| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Hanyang University Hospital 133-792 Seoul, Korea
I read with interest the article written by Dr. Cha et al. [1]. Although these authors' trial to find a more reliable modality for guidance during radiofrequency ablation is one of the key issues for a successful ablation, I have to address some limitations in study design of their investigation that may draw to an impetuous conclusion.
First, any investigation, especially in an animal model, should be designed after careful consideration of a clinical situation. In my experience performing radiofrequency ablation using sonographic guidance, monitoring during ablation seems to be not so critical for the success of the ablation. After the electrode has been placed accurately in the estimated location, any manipulation of the electrode, including repositioning, is unnecessary during the ablation even if there are no artifacts due to microbubble and the radiofrequency current itself. Instead, scanning immediately after ablation (just after the radiofrequency generator is turned off) is essential for estimation of the ablated area and planning the subsequent ablation because all the ablated lesions show maximal hyperechogenicity immediately after the ablation [2, 3]. In this regard, the crossover technique for CT versus sonography immediately after the ablation in their study has an intrinsic limitation for simulating clinical situation practically. The poorly defined hypoechoic zone that is shown in Figure 1G is a self-explanatory example that sonographic scanning was performed when echogenic microbubbles disappeared after CT scanning [1].
Second, I do not think the hypoechoic radiofrequency lesion shown in Figure 1C is a true ablated area because the ablation time was only 2 min. How can the hypoechoic zone at 2 min after the start of ablation (Fig. 1C) be similar in size to the hypoechoic zone at 13 min after the start of ablation [1] (Fig. 1G)? I believe the hypoechoic area could have been caused by artifact from deployed prongs of the radiofrequency electrode or by periportal creeping of microbubbles. These facts might have contributed to sonography having the poorest correlation to lesion size in the study performed by Dr. Cha et al. [1].
I think it will be not easy to clarify which modality is best for guidance even with a welldesigned experimental study. Therefore, the conclusion from the article by Dr. Cha et al. [1] should have been carefully addressed to the readers of AJR.
References
University of Wisconsin Madison, WI 53792
We thank Dr. Rhim for his interest and careful reading of our article [1]. In his letter, Dr. Rhim raises three concerns about our recently published article comparing CT and sonography for guidance of radiofrequency ablation [1]. The first issue questions the need for monitoring during radiofrequency ablation. We disagree with his assertion that "...monitoring during ablation seems to be not so critical for the success of the ablation." In our opinion, the two main challenges facing the use of radiofrequency today are the ability to reproducibly kill all tissue within a targeted zone and the ability to precisely direct the application of radiofrequency energy. As radiofrequency technology improves and as ablation zones become larger, monitoring of the radiofrequency lesion becomes even more important both to ensure adequate coverage of the tumor and to avoid important normal structures that do not self-protect (e.g., bile ducts, extrahepatic organs).
The high local failure rates reported for percutaneous radiofrequency for liver metastases [2, 3] point to the critical need for improved intraprocedural monitoring. Local failure in these studies happened despite sonographic imaging that erroneously suggested that the entire tumor had been ablated. Whatever sonographic criteria were used to reach this conclusion (including assuming that hyperechogenicity immediately after radiofrequency correlates with necrosis) have been shown to be inadequate. In fact, the article quoted by Dr. Rhim [2] showed a 54.5% local recurrence rate for metastases ablated by radiofrequency, and another group specifically states that the "US appearance of treated lesions was not helpful for differentiating ablated from viable tissue" [3]. These clinical results have led to an interest in investigating other monitoring methods such as MR imaging [4] or contrast-enhanced sonography [5].
The second concern raised by Dr. Rhim is that, based on Dr. Rhim's experience, radiofrequency lesions show maximal echogenicity immediately after ablation and that this area corresponds to the zone of necrosis. This statement is not supported by previous authors [3]. Also implied in this statement is that the use of CT as the initial postablation imaging modality in half of our cases (crossover design) biased our results. In our study, all CT scans were completed within approximately 1 min postablation. This was possible because we obtained prescout images of the animals and had all imaging parameters entered from prior scanning. Therefore, sonographic imaging of radiofrequency lesions commenced either immediately after ablation (in half of the cases) or a maximum of 1-2 min after ablation (and after CT) in the other half. In our clinical cases, we have also observed the hyperechogenicity that results from radiofrequency ablation, and we believe that it is still generally present this soon after the ablation. The crossover experimental design was important because if sonography was the initial postablation imaging modality in all cases, it would have potentially biased against CT.
The last point raised by Dr. Rhim concerns the hypoechoic zone in Figure 1C. Dr. Rhim states that it is not possible that the hypoechoic zone represents the zone of necrosis because it has not changed over time. Unfortunately, this will need to remain an academic argument because a specimen for pathology could be taken only at 13 min, whereas the image is from 2 min postablation.
The purpose of our study was to determine whether CT findings during and immediately after radiofrequency ablation more closely correspond to pathologic findings than sonography. Although CT scans did appear to correlate more closely with pathologic specimens than sonograms, there are many factors that will determine how a particular case is best monitored. Sonography has many unique strengths for performing radiofrequency ablation that cannot be duplicated by CT. Regardless, the need for improved monitoring tools for ablation methods is clear, and we look forward to future work that will increase confidence that targeted tumor tissue has been completely ablated with minimal collateral damage.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
