost linear for the apixaban concentration range of 0–500 ng/ml.

Figure 1.

Effect of an increasing concentration of apixaban on coagulation time using 2 reagents: Shinplastin Excel S and Coagpia PT-N. Triniclot PT Excel S^® (trade name in Japan: Shinplastin Excel S^®) is the more sensitive reagent for the measurement of prothrombin time (PT) under apixaban therapy. The reagent used by us in clinical practice (ie, Coagpia PT-N), revealed a similar sensitivity for PT with the use of apixaban.

Correlation of PT Values Between Reagents

In 50 patients (5 mg bid: n=46, 2.5 mg bid: n=4), the PT values were measured using both reagents and compared by Pearson product-moment method. PT values significantly correlated between reagents (r=0.97; Figure 2). The proportion of outpatients whose PT value was above the normal range was 80% and 82% when using the Coagpia PT-N^® reagent and Triniclot Excel S reagent, respectively. In addition, the variability of the PT value between reagents tended to be slightly larger as its absolute value becomes larger.

Figure 2.

Correlation between the 2 reagents for prothrombin time values.

Distribution of PT

In the outpatient clinic, PT values in patients taking a standard dose (5 mg bid: n=40) were significantly prolonged compared with those before the administration of apixaban and were widely distributed with a large interindividual variability (10.8–13.4 s vs. 12.8–19.0 s, respectively) (Figure 3A). PT values were also prolonged with the reduced dose (2.5 mg bid: n=20) compared with before the administration of apixaban and were widely distributed with a large interindividual variability (10.3 s–13.4 s vs. 12.0 s–15.7 s) (Figure 3B).

Figure 3.

Distribution of prothrombin time (PT) values before and after the administration of apixaban in the outpatient clinic where blood samples were collected approximately 2–3 h after drug intake. (A,B) Each line corresponds to an individual patient before the administration of apixaban (Before) and 1–6 months (M) after first-time administration at the outpatient clinic.

In hospitalized patients, PT values with the standard dose of apixaban (5 mg bid; n=17) at peak (15.4±1.7 s; range: 12.8–19.9 s) were significantly prolonged compared with those at trough (14.1±1.5 s; range: 11.7–18.6 s, P<0.05). In addition, both peak and trough values were significantly higher than before the administration of apixaban (12.2±0.8 s; range: 10.8–13.9 s, P<0.001) (Figure 4A). The variability in trough and peak PT values appeared to be reflecting interindividual variability, as the trough-peak response in each patient showed a similar response to administration of apixaban. The degree of PT prolongation from peak to trough (1.3±0.9 s) was significantly smaller than from peak to first-time drug intake (3.2±1.5 s, P<0.001) (Figure 4B).

Figure 4.

(A) Hospitalized patients: plot of prothrombin time (PT) values at peak and trough. The peak time is 3 h after drug intake and the trough time is just before apixaban intake during steady hospitalization. Each line corresponds to an individual patient. PT values with the standard dose of apixaban (5 mg bid; n=17) at peak (15.4±1.7 s; range: 12.8–19.9 s) were significantly prolonged compared with those at trough. (B) Degree of PT prolongation from peak to trough (peak-trough), and from peak to first-time drug intake (peak-before).

In the outpatient clinic, PT values in patients taking 5 mg apixaban bid as an initial dose (n=5, 16.7±1.2 s) were reduced by taking 2.5 mg apixaban bid (13.7±1.3 s) in patients who reached the age of 80 years during the treatment and with a body weight ≤60 kg The degree of prolongation of PT values with apixaban 2.5 mg bid showed an approximately middle value between the PT value with 5 mg bid and that before the administration of apixaban (11.4±0.8 s), indicating that PT values were dose-dependent within the same patient sampled at similar time points (Figure 5).

Figure 5.

Effect of dose change (from 5 mg bid to 2.5 mg bid in the outpatient clinic). Each line corresponds to an individual patient. Before: before first-time administration of apixaban.

Discussion

To the best of our knowledge, this is the first report to describe the distribution of PT values in Japanese NVAF patients. In the spike-in experiments using pooled citrated normal human platelet-poor plasma, apixaban prolonged PT values more than the normal range in a concentration-dependent manner. In addition, there was a good linear correlation in PT values between the 2 reagents (r=0.97). In the outpatient clinic, PT values in patients taking 5-mg apixaban bid were significantly prolonged, with a wide inter- and intraindividual variability, perhaps because of differences in individual metabolic activity and to the time of sampling after drug intake. In contrast, peak and trough PT values in hospitalized patients appeared to be well correlated. In addition, a dose change from 5 mg bid to 2.5 mg bid of apixaban approximately halved the degree of PT prolongation within this cohort of NVAF patients. These findings suggest that the PT value calculated using these reagents of interest is closely associated with apixaban anticoagulant activity in the described conditions in Japanese NVAF patients.

PT is broadly used to evaluate the effect of vitamin K antagonist therapy (VKA). The internationalized ratio of PT values is a well-accepted parameter for dose optimization of VKA therapy.^13,14 Although NOACs, including dabigatran, rivaroxaban, and apixaban, have been reported to prolong coagulation times (PT and aPTT), neither PT nor aPTT is useful in predicting non-VKA-associated anticoagulant activity, because these values showed wide variability depending on the reagents used for the coagulation assay, the patient, and sampling time points for an individual patient.^15–18 As shown in previous studies, PT values are not sensitive enough to estimate apixaban plasma concentrations, except when measured using Triniclot PT Excel S (Trinity Biotech, Bray, Ireland).^6,15 A cohort study of patients under rivaroxaban therapy in Japan also reported that the PT value had wide interindividual and peak-trough variability, depending on the sampling time point after drug intake.¹⁰

In the present study, we first compared the PT values using 2 specific coagulation assay reagents (Triniclot PT Excel S; trade name in Japan: Shinplastin Excel S^® and Coagpia PT-N), both showing a high sensitivity for an increasing concentration of apixaban spiked in the pooled citrated normal human platelet-poor plasma. The positive correlation of PT values between the reagents suggests that both can predict apixaban anticoagulant activity.

PT values in the outpatient clinic showed wide interindividual variability, which we suspect reflects several clinical factors, including renal and hepatic function, age, and body weight. In contrast, PT values at peak and trough with apixaban intake in patients with steady hospitalization showed predictable PT prolongation. In addition, a well-predicted change in PT values from 5 mg bid to 2.5 mg bid also revealed well-correlated dose-response characteristics of apixaban. The wide intraindividual variability among the outpatients may be associated with variable sampling time points after apixaban intake.

In addition, the degree of PT prolongation from peak to trough was significantly smaller than that from peak to first-time drug intake under apixaban therapy. This contrasts with rivaroxaban, another Xa inhibitor available in Japan, and may be associated with apixaban having multiple metabolic pathways, small volume distribution, a half-life of almost 12 h, and being taken twice daily.^19,20 These characteristics may be associated with the safety profile of apixaban compared with VKAs or other NOACs. Taken together, these findings suggest that PT measurement with a specific reagent is viable for predicting apixaban anticoagulant activity the PT values are measured under the described conditions.

Conclusions

In this study, we showed a distribution pattern of PT values in Japanese patients with NVAF under apixaban therapy in daily clinical practice. PT values were prolonged after administration of apixaban, and the degree of prolongation was dose-dependent. The data suggest that measuring the PT value using specific reagents can approximately predict apixaban anticoagulant activity.

Disclosures

Conflicts of Interest: None declared.

References

• 1.    Granger  CB,  Alexander  JH,  McMurray  JJ,  Lopes  RD,  Hylek  EM,  Hanna  M, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2011; 365: 981–992.
• 2.    Ogawa  S,  Shinohara  Y,  Kanmuri  K. Safety and efficacy of the oral direct factor Xa inhibitor apixaban in Japanese patients with non-valvular atrial fibrillation: The ARISTOTLE-J study. Circ J 2011; 75: 1852–1859.
• 3.    Raghavan  N,  Frost  CE,  Yu  Z,  He  K,  Zhang  H,  Humphreys  WG, et al. Apixaban metabolism and pharmacokinetics after oral administration to humans. Drug Metab Dispos 2009; 37: 74–81.
• 4.    Gong  IY,  Kim  RB. Importance of pharmacokinetic profile and variability as determinants of dose and response to dabigatran, rivaroxaban, and apixaban. Can J Cardiol 2013; 29: S24–S33.
• 5.    Frost  C,  Wang  J,  Nepal  S,  Schuster  A,  Barrett  YC,  Mosqueda-Garcia  R, et al. Apixaban, an oral, direct factor Xa inhibitor: Single dose safety, pharmacokinetics, pharmacodynamics and food effect in healthy subjects. Br J Clin Pharmacol 2013; 75: 476–487.
• 6.    Douxfils  J,  Chatelain  C,  Chatelain  B,  Dogné  JM,  Mullier  F. Impact of apixaban on routine and specific coagulation assays: A practical laboratory guide. Thromb Haemost 2013; 110: 283–294.
• 7.    Hori  M,  Matsumoto  M,  Tanahashi  N,  Momomura  S,  Uchiyama  S,  Goto  S, et al. Rivaroxaban vs. warfarin in Japanese patients with atrial fibrillation: The J-ROCKET AF study. Circ J 2012; 76: 2104–2111.
• 8.    Samama  MM,  Amiral  J,  Guinet  C,  Le Flem  L,  Seghatchian  J. Monitoring plasma levels of factor Xa inhibitors: How, why and when? Expert Rev Hematol 2013; 6: 155–164.
• 9.    Suzuki  S,  Yamashita  T,  Ohtsuka  T,  Sagara  K,  Uejima  T,  Oikawa  Y, et al. Prevalence and prognosis of patients with atrial fibrillation in Japan: A prospective cohort of Shinken Database 2004. Circ J 2008; 72: 914–920.
• 10.    Suzuki  S,  Otsuka  T,  Sagara  K,  Kano  H,  Matsuno  S,  Kato  Y, et al. Rivaroxaban in clinical practice for atrial fibrillation with special reference to prothrombin time. Circ J 2014; 78: 763–766.
• 11.    Lopes  RD,  Alexander  JH,  Al-Khatib  SM,  Ansell  J,  Diaz  R,  Easton  JD, et al. Apixaban for Reduction In STroke and Other ThromboemboLic Events in atrial fibrillation (ARISTOTLE) trial: Design and rationale. Am Heart J 2010; 59: 331–339.
• 12.    Cockcroft  DW,  Gault  MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16: 31–41.
• 13.    Inoue  H,  Okumura  K,  Atarashi  H,  Yamashita  T,  Origasa  H,  Kumagai  N, et al. Target international normalized ratio values for preventing thromboembolic and hemorrhagic events in Japanese patients with non-valvular atrial fibrillation: Results of the J-RHYTHM Registry. Circ J 2013; 77: 2264–2270.
• 14.    Lip  GY,  Lane  DA. Stroke prevention with oral anticoagulation therapy in patients with atrial fibrillation: Focus on the elderly. Circ J 2013; 77: 1380–1388.
• 15.    Gouin-Thibault  I,  Flaujac  C,  Delavenne  X,  Quenet  S,  Horellou  MH,  Laporte  S, et al. Assessment of apixaban plasma levels by laboratory tests: Suitability of three anti-Xa assays: A multicentre French GEHT study. Thromb Haemost 2014; 111: 240–248.
• 16.    Samama  MM,  Martinoli  JL,  LeFlem  L,  Guinet  C,  Plu-Bureau  G,  Depasse  F, et al. Assessment of laboratory assays to measure rivaroxaban: An oral, direct factor Xa inhibitor. Thromb Haemost 2010; 103: 815–825.
• 17.    Baglin  T,  Hillarp  A,  Tripodi  A,  Elalamy  I,  Buller  H,  Ageno  W, et al. Measuring Oral Direct Inhibitors of thrombin and factor Xa: A recommendation from the Subcommittee on Control of Anticoagulation of the Scientific and Standardisation Committee of the International Society on Thrombosis and Haemostasis. J Thromb Haemost 2013; 11: 756–760.
• 18.    Suzuki  S,  Otsuka  T,  Sagara  K,  Matsuno  S,  Funada  R,  Uejima  T, et al. Dabigatran in clinical practice for atrial fibrillation with special reference to activated partial thromboplastin time. Circ J 2012; 76: 755–757.
• 19.    Barrett  YC,  Wang  J,  Song  Y,  Pursley  J,  Wastall  P,  Wright  R, et al. A randomised assessment of the pharmacokinetic, pharmacodynamic and safety interaction between apixaban and enoxaparin in healthy subjects. Thromb Haemost 2012; 107: 916–924.
• 20.    Zhang  D,  Frost  CE,  He  K,  Rodrigues  AD,  Wang  X,  Wang  L, et al. Investigating the enteroenteric recirculation of apixaban, a factor Xa inhibitor: Administration of activated charcoal to bile duct-cannulated rats and dogs receiving an intravenous dose and use of drug transporter knockout rats. Drug Metab Dispos 2013; 41: 906–915.