ORIGINAL ARTICLE
Comparison of Anti-Xa and Activated Partial Thromboplastin Time Monitoring for Heparin Dosing in Patients With Cirrhosis Amaris Fuentes, PharmD, BCPS,* Sherilyn Gordon-Burroughs, MD, FACS,‡ Jeffrey B. Hall, PharmD,* David R. Putney, PharmD, BCPS, AQ Cardiology,* and Howard P. Monsour, Jr, MD† Key Words: anticoagulation, hematology, liver
Backgroud: Cirrhosis of the liver results in complex hemostatic changes that place patients at risk for both bleeding and thrombotic events. This study evaluates the adverse effects of anticoagulation with unfractionated heparin among patients with cirrhosis and analyzes the discrepancy between anti-Xa and activated partial thromboplastin time (aPTT) values for heparin monitoring among cirrhotics.
Methods: Patients with cirrhosis receiving unfractionated heparin were matched 2:1 to patients without evidence of cirrhosis anticoagulated with unfractioned heparin. Markers of bleeding events including blood product administration and use of heparin reversal were analyzed between groups. Patients from both groups with aPTT and anti-Xa values obtained at the same time were also analyzed.
Results: A higher incidence of blood product administration or use of heparin reversal was observed among patients with cirrhosis [35/ 105 (33.3%) versus 37/210 (17.6%), P = 0.002]. This finding was consistent among those receiving anticoagulation through an established anti-Xa–based heparin dosing protocol [23/62 (37.1%) versus 25/124 (20.2%), P = 0.013]. A decrease in hemoglobin greater than 2 g/dL or a platelet decrease 50% or greater from baseline was also more frequently identified among cirrhotics when receiving heparin therapy [20/105 (19%) versus 23/210 (11%), P = 0.049 and 21/105 (20%) versus 12/210 (6%), P , 0.001, respectively]. A total of 88 correlated anti-Xa and aPTT values from 35 patients with cirrhosis demonstrated supratherapeutic aPTT values for anti-Xa levels within the therapeutic range (P , 0.001). This discrepancy was not observed among controls.
Conclusions: A greater use of blood products among the cirrhotic population may indicate potential bleeding events on therapy. A discrepancy in correlated anti-Xa and aPTT values among patients with cirrhosis may explain the propensity for adverse effects. Further study is required to identify effective heparin anticoagulation monitoring strategies in liver disease. Received for publication March 18, 2014; accepted March 26, 2014. From the Departments of *Pharmacy; †Medicine; and ‡Surgery, Houston Methodist Hospital, TX. Interim results presented as a poster at the 42nd Annual Critical Care Congress of the Society of Critical Care Medicine, San Juan, Puerto Rico on January 21, 2013. The authors declare no conflict of interest. Correspondence: Amaris Fuentes, PharmD, BCPS, Department of Pharmacy, Houston Methodist Hospital, 6565 Fannin St, DB1-09, Houston, TX 77030 (e-mail: [email protected]). Copyright © 2014 Wolters Kluwer Health, Inc. All rights reserved.
40
(Ther Drug Monit 2015;37:40–44)
R
educed synthetic function in chronic liver disease is associated with complex hemostatic changes characterized by a decrease in procoagulant and anticoagulant factors. Despite these discrepancies, a delicate balance is maintained placing these patients at risk for both bleeding and thrombotic events with mounting evidence refuting the notion of autoanticoagulation among patients with cirrhosis.1,2 In fact, portal vein thrombosis is observed at a rate of 5%–15% among non-neoplastic liver disease patients along with other venous or arterial thrombotic complications such as lower extremity deep venous thrombosis or catheter-related thrombosis.3,4 Laboratory tests used for monitoring anticoagulation may reflect a hypocoagulable state limiting an accurate assessment of the safety and effectiveness of therapy. In particular, the activated partial thromboplastin time (aPTT) may be prolonged at baseline providing limited guidance in heparin anticoagulation management.1,2,5 The anti–factor Xa assay is an alternative for heparin dose monitoring, which accounts for the amount of drug present through the measurement of free factor Xa.5 Limited information, however, is available on the utility of the anti–factor Xa assay among patients with liver disease6–8 Data analyzing the in vitro effects of low-molecular weight heparins indicate an underestimation of therapy effect by the anti–factor Xa assay in patients with cirrhosis leading to potentially dangerous dose increases. Decreased production of the anticoagulant antithrombin, potentiated by heparin compounds, has been implicated in this effect.2,8,9 To date, no studies have evaluated the relationship between anti-Xa and aPTT values for heparin anticoagulation among patients with cirrhosis. This retrospective evaluation assesses potential adverse effects of heparin use for the management of thrombotic events among cirrhotics, and the discrepancy between anti–factor Xa and aPTT values in their use as measures of therapeutic effect.
MATERIALS AND METHODS Patients and Study Design A single center retrospective analysis was conducted among patients admitted to a tertiary care, academic medical Ther Drug Monit Volume 37, Number 1, February 2015
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 1, February 2015
Anti-Xa and Partial Thromboplastin Time in Cirrhosis
center with a liver transplantation program between January 2008 and September 2011. This study was approved by the institutional review board with a waiver of informed consent. Patients with cirrhosis receiving full-dose unfractionated heparin anticoagulation were identified through a search of the University Health Consortium Clinical Database/ Resource Manager (UHC CDB/RM) for ICD-9 codes for liver cirrhosis and a charge code for heparin 25,000 units in 500 mL dextrose 5% water. The UHC CDB/RM contains patient demographics, diagnosis and procedure codes, and other information about patient encounters at academic medical centers participating in UHC. Underlying etiologies of cirrhosis were obtained through a review of hospital medical records. Case matching was performed in a 2:1 fashion to patients receiving heparin without documented cirrhosis between January 2008 and September 2011. These patients were identified through an internal database of patients receiving heparin anticoagulation through an established dosing protocol with anti–factor Xa–based adjustments. Matching was performed on the basis of heparin indication divided into either thrombotic (ie, deep vein thrombosis, pulmonary embolism, and portal vein thrombosis) or other indications requiring anticoagulation (ie, acute coronary syndrome, atrial fibrillation, and hypercoagulable states). Additionally, patients were matched based on weight (610%), age (610 years), and gender.
regression models were completed for a comparison of antiXa and aPTT values between study populations. Secondary outcomes were compared using the x2 test and Fischer exact test for categorical data. Time to therapeutic anti-Xa and dose at the first therapeutic anti-Xa was compared using the Mann– Whitney U test. Significant differences were defined as values of P less than 0.05.
Outcomes
Outcomes
The use of packed red blood cells, fresh frozen plasma (FFP), platelets, and cryoprecipitate or the use of protamine for heparin reversal was evaluated as markers for bleeding complications. Additional outcomes included the incidence of a decrease in hemoglobin greater than 2 g/dL within 24 hours, decrease in platelets greater than 50% from baseline, time to first therapeutic anti-Xa value, and heparin dose at first therapeutic anti-Xa value. A subgroup analysis was performed on patients with cirrhosis managed solely by the anti– factor Xa–based heparin dosing protocol in comparison to case-matched controls. Anti-Xa levels and aPTT values obtained within 15 minutes of each other were analyzed following a previously described methodology.9 Each was categorized as therapeutic (anti-Xa: 0.3–0.7 units/dL; aPTT: 61–112 seconds), subtherapeutic (anti-Xa: ,0.3 units/dL; aPTT: ,61 seconds), or supratherapeutic (anti-Xa: .0.7 units/dL; aPTT: .112 seconds). The number of anti-Xa and aPTT values in accordance (ie, both therapeutic, subtherapeutic, or supratherapeutic) was compared between cirrhotics and noncirrhotics then analyzed by linear regression. Laboratory agents used for the measurement of anti-Xa and aPTT values remained consistent through the study period.
Thirty-five patients with cirrhosis and 37 patients without cirrhosis received transfusion of blood products or protamine for heparin reversal [35/105 (33.3%) versus 37/210 (17.6%), P = 0.002]. Significantly more patients with cirrhosis required packed red blood cells and FFP transfusions [25/ 105 (23.8%) versus 29/210 (13.8%), P = 0.03 and 12/105 (11.4%) versus 10/210 (4.8%), P = 0.03]. Limited use of protamine was observed with 1 patient receiving heparin reversal in the case–match arm (Table 2). A decrease in hemoglobin $2 g/dL in 24 hours was observed in 20 patients with cirrhosis in comparison with 23 patients without cirrhosis [20/105 (19%) versus 23/210 (11%), P = 0.05]. A platelet decrease $50% from baseline occurred in 21 patients with cirrhosis and 12 patients without cirrhosis [21/105 (20%) versus 12/210 (6%), P , 0.001]. Among the patients achieving a therapeutic anti–factor Xa value in each group, the time to therapeutic anti–factor Xa was significantly longer among the patients with cirrhosis with comparable heparin dosing (14:08 hours:minutes [IQ 7:54-25:25] versus 9:08 [IQ 5:43-23:18] and 1050 units/h [IQ 850-1350] versus 1000 units/h [IQ 800-1250], P = 0.29) (Table 2).
Statistical Analysis Data analysis was conducted with MiniTab 16 (Minitab Inc., State College, PA). Continuous values were tested for normality using the Anderson–Darling equation. Baseline characteristics are described as a mean 6 SD or median and interquartile range contingent on distribution. Linear
Sixty-two patients with cirrhosis managed on our anti–factor Xa–based heparin dosing protocol and their 124 case-matched controls were identified for analysis. A greater percentage of patients with cirrhosis required transfusions or protamine for heparin reversal [23/62 (37.1%) versus 25/124 (20.2%), P = 0.01]. Patients with cirrhosis also required
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
41
RESULTS One hundred thirty-seven patients with documented cirrhosis were identified for study inclusion from the UHC CDB/RM. On review of patient encounters, 22 were excluded for lack of documented heparin indication and 10 patients for lack of laboratory monitoring through the use of either an aPTT or anti–factor Xa level. Table 1 summarizes the baseline characteristics of the 105 patients with cirrhosis and 210 case-matched controls included in the study. The population was demographically well matched with heparin initiated for thrombotic indications in 41% of patients. Among the cirrhotic population, the average Model for End-Stage Liver Disease score was 17.5. Hepatitis C infection and alcoholic liver disease were the etiology of cirrhosis in 31.4% and 22.9% of patients with cirrhosis, respectively. Among patients with cirrhosis, 59% were managed on the anti–factor Xa–based heparin dosing protocol with at least 1 anti–factor Xa values available in 68% of patients.
Subgroup Analysis
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 1, February 2015
Fuentes et al
TABLE 1. Baseline Characteristics
Age (yrs) Weight (kg) Height (inches) Male (%) Heparin indication Thrombosis* (%) Other† (%) Protocol managed (%) MELD score‡ INR§ aPTT (sec)¶ Patients with anti-Xa values (%) Patients with aPTT values (%) Days of heparin therapy
Cirrhotic (n = 105)
Noncirrhotic (n = 210)
P
61.4 6 11.9 79.7 6 22.0 67.5 6 4.5 68 (65%)
63 6 12.2 80.3 6 22.0 68 6 6 137 (65%)
0.2 0.98 0.87 0.93
43 (41%) 62 (59%) 62 (59%)
86 (41%) 124 (59%) 210 (100%)
.0.99 .0.99 ,0.001
17.5 6 7.7 1.6 (1.4–1.9) 38.9 (33.4–50.0) 71 (68%)
N/A 1.2 (1.1–1.4) 36.6 (31.1–45.8) 210 (100%)
N/A ,0.001 0.04 ,0.001
90 (86%)
29 (14%)
,0.001
4 (2–7)
3 (1–6)
0.02
*Pulmonary embolism, portal vein thrombosis, and deep vein thrombosis. †Acute coronary syndrome, atrial flutter/fibrillation, hypercoagulable state, and stroke. ‡MELD scores available for 104 patients with cirrhosis. §INR available for 105 patients with cirrhosis and 209 patients without cirrhosis. ¶aPTT available for 91 patients with cirrhosis and 146 patients without cirrhosis. INR, International normalization ratio; MELD, Model for End Stage Liver Disease.
a greater frequency of FFP transfusions [9/62 (14.5%) versus 7/124 (5.6%), P = 0.04]. Observed decreases of hemoglobin were not different between the 2 groups [10/62 (16%) versus 12/124 (10%), P = 0.20], however, a decrease in platelets $50% from baseline occurred more frequently among patients with cirrhosis [11/62 (18%) versus 7/124 (6%), P = 0.02]. The difference
in time to therapeutic anti–factor Xa was longer among patients with cirrhosis at 14:33 hours:minutes (IQ, 7:37-25:37) in comparison with 11:58 hours:minutes (IQ, 6:11-23:18) among noncirrhotics (P = 0.18). Similar heparin dosing at the time of therapeutic anti–factor Xa was seen between study groups (1025 units/h [IQ 650-1350] versus 1025 units/h [IQ 400-1300], P = 0.83).
Correlation of aPTT and Anti-Xa Values A total of 170 values from 84 patients were identified in the noncirrhotic group. Using the line of best-fit, anti–factor Xa values of 0.3–0.7 units/mL established for heparin dosing corresponded to aPTT values of 75–125 seconds, closely approximating the hospital laboratory specific aPTT therapeutic range of 61–112 seconds. Among the cirrhotic population, 88 values from 35 patients were identified. Therapeutic anti– factor Xa values correlated to higher aPTT values of 100–200 seconds. Linear regression analysis demonstrated a significant difference between the patients with cirrhosis and patients without cirrhosis (P , 0.001) (Fig. 1). Anti-Xa and aPTT values were in higher accordance among patients without cirrhosis (48% cirrhotics versus 61% noncirrhotics, P = 0.05) (Table 3).
DISCUSSION Reduced hepatic synthetic function diminishes the production of clotting factors and anticoagulants. Overall bleeding risks, however, are minimally increased and thrombotic complications such as portal vein thrombosis are evident.1,2 Clot formation in the portal vein can ultimately worsen cirrhosis by decreasing portal blood flow and increasing portal pressures, and can complicate future liver transplantation.10 Use of heparin anticoagulation in liver disease poses unique challenges. Traditional tests used to guide therapy
TABLE 2. Secondary Outcomes Primary Study Population Cirrhotic (n = 105), n (%)
Protocol-Managed Population
Noncirrhotic (n = 210), n (%)
Patients receiving transfusion or 35 (33.3) 37 protamine Patients transfused 35 (33.3) 36 PRBC 25 (23.8) 29 Platelet 9 (8.6) 8 FFP 12 (11.4) 10 Cryoprecipitate 3 (2.9) 5 Patients receiving protamine 0 (0) 1 Hg decrease $2 g/dL 20 (19) 23 Platelets decrease $50% from 21 (20) 12 baseline Patients with therapeutic anti-Xa 59 (56) 184 Time to therapeutic anti-Xa (h) 14:08 (IQ, 7:54-25:25)* 9:40 Dose at therapeutic anti-Xa (units/h) 1050 (IQ, 850–1350)* 1000
P
Cirrhotic (n = 62), n (%)
Noncirrhotic (n = 124), n (%)
(17.6)
0.002
23 (37.1)
25 (20.2)
(17.2) (13.8) (3.8) (4.8) (2.4) (0.5) (11) (6)
0.001 0.03 0.08 0.03 .0.99 .0.99 0.05 ,0.001
23 (37.1) 16 (25.8) 6 (9.7) 9 (14.5) 2 (3.2) 0 (0) 10 (16) 11 (18)
25 20 4 7 4
(20.2) (16.1) (3.2) (5.6) (3.2) 0 (0) 12 (10) 7 (6)
P 0.01 0.01 0.12 0.09 0.04 .0.99 .0.99 0.20 0.02
(88) ,0.001 54 (87) 108 (87) .0.99 (IQ, 5:43-23:18)* 0.04 14:33 (IQ, 7:37-25:37)† 11:58 (IQ, 6:11-23:18)† 0.18 (IQ, 800–1250)* 0.29 1025 (IQ, 650–1350)† 1025 (IQ, 400–1300)† 0.83
*Therapeutic anti-Xa values achieved in 59 patients with cirrhosis and 184 patients without cirrhosis. †Therapeutic anti-Xa values achieved in 54 patients with cirrhosis and 108 patients without cirrhosis.
42
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 1, February 2015
FIGURE 1. A and B, Comparison of aPTT and anti-Xa values obtained simultaneously.
may be limited in this population, as they reflect both an unquantifiable coagulation defect as well as medication effects.2,5 Limited experience with the anti–factor Xa is noted in the literature for use with either unfractionated or low-molecular weight heparin therapy among patients with liver disease2,5,6 Evidence from in vitro laboratory
Anti-Xa and Partial Thromboplastin Time in Cirrhosis
investigations also suggests poor correlation of anti–factor Xa values and therapeutic anticoagulant effect among cirrhotics.6–8 A higher number of bleeding events were observed among patients with cirrhosis compared with patients without cirrhosis identified by the transfusion of blood products, use of protamine for heparin reversal, and alterations of hematological values. This difference was observed despite a majority of patients with cirrhosis receiving heparin therapy by a specified dosing protocol. The number of patients with cirrhosis managed off-protocol was attributed to a variable practice among members of the institution’s transplant team whereby conservative aPTT values are maintained during heparin anticoagulation to reduce the risk of bleeding events. Similar bleeding events were identified when evaluating the subgroup of patients with cirrhosis managed solely by anti– factor Xa levels compared with noncirrhotics. Our analysis of the relationship between anti-Xa and aPTT values among patients with cirrhosis during heparin anticoagulation offers a possible explanation for the increased use of blood products. Among patients without cirrhosis, therapeutic anti-Xa values closely approximated established institutional therapeutic aPTT values of 61–112 seconds. Much higher aPTT values (100–200 seconds), however, were seen with corresponding therapeutic anti-Xa values in patients with cirrhosis. If conversely analyzed, therapeutic aPTT values of 61–112 seconds corresponded to much lower anti-Xa values of 0.15–0.35 units/mL suggesting an altered relationship as a result of decreased synthesis of most procoagulants in cirrhotics. This study was limited to a description of the discrepancy between anti-Xa and aPTT values among patients with cirrhosis. Additional analysis of safety aspects of unfractionated heparin use was done as surrogate markers of clinically significant bleeding. Correlated anti-Xa and aPTT values obtained within the 15-minute window, however, were only available in a small percentage of patients. Transfusions were also administered based on the clinical judgment of physician involved in the
TABLE 3. Relationship of Anti-Xa and aPTT Values Among Patients With Cirrhosis and Patients Without Cirrhosis Cirrhotics aPTT Values
Anti-Xa values Supratherapeutic values (.0.7 units/dL) Therapeutic values (0.3–0.7 units/dL) Subtherapeutic values (,0.3 units/dL)
Subtherapeutic Values (,61 s), n (%)
Therapeutic Values (61–112 s), n (%)
Supratherapeutic Values (.112 s), n (%)
2 (2) 15 (17) 26 (30)
27 (31) 15 (17) 1 (1)
1 (1) 1 (1) 0 (0)
Noncirrhotics aPTT Values
Anti-Xa values Supratherapeutic values (.0.7 units/dL) Therapeutic values (0.3–0.7 units/dL) Subtherapeutic values (,0.3 units/dL)
Subtherapeutic Values (,61 s), n (%)
Therapeutic Values (61–112 s), n (%)
Supratherapeutic Values (.112 s), n (%)
6 (4) 16 (9) 43 (25)
21 (12) 48 (28) 19 (11)
13 (8) 3 (2) 1 (1)
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
43
Fuentes et al
Ther Drug Monit Volume 37, Number 1, February 2015
medical management of the patient. Analysis was conducted in a retrospective manner, which can introduce selection bias and limits the determination of causal relationships. Until coagulation tests better suited for monitoring in this population enter clinical practice, further study is required to establish safe and effective anticoagulation strategies, as clinical efficacy could not be accurately evaluated in a retrospective fashion.
vein thrombosis in patients with liver cirrhosis. J Clin Gastroenterol. 2010;44:448–451. Sogaard KK, Horvath-Puho E, Gronbaek H, et al. Risk of venous thromboembolism in patients with liver disease: a nation wide populationbased case-control study. Am J Gastroenterol. 2009;104:96–101. Greaves M. Limitations of the laboratory monitoring of heparin therapy. Thromb Haemost. 2002;87:163–164. Bechmann LP, Sichau M, Wichert M, et al. Low-molecular-weight heparin in patients with advanced cirrhosis. Liver Int. 2011;31:75–82. Sette H, Hughes RD, Langley PG, et al. Heparin response and clearance in acute and chronic liver disease. Thromb Haemost. 1985;54:591–594. Senzolo M, Rodriguez-Castro KI, Radu RC, et al. Increased anticoagulant response to low-molecular-weight heparin in plasma from patients with advanced cirrhosis. J Thromb Haemost. 2012;10:1823–1829. Brill-Edwards P, Ginsberg JS, Johnston M, et al. Establishing a therapeutic range for heparin therapy. Ann Intern Med. 1993;119:104–109. Charco R, Fuster J, Fondevila C, et al. Portal vein thrombosis in liver transplantation. Transplant Proc. 2005;37:3904–3905.
REFERENCES 1. Tripodi A, Primignani M, Chantarangkul V, et al. An imbalance of provs anti-coagulation factors in plasma from patients with cirrhosis. Gastroenterology. 2009;137:2105–2111. 2. Dabbagh O, Oza A, Prakash S. Coagulopathy does not protect against venous thromboembolism in hospitalized patients with chronic liver disease. Chest. 2010;137:1145–1149. 3. Amitrano LA, Guardascione MA, Menchise A, et al. Safety and efficacy of anticoagulation therapy with low molecular weight heparin for portal
44
4.
5. 6. 7. 8.
9. 10.
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Comparison of Anti-Xa and Activated Partial Thromboplastin Time Monitoring for Heparin Dosing in Patients With Cirrhosis Amaris Fuentes, PharmD, BCPS,* Sherilyn Gordon-Burroughs, MD, FACS,‡ Jeffrey B. Hall, PharmD,* David R. Putney, PharmD, BCPS, AQ Cardiology,* and Howard P. Monsour, Jr, MD† Key Words: anticoagulation, hematology, liver
Backgroud: Cirrhosis of the liver results in complex hemostatic changes that place patients at risk for both bleeding and thrombotic events. This study evaluates the adverse effects of anticoagulation with unfractionated heparin among patients with cirrhosis and analyzes the discrepancy between anti-Xa and activated partial thromboplastin time (aPTT) values for heparin monitoring among cirrhotics.
Methods: Patients with cirrhosis receiving unfractionated heparin were matched 2:1 to patients without evidence of cirrhosis anticoagulated with unfractioned heparin. Markers of bleeding events including blood product administration and use of heparin reversal were analyzed between groups. Patients from both groups with aPTT and anti-Xa values obtained at the same time were also analyzed.
Results: A higher incidence of blood product administration or use of heparin reversal was observed among patients with cirrhosis [35/ 105 (33.3%) versus 37/210 (17.6%), P = 0.002]. This finding was consistent among those receiving anticoagulation through an established anti-Xa–based heparin dosing protocol [23/62 (37.1%) versus 25/124 (20.2%), P = 0.013]. A decrease in hemoglobin greater than 2 g/dL or a platelet decrease 50% or greater from baseline was also more frequently identified among cirrhotics when receiving heparin therapy [20/105 (19%) versus 23/210 (11%), P = 0.049 and 21/105 (20%) versus 12/210 (6%), P , 0.001, respectively]. A total of 88 correlated anti-Xa and aPTT values from 35 patients with cirrhosis demonstrated supratherapeutic aPTT values for anti-Xa levels within the therapeutic range (P , 0.001). This discrepancy was not observed among controls.
Conclusions: A greater use of blood products among the cirrhotic population may indicate potential bleeding events on therapy. A discrepancy in correlated anti-Xa and aPTT values among patients with cirrhosis may explain the propensity for adverse effects. Further study is required to identify effective heparin anticoagulation monitoring strategies in liver disease. Received for publication March 18, 2014; accepted March 26, 2014. From the Departments of *Pharmacy; †Medicine; and ‡Surgery, Houston Methodist Hospital, TX. Interim results presented as a poster at the 42nd Annual Critical Care Congress of the Society of Critical Care Medicine, San Juan, Puerto Rico on January 21, 2013. The authors declare no conflict of interest. Correspondence: Amaris Fuentes, PharmD, BCPS, Department of Pharmacy, Houston Methodist Hospital, 6565 Fannin St, DB1-09, Houston, TX 77030 (e-mail: [email protected]). Copyright © 2014 Wolters Kluwer Health, Inc. All rights reserved.
40
(Ther Drug Monit 2015;37:40–44)
R
educed synthetic function in chronic liver disease is associated with complex hemostatic changes characterized by a decrease in procoagulant and anticoagulant factors. Despite these discrepancies, a delicate balance is maintained placing these patients at risk for both bleeding and thrombotic events with mounting evidence refuting the notion of autoanticoagulation among patients with cirrhosis.1,2 In fact, portal vein thrombosis is observed at a rate of 5%–15% among non-neoplastic liver disease patients along with other venous or arterial thrombotic complications such as lower extremity deep venous thrombosis or catheter-related thrombosis.3,4 Laboratory tests used for monitoring anticoagulation may reflect a hypocoagulable state limiting an accurate assessment of the safety and effectiveness of therapy. In particular, the activated partial thromboplastin time (aPTT) may be prolonged at baseline providing limited guidance in heparin anticoagulation management.1,2,5 The anti–factor Xa assay is an alternative for heparin dose monitoring, which accounts for the amount of drug present through the measurement of free factor Xa.5 Limited information, however, is available on the utility of the anti–factor Xa assay among patients with liver disease6–8 Data analyzing the in vitro effects of low-molecular weight heparins indicate an underestimation of therapy effect by the anti–factor Xa assay in patients with cirrhosis leading to potentially dangerous dose increases. Decreased production of the anticoagulant antithrombin, potentiated by heparin compounds, has been implicated in this effect.2,8,9 To date, no studies have evaluated the relationship between anti-Xa and aPTT values for heparin anticoagulation among patients with cirrhosis. This retrospective evaluation assesses potential adverse effects of heparin use for the management of thrombotic events among cirrhotics, and the discrepancy between anti–factor Xa and aPTT values in their use as measures of therapeutic effect.
MATERIALS AND METHODS Patients and Study Design A single center retrospective analysis was conducted among patients admitted to a tertiary care, academic medical Ther Drug Monit Volume 37, Number 1, February 2015
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 1, February 2015
Anti-Xa and Partial Thromboplastin Time in Cirrhosis
center with a liver transplantation program between January 2008 and September 2011. This study was approved by the institutional review board with a waiver of informed consent. Patients with cirrhosis receiving full-dose unfractionated heparin anticoagulation were identified through a search of the University Health Consortium Clinical Database/ Resource Manager (UHC CDB/RM) for ICD-9 codes for liver cirrhosis and a charge code for heparin 25,000 units in 500 mL dextrose 5% water. The UHC CDB/RM contains patient demographics, diagnosis and procedure codes, and other information about patient encounters at academic medical centers participating in UHC. Underlying etiologies of cirrhosis were obtained through a review of hospital medical records. Case matching was performed in a 2:1 fashion to patients receiving heparin without documented cirrhosis between January 2008 and September 2011. These patients were identified through an internal database of patients receiving heparin anticoagulation through an established dosing protocol with anti–factor Xa–based adjustments. Matching was performed on the basis of heparin indication divided into either thrombotic (ie, deep vein thrombosis, pulmonary embolism, and portal vein thrombosis) or other indications requiring anticoagulation (ie, acute coronary syndrome, atrial fibrillation, and hypercoagulable states). Additionally, patients were matched based on weight (610%), age (610 years), and gender.
regression models were completed for a comparison of antiXa and aPTT values between study populations. Secondary outcomes were compared using the x2 test and Fischer exact test for categorical data. Time to therapeutic anti-Xa and dose at the first therapeutic anti-Xa was compared using the Mann– Whitney U test. Significant differences were defined as values of P less than 0.05.
Outcomes
Outcomes
The use of packed red blood cells, fresh frozen plasma (FFP), platelets, and cryoprecipitate or the use of protamine for heparin reversal was evaluated as markers for bleeding complications. Additional outcomes included the incidence of a decrease in hemoglobin greater than 2 g/dL within 24 hours, decrease in platelets greater than 50% from baseline, time to first therapeutic anti-Xa value, and heparin dose at first therapeutic anti-Xa value. A subgroup analysis was performed on patients with cirrhosis managed solely by the anti– factor Xa–based heparin dosing protocol in comparison to case-matched controls. Anti-Xa levels and aPTT values obtained within 15 minutes of each other were analyzed following a previously described methodology.9 Each was categorized as therapeutic (anti-Xa: 0.3–0.7 units/dL; aPTT: 61–112 seconds), subtherapeutic (anti-Xa: ,0.3 units/dL; aPTT: ,61 seconds), or supratherapeutic (anti-Xa: .0.7 units/dL; aPTT: .112 seconds). The number of anti-Xa and aPTT values in accordance (ie, both therapeutic, subtherapeutic, or supratherapeutic) was compared between cirrhotics and noncirrhotics then analyzed by linear regression. Laboratory agents used for the measurement of anti-Xa and aPTT values remained consistent through the study period.
Thirty-five patients with cirrhosis and 37 patients without cirrhosis received transfusion of blood products or protamine for heparin reversal [35/105 (33.3%) versus 37/210 (17.6%), P = 0.002]. Significantly more patients with cirrhosis required packed red blood cells and FFP transfusions [25/ 105 (23.8%) versus 29/210 (13.8%), P = 0.03 and 12/105 (11.4%) versus 10/210 (4.8%), P = 0.03]. Limited use of protamine was observed with 1 patient receiving heparin reversal in the case–match arm (Table 2). A decrease in hemoglobin $2 g/dL in 24 hours was observed in 20 patients with cirrhosis in comparison with 23 patients without cirrhosis [20/105 (19%) versus 23/210 (11%), P = 0.05]. A platelet decrease $50% from baseline occurred in 21 patients with cirrhosis and 12 patients without cirrhosis [21/105 (20%) versus 12/210 (6%), P , 0.001]. Among the patients achieving a therapeutic anti–factor Xa value in each group, the time to therapeutic anti–factor Xa was significantly longer among the patients with cirrhosis with comparable heparin dosing (14:08 hours:minutes [IQ 7:54-25:25] versus 9:08 [IQ 5:43-23:18] and 1050 units/h [IQ 850-1350] versus 1000 units/h [IQ 800-1250], P = 0.29) (Table 2).
Statistical Analysis Data analysis was conducted with MiniTab 16 (Minitab Inc., State College, PA). Continuous values were tested for normality using the Anderson–Darling equation. Baseline characteristics are described as a mean 6 SD or median and interquartile range contingent on distribution. Linear
Sixty-two patients with cirrhosis managed on our anti–factor Xa–based heparin dosing protocol and their 124 case-matched controls were identified for analysis. A greater percentage of patients with cirrhosis required transfusions or protamine for heparin reversal [23/62 (37.1%) versus 25/124 (20.2%), P = 0.01]. Patients with cirrhosis also required
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
41
RESULTS One hundred thirty-seven patients with documented cirrhosis were identified for study inclusion from the UHC CDB/RM. On review of patient encounters, 22 were excluded for lack of documented heparin indication and 10 patients for lack of laboratory monitoring through the use of either an aPTT or anti–factor Xa level. Table 1 summarizes the baseline characteristics of the 105 patients with cirrhosis and 210 case-matched controls included in the study. The population was demographically well matched with heparin initiated for thrombotic indications in 41% of patients. Among the cirrhotic population, the average Model for End-Stage Liver Disease score was 17.5. Hepatitis C infection and alcoholic liver disease were the etiology of cirrhosis in 31.4% and 22.9% of patients with cirrhosis, respectively. Among patients with cirrhosis, 59% were managed on the anti–factor Xa–based heparin dosing protocol with at least 1 anti–factor Xa values available in 68% of patients.
Subgroup Analysis
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 1, February 2015
Fuentes et al
TABLE 1. Baseline Characteristics
Age (yrs) Weight (kg) Height (inches) Male (%) Heparin indication Thrombosis* (%) Other† (%) Protocol managed (%) MELD score‡ INR§ aPTT (sec)¶ Patients with anti-Xa values (%) Patients with aPTT values (%) Days of heparin therapy
Cirrhotic (n = 105)
Noncirrhotic (n = 210)
P
61.4 6 11.9 79.7 6 22.0 67.5 6 4.5 68 (65%)
63 6 12.2 80.3 6 22.0 68 6 6 137 (65%)
0.2 0.98 0.87 0.93
43 (41%) 62 (59%) 62 (59%)
86 (41%) 124 (59%) 210 (100%)
.0.99 .0.99 ,0.001
17.5 6 7.7 1.6 (1.4–1.9) 38.9 (33.4–50.0) 71 (68%)
N/A 1.2 (1.1–1.4) 36.6 (31.1–45.8) 210 (100%)
N/A ,0.001 0.04 ,0.001
90 (86%)
29 (14%)
,0.001
4 (2–7)
3 (1–6)
0.02
*Pulmonary embolism, portal vein thrombosis, and deep vein thrombosis. †Acute coronary syndrome, atrial flutter/fibrillation, hypercoagulable state, and stroke. ‡MELD scores available for 104 patients with cirrhosis. §INR available for 105 patients with cirrhosis and 209 patients without cirrhosis. ¶aPTT available for 91 patients with cirrhosis and 146 patients without cirrhosis. INR, International normalization ratio; MELD, Model for End Stage Liver Disease.
a greater frequency of FFP transfusions [9/62 (14.5%) versus 7/124 (5.6%), P = 0.04]. Observed decreases of hemoglobin were not different between the 2 groups [10/62 (16%) versus 12/124 (10%), P = 0.20], however, a decrease in platelets $50% from baseline occurred more frequently among patients with cirrhosis [11/62 (18%) versus 7/124 (6%), P = 0.02]. The difference
in time to therapeutic anti–factor Xa was longer among patients with cirrhosis at 14:33 hours:minutes (IQ, 7:37-25:37) in comparison with 11:58 hours:minutes (IQ, 6:11-23:18) among noncirrhotics (P = 0.18). Similar heparin dosing at the time of therapeutic anti–factor Xa was seen between study groups (1025 units/h [IQ 650-1350] versus 1025 units/h [IQ 400-1300], P = 0.83).
Correlation of aPTT and Anti-Xa Values A total of 170 values from 84 patients were identified in the noncirrhotic group. Using the line of best-fit, anti–factor Xa values of 0.3–0.7 units/mL established for heparin dosing corresponded to aPTT values of 75–125 seconds, closely approximating the hospital laboratory specific aPTT therapeutic range of 61–112 seconds. Among the cirrhotic population, 88 values from 35 patients were identified. Therapeutic anti– factor Xa values correlated to higher aPTT values of 100–200 seconds. Linear regression analysis demonstrated a significant difference between the patients with cirrhosis and patients without cirrhosis (P , 0.001) (Fig. 1). Anti-Xa and aPTT values were in higher accordance among patients without cirrhosis (48% cirrhotics versus 61% noncirrhotics, P = 0.05) (Table 3).
DISCUSSION Reduced hepatic synthetic function diminishes the production of clotting factors and anticoagulants. Overall bleeding risks, however, are minimally increased and thrombotic complications such as portal vein thrombosis are evident.1,2 Clot formation in the portal vein can ultimately worsen cirrhosis by decreasing portal blood flow and increasing portal pressures, and can complicate future liver transplantation.10 Use of heparin anticoagulation in liver disease poses unique challenges. Traditional tests used to guide therapy
TABLE 2. Secondary Outcomes Primary Study Population Cirrhotic (n = 105), n (%)
Protocol-Managed Population
Noncirrhotic (n = 210), n (%)
Patients receiving transfusion or 35 (33.3) 37 protamine Patients transfused 35 (33.3) 36 PRBC 25 (23.8) 29 Platelet 9 (8.6) 8 FFP 12 (11.4) 10 Cryoprecipitate 3 (2.9) 5 Patients receiving protamine 0 (0) 1 Hg decrease $2 g/dL 20 (19) 23 Platelets decrease $50% from 21 (20) 12 baseline Patients with therapeutic anti-Xa 59 (56) 184 Time to therapeutic anti-Xa (h) 14:08 (IQ, 7:54-25:25)* 9:40 Dose at therapeutic anti-Xa (units/h) 1050 (IQ, 850–1350)* 1000
P
Cirrhotic (n = 62), n (%)
Noncirrhotic (n = 124), n (%)
(17.6)
0.002
23 (37.1)
25 (20.2)
(17.2) (13.8) (3.8) (4.8) (2.4) (0.5) (11) (6)
0.001 0.03 0.08 0.03 .0.99 .0.99 0.05 ,0.001
23 (37.1) 16 (25.8) 6 (9.7) 9 (14.5) 2 (3.2) 0 (0) 10 (16) 11 (18)
25 20 4 7 4
(20.2) (16.1) (3.2) (5.6) (3.2) 0 (0) 12 (10) 7 (6)
P 0.01 0.01 0.12 0.09 0.04 .0.99 .0.99 0.20 0.02
(88) ,0.001 54 (87) 108 (87) .0.99 (IQ, 5:43-23:18)* 0.04 14:33 (IQ, 7:37-25:37)† 11:58 (IQ, 6:11-23:18)† 0.18 (IQ, 800–1250)* 0.29 1025 (IQ, 650–1350)† 1025 (IQ, 400–1300)† 0.83
*Therapeutic anti-Xa values achieved in 59 patients with cirrhosis and 184 patients without cirrhosis. †Therapeutic anti-Xa values achieved in 54 patients with cirrhosis and 108 patients without cirrhosis.
42
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 1, February 2015
FIGURE 1. A and B, Comparison of aPTT and anti-Xa values obtained simultaneously.
may be limited in this population, as they reflect both an unquantifiable coagulation defect as well as medication effects.2,5 Limited experience with the anti–factor Xa is noted in the literature for use with either unfractionated or low-molecular weight heparin therapy among patients with liver disease2,5,6 Evidence from in vitro laboratory
Anti-Xa and Partial Thromboplastin Time in Cirrhosis
investigations also suggests poor correlation of anti–factor Xa values and therapeutic anticoagulant effect among cirrhotics.6–8 A higher number of bleeding events were observed among patients with cirrhosis compared with patients without cirrhosis identified by the transfusion of blood products, use of protamine for heparin reversal, and alterations of hematological values. This difference was observed despite a majority of patients with cirrhosis receiving heparin therapy by a specified dosing protocol. The number of patients with cirrhosis managed off-protocol was attributed to a variable practice among members of the institution’s transplant team whereby conservative aPTT values are maintained during heparin anticoagulation to reduce the risk of bleeding events. Similar bleeding events were identified when evaluating the subgroup of patients with cirrhosis managed solely by anti– factor Xa levels compared with noncirrhotics. Our analysis of the relationship between anti-Xa and aPTT values among patients with cirrhosis during heparin anticoagulation offers a possible explanation for the increased use of blood products. Among patients without cirrhosis, therapeutic anti-Xa values closely approximated established institutional therapeutic aPTT values of 61–112 seconds. Much higher aPTT values (100–200 seconds), however, were seen with corresponding therapeutic anti-Xa values in patients with cirrhosis. If conversely analyzed, therapeutic aPTT values of 61–112 seconds corresponded to much lower anti-Xa values of 0.15–0.35 units/mL suggesting an altered relationship as a result of decreased synthesis of most procoagulants in cirrhotics. This study was limited to a description of the discrepancy between anti-Xa and aPTT values among patients with cirrhosis. Additional analysis of safety aspects of unfractionated heparin use was done as surrogate markers of clinically significant bleeding. Correlated anti-Xa and aPTT values obtained within the 15-minute window, however, were only available in a small percentage of patients. Transfusions were also administered based on the clinical judgment of physician involved in the
TABLE 3. Relationship of Anti-Xa and aPTT Values Among Patients With Cirrhosis and Patients Without Cirrhosis Cirrhotics aPTT Values
Anti-Xa values Supratherapeutic values (.0.7 units/dL) Therapeutic values (0.3–0.7 units/dL) Subtherapeutic values (,0.3 units/dL)
Subtherapeutic Values (,61 s), n (%)
Therapeutic Values (61–112 s), n (%)
Supratherapeutic Values (.112 s), n (%)
2 (2) 15 (17) 26 (30)
27 (31) 15 (17) 1 (1)
1 (1) 1 (1) 0 (0)
Noncirrhotics aPTT Values
Anti-Xa values Supratherapeutic values (.0.7 units/dL) Therapeutic values (0.3–0.7 units/dL) Subtherapeutic values (,0.3 units/dL)
Subtherapeutic Values (,61 s), n (%)
Therapeutic Values (61–112 s), n (%)
Supratherapeutic Values (.112 s), n (%)
6 (4) 16 (9) 43 (25)
21 (12) 48 (28) 19 (11)
13 (8) 3 (2) 1 (1)
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
43
Fuentes et al
Ther Drug Monit Volume 37, Number 1, February 2015
medical management of the patient. Analysis was conducted in a retrospective manner, which can introduce selection bias and limits the determination of causal relationships. Until coagulation tests better suited for monitoring in this population enter clinical practice, further study is required to establish safe and effective anticoagulation strategies, as clinical efficacy could not be accurately evaluated in a retrospective fashion.
vein thrombosis in patients with liver cirrhosis. J Clin Gastroenterol. 2010;44:448–451. Sogaard KK, Horvath-Puho E, Gronbaek H, et al. Risk of venous thromboembolism in patients with liver disease: a nation wide populationbased case-control study. Am J Gastroenterol. 2009;104:96–101. Greaves M. Limitations of the laboratory monitoring of heparin therapy. Thromb Haemost. 2002;87:163–164. Bechmann LP, Sichau M, Wichert M, et al. Low-molecular-weight heparin in patients with advanced cirrhosis. Liver Int. 2011;31:75–82. Sette H, Hughes RD, Langley PG, et al. Heparin response and clearance in acute and chronic liver disease. Thromb Haemost. 1985;54:591–594. Senzolo M, Rodriguez-Castro KI, Radu RC, et al. Increased anticoagulant response to low-molecular-weight heparin in plasma from patients with advanced cirrhosis. J Thromb Haemost. 2012;10:1823–1829. Brill-Edwards P, Ginsberg JS, Johnston M, et al. Establishing a therapeutic range for heparin therapy. Ann Intern Med. 1993;119:104–109. Charco R, Fuster J, Fondevila C, et al. Portal vein thrombosis in liver transplantation. Transplant Proc. 2005;37:3904–3905.
REFERENCES 1. Tripodi A, Primignani M, Chantarangkul V, et al. An imbalance of provs anti-coagulation factors in plasma from patients with cirrhosis. Gastroenterology. 2009;137:2105–2111. 2. Dabbagh O, Oza A, Prakash S. Coagulopathy does not protect against venous thromboembolism in hospitalized patients with chronic liver disease. Chest. 2010;137:1145–1149. 3. Amitrano LA, Guardascione MA, Menchise A, et al. Safety and efficacy of anticoagulation therapy with low molecular weight heparin for portal
44
4.
5. 6. 7. 8.
9. 10.
Copyright Ó 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
