Written reply: “Optimal multibipolar radiofrequency parameters should overcome heat-sink effect”

Sir,

We took note of the letter to the editor by Hocquelet et al. [1] regarding our article “The vascular cooling effect in hepatic multipolar radiofrequency ablation leads to incomplete ablation ex vivo” with great interest. Hocquelet et al. showed in their large multicentre study that multipolar radiofrequency ablation (RFA) outclasses monopolar RFA, regarding primary RF success and sustained local tumour response next to hepatic vessels [1]. However, his working group also pointed out that ablations near large vessels were independent factors associated with an increased risk for RFA failure. This is in accordance with our results [2]. We observed a vascular cooling effect around an artificial vessel in our ex vivo study. We choose this setting to examine RFA in a standardised setting, which is not influenced by biological variations. However, further in vivo studies will be necessary to examine exactly the heat-sink effect, described by both working groups. An optimal selection of ablation parameters and applicator positioning, using multipolar RFA, should considerably reduce incomplete ablation rates. Nevertheless, heat-sink effects have to be considered in multipolar RFA as well. Hocquelet et al. [1] addressed several issues in their letter to the editor:

Ad 1): We agree that bipolar applicators with an active length of 20 mm (T-20) should not normally be used in clinical for the treatment of hepatic malignancies, since ablations are becoming small. We used applicators with an active tip length of 20 mm (T-20) in our ex vivo setting to avoid artefacts at the liver border. These artefacts would occur in the absence of a normal liver perfusion, as RFA lesions tend to get large in the relatively small porcine liver. In our experimental setting we also used the recommended starting power of 1 Watt/mm (according to 60 Watt). However, it has to be distinguished between preselected starting power (manufacturer’s recommendation: 1 Watt/mm) and actual power output. The RFA-generator chooses optimal power output according to the actual impedance of the tissue in order to avoid dehydration and carbonisation of the tissue (the so-called “Resistance Controlled Automatic Power” – RCAP mode). In our experience (experimental and clinical), the generator never uses the preselected maximal power output. We agree that energy delivery is dependent on active tip length. However, the actual power output is usually considerably lower than 1 Watt/mm.

Ad 2): The authors addressed are very interesting issue here. In clinical practice, we choose output power and total energy according to the dosimetry table of the manufacturer, instead of waiting for the automatic shutdown by the generator. Impedance of the ablated tissue increases slowly by using the RCAP-mode and therefore ablations longer than 1 h may arise in clinical practice (in an ex vivo setting ablations more than 120 min are not uncommon without automatic shutdown). Therefore, exact dosimetry tables for each applicator size and applicator distance seem to be mandatory.

Ad 3): There is some evidence that vessels <3 mm in diameter and within an ablation zone occlude during ablation [3]. Therefore, these vessels do not seem to be crucial for vascular cooling effects. However, vessels next to the ablation zones take part in the diffuse cooling effect and therefore limit ablation size. Additionally, it has to be examined, which is the optimal applicator position in order to reduce vascular cooling effects as far as possible. In vivo studies would be of great interest.

Ad 4): A percutaneous puncture of the portal or hepatic vein under ultrasound guidance may be a possible alternative for an experienced physician. However, an ultrasound guided puncture of the portal or hepatic vein remains demanding and is not always possible. Additionally, a considerable vascular cooling effect of the arterial perfusion persists performing a selective balloon occlusion of the portal or hepatic vein [4].

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.