Do Normal d-Dimer Levels Reliably Exclude Cerebral Sinus Thrombosis? A Solution of Problems?
To the Editor:
We thank Dr Kosinski et al1 for their fine work on d-dimers in cerebral sinus thrombosis (CST) which confirms earlier results from Switzerland by Lalive et al.2 CST often offers a chameleon of symptoms ranging from headache to focal neurologic deficits and thus causes diagnostic difficulties. Both studies suggest that normal levels could reliably rule out CST in patients with symptoms lasting no longer than 2 weeks. In contrast to the high sensitivity stood the rather low positive predictive value (55.7% in Kosinski’s study). We also have to bear in mind that patients with pregnancy or malignancy (among other causes of elevated levels) were excluded in this study representing a group with elevated d-dimer levels and a higher risk of developing CST. In addition, all patients were primarily seen by neurologists. Back in everyday practice, the positive predictive value appears to be lower. Since the Stroke publication in December 2004, patients were transferred to us in most cases with “classic” headache-like migraine or tension headache but (unfortunately) in connection with elevated d-dimers. None of them showed further clinical signs of CST, but elevated d-dimers prompted further investigations in some of them, mainly MRI or contrast computer tomography. None of them showed a pathologic finding responsible for the symptoms. This illustrates that the studies on one hand helped to exclude CST and on the other hand caused further expensive and time-consuming examinations. In earlier days, no further investigations would have been performed.
Thus, we should not forget that multiple reasons can cause elevated d-dimer levels, and the positive predictive value for sinus thrombosis should even be lower than reported.
Kosinski CM, Mull M, Schwarz M, Koch B, Biniek R, Schlafer J, Milkereit E, Willmes K, Schiefer J. Do normal d-dimer levels reliably exclude cerebral sinus thrombosis? Stroke. 2004; 35: 2820–2825.
Lalive PH, de Moerloose P, Lovblad K, Sarasin FP, Mermillod B, Sztajzel R. Is measurement of d-dimer useful in the diagnosis of cerebral venous thrombosis? Neurology. 2003; 61: 1057–1060.
We thank Vatankhah et al for their comment because they are addressing a common problem regarding the use of diagnostic tools with a primarily high negative but poor positive predictive value. If such a test is used rather broadly and in an unselected patient population as if it were a screening test it may indeed lead to confusion and prove useless.
In fact, the difficulty of false-positive results using d-dimers has been studied quite extensively in the more frequent situation of patients suspected of other thromboembolic diseases, such as deep venous thrombosis or pulmonary embolism. Although broadly used in these patients, the positive predictive value of d-dimers is rather low, in most studies even lower than it was in our study on patients suspected of having cerebral sinus thrombosis (CST). To overcome this problem many studies suggest the combination of d-dimers with the use of clinical prediction rules, such as the Wells score.1 Following these clinical predictions, Linkins et al2 even suggest the use of different d-dimer cut-off values with a relatively high d-dimer cut-off point (lower sensitivity and higher specificity) in patients with a low clinical pretest probability. For patients suspected of having CST, however, no such pretest probability scoring exists, and because of its rarity it is very unlikely that a similar approach will ever be examined in a proper study for CST.
Thus, so far neither from our3 nor from any other study can we conclude that the use of d-dimers as a routine test in all headache patients coming to emergency wards can be recommended. d-dimers should only be used for exclusion in patients suspected by a neurologist to have CST. Patients with “classic” headache- like migraine, as Vatankhah et al stated, who happen to have elevated d-dimers do not automatically need to undergo further neuroradiological diagnostics. We cannot say it any clearer than Ziegler et al4 stated for the interpretation of elevated d-dimer values: “… this parameter is not suited for routine screening. Its high predictive-negative value is proved for the exclusion of thrombosis or pulmonary embolism in case of negative test result. Because a range of diseases and physiological conditions lead to increased d-dimer values, a positive d-dimer result does not verify the diagnosis of thromboembolism.”
Linkins LA, Bates SM, Ginsberg JS, Kearon C. Use of different d-dimer levels to exclude venous thromboembolism depending on clinical pretest probability. J Thromb Heamost. 2004; 2: 1256–1260.
Kosinski CM, Mull M, Schwarz M, Koch B, Biniek R, Schläfer J, Milkereit E, Willmes K, Schiefer J. Do normal d-dimer levels reliably exclude cerebral sinus thrombosis?. Stroke. 2004; 35: 2820–2825.