Use and effectiveness of a two-level initiation strategy for fixed-dose prothrombin complex concentrate according to the initial international normalized ratio in an emergency department in Japan.

AIM: Prothrombin complex concentrate (PCC) was recently approved for patients on warfarin therapy with international normalized ratios (INRs) exceeding 2 in Japan. However, rapid normalization of INR is necessary even in patients who do not meet the aforementioned criteria. We previously found that a fixed PCC dose of 500 IU is insufficient in some patients with INR elevation but is effective in patients with INR less than 2.5. On the basis of the results, we revised the protocol to administer a PCC dose of 500 IU to patients with INR less than 2.5 or 1,000 IU to patients with higher INRs. This study aimed to validate this revised protocol at an emergency department (ED) in Japan.
METHODS: We retrospectively collected data for all patients who received PCC in accordance with the revised protocol at our ED between October 2014 and December 2017 (period B) and compared the findings with those in the previous period (January 2013 to September 2014, period A).
RESULTS: In total, 15 and 11 patients received PCC without complications during periods A and B, respectively. All but one patient obeyed the protocol during period B. The average INRs at baseline and within 120 min after PCC infusion were 2.58 and 1.39, respectively, in period A (n = 9), versus 2.54 and 1.28, respectively, in period B (n = 8). Significantly more patients exhibited optimal responses (INR < 1.35) during period B (7/8) than during period A (3/9, P = 0.049).
CONCLUSION: Our revised protocol effectively normalized INR.

Introduction

Warfarin has a narrow therapeutic window, and it can cause a variety of hemorrhages. The annual incidence of major bleeding, which can be potentially life‐threatening, ranges from 1.1% to 1.5%. , , Because this risk of hemorrhage increases in parallel with increases in the international normalized ratio (INR), rapid reversal of long‐term oral anticoagulant therapy (OAT) is crucial in cases of acute major bleeding or emergency invasive intervention in patients with elevated INRs.

In recent clinical guidelines, including those published in Japan, prothrombin complex concentrate (PCC) has been recommended for the rapid reversal of OAT. , , , , , , , Two types of PCCs are currently available in Japan: PPSB‐HT Nichiyaku (Nihon Pharmaceutical, Tokyo, Japan) and Kcentra (CLS Behring, Tokyo, Japan); however, only Kcentra has been approved for the reversal of OAT. In addition, its use is limited to patients with INR over 2. However, rapid normalization of INR during hemostatic management is necessary even in patients with critical bleeding who do not meet the treatment criterion. , Although PCC has emerged as the preferred option in emergency settings, the optimal dosing strategy, especially in Japanese patients, remains unknown. We reported 15 cases of PCC treatment in the emergency department (ED) in 2015 and found that a fixed PCC dose of 500 IU is insufficient in some patients with INR elevation. Conversely, this dose was effective in patients with INR less than 2.5. On the basis of these findings, in October 2014, our institution implemented a two‐level fixed‐dose initiation strategy for PCC featuring an initial dose of 500 or 1,000 IU according to the initial INR (Fig. 1).

Fig. 1

Two‐level fixed‐dose initiation strategy for PCC in our institution. PCC is used at an initial dose of 500 or 1,000 IU in accordance with the initial INR in patients requiring rapid OAT reversal. INR, international normalized ratio; PCC, prothrombin complex concentrate; OAT, oral anticoagulant therapy.

The present study aimed to validate this revised protocol of PCC administration for rapidly normalizing INR and achieving control of hemorrhage at an ED in Japan.

Methods

This was a retrospective, single‐center study of consecutive patients who received OAT and/or exhibited INR elevation. Because of the hospital’s policy for off‐label use, the patients who received PPSB‐HT were specified by the attending physician. Hence, we selected all patients who received PPSB‐HT in the ED and completed a chart review study. Eligible patients were those admitted to our ED who required acute reversal of INR elevation using PCC from April 2012 to December 2017. Patients were eligible for study entry if they (i) required INR normalization as a result of acute bleeding or (ii) underwent an emergency surgical or urgent invasive diagnostic intervention. We divided these patients into two groups according to the use of a single fixed dose of PCC (period A, from January 2013 to September 2014) or a two‐level fixed dose (period B, from October 2014 to December 2017). The results evaluated in period A were reported in our previous study, and this study was performed using the same data.

As previously described, we used a commercially available four‐factor PCC formulation (PPSB‐HT). For each patient, the decision to administer PCC with vitamin K and/or fresh frozen plasma, as well as the decision regarding the administered dose, was made by the attending physician. INR and blood samples were examined for all patients on admission, and INR after PCC administration was measured within 120 min if possible. We used Wilcoxon’s signed‐rank test for analysis using STATISTICA software (StatSoft, Tulsa, OK, USA). A P‐value less than 0.05 was considered significant. The primary study endpoint was normalization of INR within 120 min after the end of the PCC infusion. Although the optimal target of INR was defined by each study, we defined INR normalization as a value of less than 1.35, in accordance with the Japanese Guidelines for the Management of Stroke (2015). In addition, hemostatic efficacy, outcomes, and thrombotic events were evaluated as described in our previous study. According to a previous report, the attending physician’s rating categories for clinical hemostatic efficacy were as follows: 1 = very good (prompt cessation of existing bleeding and/or a rapid decline in INR), 2 = satisfactory (>1–2‐h delay in bleeding cessation and a decrease in INR), 3 = questionable (>2‐h delay in the cessation of bleeding and a decrease in INR or the efficacy of PCC could not be determined), and 4 = none (no effect on bleeding and INR). The corresponding definitions among patients undergoing invasive interventional procedures were as follows: 1 = very good (normal hemostasis during the procedure), 2 = satisfactory (mildly abnormal intraprocedural hemostasis as judged by the quantity or quality of blood loss), 3 = questionable (moderate abnormality), and 4 = none (severe hemostatic abnormality).

Results

Clinical characteristics

The baseline clinical characteristics of the patients are presented in Table 1. A total of 26 patients (period A, 15; period B, 11), all of whom were Japanese, met the inclusion criteria. The male‐to‐female ratios in periods A and B were 7:8 and 6:5, respectively, and the average patient ages in these periods were 71.4 and 66.8 years, respectively. The average body weights in periods A and B were 58.0 and 52.0 kg, respectively. Body weight was less than 70 kg in all but one patient. The baseline average INRs in periods A and B were 2.20 (range, 1.04–4.14; median, 1.95) and 2.41 (range, 0.94–6.98; median 1.94), respectively. Sex, age, body weight, and baseline INR did not significantly differ between the periods.

Table 1

The baseline characteristics of 26 patients

Characteristic*	Period A n = 15	Period B n = 11	P
Gender, n (%)
Male	7 (47)	6 (55)	0.69
Female	8 (53)	5 (45)
Age (years), median (years)	71.4 (71)	66.4 (68)	0.27
Age, years, n (%)
<60	1 (7)	3 (27)
60–69	4 (27)	3 (27)
70–79	7 (47)	3 (27)
80–89	3 (20)	2 (18)
Body weight (kg), median (kg)	58 (57)	52 (53)	0.12
Body weight (kg), n (%)
<50	2 (13)	3 (27)
50–59	7 (47)	5 (45)
60–69	5 (33)	3 (27)
≥70	1 (7)	0 (0)
INR, median	2.2 (1.95)	2.41 (1.94)	0.52
INR, n (%)
<2	8 (53)	6 (55)
2–3	4 (27)	3 (27)
3–4	2 (13)	1 (9)
≥4	1 (7)	1 (9)
Indication for OAT † , n (%)
Atrial fibrillation	11 (79)	6 (55)
Heart valve replacement	2 (14)	4 (36)
Deep venous thrombosis	1 (7)	0 (0)
Without OAT † , n (%)	0 (0)	1 (9)
Medication for OAT † , ‡ , n (%)
Warfarin	13 (93)	10 (90)
Apixaban (Eliquis)	1 (7)	0 (0)

INR, international normalized ratio; OAT, oral anticoagulant therapy.

All patients were Japanese.

Total number of patients receiving OAT: 14.

Four patients also received concomitant antiplatelet agent therapy.

Indications for prothrombin complex concentrate administration

Details regarding the indications for PCC administration are listed in Table 2. Almost all patients (period A, 93.3%; period B, 81.8%) were administered PCC for acute bleeding. As noted in our previous report, there were more indications for trauma and iatrogenic events (period A, 66.7%; period B, 54.5%) than for endogenous bleeding.

Table 2

Indication for prothrombin complex concentrate administration

Indication, n (%)	Period A n = 15	Period B n = 11
Bleeding	14 (93)	9 (82)
Trauma	7 (47)	3 (27)
Intracranial hemorrhage	4(27)*	2(18) †
Soft tissue hemorrhage	1 (7)	1 (9)
Bleeding from multiple sites	2 (13) ‡	0 (0)
Endogenous	4 (27)	3 (27)
Intracranial hemorrhage	1 (7)	2 (18)
Aortic rupture	2 (13)	0 (0)
Pulmonary hemorrhage	1 (7)	1 (9)
Iatrogenic	3 (20)	3 (27)
Cardiac tamponade, cardiac injury	2 (13)	2 (18)
Retroperitoneal bleeding	1 (7)	1 (9)
Interventional procedure	1 (7)	1 (9)
Abdominal surgery	1(7) §	1 (9) §
Prophylaxis	0 (0)	1 (9)

One for acute epidural hemorrhage, two for acute subdural hemorrhage, and one for chronic subdural hemorrhage.

One case each of acute subdural hemorrhage and chronic subdural hemorrhage.

One case each of traumatic subarachnoid hemorrhage, soft tissue hemorrhage, hemothorax, retroperitoneal bleeding, intra‐abdominal bleeding, and traumatic aortic dissection.

Obstructive ileus.

Clinical course and outcomes

A summary of patient outcomes is presented in Table 3. The average doses per kilogram of body weight in periods A and B were 8.98 (range, 6.3–13.1; median, 8.77) and 10.8 IU (range, 7.6–18.9; median, 10.0), respectively. In total, 10/15 (66.7%) patients received concomitant agents in period A, and all but one patient in period B (90.9%) received concomitant agents. Overall, INR uniformly declined from 2.20 to 1.26 in period A and from 2.41 to 1.26 in period B after PCC administration. However, the timing of follow‐up blood sampling varied. Approximately 75% of the patients who received PCC achieved relatively good (very good and satisfactory) hemostatic efficacy in each period. The in‐hospital mortality rates in periods A and B were 13.3 and 9%, respectively. None of the patients had viral transmission or acute thromboembolic complications.

Table 3

Summary of 26 patients

Case	Age (sex)	OAT	Indication	PCC dose (IU)	BW (kg)	IU/kg	Concurrent medication	FFP (unit)	PC (unit)	Cryo (unit)	Pre‐INR	Post‐INR	Elapsed time	INR next day	Hemostatic efficacy	Treatment	Outcome	ICU stay (days)
1	71 (M)	WF	AEDH	500	52.2	9.6	VK	—	—	—	2.01	1.45	90 min	1.21	Very good	Conservative	Survived	1
2	76 (M)	WF	ASDH	500	64	7.8	—	—	—	—	1.71	1.14	100 min	1.08	Very good	Conservative	Survived	12
3	72 (F)	WF	ASDH	500	40	12.5	VK, FFP	2	—	—	4.14	1.59	35 min	1.28	Questionable	Operation	Survived	6
4	67 (M)	WF	CSDH	500	80	6.3	VK	—	—	—	1.46	1.17	24 h	1.17	Satisfactory	Operation	Survived	8
5	67 (F)	WF	Soft tissue hemorrhage	500	52	9.6	—	—	—	—	3.29	1.4	120 min	2.04	Very good	Conservative	Survived	4
6	67 (F)	WF	Multiple hemorrhage	500	60	8.3	VK, PC	—	10	—	1.63	1.04	120 min	1.05	Satisfactory	IVR	Survived	4
7	69 (M)	—	Multiple hemorrhage	500	65	7.7	VK, FFP, PC, Cryo	40	20	20	1.94	0.75	5.5 h	1.06	None	IVR	Died	6
8	75 (M)	WF	ICH	500	57	8.8	VK	—	—	—	2.11	1.23	30 min	0.98	Very good	Conservative	Survived	4
9	80 (F)	WF	Aortic rupture	500	52	9.6	VK	—	—	—	1.52	1.15	15 h	2.05	Questionable	Operation	Survived	60
10	84 (M)	WF	Aortic rupture	500	58	8.6	VK	—	—	—	1.95	1.21	10 h	1.21	Questionable	Operation	Survived	29
11	71 (F)	WF	Pulmonary hemorrhage	500	57	8.8	VK	—	—	—	1.88	1.11	7 h	1.28	Satisfactory	Conservative	Died	43
12	77 (F)	WF	Cardiac tamponade	500	66	7.6	FFP	2	—	—	2.37	1.41	35 min	1.41	Satisfactory	Drainage	Survived	15
13	70 (M)	EL	Cardiac tamponade	500	66	7.6	—	—	—	—	1.04	0.99	12 h	1.04	Satisfactory	Drainage	Survived	14
14	84 (F)	WF	Retroperitoneal bleeding	500	38	13.1	—	—	—	—	3.1	1.55	120 min	1.23	Satisfactory	IVR	Survived	2
15	41 (F)	WF	Abdominal surgery	500	57	8.8	—	—	—	—	2.84	1.71	20 min	1.11	Very good	Operation	Survived	7
16	87 (M)	WF	Soft tissue hemorrhage	500	56	8.9	—	—	—	—	0.97	0.87	18 h	0.87	Very good	IVR	Survived	3
17	73 (F)	WF	Cardiac injury	500	34	14.7	VK	—	—	—	1.41	1.22	120 min	1.19	Very good	Operation	Survived	8
18	62 (F)	WF	CSDH	500	50	10	—	—	—	—	1.71	1.14	90 min	1.26	Very good	Operation	Survived	14
19	56 (F)	WF	ICH	500	56	8.9	VK	—	—	—	1.52	1.19	20 min	1.51	Very good	Conservative	Survived	24
20	77 (M)	WF	ASDH	500	63	7.9	VK	—	—	—	2.1	1.3	120 min	1.19	Satisfactory	Operation	Survived	53
21	73 (F)	WF	Abdominal surgery	500	40	12.5	VK	—	—	—	1.94	1.41	30 min	1.55	Satisfactory	Operation	Survived	30
22	43 (M)	WF	Retroperitoneal bleeding	1,000	66	7.6	FFP, PC, Cryo	6	10	16	2.71	1.34	110 min	1.41	Satisfactory	Conservative	Survived	3
23	68 (M)	WF	ICH	500	60	8.3	VK	—	—	—	3.14	1.49	120 min	1.49	Questionable	Operation	Died	2
24	68 (M)	WF	Prophylaxis	1,000	53	18.86	VK, FFP	2	—	—	6.98	1.34	120 min	1.18	NA	Conservative	Survived	42
25	90 (M)	WF	Cardiac tamponade	500	44	11.36	VK, FFP	6	—	—	2.62	1.3	12 h	1.3	Very good	Drainage	Survived	90
26	33 (F)	—	Pulmonary hemorrhage	500	50	10	VK, FFP, PC, Cryo	10	10	12	1.41	1.28	120 min	1.31	Questionable	Operation	Survived	90+

AEDH, acute epidural hemorrhage; ASDH; acute subdural hemorrhage; BW, body weight; Cryo, cryoprecipitate; CSDH, chronic subdural hemorrhage; EL, Apixaban (Eliquis); FFP, fresh frozen plasma; ICH, intracranial hemorrhage; ICU, intensive care unit; INR, international normalized ratio; IVR, interventional radiology; NA, not available; OAT, oral anticoagulation therapy; PC, platelet concentration; PCC, prothrombin complex concentrate (PPSB‐HT); VK, vitamin K; WF, warfarin.

International normalized ratio values

As expected, INR dramatically declined after PCC administration (INR decreased on average from 2.29 to 1.26). However, these values varied with the timing of follow‐up blood sampling. Therefore, to simply focus on the efficacy of PCC, we examined patients who underwent follow‐up INR measurements within 120 min of infusion and excluded patients with protocol violations. Accordingly, six and two patients were excluded in periods A and B, respectively. One patient in period B was additionally excluded because the initial PCC dose was insufficient for the initial INR. In this subgroup analysis (period A, nine patients; period B, eight patients), the follow‐up INR decreased on average from 2.58 to 1.39 in period A and from 2.47 to 1.26 in period B. In addition, the proportion of patients who achieved INR less than 1.35 was significantly higher in period B than in period A (33% [3/9] versus 88% [7/8], P = 0.049; Table 4).

Table 4

Comparison of patient characteristics and INR changes (evaluated within 120 min)

Characteristic*	Period A n = 9	Period B n = 8 †	P
Gender, n (%)
Male	3 (33)	3 (38)	0.4
Female	6 (67)	5 (63)
Age (years), median (years)	70 (72)	60.1 (62)	0.25
Age, years, n (%)
<60	1 (11)	3 (38)
60–69	2 (22)	2 (25)
70–79	5 (56)	3 (38)
≥80	1 (11)
Body weight (kg), median (kg)	54.6 (57)	58 (51.5)	0.58
Body weight (kg), n (%)
<50	2 (22)	1 (13)
50–59	4 (44)	4 (50)
60–69	3 (33)	2 (25)
≥70	0 (0)	1 (13)
PCC dose (IU/kg), median (IU/kg)	9.46 (8.77)	11.3 (10.0)	0.2
Baseline INR, median	2.58 (2.37)	2.47 (1.83)	0.17
Post‐treatment INR, median	1.39 (1.41)	1.28 (1.28)	0.19
Achieved INR < 1.35, n (%)
Yes	3 (33)	7 (88)	0.049
No	6 (67)	1 (13)

INR, international normalized ratio.

All patients were Japanese.

Excluded for protocol violation.

Discussion

This retrospective study illustrated that (i) even among patients with INR less than 2.0, rapid OAT reversal is sometimes required, especially in the ED, and (ii) the potential effectiveness of our revised two‐level initiation strategy of fixed‐dose PCC based on the individual INR was supported.

Currently, PCC (Kcentra) is commercially available for the rapid reversal of OAT in Japan. According to the product labeling, the PCC dose should be individualized on the basis of the patients’ baseline INR and body weight; however, PCC was approved only for patients with INR over 2.0. In other words, PCC must be used off‐label in patients with INR 2.0 or less. Our study revealed that PCC administration for trauma as well as iatrogenic events resulted in satisfactory efficacy in clinical practice. In patients without hypofibrinogenemia (e.g., intracranial hemorrhage, cardiac tamponade, soft tissue injury) in whom mechanical hemostasis cannot be achieved, PCC may be effective even when INR is not elevated. , In fact, the joint commission stated that for hemorrhagic stroke in the setting of warfarin‐induced coagulopathy, reversal should be initiated for patients with INR over 1.4 or 1.35, favoring PCC for reversal, but guidance on dosing for PCC is lacking.

Meanwhile, the strategy for determining the optimal dose of PCC has not been reported. Although several studies attempted to determine the optimal dose of PCC using a variety of dosing regimens, no dosing strategy has proven superior. Because the recommended dose of PCC is 25–50 IU/kg, which was originally selected for the treatment of hemophilia, many previous reports used that dosing regimen. , In fact, Kcentra adopted this dosing regimen (e.g., initial INR = 2–3.9, 25 IU/kg; INR = 4–6, 35 IU/kg, INR > 6, 50 IU/kg). Although the balance between therapeutic efficacy and safety is a critical component in evaluating the use of PCC, cost cannot be overlooked. As we previously demonstrated, an extremely low fixed dose of 500 IU is likely inadequate for successful OAT reversal in patients with moderate INR elevation (initial INR > 2.5); however, these doses (25–50 IU/kg) may be excessive in some patients, thereby increasing costs and the risk of thrombogenicity. Recent fixed‐dose PCC protocols for OAT reversal that exhibited reasonable efficacy adopted a relatively higher dose (1,500–2,000 IU) than our institutional protocol. , , According to our previous results, the efficacy for normalization of INR using a fixed PCC dose of 500 IU could not be confirmed in Japanese patients. In this study, we used a fixed PCC dose of 1,000 IU in patients with initial INR over 2.5 and achieved good laboratory responses.

We experienced no complications related to PCC administration. Two major complications of PCC administration are viral transmission and thromboembolic complications. PCC use is associated with the emergence of serious thromboembolic events, especially at a high dose. Therefore, care must be exercised during PCC use, and excessive doses should be avoided if possible.

In this study, almost all patients weighed less than 70 kg. Because the response of PCC is related to the patient weight and initial INR, a higher dose (>1,000 IU/patient) should be considered in heavier patients (e.g., >80 kg) with an excessive INR, even among Japanese patients. Moreover, INR may not be the best surrogate marker for the reversal of anticoagulation at present because patients who did not display INR optimization had relatively good clinical hemostatic efficacy. We should validate efficacy using both laboratory values and thromboelastogram in future studies.

This study had several limitations. First, and most importantly, this study was retrospectively performed in a single institution with a small number of patients, and the findings of this study cannot be generalized. The treatment policy was selected by individual physicians; therefore, potential biases cannot be eliminated. The majority of patients received other reversal agents, particularly fresh frozen plasma. Therefore, it is difficult to conclude whether the revised two‐level initiation strategy itself is effective. Second, the patients examined in this study were relatively old with a low body weight, and these results may not be adequate for heavier patients. Finally, the complexity of the disease varied in our study, especially among patients with trauma‐related bleeding, and optimal target of INR was not generalized. Prospective trials involving large numbers of patients who require urgent normalization of INR are required to overcome these limitations and clarify the remaining unsolved issues.

In conclusion, a two‐level fixed dose of PCC according to the initial INR is a potentially reasonable strategy for normalizing INR in a Japanese ED. However, patients with supratherapeutic initial INR and/or higher weight may require a higher or supplemental dose. Additional studies with hemostatic outcome measures are needed to further elucidate the efficacy of a relatively low fixed‐dose PCC regimen in Japan.

Disclosure

Approval of the Research Protocol: This study was performed in accordance with the International Conference on Harmonization Good Clinical Practice Guidelines and the 1996 Declaration of Helsinki. The study protocol including the administration of PCC was approved by the ethics committees of our institution (approved number H2020‐189).

Informed Consent: We posted information about this study on the hospital website and gave participants the opportunity to opt out, and those who did not opt out were considered to have provided tacit consent for study participation.

Registry and the Registration No. of the study/Trial: N/A.

Animal Studies: N/A.

Conflict of Interest: None declared.