Enhancing delivery of doxorubicin using triolein emulsion in rabbit VX2 liver tumor model

Abstract

BACKGROUND

A previous study reported that a triolein emulsion (TE) injection to the liver markedly enhanced the concentration of doxorubicin by near two-fold. Hence, a TE injection could serve as a beneficial supplementary treatment to enhance the chemotherapeutic efficacy in liver cancer.

AIM

To evaluate the antitumor effects of doxorubicin with and without a TE infusion to the liver in the VX2 hepatic tumor model of rabbit.

METHODS

VX2 tumor tissues were inoculated into the liver of 16 rabbits. Seventeen days after implantation, contrast-enhanced computed tomography (CECT) was performed to localize the hepatic tumors. The rabbits were divided into two groups: TE with doxorubicin (TE-Doxo) (n = 8) and normal saline (NS) with doxorubicin (NS-Doxo) (n = 6). Intra-arterial transcatheter administration was performed via the transauricular approach. Two rabbits in the group without TE failed to undergo the transauricular approach. Follow-up CECT was performed nine days after transcatheter arterial chemoembolization (TACE). The tumor volumes were measured on the pre- and post-TACE CT scans, and the growth ratio was estimated. Statistical analysis was performed (nonparametric Mann-Whitney U tests and the Wilcoxon signed-rank test).

RESULTS

The mean tumor volume was significantly smaller in the TE-Doxo group (from 1913.09 ± 489.58 mm³ to 1229.75 ± 395.61 mm³) than in the NS-Doxo group (from 874.36 ± 361.65 mm³ to 1574.17 ± 890.12 mm³). The mass growth ratio was also significantly reduced in the TE-Doxo group (-41.68% ± 7.32%) than in the NS-Doxo group (76.59% ± 77%; P = 0.002).

CONCLUSION

The TE infusion enhanced the antitumor efficacy of doxorubicin, significantly reducing the tumor volume and growth ratio in the VX2 hepatic tumor model. These results suggest that a TE infusion may be an effective adjuvant treatment to improve the TACE outcomes in hepatic tumor.

Key Words: Transcatheter arterial chemoembolization; Hepatocellular carcinoma; VX2 tumor; Liver; Doxorubicin; Triolein emulsion

Core Tip: The infusion of triolein emulsion into the hepatic arteries increased the liver vascular permeability in the sinusoidal capillaries by transiently disrupting endothelial barriers, leading to a minimal transient decrease in liver function without specific histopathological changes in the rabbit livers. Consequently, any triolein-induced endothelial loosening can substantially enhance drug extravasation into liver tissue. This study aimed to evaluate the antitumor effects of doxorubicin by comparing the changes in tumor volume and the growth ratios between the triolein emulsion with doxorubicin and normal saline with doxorubicin groups via a hepatic arterial infusion in a rabbit liver VX2 tumor model.

INTRODUCTION

Hepatocellular carcinoma (HCC) ranks among the most common malignant tumors worldwide, responsible for roughly 500000 cancer-related deaths each year[1-3]. While surgical treatment is considered the standard approach, only 10%-20% of patients qualify for hepatic resection or transplantation due to factors such as advanced disease stage or compromised liver function[3-5]. For patients with unresectable HCC, multiple therapeutic options have been investigated, with transcatheter arterial chemoembolization (TACE) being one of the most widely adopted. Agents frequently employed in TACE include doxorubicin, cisplatin, and 5-fluorouracil. However, the clinical benefits of TACE remain modest, and overall outcomes for HCC continue to be unsatisfactory[6]. Therefore, the development of innovative treatment strategies is required to improve the management of HCC.

Doxorubicin, one of the first and most commonly used chemotherapeutic agents for HCC, is often used in intermediate HCC cases via angiography-guided transcatheter therapy, a technique known as TACE[7-12]. This approach has two primary objectives: Delivering the chemotherapeutic agent directly into the tumor and inducing embolization of tumor artery[7,13]. This technique increases local drug concentrations by delivering the drug close to the tumor, thereby enhancing the antitumor efficacy while minimizing systemic exposure and side effects[14].

Our previous studies[1,15] demonstrated that infusion of a triolein emulsion (TE) into the rabbit liver increased vascular permeability, while causing only a transient and minor reduction in liver function without notable cellular injury. Moreover, administration of TE through the hepatic arteries produced nearly a two-fold elevation in the concentration of doxorubicin.

This study assessed the antitumor effects of doxorubicin by comparing the changes in tumor volume and the growth ratios between the TE with doxorubicin (TE-Doxo) and normal saline (NS) with doxorubicin (NS-Doxo) groups via a hepatic arterial infusion in a tumor model.

MATERIALS AND METHODS

Animal preparation and VX2 liver tumor model

All experiments were carried out in compliance with the institutional guidelines for laboratory animal protocols (approval No. PNUYH-2019-073). The animal facility was maintained at a temperature of 22 ± 2 °C, with relative humidity controlled at 50% ± 15% and a 12-hour light/dark cycle. Prior to experimentation, the animals were kept in an environmentally regulated breeding room and provided with standard laboratory chow and water for three days. Throughout the study period, all rabbits were monitored daily by a veterinarian for general health status, including activity level, behavior, and signs of adverse events. Body weight was measured at the beginning of the study, and any abnormal clinical findings were reported during the study.

Prior to drug administration, the animals were fasted for 12 hours but allowed free access to water. Sixteen New Zealand White rabbits, weighing between 2.6 kg and 3 kg, were obtained from Samtao (Osan, South Korea) for use in this study. Anesthesia was induced by intramuscular injection of ketamine hydrochloride (2.5 mg/kg; Huons, Jecheon, Republic of Korea) combined with xylazine (0.125 mg/kg; Bayer Korea, Seoul, Republic of Korea) into the groin muscles. During the procedures, the rabbits were ventilated with room air. Body temperature was monitored using a rectal probe (MGA-III 219; Shibaura Electronics, Tokyo, Japan) and maintained between 35.5 °C and 36.5 °C with a heating pad. Following anesthesia, each rabbit was placed in the supine position on a custom-built wooden restraining table, with all four limbs secured using straps.

The VX2 carcinoma strain was sustained through serial transplantation into the hind limb of a carrier rabbit. Under anesthesia with ketamine chlorine hydride (HCl) and xylazine HCl, a midline subxiphoid abdominal incision was performed, and 0.1 mL of minced VX2 carcinoma tissue was implanted directly into the subcapsular parenchyma of the left medial hepatic lobe. All surgical procedures were conducted under aseptic conditions. Tumors were then allowed to develop for two to three weeks until reaching a diameter of 15-30 mm.

Transauricular approach for TACE, infusion of TE and doxorubicin

The dorsal surfaces of the ears were shaved, and dressings were applied to the puncture sites. To enlarge the lumen, the central auricular artery was repeatedly tapped, after which it was punctured at random in either the right or left ear using a 20-gauge intravenous catheter (Insyte, Becton Dickinson, Sandy, UT, United States). Once the plastic sheath was advanced and the inner stylet needle removed, a 2.0-microcatheter (Progreat α^®; Terumo, Tokyo, Japan) equipped with a 0.016-inch guidewire (GT Wire; Terumo, Tokyo, Japan) was introduced through the IV catheter and advanced to the aortic arch. Manipulation of the microcatheter and guidewire was performed under fluoroscopic guidance to enable selection of the celiac trunk. Angiography was then carried out via manual injection of contrast medium (Xenetix^®; provided by Guerbet Korea, Seoul, Korea), and under fluoroscopic control, the microcatheter tip was positioned within the left hepatic artery (Figure 1).

Figure 1 Angiography of VX2 tumor via the hepatic artery in a rabbit liver. The hepatic angiogram demonstrates tumor staining in the left lobe of the liver. The catheter tip is positioned in the left hepatic artery through the celiac axis using the trans-auricular approach. Triolein emulsion or normal saline, along with doxorubicin, were administered through a microcatheter.

Sixteen rabbits were allocated into two groups according to the intra-arterial agents administered: TE-Doxo (20 mL, n = 8) and NS-Doxo (20 mL, n = 6). The TE procedure followed the method described by Kim et al[16]. Each rabbit received a 0.6% TE solution (20 mL) infused via microcatheter at a rate of 4 mL/minute. In the control group (n = 6), 20 mL of NS was infused instead of TE. Immediately following TE or NS infusion, doxorubicin hydrochloride (2.4 mg/kg; Boryung Pharmacy, Kyungkido, South Korea) was administered through the same catheter at the same infusion rate in all groups.

Doxorubicin concentrations in the liver and VX2 tumors were assessed using fluorometric analysis. Liver tissue (100 mg) and tumor samples were collected from two rabbits in the TE-Doxo group. Fluorescence intensity was measured in the supernatants with a fluorometer (Synergy H1, BioTek, Winooski, VT, United States) using excitation/emission wavelengths of 470 nm/585 nm.

Analysis of antitumor effect on computed tomography images

Following VX2 carcinoma implantation, tumors were allowed to grow for two to three weeks until reaching 15-30 mm in diameter. Liver computed tomography (CT) scans (SOMATOM Definition AS+, Siemens, Germany) were performed two to three weeks post-implantation with the animals in the supine position, on the same day as intra-arterial agent administration. A follow-up contrast-enhanced CT (CECT) was obtained nine days after TACE. For CECT, 13 mL of contrast medium (iopamidol; Scanlux^®; Sanochemia Pharmazeutika AG, Leitha, Austria) was injected via the auricular vein at 0.5 mL/second. Using bolus tracking, hepatic arterial phase and portal venous phase images were acquired approximately 7-8 seconds and 24-25 seconds after contrast injection, respectively. CT parameters included 3-mm collimation, 1.5 pitch, and 1-mm reconstruction intervals. Tumor location and size were measured in consensus by two authors. Tumor volume (V) was calculated using the formula: V = L × S²/2, where L and S represent the longest and shortest tumor diameters, respectively[3,17]. Tumor volumes were compared between baseline CT and post-TACE CT scans. Growth ratios were determined using the formula: (V9/Vb - 1) × 100, where V9 is the tumor volume nine days after TACE and Vb is the baseline tumor volume[17].

Statistical analysis

Data analysis was performed using SPSS Statistics (version 26.0; IBM Corp., Armonk, NY, United States). A P value < 0.05 was considered statistically significant. Categorical variables were expressed as n (%). Due to the small sample size, the Wilcoxon signed-rank test and the nonparametric Mann-Whitney U test were applied.

RESULTS

All rabbits tolerated the experimental procedures without severe adverse events. Mild body weight loss was observed in some animals during the post-TACE period; however, no significant behavioral changes or signs of systemic toxicity were noted. TACE was successfully performed on 14 rabbits, whereas the trans-auricular approach failed in two rabbits in the NS group. Selective accumulation was observed in the hepatic tumors treated with contrast-mixed doxorubicin and either TE or NS during TACE (Figure 1). The mean doxorubicin concentrations in two rabbits of TE-Doxo group were 28680.7 in right lobe, 30207.7 in left lobe, 26116.0 in the central area and 42149.2 in the periphery of liver tumor. The ratio of men doxorubicin concentration was 1.00 in right lobe, 1.05 in left lobe, 0.91 in the central area of tumor, and 1.47 in the periphery of liver tumor. The volume changes of VX2 liver tumors from pre-TACE CT to post-TACE CT scans are summarized as follows. In the TE-Doxo group, the tumor volumes were 1913.09 ± 489.58 mm³ before TACE and 1229.75 ± 395.61 mm³ after TACE. In the NS-Doxo group, the tumor volumes were 874.36 ± 361.65 mm³ before TACE and 1574.17 ± 890.12 mm³ after TACE (Figure 2). The change in VX2 liver tumor volume was markedly smaller in the TE-Doxo group compared with the NS-Doxo group (Table 1). Statistical analysis using the Wilcoxon signed-rank test confirmed a significant difference in tumor volume change between the two groups (P = 0.028). The growth ratios of VX2 liver tumors, measured from pre-TACE to post-TACE CT scans, were -41.68 ± 7.32 in the TE-Doxo group and 76.59 ± 77.51 in the NS-Doxo group.

Figure 2 Representative computed tomography images of VX2 liver tumors before and after transcatheter arterial chemoembolization. A and B: In the normal saline with doxorubicin group, pre-transcatheter arterial chemoembolization (TACE) (A) and post-TACE (B) images show a marked enlargement in tumor volume (arrows) and a significantly higher tumor growth ratio; C and D: In the triolein emulsion with doxorubicin group, pre-TACE (C) and post-TACE (D) images demonstrate a smaller reduction in tumor volume (arrows) and a significantly lower tumor growth ratio. All panels depict the largest tumor diameter slice for accurate volume estimation, and the tumor locations are consistent across pre- and post-TACE images.

Table 1 VX2 tumor volume of pre-transcatheter arterial chemoembolization computed tomography and post-transcatheter arterial chemoembolization computed tomography after doxorubicin transcatheter arterial chemoembolization with and without triolein emulsion in a rabbit liver, mean ± SD.

	Tumor volume; (V9/Vb - 1) × 100
	Pre-TACE CT (Vb)	Post-TACE CT (V9)
TE-Doxo (n = 8)	1913.09 ± 489.58	1229.75 ± 395.61
NS-Doxo (n = 6)	874.36 ± 361.65	1574.17 ± 890.12

TE-Doxo: Triolein emulsion with doxorubicin group; NS-Doxo: Normal saline with doxorubicin group; TACE: Transcatheter arterial chemoembolization; CT: Computed tomography.

Tumor growth ratios in the TE-Doxo group were markedly lower than those observed in the NS-Doxo group, demonstrating significant suppression of tumor progression in rabbits treated with doxorubicin combined with TE compared to those receiving NS-Doxo (Figure 2 and Table 2). Statistical analysis using the nonparametric Mann-Whitney U test confirmed a significant difference in growth ratios between the two groups (P = 0.002).

Table 2 Longitudinal changes in the growth ratio of VX2 tumor volume measured by computed tomography before and 9 days after doxorubicin-based transcatheter arterial chemoembolization in rabbit liver tumors treated with triolein emulsion and normal saline, mean ± SD.

	Growth ratio
TE-Doxo (n = 8)	-41.68 ± 7.32
NS-Doxo (n = 6)	76.59 ± 77.51

Negative values indicate tumor volume reduction, whereas positive values indicate tumor volume increase. TE-Doxo: Triolein emulsion with doxorubicin group; NS-Doxo: Normal saline with doxorubicin group.

DISCUSSION

VX2 rabbit liver tumor model

Rabbit VX2 tumors have long been employed as experimental models to assess a variety of therapeutic approaches, particularly radiofrequency ablation and diverse transarterial embolization techniques[18]. The VX2 carcinoma model was first established by Rous and Beard[19]. Originating from a skin cancer in cottontail rabbits, it was later shown to be transplantable across all domestic rabbit strains. The tumor develops following infection with the cottontail rabbit papillomavirus[18], and this model has been widely applied in studies of multiple human cancers, including those of the lung, bladder, head and neck, breast, liver, and kidney. Although VX2 carcinoma does not arise from hepatocytes, it is frequently utilized as a surrogate model for HCC due to its exclusive hepatic arterial blood supply and rapid growth characteristics[3].

The VX2 rabbit liver tumor model is widely used in preclinical studies of locoregional therapies due to its reproducible tumor growth, hypervascularity, and suitability for catheter-based interventions[20,21]. However, VX2 tumors differ from human HCC in origin, microenvironment, and immune context, limiting direct extrapolation of efficacy outcomes[21]. Previous studies demonstrate that VX2-based interventions can reliably model treatment-induced changes in tumor perfusion and viability, providing meaningful comparative data[21,22]. Accordingly, the present findings should be interpreted as preclinical, proof-of-concept evidence demonstrating the feasibility and localized effects of TE-assisted drug delivery rather than definitive predictors of clinical efficacy. Taken together, these results support further evaluation in models with closer resemblance to human HCC and eventual clinical translation[20-22].

TACE with doxorubicin and TE infusion

HCC is a highly aggressive malignancy with a poor prognosis, often leaving patients with only a few months of survival following diagnosis. Surgical resection of the primary tumor remains the most effective treatment option; however, only 10%-30% of patients present with disease amenable to surgery. For unresectable HCC, TACE is employed as a therapeutic strategy. The principle of TACE lies in delivering a concentrated dose of chemotherapeutic agents directly to the liver, as most hepatic tumors are supplied predominantly by the hepatic artery[17].

TACE is widely recognized as a locoregional therapy for HCC, achieving tumor response rates of 60%-80%[11,23] and providing survival benefits[4,6,23]. In this procedure, chemotherapy is administered directly to the tumor, with doxorubicin an anthracycline being the most commonly used agent[23]. Conventional TACE involves mixing chemotherapeutic drugs with ethiodized oil to create an emulsion. Tumor cell death is thought to result from both ischemia caused by embolization and the cytotoxic effects of chemotherapy[23]. Nonetheless, the exact mechanisms by which TACE induces tumor necrosis remain unclear, and the biological processes underlying cell death after TACE are not yet fully understood.

Typically, iodized oil (Lipiodol; Laboratoire Guerbet, Aulnay-Sous-Bois, France) is combined with anticancer agents to form oil-in-water or water-in-oil emulsions, though these mixtures are often unstable[17]. Yoon et al[17] reported that when TACE was performed with lipiodol alone, the VX2 tumor grew significantly more than when lipiodol was mixed with a highly lipophilic anticancer drug, such as paclitaxel. Therefore, the combined effects of an embolic material and anticancer drugs are believed to be more effective. This method differs from the present study, which did not use a mixture of embolic material and an anticancer drug. Instead, a TE was injected to increase the liver vascular permeability, followed by a doxorubicin infusion to enhance the tumor uptake[1]. This approach led to an approximately two-fold increase in the doxorubicin uptake in the liver, resulting in a decreased tumor volume and growth ratio in the VX2 tumor model. The present study was not designed to directly elucidate the mechanisms underlying TE-enhanced doxorubicin efficacy. However, prior experimental evidence suggests that TE induces a transient and reversible increase in vascular permeability through temporary endothelial barrier modulation, which may facilitate enhanced locoregional drug delivery. In previous hepatic arterial infusion studies, TE increased intrahepatic doxorubicin concentrations without significantly affecting drug levels in off-target organs, indicating a localized delivery effect rather than systemic redistribution. While the relative contributions of increased permeability, local drug retention, and altered regional pharmacokinetics cannot be distinguished based on the current data, these observations provide a biologically plausible explanation for the improved tumor control observed. Conceptually, this delivery strategy parallels recent refinements in TACE, which emphasize optimization of intratumoral drug exposure while minimizing systemic toxicity and excessive embolization[24,25].

We conducted multiple animal experiments using TE in the brain, eye, testis, and liver organs that possess blood-organ barriers designed to protect against harmful substances. Injection of TE disrupted these barriers and increased vascular permeability, which appeared as contrast enhancement on CT or magnetic resonance imaging scans. Although the rise in vascular permeability occurred immediately after TE administration and persisted for one to three days, it was reversible and returned to baseline within seven days[15,16]. Infusion of TE into rabbit livers caused only a minimal, transient decline in liver function without notable histopathological alterations[15]. Moreover, TE infusion into the hepatic arteries significantly elevated doxorubicin concentrations, approximately doubling them compared with NS infusion. In contrast, in the lungs an off-target organ the mean doxorubicin levels did not differ significantly between the TE and control groups[1]. However, we did not assess directly on systemic toxicity or off-target organ effects following TE administration in present study. This contextualizes the need for future safety studies in accordance with established protocols.

Evaluation of antitumor effects with CT analysis

Volume change of VX2 liver tumors: In this study, the reduction in VX2 liver tumor volume was significantly greater in the TE-Doxo group compared with the NS-Doxo group (Table 1). Yoon et al[17] observed that the concentration of anticancer drugs within tumor nodules gradually declined, reaching levels similar to those in non-tumorous liver tissue by seven days after TACE, while drug concentrations in non-tumorous liver tissue consistently remained below 10 μg/g throughout the experiment. In our earlier work[1,15], infusion of TE into the hepatic arteries transiently disrupted endothelial barriers at the sinusoidal capillaries, thereby increasing vascular permeability. This effect produced only a minimal, short-lived decline in liver function without notable histopathological changes in rabbit livers. The phenomenon is further accentuated by the distinctive architecture of hepatic sinusoids, which feature a relatively wide lumen (often exceeding 30 μm in diameter), irregular and discontinuous endothelial lining, and large intercellular gaps of about 100 nm. Additionally, the basal lamina is incomplete, and there is no well-defined morphological barrier separating the sinusoids from the perisinusoidal space. These features allow plasma contact with hepatocytes and facilitate efficient transvascular exchange. Consequently, any triolein-induced endothelial loosening can substantially enhance drug extravasation into liver tissue[1,15]. This resulted in an approximately two-fold increase in the doxorubicin concentration. These findings suggest that follow-up CECT at nine days after TACE would have maintained higher doxorubicin levels in the VX2 liver nodules than in the non-tumorous liver tissues and the NS-Doxo group. Consequently, the volume change of VX2 liver tumors was remarkably smaller in the TE-Doxo group than in the NS-Doxo group.

The rabbits were randomly assigned to the TE-Doxo and NS-Doxo groups. The observed difference in pre-TACE tumor volume of two groups is attributable to natural variability inherent in randomization rather than a systematic bias. Baseline tumor size is a critical determinant of tumor growth kinetics, drug distribution, and therapeutic response in solid tumors. Larger tumors often exhibit altered vascular architecture, central necrosis, and heterogeneous drug penetration, which may confound treatment efficacy assessment in preclinical models[26,27]. Therefore, the larger initial tumor volume in the TE group represents a potential limitation of this study. Importantly, despite this unfavorable baseline condition, the TE group demonstrated a greater antitumor effect compared with the NS group, suggesting that the observed therapeutic benefit is unlikely to be explained solely by baseline tumor size differences.

Growth ratios of VX2 hepatic tumors: From pre-TACE to post-TACE CT scans, the growth ratios of VX2 liver tumors were -41.68 ± 7.32 in the TE-Doxo group and 76.59 ± 77.51 in the NS-Doxo group. These findings indicate that TE combined with doxorubicin exerted a stronger antitumor effect on transplanted VX2 liver tumors. Tumor growth was markedly suppressed, and the proportion of viable tumor tissue was substantially lower in the TE-Doxo group compared with the NS-Doxo group. Yoon et al[17] demonstrated that high intratumoral concentrations of anticancer drugs can be sustained for at least five days following a single TACE session. In our earlier study[1], TE infusion into the hepatic arteries produced a significant two-fold increase in doxorubicin concentration. This elevation in drug levels explains why doxorubicin concentrations in VX2 liver tumors were higher in the present study than in Yoon et al’s VX2 model[17], and why anticancer drug levels persisted for more than seven days after TACE.

This increase in drug concentration explains why the doxorubicin levels in VX2 liver tumors were higher in the present study than in Yoon et al’s VX2 study[17], and why the concentrations of anticancer persisted more than seven days after TACE. VX2 tumors, when untreated, typically exhibit exponential growth. However, the linear growth assumption is commonly applied in preclinical TACE studies using the VX2 liver tumor model to facilitate comparison of relative tumor growth between treatment groups over a limited follow-up period[17]. Yoon et al[17] evaluated tumor response in VX2 rabbit liver tumors after TACE and applied a similar linear growth assessment over early post-treatment intervals, reasoning that short-term measurement reduce the impact of exponential growth deviations and allow practical calculation of growth ratios.

Despite the promising findings, this study has certain limitations. The primary concern is the relatively small number of animals used in the VX2 tumor model, which may have reduced the statistical strength of the results. Second, the efficacy of TE was not compared directly with that of the standard lipiodol-based TACE protocol. A head-to-head comparison among TE-Doxo, lipiodol with doxorubicin, and doxorubicin alone would provide valuable insights. Third, further studies using untreated NS-control group and other chemotherapeutic agents are necessary to allow proper interpretation of the combined and independent effects of TE and doxorubicin and to determine if the observed effects are specific to doxorubicin or can be generalized to other anticancer agents. Fourth, rabbits were assigned to the TE-Doxo and NSDoxo groups using a randomization procedure after confirmation of VX2 tumor implantation. As group allocation was not stratified by tumor volume, an unintended imbalance in baseline tumor size occurred between the two groups. Future studies employing stratified randomization or tumor volume–matched groups will be conducted to further validate the therapeutic effect of TE. Fifth, the comprehensive assessment of systemic toxicity, off-target embolization, or extra-hepatic organ effects was not performed in this study. We clarify that future work will incorporate dedicated safety and toxicity assessments, including serial blood chemistry, histopathology, and imaging to monitor embolization patterns, to fully characterize the risk profile of TE administration.

CONCLUSION

Infusion of TE combined with doxorubicin through the hepatic arteries significantly reduced VX2 liver tumor volumes in rabbits on post-TACE CT scans, showing a clear difference compared with the NS control group. Tumor growth ratios measured from pre- to post-TACE CT were also markedly lower in the TE group than in the NS group. These findings indicate that TE enhances vascular permeability and facilitates greater uptake of doxorubicin in the VX2 liver tumor model. Such results provide valuable insights into the relationship between vascular permeability and liver tumors, and may guide future investigations into the therapeutic effects and optimal dosing of TE as an adjuvant chemotherapy strategy for liver cancer treatment.