When RFA is used to treat medium or large HCC, local recurrence and progression occurs frequently due to the remaining of microscopic residual tumors at the periablational zone [
33]. Even more, some recurrent HCC after suboptimal RFA could progress rapidly showing the infiltratively growing pattern [
34], albeit the mechanism underlying this phenomenon remains unclear. Here, we demonstrate that collagen I, apparently accumulated at the border of ablation zone after RFA, could aggravate the malignant behaviors of heat-treated residual HCC. This unwanted “off-target” effect of thermal ablation provides a new explanation why insufficient RFA could promote the aggressive progression of residual HCC. Second, we propose that sorafenib could reverse this detrimental pro-tumor effect, suggesting a potential treatment approach to thwart residual tumor progression after incomplete RFA. Our findings have clinical implications in improving the therapeutic outcome of RFA.
HCC occurs in the setting of cirrhotic liver with ECM richness [
35,
36]. In previous studies, massive collagen deposit was observed at the periphery of periablation zone after RFA of liver tissue [
17]. Among the major ECM proteins, the increased collagen I is associated with fibrotic diseases and tumor development [
37‐
40]. Tumor cells or stromal cells (e.g. hepatic stellate cells) may be the source of collagen I [
41,
42]. According to literature [
42,
43], we assume that collagen I in HCC is mainly produced by activated hepatic stellate cells. After thermal ablation, stromal cells will be recruited around the ablative zone [
44] resulting in the increased collagen I deposition at the periablational zone. In the present study, we provided the evidence of the crosstalk between the increased collagen I after RFA and residual HCC cells, that is, collagen I promotes the proliferation, EMT, and progenitor-like characteristics of heat-exposed surviving HCC cells via ERK activation. Also, we show that sorafenib could thwart collagen I-enhanced progression of heat-treated residual HCC cells, indicating a new treatment strategy to inhibit progression of viable HCC after RFA. In line with the previous study, abnormal activation of ERK signaling is correlated with local recurrence of residual HCC after sublethal heat treatment [
9,
29‐
31]. Sorafenib shows survival benefits in advanced HCC patients by blocking many kinase pathways including RAF/MEK/ERK [
32,
45‐
47]. Xu et al. reported that sorafenib inhibited residual HCC progression after incomplete RFA in an animal HCC model [
48]. In this study, we further elucidate that sorafenib could block the cross-talk between ECM protein collagen I and residual HCC cells through disrupting ERK signaling.
This study has several limitations. First, besides collagen I, we could not exclude the other factors in post-inflammation reaction after RFA that influence tumor progression of residual HCC [
14,
15,
49,
50], such as RFA-induced tumor-specific T-cell reaction, a Th1 cytokine pattern after RFA, heat shock proteins, periablational cellular infiltration. Second, we did not identify the source of collagen I. The thermal ablative environment may stimulate the production of collagen I from tumor cells or stromal cells. Third, we employed a subcutaneous tumor model of implanting heat-treated residual HCC cells in mice to study the response to sorafenib therapy. This model is the absence of liver microenvironment. Better animal models (rabbit VX2 hepatoma, a MDR2-knockout inflammation-induced HCC model resembling human HCC) are needed to verify our findings.