Helical Chest CT in the diagnosis of acute Pulmonary Embolism Hrudaya Nath, MD, Mitchell Olman, MD
Acute pulmonary embolism (PTE) is a common disease and if left untreated leads to significant morbidity and mortality. It is estimated that more than 500,000 episodes of pulmonary embolism occur each year in this country, accounting for 50,000 deaths annually. Antemortem diagnosis of PTE is significantly lower and it is estimated that fewer than 30% of PTE found at autopsy were suspected during life. PTE is a treatable condition; however, its treatment, anticoagulation, carries its own significant morbidity and mortality.
The highly variable and heterogeneous clinical syndrome of PTE results in its underdiagnosis. Imaging plays a key role in its diagnosis. Pulmonary angiography is considered as the gold standard for the 'in vivo' diagnosis of PTE; however, the test is expensive, invasive and is universally underutilized, including in our own institution. Also, its interpretation in cases with minimal (involving only the subsegmental arteries) PTE is less definitive. Radionuclide V/Q scanning is traditionally considered as the primary screening method for PTE, and this is a time tested method. Based on the results of PIOPED data, high probability scans have ~ 90% concordance with angiographically positive PTE. A normal V/Q scan has negligible incidence of PTE. Low probability scans have about 10% reported incidence of angiographically positive PTE. However, this technique has several problems, including the inability to obtain ventilation scans in intubated patients (e.g., ICU patients on ventilators). The most important limitation is the approximately 35-50 % incidence of 'indeterminate' probability scans. In this large subset of patients only about a third have angiographically positive PTE, and another imaging test is necessary to identify patients with true positive PTE so that an excessive number of patients do not receive needless anticoagulation. Ultrasound examination of the veins of the lower extremity was introduced as an ancilliary test following the realization that the PTE is a marker for venous thromboembolism and these emboli to the lungs arise most commonly from the lower extremity veins. However, the incidence of a positive examination is very low in the absence of clinical symptomatology of deep venous thrombosis, a figure of about 10% in our department.
The problem with the interpretation of V/Q scanning is caused by the 'inference' of PTE based on ventilation/perfusion mismatch and not the demonstration of the intravascular clot; hence, the superiority of pulmonary angiography which actually demonstrates the clot. Helical chest CT (HCCT) following rapid intravenous injection of contrast material is a modification of the conventional chest CT in which with software and hardware modifications, a volumetric data acquisition takes place allowing faster scan acquisition and subsequent image manipulation. Several recent studies have shown that this method provides an excellent, but not a perfect demonstration of the pulmonary arteries and the presence of an intravascular thrombus (fig). The technique has shown promise in many studies published since 1996. The sensitivity of the CT in detecting PTE ranged from 85%-90% and the specificity was more than 90%. Most of the missed emboli were in the subsegmental vessels. There is some consensus that these isolated subsegmental emboli in patients with adequate cardiopulmonary reserve and absence of deep venous thrombosis may not be clinically significant, thus making it possible to use HCCT either as a screening test (instead of V/Q scanning) or to replace pulmonary angiography in selected patients. In a study published in 1997, a cohort of 140 patients had HCCT and V/Q scanning among which 46 patients had proven PTE. HCCT identified more PTE (85% vs 67%) than V/Q scan and they were equally specific (95% vs 94%). In 20 cases V/Q scan results were indeterminate, and among these HCCT correctly identified emboli in five, excluded emboli in eleven and falsely diagnosed PTE in two. In another study published in 1997, 164 consecutive patients with indeterminate V/Q scan, and negative venous US were studied by HCCT. PTE was found in 39 patients at HCCT and pulmonary angiography in one who had a negative HCCT. All of the other patients had clinical follow up for three months. Three patients subsequently developed DVT who may have had PTE during the episode of indeterminate V/Q scan; two additional patients suffered PTE during the follow up. Thus HCCT missed thromboembolism in six patients (5.4% [95% confidence interval 9.7%]).
Because of the above background, we embarked on evaluating HCCT for the detection and screening of PTE. Two separate groups of patients are being evaluated. In an ongoing, prospective, randomized study (Investigators: MA Olman, Hrudaya Nath, KM Willie, JA Ball), patients are randomized to either V/Q scanning or HCCT. If the screening test is not positive for PTE, the patient has US examination of the lower extremity veins for DVT. If that is also negative and the patient has normal cardiopulmonary reserve, a follow up US examination is performed within 10 days and a clinical follow up at three and six months. If the cardiopulmonary reserve (CPR) is inadequate, a pulmonary angiogram is performed. So far, 60 patients have been randomized and completed six month followup (the target number of patients is 200). Thus far, the frequency of positive initial evaluation is equal (12%) for both the HCCT and V/Q scan groups. In the patients with good CPR 25/44 (56%) underwent HCCT and 19/44 (43%) had V/Q scanning. There have been three deaths in the HCCT group, none of which was attributable to PTE. There have been no deaths nor recurrence of PTE or DVT in the V/Q scan group. In the patients with poor CPR (n=18), there has been 1 death, possibly attributable to PTE in the CT group and none in the radionuclide group.
Fig 1.HCCT scan at the level of left main
pulmonary artery. There are multiple intraluminal filling
defects in the left pulmonary artery (arrows), extending
into the lower lobe..
Fig 2. HCCT scan at the level of the right lower lobe
branch. The embolus in the lower lobar artery is well seen
(arrow).
In another study of patients in surgical and trauma/burn ICU (Investigators: A May, H Nath, G Vangelisi), 22 consecutive patients with suspected acute PTE were investigated with HCCT between 8/1997 and 7/1998. This group of patients are difficult to screen using V/Q scans because of the often abnormal chest radiographs and the inability to obtain ventilation scans because the patients are ventilator dependent. The HCCT was positive for PTE in four patients, one of which was corroborated with a high probability V/Q scan. It was suggestive in one patient, but a pulmonary angiogram was normal. The HCCT was negative in 17 patients. In one patient, the negative finding was proven by a pulmonary angiogram. Lower extremity US examination was performed in 8/17 patients, all of which were normal. In 11/17 patients unexpected findings were found on the HCCT, including pneumonia, large pleural effusions, ARDS and hemopericardium. These findings influenced treatment decisions in six. The average hospital stay of these 17 patients was 64.34 days, during which time they did not develop PTE or DVT.
These two studies bolster our confidence in the use of HCCT for screening and diagnosis of acute PTE. We therefore propose the following process for choosing HCCT in clinical practice, based on the patient's cardiopulmonary status and the presence of pulmonary and/or pleural abnormalities on the chest radiographs that may interfere with interpretation of V/Q scans. The randomized trial in progress at UAB is specifically designed to validate the recommendations given below, and we solicit your help in recruiting patients for this study:
1. Results of theV/Q scans (high probability or normal scans) are more likely to be definitive in patients with normal chest radiographs, no major heart or lung disease and adequate CPR. Therefore, these patients should have V/Q scan as the initial screening test. If the V/Q scan is indeterminate or low probability and the result discordant with the clinical probability of PTE, additional imaging such as US examination for DVT or pulmonary angiography may be indicated.
2. In patients with adequate cardiopulmonary function, but with abnormal chest radiographs, V/Q scanning is less likely to provide a definitive answer and this group of patients are more likely to benefit from HCCT. If the patient also has suspected DVT we suggest beginning with sonography of the veins. In the others the initial test should be HCCT; a negative scan should be followed by an US of the lower extremity veins (and a followup US if the intial test is negative). If both US examinations are negative, the patient can be clinically followed.
3. Patients with significant heart or lung disease and inadequate CPR are at a high risk of fatal outcome from a recurrent PTE. Therefore, the presence or absence of PTE must be established with the greatest possible confidence in these patients. Presently there is no consensus about the best initial screening test. The randomized trial in progress is designed to test the role of HCCT vs. V/Q scanning. Of course, if the patient has clinical evidence of DVT, US of the lower extremity should be the initial test. Perhaps, based on the published material which reports a higher probability of detecting thrombi in the pulmonary vasculature, HCCT may be a better initial screening test. However, at this point the negative predictive value of HCCT is not known. In contrast to the patients with adequate CPR, the presence of subsegemental PTE may be important in this group of patients; unfortunately, this may not be diagnosable either by V/Q scanning or HCCT. Therefore, if HCCT does not show PTE and venous US is negative, pulmonary angiography should be performed to confidently exclude PTE.
The above recommendations are also in accordance with the suggestion from Society of Thoracic Radiology as well as the European Society of Thoracic Imaging.
