Risk Analysis of Pulmonary Embolism by CT Imaging

Author: Steve Chukwulebe, MD (EM Resident Physician, PGY-1, NUEM) // Edited by: Michael Macias, MD (EM Resident Physician, PGY-3, NUEM) // Expert Reviewer: Mark Courtney, MD

Citation: [Peer-Reviewed, Web Publication] Chukwulebe S, Macias M (2016, February 23). Risk Stratification of Pulmonary Embolism By CT Imaging. [NUEM Blog. Expert Peer Review by Courtney DM]. Retrieved from http://www.nuemblog.com/blog/ct-in-pe/

The Case

A 55 year old male cross country truck driver presents to your emergency department with pleuritic chest pain, dry cough, and dyspnea.  Upon examination he is tachycardic but his vitals are otherwise stable.  A CTA shows an acute PE with flattening of the interventricular septum.  In the setting of a stable blood pressure, this patient meets criteria for submassive PE.

What can we infer from a radiologist’s comments about the right heart in terms of risk stratification and management?



Definition of moderate/high risk sub-massive PE

Moderate to high risk sub-massive PE must include an acute PE diagnosis on CT angiography in the setting of a stable patient (systolic blood pressure greater than 90 mmHg) with some evidence of right heart strain [1]. This evidence can include:

  • New EKG changes
    • Atrial arrhythmias
    • New right bundle branch block
    • Inferior Q-waves (leads II, III, and aVF)
    • Anterior ST-segment changes and T-wave inversion
    • S1Q3T3 pattern
  • Troponin elevation (troponin I > 0.4 ng/mL; tropoinin T > 0.1 ng/mL)
  • BNP elevation (BNP > 90 pg/mL; proBNP > 500 pg/mL)
  • Signs of right heart strain on imaging (RV/LV ratio > 0.9)

Signs of Right Heart Strain on CT [2,3]

  • Right ventricular enlargement (RV/LV ratio > 0.9)
  • Pulmonary trunk enlargement (ie. bigger than the aorta)
  • Abnormal interventricular septum ( flattening of the septum or paradoxical septal bowing)
  • Other features:
    • Inferior vena caval contrast reflux
    • Dilated azygous venous system
    • Dilated hepatic veins with/without contrast reflux 

Review of the Literature


Right Ventricular Size on CT Moderately Correlates With Echo Findings

Many studies have attempted to quantify right heart strain by measuring the ratio between the diameter of the right and left ventricles, also known as the RV/LV ratio, and looking for correlation with outcomes. While echocardiography initially was the gold standard for this measurement given it’s dynamic imaging capabilities, one study examined if this same ratio seen on CT correlated with echocardiogram findings. A study in the Journal of Thoracic Imaging in 2014 identified a cohort of 777 individuals with a diagnosis of acute PE on admission, who also underwent an echocardiographic examination within 48 hours either before or after the initial CT scan [4].  This study showed that the RV/LV diameter ratio was moderately correlated with an RV size determined by the echo.  However, the study also showed that the sooner a patient received an echo after the CT was performed, the better the correlation of the RV size on echo was to the RV/LV ratio on CT (Spearman correlation coefficient of 0.6916 vs. 0.5396 at 6 hours and 48 hours respectively).  

Evidence of Right Heart Strain on CT Correlates with Elevation in Biomarkers

Prior meta-analysis data has shown that an increase of troponins I and/or T was associated with a higher mortality than patients with normal troponin levels, even in those patients that were hemodynamically stable [6]. Taking this into account, Seon et al investigated whether CT evidence of right heart strain could predict elevation in biomarkers and thus act as a risk stratification tool [5]. This study divided 80 patients into two groups.  One group with evidence of RV dysfunction on a transthoracic echo and a second group without.  The authors then showed that signs of right heart strain (RV/LV ratio, CT index of PA clot load, contrast reflux to the inferior vena cava, and ventricular septal bowing) were significantly correlated with higher levels of NT-pro-BNP and Troponin-T.  Ultimately, they derived an ROC curves that suggests the optimal cut-off value to predict RV dysfunction was 1.12 for RV/LV ratio on CT (sensitivity: 89.8%, specificity: 77.4%).

RV/LV Ratio Predicts Short Term PE Related Death

Saddle Embolism Status Post Surgical Removal [17] 

Saddle Embolism Status Post Surgical Removal [17] 

Two studies show a significant difference in the RV/LV diameter ratio when comparing patients that survive the first few months after the diagnosis of PE with those that die.  One article in the Journal of Thoracic Imaging from 2014 shows a significant difference in the RV/LV ratio in patients that have a “PE related death” prior to 30 days after the initial diagnosis, 1.15 ± 0.34 and 1.05 ± 0.26 (no PE related death) with a p value of 0.026 [7].  That finding is echoed in another study in European Radiology in 2015 showing a similar increase in the RV/LV ratio in patients that died 60 days after initial diagnosis, 1.21 ± 0.29 vs. 1.04 ± 0.19 with a 0.04 p value [8].  The authors of the latter also note that an optimal ROC point for predicting 60 day mortality would be 1.04 (sensitivity = 71.4%, specificity = 65.3%).  However, in a logistic regression analysis, ratio alone was not an independent predictor of mortality.  A study following patients even further to 12 years after diagnosis shows no association between RV/LV ratio and all-cause mortality [9].

While most of the focus has been on the RV/LV ratio given that it is a quantifiable metric of right heart strain, other studies tried to determine the association of septal bowing with various outcomes.  One article in Radiology published in 2007 suggests that ventricular septal bowing had a low sensitivity (21% and 18%) and high specificity (88% and 87%) for predicting death due to PE within 30 days of the initial CT [10].  Though this study was a retrospective analysis of a large number of patients, 1193, only two observers were used in the initial analysis.  Observer one showed a significant association between presence of ventricular septal bowing and 30 day mortality while observer two did not, odds ratios of 1.98 vs. 1.52, and p values of 0.04 vs. 0.22 respectively.  Since observers 1 and 2 did not agree on septal bowing 10% of the time, the study authors performed a post-hoc analysis with a third observer as a tie breaker in those 122 cases.  Thus, while septal bowing may be a specific marker for 30 day mortality, this study calls into question the inter-observer reliability of a CT scan read suggesting right heart strain.  

One article in the American Journal of Roentgenology evaluates solely the reliability between two observers [11].  This was a smaller study, looking at only 50 patients with acute PE.  It showed that in regards to septal bowing and even IVC reflux there was only fair to moderate agreement in the findings between the two observers.  Yet both observers were able to agree on the RV/LV diameter ratio with a correlation coefficient of 0.93 and a p value < 0.001.

Take Home Points

  • CT scan suggesting any evidence of right heart strain identifies patients with moderate to high risk submassive PE
  • Right heart strain on CT is associated both with RV dilation on echocardiogram and elevations in troponin and BNP biomarkers.  
  • The presence of right heart strain on CT has been variably associated with short term PE related death.  
  • Caveats: No studies have shown that any one finding on CT can predict mortality.  And moreover, with the exception of the RV/LV diameter ratio, different radiologists may disagree about the presence or absence of the other markers for right heart strain far too frequently. 
In light of this evidence, further studies may be needed to determine the clinical significance of these findings in the acute setting of the emergency department.

Expert Review


This is a great summary of how to approach image based risk stratification of sub-massive PE.   The first step after diagnosis of acute PE is risk stratification into one of 3 categories: massive PE (PE with shock/hypotension), sub-massive PE (PE without shock but with right heart strain) and non high risk PE (everything else). This post quotes the consensus paper by Jaff et al and readers are best advised to check out that article in detail for a “deep dive” not just on risk stratification of PE,  but treatment as well [1].

There are a few important points that merit additional consideration:

In aggregate, mortality from “all-comers” with PE is not high. 

Despite being taught as a lethal and to be feared diagnosis, the vast majority of patients survive and mortality is way behind severe sepsis/septic shock, traumatic shock, and acute surgical aortic emergencies.  In the EMPEROR study, PE attributed mortality was 1%.  All cause mortality was 5.4% at 30 days [12]. This is similar to mortality from community acquired pneumonia or myocardial infarction. Though it is true that on the balance one would rather have a PE without elevated biomarkers (troponin especially) and without RV dilation on CT or echo,  one should have caution in not reading too much into these findings and their impact on ED decision making for the following reasons:

There is still significant equipoise associated with decisions around treatment of sub-massive PE.

Treatment with tPA has not been shown to be associated with mortality benefit in sub-massive PE.  There are two relatively recent trials that show treatment of sub-massive PE with systemic single bolus tenectaplase is associated with reduced short term hemodynamic compromise and improvement in functional outcome, but these experiences have come at a cost of increased major bleeding [13,14].  Recently, trends (though not supported by strong evidence) have favored more and more use of catheter based treatment of sub-massive PE with ultrasound facilitated delivery of tPA locally or some other combination of mechanical and continuous catheter based local therapy [15]. It is my belief that the individual response to sub-massive PE is very patient specific.  Some patients with similar clot burden can have significantly different work of breathing, hypoxemia, dizziness, and overall threat to recovery.  At the same time the patients can have significantly different risk of bleeding associated with catheter or systemic tPA.  The decision to treat an individual patient with sub-massive PE with lysis can not be protocoled or flow-charted with exactitude.

Despite all this, recognition of RV strain is important as it identifies patients who may more likely benefit from catheter or systemic clot lysis and who may be at increased risk of deterioration, sustained hypoxemia, or negative impact on functional capacity. I consider my patients with RV strain on echo or CT to be candidates for lysis until proven otherwise by patient wishes, discussion with their primary doctors, contra-indication, or bleed risk.  

We are lacking prediction tools for short term outcomes.  The blog post summary notes “short-term outcomes” as 30 and 60 days but what is really needed is 24-72 hour outcomes.  It is unlikely RV/LV ratio alone will answer this question, but what ED physicians and intensivists really want to know is “does this patient need to go to the ICU?”  Unfortunately, imaging will not alone answer this and existing literature is devoid of true “short term” prediction tools for PE. One thing that can be certain is that the abnormal troponin in a patient with RV dilation on echo or CT should not be assumed to be just a “troponin leak.”  They are at increased risk of adverse outcome relative to a patient with a normal troponin.

There is one article that is missing from this analysis but might be of use to someone without echo, or someone with radiologists uninterested in reporting RV/LV ratio and that is an article by Kline that reported similar sensitivity between a panel of bedside tests (troponin T, ECG , oxygen saturation >= 95)  and echo in identifying patients at risk for both short term and 6 month important clinical adverse events [16]. Though becoming irrelevant in the modern hyper-imaged ED environment, this paper still is important in framing what adverse outcomes patients and physicians likely care about, and to what degree echo and various other testing approaches predict these outcomes.  


I think the general message here is:  

  1. Be able to identify sub-massive PE by RV strain on CT or echo.
  2. A simple approach is to look for an RV that is bigger than the LV (or as mentioned a ratio >0.9 has been used as a cutoff).
  3. Realize that CT may be able to provide similar data as echo but there may be some issues with inter-observer agreement.
  4. We still have a lot of work to do with respect to prediction of short-term outcomes and integrating this understanding into treatment decision-making, which is inherently very patient specific.


D. Mark Courtney MD MSCI
Director of Research; Associate Professor; Department of Emergency Medicine Northwestern University, Feinberg School of Medicine [Pubmed]


  1. Jaff MR, Mcmurtry MS, Archer SL, et al. Management of massive and submassive pulmonary embolism, iliofemoral deep vein thrombosis, and chronic thromboembolic pulmonary hypertension: a scientific statement from the American Heart Association. Circulation. 2011;123(16):1788-830.
  2. Kang DK, Thilo C, Schoepf UJ, et al. CT signs of right ventricular dysfunction: prognostic role in acute pulmonary embolism. JACC Cardiovasc Imaging. 2011;4(8):841-9.
  3. "Right Heart Strain | Radiology Reference Article | Radiopaedia.org." Radiopaedia Blog RSS. Web. 17 Dec. 2015.
  4. Wake N, Kumamaru KK, George E, et al. Computed tomography and echocardiography in patients with acute pulmonary embolism: part 1: correlation of findings of right ventricular enlargement. J Thorac Imaging. 2014;29(1):W1-6.
  5. Seon HJ, Kim KH, Lee WS, et al. Usefulness of computed tomographic pulmonary angiography in the risk stratification of acute pulmonary thromboembolism. Comparison with cardiac biomarkers. Circ J. 2011;75(2):428-36.
  6. Jiménez D, Díaz G, Molina J, et al. Troponin I and risk stratification of patients with acute nonmassive pulmonary embolism. Eur Respir J. 2008;31(4):847-53.
  7. George E, Kumamaru KK, Ghosh N, et al. Computed tomography and echocardiography in patients with acute pulmonary embolism: part 2: prognostic value. J Thorac Imaging. 2014;29(1):W7-12.
  8. Javadrashid R, Mozayan M, Tarzamni MK, Ghaffari MR, Fouladi DF. Spiral computed tomographic pulmonary angiography in patients with acute pulmonary emboli and no pre-existing comorbidity: a prospective prognostic panel study. Eur Radiol. 2015;25(1):147-54.
  9. Morris MF, Gardner BA, Gotway MB, Thomsen KM, Harmsen WS, Araoz PA. CT findings and long-term mortality after pulmonary embolism. AJR Am J Roentgenol. 2012;198(6):1346-52.
  10. Araoz PA, Gotway MB, Harrington JR, Harmsen WS, Mandrekar JN. Pulmonary embolism: prognostic CT findings. Radiology. 2007;242(3):889-97.
  11. Kang DK, Ramos-duran L, Schoepf UJ, et al. Reproducibility of CT signs of right ventricular dysfunction in acute pulmonary embolism. AJR Am J Roentgenol. 2010;194(6):1500-6.
  12. Pollack CV et al. Clinical characteristics, management, and outcomes of patients diagnosed with acute pulmonary embolism in the emergency department: initial report of EMPEROR (Multicenter Emergency Medicine Pulmonary Embolism in the Real World Registry). J Am Coll Cardiol. 2011 Feb 8;57(6):700-6. 

  13. Meyer G et al. Fibrinolysis for patients with intermediate-risk pulmonary embolism. N Engl J Med. 2014 Apr 10;370(15):1402-11. 

  14. Kline JA et al. Treatment of submassive pulmonary embolism with tenecteplase or placebo: cardiopulmonary outcomes at 3 months: multicenter double-blind, placebo-controlled randomized trial. J Thromb Haemost. 2014 Apr;12(4):459-68. 

  15. Kou et al. Catheter-directed therapy for the treatment of massive pulmonary embolism: systematic review and meta-analysis of modern techniques. J Vasc Interv Radiol. 2009 Nov;20(11):1431-40.

  16. Kline et al. Surrogate markers for adverse outcomes in normotensive patients with pulmonary embolism. Crit Care Med. 2006 Nov;34(11):2773-80.

  17. Dawn S. Hui and P. Michael McFadden (2013). Contemporary Surgical Management of Acute Massive Pulmonary Embolism, Principles and Practice of Cardiothoracic Surgery, Dr. Michael Firstenberg (Ed.), InTech, DOI: 10.5772/53969. Available from: http://www.intechopen.com/books/principles-and-practice-of-cardiothoracic-surgery/contemporary-surgical-management-of-acute-massive-pulmonary-embolism