Comparison of Several Activated Partial Thromboplastin Time Methods ROBERT J. MOWN, M.D., AND DOROTHY WILLOUGHBY, B.S., M.T. (ASCP) Department of Pathology, Harbor General Hospital, UCLA School of Medicine, Torrance, California 90509 ABSTRACT Morin, Robert J., and Willoughby, Dorothy: Comparison of several activated partial thromboplastin time methods. Am J Clin Pathol 64: 241-247, 1975. Activated partial thromboplastin times (APTT's) performed with a semi-automated electrical-conductivity type of clot timer on plasmas from patients with hepatic disease and intravascular coagulation, and on warfarin or heparin therapy, were significantly lower than when done on the same plasmas with either a manual optical method or an automated optical-endpoint instrument. Results of APTT's done on normal plasmas by the three methods were not significantly different. Substitution of different activatorphospholipid reagents resulted in some variability in results, but these differences were less than those between the different types of methods. APTT's done with both the electrical clot timer and the automated optical instrument on prepared plasmas containing or % of factor II, V, VIII, IX, or X revealed shorter times with the electrical clot timer only in the case of factor II- and factor V-deficient plasmas. APTT's done on normal plasmas to which 0.1 or 0.3 units per ml. of heparin had been added in vitro also were shorter with the electrical clot timer than with the automatic optical instrument. Prothrombin times done on normal and abnormal control plasmas and on a series of plasmas from patients on warfarin therapy showed no significant difference between the two methods. (Key words: Activated partial thromboplastin time, optical, automated, electrical; Liver disease; Intravascular coagulation; Warfarin; Heparin; Factor deficiencies.) THE ACTIVATED PARTIAL thromboplastin nical variables affecting the APTT are time (APTT) has been found to be of value venipuncture technic, 4 collection of in screening for congenital or acquired samples in citrate us. oxalate, 10 type of actiintrinsic coagulation factor deficiencies, 7 vator-phospholipid reagent utilized, 7,8 in monitoring heparin therapy, 3 and as and use of frozen rather than fresh an adjunct to the prothrombin time for plasma. 5 Recent studies have demonstrated following the reductions in factor IX acti- that an automated APTT method utilizing vity induced by anticoagulant therapy with an optical end-point produced APTT coumarin-type agents. Among the tech- results that correlated well with those of a manual optical method and showed some- Received November 11, 1974; received revised what greater precision. 1 Another previous manuscript January 21, 1975; accepted for publica- ^. r,,,,,. don January 21, 1975. study comparing factor VIII assays based Address reprint requests to Dr. Morin. upon the APTT as done by a manual 241
242 MORIN AND WILLOUGHBY A.J.C.P. Vol.64 method and a semiautomatic electrical clot timer (Fibrometer) showed good correlation between the two methods. 9 Other laboratories, however, have noted difficulties in reproducibility of APTT's performed with the Fibrometer, 6 but these have not yet been systematically studied and documented. The purposes of the present study were to evaluate the performance of APTT's done using a new automated optical-end-point instrument (Electra 600D) compared with a manual method, to compare results by the former two methods with those of the Fibrometer method, and to explore the mechanisms for possible differences among these methods. Materials and Methods Patient blood samples were collected in Vacutainers containing 1 part 3.8% citrate per 9 parts blood. Plasmas were separated from cells by centrifugation at 3,000 r.p.m. (1,000 Xg) for 5 min. Activated partial thromboplastin times were performed in sequence on the same plasmas using an automated method with an optical end-point (Electra 600D, Medical Laboratory Automation, Inc.), with the semiautomatic conductivity-end-point Fibrometer (Baltimore Biological Laboratories), and also using a manual optical method. With the Electra 600D method, duplicate 0.1-ml. aliquots of patients' plasmas were placed in 10 x 75-mm. glass tubes, 0.1 ml. of activated cephaloplastin reagent (Dade) added manually, and the combined plasma and reagent mixed by vortex and placed in the sample turntable, where it was kept at 4 C. until preincubation at 37 C. for 5.7 minutes. One tenth milliliter of 0.025 M CaCl 2 was automatically forcibly dispensed by peristaltic pump into each tube, with mixing. The Electra 600D was used in the "PTT" mode with the "B" lamp. This instrument uses a tungsten light source with a red transmission filter, and the time required to produce a change in optical density (induced by clot formation) is printed automatically, following which the next sample is automatically advanced to the reaction station. In some experiments the "APTT" reagent (General Diagnostics Division of Warner-Lambert) was substituted for the activated cephaloplastin reagent. For the Fibrometer method, duplicate 0.1-ml. aliquots of plasma were each mixed with 0.1 ml. of activated cephaloplastin reagent in plastic cups ("Fibrotubes") and incubated at 37 C. for 5 min. After addition of 0.1 ml. of 0.025 M CaCl 2, the end points of clot formation were detected by changes in electrical conductivity, and registered automatically. The Fibrometer probes were cleaned between successive determinations. In some experiments Hyland Partial Thromboplastin reagent or Fibrolet reagent (BBL) was substituted for the Dade reagent. Activated partial thromboplastin times on each plasma were repeated in duplicate on both the Electra 600D and the Fibrometer at one-hour intervals following the initial determinations. For the manual method, 0.1-ml. portions of activated cephaloplastin reagent were placed into 12 x 75-mm. glass tubes, followed by duplicate 0.1-ml. aliquots of each patient plasma. After incubation in a water bath for 5 min., 0.1 ml. 0.025 M CaCl 2 was added to each, mixed, and after 20 seconds observed under a magnifier lamp until visible clot formation had occurred. Activated partial thromboplastin times by the Electra and Fibrometer methods were also done on pooled normal plasmas that had been mixed with plasmas deficient in factor II, V, VIII, IX, or X (Dade) such that the final mixtures contained either or % of each individual factor. To other aliquots of pooled normal plasma, 0.1 or 0.3 unit per ml. of heparin (Upjohn, Heparin Sodium) was added, and APTT's done as on the above-mentioned factor-deficient plasmas.
August 1975 COMPARISON OF APTT METHODS 243 Table 1. Comparison of Activated Partial Thromboplastin Times Performed with the Electra, Fibrometer, and Manual Methods* Electra Fibrometer Electra Fibrometer Manual 1 Hr. 1 Hr. Normal 32 31 33 33 34 35 29 29 32 30 29 32 25 28 34 30 26 24 28 36 25 23 30 32 34 Hepatic disease 65 64 38 43 66 75 68 64 41 48 80 55 82 84 61 48 42 53 49 55 78 59 81 67 57 Intravascular coagulation 65 49 65 69 52 51 48 48 56 56 62 40 57 65 39 93 70 104 90 82 44 45 48 1 37 54 Warfarin 99 68 94 107 71 62 44 66' 65 47 57 34 67 55 38 58 34 57 60 39 44 32 41 ' 43 42 Heparin 54 39 47 " ; 59 38 52 120 49 73 51 128 50 114 53 79 55 30 60 ' 53 36 79 56 76 85 62 * Values given are the mean numbers of seconds from duplicate determinations. All APTT's in these experiments were done using the activated cephaloplastin (Dade) reagent. Prothrombin times using both the Electra and Fibrometer methods were also performed on a series of plasmas from warfarin-treated patients. Using the Electra method (with the instrument set at the PT mode with the "A" lamp), duplicate 0.1-ml. aliquots of plasmas in 10 X 75- mm. glass tubes were placed in the turntable, maintained at 4 C. and incubated at 37 C. for 3 min., followed by automatic addition of 0.2 ml. of the thromboplastincalcium reagent (Simplastin, General Diagnostics) and automatic printout of times required to reach the optical end point. For the Fibrometer prothrombin time method, 0.2-ml. portions of Simplastin were pre-warmed at 37 C. for 5 min. in plastic cups ("Fibrotubes"), followed by addition of 0.1-ml. aliquots of plasmas prewarmed at 37 C. for 5 min. Times required for clot formation to produce the changes in electrical conductivity were registered automatically on the digital readout dial. Standard deviations, regression analyses and t tests were all determined using a PDP-8L computer. Results and Discussion The results of APTT's performed on the same plasmas Using the Electra, Fibrometer, and manual tilt-tube methods are shown in Table 1. Five replicate normal and abnormal quality control plasmas analyzed in parallel with this series had the following results (seconds, means ± standard deviations):control normal plasmas (Dade), Electra 29.2 ± 0.9 (coefficient of variation 3.1%), Fibrometer 30.3
244 MORIN AND WILLOUGHBY A.J.C.P. Vol.64 Table 2. Comparison of the Activated Partial Thromboplastin Times with the Fibrometer and Electra Methods Using Different Commercial Reagents Electra Fibrometer General Dade Hyland BBL Dade Diagnostics Normal 26 28 32 24 27 33 31 31 26 28 34 27 28 31 29 Hepatic disease 68 61 59 93 96 43 47 39 74 70 34 35 36 44 43 Intravascular Coagulation 43 47 40 57 54 48 56 46 79 83 61 65 59 73 77 Warfarin 41 38 36 58 53 48 55 43 83 86 36 39 35 69 72 Heparin 44 52 43 47 52 56 60 62 84 78 62 68 56 106 95 ± 1.3 (C.V. 4.3%), manual 30.9 ± 1.7 (C. V. 5.5%); coagulation control, abnormal (Dade), Electra 46.9 ± 2.3 (C.V. 4.9%), Fibrometer 45.4 ± 2.5 (C.V. 5.5%), manual 47.4 ± 2.8 (C.V. 5.9%). APTT results on the five normal patient plasmas tested did not differ significantly when performed by the Electra, Fibrometer, and manual methods. APTT's on plasmas from patients with hepatic disease or on warfarin therapy were all consistently lower (13-42%) when done by the Fibrometer method compared with the Electra method. Results with the manual method were not significantly different from those with the Electra. Most plasmas from patients with the Intravascular coagulation syndrome and from patients on heparin therapy also showed significantly shorter APTT's with the Fibrometer method compared with the Electra method, (22-45%), although with two of the intravascular coagulation plasmas and one of the heparin plasmas the results with the two methods were not significantly different. Mean abnormal APTT's from all four categories were 27% lower when done by the Fibrometer method compared with the Electra method. The mean ± standard deviation for the 20 abnormal APTT's done by the manual method was 68.3 ±21.6 seconds, and for the Electra method, 66.5 ± 19.7 seconds. Regression analysis showed the correlation coefficient (r) between these two methods to be 0.961 (significant at P<0.01), with a slope of 0.886 and a y intercept of 6.00. The paired t test value was 1.476, indicating no significant difference between results performed by these two methods. For APTT's done on the same plasma by the Fibrometer method the mean ± standard deviation was 47.4 ± 12.7 seconds. When the manual method values were compared with those of the Fibrometer by regression analysis, r = 0.822, with a slope of 0.482 and a y intercept of 14.51. The t test value was 7.02, and the differences between the paired values were significant at P<0.01.
August 1975 COMPARISON OF APTT METHODS 245 When the same Electra APTT values were compared with those of the Fibrometer by regression analysis, r = 0.857, with a slope of 0.550 and a y intercept of 10.85. The t test value was 7.75, and the paired values were significantly different at P< 0.001. Since the Fibrometer APTT's were done approximately 10-15 minutes following each APTT done on the Electra, APTT's on both instruments were repeated one hour later to determine any possible influence of time on the observed variations. These one-hour APTT's showed some small differences when compared with the initial determinations, but these were smaller than the differences between the methods, and the Fibrometer times generally tended to increase with time rather than to decrease. These alterations with time were possibly due to changes in the ph of the plasma of standing 2 or to partial loss of the labile factor V. It has been observed that the quality of the end-point clot varies considerably with the type of activator-phospholipid reagent utilized. 7 In a second series of experiments, therefore, several commercial reagents were utilized to perform similar APTT determinations on the same nbrmal and abnormal plasmas. APTT's done using the Electra with the Dade activated cephaloplastin reagent (ellagate activator) compared with the General Diagnostics APTT reagent (microsilicon activator) showed no significant difference (Table 2). Reagents containing particle activators could not be used with the Electra due to turbidity interference with the opticalend-point detection system. APTT's on abnormal plasmas done using the Fibrometer method with the Dade activated cephaloplastin, Hyland Partial Thromboplastin reagent (kaolin activator), and BBL Fibrolet reagent (celite activator) revealed significant variability in results (2.8-13.2%, mean 7.8%), but these variations were of Table 3. Effects of Factor Deficiencies and Heparin in vitro on Activated Partial Thromboplastin Times by the Electra and Fibrometer Methods Factor II Factor V Factor VIII Factor IX Factor X Heparin % of Normal 0 U. per ml. 0.1 U. per ml. 0.3 U. per ml. Electra 28 ± 1.6 54 ±2.5 120 ±6.5 29 ± 1.3 59 ±2.6 114 ±6.4 33 ± 1.6 46 ±2.7 66 ± 3.4 29 ± 1.7 50 ± 2.4 58 ± 2.4 31 ± 1.8 52 ±2.1 67 ±3.3 30 ± 48 ±2.6 109 ± Fibrometer 27 32 51 25 34 44 29 45 58 33 49 55 31 57 63 ± 1.8 ± 1.8 ±2.3 ± 1.4 ± 1.4 ±2.0 ± 1.2 ± 1.7 ± 1.7 ± 1.4 ±2.2 ± 2.9 ± 1.5 ±2.5 ± 3.4 31 ±2.1 29 ± 1.7 62 ± 3.1 lesser magnitude than the mean 27% difference between the Electra and Fibrometer methods observed in the initial experiments. To investigate the mechanisms of the variations between the optical and electrical methods, APTT's were also done in quadruplicate with each method on normal plasmas adjusted to contain, 5 or 1% of factors II, V, VIII, IX, or X, and no heparin, 0.1 unit heparin per ml., or 0.3 units heparin per ml. As can be seen in Table 3, APTT's done on plasmas containing 5 or 1% of either factors II or factor V showed markedly different values when done by the Fibrometer method compared with the Electra. APTT's on plasmas deficient in factors VIII, IX, and X showed no significant differences between the two methods. Plasmas heparinized by in-vitro addition of either 0.1 or 0.3 units heparin
246 MORIN AND WILLOUGHBY A.J.C.P. Vol. 64 Table 4. Prothrombin Times of Quality Control Plasmas and Plasmas from Warfarin-treated Patients as Determined by Electra and Fibrometer Methods* Verify normal Verify abnormal I Verify abnormal II Patient 1 Patient 2 Patient 3 Patient 4 Patient 5 Patient 6 Patient 7 Patient 8 Patient 9 Patient 10 Electra 12.1 ± 0.5 19.0 ± 1.3 34.3 ± 1.9 15.5 ± 1.2 16.1 ± 1.5 17.5 ± 1.8 20.9 ± 1.6 23.5 ± 2.3 25.1 ± 2.7 25.3 ± 2.7 29.3 ± 2.3 31.6 ±2.3 36.8 ±2.1 Fibrometer 12.5 ± 0.4 19.7 ± 1.2 35.2 ± 2.2 15.5 ± 1.1 16.8 ± 1.5 17.4 ± 1.7 23.2 ± 2.3 28.5 ± 2.9 29.0 ± 2.3 29.0 ± 2.3 31.2 ± 2.4 32.8 ± 2.6 38.4 ± 3.0 * Means ± standard deviations of quadruplicate determinations on the same plasmas. Verify samples are General Diagnostics quality control plasmas. per ml. plasma showed significantly shorter APTT's when done by the Fibrometer. method compared with the Electra. Contrary to results observed with the APTT's, prothrombin times, whether done on normal and abnormal control plasmas, or plasmas from patients receiving warfarin therapy, showed no significant differences between the Electra and Fibrometer methods. Overall means ± standard deviations for the ten abnormal plasmas shown in Table 4 when done by the Electra method were 24.2 ± 7.0 seconds, and by the Fibrometer method, 26.0 ± 7.6 seconds. In addition to having a different means of end-point detection, the Fibrometer method differs from the automated and manual optical methods in that the former is performed in plastic cups rather than in glass tubes, there is introduction of two metal probes into the plasma, and there is constant mixing. Although the use of plastic tubes rather than glass would not be anticipated to result in shorter APTT's, the constant rapid mixing action of the probe might induce more rapid intermolecular interaction between the plasma coagulation factors, thus shortening the clotting times. APTT's on plasmas deficient in factors II and V seem to be specifically accelerated with the Fibrometer method, and this may be the basis for the shorter APTT's observed when plasmas from patients who have hepatic disease or intravascular coagulation and plasmas from warfarin-treated patients are analyzed by this method. Factor VIII deficiency does not seem to cause a significant difference between results with the two types of methods, a finding consistent with a previous study in which good reproducibility and a high degree of correlation was found between factor VIII assays when done by a Fibrometer method compared with a manual optical method. 9 No difference in prothrombin times was found with the two methods, suggesting that mechanisms in addition to differential sensitivities to factors II and V must be operating to cause the observed difference in APTT's. The present study does not provide evidence to indicate that one type of method may be superior to another, but does suggest that results of APTT's in many instances are not comparable when performed by the different methods, and that these methods should not be used interchangeably within the same hospital or outpatient laboratory. Acknowledgments. Sandra Butts, Betty Davis, Ary Epstein, Ingrid Jensen, Doris Krimgold, Cathy Leeds, Amor Lesaca, Cathy, Mayerchak, and Mary Lou Salcedo provided technical assistance. References Davey FR, Oates RP: Evaluation of an automated method for the determination of the activated partial thromboplastin time. Am J Clin Pathol 61:834-838, 1974 Han P, Ardlie NG: The influence of ph, temperature, and calcium on platelet aggregation:
August 1975 COMPARISON OF APTT METHODS 247 Maintenance of environmental ph and platelet function for in vitro studies in plasma. Br J Haematol 26:373-389, 1974 3. Hirsh J, O'Sullivan EF, Gallus AS: The activated partial thromboplastin time in the control of heparin treatment. Aust Ann Med 4:334-337, 1970 4. McPhedran P, Clyne LP, Ortoli NA, et al: Prolongation of the activated partial thromboplastin time associated with poor venipuncture technic. Am J Clin Pathol 62:16-20, 1974 5. Morin RJ, Richards D: Effect of freezing heparinized plasma on the activated partial thromboplastin time. Am J Clin Pathol 61:823-827, 1974 6. Palkuti H, Longberry J: Activated partial thromboplastin reagents. Am J Clin Pathol 61:447-448, 1974 7. Poller L, Thomson JM: The partial thromboplastin (cephalin) time test. J Clin Pathol 25:1038-1044, 1972 8. Sibley C, Singer JW, Wood RJ: Comparison of activated partial thromboplastin reagents. Am J Clin Pathol 59:581-586, 1973 9. Simone JV, Vanderheiden J, Abildgaard CF: A semiautomatic one-stage factor VIII assay with a commercially prepared standard. J Lab Clin Med 69:706-712, 1967 10. Soloway HB, Cox SP, Donahoo JV: Sensitivity of the activated partial thromboplastin time to heparin. Am J Clin Pathol 59:760-762, 1973 11. Soloway HB, Cornett BM, Grayson JW Jr: Comparison of various activated partial thromboplastin reagents in the laboratory control of heparin therapy. Am J Clin Pathol 59:587-590, 1973 12. Stuart RK, Michel A: Monitoring heparin therapy with the activated partial thromboplastin time. Can Med Assoc J 104:385-388, 1971