Masahiro Ieko, Department of Internal Medicine, School of Dentistry, Health Sciences University of Hokkaido, 1757-Kanazawa, Ishikari-Tobetsu, Hokkaido, 061-0293, Japan. Tel: +81 133 23 3070; fax: +81 133 23 1534. E-mail: email@example.com
Summary. Background: Lupus anticoagulant (LA) is an antibody that interferes with phospholipid-dependent coagulation reactions. Activated partial thromboplastin time (APTT) is widely used as a test for LA screening. APTT reagents are composed of activators, such as silica or ellagic acid, and phospholipids, and APTT reagents with silica are recommended for LA screening because of greater sensitivity. However, the effects of activators on LA activity have not been adequately investigated.
Objectives: In this study, we examined whether an ellagic acid-based reagent was highly sensitive to LA in a low phospholipid condition and useful for LA screening.
Methods: Silica-based (SL) and ellagic acid-based (EA) reagents were prepared in-house with the same composition and concentration of phospholipids, while the commercial APTT reagents APTT-SLA (SLA), Actin FSL (FSL), APTT-SP (SP) and PTT-LA (PTT) were also included in the study.
Results: The normal reference ranges for SL and EA were 30.1–47.0 and 28.0–40.2 s, respectively, while the cut-off index values for circulating anticoagulant activity (ICA) calculated from the results obtained with SL, EA, SLA, FSL, SP and PTT were 12.9, 11.5, 13.2, 15.6, 14.3 and 14.0, respectively. The sensitivity of those reagents based on those cut-off values was 91%, 96%, 68%, 46%, 91% and 86%, respectively.
Conclusions: Our results showed that the ellagic acid-based reagent was more sensitive to LA than silica-based reagents in a low phospholipid condition and had adequate sensitivity to detect LA. We concluded that the sensitivity of APTT reagents for LA is dependent on phospholipid concentration and not the activator.
Lupus anticoagulant (LA) is an antibody that interferes with one or more in vitro coagulation reactions, which are dependent on interactions with protein-phospholipid complexes. Due to the involvement of phospholipids in these epitopes, lupus anticoagulant is usually grouped in the antibody family that is known as antiphospholipid antibodies . Antiphospholipid antibodies have clinical significance because of their association with thrombosis, obstetrical complications, neurological problems and cutaneous manifestations. A widely used test for LA screening is activated partial thromboplastin time (APTT). Most commercial APTT reagents are composed of activators, such as silica or ellagic acid, along with various phospholipids. The sensitivity and specificity of APTT reagents to heparin, LA and coagulation factor deficiencies vary, because of differences in activators employed, such as silica and ellagic acid, as well as sources, properties and total concentrations of phospholipids. These points have been well documented in past reports with regard to sensitivity to LA [2,3].
The Subcommittee on Lupus Anticoagulants and Antiphospholipid Antibodies of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis has recommended criteria for diagnosis of LA. The recommendation is to perform two different tests that represent different assay principles. Diluted Russell Viper Venom time (dRVVT) is widely used in clinical laboratories and believed to be specific for detecting LA in patients who are thought to be at high risk of developing thrombosis. Any APTT test with silica as an activator and low phospholipid content is the second test of choice, because of its sensitivity for LA. Ellagic acid as an activator is not recommended, as it is insensitive to LA . Tripodi et al.  investigated the responsiveness of commercial APTT reagents for screening of LA-positive plasma samples and showed that the sensitivity of silica-based reagents to LA was higher than that of ellagic acid-based reagents.
However, the effects of activators, such as silica and ellagic acid, and phospholipid behavior on LA detection have not been adequately investigated. In addition, there is no comparative study of silica- and ellagic acid-based APTT reagents with the same composition and concentration of phospholipids with regard to LA sensitivity. According to the UK NEQAS proficiency testing exercise and questionnaire, 51 of 260 centers (19.6%) employed an APTT reagent that utilized ellagic acid activation . The purpose of the present study was to compare the sensitivity to LA of two types of APTT reagents prepared from silica and ellagic acid, respectively, with the same phospholipid condition. We also examined whether an ellagic acid-based reagent has a high sensitivity to LA in a low phospholipid condition to determine its usefulness for LA screening.
Materials and methods
Five different types of silica-based (SL) reagents and an ellagic acid-based (EA) reagent were prepared in our laboratory. In addition to these in-house reagents, APTT-SLA (SLA; Sysmex Corporation, Kobe, Japan), Actin FSL (FSL; Siemens Healthcare Diagnostics Inc., Marburg, Germany), APTT-SP (SP; Instrumentation Laboratory Company, Bedford, MA, USA) and PTT-LA (PTT; Diagnostica Stago, Asnieres, France) were included in this study as commercial APTT reagents. The characteristics of these reagents are shown in Table 1.
Table 1. Activators and phospholipid sources in each APTT reagent
Ellagic acid-based reagent
Phenol (Kishida Chemical Co. Ltd, Osaka, Japan), polyethylene glycol 6000 (Kishida Chemical Co. Ltd) and ellagic acid (Tokyo Chemical Industry Co., Ltd, Tokyo, Japan) of the highest commercially available grade were obtained. 2-amino-2-hydroxymethyl-1,3-propanediol (TRIS) was purchased from Nacalai Tesque Inc. (Kyoto, Japan). N-(2-hydroxyethyl) piperazine-N’-2-ethanesulfonic acid (HEPES) was purchased from Dojindo Laboratories (Kumamoto, Japan). Five different types of colloidal silica (A∼E) were purchased from Sigma-Aldrich® Corporation (St Louis, MO, USA).
The phospholipid solution, kindly provided by Sysmex Corporation, included the following three synthetic phospholipids: 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), 1,2-dioleoyl-sn-glycero-3-phosphocoline (DOPC), and 1,2-dioleoyl -sn-glycero-3-phospho-L-serine (DOPS). The composition of the phospholipid solution was the same as that of SLA, a commercial reagent available in Japan .
Preparation of APTT reagents
The ellagic acid-based reagent was prepared using the following method. Ellagic acid was dissolved in 0.35% (w/v) phenol to a final concentration of 0.1 mm and then mixed with a buffer solution composed of 50 mm HEPES and 25 mm TRIS (pH 7.35). A phospholipid solution and 1.0% (w/v) polyethylene glycol 6000 was added; then the mixture was suspended. SLA phospholipid solution was diluted twice, and the phospholipid concentration in the ellagic acid-based reagent was half of that in SLA.
In addition, five silica-based reagents were prepared using the same method with five different types of silica, Ludox® TMA colloidal silica (A), Ludox® TM-40 colloidal silica (B), Ludox® AM-30 colloidal silica (C), Ludox® LS colloidal silica (D) and Ludox® SM-30 colloidal silica (E) (Sigma-Aldrich® Corporation). The silica particles were sized 22, 22, 12, 12 and 7 nm, respectively, in diameter. The final concentration of silica was 0.45 mg mL−1 and the phospholipid concentration in the silica-based reagent was the same as that in the ellagic acid-based reagent.
Plasma (0.05 mL) was incubated for 1 min at 37 °C in a cuvette, to which 0.05 mL of APTT reagent was added, and then incubated for 3 min at 37 °C. CaCl2 (25 mm, 0.05 mL) was then added and clotting time was calculated. All APTT tests were performed in a cuvette made of polystyrene resin  using a Coagrex-800 (Shimadzu Corporation, Kyoto, Japan), which is a real-time random access intelligent coagulation analyzer with both light scattering and absorbance detection capabilities.
Normal (Coagtrol N; Sysmex Corporation) and LA-positive (Sunfco, Osaka, Japan) plasma samples were used to determine the normal clotting time and LA sensitivity of the five silica-based reagents. Clotting time was measured twice and the mean values of duplicate experiments were used.
We enrolled 22 LA-positive Japanese patients (mean age 46.0 years old; 3 men, 19 women) with various diseases including antiphospholipid syndrome, who were being treated in our medical institution. For confirmation of LA, a Staclot LA (Diagnostica Stago, Asnieres, France) examination was performed. In addition, we tested plasma samples from 41 LA-negative patients with prolonged APTT whose underlining diseases were hemophilia A, von Willebrand disease, acquired hemophilia A and liver dysfunction, and from 41 healthy subjects with normal coagulation test results as controls. Blood for coagulation studies was collected into tubes; then nine parts were added to one part of 3.13% trisodium citrate, a concentration that is very common in Japan. Platelet-poor plasma samples were obtained by single centrifugation at 1500 × g for 20 min at room temperature without taking plasma near buffy coat. Supernatants were collected in plastic capped tubes and stored frozen at −80 °C until testing. The normal reference range for each reagent was established by the geometric mean value ± 2 SD obtained from the clotting times of the samples from the 41 healthy subjects. The study protocol was approved by the Research Ethics Committee of Health Sciences University of Hokkaido. Written informed consent for all procedures was obtained from all patients and control subjects, prior to beginning the study. Clotting time for the samples from the healthy and LA-positive and LA-negative plasma samples was determined once.
Index of LA sensitivity
The index for circulating anticoagulant activity (ICA) was used to show the sensitivity of each APTT reagent to LA . ICA = (b−c)/a × 100, where a = APTT of patient plasma sample, b = APTT of 1:1 mixture of patient plasma and normal plasma, and c = APTT of normal plasma. CRYOcheck Pooled Normal Plasma (Precision BioLogic Inc., Dartmouth, Canada) was used as normal plasma mixed with the LA-positive and LA-negative samples. The ICA values for the 22 LA-positive plasma samples were calculated to compare the LA sensitivity of the reagents, while those of the 41 LA-negative plasma samples were calculated to determine the ICA cut-off value.
Data for the various parameters were compared using a Wilcoxon signed rank test. P-values below 0.05 were considered to be statistically significant.
Comparisons of characteristics of five silica-based reagents
Five silica-based reagents (A, B, C, D and E) were compared with the normal clotting time and the ICA as LA sensitivity. The ICA values of A, B, C, D and E were 37.0, 36.6, 36.0, 32.6 and 26.0, respectively, while the normal clotting times were 60.9, 43.1, 37.6, 51.4 and 38.9 s, respectively. The LA sensitivity of A, B and C was higher than that of D and E. The normal clotting time of C was 37.6 s, which was the closest normal clotting time to PTT-LA (35.9 s) and more acceptable than that of A and B.
Normal reference range determination
The normal reference ranges of SL and EA obtained with the 41 normal plasma samples were 30.1–47.0 and 28.0–40.2 s, respectively. In addition to the in-house reagents, those of SLA, FSL, SP and PTT were 25.6–32.8, 25.0–33.9, 30.4–46.1 and 33.6–48.1 s, respectively.
ICA cut-off value
Determination of the ICA cut-off value was established using 41 APTT prolonged LA-negative plasmas. The mean value + 2 SD of ICA was used as the cut-off for each reagent, which is shown in Table 2. The ICA cut-off values calculated for SL, EA, SLA, FSL, SP and PTT were 12.9, 11.5, 13.2, 15.6, 14.3 and 14.0, respectively.
Table 2. ICA cut-off value for each APTT reagent
Mean + 2SD
Determination of ICA cut-off values was performed using 41 APTT prolonged LA-negative plasma samples. The mean value + 2 SD of ICA was used as the cut-off value. APTT reagent abbreviations are the same as in Table 1.
Comparison of APTT values between normal, LA-negative and LA-positive samples
The APTT values for the normal, LA-negative and LA-positive samples were compared for each reagent (Fig. 1). The middle bars, lower bars and upper bars indicate median, minimum and maximum values in each reagent, respectively. The medians of SL, EA, SLA, FSL, SP and PTT with normal samples were 38.0, 35.2, 29.6, 29.7, 38.4 and 41.0. Those of LA-negative samples were 51.0, 43.0, 36.5, 35.1, 47.9 and 55.0, and those of LA-positive samples were 67.6, 69.6, 43.2, 45.0, 65.2 and 73.9, respectively.
Comparison of ICA values between LA-negative and LA-positive samples among reagents
The ICA values for the LA-negative and LA-positive samples were compared among the two in-house reagents and four commercial reagents (Fig. 2). The value of each bar indicates the mean of each reagent. The means of SL, EA, SLA, FSL, SP and PTT with the LA-negative samples were 4.9, 5.0, 7.0, 9.1, 7.4 and 6.1, respectively, while those with the LA-positive samples were 24.6, 33.6, 18.8, 16.4, 25.7 and 25.2, respectively. For each reagent, the ICA value of the LA-positive samples was significantly higher than that of the LA-negative samples (P <0.01). In LA-positive samples, the ICA value of EA was significantly higher with all six reagents (P <0.01), while there was no significant different between SL and PTT. The mean values of the commercial ellagic acid reagents (SLA and FSL) were lower than those of the commercial silica reagents (SP and PTT).
Sensitivity to and specificity for LA
The sensitivity and specificity of each reagent for LA samples are shown in Table 3. The sensitivity of SL, EA, SLA, FSL, SP and PTT was 91%, 96%, 68%, 46%, 91% and 86%, respectively, while specificity was 100%, 98%, 98%, 98%, 100% and 95%, respectively. The sensitivity of EA to LA was higher than that of SL and PTT, while SLA and FSL each had sensitivity lower than SP and PTT.
Table 3. Sensitivity to and specificity for LA
Sensitivity and specificity for LA-positive samples were calculated using the ICA cut-off value for each reagent. APTT reagent abbreviations are the same as shown in Table 1.
In the present study, we prepared two types of APTT reagents, silica-based and ellagic acid-based, and examined whether sensitivity to LA was dependent on the type of activator used under the same phospholipid condition. The normal reference range of SL was found to be 30.1–47.0 s, while those of SP and PTT, commercial reagents prepared from silica, were 30.4–46.1 and 33.6–48.1 s. The normal reference range of SL was similar to that of SP and PTT, thus SL was considered acceptable for use in an APTT reagent. In addition to SL, EA can also be used, because the normal reference range was acceptable (28.0–40.2 s), and fell between that of the commercial ellagic acid- and silica-based reagents. The phospholipid concentration of EA was half of that in SLA, thus the difference in normal reference range between EA and SLA was dependent on phospholipid concentration. As the phospholipid concentration of the reagent increased, the normal reference range narrowed.
The ICA cut-off value for each reagent was established using 41 APTT prolonged LA-negative plasma samples (Table 2). In this study, the mean value + 2 SD was used as the cut-off. An ICA cut-off value of 15 was previously reported by Rosner et al. , which was calculated from kaolin clotting time. However, the present ICA cut-off value calculated using APTT was lower, except for FSL. As the cut-off values differed, it was considered important to determine separate values for each reagent, as it may be related to activators and phospholipids present in the reagent.
The APTT values of normal, LA-negative and LA-positive samples were compared in each reagent (Fig. 1). The LA-positive APTT values of SL and EA were clearly separated from those of normal samples. In the commercial reagents, including SLA, FSL, SP and PTT, the minimum APTT value of LA-positive samples is the same level as the maximum value of normal samples in each reagent. SL and EA would show a highly prolonged APTT value in LA-positive samples, and be useful as an LA screening reagent. Our centrifugation method confirmed that there was not any significant difference between our preparation and the ISTH recommended preparation for detecting LA .
The ICA value of the LA-positive samples was compared with that of the LA-negative samples (Fig. 2). The ICA cut-off value of the EA reagent for the LA-positive samples was clearly different from that for the LA-negative samples. On the other hand, the differences in ICA values between LA-positive and LA-negative samples of the other reagents were not similar to those of the EA reagent. We concluded that LA-positive samples can be clearly detected by use of the EA reagent.
When the ICA values of the LA-positive samples were compared, those of EA were higher than those of the other five reagents, including SL. The phospholipid concentration of EA is half of that in SLA, thus the difference in ICA values between EA and SLA is dependent on the concentration of phospholipids. On the other hand, the ICA value of EA was higher than that of SL, indicating that the sensitivity of APTT reagents to LA is not dependent on the kind of activator utilized. In addition, SL and EA were prepared with the same phospholipid concentration, thus ellagic acid as an activator might be more suitable for LA detection than silica.
Tripodi et al.  investigated LA sensitivity of the commercial reagents Pathromtin SL, Synthasil IL, APTT LT and KPTT, in addition to PTT, FSL and SP, used in the present study, and utilized FSL as an ellagic acid reagent. Their results showed that the LA sensitivity of FSL was lower than that of PTT and SP, silica-based reagents. However, FSL was the only ellagic acid-based reagent used and its sensitivity was lower than that of the silica-based reagents. In the present study, the sensitivity to LA of three ellagic acid-based reagents was investigated and comparisons were made with silica-based reagents. FSL showed the lowest LA sensitivity of all six reagents examined. In contrast, that of EA was highest among the six reagents and showed notable sensitivity to LA as compared with the silica-based reagents. There would be a wide variation in phospholipid composition and concentrations between commercial APTT reagents. Most commercially available APTT reagents with high sensitivity to LA might be prepared with a low concentration of phospholipids.
In conclusion, ellagic acid-based reagents showed high sensitivity to LA as compared with silica-based reagents in a low phospholipid condition and had adequate sensitivity to detect LA. We conclude that the sensitivity to LA of an APTT reagent is dependent on the phospholipid concentration and not the activator employed.
We would like to thank H. Juraku for technical assistance and gratefully acknowledge Y. Ishida for kind help in preparation of the manuscript. We are also thankful for the support of T. Taniguchi of Sysmex Corporation.
Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.