IRE has attracted increasing interest because the non-thermal properties of this method  permit ablation of tumours adjacent to vital structures . Previous studies have shown that the ultrasound findings of IRE applied to the liver will change from a few seconds to at least a few hours with time [3, 17, 18]. Our results show the evolution of post-IRE ablation patterns over a clinically relevant period.
As noted in an animal-based experiment, a strong association exists between conventional US and histopathology . In our study, on conventional US, the IRE ablation zones of most patients appeared as a developing hypo-echoic area that demonstrated an increasingly hyperechogenic ablation zone starting one day after the procedure.
Our US findings after IRE ablation evolved sequentially over time. The findings showed good correlations with those in previous studies [5, 19]. In the study, specimens obtained immediately after ablation showed that the enlarged oedematous sinusoidal space was mainly filled with fluid, and almost no haemorrhagic infiltration was observed. As time passed, the haemorrhagic infiltrate became more dominant, which might be due to the widened fluid-filled sinusoids at the beginning of the procedure. The authors attribute this hyper-echogenicity to red blood cell accumulation over time. The degree of erythrocyte infiltration was qualitatively related to echogenicity. Lee et al. also reported the status of 55 ablation zones immediately after ablation and 1 day after ablation and showed that the immediate low echogenicity of the treatment area was converted into total hyper-echogenicity on the 1st postoperative day . Appelbaum et al. speculated that red blood cells progressively infiltrated into deep regions of the ablated zone such that the hyper-echoic rim on US at 90–120 min transitioned to complete hyper-echogenicity within 24 h . Concordance exists between our findings and those of previous studies describing the characteristics of images obtained after ablation.
In CEUS, most of the ablated zones showed hypo-enhancement immediately after IRE. Chung et al. found the following zones with different enhancement patterns on CT perfusion images in normal porcine liver: an inner non-enhanced zone; a middle well-defined progressive internal enhancement zone, and an outer ill-defined arterial enhancement zone. On histopathology, the inner and middle zones accounted for the extent of cell death . The histological examination suggested that the apoptotic process was involved, with complete cell death caused by IRE according to pathophysiology.
The CEUS images of the IRE ablation zones in normal liver tissue differed from the images of liver tumours treated with IRE. Our study also found that 12 ablated zones showed hypo-enhancement immediately after IRE but became non-enhanced one day after IRE for various reasons. Lee et al. attributed the focal hyper-attenuation to the release of contrast medium into an ablation defect caused by an IRE-induced microvasculature leakage within the defect zone . Because the zones became non-enhanced on the follow-up CEUS images, the hyper-enhancement was probably caused by extraluminal contrast material. Another opinion concerns differences in contrast agent concentrations. Guo et al. suggested that nanometre-scale pores in the tissue cell membrane caused contrast agent to accumulate in the IRE zone and allowed the contrast agent to be internalized into the intracellular environment rather than remaining extracellularly , which complicates evaluation of whether the viable tissue is a residual tumour after ablation. Therefore, follow-up CEUS is necessary to assess the viable portion. However, unlike normal regeneration activity, the residual tumour can continue growing and results in a newly enhanced region.
Most of the non-enhancement pattern on CEUS 1 month after ablation is a sign of effective IRE treatment, but false-negative results may occur. Such non-enhancement patterns are different to those of the thermal ablation zone after complete ablation. Previous histologic examinations of the ablation zones showed cell death caused by apoptosis . Cell death is observed with full preservation of the peri-ablative zone structures, such as blood vessels and bile ducts.
In the present study, only one patient continuously showed an enhanced ablation zone immediately and on follow-up CEUS images. The ablation zone became hyper-enhanced during the early arterial phase, slightly washed out in the late arterial phase and appeared hypo-enhanced in the late phase. To elucidate the reason for this pattern, the patient underwent liver biopsy one month after IRE. The physiology results showed small patchy necrosis, inflammatory fibrosis, tissue hyperplasia and foam cell aggregation. The result ruled out the possibility of recurrence. A reasonable explanation for this finding is hyperplasia of the inflammatory tissue. This result can cause confusion in clinical practice. Therefore, a sequential follow-up is essential. Further investigation is needed to study the histological and cytological mechanisms underlying this process.
In our study, only 2 ablation zones still appeared as hypo-echoic areas, and both zones were in the same patient. After 1 month of follow-up, recurrence occurred near one of the zones treated with IRE ablation. In a previous multi-institutional review from 2009 through 2012, 31% of the patients had recurrence during a median follow-up of 18 months. Among the study population, 31% of the patients had recurrence, and 10.7% had local recurrences at the ablated site . Previous studies have attributed such recurrences to electric field sinks resulting from the heterogeneous structure and conductivity of the liver . Later work from the authors indicated that the IRE-treated extracellular matrix (ECM) provides an environment for activation and differentiation of progenitor cells , but the mechanism is not completely understood; in contrast, some studies have suggested that abnormal ECM affects cancer progression by directly promoting cellular transformation and metastasis as well as tumour-associated angiogenesis and inflammation, leading to the generation of a tumorigenic microenvironment . The role of the IRE-spared tumour matrix in follow-up recurrences requires further research. Another theory is related to the size of the treated liver tumours. Niessen et al. found that large tumour volumes (> 5 cm3) portended early local recurrence .
In our study, the intra-hepatic blood vessels and bile duct remained almost completely intact after IRE. The hepatohilar bile duct of only one patient showed an unclear contrast agent pattern immediately after IRE. After 15 min, the phenomenon disappeared. However, a consensus on the mechanism of IRE has not yet been reached. This effect may be due to the high proportion of collagenous connective tissue. A further hypothesis is that gap junctions, which are present in large numbers in the muscularis propria of blood vessels and the bile duct walls, may act as a conductive structure for the electrical currents; thus, the current can pass from cell to cell without causing destruction of the cell membrane . Whether the effects of IRE are caused by a thermal or non-thermal mechanism remains unclear.
This study has several limitations. First, our study had a small sample size over a span of more than 2 years. Second, 9 lesions of the tumour were not diagnosed histologically before IRE. The last limitation was that almost all of the patients included in our study group previously underwent right or left liver resection and thermal ablation or catheter chemical ablation. The origin of the ablation zones was heterogeneous. However, this circumstance mirrors the status of our clinical treatment strategies. IRE as a novel technique was not the first choice in our hospital. Only patients who experienced tumour recurrence after liver resection and thermal ablation were considered for IRE. Because of this situation, assessing the images of the ablated zones during subsequent follow-up imaging examinations is important. We conclude that CEUS may be a useful tool for assessing the characteristics of post-IRE ablation changes and is an effective method to evaluate the therapeutic efficacy 1 month after ablation. Further studies are needed to evaluate more patients to precisely depict the appearance of hepatic zones on CEUS.