Controversies in Diagnosis of Pulmonary Embolism

Transcription

1 Original Articles Controversies in Diagnosis of Pulmonary Embolism Clinical and Applied Thrombosis/Hemostasis 17(2) ª The Author(s) 2011 Reprints and permission: sagepub.com/journalspermissions.nav DOI: / Paul D. Stein, MD 1,2, H. Dirk Sostman, MD 3, James E. Dalen, MD 4, Dale L. Bailey, PhD 5, Marika Bajc, MD, PhD 6, Samuel Z. Goldhaber, MD 7, Lawrence R. Goodman, MD 8, Alexander Gottschalk, MD 9y, Russell D. Hull, MBBS, MSc 10, Fadi Matta, MD 1, Massimo Pistolesi, MD 11, Victor F. Tapson, MD 12, John G. Weg, MD 13, Philip S. Wells, MD 14, and Pamela K. Woodard, MD 15, For the Consensus Group Abstract The approach to the diagnosis of acute pulmonary embolism (PE) is under constant revision with advances in technology, noninvasive approaches, and increasing awareness of the risks of ionizing radiation. Optimal approaches in some categories of patients are controversial. Data are insufficient for evidence-based recommendations. Therefore, this survey of investigators in the field was undertaken. Even among experts there were marked differences of opinion regarding the approach to the diagnosis of acute PE. Although CT pulmonary angiography was usually the imaging test of choice, the respondents were keenly aware of the dangers of ionizing radiation. In view of advances in scintigraphic diagnosis since the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) trial, ventilation/perfusion (V/Q) lung scans or perfusion scans alone and single photon emission computed tomography (SPECT) V/Q lung scans are often recommended. The choice depends on the patient s age, gender, and complexity of the findings on the plain chest radiograph. Keywords pulmonary embolism, venous thromboembolism, CT angiography, SPECT scintigraphy, V/Q scans, venous ultrasound Methods The Consensus Group consists of investigators who have published in the field of PE. This consensus report is unsponsored; we do not represent any organization. The opinions are those of the 33 investigators who responded to an invitation to 72 investigators to participate in the survey (Appendix). The investigators were selected from those who published on the subject of diagnosing venous thromboembolism (VTE). We did not attempt nor would it be realistic to obtain a consensus statement. Such a statement would be attenuated by compromise. Majority and minority opinions, therefore, were tabulated and presented. Questions Regarding the Diagnosis of Acute PE Question 1. Should CT venography be obtained with 64-detetor CTA in patients with suspected acute PE? Background question 1. In the PIOPED II 1, the sensitivity for detecting PE with a CT venography in combination with CTA (typically 4-detector) was higher, 90%, than with CTA alone, 83%. Computed tomographic venography and compression 1 Department of Internal Medicine and Research and Advanced Studies Program, Michigan State University College of Osteopathic Medicine, East Lansing, MI, USA 2 Dept of Research, St. Mary Mercy Hospital, Livonia, MI, USA 3 Office of the Dean, Weill Cornell Medical College and Methodist Hospital, Houston, TX, USA 4 College of Medicine, University of Arizona, Tucson, AZ, USA 5 Department of Nuclear Medicine, University of Sydney, Australia 6 Department of Clinical Physiology, Lund University Hospital, Lund, Sweden 7 Department of Medicine, Brigham and Women s Hospital, Harvard Medical School, Boston, MA, USA 8 Department of Radiology, Medical College of Wisconsin, Milwaukee, WI, USA 9 Department of Radiology, Michigan State University, East Lansing, MI, USA 10 Department of Medicine, University of Calgary, Calgary, Alberta, Canada 11 Department of Internal Medicine, University of Florence, Florence, Italy 12 Department of Medicine, Duke University, Durham, NC, USA 13 Department of Medicine, University of Michigan, Ann Arbor, MI, USA 14 Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Canada 15 Department of Radiology, Washington University, St Louis, MO, USA y Deceased Corresponding Author: Paul D. Stein, Michigan State University College of Osteopathic Medicine, Venous Thromboembolism Research Unit, Woodward Ave, Pontiac, MI 48341, USA steinp@trinity-health.org

2 Stein et al 141 ultrasound (US) showed similar results in diagnosing or excluding deep venous thrombosis (DVT). 2 The radiation dose of CT venography 3 of the pelvic and leg veins was calculated as 9.1 msv and 8.3 msv for CT. 4 The dose could be reduced to 2.0 msv using 5 mm axial discontinuous sections with 15 mm skip, automated current modulation, and scanning from the acetabulum to the knees rather than the iliac crest to the knees. 3 Neither exclusion of the pelvic veins 2 nor discontinuous imaging 3 resulted in the loss of diagnostic sensitivity. Among 1903 patients with suspected PE who were evaluated by 64-detector CT, 206 (10.8%) were shown to have PE by CTA and an additional 25 (1.3%) had VTE based on a positive CT venogram with a negative CTA. 5 A1.3% yield would seem poorly cost-effective. On the other hand, among the 231 patients shown to have VTE, 25 (10.8%) were diagnosed by CT venography alone. From this viewpoint, the proportion diagnosed by CT venography is sufficiently high to merit consideration of its use. Answers to question 1 (32 responded; Figure 1). Yes. CT venography should be obtained in combination with CTA. The risk from radiation is small compared to the benefits of additional diagnoses of VTE; n ¼ 6(19%). No. CT venography is unnecessary with CTA. The risk from radiation outweighs the benefits of additional diagnoses of VTE; n ¼ 14 (44%). Computed tomographic venography should be obtained or avoided based on the risk benefit ratio in an individual patient. For example, a patient with a high risk of lower extremity DVT or an elderly patient with low risk of radiation effect and limited cardiopulmonary reserve should both have CT venography routinely; n ¼12 (38%). Comments regarding answers to question 1. Some commented that they recommend obtaining venous ultrasound of the lower extremities instead of CT venography. Ultrasound would be indicated particularly in patients with a negative CTA and high clinical suspicion of DVT or PE. These results are consistent with data based on 2005 Medicare 5% research identifiable claims data that showed that 15% of patients who had CTAs also had CV venography, and 17% also had venous ultrasound. 6 Question 2. Is a perfusion lung scan (without a ventilation scan) an appropriate first choice for diagnostic imaging in patients with a normal or nearly normal chest radiograph after triage with D-dimer and clinical assessment? Background to question 2. The V/Q lung scan fell into disfavor after the PIOPED. 7 The majority of patients had either an intermediate probability assessment (39%) or low probability assessment (34%), both of which are indeterminate for PE. 7 Now, two decades since PIOPED was published, advances have been made in imaging equipment, improved methods of interpretation, and new radiopharmaceuticals. Improved diagnostic criteria include the revised PIOPED criteria, 8 CTV should accompany CTA (%) Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISAPED) criteria, 9 very low probability interpretation, 10 mismatched vascular defects, 11 stratification according to prior cardiopulmonary disease, 12 stratification of the number of mismatches required for diagnosis according to clinical assessment, 13 and the new European guidelines. 14,15 With such advances, and recognizing the risk of radiation with CTA, radionuclear imaging is receiving renewed interest. Among patients in PIOPED II, most of whom were outpatients, 72% had a normal or nearly normal chest radiograph. 16 Among those with a normal chest radiograph, only 11% had nondiagnostic perfusion scans 16 using modified PIOPED II criteria. There were no nondiagnostic perfusion scans among patients with a normal or nearly normal chest radiograph when interpreted by the PISAPED criteria. 16 Readers using the PIOPED II criteria categorized the chest radiographs as normal/near normal if there was no parenchymal opacity apart from a few small (<1 cm) nodules. 16 Normal/ nearly normal chest radiographs also included those with diffuse lung disease with low abundance and opacity of radiographic abnormalities, no bullae, no pleural effusion or opacity greater than costophrenic sulcus, and no extrapulmonary (e.g., cardiac or mediastinal contour) abnormality large enough to obscure most of the lungs. Oligemia was not considered abnormal. Readers using the PISAPED criteria categorized the chest radiographs as normal if none of the following were present: enlargement of the heart or hilar vessels; elevated diaphragm (unilateral or bilateral); pleural effusion (including intrafissural liquid); increased lung density (focal or diffuse); pulmonary edema; oligemia with or without pleonemia in the contralateral lung; consolidation suggestive of infarction; emphysema; or fibrothorax. 16 In patients with a normal or nearly normal chest radiographs in whom the 11% with nondiagnostic perfusion scans were excluded, sensitivity using revised PIOPED criteria was 86% and specificity was 93%. 16 Using PISAPED criteria, sensitivity was 72% and specificity was 97% with no nondiagnostic readings. 16 It may be, therefore, that with updated techniques, perfusion scintigraphy combined with a normal chest radiograph can provide 44 CTV yes CTV no Individual assessment Figure 1. Proportion of respondents to question 1 who indicated computed tomographic venography (CTV) should or should not accompany CT pulmonary angiography (CTA). 38

3 142 Clinical and Applied Thrombosis/Hemostasis 17(2) First imaging if CXR normal (%) diagnostic accuracy similar to CT angiography at a lower cost and with a lower radiation dose. 16 Answers to question 2 (33 responded; Figure 2). Yes. Perfusion lung scans should be obtained as a first imaging test in outpatients with a normal chest radiograph; n ¼ 15 (45%). No. Perfusion lung scans typically should not replace CTA as the first imaging test; n ¼ 14 (42%). Available evidence is contradictory and issue is at equipoise ; n ¼ 4(12%). Comments regarding answers to question 2. Several responded that they would opt for perfusion imaging only if the patient were pregnant or young or had a contraindication to CTA, as with chronic kidney disease. Regarding pregnant women, a perfusion scan only (preferably SPECT) with reduced dose of isotope was recommended by the European Association of Nuclear Medicine. 14 Some suggested that patients with emphysema, chronic obstructive pulmonary disease, or poorly controlled asthma may require a ventilation scan in addition to a perfusion scan even if the chest radiograph appears nearly normal. Some suggested use of the PISAPED criteria for interpretation. Data based on 2005 Medicare claims showed that among patients who had chest imaging for suspected PE, only 4% had perfusion scans without ventilation scans. 6 This database also showed that 72% of patients with suspected PE had a chest radiograph on the same day as PE symptoms or diagnosis. 6 Question 3. Should SPECT V/Q scans be substituted for planar V/Q scans in patients with suspected acute PE? Background to question 3. Single photon emission CT V/Q, compared with traditional planar V/Q imaging, offers the advantages of tomographic sections. 17 Dual- and tripleheaded gamma cameras with ultra-high-resolution collimators subsequently have been developed. New radiopharmaceuticals for ventilatory studies also have been developed, prominent 12 CTA Q scan Uncertain Figure2. Proportion of respondents to question 2 who indicated whether computed tomographic pulmonary angiography (CTA) or perfusion (Q) lung scan should be the first imaging test in patients with suspected pulmonary embolism who have a normal chest radiograph (CXR). among which is 99m Tc Technegas (Cyclomedica, Lucas Heights, Australia). It consists of ultrafine carbon particles that behave physiologically like a gas, 18 but it has not yet been approved by the US Food and Drug Administration. Nondiagnostic SPECT V/Q scans were reported in 3% by most investigators. 17 Among the advantages of SPECT is the avoidance of overlapping of small perfusion defects by normal tissue. 19 SPECT, in addition, has a higher contrast resolution than planar V/Q. 20 It can, therefore, detect abnormalities particularly at the subsegmental level and in the lung bases, where the segments are tightly packed. 20 Single photon emission CT V/Q scans, compared with planar V/Q scans, showed more and better delineated mismatched defects, improved quantification, and less interobserver variation. 21 Review showed that the sensitivity of SPECT was often higher than planar V/Q. 17 Specificity of SPECT was sometimes, but not always, higher than planar V/Q scans. 17 Answers to question 3 (28 responded). Yes. Evidence indicates that SPECT V/Q scans give fewer nondiagnostic images than planar V/Q scans with a higher sensitivity and specificity; n ¼ 11 (36%). No. The time and effort in obtaining SPECT V/Q scans are not worth the potential benefit; n ¼ 4 (14%). Evidence is contradictory and issue is at equipoise ; n ¼ 13 (46%). Comments regarding answers to question 3. There is no added time or effort to perform SPECT. In the opinion of 1 respondent, however, SPECT requires an adjustment of the interpretation criteria; PIOPED or PISAPED criteria cannot be simply transposed to SPECT. Time is required for clinicians to become familiar with SPECT and its interpretation. Reformatting techniques of the SPECT data may facilitate interpretation during this adjustment period. Question 4. Is there need for a prospective study of SPECT V/Q imaging versus planar V/Q imaging? Answers to question 4 (32 responded). Yes. n ¼ 25 (78%); no. n ¼ 7(22%). Comments regarding answers to question 4. The majority who voted no was from Europe or Australia, where SPECT is frequently used. In their experience, the clinical evidence is overwhelming, compelling that SPECT is superior to planar imaging. Some voted yes because most published investigations are retrospective. Question 5. Would SPECT perfusion lung scans alone (without ventilation imaging) be a first choice for imaging in patients with suspected acute PE and a normal or nearly normal chest radiograph? Background to question 5. Perfusion imaging alone, in comparison to V/Q scintigraphy, is more rapidly obtained, exposes the patient to less radiation, and is less expensive. We are not

4 Stein et al 143 aware of literature showing the accuracy of SPECT perfusion scans used in patients with a normal or nearly normal chest radiograph. Planar perfusion scans in patients with a normal chest radiograph showed good results. 16 Single photon emission CT V/Q scans, irrespective of the findings on the chest radiograph, also showed good results. 17 Answers to question 5 (28 responded). Yes. Although unproven, the potential advantages of SPECT perfusion scans are apparent; n ¼ 7(25%). Yes, SPECT perfusion scans are appropriate, but only if the PISAPED criteria are used; n ¼ 2(7%). No. SPECT perfusion scans in patients with a normal chest radiograph are untested; n ¼ 15 (54%). Available evidence is contradictory and issue is at equipoise ; n ¼ 4(14%). Comments regarding answers to question 5. Single photon emission CT may be difficult to interpret with perfusion imaging alone and ventilation imaging is needed. The amount of additional radiation with ventilation imaging is small, and it is particularly important in patients with pneumonia, chronic obstructive pulmonary disease, and heart failure. 22 Ventilation imaging is easy to obtain, particularly if one has access to 99m Tc Technegas. Question 6. Is there a need for a prospective study of the accuracy of SPECT perfusion scans compared with planar perfusion scans in patients with a normal chest radiograph? Answers to question 6 (31 responded). Yes. n ¼ 22 (71%); no. n ¼ 9(29%). Comments to question 6. If SPECT ventilation scans can be safely omitted, scanning time can be nearly halved. Therefore, such a study would be useful. Single photon emission CT perfusion scans could be obtained in critically ill patients, which adds to its potential importance. However, such a study may not be necessary, because it is generally agreed that SPECT is superior to planar imaging. Question 7. Should gadolinium-enhanced magnetic angiography be recommended for patients with suspected PE? Background to question 7. Magnetic resonance (MR) angiography in PIOPED III was shown to be technically inadequate in 25% of patients. 23 The proportion of technically inadequate images at various centers, however, ranged from 11% to 52%. Technically adequate MR angiography had a sensitivity of 78% and specificity of 99%. The combination of a technically adequate MR angiography/mr venography had a higher sensitivity of 92% and specificity was 96%, but 52% of patients had technically inadequate tests. 23 The PIOPED III investigators recommended that MR angiography should only be considered at centers that routinely perform it well, and for patients who have contraindications to standard tests. Finally, the development of nephrogenic systemic fibrosis has been correlated with the increasing use of gadolinium-based MR angiography contrast agents in patients with kidney disease. The risk of nephrogenic systemic fibrosis, however, is low, even in patients with chronic kidney disease. 24,25 Even so, it is recommended that in patients with moderate-to-end stage kidney disease, imaging methods other than MR angiography with a gadoliniumbased contrast agent should be selected whenever possible. 24 Answers to question 7 (32 responded). Yes. The literature is contradictory but MR angiography and MR venography are safe tests in patients with adequate renal function and are accurate; n ¼ 0(0%). No. Gadolinium-basedd MR angiography/mr venography is not a recommended imaging test for PE in view of many other alternative approaches; n ¼ 13 (41%). Gadolinium-based MR angiography/mr venography is an appropriate alternative imaging test for PE at centers with experience with it in patients with suspected PE and demonstrated capability of obtaining a high proportion of technically adequate images; n ¼ 19 (59%). Comments to question 7. Further technological refinement and development may make MR angiography the diagnostic test of the future because of no radiation, and the possibility of showing central pulmonary vessels, peripheral perfusion, and pulmonary veins. At the present time, there are major limitations in availability and cost. Question 8. Should serial noninvasive leg tests be recommended with today s state-of- the-art noninvasive imaging in patients with suspected acute PE? Background to question 8. Serial noninvasive leg tests in patients with suspected PE who had nondiagnostic V/Q scans can be used to safely exclude PE. Among patients with suspected PE who had a nondiagnostic V/Q scan, normal cardiorespiratory reserve, and negative serial impedance plethysmograms, PE at 3 months follow-up occurred in only 0.6% (4 of 627). 26 In those with suspected PE who had a low probability clinical assessment with a nondiagnostic V/Q scan and negative serial compression ultrasound studies, PE at 3 months follow-up occurred in only 0.5% (2 of 443). 27 In such patients with an intermediate probability clinical assessment, PE occurred in 0.4% (1 of 248). 27 Even a single negative compression ultrasound in patients with suspected DVT, if it included both the calf veins and thigh veins, resulted in few occurrences of PE on follow-up of untreated patients. 28 Answers to question 8 (33 responded; Figure 3). Yes. Serial noninvasive leg tests are a useful option after triage with objective clinical assessment and D-dimer as an alternative to lung imaging; n ¼ 6(18%). No. With many safe and accurate imaging options available, serial noninvasive leg tests are rarely if ever indicated; n ¼ 23 (70%). Serial noninvasive leg tests can be used safely, but only if the patient is known to be compliant and can be expected to appear for follow-up examinations; n ¼ 4(12%).

5 144 Clinical and Applied Thrombosis/Hemostasis 17(2) Recommended serial noninvasive tests (%) Comments to question 8. Serial leg tests were more valuable in the past when most patients had an indeterminate V/Q scan and CTA was not yet available. Some would consider serial venous ultrasound tests only if there is a high clinical suspicion of DVT. Question 9. What is the most appropriate diagnostic approach to a hemodynamically unstable patient with suspected PE? Background to question 9. Recommendations by the PIOPED II investigators for patients in extremis were bedside echocardiography in combination with bedside leg ultrasonography because they are rapidly obtainable bedside tests. 29 Right ventricular enlargement or poor right ventricular function, in a proper clinical setting, can be interpreted as resulting from PE. Venous ultrasound, in the appropriate clinical setting, also indicates PE. A portable perfusion scan was recommended by some. 29 Immediate transfer to an interventional catheterization laboratory was also recommended by some. Answers to question 9 (32 responded). Immediate CTA; n ¼ 10 (31%). Transfer to cardiac catheterization laboratory for digital pulmonary angiogram and possible catheter intervention; n ¼ 1 (3%). Bedside echocardiography in combination with bedside leg ultrasonography. Right ventricular enlargement or poor right ventricular function, in a proper clinical setting, can be interpreted as resulting from PE; n ¼ 6(19%). Venous ultrasound alone. In the appropriate clinical setting, a positive venous ultrasound indicates PE; n ¼ 0(0%). Portable perfusion scan alone; n ¼ 2(6%). Any of the above depending on facilities available; n ¼ 5(16%). Any of the above except CTA, depending on the facilities available; n ¼ 6(19%).Portable perfusion scan and venous ultrasound; n ¼ 2(6%). Comments to question 9. The echocardiogram in unstable patients provides invaluable information related to cardiac function and reserve according to some. Transesophageal echocardiography is often useful. If the patient can tolerate a CTA, 70 Serial leg yes Serial leg no Individual assessment Figure 3. Proportion of respondents to question 8 who recommended serial noninvasive leg tests in patients with suspected pulmonary embolism. 12 it is the easiest and most available test. Some obtain an echocardiogram, then administer thrombolytic agents and follow thrombolytic therapy with a CTA. If the patients were not in shock, they would obtain an echocardiogram followed by a CTA prior to thrombolytic therapy. Question 10. What is the role of venous ultrasound in a hemodynamically stable pregnant woman with suspected acute PE? Background to question 10. Venous ultrasound detects DVT in 13% to 15% of patients with suspected PE 30,31 and in 29% with proven PE, 31 eliminating the need for radiographic imaging in those patients. The positivity rate, in the absence of signs or symptoms is 8% and in the absence of a history of DVT is 13%. 2 There is growing evidence that the diagnostic quality of CT scanning diminishes during pregnancy, due to physiologic changes and increased noise from dense breast tissue. 32 The risk of radiation-induced cancer is increased in pregnant women because the radiation dose absorbed by the breasts increases with breast size. 33 Potentially adverse effects of gadopentetate dimeglumine on the fetus have not been determined, and it is not known to what extent it is excreted in human breast milk. 34 Some indicate that the radiation dose to the fetus from 16-slice CTA, 0.24 to 0.47 mgy at 0 months and 0.61 to 0.66 mgy at 3 months, is of the same magnitude as a V/Q scan, 0.25 to 0.36 mgy at 0 months and 0.31 to 0.32 mgy at 3 months, or a perfusion scan alone, 0.21 mgy at 0 months and 0.30 mgy at 3 months. 35 Others indicate that the absorbed dose to the fetus is less with CTA than a perfusion scan (0.01 vs 0.12 mgy). 36 These levels are below the 50 mgy threshold dose for fetal death, gross malformation, mental retardation, or induced genetic disease Childhood malignancy, however, may increase 1 in per msv. 40 The recommendations 29 of the PIOPED II investigators for pregnant patients with suspected PE were: D-dimer with clinical assessment should be obtained. If D-dimer is positive, venous ultrasound is recommended before imaging tests with ionizing radiation. Some PIOPED II investigators recommended pulmonary scintigraphy, and some recommended a CTA. The European Association for Nuclear Medicine recommended a perfusion scan only with lower dose radioisotope. 14 Answers to question 10 (33 responded; Figure 4). Venous ultrasound should be the first test; n ¼ 21 (63%). Chest imaging should be the first test (either CTA or V/Q scan); n ¼ 9(27%). Available evidence is contradictory and issue is at equipoise ; n ¼ 3(9%). Comments to question 10. There is no excuse for not performing a venous ultrasound as a first test in stable pregnant patents, according to 1 respondent. Others would obtain

6 Stein et al 145 venous ultrasound first only if there are signs or symptom or a history of DVT. Some suggest obtaining a perfusion lung scan without a ventilation scan as the first imaging test. A low dose tracer regimen for a perfusion lung scan should be considered. Others would obtain a perfusion lung scan if the venous ultrasound were negative. This would be appropriate if a radioaerosol, such as 99m Tc-diethylenetriaminepentaacetate (DTPA), is used for ventilation imaging because it is cleared by the kidneys and results in some fetal radiation. 41 However, Technegas essentially stays in the lungs and results in no fetal radiation. 41 The fetal dose with a V/Q scan using 99m Tc Technegas and CTA are equally negligible, provided the CTA is limited to the thorax and CT venography is not obtained. 41 Others believe that rapid diagnosis is crucial and radiation is a secondary issue. If a CTA is selected, imaging should be strictly limited to the thoracic cavity, and low kvp, if applicable, should be utilized. In a survey of 47 emergency department physicians, 36% preferred venous ultrasound, 32% preferred CTA, 17% preferred V/Q lung scans, and 11% preferred perfusion scans only and as the first diagnostic tests. 42 Radiologists preferred similar proportions. 42 Some pointed out that arterial enhancement of CTAs in pregnant women is diminished, presumably due to high cardiac output. 43 Suboptimal CTAs were more than 3 times higher in pregnant patients. 44 Protocols for CTAs, therefore, should be modified in pregnant women. 32 Question 11. What is the most appropriate diagnostic approach to hemodynamically stable nonpregnant young female with suspected acute PE and a normal chest radiograph? Background to question 11. Recommendations 29 of the PIOPED II investigators for women of reproductive age were: Pregnant patients (%) CTA or V/Q scan Venous US first Uncertain Figure 4. Proportion of respondents to question 10 who recommended chest imaging with either computed tomographic pulmonary angiography (CTA) or ventilation/perfusion (V/Q) lung scan or recommended venous compression ultrasound (US) as the first diagnostic test in stable pregnant patients with suspected pulmonary embolism. If D-dimer is positive, venous ultrasound was the next diagnostic test but is optional Pulmonary scintigraphy was recommended by some PIOPED II investigators as the next imaging test. A CTA with venous ultrasound was recommended as an acceptable alternative. If a CT venogram is deemed necessary, it was considered advisable to start at the acetabulum to reduce gonadal irradiation. The risk of death from undiagnosed PE far exceeds the risk of radiation-induced malignancy. The absorbed dose to the breast with CTA has been calculated as 10 to 50 mgy. 36,45,46 The absorbed dose to the breast with a perfusion lung scan has been estimatedtobe0.28mgy. 36 Several strategies, however, including dose modulation, low kvp, and breast shields have reduced the total breast dose on CTA A concern is repeated imaging and repeated exposure to radiation. 50,51 Approximately one third of patients who received a CTA for PE in the emergency department had a second CTA within 5 years 51 ; one sixth 51 had 3. According to the linear no-threshold model, cumulative exposure over a lifetime is linearly associated with an increased risk of cancer. 52 Pulmonary scintigraphy would minimize breast radiation. In PIOPED I, a V/Q scan in patients with a normal chest radiographic was diagnostic (high probability or normal/nearly normal) in 52% of patients with suspected PE. 53 More recently, a V/Q scan was shown to be diagnostic in 91% of patients with suspected PE and a normal chest radiograph. 54 Some, using revised PIOPED criteria, found good results with perfusion scans alone in patients with a normal or nearly normal chest radiograph. 16 Only 11% had nondiagnostic perfusion scans. 16 In those in whom nondiagnostic perfusion scans were excluded, sensitivity was 86% and specificity was 93%. 16 Using PISAPED criteria diagnostic accuracy was similar, but there were no nondiagnostic perfusion scans. 16 Answers to question 11 (33 responded; Figure 5). Perfusion lung scan; n ¼ 16 (48%). Venous ultrasound followed by perfusion lung scan if ultrasound is negative; n ¼ 8(24%). Venous ultrasound followed by CTA if ultrasound is negative; n ¼ 1 (3%). Computed tomography angiogram; n ¼ 6(18%). Single venous ultrasound; n ¼ 0(0%). Serial venous ultrasound if single ultrasound is negative; n ¼ 1(3%). Available evidence is contradictory and issue is at equipoise ; n ¼ 1(3%). Comments to question 11. Some suggested using V/Q scans rather than a perfusion scan alone. Some stated that perfusion lung scans (without ventilation lung scans) have not been sufficiently validated as options. Others, to the contrary, believe that the evidence in favor of perfusion scans alone is unquestionable. Some felt so strongly, that they cannot think of a situation where CTA would be preferred in a young person with a normal chest radiograph. Others believe that rapid diagnosis is more important than exposure to radiation. Therefore, CTA should be obtained. Survey of 48 emergency department physicians and 57 radiologists showed that 76% and 70%, respectively, preferred CTA as the first-line diagnostic test. 42

7 146 Clinical and Applied Thrombosis/Hemostasis 17(2) Young female, normal CXR (%) CTA 3 US followed by CTA 48 Q scan Question 12. What is the most appropriate diagnostic approach to a hemodynamically stable nonpregnant young female with suspected acute PE and an abnormal chest radiograph? Background to question 12. Perfusion scans obtained in PIOPED II using the modified PIOPED criteria, were nondiagnostic in 45% of patients with an abnormal chest radiograph. 16 After excluding those with nondiagnostic results, the sensitivity of a PE Present perfusion scan was 78%, and the specificity of PE Absent was 91%. Using the PISAPED criteria, no patients had a nondiagnostic perfusion scan with an abnormal chest radiograph. 16 The sensitivity of a PE Present perfusion scan by PISAPED criteria was 82% and the specificity of PE Absent was 97%. Answers to question 12 (32 responded). After triage with clinical assessment and D-dimer, if the chest radiograph is abnormal and further testing is indicated, obtain: V/Q lung scan; n ¼ 6(19%). Venous ultrasound followed by a V/Q lung scan if ultrasound is negative; n ¼ 6(19%). Computed tomography angiogram; n ¼ 20 (63%). Single venous ultrasound; n ¼ 0. Serial venous ultrasounds if single ultrasound is negative; n ¼ 0. Comments to question 12. The preferred method of imaging depends on the abnormality on the chest radiograph. If there were severe lung disease, CTA would be preferred. A baseline V/Q scan would be useful after the chest radiograph clears. Questions 13 to 16: Relate to the most appropriate diagnostic approach to a hemodynamically stable patients stratified according to age, gender, and whether the chest radiograph is normal or abnormal. Background to questions 13 to 16. Computed tomography angiography is being used excessively, with CTA showing PE in 10% of patients in some emergency departments US followed by Q scan 3 3 Serial US Uncertain Figure 5. Proportion of respondents to question 11 who recommended the following first tests in a young female with a normal chest radiograph (CXR): computed tomographic pulmonary angiography (CTA), venous compression ultrasound (US) followed by CTA, perfusion (Q) lung scan, USfollowedbyQscan,orserialUS. Female glandular breast dose may be 10 to 20 mgy 36,45 and as high as 190 mgy in a full figured woman. 33 This may be equivalent to 300 to 1000 chest radiographs. 33 The lifetime attributable risk 56 of cancer following a 64-detector CT coronary angiogram in a 20-year-old woman may be 1 in 143. The risk drops sharply with the patient s age. 56 The risk estimate, however, is based on extrapolated data from Japanese atomic bomb survivors to a US population with different baseline cancer rates and differences in relative biological effects between x-rays and other types of ionizing radiation. 56 Breast irradiation with V/Q scintigraphy is much lower than with CTA, approximately 0.28 to 0.9 mgy. 57 Question 13. What is the most appropriate diagnostic approach to hemodynamically stable older men or women with a normal chest radiograph? Answers to questions 13 (33 responded). All of the following assume that triage with D-dimer and clinical assessment failed to exclude PE. The risk of radiation-induced cancer is small. CTA is the most convenient and definitive approach and should be the first imaging test in older men and women with normal chest radiograph; n ¼ 19 (58%). An effort should be made to reduce radiation in all patients. Therefore, a perfusion lung scan should be the first imaging test in older men and women with a normal chest radiograph; n ¼ 6(18%). V/Q lung scan should be the first imaging test in older men and women with a normal chest radiograph; n ¼ 6(18%). Chest imaging should be obtained only if the venous ultrasound is negative; n ¼ 1(3%). Available evidence is contradictory and issue is at equipoise ; n ¼ 1(3%). Comments to question 13. High likelihood of comorbid pathology makes CTA more attractive in older patients. The choice depends on how abnormal the chest radiograph is. A perfusion scan might be higher yield if the chest radiograph is minimally abnormal. Question 14. What is the most appropriate diagnostic approach to hemodynamically stable older men or women with an abnormal chest radiograph? Answers to questions 14 (32 responded). Computed tomography pulmonary angiography should be the first imaging test in older men and women with an abnormal chest radiograph; n ¼ 26 (81%). A V/Q lung scan should be the first imaging test in older men and women with an abnormal chest radiograph; n ¼ 3 (9%). Chest imaging should be obtained only if the venous ultrasound is negative; n ¼ 2 (6%). Available evidence is contradictory and issue is at equipoise ; n ¼ 1(3%). Comments to question 14. The risk of radiation-induced malignancy in older patients, particularly older men is small.

8 Stein et al 147 Computed tomography angiography may be useful for evaluation of the lung abnormality shown on the chest radiograph. Even so, many patients with PE who have an abnormal chest radiograph will have a V/Q mismatch elsewhere, so the V/Q scan has potential value. Question 15. What is the most appropriate diagnostic approach to hemodynamically stable young men with a normal chest radiograph? Answers to questions 15 (32 responded). Even though the effect of radiation on male reproduction is uncertain, CTA should be the first imaging test in young men with a normal chest radiograph; n ¼ 9(28%). A perfusion lung scan should be the first imaging test in a young man with a normal chest radiograph; n ¼ 11 (34%). A V/Q lung scan should be the first imaging test in young men with a normal chest radiograph; n ¼ 7(22%). Chest imaging should be obtained only if the venous ultrasound is negative; n ¼ 3(9%). Available evidence is contradictory and issue is at equipoise ; n ¼ 2(6%) Comments to question 15. In young men with a normal chest radiograph, opinions differed. Question 16. What is the most appropriate diagnostic approach to hemodynamically stable young men with an abnormal chest radiograph? Answers to questions 16 (31 responded). Although the effect on male reproduction is uncertain, CTA should be the first imaging test in young men with an abnormal chest radiograph; n ¼ 21 (68%). A V/Q lung scan should be the first imaging test in young men with an abnormal hest radiograph; n ¼ 5(16%). Chest imaging should be obtained only if the venous ultrasound is negative; n ¼ 5(16%). Comments to question 16. Whole body and chest radiation in men are of less concern than breast radiation in women. Computed tomography angiography is warranted because V/Q s scans are often nondiagnostic. Conclusion Even among experts there were marked differences of opinion regarding the optimal approach to the diagnosis of acute PE. There was no statistically significant difference in responses of radiologists or nuclear medicine physicians compared with nonradiologists. Although CTA was usually the imaging test of choice, the respondents were keenly aware of the dangers of ionizing radiation. In view of advances in scintigraphic diagnosis since PIOPED I, V/Q scans or perfusion scans alone and SPECT are often recommended. The choice depends on the patient s age, gender, and complexity of the findings on the plain chest radiograph. Appendix Nuclear Medicine/Nuclear Radiology Dale L. Bailey, University of Sydney, Australia Marika Bajc, University of Lund, Lund, Sweden R. Edward Coleman, Duke University Medical Center, NC, USA Leonard M. Freeman, Montefiore Medical Center, NY, USA Kirk A. Frey, University of Michigan, Ann Arbor, MI, USA Alexander Gottschalk, Michigan State University, East Lansing, MI, USA Radiology Lawrence R. Goodman, Medical College of Wisconsin, Milwaukee, USA David P. Naidich, New York University, NY, USA H. Dirk Sostman, Weill Cornell Medical College and Methodist Hospital, Houston, TX, USA Pamela K. Woodard, Washington University School of Medicine, St Louis, MO, USA Cardiology/Critical care/hematology/ Pulmonary/Internal medicine James E. Dalen, University of Arizona, AZ, USA James D. Douketis, McMaster University, Hamilton, Ontario, Canada. C. Gregory Elliott, University of Utah School of Medicine, UT, USA Annette Geibel, University of Freiburg, Freiburg, Federal Republic of Germany Samuel Z. Goldhaber, Harvard Medical School, Boston, MAs, USA Gregoire Le Gal, Brest University Hospital, Brest, France Charles A. Hales, Massachusetts General Hospital, Boston, MA, USA Benjamin Harris, Royal North Shore Hospital, St Leonards, Sydney, Australia Menno V. Huisman, Leiden University Medical Center, Leiden, Netherlands Russell D. Hull, University of Calgary, Calgary, Alberta, Canada Clive Kearon, McMaster University, Hamilton, Ontario, Canada Nils Kucher, University Hospital Zurich, Switzerland Kenneth V. Leeper, Emory University School of Medicine, Atlanta, GA, USA Fadi Matta, Michigan State University, East Lansing, MI, USA Massimo Miniati, University of Florence, Florence, Italy Massimo Pistolesi, University of Florence, Florence, Italy Paolo Prandoni, University of Padua, Padua, Italy Arthur A. Sasahara, Harvard Medical School, Boston, MA, USA

9 148 Clinical and Applied Thrombosis/Hemostasis 17(2) Paul D. Stein, Michigan State University, East Lansing, MI, USA Victor F. Tapson, Duke University Medical Center, NC, USA John G. Weg, University of Michigan Health System, Ann Arbor, MI, USA Philip S. Wells University of Ottawa, Ontario, Canada Surgery Thomas Wakefield, University of Michigan Medical Center, Ann Arbor, MI Declaration of Conflicting Interests The author(s) declared no potential conflicts of interests with respect to the authorship and/or publication of this article. Funding The author(s) received no financial support for the research and/or authorship of this article. References 1. Stein PD. Fowler SE, Goodman LR, et al. Multidetector computed tomography for acute pulmonary embolism. N Eng J Med. 2006; 354(22): Goodman LR, Stein PD, Matta F, et al. CT Venography and compression sonography are diagnostically equivalent: data from PIOPED II. AJR Am J Roentgenol. 2007;189(5): Goodman LR, Stein PD, Beemath A, et al. CT venography: continuous helical images versus reformatted discontinuous images using PIOPED II data. AJR Am J Roentgenol. 2007; 189(2): Begemann PG, Bonacker M, Kemper J, et al. Evaluation of the deep venous system in patients with suspected pulmonary embolism with multi-detector CT: a prospective study in comparison to Doppler sonography. J Comput Assist Tomogr. 2003;27(3): Stein PD, Matta F, Yaekoub AY, et al. CT venous phase venography with 64-slice CT angiography in the diagnosis of acute pulmonary embolism. Clin Appl Thromb Hemost. 2010;16(4): Bhargavan M, Sunshine JH, Lewis RS. Frequency of use of imaging tests in the diagnosis of pulmonary embolism: effects of physician specialty, patient characteristics, and region. AJR Am J Roentgenol. 2010;194(4): A Collaborative Study by the PIOPED Investigators. Value of the ventilation perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA. 1990;263(20): Gottschalk A, Sostman HD, Coleman RE, et al. Ventilationperfusion scintigraphy in the PIOPED study. Part II. Evaluation of the scintigraphic criteria and interpretations. J Nucl Med. 1993;34(7): Miniati M, Pistolesi M, Marini C, et al. Value of perfusion lung scan in the diagnosis of pulmonary embolism: results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISA-PED). Am J Respir Crit Care Med. 1996; 154(5): Stein PD, Gottschalk A. Review of criteria appropriate for a very low probability of pulmonary embolism on ventilation-perfusion lung scans: a position paper. Radiographics. 2000;20(1): Stein PD, Henry JW, Gottschalk A. Mismatched vascular defects: an easy alternative to mismatched segmental equivalent defects for the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest. 1993;104(5): Stein PD, Gottschalk A, Henry JW, Shivkumar K. Stratification of patients according to prior cardiopulmonary disease and probability assessment based on the number of mismatched segmental equivalent perfusion defects. Approaches to strengthen the diagnostic value of ventilation/perfusion lung scans in acute pulmonary embolism. Chest. 1993;104(5): Stein PD, Henry JW, Gottschalk A. The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest. 1993;104(5): Bajc M, Neilly JB, Miniati M, Schuemichen C, Meignan M, Jonson B. EANM Committee. EANM guidelines for ventilation/perfusion scintigraphy: part 1. Pulmonary imaging with ventilation/perfusion single photon emission tomography. Eur J Nucl Med Mol Imaging. 2009;36(8): Bajc M, Neilly JB, Miniati M, Schuemichen C, Meignan M, Jonson B. EANM guidelines for ventilation/perfusion scintigraphy: part 2. Algorithms and clinical considerations for diagnosis of pulmonary emboli with V/P(SPECT) and MDCT. Eur J Nucl Med Mol Imaging. 2009;36(9): Sostman HD, Miniati M, Gottschalk A, et al. Sensitivity and specificity of perfusion scintigraphy combined with chest radiography for acute pulmonary embolism in PIOPED II. J Nucl Med. 2008;49(11): Stein PD, Freeman LM, Sostman HD, et al. SPECT in acute pulmonary embolism. J Nucl Med. 2009;50(12): Leblanc M, Leveill ee F, Turcotte E. Prospective evaluation of the negative predictive value of V/Q SPECT using 99mTc-Technegas. Nucl Med Commun. 2007;28(8): De Geeter FW. Tomographic imaging in the diagnosis of pulmonary embolism: still, we do not know. J Nucl Med. 2005;46(12): Reid JH, Coche EE, Inoue T, et al, International Atomic Energy Agency Consultants Group. Is the lung scan alive and well? Facts and controversies in defining the role of lung scintigraphy for the diagnosis of pulmonary embolism in the era of MDCT. Eur J Nucl Med Mol Imaging. 2009;36(3): Bajc M, Olsson CG, Olsson B, Palmer J, Jonson B. Diagnostic evaluation of planar and tomographic ventilation/perfusion lung images in patients with suspected pulmonary emboli. Clin Physiol Funct Imaging. 2004;24(5): Bajc M, Olsson B, Palmer J, Jonson B. Ventilation/Perfusion SPECT for diagnostics of pulmonary embolism in clinical practice. J Intern Med. 2008;264(4): Stein PD, Chenevert TL, Fowler SE, et al. Gadolinium enhanced magnetic resonance angiography for acute pulmonary

10 Stein et al 149 embolism: results of the PIOPED III Trial. Ann Intern Med. 2010;152(7): Kanal E, Barkovich AJ, Bell C, et al. ACR guidance document for safe MR practices: Am J Roentgenol. 2007;188(6): Chrysochou C, Power A, Shurrab AE, et al. Low risk for nephrogenic systemic fibrosis in nondialysis patients who have chronic kidney disease and are investigated with gadolinium-enhanced magnetic resonance imaging. Clin J Am Soc Nephrol. 2010; 5(3): Hull RD, Raskob GE, Ginsberg JS, et al. Noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med. 1994:154(3): Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med. 2001;135(2): Johnson SA, Stevens SM, Woller SC, et al. Risk of deep vein thrombosis following a single negative whole-leg compression ultrasound: a systematic review and meta-analysis. JAMA. 2010;303(5): Stein PD, Woodard PK, Weg JG, et al. Diagnostic pathways in acute pulmonary embolism: Recommendations of the PIOPED II investigators. Am J Med. 2006;119(12): Perrier A, Roy PM, Aujesky D, et al. Diagnosing pulmonary embolism in outpatients with clinical assessment, D-dimer measurement, venous ultrasound, and helical computed tomography: a multicenter management study. Am J Med. 2004;116(5): Turkstra F, Kuijer PM, van Beek EJ, et al. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med. 1997;126(10): Schaefer-Prokop C, Prokop M. CTPA for the diagnosis of acute pulmonary embolism during pregnancy. Eur Radiol. 2008; 18(12): Milne ENC. Female breast radiation exposure. AJR Am J Roentgenol. 2006;186(6):E Magnevist. Wayne, NJ: Berlex Laboratories. Published Nov, Hurwitz LM, Yoshizumi T, Reiman RE, et al. Radiation dose to the fetus from body MDCT during early gestation. Am J Roentgenol. 2006;186(3): Cook JV, Kyriou J. Radiation from CT and perfusion scanning in pregnancy. BMJ. 2005;331(7512): Scarsbrook AF, Evans AL, Owen AR, Gleeson FV. Diagnosis of suspected venous thromboembolic disease in pregnancy. Clin Radiol. 2006;61(1): Osei EK, Faulkner K. Fetal doses from radiological examinations. Br J Radiol. 1999;72(860): Chunilal SD, Bates SM. Venous thromboembolism in pregnancy: diagnosis, management and prevention. Thromb Haemost. 2009; 101(3): Valentin J. Biological effects after prenatal irradiation (embryo and fetus): ICRP Publication 90. Ann ICRP. 2003;33(1-2): Schembri GP, Miller A, Smart RC. Radiation dosimetry and safety issues in the investigation of pulmonary embolism. Semin Nucl Med. 2010;40(6): Jha S, Ho A, Bhargavan M, Owen JB, Sunshine JH. Imaging evaluation for suspected pulmonary embolism: what do emergency physicians and radiologists say? AJR Am J Roentgenol. 2010;194(1):W38-W Andreou AK, Curtin JJ, Wilde S, Clark A. Does pregnancy affect vascular enhancement in patients undergoing CT pulmonary angiography? Eur Radiol. 2008;18(12): U-King-Im JM, Freeman SJ, Boylan T, Cheow HK. Quality of CT pulmonary angiography for suspected pulmonary embolus in pregnancy. Eur Radiol. 2008;18(12): Parker MS, Hui FK, Camacho MA. Female breast radiation exposure during CT pulmonary angiography. Am J Roentgenol. 2005;185(5): International Commission on Radiological Protection. ICRP Publication 87: managing patient dose in computed tomography, 87. Ann ICRP. 2000;30(4): Campbell J, Kalra MK, Rizzo S, et al. Scanning Beyond Anatomic Limits of the Thorax in Chest CT: findings, radiation dose, and automatic tube current modulation. AJR Am J Roentgenol. 2005;185(6): Matsuoka S, Hunsaker AR, Gill RR, et al. Vascular enhancement and image quality of MDCT pulmonary angiography in 400 cases: comparison of standard and low kilovoltage settings. AJR Am J Roentgeno. l. 2009;192(6): Yilmaz MHMD, Albayram SMD, Yasar DP, et al. Female Breast radiation exposure during thorax multidetector computed tomography and the effectiveness of bismuth breast shield to reduce breast radiation dose. J Comput Assist Tomogr. 2007;31(1): Stein EG, Haramati LB, Bellin E, et al. Radiation exposure from medical imaging in patients with chronic and recurrent conditions. J Am Coll Radiol. 2010;7(5): Kline JA, Courtney DM, Beam DM, et al. Incidence and predictors of repeated computed tomographic pulmonary angiography in emergency department patients. Ann Emerg Med. 2009;54(1): Sodickson A, Baeyens PF, Andriole KP, et al. Recurrent CT, cumulative radiation exposure, and associated radiation-induced cancer risks from CT of adults. Radiology. 2009;251(1): Stein PD, Alavi A, Gottschalk A, et al. Usefulness of non-invasive diagnostic tools for diagnosis of acute pulmonary embolism in patients with a normal chest radiograph. Am J Cardiol. 1991; 67(13): Forbes KP, Reid JH, Murchison JT. Do preliminary chest X-ray findings define the optimum role of pulmonary scintigraphy in suspected pulmonary embolism? Clin Radiol. 2001;56(5): Le Gal G, Bounameaux H. Diagnosing pulmonary embolism: running after the decreasing prevalence of cases among suspected patients. J Thromb Haemost. 2004;2(8): Einstein AJ, Henzlova MJ, Rajagopalan S. Estimating risk of cancer associated with radiation exposure from 64-slice computed tomography coronary angiography. JAMA. 2007;298(3): Radiation dose to patients from radiopharmaceuticals (addendum 2 to ICRP publication 53). Ann ICRP. 1998;28(3):1-126.