Introduction

In order to answer Question 18 in this review, it is necessary to document factor V Leiden or prothrombin G20210A mutation status in a cohort of individuals with an initial episode of venous thrombosis and then follow the cohort for a period of time to determine who develops a recurrence. Selection criteria for including published studies in the present analysis require a mean of at least four years' follow-up. Five studies satisfy these criteria for factor V Leiden, along with four studies for prothrombin G20210A. All of these studies include only Caucasians (both Hispanic and non-Hispanic). Rates are likely to be different for blacks (Question 22), as the prevalence of factor V Leiden is close to zero.

Sensitivity and Specificity of Factor V Leiden Testing

Question 18 asks how often a positive test for factor V Leiden (FVL) is associated with a recurrent episode of venous thrombosis among individuals who have been diagnosed with an initial episode. This defines the clinical sensitivity of a DNA test (see Appendix for description of 2 x 2 tables and the calculation of sensitivity and specificity). Figure 18-1 shows the mean clinical sensitivity and 95 percent confidence interval among the five selected studies. The confidence interval around the group estimate is calculated using the DerSimonian and Laird random effects model with an accompanying test of heterogeneity. According to the chi-square analysis, these studies are not statistically heterogeneous in their estimate of sensitivity. A total of 1637 individuals was followed for a minimum of 4 years, 385 of whom had a factor V Leiden mutation. The overall sensitivity of factor V Leiden mutation testing for predicting a recurrent episode of venous thrombosis in these studies is 28 percent. Table 18-1 shows the sizes of the individual studies, their respective sensitivities and the 95 percent confidence intervals. Details of the studies used for these calculations are in the Appendix. The Appendix also lists studies excluded from the present analyses because study subjects were not fully confirmed to have a first thrombotic event.

Figure 18-1. Clinical Sensitivity of the Factor V Leiden Mutation Testing. Study reference numbers are from Table 18-1. The individual and overall mean sensitivities are shown as filled-in squares. Error bars are the 95 percent CI. The combined mean sensitivity is 28 percent (95 percent CI 13-35%). The chi-square for the test for heterogeneity is 3.43 and p > 0.05.

Question 19 asks how often a positive test for factor V Leiden is not present among individuals who do not experience a recurrence of venous thrombosis. This defines the clinical specificity of each DNA test . This aspect of clinical test performance can also be expressed as the false positive rate, which is 1-specificity. The false positive rate contributes to understanding test performance by directly expressing how often a mutation will be found among individuals who will not experience a recurrence of venous thrombosis. The false positive rates are shown graphically in Figure 18-2 for the five factor V Leiden studies, and the data on which they are based are shown in Table 18-1. For the five factor V Leiden studies, the overall false positive rate, adjusted for heterogeneity, is 19 percent. The chi-square for the test of heterogeneity in these studies' estimates of the false positive rate is significant, as shown (see Appendix for further discussion of heterogeneity). In addition, the sensitivity divided by the false positive rate gives the likelihood ratio. This is a useful estimate of the test's power to alter pre-test probability of an outcome. The likelihood ratio for factor V Leiden testing is 1.5, as shown in Table 18-1. This means that individuals found to have a factor V Leiden mutation are one and a half-times more likely to experience a recurrence than would have been known in the absence of testing. This knowledge does not provide strong clinical guidance.

Figure 18-2. Clinical False Positive Rate (1-Specificity) of the Factor V Leiden Mutation Among Individuals Who Will Not Experience a Recurrence. Study numbers are from Table 18-1. The individual and overall mean false positive rates are shown as filled-in squares. Error bars are the 95 percent CI. The combined mean false positive rate is 19 percent (95 percent CI 14-27%). The chi-square for the test for heterogeneity is 36.4 and p < 0.01.

Table 18-1. Factor V Leiden -- Clinical Sensitivity, False Positive Rate (1-Specificity), and Likelihood Ratio (LR) for Identifying Recurrent Episodes of Deep Venous Thrombosis

Sensitivity and Specificity of Prothrombin G20210A Mutation Testing 

Figure 18-3 shows the mean clinical sensitivity and 95 percent confidence interval for the four selected studies of prothrombin G20210A mutation testing. The criteria for selecting these studies were identical to those for the factor V Leiden studies. The asymmetric confidence interval around the estimate for the group of studies is calculated using the DerSimonian and Laird random effects model with an accompanying test of heterogeneity. A total of 1326 individuals was followed for a minimum of 4 years, 95 of whom had a prothrombin G20210A mutation. The overall sensitivity of prothrombin G20210A mutation testing for predicting a recurrence in these studies is 11 percent. The chi-square for the test of heterogeneity in these studies' estimates of the sensitivity is significant. Table 18-2 shows the size of the individual studies, their respective sensitivities and the 95 percent confidence intervals. Details of the studies used for these calculations are in the Appendix. The Appendix also lists studies excluded from the present analyses because study subjects were not fully confirmed to have a first thrombotic event.

Figure 18-3. Clinical Sensitivity of Prothrombin G20210A Mutation Testing. Study numbers are from Table 18-2. The individual and overall mean sensitivities are shown as filled-in squares. Error Bars are the 95 percent CI. The combined mean sensitivity is 11% (95% CI 6-21%). The chi-square for the test for heterogeneity is 23.3 and p <0.01.

The clinical false positive rate (1-specificity) for prothrombin G20210A mutation testing in the four selected studies is shown in Figure 18-4. The data on which this figure is based are found in Table 18-2. The overall false positive rate is six percent. The chi-square for the test of heterogeneity in these studies' estimates of the false positive rate is significant. Table 18-2 also shows that overall likelihood ratio for prothrombin G20210A testing is 1.8. This is similar to factor V Leiden testing and demonstrates that prothrombin G20210A testing does not provide strong clinical guidance.

Figure 18-4. Clinical False Positive Rate (1-Specificity) for Prothrombin G20210A Mutation Testing for Individuals Who Will Not Experience a Recurrence. Study numbers are from Table 18-2. The combined mean false positive rate is 6 percent (95 percent CI 5.7-7%). The chi-square for the test for heterogeneity is 5.48 and p <0.01.

Table 18-2. Prothrombin G20210A – Clinical Sensitivity, False Positive Rate (1-Specificity), and Likelihood Ratio (LR) for Identifying Recurrent Episodes of Deep Venous Thrombosis

Berlin, J. A., Laird, N. M., Sacks, H. S., and Chalmers, T. C. 1989. A comparison of statistical methods for combining event rates from clinical trials. Stat Med 8 (2) : 141-51.

De Stefano, V., Martinelli, I., Mannucci, P. M., Paciaroni, K., Chiusolo, P., Casorelli, I., Rossi, E., and Leone, G. 1999. The Risk of Recurrent Deep Venous Thrombosis among Heterozygous Carriers of Both Factor V Leiden and the G20210A Prothrombin Mutation. N Engl J Med 341 (11) : 801-806.

De Stefano, V., Martinelli, I., Mannucci, P. M., Paciaroni, K., Rossi, E., Chiusolo, P., Casorelli, I., and Leone, G. 2001. The risk of recurrent venous thromboembolism among heterozygous carriers of the G20210A prothrombin gene mutation. Br J Haematol 113 (3) : 630-5.

Eichinger, S., Pabinger, I., Stumpflen, A., Hirschl, M., Bialonczyk, C., Schneider, B., Mannhalter, C., Minar, E., Lechner, K., and Kyrle, P. A. 1997. The risk of recurrent venous thromboembolism in patients with and without factor V Leiden. Thromb Haemost 77 (4) : 624-8.

Kearon, C., Gent, M., Hirsh, J., Weitz, J., Kovacs, M. J., Anderson, D. R., Turpie, A. G., Green, D., Ginsberg, J. S., Wells, P., MacKinnon, B., and Julian, J. A. 1999. A comparison of three months of anticoagulation with extended anticoagulation for a first episode of idiopathic venous thromboembolism. N Engl J Med 340 (12) : 901-7.

Lindmarker, P., Schulman, S., Sten-Linder, M., Wiman, B., Egberg, N., and Johnsson, H. 1999. The risk of recurrent venous thromboembolism in carriers and non-carriers of the G1691A allele in the coagulation factor V gene and the G20210A allele in the prothrombin gene. DURAC Trial Study Group. Duration of Anticoagulation. Thromb Haemost 81 (5) : 684-9.

Margaglione, M., D'Andrea, G., Colaizzo, D., Cappucci, G., del Popolo, A., Brancaccio, V., Ciampa, A., Grandone, E., and Di Minno, G. 1999. Coexistence of factor V Leiden and Factor II A20210 mutations and recurrent venous thromboembolism. Thromb Haemost 82 (6) : 1583-7.

Miles, J. S., Miletich, J. P., Goldhaber, S. Z., Hennekens, C. H., and Ridker, P. M. 2001. G20210A mutation in the prothrombin gene and the risk of recurrent venous thromboembolism. J Am Coll Cardiol 37 (1) : 215-8.

Ridker, P. M., Miletich, J. P., Stampfer, M. J., Goldhaber, S. Z., Lindpaintner, K., and Hennekens, C. H. 1995. Factor V Leiden and risks of recurrent idiopathic venous thromboembolism. Circulation 92 (10) : 2800-2.

Rintelen, C., Pabinger, I., Knobl, P., Lechner, K., and Mannhalter, C. 1996. Probability of recurrence of thrombosis in patients with and without factor V Leiden. Thromb Haemost 75 (2) : 229-32.

Simioni, P., Prandoni, P., Lensing, A. W. A., Manfrin, D., Tormene, D., Gavasso, S., Girolami, B., Sardella, C., Prins, M., and Girolami, A. 2000. Risk for subsequent venous thromboembolic complications in carriers of the prothrombin or the factor V gene mutation with a first episode of deep-vein thrombosis. Blood 96 (10) : 3329-3333.

APPENDIX