Objective To determine the cause of a markedly raised D-dimer among patients in whom a diagnosis of pulmonary embolism (PE) has been excluded by CT pulmonary angiogram (CTPA) with particular reference to new cases of cancer and aortic dissection.

Methods One thousand consecutive patients, suspected of PE, who had undergone CTPA and for whom a D-dimer had been requested, were seen between 2012 and 2016. Retrospectively we examined the case records of all those in the top quintile of the D-dimer distribution whose CTPA was negative for PE. D-dimer in the top quintile ranged from 7.5 to 260 times upper limit normal.

Results Eighty-five patients fulfilled our inclusion criteria. The likely causes of their very high D-dimer were infection (n=35, 41.2%), cardiovascular disease (n=12, 14.1% including two patients with previously undiagnosed aortic dissection), surgery or trauma (n=12, 14.1%), new or active cancer (n=9, 10.6% comprising six new cancers and three patients with cancers diagnosed previously that were considered to be active) and miscellaneous causes (n=17, 20.0%). Thirty-five patients (43.5%) died over a 2-year follow-up. Kaplan-Meier survival analysis showed poorer outcomes for patients with new or active cancer, when compared with those with no known cancer (p<0.001).

Conclusions We have shown that a small proportion of patients suspected of PE whose D-dimers are markedly elevated have diagnoses we would not want to miss including previously unsuspected cancer and aortic dissection. Further studies will be required to define the optimal workup of patients with extremely high D-dimer who do not have venous thromboembolism