Delayed Intratumoral Hemorrhage after Drug-Eluting Bead Transarterial Chemoembolization for Hepatocellular Carcinoma

Three types of calibrated microspheres have recently been approved in Japan as embolic agents for hypervascular tumors and arteriovenous malformations. To date, two types of embolic microspheres capable of being loaded with drug have been introduced: DC Bead® (Biocompatibles, Farnham, UK) and HepaSphere® (Biosphere Medical, Roissy CDG Cedex, France) [1]. In a phase II randomized controlled study of transarterial chemoembolization (TACE) using a drug-eluting bead (DEB-TACE) for hepatocellular carcinoma (HCC), the PRECISION-V study [2], the response rate after 6 months was 52% in the DEB-TACE group and 44% in the TACE group using a Lipiodol® (Laboratoire Guerbet, Aulnay-Sous-Bois, France)/anticancer drug mixture solution (Lip-TACE), showing no significant difference or superiority of DEB-TACE. However, complication by postembolization syndrome, such as abdominal pain, fever, nausea, vomiting and liver enzyme elevation, developed in 20-80% after Lip-TACE [3,4], whereas symptoms developed in only 18-37% after DEB-TACE [5,6], showing that the complications were milder than those after Lip-TACE. On the other hand, DEB-TACE induced liver abscesses at a relatively high frequency, which could be fatal [5,7,8,9]. Adverse events after DEB-TACE have been reported occasionally [5,7,8,9,10,11], but as it has just been introduced in Japan, many points regarding adverse events are unclear with respect to the developmental mechanism and risk factors.

We encountered a HCC patient who underwent DEB-TACE and experienced delayed intratumoral hemorrhage. Herein, we report on this case and present a review of the literature.

An 81-year-old male regularly visited his physician for treatment of hepatitis C virus-related cirrhosis. On routine abdominal ultrasonography (US), a mass was detected in S5 of the liver. On esophagogastroduodenoscopy, a type 0-IIa early gastric carcinoma with a diameter of 25 mm was detected in the lesser curvature of the middle body. The patient was referred to our hospital for further examination and treatment. During the first examination, the liver function was classified as Child-Pugh class A, and the tumor markers were normal (AFP level: 3.4 ng/ml, AFP-L3 fraction level <0.5%), but the PIVKA II had a level of 118 mAU/ml, showing a mild elevation. The mass in S5 of the liver was 35 mm in diameter. It was capsulated and enhanced in the arterial phase on dynamic computed tomography (CT) (fig. 1), and showed a mosaic pattern accompanied by a marginal hypoechoic zone on gray-scale US. The mass was enhanced in the vascular phase (0-40 s) on contrast-enhanced US (CEUS) using Sonazoid® (Daiichi Sankyo, Tokyo, Japan), and showed a hypoechoic area in the postvascular phase (10 min after an intravenous injection of Sonazoid). Based on the tumor marker and imaging findings, the patient was diagnosed with HCC. DEB-TACE was performed because the patient declined to undergo surgical resection of HCC.

Access for TACE was performed under sterile conditions and under local anesthesia, via the right femoral artery using a 3-Fr sheath (Medikit, Tokyo, Japan) and in a retrograde fashion. An intensely stained tumor was detected in the S5 region on celiac arterial angiography. When a microcatheter was advanced to the anterior segmental artery of the right hepatic artery and contrast imaging was applied, an intensely stained tumor was observed, with A5 as the feeding artery. A solution impregnated with 100-300 μm DC Bead and 50 mg epirubicin hydrochloride, with a DC Bead volume of 0.35 ml, was administered via the feeding artery to perform DEB-TACE. Disappearance of the intensely stained tumor image was confirmed by right hepatic arterial angiography, and the treatment was completed. The course after DEB-TACE was favorable, and no adverse events above grade 3 according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.0 were observed. Endoscopic submucosal dissection (ESD) of the early gastric carcinoma was performed about 1 month after DEB-TACE. Right hypochondriac pain suddenly developed 3 days after ESD, but no adverse events were assumed to be caused by ESD (i.e., free air, noted on dynamic CT). However, the tumor diameter had increased from that before DEB-TACE, and the tumor showed a high-intensity area on unenhanced CT (fig. 2a), which was not enhanced in the arterial phase on dynamic CT, suggesting intratumoral hemorrhage. The hemodynamics of the tumor were followed using CEUS over the time before and after DEB-TACE. On CEUS carried out on the day after DEB-TACE, the whole intratumoral enhancement decreased, and nonenhanced patchy regions, assumed to be necrosis, were noted in the tumor in the vascular phase. On CEUS performed 4 weeks after DEB-TACE, i.e., immediately before ESD, the tumor size was 33 × 31 mm, and the enhancement area was increased in the vascular phase (fig. 3) compared with the CEUS imaging of the day after DEB-TACE. However, the tumor had enlarged to 41 × 36 mm on CEUS when the right hypochondriac pain developed (5 weeks after DEB-TACE), and changes in the echogenicity on gray-scale US and nonenhancement of almost the entire tumor in the vascular phase on CEUS were noted (fig. 2b). Based on the above findings, the cause of right hypochondriac pain may have been DEB-TACE-associated intratumoral hemorrhage. Since the hemorrhage was limited to inside the tumor, it stopped spontaneously with rest.

The advantages of DEB-TACE are as follows: the infusion dose of the chemotherapeutic agent can be modified by adjusting the chemotherapeutic agent carrier, while altering the depth of embolization with beads, the duration of tumor exposure to chemotherapeutic agent can be fine-tuned through sustained release of the chemotherapeutic agent, and systemic adverse reactions of chemotherapeutic agent can be attenuated by continuous release of high-dose chemotherapeutic agents into the tumors [12]. DEB has proven to be extremely potent and safe by allowing large amounts of the chemotherapeutic agent to concentrate within the tumor over time, thereby maximizing the cytotoxic effect, which also makes it attractive as an effective treatment for large lesions [11]. Favorable therapeutic effects have been reported occasionally [8,9] and have increased the expectations of HCC treatment with DEB-TACE; however, the beads exhibit characteristic physical properties and behavior that are different from those of gelatin particles. It is important to fully investigate these characteristics to avoid inadequacy of the therapeutic effect and to prevent the onset of adverse events.

Previous studies have reported that the incidence of mild adverse events, such as postembolization syndrome, is lower in the DEB-TACE group compared with the Lip-TACE group [3,4,5,6]. Meanwhile, liver abscesses were induced at a relatively high frequency (about 1.6-7.4%) [5,7,8,9]. Malagari et al. [7] reported that the 30-day mortality was 1.26% among 237 HCC patients treated with DEB-TACE with DC Bead loaded with doxorubicin, and that the cause was liver abscess in all cases. Regarding other adverse events, acute pancreatitis and cholecystitis developed with incidences of 1-5 and 1-6.3%, respectively. Previous studies have shown that rare adverse events included liver failure and infection, bile duct injury, upper gastrointestinal bleeding, pulmonary embolism, splenic infarction, or spinal embolization [1,7,10,11]. Unlike Lip-TACE, many adverse events occurred due to ischemic changes induced by the inflow of beads into organs other than the HCC lesion, which is characteristic of DEB-TACE.

Tumor hemorrhage has also been reported after DEB-TACE [1,7,10,11,13] with large tumors, with a diameter of 5-16 cm, and the tumor locations were subcapsular in all previous reports. To our knowledge, there has been no case of a tumor with a diameter <5 cm distinct from the subcapsular, as was observed in our patient.

There has been no report of a mechanism underlying the DEB-TACE-specific tumor hemorrhage. In general, the mechanism of tumor rupture after TACE might be related to increased intratumoral pressure as a result of rapid edematic expansion due to tumor necrosis or vascular injury, secondary to embolization [14]. The risk is particularly high in subcapsular lesions, and we must be careful because the spread of hemorrhage in the abdominal cavity cannot only lead to hemorrhagic shock, but also peritoneal metastasis. Unlike the reported cases of tumor hemorrhage after DEB-TACE, in the present case, the tumor diameter was not large and the tumor was distinct from the subcapsular region. However, intratumoral hemorrhage occurred about 1 month after DEB-TACE. Local pooling of the contrast agent within the target tumor could occur during DEB-TACE, and Seki et al. [10] reported that it occurred in 25.2% of 135 HCC patients treated with DEB-TACE (epirubicin-loaded HepaSphere). While the cause of this local pooling remains unclear, it is similar to extravasation or blood sinus [15], suggesting that it was a local hemorrhage due to intratumor vascular collapse as a consequence of the treatment. Injected beads can easily escape into the pooling space. When local pooling of the contrast agent was noted, Seki et al. [10] added a small amount of gelatin sponge particles, with a diameter ranging from 1 to 2 mm, to the feeding artery until the pooling disappeared. Based on the above findings, the observation of local pooling of the contrast agent within the target tumor during treatment suggests that intratumoral pressure has already increased, which was unlikely in our patient. Treatment-induced intratumoral hemorrhage continued immediately after DEB-TACE because no pooling was noted during treatment or on confirmation angiography following treatment. We considered that embolization was sufficient during treatment. However, it is possible that the amount of DC Bead was insufficient and embolization was incomplete. It was assumed that incomplete embolization subsequently caused intratumoral vascular collapse, i.e., a phenomenon similar to pooling, and that the hemorrhage expanded to the necrotic intratumoral area by sustained release of the chemotherapeutic agent and embolization, leading to complication of intratumoral hemorrhage despite the low risk of tumor hemorrhage.

DEB-TACE has just been introduced in Japan. Many points remain unclear, i.e., the appropriate bead sizes for each case and the level of embolization, but it may be important to apply sufficient embolization because insufficient embolization is likely to lead to tumor hemorrhage. DEB-TACE is considered to rarely cause adverse events, but fatal complications have been reported [5,7,8,9], and reports of adverse events may increase with augmentation of the numbers of DEB-TACE-treated patients. If reports of tumor hemorrhage after DEB-TACE increase, the mechanism of intratumoral hemorrhage in cases similar to our patient may be elucidated.

Intratumoral hemorrhage was suspected due to the presence of a high-intensity area in the tumor on unenhanced CT and a rapid increase in the tumor diameter and accompanying enlargement of the unenhanced area on CEUS performed before and after DEB-TACE. Based on these findings, intratumoral hemorrhage was diagnosed. It was difficult to assess the therapeutic effect of DEB-TACE by CT, based on Lipiodol accumulation, unlike that after conventional Lip-TACE, and evaluation of the intratumoral hemodynamics could be important. CEUS is superior in evaluating intratumoral hemodynamics in real time, and the utility of CEUS to evaluate the therapeutic effect after DEB-TACE was suggested.

We encountered a patient who experienced intratumoral hemorrhage about 1 month after DEB-TACE. DEB-TACE for HCC is a new treatment method in Japan, and reports of adverse events may increase in the future. To obtain the therapeutic effect of DEB-TACE while preventing the adverse events, it may be important to understand the characteristics of the beads themselves and to apply the appropriate embolization to each individual case. It was also suggested that CEUS is useful to evaluate the therapeutic effect after DEB-TACE.