Haemostasis 1990;20:208-214
© 1990 S. Karger AG, Basel 0301 -0147/90/0204-0208S2.75/0
Comparison of Various Screening and Confirmatory Tests for the Detection of the Lupus Anticoagulant R.R. Forastiero, C.R. Falcôn, L.O. Carreras Hemostasis and Thrombosis Section, Division of Hematology, Hospital de Clinicas, University of Buenos Aires, Argentina
Key Words. Lupus anticoagulant • Activated partial thromboplastin time • Dilute Russell viper venom time • Platelet neutralization procedure
Introduction The lupus anticoagulant (LA) is an autoantibody directed against negatively charged phospholipids [1-3]. It is usually detected by prolongation of phospholipiddependent coagulation assays. LA was first described in patients with systemic lupus erythematosus (SLE) [4], However, it was also reported in other autoimmune disor
ders, neoplasms, infectious diseases, certain drug-induced disorders and in otherwise normal individuals [5, 6]. The presence of LA has been correlated with an increased risk to develop thrombo sis, recurrent fetal loss and thrombocytope nia [7-11], These associations have stressed the importance of diagnosing LA. Several coagulation tests were used in or der to detect LA (screening tests) and to con
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Abstract. Several assay systems have been proposed for detection of the lupus anticoagu lant (LA). We compared several screening and confirmatory tests used for the detection of LA in 108 patients. LA was detected in 52 plasmas. The activated partial thromboplastin time (APTT) and the dilute Russell viper venom time (DRVVT) were the most sensitive screening tests as compared to the kaolin clotting time (p < 0.001). The platelet neutraliza tion procedure in both the APTT and DRVVT systems was superior to APTT performed with high phospholipid concentration (p < 0.01) and tissue thromboplastin inhibition (p < 0.001) as confirmatory tests. There was an association between the presence of LA and antiphospholipid antibodies detected by the enzyme-linked immunosorbent assay (p < 0.001). In summary, our results show that APTT may be a sensitive test for the detection of LA when an appropriate reagent is employed, and that freeze-thawed platelets are more effective than phospholipids to neutralize LA activity.
firm the antiphospholipid nature of the in hibitor (confirmatory procedures) [12-16]. Nevertheless, there are few large studies comparing these tests. The aim of our study was to compare several screening and confir matory tests used for the detection of LA. We have studied 108 patients, most of them from a high-risk population. The presence of clinical complications in LA-positive pa tients and the association with antiphospho lipid antibodies (APA) detected by an en zyme-linked immunosorbent assay (ELISA) method were also evaluated.
Patients and Methods Patients One hundred and eight selected patients were studied. The clinical features of our population were the following: SLE in 31, other related rheumatic dis eases (rheumatoid arthritis, Sjogren, scleroderma) in 7, unexplained thrombosis (arterial or venous) in 17, recurrent fetal loss in 26, neoplasms in 5, autoim mune thrombocytopenic purpura in 3, autoimmune hemolytic anemia in 4, infectious diseases in 3, mis cellaneous (e.g. preoperative patients, incidental find ings) in 12. Clinical complications: thrombotic events (arterial or venous) and recurrent fetal loss were present in 12 patients of the SLE group (two patients had both types of complications) and in 1 patient of the group with other rheumatic diseases. Normal control blood samples were obtained from healthy blood donors. Plasma samples from patients with factor VIII deficiency, factor VIII inhib itor, factor IX deficiency, patients under treatment with acenocoumarol, or heparinized plasma samples (0.1-0.5 U/ml) were also studied. Samples Blood was obtained by clean venipuncture from the antecubital vein and collected into plastic tubes containing 0.129 M sodium citrate in a ratio of 9 parts blood to 1 part anticoagulant, and then centri fuged twice at 4 °C and 2,500 g for 15 min. Plateletpoor plasma (PPP) was obtained and assayed imme diately. Blood samples for serum tests were collected
209
into glass tubes without anticoagulant at the same time as plasma samples were drawn. The blood was allowed to clot and then was centrifuged at 1,000 g for 10 min to obtain serum. Coagulation Studies All coagulation studies were performed in dupli cate. Human brain partial thromboplastin (HPT) was used as the source of phospholipids. HPT was pre pared in our laboratory according to a method de scribed by Larrieu and Willard [17]. Briefly, 1 g of acetone-extracted human brain was gently mixed with 50 ml chloroform during 30 min at room tempera ture. The suspension was filtered and then evaporated at 56 °C. Fifty milliliters of sodium chloride (0.9 g%) was then added. A final concentration of 0.6 mg% was obtained. Screening Tests Activated Partial Thromboplastin Time. Activated partial thromboplastin time (APTT) reagent was pre pared mixing equal volumes of HPT 1:75 in barbital buffer pH 7.6 and 1.5% kaolin in sodium chloride 0.15 M. Test PPP (0.1 ml) was incubated with 0.1 ml of the APTT reagent during 2 min at 37 °C. Clotting was initiated by the addition of 0.1 ml of 0.025 M CaCli. The normal range for control plasmas was 34-50 s. Forty-five plasmas with factor deficiencies were tested in a 1:1 mixture with normal plasma. The APTT of the mixture was in all cases within 5 s of the control plasma. Thus, the prolongation of the APTT was consid ered to be due to an inhibitor if it was not corrected in a 1:1 mixture with normal plasma (greater than 5 s over the control). A similar criterion was proposed by Green et al. [18] on acquired inhibitors of coagula tion. Kaolin Clotting Time. This test was performed as described by Exner et al. [ 13]. PPP (0.2 ml) was incu bated with 0.1 ml kaolin (20 mg/ml) at 37 ° C during 3 min, and then 0.025 M CaCL was added. The kaolin clotting time (KCT) was measured. The results were expressed by the calculated index according to the formula described by Rosner et al. [19]: (KCT 1:1 mixture) - (KCT control plasma) w 1An KCT test plasma KCT index greater than 15 was considered a positive result.
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Detection of Lupus Anticoagulant
210
Confirmatory Tests Platelet Neutralization Procedure in the APTT Sys tem. A reaction mixture of 0.1 ml APTT reagent, 0.1 ml PPP and 0.1 ml platelet suspension obtained as described by Triplett et al. [14] was incubated at 37 °C during 2 min. Following incubation, 0.1 ml of 0.025 M CaCL was added and the time to clot forma tion was determined in duplicate. A shortening greater than 7 s from the baseline APTT was consid ered as indicative of the antiphospholipid nature of the inhibitor. Platelet Neutralization in the DR VVT System. The DRVVT was performed with substitution of phospho lipids for platelets prepared as described by Triplett et al. [14], Platelets were diluted to a concentration giv ing a DRVVT from 38-50 s in control plasma. The test indicated the presence of an LA if there was cor rection to normal values from the baseline DRVVT. Neutralization by High Phospholipid Concentra tion. Neutralization by high phospholipid concentra tion (HPC-APTT) was performed as described by Rosove et al. [16], but using HPT diluted 1:18 as the source of phospholipids. A difference between patient/normal (P/N) APTT performed with standard phospholipid concentration and the same test with 4-fold increased phospholipid concentration, greater than 0.18 was considered a positive result. Tissue Thromboplastin Inhibition. This test was performed as described by Schleider et al. [5]. Human brain thromboplastin diluted 1:1,000 in sodium chlo ride (0.15 M) was used. The test was considered pos itive if the ratio between P/N with dilute thrombo plastin (10-3) and P/N with undiluted thromboplastin was greater than 1.25 [19].
The diagnosis of LA was based tion of at least one of the screening rection in the mixture with normal tive results in at least one of the says.
on the prolonga tests, lack of cor plasma and posi confirmatory as
Detection o f APA by ELISA APA, IgG and IgM isotype were measured using an ELISA assay as described by Loizou et al. [20]. We detected the prevalence of APA using 5 different anti gens: cardiolipin, phosphatidylserine, phosphatidylinositol, HPT and bovine brain partial thromboplas tin (Thrombofax®; Ortho Diagnostics, Raritan, N.J., USA). The purified phospholipids were purchased from Sigma. St. Louis, Mo., USA. Statistical Analysis The x2 test was used when comparing the clinical features of patients with laboratory findings. The cor relation coefficients of the different coagulation tests were calculated using a Spearman rank. The Pearson correlation coefficient was obtained to compare the immunological assays with the APTT prolongation, and determined from the numerical data.
Results LA was detected in 52 plasmas (48%). APTT was prolonged in 50 of 52 (96.1%), DRVVT in 48 of 52 (92.3%) and KCT in 33 of 52 (63.5%). We found a good agreement between APTT and DRVVT (r = 0.84), and only moderate agreement between these tests and KCT (r values around 0.60; table 1). In our population, 4 of 108 patients had only one screening test abnormal (3 APTT, 1 DRVVT). APTT and DRVVT were more sensitive than KCT (p < 0.001) as screening tests. The ‘cofactor phenomenon’ was consid ered to be present when a further prolonga tion of the patient’s plasma clotting time was obtained after mixing with control plasma. It was demonstrated in 5 patients in the 4:1 mixture of patient and normal plasma in the
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Dilute Russell Viper Venom Time. Dilute Russell viper venom time (DRVVT) was performed by a modification of the technique described by Thiagarajan et al. [12], HPT was used as the source of phos pholipids, diluted 1:150 with barbital buffer pH 7.6. The viper venom was purchased from Sigma, St. Louis, Mo., USA, reconstituted as suggested by the manufacturer and further diluted 1:100 in barbital buffer pH 7.6. PPP (0.1 ml) was incubated with 0.1 ml viper venom and 0.1 ml of the dilute HPT at 37 °C during lmin, and then 0.025 M CaCL was added. The nor mal range for control plasmas was 38-50 s. The test was considered positive if it was prolonged and not corrected in a 1:1 mixture with normal plasma.
Forastiero/Falcdn/Carreras
211
Detection of Lupus Anticoagulant
Table 1.Correlation between coagulation tests used for the diagnosis of the LA Confirmatory tests
Screening tests
APTT KCT DRVVT PNP PN-DRVVT HPC-APTT TTI
APTT
KCT
DRVVT
1
0.63 1
0.84 0.69 1
PNP
PN-DRVVT
HPC-APTT
TTI
1
0.80 1
0.76 0.72 1
0.48 0.46 0.28 1
PNP = Platelet neutralization procedure in the APTT system; PN-DRVVT = platelet neutralization in the DRVVT system. Values are correlation coefficients obtained by Spearman rank.
Table 2. Evaluation of the specificity of confirmatory tests for LA in the presence of anticoagulants or factor deficiencies
Heparin, 0.1-0.5 U/ml Oral anticoagulants Factor VIII deficiency Factor VIII inhibitor Factor IX deficiency
n
PNP
PN-DRVVT
HPC-APTT
TTI
10 15 7 1 1
10 0 0 0 0
6 0 0 0 0
4 1 0 0 0
8 1 0 0 0
APTT system. This phenomenon was more pronounced in 3 of them using a 1; 1 mixture. Two of the patients showing a cofactor phe nomenon, had normal APTT. In these patients, the cofactor effect was also detected in the DRVVT system, using a 1:1 mixture of patient:control plasma. Concerning confirmatory tests, platelet neutralization was positive in 46 out of 50 positive patients (92%) in the APTT system and in all the patients with an abnormal DRVVT (100%). On the other hand, HPC-
APTT was positive in 36 of 50 (72%) and tissue thromboplastin inhibition (TTI) in 24 of 52 (46%) LA-positive patients. The corre lation coefficients between these tests are shown in table 1. The PN in both the APTT and DRVVT systems was more sensitive than HPC-APTT (p < 0 .0 1 ) and TTI (p < 0.001) as confirmatory tests. Table 2 shows the specificity evaluation of the confirmatory procedures. APA were detected in 50 patients with LA (96.2%) and in 37 patients without LA
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Results are number of cases with positive results in each test. PNP - Platelet neutralization procedure in the APTT system; PN-DRVVT = platelet neutralization in the dilute Russell viper venom time.
212
Table 3. Correlation between levels of APA and prolongation of the APTT
IgG isotypc APTT IgM isotype APTT
ACA
APSA APIA ATA
AHPTA
0.54
0.48
0.43
0.43
0.55
0.44
0.42
0.4!
0.41
0.49
Values are Pearson correlation coefficients (r). ACA = Anticardiolipin antibodies; APSA = antiphosphatidylserine antibodies; APIA = antiphosphatidylinositol antibodies; ATA = antibovine brain partial thromboplastin antibodies; AHPTA = antihuman brain partial thromboplastin antibodies.
(66 %). These results show an association be tween LA and the presence of APA (p < 0.001). However, the Pearson correlation coefficients demonstrate only a moderate agreement between APTT prolongation and ELISAs (r values around 0.50; table 3). Regarding the clinical associations, LA was detected in 7 of 17 patients (41 %) with unexplained thrombosis, in 4 of 26 patients (15 %) with recurrent fetal loss, and in 20 of 38 patients (52%) with rheumatic diseases. In the group with rheumatic diseases, LA was detected in 9 of 11 patients (81.8%) with APA-related disorders and in 11 of 27 pa tients (40.7%) without complications. There was a statistically significant asso ciation between the detection of LA and the presence of clinical complications in this group of patients (p < 0.001).
Discussion LA is one of the most frequent abnormal ities found when a prolonged APTT is ex plored in a routine laboratory. In recent
years, increasing interest for the detection of LA was stimulated by its reported associa tion with thromboembolic events and recur rent fetal loss [7-11]. Different criteria were used to diagnose LA. In an effort to standardize the diagnosis, a group of criteria were established at a working party on acquired inhibitors of co agulation of the International Committee on Thrombosis and Haemostasis [18], These criteria have been widely used, but recently, several investigators realized that they may be too restrictive for the diagnosis of LA. The test systems used can be divided into screening procedures (phospholipid-depen dent coagulation tests) and confirmatory procedures that allow to demonstrate that the inhibitor is directed to phospholipids and not to a specific coagulation factor. Sev eral screening [12, 13] and confirmatory tests [14-16] were proposed by different au thors. However, there are few large compar ative studies [21-24], Regarding the screening tests, they may vary in their sensitivity for the detection of LA, a particular system, however, may be considerably influenced by the reagents em ployed. For instance, the varying sensitivity of various APTT reagents for the detection of LA may be due to different phospholipid composition. In our study, APTT and DRVVT were more sensitive screening tests than KCT (p < 0.001), even employing the index for circulating anticoagulant activity proposed by Rosner et al. [19]. Exner et al. [13] showed that by using highly centrifuged plas ma, the KCT became very dependent on the number of residual platelets. Although we centrifuged plasma samples twice, the filter ing of samples is reported to considerably improve KCT sensitivity [21].
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Forastiero/Falcôn/Carreras
Detection of Lupus Anticoagulant
Some controversies still exist with regard to the screening test of choice for the detec tion of LA. KCT [ 13] and DRVVT [ 12] were proposed as the most sensitive. However, our results show that APTT performed with HPT as the source of phospholipids may also be very sensitive. The APTT is widely used as basic coagulation test in routine hospital laboratories, and does not require plasma fil tration for detecting LA. Thus, the use of a sensitive APTT reagent may facilitate the diagnosis of LA. With regard to confirma tory tests, freeze-thawed platelets were highly effective to neutralize or ‘by-pass’ the LA activity, both in the APTT and in the DRVVT systems. The controversial results reported by different laboratories suggested the need for standardization of reagents and test systems. The widespread availability of sensitive reagents and reference plasmas would be of great interest for routine labora tories to improve the detection of LA. References 1 Feinstein DI, Rapaport SI: Acquired inhibitors of blood coagulation. Prog Hemost Thromb 1972; 1: 75-95. 2 Thiagarajan P, Shapiro SS, De Marco L: Mono clonal immunoglobulin MX coagulation inhibitor with phospholipid specificity. Mechanism of a lu pus anticoagulant. J Clin Invest 1980:66:397— 405. 3 Pengo V, Thiagarajan P, Shapiro SS, et al: Immu nological specificity and mechanism of action of IgG lupus anticoagulants. Blood 1987;70:69-76. 4 Conley CL, Hartmann RC: A hemorrhagic disor der caused by circulating anticoagulant in patients with disseminated lupus erythematosus. J Clin Invest 1952;31:621-622. 5 Schleider MA, Nachman RL, Jaffe EA, et al: A clinical study of the lupus anticoagulant. Blood 1976;48:499-509. 6 Gastineau DA, Kazmier FJ, Nichols WL, et al: Lupus anticoagulants: An analysis of the clinical
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Concerning confirmatory procedures, these are based on the demonstration of an inverse relationship between the phospho lipid content of the assay system and the degree of prolongation of coagulation tests. Thus, neutralization of LA activity may be obtained with HPC, or additional prolonga tion of coagulation tests occurs when a dilute phospholipid is used. Different confirmatory systems may vary in their specificity for LA. In our study, as confirmatory tests for the diagnosis of LA, platelet neutralization procedure and platelet neutralization by DRVVT were superior to HPC-APTT (p < 0.01) and TTI (p < 0.001). Our results con firm a previous report which showed the ability of platelets activated by freeze-thaw ing to bypass the inhibitor activity in a more effective way than phospholipids [25]. Using freeze-thawed platelets in the DRVVT, we obtained similar results to those reported by Thiagarajan et al. [12], where platelets were stimulated by calcium ionophore A 23187. In our study we observed both false-positive and false-negative results of the TTI. This test proved to be less specific than the other procedures for the detection of LA. Similar results were reported by other groups [14, 16], In our population, we found an associa tion between the detection of LA and the presence of APA measured by ELISA (p < 0.001). A similar prevalence of the different APA was found in LA-positive and LA-nega tive patients (data not shown). We observed a similar correlation with antibodies to other negatively charged phospholipids, including phosphatidylserine. Among patients with au toimmune diseases, we found an association between the presence of LA and APA-related clinical complications.
213
7
8 9
10
11
12
13
14
15
16
17
and laboratory features of 219 cases. Am J Hema tol 1985;19:265-275. Lechner K: Lupus anticoagulants and thrombosis; in Verstraete M, Vermylen J, Lijnen R, Amout J (eds): Thrombosis and Haemostasis 1987. Leu ven, Leuven University Press, 1987, pp 525— 547. Branch DW: Immunological disease and fetal death. Clin Obstet Gynecol 1987;30:295-311. Carreras LO, Vermylen J, Spitz B, et al.: Lupus anticoagulants and inhibition of prostacyclin for mation in patients with repeated abortion, intra uterine growth retardation, and intrauterine death. Br J Obstet Gynaecol 1981;88:890-894. Carreras LO, Vermylen J: Lupus anticoagulant and thrombosis: Possible role of inhibition of prostacyclin formation. Thromb Haemost 1982; 48:38-40. Elias M, Eldor A: Thromboembolism in patients with the ‘lupus’-type-circulating anticoagulant. Arch Intern Med 1984;144:510-515. Thiagarajan P, Pengo V, Shapiro SS: The use of the dilute Russell viper venom time for the diag nosis of lupus anticoagulants. Blood 1986;68: 869-874. Exner T, Rickard KA, Kronenberg H: A sensitive test demonstrating lupus anticoagulant and its be havioural patterns. Br J Haematol 1978:40:143— 151. Triplett DA, Brandt JT, Kaczor D, et al: Labora tory diagnosis of lupus inhibitors: A comparison of the tissue thromboplastin inhibition procedure with a new platelet neutralization procedure. Am J Clin Pathol 1983;79:678-682. Alving BN, Baldwin PE, Richards RL, et al: The dilute phospholipid APTT: A sensitive assay for verification of lupus anticoagulants. Thromb Hae most 1985;54:709-712. Rosove MH, Ismail M, Koziol BJ, et al: Lupus anticoagulants: Improved diagnosis with a kaolin clotting time using rabbit brain phospholipid in standard and high concentrations. Blood 1986;68: 472-478. Larrieu MJ, Willard C: Utilization de la céphaline dans les tests de coagulation. Rev Hématol 1957; 12:199-201.
Forastiero/Falcón/Carreras
18 Green D, Hougie C, Kazmier FJ, et al: Report of the working party on acquired inhibitors of coag ulation: Studies of the ‘lupus’ anticoagulant. Thromb Haemost 1983;49:144-146. 19 Rosner E, Pauzner R, Lusky A, et al: Detection and quantitative evaluation of lupus circulating anticoagulant activity. Thromb Haemost 1987;57: 144-147. 20 Loizou S, Mc Créa JD, Rudge AC, et al: Measure ment of anticardiolipin antibodies by an enzymelinked immunosorbent assay (ELISA): Standard ization and quantitation of results. Clin Exp Im munol 1985;62:738-745. 21 Exner T: Comparison of two simple tests for the lupus anticoagulant. Am J Clin Pathol 1985;83: 215-218. 22 Lesperance B, David M, Rauch J, et al: Relative sensitivity of different tests in the detection of low-titer lupus anticoagulants. Thromb Haemost 1988;60:217-219. 23 Derksen RHWM, Hasselaar P, Blokzijl L, et al: Coagulation screen is more specific than the anti cardiolipin antibody ELISA in defining a throm botic subset of lupus patients. Ann Rheum Dis 1988;47:364-371. 24 Lo SCL, Oldmeadow MJ, Howard MA, et al: Comparison of laboratory tests used for identifi cation of the lupus anticoagulant. Am J Haematol 1989;30:213-220. 25 Howard MA, Firkin BG: Investigations of the lupus-like inhibitor bypassing activity of platelets. Thromb Haemost 1983;50:775-779.
Received: November 10, 1989 Received in revised form: January 10, 1990 Accepted: January 23, 1990 L.O. Carreras División Hematología Hospital de Clínicas Universidad de Buenos Aires Córdoba 2351 1120 Buenos Aires (Argentina) Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
214
© 1990 S. Karger AG, Basel 0301 -0147/90/0204-0208S2.75/0
Comparison of Various Screening and Confirmatory Tests for the Detection of the Lupus Anticoagulant R.R. Forastiero, C.R. Falcôn, L.O. Carreras Hemostasis and Thrombosis Section, Division of Hematology, Hospital de Clinicas, University of Buenos Aires, Argentina
Key Words. Lupus anticoagulant • Activated partial thromboplastin time • Dilute Russell viper venom time • Platelet neutralization procedure
Introduction The lupus anticoagulant (LA) is an autoantibody directed against negatively charged phospholipids [1-3]. It is usually detected by prolongation of phospholipiddependent coagulation assays. LA was first described in patients with systemic lupus erythematosus (SLE) [4], However, it was also reported in other autoimmune disor
ders, neoplasms, infectious diseases, certain drug-induced disorders and in otherwise normal individuals [5, 6]. The presence of LA has been correlated with an increased risk to develop thrombo sis, recurrent fetal loss and thrombocytope nia [7-11], These associations have stressed the importance of diagnosing LA. Several coagulation tests were used in or der to detect LA (screening tests) and to con
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Abstract. Several assay systems have been proposed for detection of the lupus anticoagu lant (LA). We compared several screening and confirmatory tests used for the detection of LA in 108 patients. LA was detected in 52 plasmas. The activated partial thromboplastin time (APTT) and the dilute Russell viper venom time (DRVVT) were the most sensitive screening tests as compared to the kaolin clotting time (p < 0.001). The platelet neutraliza tion procedure in both the APTT and DRVVT systems was superior to APTT performed with high phospholipid concentration (p < 0.01) and tissue thromboplastin inhibition (p < 0.001) as confirmatory tests. There was an association between the presence of LA and antiphospholipid antibodies detected by the enzyme-linked immunosorbent assay (p < 0.001). In summary, our results show that APTT may be a sensitive test for the detection of LA when an appropriate reagent is employed, and that freeze-thawed platelets are more effective than phospholipids to neutralize LA activity.
firm the antiphospholipid nature of the in hibitor (confirmatory procedures) [12-16]. Nevertheless, there are few large studies comparing these tests. The aim of our study was to compare several screening and confir matory tests used for the detection of LA. We have studied 108 patients, most of them from a high-risk population. The presence of clinical complications in LA-positive pa tients and the association with antiphospho lipid antibodies (APA) detected by an en zyme-linked immunosorbent assay (ELISA) method were also evaluated.
Patients and Methods Patients One hundred and eight selected patients were studied. The clinical features of our population were the following: SLE in 31, other related rheumatic dis eases (rheumatoid arthritis, Sjogren, scleroderma) in 7, unexplained thrombosis (arterial or venous) in 17, recurrent fetal loss in 26, neoplasms in 5, autoim mune thrombocytopenic purpura in 3, autoimmune hemolytic anemia in 4, infectious diseases in 3, mis cellaneous (e.g. preoperative patients, incidental find ings) in 12. Clinical complications: thrombotic events (arterial or venous) and recurrent fetal loss were present in 12 patients of the SLE group (two patients had both types of complications) and in 1 patient of the group with other rheumatic diseases. Normal control blood samples were obtained from healthy blood donors. Plasma samples from patients with factor VIII deficiency, factor VIII inhib itor, factor IX deficiency, patients under treatment with acenocoumarol, or heparinized plasma samples (0.1-0.5 U/ml) were also studied. Samples Blood was obtained by clean venipuncture from the antecubital vein and collected into plastic tubes containing 0.129 M sodium citrate in a ratio of 9 parts blood to 1 part anticoagulant, and then centri fuged twice at 4 °C and 2,500 g for 15 min. Plateletpoor plasma (PPP) was obtained and assayed imme diately. Blood samples for serum tests were collected
209
into glass tubes without anticoagulant at the same time as plasma samples were drawn. The blood was allowed to clot and then was centrifuged at 1,000 g for 10 min to obtain serum. Coagulation Studies All coagulation studies were performed in dupli cate. Human brain partial thromboplastin (HPT) was used as the source of phospholipids. HPT was pre pared in our laboratory according to a method de scribed by Larrieu and Willard [17]. Briefly, 1 g of acetone-extracted human brain was gently mixed with 50 ml chloroform during 30 min at room tempera ture. The suspension was filtered and then evaporated at 56 °C. Fifty milliliters of sodium chloride (0.9 g%) was then added. A final concentration of 0.6 mg% was obtained. Screening Tests Activated Partial Thromboplastin Time. Activated partial thromboplastin time (APTT) reagent was pre pared mixing equal volumes of HPT 1:75 in barbital buffer pH 7.6 and 1.5% kaolin in sodium chloride 0.15 M. Test PPP (0.1 ml) was incubated with 0.1 ml of the APTT reagent during 2 min at 37 °C. Clotting was initiated by the addition of 0.1 ml of 0.025 M CaCli. The normal range for control plasmas was 34-50 s. Forty-five plasmas with factor deficiencies were tested in a 1:1 mixture with normal plasma. The APTT of the mixture was in all cases within 5 s of the control plasma. Thus, the prolongation of the APTT was consid ered to be due to an inhibitor if it was not corrected in a 1:1 mixture with normal plasma (greater than 5 s over the control). A similar criterion was proposed by Green et al. [18] on acquired inhibitors of coagula tion. Kaolin Clotting Time. This test was performed as described by Exner et al. [ 13]. PPP (0.2 ml) was incu bated with 0.1 ml kaolin (20 mg/ml) at 37 ° C during 3 min, and then 0.025 M CaCL was added. The kaolin clotting time (KCT) was measured. The results were expressed by the calculated index according to the formula described by Rosner et al. [19]: (KCT 1:1 mixture) - (KCT control plasma) w 1An KCT test plasma KCT index greater than 15 was considered a positive result.
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Detection of Lupus Anticoagulant
210
Confirmatory Tests Platelet Neutralization Procedure in the APTT Sys tem. A reaction mixture of 0.1 ml APTT reagent, 0.1 ml PPP and 0.1 ml platelet suspension obtained as described by Triplett et al. [14] was incubated at 37 °C during 2 min. Following incubation, 0.1 ml of 0.025 M CaCL was added and the time to clot forma tion was determined in duplicate. A shortening greater than 7 s from the baseline APTT was consid ered as indicative of the antiphospholipid nature of the inhibitor. Platelet Neutralization in the DR VVT System. The DRVVT was performed with substitution of phospho lipids for platelets prepared as described by Triplett et al. [14], Platelets were diluted to a concentration giv ing a DRVVT from 38-50 s in control plasma. The test indicated the presence of an LA if there was cor rection to normal values from the baseline DRVVT. Neutralization by High Phospholipid Concentra tion. Neutralization by high phospholipid concentra tion (HPC-APTT) was performed as described by Rosove et al. [16], but using HPT diluted 1:18 as the source of phospholipids. A difference between patient/normal (P/N) APTT performed with standard phospholipid concentration and the same test with 4-fold increased phospholipid concentration, greater than 0.18 was considered a positive result. Tissue Thromboplastin Inhibition. This test was performed as described by Schleider et al. [5]. Human brain thromboplastin diluted 1:1,000 in sodium chlo ride (0.15 M) was used. The test was considered pos itive if the ratio between P/N with dilute thrombo plastin (10-3) and P/N with undiluted thromboplastin was greater than 1.25 [19].
The diagnosis of LA was based tion of at least one of the screening rection in the mixture with normal tive results in at least one of the says.
on the prolonga tests, lack of cor plasma and posi confirmatory as
Detection o f APA by ELISA APA, IgG and IgM isotype were measured using an ELISA assay as described by Loizou et al. [20]. We detected the prevalence of APA using 5 different anti gens: cardiolipin, phosphatidylserine, phosphatidylinositol, HPT and bovine brain partial thromboplas tin (Thrombofax®; Ortho Diagnostics, Raritan, N.J., USA). The purified phospholipids were purchased from Sigma. St. Louis, Mo., USA. Statistical Analysis The x2 test was used when comparing the clinical features of patients with laboratory findings. The cor relation coefficients of the different coagulation tests were calculated using a Spearman rank. The Pearson correlation coefficient was obtained to compare the immunological assays with the APTT prolongation, and determined from the numerical data.
Results LA was detected in 52 plasmas (48%). APTT was prolonged in 50 of 52 (96.1%), DRVVT in 48 of 52 (92.3%) and KCT in 33 of 52 (63.5%). We found a good agreement between APTT and DRVVT (r = 0.84), and only moderate agreement between these tests and KCT (r values around 0.60; table 1). In our population, 4 of 108 patients had only one screening test abnormal (3 APTT, 1 DRVVT). APTT and DRVVT were more sensitive than KCT (p < 0.001) as screening tests. The ‘cofactor phenomenon’ was consid ered to be present when a further prolonga tion of the patient’s plasma clotting time was obtained after mixing with control plasma. It was demonstrated in 5 patients in the 4:1 mixture of patient and normal plasma in the
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Dilute Russell Viper Venom Time. Dilute Russell viper venom time (DRVVT) was performed by a modification of the technique described by Thiagarajan et al. [12], HPT was used as the source of phos pholipids, diluted 1:150 with barbital buffer pH 7.6. The viper venom was purchased from Sigma, St. Louis, Mo., USA, reconstituted as suggested by the manufacturer and further diluted 1:100 in barbital buffer pH 7.6. PPP (0.1 ml) was incubated with 0.1 ml viper venom and 0.1 ml of the dilute HPT at 37 °C during lmin, and then 0.025 M CaCL was added. The nor mal range for control plasmas was 38-50 s. The test was considered positive if it was prolonged and not corrected in a 1:1 mixture with normal plasma.
Forastiero/Falcdn/Carreras
211
Detection of Lupus Anticoagulant
Table 1.Correlation between coagulation tests used for the diagnosis of the LA Confirmatory tests
Screening tests
APTT KCT DRVVT PNP PN-DRVVT HPC-APTT TTI
APTT
KCT
DRVVT
1
0.63 1
0.84 0.69 1
PNP
PN-DRVVT
HPC-APTT
TTI
1
0.80 1
0.76 0.72 1
0.48 0.46 0.28 1
PNP = Platelet neutralization procedure in the APTT system; PN-DRVVT = platelet neutralization in the DRVVT system. Values are correlation coefficients obtained by Spearman rank.
Table 2. Evaluation of the specificity of confirmatory tests for LA in the presence of anticoagulants or factor deficiencies
Heparin, 0.1-0.5 U/ml Oral anticoagulants Factor VIII deficiency Factor VIII inhibitor Factor IX deficiency
n
PNP
PN-DRVVT
HPC-APTT
TTI
10 15 7 1 1
10 0 0 0 0
6 0 0 0 0
4 1 0 0 0
8 1 0 0 0
APTT system. This phenomenon was more pronounced in 3 of them using a 1; 1 mixture. Two of the patients showing a cofactor phe nomenon, had normal APTT. In these patients, the cofactor effect was also detected in the DRVVT system, using a 1:1 mixture of patient:control plasma. Concerning confirmatory tests, platelet neutralization was positive in 46 out of 50 positive patients (92%) in the APTT system and in all the patients with an abnormal DRVVT (100%). On the other hand, HPC-
APTT was positive in 36 of 50 (72%) and tissue thromboplastin inhibition (TTI) in 24 of 52 (46%) LA-positive patients. The corre lation coefficients between these tests are shown in table 1. The PN in both the APTT and DRVVT systems was more sensitive than HPC-APTT (p < 0 .0 1 ) and TTI (p < 0.001) as confirmatory tests. Table 2 shows the specificity evaluation of the confirmatory procedures. APA were detected in 50 patients with LA (96.2%) and in 37 patients without LA
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Results are number of cases with positive results in each test. PNP - Platelet neutralization procedure in the APTT system; PN-DRVVT = platelet neutralization in the dilute Russell viper venom time.
212
Table 3. Correlation between levels of APA and prolongation of the APTT
IgG isotypc APTT IgM isotype APTT
ACA
APSA APIA ATA
AHPTA
0.54
0.48
0.43
0.43
0.55
0.44
0.42
0.4!
0.41
0.49
Values are Pearson correlation coefficients (r). ACA = Anticardiolipin antibodies; APSA = antiphosphatidylserine antibodies; APIA = antiphosphatidylinositol antibodies; ATA = antibovine brain partial thromboplastin antibodies; AHPTA = antihuman brain partial thromboplastin antibodies.
(66 %). These results show an association be tween LA and the presence of APA (p < 0.001). However, the Pearson correlation coefficients demonstrate only a moderate agreement between APTT prolongation and ELISAs (r values around 0.50; table 3). Regarding the clinical associations, LA was detected in 7 of 17 patients (41 %) with unexplained thrombosis, in 4 of 26 patients (15 %) with recurrent fetal loss, and in 20 of 38 patients (52%) with rheumatic diseases. In the group with rheumatic diseases, LA was detected in 9 of 11 patients (81.8%) with APA-related disorders and in 11 of 27 pa tients (40.7%) without complications. There was a statistically significant asso ciation between the detection of LA and the presence of clinical complications in this group of patients (p < 0.001).
Discussion LA is one of the most frequent abnormal ities found when a prolonged APTT is ex plored in a routine laboratory. In recent
years, increasing interest for the detection of LA was stimulated by its reported associa tion with thromboembolic events and recur rent fetal loss [7-11]. Different criteria were used to diagnose LA. In an effort to standardize the diagnosis, a group of criteria were established at a working party on acquired inhibitors of co agulation of the International Committee on Thrombosis and Haemostasis [18], These criteria have been widely used, but recently, several investigators realized that they may be too restrictive for the diagnosis of LA. The test systems used can be divided into screening procedures (phospholipid-depen dent coagulation tests) and confirmatory procedures that allow to demonstrate that the inhibitor is directed to phospholipids and not to a specific coagulation factor. Sev eral screening [12, 13] and confirmatory tests [14-16] were proposed by different au thors. However, there are few large compar ative studies [21-24], Regarding the screening tests, they may vary in their sensitivity for the detection of LA, a particular system, however, may be considerably influenced by the reagents em ployed. For instance, the varying sensitivity of various APTT reagents for the detection of LA may be due to different phospholipid composition. In our study, APTT and DRVVT were more sensitive screening tests than KCT (p < 0.001), even employing the index for circulating anticoagulant activity proposed by Rosner et al. [19]. Exner et al. [13] showed that by using highly centrifuged plas ma, the KCT became very dependent on the number of residual platelets. Although we centrifuged plasma samples twice, the filter ing of samples is reported to considerably improve KCT sensitivity [21].
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Forastiero/Falcôn/Carreras
Detection of Lupus Anticoagulant
Some controversies still exist with regard to the screening test of choice for the detec tion of LA. KCT [ 13] and DRVVT [ 12] were proposed as the most sensitive. However, our results show that APTT performed with HPT as the source of phospholipids may also be very sensitive. The APTT is widely used as basic coagulation test in routine hospital laboratories, and does not require plasma fil tration for detecting LA. Thus, the use of a sensitive APTT reagent may facilitate the diagnosis of LA. With regard to confirma tory tests, freeze-thawed platelets were highly effective to neutralize or ‘by-pass’ the LA activity, both in the APTT and in the DRVVT systems. The controversial results reported by different laboratories suggested the need for standardization of reagents and test systems. The widespread availability of sensitive reagents and reference plasmas would be of great interest for routine labora tories to improve the detection of LA. References 1 Feinstein DI, Rapaport SI: Acquired inhibitors of blood coagulation. Prog Hemost Thromb 1972; 1: 75-95. 2 Thiagarajan P, Shapiro SS, De Marco L: Mono clonal immunoglobulin MX coagulation inhibitor with phospholipid specificity. Mechanism of a lu pus anticoagulant. J Clin Invest 1980:66:397— 405. 3 Pengo V, Thiagarajan P, Shapiro SS, et al: Immu nological specificity and mechanism of action of IgG lupus anticoagulants. Blood 1987;70:69-76. 4 Conley CL, Hartmann RC: A hemorrhagic disor der caused by circulating anticoagulant in patients with disseminated lupus erythematosus. J Clin Invest 1952;31:621-622. 5 Schleider MA, Nachman RL, Jaffe EA, et al: A clinical study of the lupus anticoagulant. Blood 1976;48:499-509. 6 Gastineau DA, Kazmier FJ, Nichols WL, et al: Lupus anticoagulants: An analysis of the clinical
Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
Concerning confirmatory procedures, these are based on the demonstration of an inverse relationship between the phospho lipid content of the assay system and the degree of prolongation of coagulation tests. Thus, neutralization of LA activity may be obtained with HPC, or additional prolonga tion of coagulation tests occurs when a dilute phospholipid is used. Different confirmatory systems may vary in their specificity for LA. In our study, as confirmatory tests for the diagnosis of LA, platelet neutralization procedure and platelet neutralization by DRVVT were superior to HPC-APTT (p < 0.01) and TTI (p < 0.001). Our results con firm a previous report which showed the ability of platelets activated by freeze-thaw ing to bypass the inhibitor activity in a more effective way than phospholipids [25]. Using freeze-thawed platelets in the DRVVT, we obtained similar results to those reported by Thiagarajan et al. [12], where platelets were stimulated by calcium ionophore A 23187. In our study we observed both false-positive and false-negative results of the TTI. This test proved to be less specific than the other procedures for the detection of LA. Similar results were reported by other groups [14, 16], In our population, we found an associa tion between the detection of LA and the presence of APA measured by ELISA (p < 0.001). A similar prevalence of the different APA was found in LA-positive and LA-nega tive patients (data not shown). We observed a similar correlation with antibodies to other negatively charged phospholipids, including phosphatidylserine. Among patients with au toimmune diseases, we found an association between the presence of LA and APA-related clinical complications.
213
7
8 9
10
11
12
13
14
15
16
17
and laboratory features of 219 cases. Am J Hema tol 1985;19:265-275. Lechner K: Lupus anticoagulants and thrombosis; in Verstraete M, Vermylen J, Lijnen R, Amout J (eds): Thrombosis and Haemostasis 1987. Leu ven, Leuven University Press, 1987, pp 525— 547. Branch DW: Immunological disease and fetal death. Clin Obstet Gynecol 1987;30:295-311. Carreras LO, Vermylen J, Spitz B, et al.: Lupus anticoagulants and inhibition of prostacyclin for mation in patients with repeated abortion, intra uterine growth retardation, and intrauterine death. Br J Obstet Gynaecol 1981;88:890-894. Carreras LO, Vermylen J: Lupus anticoagulant and thrombosis: Possible role of inhibition of prostacyclin formation. Thromb Haemost 1982; 48:38-40. Elias M, Eldor A: Thromboembolism in patients with the ‘lupus’-type-circulating anticoagulant. Arch Intern Med 1984;144:510-515. Thiagarajan P, Pengo V, Shapiro SS: The use of the dilute Russell viper venom time for the diag nosis of lupus anticoagulants. Blood 1986;68: 869-874. Exner T, Rickard KA, Kronenberg H: A sensitive test demonstrating lupus anticoagulant and its be havioural patterns. Br J Haematol 1978:40:143— 151. Triplett DA, Brandt JT, Kaczor D, et al: Labora tory diagnosis of lupus inhibitors: A comparison of the tissue thromboplastin inhibition procedure with a new platelet neutralization procedure. Am J Clin Pathol 1983;79:678-682. Alving BN, Baldwin PE, Richards RL, et al: The dilute phospholipid APTT: A sensitive assay for verification of lupus anticoagulants. Thromb Hae most 1985;54:709-712. Rosove MH, Ismail M, Koziol BJ, et al: Lupus anticoagulants: Improved diagnosis with a kaolin clotting time using rabbit brain phospholipid in standard and high concentrations. Blood 1986;68: 472-478. Larrieu MJ, Willard C: Utilization de la céphaline dans les tests de coagulation. Rev Hématol 1957; 12:199-201.
Forastiero/Falcón/Carreras
18 Green D, Hougie C, Kazmier FJ, et al: Report of the working party on acquired inhibitors of coag ulation: Studies of the ‘lupus’ anticoagulant. Thromb Haemost 1983;49:144-146. 19 Rosner E, Pauzner R, Lusky A, et al: Detection and quantitative evaluation of lupus circulating anticoagulant activity. Thromb Haemost 1987;57: 144-147. 20 Loizou S, Mc Créa JD, Rudge AC, et al: Measure ment of anticardiolipin antibodies by an enzymelinked immunosorbent assay (ELISA): Standard ization and quantitation of results. Clin Exp Im munol 1985;62:738-745. 21 Exner T: Comparison of two simple tests for the lupus anticoagulant. Am J Clin Pathol 1985;83: 215-218. 22 Lesperance B, David M, Rauch J, et al: Relative sensitivity of different tests in the detection of low-titer lupus anticoagulants. Thromb Haemost 1988;60:217-219. 23 Derksen RHWM, Hasselaar P, Blokzijl L, et al: Coagulation screen is more specific than the anti cardiolipin antibody ELISA in defining a throm botic subset of lupus patients. Ann Rheum Dis 1988;47:364-371. 24 Lo SCL, Oldmeadow MJ, Howard MA, et al: Comparison of laboratory tests used for identifi cation of the lupus anticoagulant. Am J Haematol 1989;30:213-220. 25 Howard MA, Firkin BG: Investigations of the lupus-like inhibitor bypassing activity of platelets. Thromb Haemost 1983;50:775-779.
Received: November 10, 1989 Received in revised form: January 10, 1990 Accepted: January 23, 1990 L.O. Carreras División Hematología Hospital de Clínicas Universidad de Buenos Aires Córdoba 2351 1120 Buenos Aires (Argentina) Downloaded by: King's College London 137.73.144.138 - 1/26/2019 12:52:33 AM
214
