Andrew Stone, MD
Providence VA Medical Center
Providence, Rhode Island, USA

References

Thrombolytics in Pulmonary Embolism: Risk Stratification and Timing

Treatment options for pulmonary embolism (PE) are limited. They include full-dose anticoagulation alone or combined with thrombolytic therapy. Surgical thrombectomy and catheter embolectomy also are options. The high mortality rate of patients with hemodynamically unstable PE warrants aggressive treatment (thrombolytic therapy unless contraindicated, or surgical thrombectomy or catheter embolectomy). Hemodynamically stable patients without right ventricular (RV) strain have low mortality rates (1.5%) and do well with anticoagulation alone.(1) Controversy remains about how to assess, diagnose and treat patients who are hemodynamically stable and have RV strain (e.g., submassive PE). Konstantinides et al (2) demonstrated that among 256 patients with submassive PE randomized to receive heparin plus alteplase or heparin plus placebo, the time to administer alteplase did not change morbidity and mortality rates. Those patients in the heparin and placebo group who required escalation of treatment with later administration of alteplase had similar morbidity and mortality rates compared to those who were randomized to receive heparin plus alteplase at the start of the trial. Mortality was 3.4% and 2.2% in the heparin plus alteplase group compared to the heparin plus placebo group (P=0.71), respectively.

It is unclear when more aggressive treatment is indicated for hemodynamically stable patients with RV dilatation and hypokinesis. Much debate exists about whether the benefits of thrombolytics outweigh the risks. Some evidence suggests that mortality rates can be high in this group (range 5.0%-15.6%). (1,3) It may be important to identify high-risk patients with submassive PE who may benefit from thrombolytic therapy.

Diagnosing Right Ventricular Dysfunction
With a full arsenal of readily available diagnostic tools, patients with RV dysfunction can be stratified according to risk. In a prospective multivariate analysis, Geibel et al (4) evaluated 508 patients with acute massive or submassive PE. They suggested that the presence of electrocardiographic (EKG) abnormalities was associated with an increased risk of death (odds ratio [OR], 2.56; 95% confidence interval [CI], 1.49-4.57; P<0.001). An EKG may be used to screen hemodynamically stable PE patients for RV dysfunction.

Retrospective studies have demonstrated that multidetector computed tomography (CT) is becoming more accurate at diagnosing both pulmonary embolism and right heart dysfunction. Schoepf et al (3) found that of 431 patients diagnosed with PE on chest CT, the 30-day mortality was 15.6% in patients with RV enlargement (defined as right/left ventricular dimension ratio >0.9) compared with 7.7% in patients without RV enlargement (hazard ratio, 5.17; 95% CI, 1.63-16.35; P=0.005). Van der Meer et al (5) determined the negative predictive value (NPV) for PE-related mortality with normal RV size on chest CT was almost 100%. Positive predictive value (PPV) was 10%. Further investigation is required before the routine use of CT scans to diagnose RV dysfunction, but preliminary data are promising. Cardiac biomarkers have been used to stratify risk. Kucher and Goldhaber (6) demonstrated elevated cardiac troponin I and troponin T levels in 11% to 50% of acute PE patients. RV dysfunction detected echocardiographically correlated with cardiac troponin elevation. Normal levels ruled out RV dysfunction with a NPV of 97% to 100%; PPV was 12% to 44%. Douketis et al (7) found that among 458 patients with hemodynamically stable PE, 14% had elevated troponin I levels (>0.5 ng/mL). This was associated with increased all-cause mortality (OR 3.5; 95% CI, 1.0-11.9). Elevated troponin level alone does not reliably predict adverse outcomes from acute PE, but it can predict those patients without RV dysfunction. Kostrubiec et al (8) examined the prognostic value of both troponin T and NT-proBNP, a natriuretic peptide. In 100 normotensive patients admitted with PE, those with NT-proBNP levels <600 pg/mL had no deaths or serious complications during 40-day follow-up. Those with NT-proBNP levels >600 pg/mL but troponin T levels <0.07 ng/mL had intermediate risk of fatal outcome. Patients with NTproBNP and troponin T elevation had a PE-related death rate of 33%.

Markers are helpful in screening patients for RV dysfunction, but an echocardiogram is required for diagnosis. Echocardiographic findings of RV dysfunction include RV enlargement and hypokinesis, leftward septal shift, and evidence of pulmonary hypertension. Binder et a (l9) demonstrated that the combination of biomarkers and echocardiography can identify patients at high risk. Of 124 patients, those with an elevated NT-proBNP level (>1000 pg/mL) and an abnormal echocardiogram were 12 times more likely to have increased rates of hospital death or complications (P=0.002). The risk of adverse outcome was increased 10-fold in patients with an elevated troponin T level and RV dysfunction (P=0.004).

The Significance of the Research
The clinical course of PE with RV dysfunction is either improved with supportive care or deterioration, requiring pressors and mechanical ventilation, results in death. Belenkie et al (10,11) induced PE progressively in dogs, creating hemodynamic instability.

As the PE burden increased, the transseptal pressure gradient decreased with a significant increase in septum-to-RV free wall diameter and a decrease in septum-to-LV free wall diameter. Leftward shift of the ventricular septum reduced stroke volume. Of interest, volume loading for resuscitation led to hemodynamic deterioration with large PE. Although these are animal models, the findings suggest that volume loading may not be beneficial in patients with PE. As such, if volume loading is potentially harmful, RV strain may be significant enough to consider use of thrombolytic therapy in hemodynamically stable patients as more fluid may result in deterioration.

Our ability to recognize which patients with RV dysfunction may deteriorate with the administration of intravenous fluids is poor, and further research should be directed at identifying patients who may be harmed with fluid administration. The algorithm in Figure 1 may be helpful for risk stratification in patients with PE.

The placement of a “sepsis catheter” in patients with PE is analogous to early goal-directed therapy for sepsis. In hemodynamically stable patients who have the potential to deteriorate, measurement of central venous pressure and central venous oxygen saturation (ScvO2) may help determine the need for thrombolytic therapy.Those with low central venous pressure and ScvO2 could be volume-loaded to euvolemia. However, patients who are volume-loaded but have persistently low ScvO2 might benefit from thrombolytic therapy, as this is indicative of persistently low cardiac output and higher risk of deterioration.

In conclusion, despite the current available armamentarium for assessing patients with PE, the decision for thrombolytics is not an easy one. The algorithm in Figure 1 may help in risk stratification of patients to appropriate therapy and ultimately help decrease mortality rates.