A weight-adjusted bolus of heparin (70 to 100 IU/kg) is recommended to achieve adequate anticoagulation during percutaneous coronary intervention (PCI). Proper dosing is mandatory to avoid bleeding and thrombotic complications. We investigated whether sensitivity to heparin is affected by difference in race. We performed a retrospective study with 874 consecutive PCI cases in our catheterization laboratory. The amount of initial heparin bolus (international units) per weight and subsequent activated clotting time (ACT; seconds) were obtained. Patients were divided into 4 categories based on race: Asian, African-American, European-American, and Hispanic. Multiple regression analysis was performed to validate the variables that determine the ACT. After adjusting for these variables, analysis of variance revealed the presence of a significant racial difference in ACT (p = 0.002). Successively, Student-Newman-Keuls test and Bonferroni t test revealed that Asian patients have higher ACT levels compared to other racial groups (p <0.03 for Asian vs others, p >0.26 between non-Asian groups). There was a positive relation between ACT and Asian race (p = 0.0004). Further analyses showed that Asians require 10 IU/kg less heparin per weight than other racial groups to achieve the same goal of ACT. In conclusion, decreased heparin dosing should be considered for Asian patients undergoing PCI.
During percutaneous coronary intervention (PCI), unfractionated heparin is administered to prevent thrombus formation. A strong inverse relation is demonstrated between the activated clotting time (ACT) and abrupt vessel closure, whereas higher ACT levels are associated with bleeding complications. Current PCI guidelines recommend an initial bolus of heparin 70 to 100 IU/kg to reach an ACT of 300 to 350 seconds with the Hemochron device. Racial difference in genetic polymorphism that may affect heparin binding with antithrombin III has been reported. However, there has been no clinical study that directly compared heparin sensitivity among different racial groups in vivo. We hypothesized that race might affect the heparin susceptibility determined by ACT.
Methods
This retrospective study was performed at an urban teaching hospital in New York. All patients who underwent elective or ad hoc PCI from December 2007 to October 2008 at our institution were eligible for the study. Patients who presented with acute coronary syndrome and/or were started on anticoagulation or glycoprotein IIb/IIIa inhibitors before the procedure were excluded. Clinical and demographic information, heparin dosing (international units per kilogram) calculated from the heparin dose and patient weight, and ACT (seconds) were collected from 874 consecutive PCI cases in our catheterization laboratory. Data of in-hospital events including death, myocardial infarction, urgent target vessel revascularization, and bleeding were obtained. PCI was performed at the operator’s discretion, and the amount of heparin given was decided by the operator with approximations for upper weight ranges. ACTs were determined in the cardiac catheterization laboratory. At least 2 ml of blood was collected from the sheath for ACT measurement 5 to 10 minutes after heparin was administered. ACT was measured in the procedure room using GEM PCL plus (Instrumentation Laboratory, Barcelona, Spain). The device was standardized weekly using whole-blood quality-control products from the manufacturer to keep the coefficient of variation <10%. The first ACT after the heparin bolus was used for study purposes.
Patients were divided into 4 racial groups: Asian, African-American, European-American, and Hispanic. The first multivariable regression analysis without effect of race was performed between ACT and heparin per weight, diabetes status, age, gender, patient weight, serum creatinine, total protein, current smoking, and time from heparin injection to ACT measurement. After adjusting for effects of these variables, analysis of variance was performed for residuals to see if there was remnant racial difference in ACT. Student-Newman-Keuls test and Bonferroni t test were performed successively to reveal which racial groups had different responses to heparin. The second multiple regression analysis, taking the difference of race into consideration, was performed to establish a better correlation formula.
Baseline characteristics were expressed as percentage for discrete variables or as mean ± SD for continuous variables. Discrete variables were compared by chi-square or Fisher’s exact tests. Continuous variables were compared by analysis of variance. Independence of each variable was evaluated by a variance inflation factor. A variance inflation factor <5 was used as a cutoff for independence. A p value <0.05 was considered statistically significant. Statistical analysis was performed using Excel (Microsoft, Seattle, Washington) with statistical analysis add-ons.
Results
The study population consisted of 874 patients who underwent PCI. Table 1 presents the main patient characteristics. Data of weight or heparin dose were missing for 13 patients (3 Asians, 4 African-Americans, 2 Hispanics, and 4 European-Americans).
| Variable | Asian | African-American | Hispanic | European-American | p Value |
|---|---|---|---|---|---|
| (n = 194) | (n = 220) | (n = 220) | (n = 242) | ||
| Mean age (years) | 64 ± 11 | 66 ± 13 | 66 ± 10 | 66 ± 12 | 0.16 |
| Men | 140 (72%) | 125 (57%) | 116 (53%) | 175 (72%) | <0.05 |
| Diabetes mellitus | 92 (47%) | 123 (56%) | 131 (60%) | 83 (34%) | <0.05 |
| Current smoker | 104 (53%) | 60 (27%) | 35 (16%) | 58 (24%) | <0.05 |
| Serum creatinine (mg/dl) | 0.87 ± 0.13 | 0.86 ± 0.13 | 0.88 ± 0.12 | 0.90 ± 0.12 | <0.05 |
| Total protein (mg/dl) | 7.2 ± 0.4 | 7.3 ± 0.5 | 7.2 ± 0.5 | 7.2 ± 0.5 | 0.07 |
| Heparin (IU) | 4,263 ± 874 | 4,925 ± 814 | 4,557 ± 945 | 4,927 ± 1,050 | <0.001 |
| Weight (kg) | 71 ± 14 | 84 ± 17 | 78 ± 17 | 87 ± 20 | <0.001 |
| Heparin per weight (IU/kg) | 60 ± 10 | 60 ± 8 | 60 ± 10 | 57 ± 9 | 0.001 |
| Measurement time (minutes) | 7.8 ± 2.1 | 7.9 ± 1.8 | 7.5 ± 2.3 | 7.8 ± 2.0 | 0.20 |
| Activated clotting time (seconds) | 293 ± 49 | 282 ± 49 | 277 ± 50 | 272 ± 53 | <0.001 |
The first multiple regression analysis before including effect of race revealed a positive relation between ACT and heparin per weight (p <0.000 01), whereas the other factors were not significantly related to ACT ( Table 2 ). Thus, the best model for estimating ACT level without factor of race was as follows ( Figure 1 ): ACT = 1.54 × (heparin per weight) + 189 (n = 874, adjusted R 2 = 0.071).
| Variable | Coefficient | 95% Confidence Interval | p Value | Variance Inflation Factor |
|---|---|---|---|---|
| Age (years) | +0.08 | −0.23 to +0.39 | 0.61 | 1.15 |
| Male gender | −2.93 | −10.1 to +4.2 | 0.41 | 1.10 |
| Diabetes mellitus | +0.18 | −6.46 to +6.82 | 0.96 | 1.03 |
| Current smoking | +1.18 | −5.76 to +8.13 | 0.73 | 1.02 |
| Serum creatinine (mg/dl) | −0.96 | −27.0 to +25.1 | 0.94 | 1.00 |
| Total protein (mg/dl) | −3.79 | −10.2 to +2.65 | 0.24 | 1.01 |
| Weight (kg) | −0.003 | −0.22 to +0.21 | 0.98 | 1.42 |
| Measurement time (minutes) | +0.94 | −0.64 to +2.52 | 0.24 | 1.00 |
| Heparin per weight (IU/kg) | +1.53 | +1.12 to +1.93 | <0.00001 | 1.21 |
To adjust for possible confounding factors, the residual for each patient (i.e., difference between observed and estimated ACTs by the first multiple regression analysis) was calculated. Analysis of variance for residuals revealed a significant p value of 0.002, displaying remnant racial difference in achieved ACT levels. Successively, Student-Newman-Keuls test and Bonferroni t test revealed that Asian patients achieved higher ACT levels compared to other racial groups (p <0.03). There was no significant difference between non-Asian racial groups by the 2 methods ( Figure 2 ).
