| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Johns Hopkins Hospital Baltimore, MD 21287
The study by Dr. Kreitner et al. [1] concerning contrast-enhanced three-dimensional MR angiography (MRA) and conventional digital subtraction angiography (DSA) is yet another example of a novel technique being compared with a conventional technique that has been inadequately performed. This is a recurring problem with studies comparing MRA with angiography and has been criticized previously [2].
Dr. Kreitner and his colleagues [1] compared a good-quality MRA technique with a poor-quality DSA technique. The quality of the DSA they described is unsatisfactory because DSA was performed using a non-standardized technique and frequently using nonselective injections. It has been well shown [3, 4] that a nonselective injection (e.g., aortic injection) is inadequate for many patients who have severe peripheral vascular disease of the type reported in their article. It is not surprising that Kreitner et al. found that in this group of patients the correlation between angiography and MRA was poor. Even in the 10 patients who underwent selective angiography with the catheter advanced to the common femoral artery or more distally, imaging appears to have been substandard because an inadequate volume of contrast medium was administered. The authors stated that they tried to minimize the amount of contrast material used in an effort to reduce the incidence of renal insufficiency in this high-risk population. This effort is not doing their patients any favor if it leads to an inappropriate amputation, as has been pointed out before [4].
In their discussion, Dr. Kreitner et al. [1] refer to previous data presented by Dr. Brophy et al. (presented at the Radiological Society of North America meeting, November 1997) concerning the superior delineation of popliteal vessels by selective intraarterial DSA when compared with contrast-enhanced three-dimensional MRA. The information about that patient population is, in fact, available and has been reported previously [4, 5]. More than 70% of the patients in Brophy et al.'s study had diabetes with critical (limb-threatening) ischemia, mainly manifested as rest pain, ulceration, and often gangrene. This population, therefore, represents the most severe part of the spectrum of peripheral vascular disease.
The images presented by Kreitner et al. [1] provide some insight into why and how DSA performed poorly in their study. These images show inappropriate foot positioning with noisy images, as in Figure 1, and excessive plantar flexion and poor image collimation, as in Figures 2 and 5. The apparently false-positive occlusion of the dorsalis pedis artery shown in Figure 3 is a good example of why nonselective angiography should not be performed in these patients. The dorsalis pedis artery was not shown but the vessels in the foot, which should have been visible if an adequate volume of contrast material had been injected, are also not seen. These vessels are shown on MRA and also on the post-operative angiogram, which shows a patent dorsalis pedis artery of a size that should be shown on selective DSA. In addition, Kreitner and colleagues did not use vasodilators even when no contrast material was shown in the foot.
It is sad that the optimum technique for peripheral DSA in patients with severe critical ischemia is still not practiced at many institutions. Although MRA may be superior to DSA at these institutions, Kreitner et al.'s study is not evidence that MRA is superior to the best quality selective intraarterial DSA. It would be more appropriate for these institutions to improve their angiographic technique and to evaluate new MRA techniques.
Severe peripheral ischemia of the type discussed here is a common problem, especially among patients with diabetes. It is the responsibility of angiographers to be fully aware of the need to perform the highest quality arteriographic examination. In this patient population, achieving this level of quality requires the use of DSA, selective injections of an adequate amount of contrast medium, excellent radiographic technique, correct positioning, and the appropriate use of vasodilators. Anything less puts patients at risk for unnecessary amputation [2]. This standard should be required of studies that compare DSA with new imaging techniques.
References
Johannes Gutenberg-University D-55131 Mainz, Germany
We thank Dr. Hartnell for his interest and comments regarding our recent article [1] on contrast-enhanced three-dimensional MR angiography (MRA) and conventional digital subtraction angiography (DSA). His major criticism is that he has the impression that our DSA technique was inadequately performed.
In our discussion [1], we admitted that one major limitation of our work is that we did not use a uniform DSA technique. This lack of uniformity resulted from the fact that many patients had severe bilateral arterial occlusive disease. We respectfully disagree with Dr. Hartnell about the risk of renal failure. We consider a risk of renal failure in patients with diabetes and azotemia of up to 30% as not negligible [2]. Therefore, the performance of the DSA technique was influenced by efforts to minimize the amount of contrast medium.
Dr. Hartnell regards the total volume applied in the 10 patients who underwent a selective DSA technique as inadequate. However, compared with the work of Gates and Hartnell [3], the amount of nondiluted contrast agent in our study corresponds to their recommendations.
Dr. Hartnell claims that the nature of his patient population was available to us [3, 4]. This assertion is not true because the last work [3] was published at the same time as our article and, therefore, was not available in the literature at the time of preparation of our manuscript. One additional point is, even now, not clear: namely, the frequency of long-distance occlusions of the thigh vessels in his patient population. In our opinion, this frequency might decisively influence the ability to detect patent pedal vessel segments, and we consider it the main reason why MRA was superior to DSA in our study. However, because of the limited patient population, this aspect of our study surely has to be worked out in larger studies.
Concerning patient and foot positioning, we are well aware of the occluding effect of the retinaculum extensorum on the distal anterior tibial artery in plantar flexion. It is our practice to position the foot in a way most comfortable for the patient to minimize patient movement that leads to misregistration artifacts. If one compares the DSA images with the MRA images, there is often plantar flexion during both imaging procedures. This finding could be another argument for MRA because the depiction of a patent dorsal pedal artery is apparently not dependent on patient positioning in dorsiflexion of the foot [5].
We do not agree with Dr. Hartnell about the use of vasodilators. There is no convincing evidence from the current literature that vasodilative drugs improve image quality. If one considers the mean transcutaenous oxygen pressure of 17 mm Hg in our patient population, it is justified to postulate that the pedal vessels were dilated to a maximum diameter to make up for impaired inflow.
Dr. Hartnell suggests that because of our inadequate DSA technique, there might be an unnecessarily high rate of amputation at our institution. Actually, we treat approximately 1200 patients with infrainguinal bypass grafts, including 159 pedal bypass grafts. Seventy percent of the patients have diabetes. The primary rate of major amputations is approximately 5%. These figures should make it clear that there cannot be any substantial deficits in the preoperative imaging of these patients at our institution.
We are fully aware of the criteria of high-quality angiography in patients with diabetes who have severe peripheral ischemia. We consider intraarterial DSA with selective injections and adequate foot positioning as the golden standard at our institution. However, we have to admit that sometimes the better is the enemy of the good.
We deeply regret that Dr. Hartnell has not carefully read the summary of our article [1], in which we try to tone down our results—because of our selection bias—to indicate that we consider MRA as a substantial part of the diagnostic algorithm in cases for which DSA provides inadequate visualization of the pedal arteries. We have the impression that by doing so we could further reduce our rate of primary major amputations.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
