A Novel Correlation of Preoperative Gd-EOB-DTPA-contrast-enhanced MRI with FGFR4 Expression and Its Value in Targeted Therapy for Hepatocellular Carcinoma

Purpose: To assess the relationship between preoperative gadolinium ethoxy-benzyl diethylenetriaminepentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) features and fibroblast growth factor receptor 4 (FGFR4) gene expression in hepatocellular carcinoma (HCC). Materials and methods: Fifty-nine HCC patients (54 males, 5 females) who underwent preoperative enhanced MRI were retrospectively enrolled in this study. Quantitative and qualitative features of Gd-EOB-DTPA-enhanced MRI were analyzed in these pathologically confirmed HCC patients. Immunohistochemistry (IHC) and reverse transcription-polymerase chain reaction (RT-PCR) were performed to determine the mRNA and protein levels of FGFR4 in HCC. The relationship between these image features and the level of FGFR4 gene expression in HCC was evaluated by correlation analysis. Results: The FGFR4 mRNA and protein expression has significant correlation with the change of signal intensity in the phase of hepatobiliary ， IHC analysis revealed significant correlation between the protein expression of FGFR4 and the qualitative enhanced MRI feature, mainly the manifestation of the intratumoral vessels at the arterial phase. Furthermore, the presence of intratumoral vessels (P =0.034 ， OR=4.71) and heterogeneous signal performance in the hepatobiliary phase (P =0.008 ， OR=4.2) were identified as independent indicators for high FGFR4 expression in HCC. Conclusions: The findings demonstrate novel correlation between enhanced MRI features and FGFR4 gene expression, suggesting the heterogeneous signal intensity at the phase of hepatobiliary and the present of intratumoral vessels in the arterial phase as indicators for high FGFR4 expression in HCC. Our study may have clinical implication that enhanced MRI holds promise as useful modality in treatment selection of targeted therapies to HCC patients. Taken together, the findings of this study have indicated that signal intensity heterogeneous in the hepatobiliary phase and the presence of Intratumoral vessels in the arterial phase on enhanced MRI reflect the high FGFR4 expression in HCC. Considering FGFR4 as an independent predictor of targeted drug therapy such as Lenvatinib as well as a variety of small molecule drug targets, Gd-EOB-DTPA-enhanced MRI is potential modality in treatment selection and efficacy evaluation of targeted therapies for HCC patients. Further researches are needed to validate the findings in the future studies, and eventually to improve our care for patients with HCC.


Introduction
HCC is the fifth most common cancer worldwide and the fourth leading cause of cancer-related death, mainly attributed to relatively high incidence of hepatitis B virus (HBV)-induced HCC and delayed diagnosis of HCC at late stages [1,2]. Among the treatments for advanced HCC, targeted cancer therapy, which specifically interferes with target molecules to block the tumor growth and spread, has emerged as a new generation of treatment.
Fibroblast growth factor receptor 4 (FGFR4) , a member of the FGFR (1)(2)(3)(4) family, is a kind of tyrosine kinase receptor with the ligand fibrocyte growth factor 19(FGF19) that mediates cellular effects by stimulating autophosphorylation in reverse transcription and downstream MAP kinase 、 Akt signaling pathways [3]. It has been shown that the interactivity between FGF19 and its specific receptor FGFR4 contributes to the tumorigenesis of liver cancer [4]. Recently, several FGFR4 inhibitors, with different binding patterns and selectivity, have been developed and assessed in clinical trials for the treatment to HCC with abnormally high FGFR4 expression [5].
For instance, Lenvatinib is a potent tyrosine receptor kinase inhibitor targeting the vascular endothelial growth factor receptor (VEGFR1-3), fibroblast growth factor receptor1-4 (FGFR1-4), KIT and RET. Compared with sorafenib, Lenvatinib is the second and most commonly used first-line targeted drug for HCC approved by the American Food and Drug Administration (FDA) [6]. Lenvatinib is not inferior to sorafenib for improving tumor-free to progression and overall survival rate, is indeed superior to sorafenib and is more cost effective [7]. Although targeted drug therapy can improve the outcome of a proportion of patients with advanced HCC, it has been noted that not all patients benefit from Lenvatinib.
Differences in the clinical prognosis in patients receiving the targeted cancer therapy may be due to the heterogeneity of target genes for HCC [8].
It has been noted that FGFR4 gene express level is an independent predictor for reaction to Lenvatinib treatment and is related with longer progression-free survival and good objective response rate [9]. The study showed that FGFR4-positive tissues were related with longer progression-free survival (2.5vs5.5 months, P =0.01) and better objective response rates (31%vs81%, P =0.006). Therefore, these findings highlight importance of detecting the expression of target genes before targeted cancer therapy to achieve high therapeutic efficacy at low cost.
Currently, preoperative genetic testing for HCC is mainly performed by biopsy. However, because of the genetic heterogeneity in cancer tissues, biopsy specimen may not reflect the overall tumor characteristics. In addition, the procedures are invasive, leading to increase in the risk of procedure-associated complications (e.g. hemorrhage, tumor spread).
Notably, preoperative enhanced MRI allows examination of the overall tumor tissues, for which their corresponding characteristics can be better analyzed [10]. In fact, some MRI features of cancer tissues were significantly correlated with the mutated B-RAF and RAF-1 gene expression [11]. Hectors et al [12]proposed that the association of 28 imaging signs could predict the expression of 74% of 6732 genes, and many global gene expression profiles for HCC were related to their imaging characteristics. A previous study has revealed a correlation between contrast-enhanced MRI features on T1-weighted/diffusion weighted images ,and programed death-1 (PD -1)/programed death-ligand 1 (PD-L1) and cytotoxic T-lymphocyte-associated antigen 4 (CTLA4), the targets for immunotherapeutic cancer treatments [13].

Study patients
In this study, we retrospectively collected and analyzed 59 patients with  were all at the solid component level of HCC to avoid necrosis, blood vessels and artifacts. T1 relaxation reduction rate T1D% was calculated in the hepatobiliary phase: T1D%=(T1N-T1E)/T1N [11].

Immunohistochemical analysis
Immunohistochemistry (IHC) was used to detect the protein expression level of FGFR4 in HCC specimens. The primary antibody is mouse monoclonal antibody (5B5-ab44971 /FGFR4 antibody) used for immunohistochemical staining of all nodular specimens with FGFR4，The second antibody is goat anti-mouse/rabbit IgG polymer (SP-9000). Intratumoral specimens were taken from 1-5 sites of the primary lesion to avoid necrosis and bleeding. All specimens were stained with FGFR4 immunohistochemistry and the FGFR4 positive staining was defined as brown / yellow staining of the cell cytoplasm and membrane. In this study, after hematoxylin reverse staining, the percentage of FGFR4 positive cells was quantified microscopically by two independent examinators. The IHC staining of the FGFR4 protein expression was categorized: Grade1with staining in <10%, Grade 2 with staining in 10%-50%, and Grade3 with staining in ≥50% for tumor cells [9] (Fig. 2).

RNA extraction and RT-PCR
Total RNA was extracted from fresh collected specimens using Trizol reagent (Invitrogen) in strict accordance with standard operating instructions. Reverse transcription -polymerase chain reaction (RT-PCR) was performed to amplify target RNA, during which the following primers were used: FGFR4: forward ATTGCCAGCTTCCTACCTGAG, reverse GGGGTAACTGTGCCTATTCG. The internal reference primer is GADPH. SYBR primers and miRNA reverse transcription polymerase chain reaction kits are used for polymerase chain reaction amplification.
Then Ct value of the sample was obtained according to the analytical dissolution curve and the real-time amplification curve. The FGFR4 gene expression in the HCC tissues was found to be relative to that of the adjacent tissues(△Ct caliv), as calculated using the following formula: 2 -△△ Ct Where, △△C t= △Ct ca-△Ct caliv， △Ct ca = Ct ca-Ct GADPH， △Ct caliv = △Ct caliv-Ct GADPH [14].

Discussion
FGFR4 is the target of the first-line targeted drug Lenvatinib in targeted cancer therapy, and can be used as an independent predictor of the therapeutic response to Lenvatinib [9]. To date, the relationship between the Extensive previous studies have demonstrated that FGFR4 plays a key role in multiple cellular processes, and is involved in the development and progression of various cancers [15], such as malignant melanoma [16], breast cancer [17] and renal cell carcinoma [18] ， as well as HCC [19], There are possibilities that the small sample size was relative small, or the relatively small number of T1mapping tests conducted in this study.
Lenvatinib and some other anticancer drugs specifically targeting FGFR4 prevent angiogenesis and inhibit tumor growth through downstream MAP kinase and Akt signaling pathways [19]. Therefore, the use of Lenvatinib or specific FGFR4 inhibitors may have better therapeutic effect in patients with high expression of FGFR4 gene. Previous studies on the correlation between the expression of this molecule and imaging were limited and the findings were inconsistent in previous studies [22][23][24][25]

Consent for publication
Not applicable.
provided experimental guidance and participated in the specimen collection.

Data Availability
The data supporting the conclusions of this study are available from the corresponding author upon a reasonable request.

Conflicts of Interests
The authors declare no competing interests.              was statistically different between the two groups (P=0.041); (B) FGFR4 mRNA expression levels between the hypointense group and the heterogeneous group in the hepatobiliary phase. The FGFR4 mRNA expression was statistically different between the two groups (*P=0.005).