Regression line for Thrombosil I on ACL 3000 coagulometer. The activated partial thromboplastin time (aPTT) therapeutic range is determined by the points (dashed lines) on the regression line (solid line) that correspond to anti–factor Xa heparin levels of 0.3 to 0.7 U/mL. See the "Procedures" subsection of the "Patients and Methods" section and Table 1 for manufacturers of reagents and coagulometers.
Activated partial thromboplastin time (aPTT) ratios corresponding to therapeutic anti–factor Xa heparin levels of 0.3 to 0.7 U/mL for various reagents on different coagulometers. See the "Procedures" subsection of the "Patients and Methods" section and Table 1 and Table 2 for manufacturers of reagents and coagulometers, respectively.
Activated partial thromboplastin time (aPTT) ratios corresponding to therapeutic anti–factor Xa heparin levels of 0.3 to 0.7 U/mL for various reagents on each of the 4 coagulometers used. See the "Procedures" subsection of the "Patients and Methods" section and Table 1 and Table 2 for manufacturers of reagents and coagulometers, respectively.
Bates SM, Weitz JI, Johnston M, Hirsh J, Ginsberg JS. Use of a Fixed Activated Partial Thromboplastin Time Ratio to Establish a Therapeutic Range for Unfractionated Heparin. Arch Intern Med. 2001;161(3):385-391. doi:10.1001/archinte.161.3.385
The commonly recommended therapeutic range for patients receiving unfractionated heparin of 1.5 to 2.5 times the control activated partial thromboplastin time (aPTT) is not universally applicable. It has been suggested that the therapeutic range for each aPTT reagent should be based on plasma heparin levels. We sought to identify an aPTT ratio that corresponds to therapeutic anti–factor Xa heparin levels for combinations of several reagents and coagulometers that are commonly used.
Citrated plasma was collected from 126 unselected patients receiving unfractionated heparin. Four automated coagulometers and 6 commercial aPTT reagents were used to measure the aPTT. Plasma anti–factor Xa levels were measured by means of a commercially available assay. The relationship between the aPTT results and anti–factor Xa heparin levels for each reagent-coagulometer combination was determined by linear regression analysis, and the aPTT results corresponding to therapeutic anti–factor Xa heparin levels were calculated.
For all reagent-coagulometer combinations studied, an aPTT ratio of 1.5 resulted in anti–factor Xa heparin levels considerably below the lower limit of the therapeutic range. When the aPTT was performed on any of the coagulometers assessed with the use of Actin (Dade Diagnostics, Aguada, Puerto Rico) and IL Test (Instrumentation Laboratories, Fisher Scientific, Unionville, Ontario) reagents, aPTT ratios necessary to achieve therapeutic anti–factor Xa heparin levels approximated 2.0 to 3.5.
For laboratories that cannot perform heparin levels, the use of less responsive reagents and any of the coagulometers studied, along with target aPTT ratio between 2.0 and 3.5, appears to be a reasonable alternative.
ALTHOUGH THE use of low-molecular-weight heparin is increasing, unfractionated heparin is still widely used to treat venous and arterial thromboembolic disorders. Because the anticoagulant response to unfractionated heparin varies among patients1- 4 and its efficacy and safety are thought to be optimal when a target therapeutic range is achieved,5- 8 laboratory monitoring with dose adjustment is necessary to ensure that an appropriate level of anticoagulation is given. The activated partial thromboplastin time (aPTT), a clotting assay that reflects the ability of the heparin-antithrombin complex to inactivate thrombin, factor Xa, and other coagulation enzymes within the intrinsic pathway, is the most widely used laboratory test for monitoring heparin therapy9 because it is widely available, rapid, easily automated, simple to perform, and relatively inexpensive.
There are, however, several problems associated with the use of the aPTT to monitor heparin therapy. No aPTT standard exists. Most medical textbooks and many experts recommend a therapeutic range of 1.5 to 2.5 times the control value (the mean aPTT obtained by testing a minimum of 20 plasma samples from healthy persons).8,10 This recommendation is based largely on 2 studies. In the first, a prospective cohort study by Basu and colleagues,8 patients with recurrent thromboembolism during heparin treatment were more likely to have an aPTT less than 1.5 times the control value than those without recurrence. The second study demonstrated that a heparin level range of 0.2 to 0.4 U/mL, as measured by protamine sulfate titration, was most effective at inhibiting thrombus growth in an animal model.11 A level of less than 0.15 U/mL was associated with increased fibrinogen accretion, whereas a heparin level of greater than 0.5 U/mL was associated with an increased risk of bleeding.11 This range of heparin level based on protamine sulfate titration corresponded to an aPTT range of 1.5 to 2.5 times control for the reagent used at that time.8 However, the responsiveness of different commercial aPTT reagents and different lots of the same reagent is variable,9,12- 21 and an aPTT ratio of 1.5 to 2.5 times control is not likely to be universally applicable.
The potential to inappropriately dose patients with heparin is a consequence of this variability in responsiveness to heparin. Two strategies have the potential to solve the problem of variability in therapeutic aPTT ranges for different aPTT reagents and coagulometers. The first is to abandon the aPTT and routinely monitor heparin therapy by protamine sulfate titration (therapeutic range, 0.2-0.4 U/mL)11 or by chromogenic anti–factor Xa assay (therapeutic range, 0.3-0.7 U/mL).10 Neither method is practical on a routine basis because both are expensive and many laboratories are not equipped to perform them regularly. The second solution is to continue using the aPTT, but to establish an accurate therapeutic range for each reagent and each coagulometer by performing aPTT and heparin level measurements by anti–factor Xa or protamine sulfate titration on the plasma of at least 30 heparin-treated patients and, by linear regression, calculating the aPTT range that corresponds to the heparin level therapeutic range.15 This, however, is onerous and does not obviate the need to measure heparin levels. Therefore, to optimize the monitoring of heparin therapy, a simple means of determining the aPTT therapeutic range is critically needed.
The purpose of this study was to simplify heparin monitoring. To accomplish this, citrated plasma was collected from patients receiving unfractionated heparin and the aPTT was measured by means of commonly used aPTT reagents and coagulometers. Concomitant anti–factor Xa heparin levels were also determined so that, for each aPTT reagent–coagulometer combination, the aPTT values that resulted in therapeutic anti–factor Xa heparin levels, as well as anti–factor Xa heparin levels corresponding to an aPTT ratio of 1.5 to 2.5, could be identified. Using this information, we sought to identify whether the use of certain reagent-coagulometer combinations would yield a single fixed aPTT ratio that closely approximates the therapeutic range as established by anti–factor Xa level measurement.
Plasma samples were collected from 126 unselected patients at 3 hospitals. Patients were receiving unfractionated heparin for the treatment of either venous thromboembolic disease or acute coronary syndromes or for the prevention of thromboembolism in the setting of atrial fibrillation. None of the patients was receiving warfarin at the time of blood sampling. Samples for the determination of control values were obtained from healthy volunteers without known coagulation abnormalities who were not receiving anticoagulants.
Venous blood samples were collected in 5-mL specimen tubes (BD Vacutainers; Becton Dickinson Co, Mountain View, Calif) prefilled with 0.5 mL of 3.2% (0.105-mol/L) buffered sodium citrate. After sedimentation of the red blood cells by centrifugation at 1700g for 15 minutes at 4°C, the harvested plasma was subjected to a second centrifugation under the same conditions to ensure complete removal of platelets. Aliquots of platelet-poor plasma were then pipetted into polystyrene tubes that were maintained at −70°C until assayed. Four automated coagulation instruments (MLA-700 [Medical Laboratory Instrumentation, Pleasantvillle, NY], ACL 3000 [Fisher Scientific, Unionville, Ontario], MDA-180 [Organon Teknika, Durham NC], and STA Compact [Diagnostica Stago, Asnières, France]) were used for aPTT determinations performed according to each manufacturer's specifications. Six commercial aPTT reagents were used (Table 1). Plasma anti–factor Xa heparin levels were measured by the method of Teien and Lie22 with the use of the ACL 3000 instrument and a commercially available assay (Stachrom Heparin; Diagnostica Stago). The therapeutic range for unfractionated heparin with this assay is 0.3 to 0.7 U/mL.10,23
For each reagent and coagulometer combination, control values were determined by calculating the mean aPTT results for plasma samples obtained from a minimum of 20 control subjects. The normal range, corresponding to the control value ± 2 SDs, was calculated for each reagent-coagulometer combination.
The relationship between the aPTT results and ex vivo anti–factor Xa heparin levels (both derived from aliquots of the same plasma sample) for each reagent-coagulometer combination was determined by linear regression analysis. The aPTT results corresponding to anti–factor Xa heparin levels of 0.3 to 0.7 U/mL were calculated from the ordinate values for the points on the regression line corresponding to these 2 heparin levels. By dividing these aPTT results by the reagent's mean control value on that coagulometer, corresponding aPTT ratios were calculated. The anti–factor Xa heparin levels corresponding to aPTT ratios of 1.5 to 2.5 were calculated from the abscissa values for the points on the regression line at these aPTT values. The Pearson correlation coefficient (r) was used to assess the extent of linear correlation between the aPTT and anti–factor Xa heparin level.
Figure 1 shows an example of a regression line. The y-intercept, slope, and Pearson correlation coefficient (r) for each regression line are listed in Table 2. The correlation between anti–factor Xa heparin levels and the aPTT was good (r = 0.64 to 0.95). However, the assumed linear relationship between anti–factor Xa heparin levels and the aPTT did not always hold true, and, as a consequence, the y-intercept on occasion differs from the control value. The control values and normal ranges for each reagent-coagulometer combination are shown in Table 3. The control aPTT values are similar regardless of the reagent or instrument used.
Table 4 shows the aPTT ratios corresponding to anti–factor Xa heparin levels of 0.3 to 0.7 U/mL for each reagent-coagulometer combination. The results obtained on the STA-Compact coagulometer with a second lot of Thrombosil I (lot ITH237) were nearly identical to those for lot ITH250 (data not shown). The 6 reagents differed in terms of heparin responsiveness, with Actin, IL Test, and Thrombosil I appearing less responsive than Actin FSL, Actin FS, and Pathromtin, in terms of the aPTT ratio necessary to attain a therapeutic anti–factor Xa heparin level. The variation in aPTT ratios necessary to achieve anti–factor Xa heparin levels of 0.3 to 0.7 U/mL appeared to be greater when different reagents were used on the same coagulometer, than when the same reagent was used with a different coagulometer (Figure 2 and Figure 3). When the aPTT was performed on any of the 4 coagulometers with the use of Actin or IL Test, aPTT ratios necessary to achieve therapeutic anti–factor Xa heparin levels approximated 2.0 to 3.5 (Figure 2 and Figure 3). The aPTT ratios corresponding to therapeutic anti–factor Xa heparin levels were higher for Actin FSL, Actin FS, and Pathromtin SL on all of the coagulometers.
Table 5 shows anti–factor Xa heparin levels corresponding to aPTT ratios of 1.5 to 2.5 for the various reagent-coagulometer combinations. Again, the results obtained on the STA Compact coagulometer with a second lot of Thrombosil I (lot ITH237) were nearly identical to those for lot ITH250 (data not shown). With aPTT ratios of 1.5 to 2.5, the therapeutic range used by many laboratories, variable anti–factor Xa heparin levels were achieved. The variation in ex vivo anti–factor Xa heparin levels achieved with a ratio of 1.5 to 2.5 appeared to be greater when different reagents were used on the same coagulometer, than when the same reagent was used with a different coagulometer (Table 5). For all reagent-coagulometer combinations, an aPTT ratio of 1.5 corresponded to anti–factor Xa heparin levels considerably below the targeted lower limit of 0.3 U/mL.
Our results suggest that, with currently available reagents and coagulometers, the aPTT ratio that corresponds to an anti–factor Xa heparin level of 0.3 to 0.7 U/mL is highly variable. As such, the common practice of recommending a single "therapeutic" aPTT ratio provides no assurance that target anti–factor Xa heparin levels will be achieved. This study also confirms the observation made by Brill-Edwards et al15 that a therapeutic range with a lower limit set by an aPTT ratio of 1.5 consistently results in subtherapeutic heparin levels. In addition, our data demonstrate that this observation holds true regardless of the coagulometer used.
The previously offered solution of establishing a therapeutic range for aPTT results for each reagent with the use of anti–factor Xa or protamine sulfate titration heparin levels as a reference standard15 is impractical for most individual laboratories, as they are not equipped to perform these assays and do not have access to plasma samples from at least 30 heparin-treated patients. Ideally, manufacturers should provide a therapeutic range based on anti–factor Xa or protamine sulfate titration heparin levels for their reagent with commonly used coagulometers, much the same way as International Sensitivity Indexes are provided for thromboplastin reagents. Failing that, central reference laboratories with access to patient plasma could provide individual laboratories with therapeutic ranges based on their reagent-coagulometer combination. Third, laboratories could choose to use less heparin-responsive reagents, such as Actin or IL Test. With these reagents, an aPTT ratio of 2.0 to 3.5 appears to encompass therapeutic anti–factor Xa heparin levels with the use of several coagulometers. The use of this aPTT ratio with more responsive reagents would be expected to result in subtherapeutic anti–factor Xa heparin levels, and for these reagents the therapeutic range would need to be established by means of ex vivo anti–factor Xa heparin levels from heparinized patients as a reference standard.
Previous work has indicated that the sensitivity of an aPTT reagent to heparin depends both on its phospholipid content and on the nature of the activator present.19,24,25 The concentration of total phospholipid in the aPTT reagent is known to govern the sensitivity of the assay to the presence of nonspecific inhibitors.26- 28 In general, the less heparin-responsive reagents for which an aPTT ratio of 2.0 to 3.5 encompasses therapeutic anti–factor Xa heparin levels are relatively insensitive to lupus anticoagulants. Consequently, laboratories that use one of these reagents may need to use a different reagent for heparin monitoring than that used for lupus anticoagulant screening.
This study has limitations. Data supporting the clinical relevance of a therapeutic range for heparin therapy based on either protamine sulfate titration or anti–factor Xa heparin levels are based on only a very small number of animal and clinical studies.8,11,29 Most of these studies have used heparin levels from protamine sulfate titration 8,11; not all authors have found anti–factor Xa heparin levels of 0.3 to 0.7 U/mL to be equivalent to protamine sulfate titration levels of 0.2 to 0.4 U/mL.18 However, in a randomized controlled trial, the incidences of recurrent thrombosis and bleeding were not significantly higher in patients whose dose of heparin was adjusted to maintain an anti–factor Xa heparin level of 0.3 to 0.7 U/mL than they were in those whose heparin dose was adjusted to maintain an aPTT equivalent to a protamine sulfate titration heparin level of 0.2 to 0.4 U/mL,29 suggesting that these 2 ranges are equivalent. Variability in heparin responsiveness among different lots of the same aPTT reagent is well documented.9,12- 21 However, unless major differences exist among the lots, this variability is unlikely to have a major impact on our findings. In support of this statement, no difference was seen between the 2 lots of Thrombosil I tested. Given the small number of reagents used in this study, it is not possible to state with certainty that all lupus anticoagulant–insensitive reagents will have the same responsiveness to heparin. Only 6 aPTT reagents were studied, and it is possible that those selected may not be representative of all reagents. Those chosen are, however, widely used.
Differences in the heparin responsiveness of aPTT reagents increase the difficulty and expense of determining the therapeutic range for heparin therapy. Therapeutic ranges for various aPTT reagent–coagulometer combinations could be provided by reagent manufacturers or central reference laboratories. Alternatively, the use of less heparin-responsive reagents, such as Actin and IL Test, and a therapeutic range corresponding to 2.0 to 3.5 times the control aPTT value might simplify the procedure in institutions that cannot measure anti–factor Xa or protamine sulfate titration heparin levels or in those where access to plasma samples from patients treated with heparin is limited.
Accepted for publication September 21, 2000.
Dr Bates is a recipient of a Research Fellowship from the Heart and Stroke Foundation of Ontario, Toronto. Drs Weitz and Ginsberg are recipients of Career Investigator Awards from the Heart and Stroke Foundation of Ontario, Ottawa. This work was supported in part by the Thrombosis Interest Group, Mississauga, Ontario.
We thank Patrick Brill-Edwards, MD, and Jim Julian, PhD, for helpful comments and suggestions.
Corresponding author and reprints: Shannon M. Bates, MDCM, Thromboembolism Unit, HSC 3W15, McMaster University Medical Centre, 1200 Main St W, Hamilton, Ontario, Canada L8N 3Z5 (e-mail: firstname.lastname@example.org).