Key Point

Several prognostic measures and scores currently being evaluated may prove to be the most useful in determining which patients with acute pulmonary embolism can be treated on an outpatient basis.

CHICAGO—Prognostic factors have been identified in acute pulmonary embolism (PE), but there is much disagreement about their accuracy and use. At the American College of Chest Physicians’ 2007 Annual International Scientific Assembly, Colonel Lisa K. Moores, MD, attempted to clarify matters by discussing the potential role of many of these factors in differentiating between higher- and lower-risk acute PE cases. Dr. Moores also explained how prognostic factors can be used to determine when acute PE may be treated on an outpatient basis.

ASSESSING SHORT-TERM OUTCOMES IN PE

In PE, increasing age is associated with worse short- and long-term outcomes, related Dr. Moores, of the Walter Reed Army Medical Center in Washington, DC. “Interestingly, several studies have shown that men have worse short-term outcomes than women do,” she remarked. Outcomes are also worse in patients with comorbid disease, such as COPD or congestive heart failure.

“To me, the EKG is the most controversial area of research, and there are quite a few studies that range from stating that it’s helpful to stating that it’s not,” Dr. Moores said. More recent studies suggest that EKG tracings may be prognostic of poor short-term outcomes in patients with objectively confirmed PE.“To me, the EKG is the most controversial area of research, and there are quite a few studies that range from stating that it’s helpful to stating that it’s not,” Dr. Moores said. More recent studies suggest that EKG tracings may be prognostic of poor short-term outcomes in patients with objectively confirmed PE.

A new complete right bundle-branch block is an EKG finding that stands out in most PE studies as being predictive of poor short-term outcomes. T-wave inversion in the right precordial lead also seems to predict such outcomes in PE. However, the prognostic value of these findings is not as great if it is unclear whether the timing of onset of right bundle-branch block or T-wave inversion is new or old.

There has been much disagreement over likely outcomes in patients with submassive PE who are not in shock but who show signs of right ventricular dysfunction. Several of the larger analyses, such as those performed by the Matisse Study investigators (Büller et al), have shown that these patients are clearly at higher risk for poor short- and intermediate-term outcomes. Incorporating these findings into clinical practice has been difficult, though, because the number of studies is relatively small, Dr. Moores remarked.

Although mainly used by radiologists, clot burden scores in conjunction with CT are being used more frequently and have shown prognostic value in PE. Scores indicating greater than 60% vascular occlusion have been associated with significantly increased mortality. There is also a correlation between the degree of vascular occlusion and right heart dysfunction.

A newer area that is being explored in the literature is the use of quantitative CT measures of PE severity, such as the right ventricular to left ventricular (RV/LV) ratio, Dr. Moores explained. There currently is no universally accepted cutoff point indicating greater severity (or even agreement on how to measure the RV/LV ratio). However, the literature has consistently shown that as the right ventricular diameter and the RV/LV ratio rise above 1, there is a strong correlation with worsening right ventricular function and acute pulmonary hypertension.

Several qualitative CT measures of PE severity are also being studied. Leftward bowing of the intraventricular septum, for example, may suggest right ventricular overload and right ventricular dysfunction, Dr. Moores pointed out.

“The data for these CT measures are less robust, as you might imagine; it’s a newer body of literature,” she said. These data, which are from nine retrospective studies with a combined population of about 2,000 patients, provide some good evidence of the prognostic value of the RV/LV ratio despite variation between studies in design and in the RV/LV cutoffs used.

OUTPATIENT THERAPY FOR LOWER-RISK PATIENTS

With regard to brain natriuretic peptide (BNP) levels, it is well known that PE mortality more than doubles with every log increase in BNP levels. However, research has also associated BNP levels below 600 pg/mL with a high negative predictive value for mortality in PE (ie, the risk of death in these cases is very low). “That may come into play later as we are trying to identify the lower-risk patients,” Dr. Moores explained.

It is possible, she suggested, that these lower-risk cases could be treated on an outpatient basis. There are not much data to guide this decision, however; only two “reasonable” studies of outpatient PE management have been conducted, said Dr. Moores. One of these studies, completed about eight years ago, evaluated outcomes in 108 patients with acute PE who were all treated with dalteparin essentially as outpatients.

Over the next 30 days, the recurrence rate was 5.6%, and bleeding incidence was 2%. Dr. Moores acknowledged that the recurrence rate was a bit high for typical low-risk acute PE patients aggressively anticoagulated in the first 24 hours, and there was no control arm. “But overall, there was no significant trouble treating this group of patients in the outpatient setting,” she said.

In a more rigorous study of 504 acute PE patients randomized to outpatient treatment with dalteparin or tinzaparin, the recurrence rate of venous thromboembolic events was 3.6% for dalteparin and 3.9% for tinzaparin—more in line with what one would expect in this population. Furthermore, outpatient therapy did not raise bleeding complication rates.

IS THE PESI MORE ACCURATE THAN THE GENEVA SCORE?

In recent years, two particularly useful scores have been developed to identify acute PE patients who are candidates for outpatient therapy—the Geneva score and the pulmonary embolism severity index (PESI), Dr. Moores said. With the Geneva score, acute PE patients are stratified into high- and low-risk groups based on clinical variables. The PESI uses those variables to assign these patients into classes I through V, with V being the highest risk group and I the lowest.

In 2007, Jiménez et al compared the Geneva score and PESI in 599 patients with objectively confirmed PE. Thirty-six percent of the patients were classified as low risk with the PESI (PESI class I or II) compared to 84% with the Geneva score. Thirty-day mortality was 0.9% for the PESI class I or II patients versus 5.6% for those deemed low risk by the Geneva score. The two scores were not associated with significantly different rates of nonfatal venous thromboembolism or major bleeding.

“The PESI appears to be more accurate and have greater clinical utility than the Geneva score,” Dr. Moores observed. CT angiography and serum biomarkers such as BNP and troponin may eventually be used to increase that clinical utility by further refining the analysis of suitability for outpatient care. CT angiography, however, will likely continue to be most useful in identifying higher-risk patients, suggested Dr. Moores.

From a purely medical standpoint, class I or II PESI patients should be considered for outpatient therapy, she concluded. Serum biomarkers may help to identify a subset of class III PESI cases that may be suitable for outpatient treatment or care in a less-monitored hospital setting.           

—Timothy Begany

Suggested Reading
Aujesky D, Perrier A, Roy PM, et al. Validation of a clinical prognostic model to identify low-risk patients with pulmonary embolism. J Intern Med. 2007;261(6):597-604.
Jiménez D, Yusen RD, Otero R, et al. Prognostic models for selecting patients with acute pulmonary embolism for initial outpatient therapy. Chest. 2007;132(1):24-30.
Moores LK. There’s no place like home. Chest. 2007;132(1):7-8.

Return to table of contents