Are biomarkers additive to pulmonary embolism severity index for severity assessment in normotensive patients with acute pulmonary embolism?
A. SINGANAYAGAM1, C. SCALLY2, M.Z. AL-KHAIRALLA2, L. LEITCH2, L.E. HILL3, J.D. CHALMERS1 and A.T. HILL1
From the 1Department of Respiratory Medicine, Royal Infirmary of Edinburgh, 51 Little France Crescent, Old Dalkeith Road, Edinburgh, EH16 4SA, 2Department of Respiratory Medicine, Ninewells Hospital, Dundee, DD19SY and 3Department of Radiology, St Johns Hospital, Livingston, EH54 6PP, UK
Downloaded from http://qjmed.oxfordjournals.org/ by guest on February 25, 2013
Address correspondence to A. Singanayagam, Department of Respiratory Medicine, Royal Infirmary of Edinburgh, 51 Little France Crescent, Old Dalkeith Road, Edinburgh, EH16 4SA, UK. email: aransinga@gmail.com Received 18 April 2010 and in revised form 26 August 2010
Summary
Background: Biomarkers and clinical prediction rules have been proposed for severity assessment in acute pulmonary embolism (PE). Aim: The aim of this study was to compare biomarkers with the PE Severity Index (PESI), a validated scoring system for predicting 30-day mortality and to determine if addition of biomarkers to PESI would improve its predictive accuracy. Study Design and Methods: We conducted a retrospective analysis of normotensive patients admitted with PE confirmed by CT pulmonary angiogram, to three teaching hospitals between January 2005 and July 2007. All patients had admission levels of D-dimer and Troponin I and calculation of PESI score on admission. The outcome of interest was 30-day mortality. Results: There were 411 patients included in the study. Patients who died had higher levels of D-dimer (median 2947 ng/ml vs. 1464 ng/ml; P = 0.02), Troponin (57.1% positive vs. 13.8%; P < 0.0001) and higher PESI scores [median 109 vs. 83; P < 0.0001], compared to survivors. PESI had superior accuracy for predicting 30-day mortality than a combination of Troponin and D-dimer (AUC 0.80 vs. 0.75). Addition of Troponin to PESI further improved the predictive value of the score (AUC 0.85 for vs. AUC 0.80 for PESI alone). Conclusion: Biomarkers and clinical prediction rules predict outcome in acute PE. Addition of troponin to the PESI scoring system improves the predictive value for 30-day mortality and may be useful for guiding initial management of patients presenting with PE.
Introduction
Acute pulmonary embolism (PE) is a potentially life-threatening condition and mortality is known to be higher in patients with extensive clot load and right ventricular (RV) strain.1,2 It has been shown that mortality tends to occur early in patients with PE, with up to 80% of deaths occurring within 2.5 h into the admission.3 Early identification of severity is clinically important as it allows stratification
! The Author 2010. Published by Oxford University Press on behalf of the Association of Physicians. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
126
A. Singanayagam et al. extent of clot and mortality risk in those with confirmed PE.27,28 The aim of this study was to compare biomarkers Troponin I and D-dimer with the PESI for prediction of 30-day mortality in patients admitted with acute PE and also to assess whether addition of biomarkers to PESI would improve its predictive accuracy.
of patients into those who may benefit from closer monitoring (e.g. high-dependency unit) and use of interventions such as thrombolysis and those in a lower risk category, in whom it could be safe to delay imaging and who would be suitable for a lower level of care or outpatient management. Patients with massive PE are identifiable by the presence of haemodynamic compromise and may benefit from interventions such as thrombolysis. However, the management of patients with sub-massive PE (presence of right-heart strain but haemodynamically stable) is less clear. Risk stratification can be made, to some extent, by identifying evidence of RV strain on clinical examination and simple bedside tests such as electrocardiogram (ECG).4,5 However, these methods are known to be unreliable for identification of PE severity.4 More accurate prognostic information can be obtained by use of echocardiography which can reliably identify RV dysfunction and has been shown to be predictive of patient outcome.68 However, in many healthcare systems, a 24-h echocardiography service is not available and this modality also suffers from technical difficulties of imaging the right heart, particularly when used by less experienced operators.9 CT-pulmonary angiographic (CTPA) features of RV dilation such as RV/left ventricle ratio compare favorably with echocardiography for evaluation of RV strain in acute PE.1016 However, immediate risk stratification also may not be achievable with radiological tests as there are often significant delays in patients undergoing CTPA. Clinical prediction scores such as the Geneva risk score17 and the PE severity index (PESI)1819 have recently been proposed as potential initial stratification tools for patients presenting with acute PE. Comparative studies have shown that PESI outperforms the Geneva score for prediction of adverse outcome.20 However, these scores suffer from complexity, relying on measurement of a number of variables, which limits their front door use in clinical practice, and they have not, so far, been widely adopted. As existing scoring systems may be too complicated for use in everyday practice and imaging techniques are not always rapidly available, it would be appealing to have a single blood test that could provide rapid prognostic information. A growing body of literature suggests that admission biomarkers such as cardiac troponins and brain natriuretic peptide (BNP) can be used to predict accompanying RV strain and adverse outcome in patients presenting with PE.2126 There is also emerging evidence that admission levels of D-dimer, a biomarker used in the investigation of suspected PE, correlate with
Methods
A retrospective study of consecutive patients admitted with PE in three teaching hospitals in East of Scotland between January 2005 and July 2007. Inclusion criteria were normotensive patients admitted with PE, confirmed by CT pulmonary angiogram and that had D-dimer and Troponin I measured on admission, which were standard admitting investigations for patients with suspected PE in these hospitals. Patients who did not have both Troponin I and D-dimer measured were not included in the analyses. The study was approved by the Lothian Research Ethics committee. Exclusion criteria were:
(i) cardiac arrest or cardiogenic shock without confirmed PE diagnosis; (ii) patients developing PE whilst already an in-patient for an alternative diagnosis or condition; (iii) a history of a known thrombotic disorder, thrombophilia or haematological malignancy; (iv) ongoing therapeutic anticoagulation at time of admission; (v) patients not actively treated with anticoagulation (palliative care); (vi) hypotension on admission (systolic blood pressure <90 mmHg); and (vii) requirement for thrombolytic therapy.
Biomarkers additive to PE severity index (see Appendix 1). Missing data was assumed to be normal in keeping with the methodology in previous studies19 and, since admitting doctors at the study hospitals are required to complete standardized admission pro-formas, <1% missing data was encountered in this study.
127
We used multivariable logistic regression to evaluate the association of PESI, positive Troponin I (>0.2 mg/l) and D-dimer with the outcome of 30-day mortality.
Outcomes
The outcome of interest was 30-day mortality.
Results
Four hundred and eleven patients met the criteria and were included in the study. Baseline characteristics of the study cohort and comparison with other studies in the literature is summarized in Table 1. The overall 30-day mortality rate was 5.1%. There were 16.1% of the total cohort who had positive Troponin I measurements. The distribution of patients according to PESI class is displayed in Table 1.
Downloaded from http://qjmed.oxfordjournals.org/ by guest on February 25, 2013
Statistical analysis
All data was analyzed and processed on SPSS version 13.0 for windows (SPSS Inc., Chicago, Ill). Descriptive statistics of demographic and clinical variables are expressed as median (inter-quartile range) (IQR) unless otherwise stated. The 2-test was used to compare categorical data between groups and the MannWhitney U test was used to compare continuous data. Sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV) and the area under the receiver operator characteristic curve (AUC) were used to assess predictive tests. We evaluated the following tests for the outcomes of interest: PESI score (as previously described18), positive Troponin I (>0.2 mg/l), D-dimer, a combination of positive Troponin I (>0.2 mg/l) and D-dimer and a combination of PESI with positive Troponin I and/or D-dimer. PESI and D-dimer were analyzed as continuous variables while Troponin I was analyzed as a categorical variable (positive vs. negative). A two-tailed P-value < 0.05 was considered statistically significant. KaplanMeier analysis was used for comparison of survival. The statistical significance was evaluated using the log-rank test.
128
A. Singanayagam et al.
Table 1 Baseline characteristic in the study population and comparion with previous studies Baseline characteristic Current study 411 43.1 55.4% 14.9% 15.5% 18.8% 27.4% 16.8% 3.1% 8.1% 5.1% 24.3% 27.7% 26.8% 13.9% 7.3% National studies (Aujesky et al.18) 10 354 39.6 52.8% 16.1% 18.2% 19.9% 29.2% 14.5% 6.9% 8.0% 9.2% 19.4% 21.5% 21.7% 16.4% 21.0% National studies (Jimenez et al.20) 599 46.1 67.0% 8.0% 10.5% 23.4% 19.4% 5.0% 0% 23.7% 7.2% 12.3% 23.7% 28.9% 21.5% 13.5%
Number of patients Gender (percentage of males) Age>65 years Cardiac failure Chronic lung disease Malignancy Pulse 110 per min Respiratory rate 30 per minute Confusion Arterial oxygen saturations < 90% on air 30-day mortality PESI 1 PESI 2 PESI 3 PESI 4 PESI 5
Figure 1. Kaplan Meier survival analysis according to PESI, Troponin and D dimer levels.
Multivariable analysis
On multivariable analysis, both PESI [adjusted odds ratio (AOR) 1.04 (1.021.06), P < 0.0001] and positive Troponin I [AOR 7.73 (2.9020.6); P < 0.0001] were independently associated with 30-day mortality. D-dimer was not independently associated with mortality, AOR 1.00 (0.991.00, P = 0.07).
129
values
for
PESI,
biomarker
and
AUC PESI Troponin alone D-dimer alone D-dimer + troponin PESI + troponin PESI + D-dimer PESI + troponin + D-dimer 0.80 0.72 0.66 0.75 0.85 0.80 0.82 (0.740.87) (0.610.83) (0.530.78) (0.620.84) (0.770.93) (0.750.88) (0.700.93)
P-value compares individual AUC values to the null hypothesis (AUC = 0.5)
Discussion
This study is the first to compare a combination of admission levels of D-dimer and Troponin I with the PE Severity Index for prediction of 30-day mortality in normotensive patients admitted with acute PE. Our data shows that the PESI score outperforms biomarkers for prediction of 30-day mortality, but that Troponin I adds additional predictive value to the clinical score. Although high D-dimer levels were associated with 30-day mortality, addition of D-dimer to PESI or Troponin I did not give additional prognostic information. D-dimer and Troponin I are relatively cheap, rapidly available blood tests that are routinely measured in patients presenting with acute cardiorespiratory symptoms. Our findings are in keeping with previous studies which have shown that levels of these biomarkers correlate with extent of PE and adverse outcome.2129 A recent study by Moores
et al. showed no additional value of adding Troponin measurement to PESI for selection of low-risk patients with acute PE.30 Our study assessed the value of adding Troponin to PESI for identification of patients at higher risk of 30-day mortality and showed a small improvement in predictive value (from AUC 0.80 to AUC 0.85). The value of using biomarkers alone for risk assessment is unclear and requires further independent assessment. Severity assessment is an important initial step in the management of acute PE. The difficulty of this initial assessment has led to development of various severity scores and prediction rules to assist clinical decision making. PESI is the most extensively validated score and has been shown to be a useful tool for predicting patients at high and low risk of mortality with AUC values ranging from 0.70 to 0.80 in derivation and validation cohorts.18,19,31,32 However, PESI is heavily influenced by age and comorbidities and may not be the optimum tool for predicting patients who deteriorate rapidly from PE. Simplicity is an important requirement for widespread acceptance of a prediction rule and although PESI has been shown to be a robust clinical tool, its calculation relies on 11 different variables and this complexity may make it difficult to implement into clinical practice. The complexity of PESI has led to recent guidelines recommending simpler severity scores,33 which incorporate cardiac biomarkers such as Troponin I and an assessment of right-heart strain and this further supports our finding and the increasing importance of biomarkers as severity assessment tools in acute PE. This study is limited by its retrospective design. As a result, we were unable to assess the predictive value of other biomarkers, which have also been shown previously to correlate with outcome in PE, such as BNP.2426 Our analysis is limited to Troponin I and D-dimer, as they are routine admission investigations for all patients with suspected PE at the hospitals in this study. We assessed Troponin I as this was the assay used at the study hospitals. Previous studies have shown that the prognostic value of troponins is consistent, regardless of whether Troponin I or Troponin T is used.22 Prospective studies are preferable because of a reduced risk of ascertainment bias. However, as studies of PE require a positive CTPA and we included consecutive patients undergoing CTPA with confirmed PE, no cases of radiologically confirmed PE were missed during the study period. Our study is also limited to patients presenting from the community with acute PE and does not consider patients who develop PE whilst an in-patient for an alternative condition.
130
A. Singanayagam et al.
10. Araoz PA, Gotway MB, Trowbridge RL, Bailey RA, Auerbach AD, Reddy GP, et al. Helical CT pulmonary angiography predictors of in-hospital mortality in patients with acute pulmonary embolism. J Thorac Imaging 2003; 18:20716. 11. Ghaye B, Ghuysen A, Willems V, Lambermont B, Gerard P, DOrio V, et al. Severe pulmonary embolism: pulmonary artery clot loads and cardiovascular parameters as predictors of mortality. Radiology 2006; 239:88491. 12. Schoepf UJ, Kucher N, Kipfmueller F, Quiroz R, Costello P, Goldhaber SZ. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism. Circulation 2004; 110:327680. 13. Collomb D, Paramelle PJ, Calaque O, Bosson JL, Vanzetto G, Barnoud D, et al. Severity assessment of acute pulmonary embolism: evaluation using helical CT. Eur Radiol 2003; 13:150814. 14. Contractor S, Maldjian PD, Sharma VK, Gor DM. Role of helical CT in detecting right ventricular dysfunction secondary to acute pulmonary embolism. J Comput Assist Tomogr 2002; 26:58791. 15. Lim KE, Chan CY, Chu PH, Hsu YY, Hsu WC. Right ventricular dysfunction secondary to acute massive pulmonary embolism detected by helical computed tomography pulmonary angiography. Clin Imaging 2005; 29:1621. 16. Singanayagam A, Chalmers JD, Scally C, Akram A, Al-Khairalla MZ, Leitch L, et al. Right ventricular dilation on CT pulmonary angiogram independently predicts mortality in pulmonary embolism. Respir Med 2010; 104:105762. 17. Wicki J, Perrier A, Perneger TV, Bournameaux H, Junod AF. Predicting adverse outcome in patients with acute pulmonary embolism: a risk score. Thromb Haemost 2000; 84:54852. 18. Aujesky D, Obrosky D, Stone RA, Auble TE, Perrier A, Cornuz J, et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Resp Crit Care Med 2005; 172:10416. 19. Aujesky D, Obrosky DS, Stone RA, Auble TE, Perrier A, Cornuz J, et al. A prediction rule to identify low-risk patients with pulmonary embolism. Arch Intern Med 2006; 166:16975. 20. Jimenez D, Yusen RD, Otero R, Uresandi F, Nauffal D, Laserna E, et al. Prognostic models for selecting patients with acute pulmonary embolism for initial outpatient therapy. Chest 2007; 132:2430. 21. Jimenez D, Diaz G, Molina J, Marti D, Del Rey J, GarciaRull S, et al. Troponin I and risk stratification of patients with acute nonmassive pulmonary embolism. Eur Respir J 2008; 31:84753. 22. Becattini C, Vedovati MC, Agnelli G. Prognostic value of troponins in acute pulmonary embolism. A Meta analysis. Circulation 2007; 116:42733. 23. Logeart D, Lecuyer L, Thabut G, Tabet JY, Chavelas C, Bonnin F, et al. Biomarker-based strategy for screening right ventricular dysfunction in patients with non-massive pulmonary embolism. Intensive Care Med 2007; 33:28692. 24. Sohne M, Ten Wolde M, Boomsma F, Reitsma JB, Douketis JD, Buller HR. Brain natriuretic peptide in haemodynamically stable acute pulmonary embolism. J Thromb Haemost 2006; 4:5526.
A combination of Troponin I and PESI has good predictive value for 30-day mortality in PE. Potentially, this combination may be useful in guiding initial management of patients presenting with PE but further independent validation is needed.
Acknowledgements
The authors acknowledge the team of radiologists at Ninewells Hospital, Royal Infirmary of Edinburgh and St Johns Hospital that reported the CT scans for patients included in the study.
Funding
Clinical Research Training Fellowship from the Medical Research Council (London, United Kingdom, to J.D.C.). Conflict of interest: None declared.
References
1. Ghuysen A, Ghaye B, Willems V, Lambermont B, Gerard P, Dondelinger RF, et al. Computed tomographic pulmonary angiography and prognostic significance in patients with acute pulmonary embolism. Thorax 2005; 60:95661. 2. Van de Meer RW, Pattynama PM, Van Strijen MJ, van-den Berg-Huijsmans AA, Hartmann IJ, Putter H, et al. Right ventricular dysfunction and pulmonary obstruction index at helical CT: prediction of clinical outcome during 3 month follow-up in patients with acute pulmonary embolism. Radiology 2005; 235:798803. 3. Stein PD. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108:97881. 4. Becattini C, Agnelli G. Acute pulmonary embolism: risk stratification in the emergency department. Intern Emerg Med 2007; 2:11929. 5. British Thoracic Society Standards of Care Committee Pulmonary Embolism Guideline Development Group. British Thoracic Society guidelines for the management of suspected acute pulmonary embolism. Thorax 2003; 58:47084. 6. Goldhaber SZ. Echocardiography in the management of pulmonary embolism. Ann Intern Med 2002; 136:691700. 7. Grifoni S, Olivotto I, Cecchini P, Pieralli F, Camaiti A, Santoro G, et al. Short term clinical outcome of patients with acute pulmonary embolism, normal blood pressure and echocardiographic right ventricular dysfunction. Circulation 2000; 101:281722. 8. Viellard-Baron A, Page B, Augarde R, Prin S, Qanadli S, Beauchet A, et al. Acute cor pulmonale in massive pulmonary embolism: incidence, Echocardiographic pattern, clinical implications and recovery rate. Intensive Care Med 2001; 27:14816. 9. Goldhaber SZ. Cardiac biomarkers in pulmonary embolism. Chest 2003; 123:17824.
131
Appendix 1
PESI
The most extensively validated existing severity score in PE is the PESI, which was developed following a study of 15 531 patients,18 It is based on 11 demographic, co-morbid and clinical variables (Table 4) and stratifies patients into five risk classes, three with low risk of 30-day mortality (Class I = 1.1%. Class II = 3.1%; Class III = 6.5%), a fourth with increased risk (10.4%) and a fifth with high risk (24.5%). PESI has now been validated in large independent populations18,19 and is proposed as an admission severity tool for use in clinical practice.18,19
Table 4 Variable Age Male sex Cancer Heart failure Chronic lung disease Heart rate > 110/min Systolic blood pressure < 100 mmHg Respiratory rate 30/min Body temperature < 368C Disorientation, lethargy, stupor, coma Arterial oxygen saturations < 90% Variables used to calculate PE severity index
Downloaded from http://qjmed.oxfordjournals.org/ by guest on February 25, 2013
Risk categories: class I, <65 points; class II, 6685 points; class III, 86105 points; class IV, 106125 points; class V, >125 points.