ADJUST-PE

Published

Righini M, et al. "Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study". JAMA. 2014. 311(11):1117-1124.
PubMed • Full text • PDF

Clinical Question

In patients older than 50 presenting to the ED with clinically-suspected PE, is using an age-adjusted D-dimer cutoff versus conventional cutoff (>500 µg/L) associated with an increased diagnostic yield?

Bottom Line

Using an age-adjusted D-dimer cutoff (age × 10 in patients >50 years) in combination with pretest probability assessment is associated with a larger number of patients in whom PE can be considered ruled out with a low likelihood of subsequent clinical venous thromboembolism.

Major Points

Pulmonary embolism (PE) is a potentially life-threatening disorder that is frequently considered in the emergency medicine differential diagnoses. Given its significant morbidity and mortality, it must be rapidly ruled out. The presentation of PE is often non-specific and diagnostic testing and treatment can be associated with significant adverse events. D-dimer testing with a threshold of 500 μg/L has been shown to be a highly sensitive test with a correspondingly high negative predictive value among patients with low pretest probability for PE. However, several studies have shown that D-dimer levels increase with age, thereby limiting the yield and cost-effectiveness of D-dimer testing among older patients. Retrospective and meta-analysis studies suggested that a simple age adjustment (age × 10 in patients >50 years) may better define an optimal D-dimer cutoff value among older patients. ADJUST-PE sought to prospectively validate this method.

Published in 2014, ADJUST-PE was a European study of 3,346 patients who presented to the ED and were suspected of having a PE, with low or intermediate probability by revised Geneva score or modified Wells score. D-dimer testing was interpreted using an age-adjusted cutoff: 500 μg/L in patients <50 years, and age × 10 in patients >50 years. Patients with D-dimer results less than this threshold were considered to have PE excluded and did not receive anticoagulation. CT pulmonary angiography was performed in patients with a positive D-dimer, and management was dictated based on the results of CT imaging. The primary outcome was the rate of symptomatic thromboembolic events during the 3-month follow-up period among patents with negative age-adjusted D-dimer testing. Among patients with a D-dimer <500 μg/L, the 3-month risk of VTE was 0.1%. In patients with a D-dimer >500 μg/L but less than the age-adjusted cutoff value, the 3-month VTE risk was 0.3%. Among elderly patients >75 years with a low/intermediate pretest probability for PE (20% of the entire study population), 6.4% had a D-dimer result <500 μg/L, and 30.0% had a D-dimer result below the age-adjusted cutoff. Among elderly patients with a negative age-adjusted D-dimer, there were no VTE events observed during the 3-month follow-up period (risk of 0.0%).

Taken together, the results of ADJUST-PE reconfirm that a "standard" D-dimer cutoff of 500 μg/L has a high negative predictive value among patients <50 years with a low/intermediate pretest probability for PE. Additionally, ADJUST-PE validates the use of an age-adjusted D-dimer threshold of age × 10 in patients >50 years of age, which is also associated with a high negative predictive value. So long as high sensitivity D-dimer assays are available (eg, ELISA-based), it is reasonable to incorporate age-adjusted D-dimer testing into the routine outpatient and ED setting.

Importantly, this study included patients with a history of VTE (14.0%), active malignancy (12.9%), recent surgery (11.8%), and concurrent estrogen use (5.5%), which gives some confidence that the age-adjusted D-dimer can be used to rule out PE among these patient populations. Nevertheless, caution must be exercised in doing so, since the numbers of patients in these subsets are relatively small.

Guidelines

European Society of Cardiology (ESC) (2014, adapted)

• The use of validated criteria for diagnosing PE is recommended (Class IA recommendation)
• It is recommended that the diagnostic strategy be based on clinical probability assessed either by clinical judgement or a validated prediction rule (Class IA)
• Plasma D-dimer measurement is recommended in outpatients/emergency department patients with low or intermediate clinical probability, or, to reduce the need for unnecessary imaging, preferably using a highly sensitive assay (Class IA)
• In low clinical probability or PE-unlikely patients, normal D-dimer level using either a highly or moderately sensitive assay excludes PE (Class IA)

In their guidelines the ESC have acknowledged the ADJUST-PE study however no explicit recommendations have been presented to address the findings of the study.

ACP (2015, adapted)

• Use age-adjusted d-dimer (age*10 ng/mL) if ≥50 years to determine whether to pursue imaging

Design

• Multicenter, multinational prospective diagnostic management outcome study
• N=3,346 patients presenting to an ED, in whom PE is suspected
• Setting: 19 hospitals in Europe
• Enrollment: 2010-2013
• Follow-up: 3 months
• Analysis: Per-protocol
• Primary outcome: 3-month VTE rate among patients with low/intermediate pretest probability for PE and a negative D-dimer test
• Secondary outcomes:
  • 3-month thromboembolic risk in patients with low/intermediate and a D-dimer between 500 and their age-adjusted cutoff value
  • 3-month thromboembolic risk in elderly patients (age >75 years)

Population

Inclusion Criteria

• Outpatients presenting to the ED with clinical suspicion of PE, defined as:
  • Acute onset of worsening shortness of breath, or
  • Chest pain without another obvious etiology

Exclusion Criteria

• PE suspicion raised more than 24 hours after admission to the hospital
• Absence of informed consent
• Receiving anticoagulation therapy for another indication
• Allergy to contrast medium
• Impaired renal function (CrCl <30 mL/min as per Cockcroft-Gault formula)
• Life expectancy <3 months
• Ongoing pregnancy
• Geographic inaccessibility for follow-up

Baseline Characteristics

• Mean age: 63 years
• Women: 56.8%
• Personal history of VTE: 14%
• Active malignancy: 12.9%
• Surgery within 1 month: 11.8%
• Estrogen use: 5.5%
• Symptoms: Chest pain 48%, dyspnea 63%, syncope 8%, hemoptysis 4%
• Vitals: HR 87 BPM, RR 19 BPM

Interventions

• Patients with clinical suspicion of PE were then stratified based on clinical probability of PE. High probability defined as simplified, revised Geneva score >5 or modified (2-level) Wells score >4
• In patients with low/intermediate or unlikely clinical probability, a D-dimer test was performed
  • Patients with D-dimer <500 µg/L or D-dimer >500 µg/L but less than age-adjusted cutoff received no further intervention during follow-up period
  • Patients with D-dimer greater than their age-adjusted cutoff underwent CT pulmonary angiography

Outcomes

Primary Outcomes

3-month VTE risk among patients with low/intermediate pretest probability and D-dimer <500 µg/L

1 of 810 patients (0.1%, 95% CI 0.0%-0.7%)

3-month VTE risk among patients with low/intermediate pretest probability and D-dimer >500 µg/L but less than age-adjusted cutoff

1 of 331 patients (0.3%, 95% CI 0.1%-1.7%)

Secondary Outcomes

3-month thromboembolic risk in elderly patients (age >75 years)

0 of 195 patients (0.0%, 95% CI 0.0%-1.9%)

Criticisms

• Two different pretest probability assessment tools (Geneva and Wells scores) and 6 different commercial D-dimer assays were used; not all patients were managed using the exact same diagnostic tests. However, despite using 6 different high sensitivity D-dimer assays, the results were homogenous across the different assays.
• Not a randomized clinical study
• Only 1 of the 7 deceased patients with D-dimer levels higher than 500 µg/L and below the age adjusted D-dimer cutoff had an autopsy
• Prevalence of PE in this study (19.0%) was higher than that usually observed in North American studies

Funding

• Various grants from European bodies which do not clearly have industry associations.
• Authors with multiple conflicts of interest