The Utility of D-Dimer in Non-Traumatic Aortic Dissection

Author: Erik Handberg, MD (EM Resident Physician, PGY-2, NUEM) // Edited by: Sushil Jain, MD (EM Resident Physician, PGY-3, NUEM) // Expert Reviewer: Matthew Pirotte, MD

Citation: [Peer-Reviewed, Web Publication] Handberg E, Jain S (2016, February 2). The Utility of D-Dimer in Non-Traumatic Aortic Dissection. [NUEM Blog. Expert Peer Review by Pirotte MJ]. Retrieved from http://www.nuemblog.com/blog/ddi-aortic-dissection/


Introduction

Histopathological image of dissecting aneurysm of thoracic aorta. 

Aortic dissection is the cleavage of the media layer of the aorta by a column of blood. Acute aortic dissection is a potentially catastrophic cardiovascular disease, and is a somewhat rare but exceedingly important diagnosis to make. The gold standards of diagnosis, CT angiography of the chest/MR angiography of the chest/Transesophageal echocardiography, are useful but concerns about radiation, cost, availability, and the low incidence of this disease may be preventing physicians from using these tests as “rule-out” first lines. A low-risk, cost-efficient diagnostic test with the ability to reliably exclude this disease would be useful. Is the D-dimer that test?


Usefulness of D-Dimer in Aortic Dissection

D-Dimer, a degradation product of cross-linked fibrin, has been reported to be elevated in acute aortic dissection, and some recent studies have suggested it may be a useful screening tool to rule-out acute aortic dissection. Initial studies were inconclusive due to limited sample sizes, variable D-Dimer cut-off values, and some studies were without a control group.  More recent studies, including a meta-analysis, have improved on or removed some of those early faults.

It is important to note that when trying to create a “rule-out” test, it is vital that the test be incredibly sensitive. It is less vital that it be specific, though the lower the specificity of the test, the less useful it will be in avoiding further diagnostic studies.


Argument for use of D-Dimer in Aortic Dissection

Shimony et al [1], in a 2011 meta-analysis which excluded studies without control groups and used a set cut-off point of 500ng/mL, found a pooled sensitivity of 97% (with confidence intervals of 0.94-0.99), with a negative predictive value of 96% (with confidence intervals of 0.93-0.98). They came to the conclusion that D-Dimer may help clinicians in deciding whether or not to pursue diagnostic imaging, also acknowledging that their sensitivity and NPV were high but not 100% and a missed diagnosis could be catastrophic.

One of the studies included in the above meta-analysis, Suzuki et al [2], a multi-center prospective study with controls, came to the conclusion that D-dimer levels may be useful to rule out acute aortic dissection if used within the first 24 hours after symptom onset. They based this on a negative predictive value of 95%. They note that this NPV may be biased based on their inclusion criteria of “suspicion for aortic dissection,” in institutions which typically care for aortic dissection. 

Sbarouni et al [3], also included in the above meta-analysis, found a sensitivity of 94% using a cutoff value of 700ng/mL, but had a small sample size and was confined to a single center. Nonetheless, the authors concluded this was sufficient sensitivity for a “rule-out” test.

Ohlmann et al [5], another in the above meta-analysis, found a sensitivity of 99%, only having a single false-negative in a patient with an intramural hematoma. This study, like the one from Wiegand [4], was a retrospective case series including only patients with imaging-confirmed aortic dissections that also had a D-dimer sent at the time of presentation.

Hazui et al [6], included in the above meta-analysis, looked specifically at differentiating patients with acute myocardial infarction and those with acute ascending aortic dissection. Within the results of their study, they report that when using a cut-off of 800ng/mL (double their institutional “upper limit of normal”), the D-Dimer was 93.8% sensitive within 90 minutes of pain onset, 90.9% sensitive in the following 90 minutes, and 100% sensitive beyond that. (This paper makes a separate but interesting point about differentiation between acute ascending aortic dissection and acute myocardial infarction with regard to appropriate therapeutic intervention, i.e., avoiding giving thrombolytic therapy to aortic dissection). This is interesting because it states above that the meta-analysis concluded that dimer would be useful in the first 24 hours but this study here shows 90% sensitivity after 90 minutes.

Akutsu K et al [7] found 100% sensitivity for D-Dimer in acute aortic dissection within the first hour using a cut-off of 500ng/mL. Eggebrecht H, et al [8] also found a sensitivity of 100%, and re-iterates the point raised by Hazui, et al that consideration for acute aortic dissection should be given due to the catastrophic effects of administering thrombolysis in that setting.  Weber et al [8] in their 2003 article in Chest also found a sensitivity of 100% (which seems to be what set the other studies into motion).


Argument against use of D-Dimer in Aortic Dissection

Suzuki et al [2] mention in their discussion that intramural hematoma, a distinct sub-entity wherein there is bleeding into the aortic wall but without communication into the lumen of the aorta, may not have D-dimer elevation but carries a similar prognosis to typical aortic dissection.

Wiegand, Koller, and Bingisser [4] in their Letter to the Editor for Swiss Medical Weekly report that in their single-center retrospective case series of 25 patients with imaging-confirmed acute aortic dissection, in whom a D-Dimer had been sent, the D-Dimer sensitivity was only 88%. They contrast that to the sensitivities listed in early studies, and state that until larger studies are available and clearer justification of a cut-off value are given, it would be unsafe to use D-Dimer as a “rule-out” tool. Notably, they do not state how many confirmed aortic dissections were present in their database without a D-dimer sent, nor give clarification on level of suspicion (low/moderate/high) for dissection at time of presentation for the cases included in their series.

Hazui et al [6] note sensitivities of 90.9%-93.8% within the first 3 hours after symptom onset, raising concern that in those cases of emergent presentation and evaluation, one or two patients in twenty may have a false negative D-dimer in setting of acute ascending aortic dissection. However, they do note a 100% sensitivity after 180 minutes, suggesting the possible utility of serial D-dimer measurement.

Paparella et al [10] report in their study that 11 of 61 patients with confirmed acute aortic dissection that had D-Dimers sent preoperatively were below a cut-off point of 400ng/mL (sensitivity of 82%), a similar finding to the smaller study by Wiegand et al. These findings are in stark contrast to those produced by the studies included in the systematic review listed above, possibly due to the systematic review's exclusion of studies without a control group.

And, ACEP’s word on D-Dimer in acute aortic dissection published in 2015 is as follows: 

Level C recommendations. In adult patients with suspected non-traumatic thoracic aortic dissection, do not rely on D-dimer alone to exclude the diagnosis of aortic dissection.


Bottom Line:

  • A negative D-dimer makes it very unlikely that someone is having an acute aortic dissection, however false negatives remain possible

  • A positive D-dimer is absolutely not diagnostic of aortic dissection, and further imaging is required to make the diagnosis

  • There are multiple D-Dimer assays currently in use, with variable sensitivities. It is important to know which your institution uses, and interpret study data with that in mind

  • ACEP says that D-Dimer is not sufficient to exclude the diagnosis of aortic dissection, so doing so would be outside the standard of care


Expert Review

As is so often the case in the modern era of emergency medicine, in the case of aortic dissection (AD) we find ourselves examining a biomarker and trying to reduce CT utilization. Dr. Handberg’s discussion of the relevant literature is complete and accurate. In the case of d-dimer (DD) for detection of aortic dissection, a quality meta-analysis is available, peer-reviewed, and published in a reputable journal. Shimony et al’s meta-analysis published in the American Journal of Cardiology concludes that “plasma DD may thus be used to identify subjects who are unlikely to benefit from further aortic imaging.” This is based primarily on their findings of a high sensitivity and a high negative predictive value [1].

Two reviews of the availably literature by emergency physicians have even more cautious conclusions. In Annals in 2011 Brown and Newman concluded that to make DD a useful test, pre-test probability would have to be very low and there are no validated instruments to establish pre-test probability in evaluation of AD [11]. Writing in EMDocs in 2014 Schaefer concluded that a well-tested clinical decision rule would be needed before using DD to exclude AD [12].

Analysis of individual papers and meta-analyses aside, Dr. Handberg appropriately points out that the American College of Emergency Physicians has a very clear position on this [13]. I would caution any practitioner of emergency medicine against adopting a practice pattern that deviates from College guidelines towards a less aggressive workup of an extremely dangerous disease process.

My take on this has always been that this is not ready for clinical use and may never be. In general if we are considering aortic dissection as a real part of a differential diagnosis then we are probably fairly worried about the patient in front of us. The mortality for this condition is extremely high, up to 27% when both type A and B dissections are considered. While we should generally be judicious in our use of CT scanning and contrast dye, a subset of patients presenting to the ED with symptoms possibly suggestive of acute AD is simply not the population in which to aggressively attempt to decrease utilization. Much time and many pages have been devoted in recent years to the evaluation of low risk chest pain patients but a patient with any of the features of aortic dissection (sudden onset pain, severe pain, hypertension, etc.) is not low risk [14].  

If aortic dissection is a real possibility I think we owe it to our patients to utilize the gold standard to exclude it.


Matthew Pirotte, MD, FACEP
Assistant Professor; Assistant Residency Director; Department of Emergency Medicine, Northwestern University, Feinberg School of Medicine [Pubmed]


References

  1. Shimony A, et al. Meta-analysis of usefulness of d-dimer to diagnose acute aortic dissection. Am J Cardiol. 2011;107(8):1227.
  2. Suzuki T, et al. Diagnosis of acute aortic dissection by D-dimer: the International Registry of Acute Aortic Dissection Substudy on Biomarkers (IRAD-Bio) experience. Circulation. 2009;119(20):2702.
  3. Sbarouni E, et al. D-dimer and BNP levels in acute aortic dissection. Int J Cardiol. 2007;122(2):170.
  4. Wiegand J, et al. Does a negative D-dimer test rule out aortic dissection? Swiss Med Wkly. 2007;137(31-32):462.
  5. Ohlmann P, et al. Diagnostic and prognostic value of circulating D-Dimers in patients with acute aortic dissection. Crit Care Med. 2006;34(5):1358.
  6. Hazui H, et al. Simple and useful tests for discriminating between acute aortic dissection of the ascending aorta and acute myocardial infarction in the emergency setting. Circ J. 2005;69(6):677.
  7. Akutsu K, et al. A rapid bedside D-dimer assay (cardiac D-dimer) for screening of clinically suspected acute aortic dissection. Circ J. 2005;69(4):397.
  8. Eggebrecht H, et al. Value of plasma fibrin D-dimers for detection of acute aortic dissection. J Am Coll Cardiol. 2004;44(4):804.
  9. Weber T, et al. D-dimer in acute aortic dissection. Chest. 2003;123(5):1375.
  10. Paparella D, et al. D-dimers are not always elevated in patients with acute aortic dissection. J Cardiovasc Med (Hagerstown). 2009;10(2):212.
  11. Brown, M.D. and D.H. Newman, Evidence-based emergency medicine. Can a negative D-dimer result rule out acute aortic dissection? Ann Emerg Med, 2011. 58(4): p. 375-6.

  12. Schaefer, T. D-dimer in aortic dissection workup. 2014[cited 2016 January 1]; Available from: http://www.emdocs.net/d-dimer-in-aortic-dissection-workup/.

  13. American College of Emergency Physicians Clinical Policies Subcommittee on Thoracic Aortic, D., et al., Clinical policy: critical issues in the evaluation and management of adult patients with suspected acute nontraumatic thoracic aortic dissection. Ann Emerg Med, 2015. 65(1): p. 32-42 e12.

  14. Hagan, P.G., et al., The International Registry of Acute Aortic Dissection (IRAD): new insights into an old disease. JAMA, 2000. 283(7): p. 897-903.

Posted on February 2, 2016 and filed under Cardiovascular, EBM.