| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
THORACIC VASCULATURE |
Department of Radiology, Vancouver General Hospital, University of British Columbia, Vancouver, British Columbia, Canada
Department of Radiology, University of New Mexico Health Science Center, Albuquerque, New Mexico
 |
Commentary |
|---|
 Top Commentary ReferencesResponseReferences |
|---|
In patients with suspected PE, it has been shown that the addition of CT venography to CT pulmonary angiography protocols enables the identification of more venous thromboembolic disease (2,4,6). In the current study, 10% (100 of 957) of patients with suspected PE had a DVT. In 6% (58 of 957), there was associated PE, but in 4% (42 of 100), the DVT was the only thrombus detected. This finding is in agreement with other studies (2,4,6) that have shown a 3%–5% frequency of isolated DVT in patients with suspected PE. When compared with use of CT pulmonary angiography alone, the addition of CT venography has been shown to increase the number of patients treated with anticoagulation by approximately 26% (7). Compared with US, CT venography also offers improved assessment of pelvic, deep femoral, and calf veins. However, it is noted that isolated thrombus in these sites is rare.
The addition of CT venography to a CT pulmonary angiographic examination doubles the effective radiation dose from approximately 5 mSv to 10 mSv (5). Within the pelvis, CT venography increases the ovarian and testicular effective radiation dose by 470 and 2,200 times, respectively, over CT pulmonary angiography (5). Gapped sections, increased pitch, and a decrease in the tube current can mitigate but not eliminate the increased radiation dose. The impact of these dose reduction maneuvers on the diagnostic accuracy of the examination has not been fully evaluated. It is estimated that the additional 5-mSv radiation dose will, on average, cause an increase in cancer mortality of 250 per million exposed (8). This additional cancer incidence must be compared with the natural incidence of fatal cancer, which is approximately 230,000 per million (9). Although the estimated 0.1% increase in lifetime cancer mortality risk associated with this radiation exposure is small, it must be compared with a leg US examination, which has no known associated risk. In addition, there is a pronounced age effect on the cancer risk from ionizing radiation (9). Patients younger than 30 years of age are up to an order of magnitude more sensitive than those older than 40 years to the risk of cancer induction from an equivalent dose of ionizing radiation. Therefore, the radiation risk to children and young adults is underestimated in the above calculation.
In the clinical management of suspected PE, it is expected that CT venography will replace leg US, resulting in improved diagnostic accuracy for pelvic and lower calf thrombus but the added risk of increased radiation exposure. This result raises several questions. Is it necessary to evaluate the lower extremity and pelvic veins in all patients who undergo CT pulmonary angiography or only in those with risk factors or symptoms? Is there a significant incremental increase in diagnostic accuracy when CT venography is substituted for leg US? Should CT venography be limited to patients over the age of 40 years for whom the radiation risk is lower? The clinical probability component of this issue warrants special comment. In our institution in 1998, 25% of patients with clinical suspicion of PE who underwent CT pulmonary angiography harbored PE. In 2001, this rate had dropped to less than 10%. At the recent Society of Thoracic Radiology meeting (San Francisco, March 24–28, 2002), this finding appeared to be a widespread phenomenon. Clearly, we are performing CT pulmonary angiography with less clinical suspicion of PE. The radiation dose implications of routinely adding CT venography to the current caseload, with a low outcome of PE, needs to be carefully considered. Finally, although the authors of the current study suggest that CT venography should be added to the CT pulmonary angiographic study at no additional charge, we doubt that this economic model will be widely accepted. The provision of this service and the acceptance of the medical litigation liability for both venous thrombosis and other detectable abnormalities justify an acceptable level of remuneration.
Published outcome data from large series that contain as many as 1,000 subjects indicate a 1%–3% frequency of recurrent PE when a negative result of CT pulmonary angiography is obtained with current clinical management (10–12). We note that these trials have excluded those patients who had positive results from leg US studies, and the extent to which routine US screening was performed is not clear. This information makes it difficult to directly apply these outcome results to a diagnostic algorithm for CTVPA. However, the current management has a low failure rate. Anecdotal evidence from one of these outcome studies (11) shows that patients can have no US evidence of DVT and a negative result of CT pulmonary angiography but still suffer a fatal recurrent PE. It is not known whether the addition of CT venography to CT pulmonary angiography with US screening, or the substitution of CT venography for US screening, will reduce this risk.
The current published CT venographic data are from observational and accuracy studies, which is appropriate given the recent description of the examination. These studies indicate that, compared with leg US, CT venography is a sensitive and specific study. However, to our knowledge, there are presently no clinical outcome studies that assess the impact of the addition of CT venography to the current clinical algorithm. Therefore, it is not known if the addition of CT venography to CT pulmonary angiography leads to a significant improvement in patient outcome over current management (ie, decrease in the frequency of recurrent PE) to justify the radiation dose and additional expense. The predictive value of a negative CT venographic study is also not known. Previous articles have indicated that leg US is a low-yield examination in the absence of risk factors or symptoms of DVT (13,14). There are no data relating the findings at CT venography to clinical risk factors or symptoms. Therefore, we do not know whether everyone or only a subset of patients should undergo the test (15,16).
To address these problems, clinical outcome trials are required. These trials will probably need large numbers of patients to show significant differences, as the frequency of recurrent PE with current management is low. Some authors have suggested that CTVPA would be useful to resolve suboptimal or inconclusive CT pulmonary angiographic studies. Determination of the value of CT venography in this situation will be even more difficult, given the low likelihood of the co-occurrence of an indeterminate CT pulmonary angiographic study and a positive CT venographic one.
In summary, CTVPA is a new, attractive examination that offers one-stop shopping for the diagnosis of venous thromboembolism. It may prove to be the fastest and most efficient strategy to diagnose venous thromboembolism. However, because of the increased radiation dose and cost, further studies are needed before it can be recommended as a routine procedure in patients who undergo CT angiography for the evaluation of suspected acute PE. These studies are required to establish the role of CTVPA in the complex algorithm of clinical, laboratory, and imaging examinations that currently exist for the diagnosis of venous thromboembolism. In these trials, outcome data will be essential to ensure that the radiation exposure and expense associated with this technique are warranted. The excellent literature review and large amount of new data presented by Katz et al substantially advances this evaluation process.
|
References |
|---|
|
TopCommentaryReferencesResponseReferences |
|---|
Department of Radiology, Roswell Park Cancer Institute, Buffalo, New York
Department of Radiology, Winthrop-University Hospital, Mineola, NY
|
Response |
|---|
|
TopCommentaryReferencesResponseReferences |
|---|
Considering the issue of radiation dose, the data presented in the commentary by Drs Mayo and Ketai are somewhat misleading. In our previous study of 650 consecutively seen patients who underwent CTVPA (2), the mean patient age was 63 years and 585 patients (90%) were over the age of 40 years. Less than 4% were younger than 30 years of age. In addition, recent literature on the age distribution of patients with venous thromboembolism shows that the condition occurs predominantly in middle-aged and especially elderly patients. Stein et al (4) demonstrated in one study of hospitalized patients that the incidence of PE increased in a direct linear relationship to patient age. Heit et al (5), in a Mayo Clinic review of 10 years of data on patients from Olmsted County with venous thromboembolism, concluded that the disorder is a "major national health problem, especially among elderly hospitalized patients," and that the "incidence rose markedly with increasing age" for both outpatients and hospitalized patients.
When one considers the usual ages of patients with suspected venous thromboembolism, the frequency of serious comorbid conditions in these patients, and the long latency period of radiation-related malignancies, the risk of inducing a clinically significant neoplasm appears to be very small. Also, the radiation dose for CT venography as quoted by Drs Mayo and Ketai relates to an analysis of a study performed with contiguous images with 8-mm collimation (6). We recommend that CT venographic images be acquired every 4–5 cm and have shown that this practice allows for accurate evaluation of lower extremity DVT compared with sonography (2). Our CT venographic protocol typically includes only seven to nine abdominal and pelvic images as opposed to 50–60 images produced in a contiguous scan with 8-mm collimation. We also believe that the ability to image the iliac veins and inferior vena cava is an important advantage of CT venography over sonography. In the aforementioned study of CTVPA (2), 11 of 89 (12%) patients with DVT had thrombus involving the abdominal and pelvic veins and four (4%) had thrombus seen only in the iliac veins or vena cava. Of course, individual institutions should consider whether it is prudent to withhold abdominal and pelvic CT venographic imaging, or CT venography entirely, in younger adult patients. This small minority could undergo venous sonography if the results from CT pulmonary angiography are negative.
Neither the breast nor the thyroid are included in the radiation field of CT venography, although both are obviously irradiated during CT pulmonary angiography. There is no hesitation by Drs Mayo and Ketai to endorse CT pulmonary angiography because of radiation safety issues. Furthermore, the radiation dose concerns raised regarding CT venography also appear somewhat misguided when one considers the much bigger problem of the unregulated proliferation of self-referred, whole-body CT screening studies that are increasingly being performed on a younger population than would typically undergo CTVPA (7,8).
Drs Mayo and Ketai also question the role that CT venography should play in the algorithm for PE. We believe it should replace sonography in most patients with suspected PE who do not specifically have signs and symptoms of DVT (although an argument could be made for also initially performing CTVPA in those latter patients as well). Previous algorithms for PE have suggested that sonography be performed after negative or equivocal results from CT pulmonary angiography (9). One important aspect of CTVPA is the ability to avoid a delay in diagnosis caused by waiting for sonography results. Perhaps in select institutions where sonography can be performed rapidly after CT this advantage is less important, but in the "real world" it frequently takes hours to have sonography performed and interpreted, especially after hours, and sonography may be less readily available than CT. At one of our institutions, for example, where radiology residents are taking increasingly busy evening and weekend call, the addition of CT venography to CT pulmonary angiography adds virtually no additional time to the performance and interpretation of the CT study from the resident’s perspective. In contrast, the time for a radiology resident to perform and interpret a leg sonogram is not trivial, especially during those hours when a sonographic technologist is not available to perform the initial examination.
We agree that the successful implementation of CT for suspected venous thromboembolism has paradoxically led to more studies being ordered and therefore to a lower yield of positive studies. It is our expectation, however, that the use of new, fast, and inexpensive D-dimer blood tests as an initial screening examination may significantly reduce the use of CTVPA owing to the very high negative-predictive value of a negative D-dimer test for venous thromboembolism (10,11).
Regarding reimbursement for CT venography, we understand the concerns of Drs Mayo and Ketai. Since the acceptance of our manuscript, one of our institutions has begun charging for a limited CT study to cover the CT venographic portion of the CTVPA examination, in addition to the charge for CT pulmonary angiography. In the United States, the Medicare reimbursement for such a limited CT study is $184.97 (as per the 2001 Medicare Part B Physicians’ fee schedule). We did not want to set an economic precedent in our initial manuscript, but we do believe that this additional limited CT charge is acceptable and addresses the concerns of Drs Mayo and Ketai. Additionally, this limited charge is actually slightly cheaper than bilateral venous Doppler sonography, which Medicare reimburses at $198.62 per examination.
Regarding the lack of outcomes studies for CTVPA, we do agree with Drs Mayo and Ketai that it would require large populations of patients and that specifically demonstrating an advantage of CT venography in patients with equivocal results from CT pulmonary angiographic studies would be difficult. Although we agree that additional large outcome studies may allow further refinement of CTVPA technique and indications, we believe that the time for CTVPA has come. CTVPA is a fairly new procedure, and there has not been the opportunity for large prospective multicenter outcomes studies; however, few of the routine CT procedures now performed frequently around the world have been justified to date by large prospective outcomes or cost-effective studies.
In conclusion, the CT venographic portion of CTVPA is as accurate as sonography and provides the treating physician with more information in a shorter period of time. When radiation exposure is minimized with the recommended 4- to 5-cm image intervals and the combined examination is restricted to patients over 40 years of age or younger adults with serious comorbid conditions or pelvic trauma (the latter group being at significant risk for pelvic DVT, which is difficult to demonstrate sonographically), the examination is appropriate, safe, and effective. Energy directed at reducing the radiation dose to the typically selected and physically ill population who may undergo a CTVPA examination would be better directed at the mass screening of asymptomatic, self-referred persons who are increasingly undergoing total body CT.
|
References |
|---|
|
TopCommentaryReferencesResponseReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOGRAPHICS | RADIOLOGY | RSNA JOURNALS ONLINE |
