| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institut Gustave-Roussy 94805 Villejuif Cedex, France
My colleagues and I read with great interest the article by Chinn et al. [1] in the March issue of the AJR. We agree with the statement that radiofrequency tissue destruction is highly modified by tissue vascularization and that portal and arterial occlusion or a combination of both may therefore alter the size or even the shape of areas of radiofrequency-induced tissue destruction.
In a recently published study [2], we pointed out the weakness of the four-prong radiofrequency needle (RITA Medical Systems, Mountainview, CA), powered by the 50-W generator, which can sometimes create discontinuous zones of necrosis when used in close proximity to vessels. Using the same system, Chinn et al. [1] encountered a high number of discontinuous lesions, radiofrequency-induced necrosis, and numerous cruciform-shaped lesions that we have never encountered in our experiments. We believe that this discrepancy may be due to the short radiofrequency delivery time applied in the study by Chinn et al. We would like to underline the fact that the experiment described by Chinn et al. used the radiofrequency system outside the time range usually recommended for complete radiofrequency ablation with this system. Radiofrequency current was delivered for 7 min at a targeted temperature of 100°C, instead of the 15 min applied in most experimental and clinical studies. Such a modification may modify the shape and size of the radiofrequency-induced lesion considerably.
The principle behind the use of expandable multiple-tine arrays (Fig. 1A) is to create small elementary radiofrequency lesions with each tine (Fig. 1B). Then, as these elementary lesions increase in size (Fig. 1C), they coalesce, forming a larger lesion that is almost round (Fig. 1D). If the tines are adequately positioned in relation to each other, the large radiofrequency-induced lesion that results will not harbor any viable tissue. If the radiofrequency current is disconnected too early, coalescence either will not occur, resulting in a discontinuous radiofrequency lesion (Fig. 1B), or will be incomplete, producing a cruciform-shaped lesion (Fig. 1C). This phenomenon is even more pronounced when using the RITA four-tine array, because it has a large space between the tines—to such an extent that the radiofrequency multiple-tine arrays now released have a shorter distance between the tines. Therefore, the short radiofrequency delivery time applied by Chinn et al. [1] may account for the high number of discontinuous and cruciform radiofrequency lesions they reported when vascular occlusion was not used. Because the system was not used adequately, it is difficult to reach definitive conclusions concerning the shape of radiofrequency-induced lesions.
|
|
|
|
The growth rate and coalescence of elementary lesions are accelerated during vascular occlusion because of lower heat sink in the tissue. In the Chinn et al. study [1], the modification of liver vascularization probably makes up for inadequate time of radiofrequency delivery. Thus, the relative influence of vascular occlusion is probably artificially increased in the conclusion of their study.
We believe that vascular occlusion is an interesting way to increase the size of radiofrequency-induced lesions, and we already use it with success in clinical practice [3], but the study by Chinn et al. probably magnifies the value of vascular occlusion in this field by using the radiofrequency system inadequately in baseline conditions (without vascular occlusion).
References
University of Wisconsin Madison, WI 53792
We thank Dr. de Baere for his letter concerning our article [1] and for his contributions to the radiofrequency literature. The main point addressed in his letter is that the time used for ablation was inadequate. We are somewhat surprised with Dr. deBaere's contention that 15 min of radiofrequency ablation using this particular generator is applied in "most" of the experimental and clinical studies found in the literature. In fact, a paucity of studies featuring this electrode is found in the medical literature, but those studies that are available generally describe shorter times for application of radiof-requency energy, similar to what was used in our study [2,3,4,5,6].
As radiofrequency energy is applied to a probe, the lesion size will
increase, but only up to a point. The ability of radiofrequency energy to heat
tissue is governed by the bio-heat equation:
 |
is density (kg/m3), c is specific heat
(J/kg·K), and k is thermal conductivity (W/m·K),
J is current density (A/m2), and E is electric
field intensity (V/m). Tbl is the temperature of blood,
bl is the blood density (kg/m3),
cbl is the specific heat of blood (J/kg·K),
wbl is blood perfusion (1/s), and hbl
is the convective heat transfer coefficient accounting for blood perfusion
[7]. If temperature is measured at increasing distances from a radiofrequency prong, at some time point, a steady state is reached at which power deposition equals power loss due to thermal conduction. The farther apart the prongs are spaced, the more power needs to be deposited to create lethal temperatures in the gap between prongs. In our opinion, the main reason for the nonideal shape of the radiofrequency lesions in this study was the use of an underpowered generator and widely spaced prongs that could not over-come the heat-sink effect of tissue perfusion, regardless of the duration of application. The manufacturers of radiofrequency equipment clearly agree with this premise. As stated in our article, "The 50-W generator and four-prong electrode used for this study...have already been replaced with seven- and nine-prong devices and 150-W generators...." If increased time of application with a 50-W generator resulted in a large, round lesion, this would surely be a less costly solution than increasing the strength of the generator. At present, all three major manufacturers of radiofrequency equipment in the United States have generators capable of producing 150-200 W. The increase in radiofrequency power was in response to the small irregular lesions, created with less powerful devices, that led to a high local recurrence rate and the need for multiple, overlapping ablations, even when treating small tumors. This problem was exacerbated when attempting to ablate lesions near major blood vessels.
Images of radiofrequency ablation in non-perfused phantoms, such as the egg-white experiment included as part of the preceding letter, are simply not relevant to radiofrequency of tissue in vivo. Indeed, Dr. de Baere has made our point for us, albeit in a more gastronomically pleasing fashion than shown in our manuscript. The egg-white ablation is similar to the devascularized liver that results from a Pringle maneuver. Performing radiofrequency ablation in egg white is much more likely to result in a large, uniform lesion than that found in a highly vascular liver in which blood vessels abut the spaces between prongs.
In summary, the bioheat equation describes a balance between power deposition and cooling that is largely related to tissue perfusion. Simply applying energy for a longer time once a steady state is reached is unlikely to create a substantially larger and more regular lesion than would otherwise form. In our study, the 50-W generator was simply not strong enough to overcome the cooling of perfused tissue in the gaps between prongs. The fact that vascular occlusion could help create a large round lesion, even with this inadequate generator, is evidence of the efficacy of radiofrequency combined with vascular occlusion.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
