Something old, something new…

Diagnosing recurrent ipsilateral deep vein thrombosis (DVT) is very challenging. In this issue of Blood, van Dam et al have prospectively tested the outcome of using magnetic resonance direct thrombus imaging (MRDTI) to exclude suspected recurrent DVT in patients with a previous ipsilateral thrombus.¹ 

Around half of patients who experience lower extremity (LE) DVT will have subsequent postthrombotic changes in LE veins on repeat imaging.²  Compression ultrasound, the most commonly used imaging test for suspected DVT, has limited ability to distinguish new DVTs from previous postthrombotic changes.

Because previous DVT is one of the most important risk factors for recurrent DVT, many patients are evaluated for new thrombotic disease. The stakes for the patients are high, making accurate diagnosis crucial. Recurrent DVT is treated with anticoagulation, which carries significant risk of bleeding. Furthermore, when DVT recurs, a decision is often made to offer indefinite (lifetime) anticoagulant therapy.³  Therefore, accurately distinguishing new DVT from damage resulting from previous disease can have lifelong implications for patients.

The most common technique for distinguishing new DVT from previous postthrombotic changes on ultrasound relies on assessing a change in the residual diameter of the vein. A new thrombus would be expected to further reduce the diameter of the vein still open to flow when compared with a previous measurement. This approach has many limitations. It requires imaging to be performed at the conclusion of anticoagulant therapy after the initial DVT, which does not always happen. Even when imaging is performed, the previous images must be available for direct comparison. In addition, the system for comparing residual diameters has a gray zone of measurement in which diagnosis remains uncertain.⁴ 

MRDTI has been proposed as a means of overcoming the limitations of ultrasound. MRDTI distinguishes a new thrombus on the basis of signal alterations from the presence of methemoglobin, which rapidly disappears as a thrombus ages.⁵  To date, small studies have explored this method, but it has not previously been tested in an adequately powered prospective clinical trial.

van Dam et al address this problem with a prospective diagnostic management study examining the rate of venous thromboembolism (VTE) in 90 days after a negative MRDTI in patients suspected of recurrent ipsilateral DVT. After an MRDTI was shown to be negative for recurrent DVT, patients received no further antithrombotic therapy, were observed for 90 days, and underwent objective testing for any symptoms of new VTE.

A management study (also referred to as diagnostic outcome study) is the best design for this question, because there is no gold standard imaging test that can serve as a basis for comparison in this clinical setting. Furthermore, management studies may be preferable to diagnostic accuracy studies (which compare the sensitivity and specificity of one diagnostic test to another) for determining the clinical standard of care, because this design better mimics clinical practice and focuses on outcomes important to patients.

The work by van Dam and colleagues is well executed and adheres to most methodologic standards of quality for this type of study design. Enrollment was sequential (important for limiting bias in the absence of randomization), the rate of patients lost to follow-up was low, and suspected recurrent DVT during the follow-up period was independently adjudicated.

The principal finding was reassuring: only 1.1% of patients with suspected recurrent DVT excluded by MRDTI had a new thrombotic event during follow-up. This rate would be consistent with the criteria for adoption suggested in the American College of Clinical Pharmacy (ACCP) guidelines. Even though the upper boundary of the 95% confidence interval was higher than the standard suggested by ACCP, this is likely attributable to the study being modestly underpowered.

There were 2 protocol modifications. The first was added to address an unanticipated frequent finding of superficial vein thrombosis (SVT) in patients. Although this modification made the study’s results less clear, it was necessary to address the possibility that use of anticoagulants for SVT would interfere with the outcome of subsequent DVTs. The other modification changed the inclusion criteria to allow enrollment of patients already receiving anticoagulation—a reasonable, pragmatic step that addressed a common characteristic of the screened population and likely makes the study more closely mimic the clinical environment. The authors managed these changes by analyzing a subgroup of patients who were enrolled and managed according to the original protocol.

Overall, this study moves MRDTI forward as a viable option for assessing suspected recurrent ipsilateral DVT. Availability of the technology and cost remain potential barriers. A larger study that includes centers from additional countries would be welcome.