Standardization of Anti-Xa Assay and its Comparison with Activated Partial Thromboplastin Time for Monitoring Unfractionated Heparin Therapy

Abstract

Monitoring of anticoagulant activity of unfractionated heparin (UFH) is primarily done with activated partial thromboplastin time (aPTT), which is affected by many factors. Anti-Xa assays are considered to overcome these factors and may provide a better method for monitoring patients on UFH with a narrow therapeutic range. This study aimed to compare the effectiveness of aPTT and anti-Xa assays in UFH monitoring. A prospective non-randomized study was carried out in two stages: first, the anti-Xa assay was standardized using kit instructions; each sample was then analyzed by both tests. The outcomes of the two assays were compared and assessed for agreement of maintaining therapeutic anticoagulant levels. These levels for anti-Xa assay were between 0.3 and 0.7 IU/ml, while it was 1.5–2.5 times the control for aPTT assay. Below this range was regarded as subtherapeutic, and above this as supratherapeutic. A total of 90 samples were tested and analyzed using both assays. Most of them (> 70%) were noted to be in subtherapeutic levels with both tests. The overall concordance was 73.3%, and the estimated kappa value was 0.483 (0.396–0.57). The correlation between aPTT and anti-Xa assay was 0.74 (p < 0.001). With anti-Xa levels in the therapeutic range, aPTT levels were in subtherapeutic in 60% and supratherapeutic in 13.3% cases. Although both the testing strategies had a good agreement and correlation, discordance was observed in interpretative values with anti-Xa levels in therapeutic range and aPTT levels in non-therapeutic range. Its clinical implications need to be evaluated further in future studies.

Introduction

Heparins are the most common anticoagulants prescribed to reduce or prevent the formation of thrombus in those at high risk of developing them. It is a heterogeneous mixture of sulfated mucopolysaccharides, primarily available in two preparations for clinical use: (i) unfractionated heparin (UFH) and (ii) low molecular weight heparin (LMWH). The average molecular weight of UFH is between 15,000 and 18,000 Daltons, whereas LMWH is 3000 Daltons [1]. UFH is frequently used for a variety of conditions, including deep vein thrombosis (DVT), pulmonary embolism (PE), venous thromboembolism (VTE), acute coronary syndrome, unstable angina, coronary angioplasty, coronary thrombolysis, preventing mural thrombosis after MI and other thrombotic diseases [2]. The binding of unfractionated heparin induces a conformational change in antithrombin (AT) and increases antithrombin's ability to block factor Xa and thrombin by 1000 folds [3]. The chain length of heparin molecules with fewer than 18 saccharides prevents them from bridging between thrombin and antithrombin. So, they are incapable of stopping thrombin. Heparin must bind AT to inhibit factor Xa, whereas thrombin must be inactivated by a complex involving heparin, AT, and thrombin. So, a smaller molecule of heparin can inactivate factor Xa, but a larger molecule of heparin is required for thrombin inactivation [2, 4]. UFH similarly inhibits thrombin and factor Xa, while LMWH has more factor Xa inhibitory activity than thrombin [5].

Heparin has a narrow therapeutic range, with inadequate doses causing thrombosis and excess doses leading to haemorrhage. Anticoagulant response to UFH is highly variable between individuals [3]. Achieving adequate anticoagulation for optimizing patient outcomes with UFH is more difficult than with LMWH. Therefore, careful laboratory monitoring is necessary. The two most common techniques for measuring heparin are the activated partial thromboplastin time (aPTT) and the anti-Xa assay. The aPTT test is easy, affordable, and widely accessible. However, it lacks standardization because the analyzer and reagent employed greatly influence the UFH-induced extension of the aPTT [6]. The accepted therapeutic range is 1.5–2.5 times the control value. Each laboratory should establish its own therapeutic range for aPTT based on its own setup. On the other hand, the chromogenic anti-Xa test is more expensive than aPTT and only available in fewer centers, even though it has a unique therapeutic range for UFH between 0.30 and 0.70 IU/ml [6].

This study aimed to standardize the anti-Xa assay and compare it with the existing aPTT assay for monitoring UFH.

Material and Methods

This was a prospective non-randomized study conducted from August 2021 to April 2022 in the Hematology Section of the Department of Pathology, JIPMER. The study was approved by the postgraduate research monitoring committee and the institute ethics committee (JIP/IEC/2021/096). The cases enrolled were consecutive routine samples sent to the coagulation laboratory from patients who received UFH during the study period. Patients on concurrent other anticoagulants like warfarin, LMWH, or direct thrombin inhibitors, as well as those with known factor deficiencies that can affect aPTT, were excluded from this study.

The investigation was carried out in two stages: first, the anti-Xa test was standardized by the kit instructions; then, each sample was analyzed by both the assays; and finally, the outcomes of the two assays were compared.

The three kits-calibrator kit, control kit, and liquid anti-Xa reagents were used to standardize anti-Xa assay. Following the instructions in the kit insert (HemosiL, Instrumentation laboratories, Lisses, France), the standard operating procedure for reconstitution, reagent preparation, and standardization was carried out. Heparin is present in calibrators 1, 2, and 3 at concentrations of 0.0, 0.8, and 2.0 IU/ml, respectively. The calibrators were loaded into the automated system (ACL TOP 300) to create the standard curve. The heparin concentration (IU/ml) on the X-axis and the optical density (mAbs/min) on the Y-axis were used to create a standard curve.

Fresh blood samples sent in a trisodium citrate tube were accepted for the comparison of the anti-Xa assay with the aPTT test. Samples that had clotted, lysed and had quantities greater than 2 ml or less than 2 ml were rejected. For the assigned samples, a lab ID was made. After being centrifuged at 3500 rpm for 10 min, platelet-poor plasma (PPP) was collected. The ACL TOP 300 had been cleaned and its quality checked, then the appropriate reagents and samples were placed into their respective slots. On the same sample, the aPTT and anti-Xa tests were run simultaneously. Anti-Xa values were generated in IU/ml while aPTT values were obtained in seconds. The therapeutic range for aPTT is 1.5–2.5 times the control value, whereas, for anti-Xa, the therapeutic range was taken as 0.3–0.7 IU/ml. Values below that were considered subtherapeutic, while those over that were considered supratherapeutic.

Relevant history was collected from the online requisition/telephonically as part of the usual routine coagulation workup. The frequency of monitoring for both tests was the same; the usual recommended time is 6 h after the initiation of UFH therapy with subsequent 4–6 hourly intervals until the therapeutic range is achieved [7]. The frequency of monitoring was according to the sample sent and lab timings.

Data entered in Microsoft excel 2013 was subsequently exported to SPSS version 20. The dose of heparin and time interval between heparin administration and sample collection is summarized as median with interquartile range (IQR). Independent continuous variables like anti-Xa and aPTT values were summarized as means with standard deviation. Categorical variables like the interpretation of anti-Xa and aPTT values were summarized as percentages. The agreement between continuous variable aPTT (second) and anti -Xa (IU/ml) was studied by Bland–Altman plot and the agreement between categorical variables i.e., therapeutic level by both tests was studied by kappa statistics. The correlation between aPTT and anti-Xa assay was studied by Pearson’s correlation. All statistical tests were carried out at a 5% level of significance.

Results

Standardization of Anti-Xa Assay

We performed four replicates of each calibrator and assessed the absorbance to standardize the anti-Xa test. The mean absorbance was calculated to depict the standard curve, and the absorbance value was transformed into a log scale (Table 1). Anti-Xa concentration (IU/ml) on the X axis and absorbance (mAbs/min) on the Y axis were used to plot the standard curve. (Fig. 1).

Table 1.

Measured absorbances of calibrator 1, calibrator 2, and calibrator 3

Calibrator	Heparin concentration (IU/ml)	Replicate 1 (mAbs/min)	Replicate 2 (mAbs/min)	Replicate 3 (mAbs/min)	Replicate 4 (mAbs/min)	Mean OD	[log10 (log10(y)] (mAbs/min)
1	0.0	1079.43	1082.53	1088.25	1070.74	1080.24	0.48
2	0.8	548.54	556.93	549.59	545.00	550.01	0.43
3	2.0	239.94	221.11	219.42	213.01	223.37	0.37

Fig. 1.

Standard curve generated with heparin concentration (IU/ml) on the X-axis vs optical density (mAbs/min) on the Y-axis

Comparison of aPTT and Anti-Xa Assay

A total of 90 blood samples from patients who received UFH sent to the coagulation laboratory during the study period were assessed by both aPTT and anti-Xa assay. Heparin doses for these 90 samples ranged from 1000 to 12,500 Units, with a median of 5000 Units and IQR of 1000 Units. The interval between the injection of heparin and blood sample collection ranged from 1 h 30 min to 10 h, with a median interval of 3 h 25 min and an IQR of 3 h.

Distribution of Therapeutic Level Versus Nontherapeutic Level by Both Tests

The anti-Xa levels in 74.44% (67/90) samples were < 0.3 IU/ml (subtherapeutic level), whereas those in 16.66% (15/90) of the samples were within the range of 0.3–0.7 IU/ml (therapeutic range). Supra-therapeutic levels of > 0.7 IU/ml were noted in 8.88% (8/90) samples. The aPTT result in 72.22% (65/90) of the sample was at a subtherapeutic level (< 1.5 times the control value), and 15.55% of samples showed a therapeutic range (result within 1.5–2.5 times the control value). In 12.22% (14/90) samples it was in the supratherapeutic range (> 15–2.5 times the control) (Fig. 2).

Fig. 2.

Distribution of therapeutic levels of heparin

Comparison Between Therapeutic Level Versus Nontherapeutic Level by Both Tests

The proportion of samples in the therapeutic versus non-therapeutic level noted using both methods was compared (Table 2). The concordance for the therapeutic level was 26.7%, while it was much higher, 82.1% for the subtherapeutic and 87.5% for the supratherapeutic level. The overall concordance between the two tests was 73.3% (n = 66), and the non-concordance was 26.7% (n = 24).

Table 2.

Proportion of samples showing therapeutic and nontherapeutic levels for both the tests

Anti-Xa	aPTT
	0	1	2	Total
0	55 (82.1)	10 (14.9)	2 (3)	67
1	9 (60)	4 (26.7)	2 (13.3)	15
2	1 (12.5)	0 (0)	7 (87.5)	8
Total	65	14	11	90

(0-subtherapeutic level, 1-therapeutic level, 2-supratherapeutic level)

The most frequent cause of the discordance between the values when the paired values were analyzed was a therapeutic anti-Xa with concomitant subtherapeutic aPTT levels. When the anti-Xa levels were noted to be in the therapeutic range, the aPTT levels were in the subtherapeutic range in 9 (60%) patients, and in supratherapeutic levels in 2 (13.3%) patients. Anti-X-a levels were below therapeutic levels in 10 (14.9%) cases when the aPTT levels were within the therapeutic range. No cases with concurrent therapeutic aPTT and supratherapeutic anti-Xa were documented.

Anti-Xa assay is considered the gold standard. The validity of aPTT test was compared with anti-Xa for detecting a non-therapeutic range vs therapeutic range; aPTT test showed a sensitivity of 86.7% and specificity of 26.7% for detecting non-therapeutic range.

Agreement Between Therapeutic Level by aPTT and Anti-Xa Assay

Agreement between therapeutic levels by both the tests was assessed using kappa statistics. Keeping anti-Xa assay as the gold standard, the weighted kappa value (95% CI) was 0.483 (0.396–0.57), which represents a good agreement.

Agreement Between aPTT and Anti-Xa Assay

Agreement between continuous variable aPTT (seconds) and anti-Xa level (IU/ml) was done using the Bland–Altman plot. On observation, all values were between ± 2SD, indicating a good level of agreement (Fig. 3).

Fig. 3.

Bland–Altman agreement among the performance of anti-Xa assay versus aPTT for UFH monitoring

Correlation Between aPTT and Anti-Xa Assay

The assessment of the correlation between aPTT and anti-Xa assay using Pearson’s correlation showed a strong positive correlation between anti-Xa level and aPTT, with an r value of 0.7443, p < 0.001 (Fig. 4).

Fig. 4.

Pearson Correlation between anti-Xa and aPTT

Discussion

UFH, when administered to a patient, inhibits factors Xa and IIa in a 1:1 ratio. Due to its variation in anticoagulant response from patient to patient and its restricted therapeutic range, careful laboratory monitoring of UFH anticoagulation is mandatory. The two commonly used techniques for monitoring heparin therapy include activated partial thromboplastin time (aPTT) testing, and (ii) the anti-Xa assay. By conducting the study, the anti-Xa assay was standardized and introduced in our hospital for UFH monitoring and evaluated for any discernible advantages over the aPTT testing.

aPTT is an “in vitro” coagulation test commonly used to assess the intrinsic and common pathways of the coagulation cascade. The testing can be affected by the type of reagent, length of incubation, the method of measuring clot formation, and the individual disease process, such as liver disease, factor deficiencies (VIII, IX, and XII), or antiphospholipid syndromes. Given these challenges, anti-Xa assay is becoming the gold standard for monitoring therapeutic anticoagulation with UFH. The chromogenic assay, which measures the residual factor Xa activity that is inversely proportional to the heparin level.

Earlier studies suggested the therapeutic anti-Xa levels of 0.3–0.7 IU/ml, whereas the accepted therapeutic range of the aPTT is 1.5–2.5 times the control value for adequate anticoagulation. Most of the samples were noted to be in a subtherapeutic range rather than a therapeutic or supratherapeutic range. The overall concordance was 73.3%, and discordance was 26.7%. It should be noted that the discrepancy was when the anti-Xa levels were in the therapeutic range, with the aPTT levels in 60% in the subtherapeutic range. If the anticoagulation titration was based on the aPTT levels, it might result in bleeding due to excessive dosing. The converse is true when anti-Xa levels are subtherapeutic with the aPTT levels in the therapeutic range, which leads to underdosing where the risk of thrombosis remains high. Previous studies reveal a concordance rate from 35 to 60% [8–10].

One of the main limitations of this study is the small sample size, with multiple samples taken from the same patients on anticoagulation for a wide range of indications. This limited the assessment of the benefit of either method of testing to titrate dosing of anticoagulation and the clinical effects of the observed levels.

In conclusion, although both the testing strategies had a good agreement and correlation, discordance was observed in interpretive values especially with anti-Xa levels in the therapeutic range, and aPTT levels in the non-therapeutic range. The clinical implications of such discordance need to be evaluated further in future studies.

Authors Contribution

VT performed the research work and VT and JA analyzed the results under the guidance of RK; RK and AK designed the study; AK was the clinician-in-charge of the patients; VT prepared the manuscript which was edited by RK and was reviewed and approved by all co-authors.

Funding

An intramural grant was obtained for purchasing an anti-Xa assay kit (JIP/Res/Intramural/Phs 2/2021–2022). JIPMER intramural fund (JIP/Res/Intramural/Phs 2/2021–2022) for MSc MLT thesis.

Declarations

Ethical Approval

JIP/IEC/2021/096.