Introduction

D-dimer is the name given to one of a family of fibrin fragments which form and circulate in the blood-stream for several days immediately following a thrombotic event. During fibrin formation, fibrinogen is converted into fibrin by enzymatic (thrombin) cleavage of the fibrinopeptides A and B. This is followed by aggregation of the resulting fibrin monomers. Linkage of C-terminal appendages of the gamma chains by factor XIIIa results in dimerization of D-domains of adjacent fibrin monomer units. Plasmin proteolysis of this ‘cross-linked’ fibrin generates fragments DD and E as terminal products (Figure 1). In contrast, proteolysis of fibrinogen or non-cross-linked fibrin results in formation of monomeric fragment D. The dimeric D-domain (‘D-dimer’) may therefore serve as an indicator of in-vivo fibrin formation. The resulting fibrin degradation products (FDPs) comprise a heterogeneous group of molecules, characterized by the presence of multiple cross-linked D-domains 1, and serving as direct biomarkers of fibrinolysis (plasmin generation) as well as indirect markers of coagulation (thrombin generation), whose half-life (similar to that of FDPs) is approximately 8 hours in individuals with normal renal function 2.

Help me, Doctor! My D-dimer is raised

All authors

Professor Giuseppe Lippi MD, Massimo Franchini, Giovanni Targher & Emmanuel J. Favaloro

https://doi.org/10.1080/07853890802161015

Published online:

08 July 2009

Figure 1.  Origin and structure of plasma D-dimer (FIIa = activated factor II (thrombin); FXIIIa = activated factor XIII; FPA = fibrinopeptide A; FPB = fibrinopeptide B).

Display full size

D-dimer measurement plays an important role in the detection of a variety of hypercoagulable states. Local fibrin formation and lysis are part of the inflammatory response, and FDPs, including D-dimer, have been shown to modulate acute phase responses and the production of a variety of systemic inflammatory mediators 3. In fact, compromising any of the components of this intricate process (by hereditary or acquired deficiency or dysfunction) will alter the rate of fibrinolysis. Although useful as prognostic indicator of disease, D-dimer testing has limited specificity because many conditions are associated with fibrin formation (Table I).

Diagnostic value of D-dimer measurement

Thrombotic events are common and these can be a source of serious morbidity and mortality. As venography, the current gold standard test for deep vein thrombosis (DVT), is both invasive and costly, non-invasive diagnostic strategies for diagnosing DVT have been developed. Because such non-invasive tests tend to be less specific for DVT, they often have to be combined to either raise the post-test probability of a disease to a level justifying treatment, or else to lower the post-test probability to a level that warrants withholding treatment. Although the optimal strategy at individual institutions is dependent on local expertise and cost, diagnostic algorithms involving clinical assessment, venous ultrasonography, and D-dimer testing have been validated in management trials of patients with suspected venous thromboembolism (VTE). Over the past decades, D-dimer has become recognized as the best over-all biomarker during the initial assessment of patients with suspected VTE, in that a negative value safely rules out an on-going thrombotic process, both DVT 3, 4 and pulmonary embolism (PE) 3, 5. A meta-analysis, summarizing data from 78 prospective studies that compared results of different D-dimer assays with findings of objective tests (e.g. compression ultrasonography, venography, lung scanning) in patients with suspected VTE, demonstrated that enzyme-linked immunosorbent assays (ELISAs) have the best sensitivity (nearly 95%) and negative likelihood ratios (about 0.1) for excluding both DVT and PE. Although none of the D-dimer assays had positive likelihood values that greatly increased the certainty of diagnosis, it was concluded that negative ELISA results are strong evidence against DVT or PE 6. This hypothesis was supported by the outcome of a more recent meta-analysis, concluding that enzyme-linked immunofluorescence assay (ELFA), microplate ELISA, and latex quantitative assays have higher sensitivity but lower specificity compared to other D-dimer assays, thus resulting in a more confident exclusion of the disease and reducing the need for additional expensive and invasive imaging analysis 7. D-dimer testing is currently included in the clinical practice guide-line from the American Academy of Family Physicians and the American College of Physicians 8 and in the Institute for Clinical Systems Improvement (ICSI) health care guide-line 9, along with a clinical prediction rule (the one developed by Wells is by far the best known and most often applied), which also takes into account signs, symptoms, and risk factors and that can be accurately applied to categorize patients as having low, moderate, or high probability of VTE. The Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II) investigators also recommend stratification of all patients with suspected PE according to an objective clinical probability assessment and D-dimer measurement by quantitative rapid ELISA or ELFA. The combination of a negative high-sensitivity D-dimer with a low or moderate clinical probability can safely exclude PE in many patients. If PE is not excluded, contrast-enhanced computed tomographic pulmonary angiography (CT angiography) in combination with venous phase imaging (CT venography) is recommended, although CT angiography plus clinical assessment remains an option 10. Moreover, in patients with suspected PE and previous VTE, a negative D-dimer test result appears to safely rule out recurrent thrombosis 11. D-dimer may also be a useful screening test for excluding cerebral sinus and venous thrombosis in patients presenting with acute headache. In this setting, sensitivity and negative predictive values are comparable to those reported in diagnosing VTE 12. Although there is no reliable trial on the effectiveness of D-dimer for diagnosing thrombosis of the upper extremities or major abdominal veins (e.g. portal or suprahepatic), it is conventionally acknowledged that there would be no difference on D-dimer concentration between such cases and leg DVT. Dynamic monitoring of D-dimer levels can also help early diagnosis of portal system thrombosis in patients with portal hypertension 13 and in those undergoing splenectomy 14. Therefore, D-dimer measurement combined with other diagnostic modalities, such as venous duplex scan or Doppler sonography, is expected to be effective also in these peculiar settings 15, 16.

Controlled and experimental observations in both humans and animals indicate that the elevation of FDPs is a constant feature of disseminated intravascular coagulation (DIC), with marked elevations occurring rapidly after initiation 17. Wada et al. 18 demonstrated that elevations in soluble fibrin, fibrin(ogen) degradation products, and D-dimer often precede overt DIC by several days. In 2001, the Scientific and Standardisation Committee (SSC) of the International Society on Thrombosis and Hemostasis (ISTH) proposed new diagnostic criteria for diagnosing overt DIC, which include results of platelet count, prothrombin time, fibrinogen level, and a marker of the presence of fibrin, which can be ascertained among soluble fibrin monomer, fibrin degradation products, and/or D-dimer. In the 2003 meeting of the SSC on DIC, D-dimer was proposed as the ideal fibrin marker, which was attributed a score of 0, 2, and 3 points to normal, moderately, and strongly increased concentrations, respectively 19. Similarly to VTE, despite D-dimer appearing to have a high sensitivity for the diagnosis of DIC, it has, however, a low specificity 20.

The role of D-dimer testing in the diagnosis of acute arterial thrombosis is less clear. Recent evidence attests that D-dimer testing would not add any clinically meaningful information to the sole determination of the cardiospecific troponins 21. However, mounting evidence suggests a possible role of D-dimer in the acute diagnosis of ischaemic stroke subtypes, especially in identifying thromboembolic and lacunar stroke 12.

Besides venous and arterial thrombosis, elevated D-dimer levels might be encountered in a number of other clinical situations and are therefore not diagnostic of any specific condition. However, they might still retain a clinical value. Recently, it has been highlighted that D-dimer might be a useful complementary tool to the current diagnostic work-up of patients with vascular disorders, such as acute aortic dissection (AAD) 22, in elderly Alzheimer care-givers who are at increased risk of cardiovascular complications and over-all mortality 23. Since the conventional plasma biomarkers for diagnosing intestinal ischaemia have insufficient accuracy, D-dimer has been proposed as an initial exclusion test in this clinical context, regardless of its low specificity 24. In patients with sickle cell disease, measures of haemolytic rate (haemoglobin and lactate dehydrogenase) significantly correlate with D-dimer 25.

Prognostic value of D-dimer measurement

These and other potential applications signal a paradigm shift in the role of D-dimer measurement in clinical practice. A large bulk of scientific evidence has recently emerged and endorses an intriguing predictive role of raised D-dimer levels, regardless of the analytical methodology and the patient setting. The measurement of D-dimer may be of clinical interest in patients with constitutional thrombophilia, as there is often no close relationship between clinical expression and genotype to indicate the presence of hypercoagulability 26. In patients with VTE, D-dimer levels at admission are associated with both the extent of the thromboembolic burden and with the prospective risk of all-cause and thrombosis-related death 27–29. The D-dimer odds ratio for mortality varies between 3 and 7, at concentrations 6–10 times above the conventional references ranges 27, 30. Moreover, D-dimer levels measured 1 month after withdrawal of oral anticoagulant therapy have a high negative predictive value for recurrence in subjects with unprovoked VTE. The authors of this study also highlighted that thrombophilic carriers with normal D-dimer levels have a very low risk of recurrence, suggesting that a negative result may be a more useful predictor 31. Furthermore, in the setting of paediatric thrombosis, the presence of elevated levels of D-dimer, either at diagnosis or following 3–6 months’ anticoagulation, has been defined as a predictor of adverse long-term outcomes 32.

D-dimer plays an integral role in monitoring acute DIC. D-dimer levels continue to rise as DIC progresses and subsequently declines with its clinical and laboratory improvement, thus having an important prognostic value 33. Elevated D-dimer levels per se have been shown to have a negative impact on survival in patients with DIC accompanying severe sepsis, presumably reflecting an on-going and unopposed activation of coagulation and consumption of procoagulant factors 34. It has recently been reported that the use of a scoring system for DIC, based on D-dimer measurement in addition to the APACHE (acute physiology and chronic health evaluation) II score, helps predict mortality better than the APACHE II score alone, especially in patients with infections 35. Accordingly, early diagnosis and proper management of pre-DIC, using combined D-dimer and FDP assays, had a positive impact on the prognosis of high-risk paediatric patients 36.

A variety of haemostatic biomarkers have emerged as interesting predictors of risk in acute coronary syndromes (ACS). These molecules are conventionally divided into those with short plasma half-life (10 min or less: fibrinopeptide A, thrombin-antithrombin complex, etc.), which reflect on-going thrombosis and may identify patients at increased risk, and those with longer half-life (>90 min: prothrombin fragment 1 + 2, D-dimer, etc.), more useful for indicating a single acute thrombotic event 2, 37. Although the measurements of haemostatic plasma markers do not seem useful in routine clinical diagnosis of acute atherosclerotic complications, they still provide valuable prospective information. Based on the premise that thrombogenesis is one of the fundamental pathological processes underlying the major complications of atherosclerosis, measurement of D-dimer might be useful to identify patients at high risk for future averse events 38. D-dimer has been associated with risk of cardiovascular disease in studies published over the past 15 years, indicating that activated coagulation and in-vivo fibrin formation and lysis may play a role in arterial thrombosis, independently of classic risk factors such as smoking and hypercholesterolaemia 39. In the Physician's Health Study, a strong association between increased D-dimer concentrations and future myocardial infarction was observed by multivariate analysis, after adjusting for non-lipid cardiovascular risk factors 40. Future ischaemic events were also predicted by raised D-dimer levels in the Caerphilly Study 41 and in a further prospective study on healthy middle-aged subjects 42. In the following meta-analysis, including seven studies, an odds ratio of 1.7 (95% CI, 1.3–2.2) was observed for subjects with D-dimer levels in the top tertile compared with those in the bottom tertile 42. Some additional studies supported this association, revealing multivariate-adjusted risk estimates ranging from 1.60 to 4.20 38. There is also mounting evidence that inflammation and hypercoagulability play an important role in the pathophysiology of acute ischaemic stroke. Accordingly, D-dimer levels, along with those of C-reactive protein and fibrinogen, appear significantly higher in patients who die compared with those who survive 43. An excess risk of vascular death or new cardiovascular events during follow-up could also be demonstrated with increasing levels of D-dimer in first-ever ischaemic stroke patients 44. Increased values of D-dimer and other haemostatic markers are independent predictors of mortality in patients with polyvascular disease, including peripheral and coronary artery disease. Therefore, they can be helpful in risk stratification or to monitor anticoagulant therapy 37. Higher D-dimer levels among individuals with peripheral arterial disease are predictive of poorer functioning 45, impaired cognitive performances 46, and are significantly associated with later occurrence of myocardial ischaemia 47. Other studies have outlined the importance of D-dimer as a prognostic factor in hypertension. Indeed, these have documented that increased D-dimer levels are independently associated with the presence of cardiac, renal, cerebrovascular, and peripheral vascular damage in patients with essential hypertension 48, 49. However, this close association suggests that D-dimer is not just a marker of atherosclerotic complications in hypertensive patients, but may somehow contribute to the pathogenesis. Although further research is needed to elucidate the complex associations between D-dimer and cardiovascular disease, the available evidence suggests a potential clinical utility for D-dimer in prediction of cardiovascular risk, in indicating patient groups for prophylaxis and monitoring of anticoagulation.

Besides VTE, arterial thrombosis, and DIC, raised D-dimer in plasma appears frequently associated with additional underlying disorders, which might activate the coagulation process without clinical evidence of thrombosis but ultimately contribute to increase the risk of all-cause death in carriers, either primarily or indirectly. Malignancies are paradigmatic, in that patients with active cancers have a worse survival in the medium term (3–6 months), either from the consequences of the cancer itself of from the substantially increased probability to develop VTE or DIC. The association between malignancy and thrombosis has been recognized since 1865, when Trousseau first reported that thrombophilia, VTE, and particularly idiopathic VTE occur rather frequently as a paraneoplastic phenomenon (Trousseau syndrome). Patients with malignancy, in particular those with solid tumours, are therefore characterized by a net haemostatic imbalance, and the tumour itself is the origin of hypercoagulability 50. In 1993, Tripodi and colleagues defined hypercoagulability in cancer as a condition of procoagulant imbalance due to heightened enzymatic activation of coagulation zymogens, but without laboratory evidence of fibrin deposition, nor clinical signs of thrombosis 51. D-dimer is over-produced in the presence of active malignancies, and its levels are increased in patients with various solid tumours including lung, prostate, cervical, ovarian, breast, and colon 52. High D-dimer levels are correlated with active malignancy and age over 65 years, both being indicators of 3-month mortality. On the other hand, active malignancy, being an in-patient at time of diagnosis, and age over 65 years are associated with higher D-dimer levels and worse 3-month survival 30. Since increased levels of FDPs are suggestive of on-going fibrinolysis during tumour progression 50, 52, the preoperative concentration of several prothrombotic markers, especially D-dimer, is associated with relatively advanced tumour stage and short postoperative survival after curative resection 52, 53. Additional investigations demonstrated that the measurement of preoperative D-dimer level would be useful in the preoperative staging, lymph node metastasis, and prediction of postoperative survival in patients with various types of cancer 52. Although the value of screening patients with increased D-dimer for occult cancer is still debated, when other physiological or pathological causes can be ruled out, the presence of malignancy may be suspected as the potential source for a substantial D-dimer elevation. The usefulness of this approach has been highlighted in a clinical investigation of patients with DVT, which demonstrated that high D-dimer concentrations at presentation or during the first days of treatment are indicators of an increased probability of overt or occult forms of cancer, especially in patients under 60 years old 54. The increased thrombotic risk in cancer patients in particular clinical circumstances, as reflected by elevated levels of D-dimer in plasma, might also represent the rationale basis for primary thromboprophylaxis in these patients, since it would substantially reduce the rate of recurrent VTE without an increase in bleeding, thereby improving the over-all quality of life 50.

Since a high D-dimer level is associated with a poor outcome in patients with traumatic intracranial haemorrhage, it can be used in addition to neurological assessment to predict the outcome 55. Acute thrombotic complications remain a constant, proportionally increasing complication before and after renal transplantation, and D-dimer appears the strongest biochemical predictor of vascular access thrombosis among patients on the renal transplantation waiting list 56.

HELLP syndrome belongs to the group of pathological disorders associated with pregnancy-induced hypertension and may accompany pre-eclampsia. Although the basic criteria for establishing the diagnosis are H for haemolysis, EL for elevated liver enzymes, and LP for low platelets, D-dimer testing may provide a useful tool for the early identification of patients with pre-eclampsia who may develop severe HELLP syndrome 57, albeit that interpretation of D-dimer values in pregnancy remains challenging 58, as further discussed. D-dimer levels are also significantly higher in ovarian hyperstimulation syndrome (OHSS) patients with unsuccessful pregnancy outcome with respect to those with successful outcome of pregnancy, indicating that a marked hypercoagulability state can be detected in these patients and might be related to the clinical outcome 59.

Plasma D-dimer levels are also associated with increased risk of over-all mortality in patients with an acute exacerbation of idiopathic pulmonary fibrosis (IPF) 60 and are significantly elevated in patients with acute graft-versus-host disease. Accordingly, the identification of haemostatic abnormalities by reliable laboratory parameters such as D-dimer might help predict poor outcomes or complications in patients with allogeneic bone marrow transplantation 61.

Elevated D-dimer levels predict all-cause mortality rates among diabetic (relative risk, 2.36) and non-diabetic (relative risk, 3.45) patients undergoing chronic haemodialysis 62. D-dimer levels positively correlate with the severity of the disease and organ dysfunction in patients with infections, suggesting that elevated D-dimer levels may reflect the extent of microcirculatory failure 63. In patients with community-acquired pneumonia, a significant relationship is observed between elevated D-dimer levels and the Pneumonia Severity Index (PSI) and APACHE II score; raised levels are associated with radiologic pneumonia extension, and D-dimer monitoring might therefore be useful for predicting clinical outcome in these patients 64. Recently, it has also been reported that D-dimer levels are drastically increased in humans infected with Sudan Ebola virus, but they are four times higher in patients with fatal cases than in patients who survived during the most acute period of the infection 65. Therefore, since D-dimer levels might predict adversely in patients with severe infections and sepsis, it is conceivable that the therapeutic strategy in these patients may also be monitored using D-dimer plasma levels.

Management of raised D-dimer in patients with thrombosis

Regardless of the underlying cause, D-dimer might also be a useful biomarker to help determine initial antithrombotic or anticoagulant therapies. The optimal duration of oral anticoagulation therapy for VTE is still controversial, but recent studies show that D-dimer levels have a predictive value for the risk of recurrence. Basically, the amount of time that subjects with an acute unprovoked VTE event spend at near-normal International Normalized Ratio (INR) values (<1.5) during the first 3 months of treatment is associated with higher D-dimer values measured during oral anticoagulant therapy and after its interruption, and is a significant risk factor for late VTE recurrence 66. Moreover, patients with an abnormal D-dimer level 1 month after the discontinuation of anticoagulation have a significant incidence of recurrent VTE, which is reduced by the resumption of anticoagulation. The risk of recurrence in patients with normal D-dimer is significantly lower 67. Accordingly, since D-dimer testing provides valuable prognostic information in patients with several vascular disorders, including VTE, DIC, and AAD, it might be used to help regulate the duration of anticoagulation.

Management of raised D-dimer in patients without clinical evidence of thrombosis

The potential predictive role of D-dimer measurement in a variety of pathologies other than VTE and DIC opens new challenging scenarios. Firstly, in otherwise asymptomatic subjects, the source of an abnormal D-dimer level should not be overlooked. Irrespectively of whether D-dimer is raised because of occult cancer, infection, or clinically undetectable thrombosis, all patients should be given a chance to receive an early diagnosis and triage, which may substantially modify the prognosis. Secondly, even outside its typical clinical setting (clinically manifest thrombosis), D-dimer level might influence initiation or duration of anticoagulant therapy. Thirdly, elevated plasma levels of D-dimer in patients with cancer are associated with decreased survival and a poor response to treatment. Therefore, pretreatment for raised D-dimer levels in cancer patients may be useful in the prediction of survival and the response to treatment 50, 68, raising the question of whether this approach might be useful and appropriate to initiate antithrombotic prophylaxis in all patients with curable cancer 50, 69. This is also true of subjects with heart failure (HF), where hypercoagulability and inflammation assessed by D-dimer and interleukin-6 (IL-6) levels are associated with an increased mortality risk, and intervention on both conditions might improve patient outcome 70. As patients with high D-dimer levels are at high risk for thrombosis of the vascular accesses, close monitoring of these patients and use of anticoagulant therapy during the waiting period prior to renal transplantation might be advisable 56. Finally, plasma D-dimer levels are associated with mortality in patients with an acute exacerbation of idiopathic pulmonary fibrosis (IPF), and anticoagulant therapy has a beneficial effect on survival in these patients 60.

Limitations of D-dimer measurement

D-dimer assays were first developed more than 2 decades ago. Since then, many quantitative and rapid tests were made available to clinical laboratories. Most assays are typically based on monoclonal antibodies reactive with epitopes found on fibrin fragment D-dimer but not on fibrinogen fragment D, other FDPs, or native fibrinogen. The antibodies react with conformational epitopes generated by factor XIII-induced linkage of the C-terminal appendages of the fibrin gamma chains of adjacent D-domains within a fibrin polymer. For some monoclonal antibodies, degradation of the cross-linked fibrin compound by plasmin is an additional requirement for generation of the epitope. Therefore, in patient samples, D-dimer antigen assays detect an array of fibrin compounds of different molecular weights, including fibrin fragment D-dimer as well as higher-molecular-weight FDPs and fibrin X-oligomers. Most D-dimer antigens in plasma represent cross-linked soluble fibrin present in circulation rather than degradation products from particulate clots 71. Assay technologies currently include manual latex agglutination assays, automated latex-enhanced light-scattering immuno-assays, enzyme-linked immuno-assays, and others.

Although the accurate and precise measurement of plasma D-dimer levels is pivotal to the diagnosis and management of several pathological disorders, considerable variability has been reported between results obtained using different methods, in different laboratories 72. Because of the complexity of the analyte of target as well as the variability in specificity of different D-dimer assays, harmonization of the test results by per method linear regression (i.e. transformation of the method-specific regression line to a reference line common for all the assays by using a set of standard preparations containing a similar variety of cross-linked fibrin compounds), seems to be a feasible opportunity, in that it can reduce the variability between the method-specific consensus values from 75% to less than 6%. Clinical validation of this concept had shown significant improvement of the test result comparability 73. Another harmonization approach, by using a calibration model using a set of common calibrant plasmas, was effective to improve agreement of D-dimer results by different methods 74. An additional problem is the use of different units in different assays, i.e. D-dimer units and fibrinogen equivalent units (FEU).

Owing to the increasing and competing demands on primary care resources and the mounting pressure from cost containment policies, inefficient or inappropriate diagnostic testing causes considerable organizational and economic problems. Therefore, the interpretation of D-dimer measurements always requires caution 75. Both the sensitivity and specificity of D-dimer measurement depend upon the prevalence of disease in the patient population, and this might explain some of the variety in results from the different studies. Although rare, and mostly using methods based on turbidimetry or latex-enhanced nephelometry, false-negative results of D-dimer testing can be encountered in patients with VTE. Non-diagnostic value may be observed when D-dimer is measured too early or too late after the development of thrombosis. Falsely negative results can also occur after initiation of anticoagulant therapy, because anticoagulants limit clot propagation and diminish D-dimer production. Also, thrombus location and size are important determinants of D-dimer levels, in as much as lower sensitivities have been reported in diagnosing calf DVT versus above-the-knee DVT, and peripheral versus central PE. Finally, interference by irrelevant (non-cross-linked) analytes should also be considered 1, 75.

It is also important to note that the clinical value of D-dimer testing might be substantially lower during specific circumstances. The markedly heterogeneous fluctuation of plasma D-dimer suggests that the postoperative activation of the haemostatic system depends on the type of, and time since, surgery, thus limiting the clinical usefulness of D-dimer testing in the diagnosing of postoperative VTE 1, 75, 76. The over-all diagnostic efficiency of D-dimer testing is also decreased in pregnancy and in patients with recent injury or trauma. In particular, the normal pregnancy causes a progressive increase in circulating D-dimer, such that only a minority of women in the second trimester and none in the third trimester have a normal D-dimer concentration 77. Epiney et al. also observed that D-dimer values might remain above the diagnostic threshold for VTE at delivery, at day 1, and at day 3 post partum. However, a sharp decrease was observed between day 1 and day 3, followed by a slight increase at day 10. At day 30 and day 45, respectively, 79% and 93% of women in the vaginal delivery group and 70% and 83% in the Caesarean group had levels below the diagnostic cut-off 78. These data are consistent with the hypothesis that more research is needed to determine how much to increase the threshold for abnormal D-dimer in pregnancy. D-dimer measurement is also of little contribution to the diagnosis of superficial thrombophlebitis (ST) 79, and its value for ruling out VTE is also substantially decreased in patients with liver disease, since the presence of ascites and hepatocellular carcinoma is constantly associated with increased values 80. Finally, although in elderly outpatients the combination of a non-high clinical probability and a normal D-dimer result is a safe strategy to rule out VTE, this strategy appears less reliable in in-patients where several comorbidities might be present. In addition, the proportion of patients of >65 years in which this strategy excludes VTE is markedly lower compared to younger patients 81.

Conclusions

Measurement of D-dimer is important for determining not only the activation of fibrinolysis but also the severity of a hypercoagulable state. The central role of D-dimer testing in the diagnostic approach of patients with suspected VTE is now well established. Among patients with suspected VTE in whom the clinical probability of venous thrombosis is judged to be low or moderate, a negative D-dimer assay result, provided by a rapid, automated, and highly sensitive assay, enables clinicians to rule out DVT and PE in less than an hour with no additional testing. However, a high proportion of patients presenting with raised D-dimer and no clear evidence of venous thrombosis may have different or coexistent underlying conditions (Figure 2). In addition to VTE and DIC, D-dimer may be a prognostic marker in other pathologies such as cardiovascular disease, cancer, acute bowel ischaemia, acute upper gastrointestinal haemorrhage, intracranial haemorrhage, cerebral infarction, atrial fibrillation, bacteraemia, and sepsis 1. The clinical significance and practical usefulness of investigating elevation of D-dimer in these and other settings is currently debated. Due to the significant risk of morbidity and mortality associated with persistently raised D-dimer values, the accurate investigation of the underlying cause(s) and the appropriate triage of the affected individuals appear to be essential. Although further studies are needed to establish whether D-dimer, alone or combined with other prognostic instruments, can be used to identify patients who are potential candidates for further triage, treatment, and even hospital stay, such potential hazards require constant re-emphasis in the teaching of post-graduates, junior doctors, and medical students. Additional investigation is needed to clarify whether raised D-dimer is an epiphenomenon in a variety of pathological disorders or, rather, it is directly involved in the pathophysiology of these disorders. Although over-treatment of patients with anticoagulant drugs, when there is only mild suspicion of blood clots, can endanger life, the enhanced risk of morbidity and mortality associated with raised D-dimer levels alone prompts further studies to assess the cost-effectiveness of more immediate action to inhibit clot spread and decrease the probability of thrombotic incidents, even in the absence of any evident thrombosis. In this circumstance, monitoring D-dimer levels might also be appropriate to confirm the effectiveness of anticoagulation.

Help me, Doctor! My D-dimer is raised

All authors

Professor Giuseppe Lippi MD, Massimo Franchini, Giovanni Targher & Emmanuel J. Favaloro

https://doi.org/10.1080/07853890802161015

Published online:

08 July 2009

Figure 2.  Suggested algorithm for investigating raised D-dimer in plasma (CAT = computerized axial tomography; DIC = disseminated intravascular coagulation; DVT = deep vein thrombosis; PE = pulmonary embolism; VTE = venous thromboembolism).

Display full size

Take-home points

• Measurement of D-dimer is important for determining not only the activation of fibrinolysis but also the severity of a hypercoagulable state.

• In patients with raised D-dimer, the underlying cause(s) should always be investigated.

• D-dimer testing is pivotal in diagnosing venous thromboembolism (VTE), disseminated intravascular coagulation (DIC), and it may also contribute to the diagnostic work-up of vein thrombosis in atypical sites, ischaemic stroke, intestinal ischaemia, and acute aortic dissection.

• D-dimer levels predict all-cause death and complications in several pathological conditions, including VTE, DIC, cardiovascular disease, cancer, severe infections, HELLP syndrome, idiopathic pulmonary fibrosis, acute graft-versus-host disease, Alzheimer's disease, and chronic haemodialysis.

• Systematic measurement of D-dimer helps determine initial antithrombotic or anticoagulant therapies.