Balloon-Occluded Transcatheter Arterial Chemoembolization (b-TACE) for Hepatocellular Carcinoma Performed with Polyethylene-Glycol Epirubicin-Loaded Drug-Eluting Embolics: Safety and Preliminary Results

Transarterial therapies represent the standard of care for intermediate hepatocellular carcinoma (HCC) not amenable to curative treatments and an option in early stage if surgery/ablation was contraindicated [1]. Balloon-assisted TACE (b-TACE) that is a TACE procedure performed with a balloon micro-catheter inflated within the lesion’s arterial feeders prior to the embolization has been recently developed in Japan by Irie et al. [2]. Balloon micro-catheter intervention works by way of blood flow modification, through a drop of intra-arterial pressure that provides a re-distribution of the flow towards areas of less resistance, in this case tumour lesions [3]. Safety and efficacy of the b-TACE procedure with lipiodol in combination with anticancer drugs have been recently confirmed by some reports [4‐6]. Several retrospective studies comparing oncological results on patients treated with b-TACE with lipiodol, with those of an historical matched cohort treated with c-TACE, demonstrated that b-TACE may obtain better tumour control over c-TACE in tumours up to 4 cm [4, 7, 8].

Notwithstanding these advantages, it is to note that lipiodol conventional TACE (c-TACE) has shown several limitations (procedure standardization, toxicity profile, pain), overcome by the introduction of drug-eluting microsphere transarterial chemoembolization (DEM-TACE) [9‐11].

On the basis of those results, the combination of DEM-TACE with the b-TACE may improve the oncological response in HCC. The choice of a different delivery system will provide the Interventional Oncologist with another opportunity to further address the specific needs of every single patient, in line with the modern concept of individualized medicine. To date, the combined use of these two techniques has not been reported.

The aim of this study is to describe our experience with b-TACE performed with polyethylene-glycol epirubicin-loaded drug-eluting embolics in HCC patients and report about the safety and preliminary oncological results at 3- and 6-month follow-ups.

This study was approved by the ethical institutional review board. Informed consent for the procedure for anonymized publication of this series of patients was obtained from all individual participants included in the study.

Between January 2018 and May 2018, we treated 22 consecutive patients (mean age 67.1 ± 14.0 years [range, 41–86 years]; 20 males) with 29 hepatocellular carcinoma (HCC) (average 1.3 HCC nodule/patient) [12, 13]. All patients were evaluated by a multidisciplinary board (composed of a transplant surgeon, an interventional radiologist, body radiologist and a hepatologist).

Inclusion criteria were: Child–Pugh score up to B8 and Barcelona Clinic Liver Cancer (BCLC) stage up to B and not eligible for curative treatments (surgical resection or percutaneous ablative treatments). Patients presenting with Child–Pugh > B8, BCLC stage C, portal vein thrombosis (defined as the complete or partial obstruction of blood flow in the portal vein, due to the presence of a chronic, acute or neoplastic thrombus in the vasal lumen) [14], extrahepatic metastasis, and high-flow arterioportal or arteriovenous shunts, previous systemic treatment, platelet count < 50,000, and bilirubin level > 3 mg/dL, were not considered suitable for the procedure.

Diagnosis of HCC was performed according to the American Association for the Study of Liver Disease guidelines [13], using multidetector computed tomography (MDCT) (14/22 patients [64%]) or contrast-enhanced magnetic resonance imaging (CE-MRI) (8/22 patients [36%]).

Clinical and demographics characteristics of the cohort as well as the indications for treatment are shown in Table 1.

Twenty-four b-TACE procedures were performed in 22 patients with 29 HCC. Patients’ indication for b-TACE was down-staging to liver transplantation in 7/22, bridging to liver transplantation in 4/22 and tumour palliation in 11/22. Two patients received a second b-TACE for treating a different HCC lesion fed by a different arterial feeder, whereas in 2 cases a second b-TACE (re-b-TACE) was performed for treating the residual vital tumour, either within the same catheter positioning or within another feeder. One patient received radiofrequency ablation (RFTA) of the residual viable tumour after 1-month follow-up. (Details on cohort follow-up are shown in Fig. 1). All these procedures were performed within 30 days of the CT scan performed 30 days after the first procedures.

All 22 patients fulfil the 1-month follow-up. Of 22 patients, 14 (63.6%) performed the 3–6-month follow-up, particularly 6/8 (75%) did not perform the second line imaging modality yet, 1/8 (12%) follow-up not performed for worsening of clinical condition, and 1/8 (12%) underwent to RFTA (Fig. 1).

The technical success of the procedure was 100%. Planned location within the feeding artery was reached in all cases. (The selected arteries expressed as the selectivity score are summarized in Table 2).

The BOASP with the micro-balloon inflated was 64.0 ± 15.0 mmHg (min 46; max 110 mmHg), while prior to inflation it was 115.4 ± 29.5 mmHg; therefore, the pressure drop average was 51.5 ± 21.5 mmHg. An inverse correlation was found between BOASP drop and tumour volume reduction (r = − 0.45, P < 0.05), and this inverse correlation remained significant considering the pre-procedural diameter of the tumour (r = − 0.45, P = 0.04).

The subgroups (diameter > 3 cm or < 3 cm) correlational analysis showed no differences between the two curves (r < 3 cm = − 0.05 and r > 3 cm = − 0.46; P > 0.05) as summarized in Fig. 2. DEM deposition showed qTCR in 14/24 (58.3%); qDCR was observed in 10/24 (41.7%), qDCRa in 2/10 (20.0%), and qDCRb in 8/10 (80.0%). In 13/24 (54.2%) of the procedures, we observed an extra-nodule DEM deposition, among these 1/13 (7.7%) was minimal (grade 1), 8/13 (61.5%) was mild (grade 2), and 4/13 (30.8%) was diffuse (grade 3).

The median epirubicin dose delivered was 62.5 mg [95% CI 55.0–100.0], the median percentage of 100 ± 25 µm microsphere administered was 100% [95% CI 100.0–100.0] (21/24 [88%] completed vial of 100 ± 25 µm), while the median percentage of 200 ± 50 µm infused was 22.5% [95% CI 7.5–100.0] (8/24 [33%] completed vial of 200 ± 50 µm).

Four AEs (15%) were reported: one pseudo-aneurysm of the inflated artery (maximum diameter 5 mm; complication grade 3), one liver abscess (maximum diameter 42 mm; complication grade 2) and two ALI [asymptomatic segmentary biliary tree dilatation (complication grade 2)].

The pseudo-aneurysm was treated using detachable coils (Concerto™, Medtronic, USA), without sequelae. The liver abscess was treated using antibiotic therapy only. The ALI was observed in a patient after a re-b-TACE.

After 8/24 (33.3%) procedure, patients experienced PES. In detail, 5/24 (20.8%) had abdominal pain (VAS score 6.7 ± 1.2), 5/24 (20.8%) had nausea, and 1/24 (4.2%) had low-grade fever.

Statistically significant increase of laboratory test was observed for ALT, AST, direct bilirubin, neutrophil percentage and white blood cells (for details, see Table 3). All changes were grade 1 according CTCAEv5. On the other hand, ALP and total bilirubin significantly decreased when comparing pre- and post-procedure laboratory results.

Per-patient analysis, at 1 month: in 7/22 patients (31.8%), there was a complete response (CR), and in 13/22 patients (59.1%), there was a partial response (PR) for a total objective response (OR; [CR + PR]) of 90.9% (20/22 patients). In two patients (2/22; 9.1%), we observed progressive disease (PD), due to the occurrence of a new tumour in another liver segment.

Per-procedure analysis, at 1 month: 10/24 (41.7%) were on CR, and 14/24 (58.3%) were on PR for an OR of 100% (24/24).

Among patients classified as partial responders, the average percentage of reduction in tumour volume was 64.9 ± 27.3% with an absolute tumour reduction volume median of 6.5 cm³ (95% CI 1.8 to 13.3). The baseline tumour volume was 17.7 cm³ (95% CI 8.3 to 38.2), and this value decreased to 1.2 cm³ (95% CI 0.0 to 8.8) at 1 month (< 0.05) (Figs. 3, 4).

Per-patients analysis at 3–6 months showed a CR of 14.3% (2 out of 14 patients) and a PR 42.9% (6 out 14) leading to an OR of 57.2% (8/14). The PD rate was 42.8% (6 out of 14), of these, 2/6 (33.3%) due to a new tumour in another segment.

Per-procedure analysis, at 3–6 months: 9/17 (52.9%) were CR, 4/17 (23.6%) were PR, and 4/17 (23.5%) were PD. The OR rate was 76.5% (13/17).

This preliminary study shows that b-TACE performed in conjunction with epirubicin-loaded DEM is a safe procedure and is associated with a high overall tumour response.

Despite the initial complexity added to the procedure by the presence of the balloon micro-catheter and the different catheter manipulation technique, the operators were able to position and inflate the balloon micro-catheter in the selected vascular segment in all cases. This is of note, considering that 70% of balloon placements were performed in a second-order hepatic artery branch or in a more distal branch. All catheterization procedures were carried out without observing a vascular complication that could stop the embolization procedure, such as vasospasm or a dissection. In all cases, the balloon micro-catheter visibility was good enough to perform procedures.

Our data on the BOASP reduction (63.2 ± 14.7 mmHg) are in line with those reported by Irie et al. [2] that identified a better lipiodol dense accumulation with a BOASP lower than 64 mmHg. In our study, we observed only one case of suboptimal reduction of the BOASP (prior to inflation it was 210 mmHg; after inflation it was 110 mmHg). This was probably due to the vascular anatomy of this patient which had previously undergone a right hepatectomy that limited the correct placement of the balloon micro-catheter proximal to all HCC feeders. This patient, as expected, experienced a partial response to the b-TACE procedure and required a second treatment.

Regarding the site of balloon micro-catheter placement, inflation and subsequent embolization, it is worth to underline that b-TACE was performed differently from standard DEM-TACE procedures. In fact, while DEM-TACE procedures are usually performed by embolizing “as super-selective as possible within all lesion’s feeders” (scores 0, 1), b-TACE in this study was performed more proximal (“as selective as possible proximal to all lesion’s feeders”) by placing the balloon micro-catheter proximal to all lesions’ feeders. This was done with the rationale that balloon micro-catheter inflation within the arterial segment proximal to the lesions’ feeders may determine a drop of the “vis a tergo”, confirmed by the drop of the BOASP, permitting flow redirection towards a low-resistance territory such as a hypervascular HCC [21]. The theoretical background underlying this procedural decision was to maximize the DEM deposition inside the tumour. This hypothesis is confirmed by the analysis performed at post-procedural non-enhanced CBCT that demonstrated presence of DEM within the target lesion. Moreover, despite that extra-nodular DEM deposition was seen in 54.1% of procedures, this was categorized as diffuse only in 4 cases and was not associated to post-procedural complications.

Of the four significant AEs, only the pseudo-aneurysm of the arterial feeder was strictly related to the use of the balloon micro-catheter. Pseudo-aneurysm formation has been reported and is correlated with the learning curve in the correct use of this device [22]. This episode occurred in fact within the first five cases of our experience. The remaining AEs represent a known complication of both the DEM-TACE procedures and b-TACE performed with lipiodol [15, 23].

Several aspects should be kept in mind when dilating the balloon micro-catheter: first, despite its compliant structure (the balloon is made of polyurethane), over-dilation may result in a vascular damage of small arteries; second, flushing the dead space of the micro-catheter balloon’s line is crucial because a suboptimal flushing may impair the correct visualization of the balloon inflated to its nominal diameter; third, during micro-balloon inflation continuous BOASP monitoring is suggested because pressure drop may occur even before the required balloon diameter is reached. This third point is of particular relevance because often there is no perfect match between the recommended injected volume and the actual calibre of the inflated micro-balloon. In our opinion, inflation should be always performed under fluoroscopy as the micro-balloon needs an appropriate time (longer than a standard angioplasty balloon) to reach the required diameter.

Concerning laboratory findings, all post-procedural modifications were grade 1 according to CTCAEv5.0, despite the high degree of tumour devascularization (average debulking percentage 61.4 ± 31.2%) and the fact that the embolization was segmental or even more proximal in 14/24 procedures (58.3%).

With the exception of abdominal pain (20.1% vs. 14% [23]), each aspect of PES had a lower incidence, as compared with the incidence reported in the existing literature, in particular fever was 4.0% versus 78% and 68% [23], and nausea was 20.1% versus 28% [24].

Due to the lack of data on the b-TACE performed with DEM, the comparison with the existing literature is influenced by several causes of bias: use of lipiodol, different drugs and population characteristics (e.g. number and dimension of treated tumours).

Despite these limitations, the objective response of this series is 90.9% and 58.3%, respectively, at 1 and 3–6 months which are consistent with the results reported by Hatanaka et al. [6] (OR of 63.6%), using miriplatin with lipiodol, and Kawamura et al. [25], Asayama et al. [26] and Minami et al. [23], reporting an OR of 59.6%, 57.1% and 56.3%, respectively. This consistency is particularly relevant as our cohort of patients presented greater tumour comparing with other studies (max diameter 32.1 mm ± 14.4 [range 12–64 mm] versus 2.0 mm ± 0.9 [23] versus 6.6 mm [range 9–40 mm] [26]). Further studies, with randomized control group, are necessary to deeply investigate this aspect.

Partial correlation analysis results between BOASP during balloon inflation and percentage volume reduction, categorized for tumour < 3 and > 3 cm (r = − 0.05; − 0.46; P > 0.05, respectively), demonstrated a moderate inverse correlation in the subset of patients having tumours > 3 cm as confirmed in Fig. 2. This finding suggests a potential major benefit of the b-TACE procedure in the cohort of patients having tumour > 3 cm.

This is of particular relevance considering the initial indication for b-TACE. Four out of five patients (80%) with a total tumour burden outside the “up-to-seven” criteria for liver transplantation [27] were down-staged and subsequently listed in the active waiting list for liver transplantation. Among these four patients, in three cases a tumour greater than 5 cm was treated with a b-TACE as a standalone treatment which achieved complete necrosis, whereas the fourth patient with a tumour greater than 5 cm, b-TACE achieved effective tumour debulking and allowed the execution of adjunctive RFA of the residual vital tissue.

Our study presents several limitations, first the small sample size sample size (22 patients), second, the non-randomized nature of this series.

In conclusion, our series showed that the combination of b-TACE and DEM was safe and obtained a CR, at 1 and 3–6 months, of 41.7% (10/24) and 52.9% (9/17) and a PR of 58.3% (14/24) and 4/17 (23.5%), respectively.