Treatment of Hepatocellular Carcinoma: Value of Percutaneous

Transcription

1 1159 Ryuji Murakami1 2 Shunji Yasuyuki Yamashita2 Tetsuya Matsukawa2 Mutsumasa Takahashi2 Katsuro Sagara3 Received June 1, 1 994; accepted after revision December 5, Department of Radiology, Kumamoto Regional Medical Center, Honjo, Kumamoto 860, Japan. 2Department of Radiology, Kumamoto University School of Medicine, Honjo, Kumamoto 860, Japan. Address correspondence to A. Murakami. 3Department of Internal Medicine, Kumamoto Aegional Medical Center, Kumamoto 860, Japan. 036i-803X/95/i645-i159 American Roentgen Ray Society Treatment of Hepatocellular Carcinoma: Value of Percutaneous Microwave Coagulation OBJECTIVE. Percutaneous microwave coagulation therapy (PMCT) is a new therapeutic technique for the treatment of solid neoplasms that uses an energy source dlfferent from those of other interstitial therapies. We report our initial experience using PMCT to treat hepatocellular carcinomas. MATERIALS AND METHODS. Nine hepatocellular carcinomas exceeding 3 cm In diameter In nine patients were treated with PMCT. Within 2 weeks before PMCT, all patients had been treated with transcatheter arterial embolization therapy, which had failed to produce complete necrosis of the tumors. PMCT was done under local anesthesia. A 14-gauge guiding needle was inserted percutaneously toward the lesion under sonographlc guidance, and a needle electrode was positioned precisely within the lesion. Microwaves of 2450 MHz in frequency were produced for 60 sec with a 60- W emission. Three to 12 microwave emissions were administered In each case. RESULTS. DynamIc CT showed that unenhanced areas Indicative of coagulation necrosis developed in all lesions. All lesions appeared smaller without enhancement: on CT, the tumor diameters (mean ± SD) were 48 ± 13 mm before treatment and 41 ± 13 mm I month after treatment. Follow-up studies showed that five lesions were controlled without any signs of recurrence. All patients tolerated the treatments well, and no serious complications occurred. CONCLUSION. Our preliminary experience suggests that PMCT may be a useful alternative to other forms of interstitial therapy for the treatment of hepatocellular carcinomas. AJR 1995;164: Curative local treatment of hepatocellular carcinomas has been accomplished by surgical resection. Unfortunately, most patients with hepatocellular carcinomas have underlying hepatic cirrhosis, and the tumors frequently recur in other portions of the liver. In addition, the morbidity and mortality associated with surgical resection are relatively high [1, 2]. Transcatheter arterial embolization therapy (TAE) with iodized oil has contributed to the improvement in therapeutic results [3, 4]. However, the therapeutic effects of TAE with iodized oil cannot be expected for tumors that do not accumulate iodized oil [5]. The local effects of sonographically guided percutaneous ethanol injection therapy (PElT) have been reported for the treatment of small tumors [6, 7]. With PElT, however, it is sometimes difficult to achieve complete necrosis of tumors exceeding 3 cm in diameter because of nonhomogeneous distribution of the ethanol. Although other interstitial techniques, such as laser therapy, cryotherapy, and interstitial radiation therapy, have been proposed [8-13], all these techniques are still under investigation. Microwave therapy has been considered an alternative form of percutaneous treatment for hepatic tumors [14]. Microwaves produce tissue necrosis by a mechanism different from those of other forms of interstitial therapy [15, 16]. The tissue around the electrode is coagulated in a spindle shape at the level of the monopolar electrode, but not beyond (Fig. 1).

2 1160 MURAKAMI ET AL. AJA:164, May 1995 Fig. 1 -Appearance of cut surface of pig liver after microwave coagulation therapy. Microwave coagulation occurred from base of monopolar electrode (arrow). Tissue around monopolar electrode is coagulated in spindle shape (arrowheads). Tissue deeper than tip of monopolar electrode (asterisk) is not coagulated. We undertook a preliminary study to evaluate the efficacy and safety of percutaneous microwave coagulation therapy (PMCT) for the treatment of hepatocellular carcinomas unresponsive to TAE. Materials and Methods PMCT was done in nine patients (eight men and one woman, years old [mean = 65 years]) with hepatocellular carcinomas exceeding 3 cm in diameter. The lesions were detectable with sonography and measured mm (mean = 48 mm) in diameter. Before treatment, histopathologic diagnosis of hepatocellular carcinoma was made by needle biopsy. The inclusion criterion for this preliminary study was the presence of hepatocellular carcinomas that were inoperable, detectable with sonography, and uncontrollable by TAE. All hepatocellular carcinomas were inoperable not only because of severe hepatic dysfunction but also because of multifocal disease. All patients had two to six (mean = three) hepatocellular carcinomas. The largest tumors were selected for treatment. This technique had been approved by the institutional review board, and informed consent was obtained from all patients. Within 2 weeks before PMCT, all patients were treated by TAE with an emulsion of doxorubicin hydrochloride (Adriamycin; Adria, Dublin, OH) in 3-5 ml of iodized oil (Lipiodol; Andre Guerbet, Aulnay-sous- Bois, France) and then with gelatin sponge particles (Gelfoam; Upjohn, Kalamazoo, MI). All patients were studied with both CT scanning and sonography within 1 week before microwave therapy to evaluate the therapeutic effects of embolization. Compared with tumor size before TAE, tumor shrinkage was not seen on either CT scans or sonograms, and the accumulation of iodized oil within the tumors was poor on CT scans. Furthermore, needle biopsy after TAE proved the presence of viable tumor cells in all lesions. For six patients, TAE was repeated two to four times but was ineffective. PElT was not indicated because the tumors exceeded 3 cm in diameter. The microwave tissue coagulator (Microtaze; Heiwa, Osaka, Japan), which emits a 2450-MHz microwave, was used in combination with a specially designed needle electrode (stereotactic percutaneous coagulation electrode; Heiwa) 1 6 mm in diameter and 25 cm long. The needle electrode consists of a stainless steel needle (external electrode) with a 1-cm-long monopolar electrode (internal electrode) at the tip (Fig. 2). Microwaves are emitted between the external and internal electrodes, which resemble magnetic dipoles; therefore, microwave coagulation proceeds from the base of the monopolar electrode. PMCT was done under continuous sonographic guidance (Fig. 3). Real-time linear-array scanners (SSD- 650CL; Aloka, Tokyo, Japan) were used with 3.5-MHz transducers (UST or UST-5037P-3.5; Aloka) for monitoring. Local anesthesia was administered at the site of needle insertion. A 14-gauge, 17-cm-long guiding needle (Quickcut-C2; Hakko, Tokyo, Japan) was inserted percutaneously toward the lesion via either a subcostal or an intercostal approach, and a needle electrode was inserted through this guiding needle. The monopolar electrode was precisely placed within the lesion. On the basis of the energy limit of the needle electrode and patients tolerance as well as previous experimental results [14], microwaves were produced for 60 sec with an emission of 60 W. After completion of microwave coagulation of the tumor, microwave emissions of 60 W were added step by step up to the liver surface dunng withdrawal of the needle electrode for about 10 sec to prevent bleeding, bile leakage, and malignant seeding. The guiding needle coupled with the needle electrode was withdrawn through the needle track in the liver. Immediately after microwave emission, characteristic hyperechogenicity developed on the real-time sonogram (see below). MultipIe punctures and multiple microwave emissions in various parts of the tumor were required until the hyperechogenic change covered the entire tumor. One to three punctures were done at each sitting, and two emissions at the deep and superficial portions were done for some punctures. The procedure was done twice within 1 week. Three to 12 microwave emissions were administered., y y Fig. 2.-Specially designed needle electrode 1.6 mm in diameter and 25 cm long. Miionopolar electrode (arrow) is attached to tip of stainless steel needle (arrowheads), which acts as external electrode. Fig. 3.-Clinicalappilcation of percutaneous microwavecoagulation therepy. Under continuous sonographlc guidance, guiding needle (arrows) is Inserted percutaneously toward lesion, and needle electrode (arrowheads) connected to coaxial cable(asterlsks) is inserted through this guiding needle.

3 AJA:164, May 1995 PERCUTANEOUS MICROWAVE COAGULATION FOR CARCINOMAS 1161 Response to treatment was evaluated with dynamic contrastenhanced CT (TCT900S scanner; Toshiba, Tokyo, Japan), which was performed with an IV bolus injection of 100 ml of 300-mg/mI iopamidol (lopamiron; Schering, Berlin, Germany). Dynamic CT studies were obtained a few days and 1 month after treatment and then every 1-3 months. We defined complete tumor necrosis on CT after therapy not only as tumor shrinkage but also as the complete disappearance of tumor vasculanty. On follow-up studies, the criterion for local recurrence was the reappearance of enhancement with or without tumor enlargement. All CT studies (before and after treatment) were prospectively analyzed with discussion by three radiologists without the knowledge of laboratory studies or clinical history. A biopsy of the treated area was done for five patients. Serum cz-fetoprotein levels were measured for the evaluation of treatment effect. Evaluation of the immediate and long-term safety of this technique included patients complaints, physical examinations, and CT studies before and after treatment. Serum biochemistry evaluations (electrolyte, urea, transaminase, and bilirubin levels) and hematologic evaluations (hemoglobin level, platelet count, and prothrombin time) were also done before and after treatment. Results In all nine patients, placement of the needle electrode within the tumor was accomplished easily. On sonograms obtained before PMCT, four lesions were of low echogenicity, three were highly echogenic, and two had mixed echogenicity. During the procedure, real-time sonography revealed expansion of the highly echogenic area with strong echogenic foci around the tip of the needle electrode (Fig. 4). These strong echogenic foci often were accompanied by acoustic shadows. However, the sonographic appearances of the treated tumors reverted to the pretreatment appearances on the following day. No bleeding occurred at the puncture sites in any of the patients. Four patients experienced slight pain in the right hypochondrial region that lasted for a few hours, and three patients had mild fevers (37-38#{176}C)that lasted for a few days. Pain tended to occur in patients with relatively large lesions : r I I -.a L-. - v. - : : m; Two patients experienced both pain and fever. In one patient, whose lesion was located near the surface of the liver adjacent to the diaphragm, ascites and pleural effusion developed after treatment. However, none of these signs or symptoms required any treatment, and no serious complications occurred. No significant changes were seen in the results of laboratory studies after PMCT. On dynamic CT scans obtained before treatment, all lesions appeared as well-enhanced areas in the early vascular phase and as low-attenuation areas in the parenchymal phase. After treatment, unenhanced areas indicative of coagulation necrosis developed, and all lesions appeared smaller without enhancement: the tumor diameters (mean ± SD) on CT scans were 48 ± 1 3 mm before treatment and 41 ± 13 mm 1 month after treatment. In summary, complete necrosis was accomplished in all lesions. A biopsy of the treated area in five patients revealed complete tumor necrosis. Serum a-fetoprotein levels were elevated (21 69 ± 2239 ng/mi) before PMCT and reduced (626 ± 448 ng/ml) after PMCT in five patients. In the other four patients, a-fetoprotein levels were within the normal range before treatment and could not be used as a tumor marker. During the follow-up period, which ranged from 4 to 9 months (mean, 6 months), the maximal reduction in tumor diameter was 30 ± 18 mm, and radiologic evidence of complete necrosis persisted in five lesions (Fig. 5). However, the remaining four lesions, which were poorly differentiated and! or were larger hepatocellular carcinomas, had recurred. Two of the four patients died 4 and 6 months after treatment as a result of the progress of massive intrahepatic tumors. Discussion PMCT is a new form of local thermotherapy. Tissue is coagulated by microwaves emitted from a needle electrode that is inserted directly into tumors. Microwave energy, a part of the : -.-;. - i--- : , -._:$t,4,. -. y : ::.* Fig year-old man with hepatocellular carcinoma treated by percutaneous microwave coagulation. Patient had been treated four times previously with arterial embolizatlon therapy with iodized oil. A, Before treatment, sonogram of liver shows hyperechoic mass 67 mm in diameter in right hepatic lobe (arrows). B, During microwave emission, sonogram shows echogenic area expending around monopolar electrode (arrowheads). C, Immediately after treatment with six microwave emissions, sonogram shows expansion of hyperechogenic area with strong echogenic foci. It is difficult to evaluate deeper portion of lesion.

4 1162 MURAKAMI ET AL. AJA:164, May 1995 C. - :: #{149}---:.. :.j. % D Fig year-oid man with hepatocellular carcinoma treated with percutaneous microwave coagulatlon. This patient had beentreated once before with aded embofizadon therapy with Iodized oil. A, Dynamic CT scan obtained before embollzatlon therapy shows well-enhanced nodule(armwheads). B, Unenhanced CT scan obtamed 7 days after embollzatlon therapy shows Incomplete accumulatlon of Iodized oil. C, Sonogram before treatment shows hyperechoic nodule 35 mm in diameter In anterior segment of right hepatic lobe (arrows). 0, Sonogram immediately after treatment (two emissions) shows marlcedly echogenic area in region corresponding to nodule seen on C. E, Dynamic CT scan obtained 4 days after treatment shows unenhanced area within and around tumor (arrows). F DynamIc CT scan obtained 9 months after treatment shows that lesion has been reduced in size, without evidence of viable tumor. electromagnetic spectrum, penetrates a few centimeters into tissue and is converted to heat [1 5]. The basic mechanism of the heat generation involves polarities in water molecules in the tissue. The polarized molecules follow the induction of the electric field of microwaves. As the electric field is altered at ultrahigh speeds (2450 MHz), microwaves generate induced heat. This heat is not emitted externally but is generated in the tissue itself. This mechanism is different from those of other forms of interstitial thermotherapy. Microwaves produce heat and cause tissue coagulation and hemostasis as well [16]. Microwaves for medical use were initially described in the 1970s for external hyperthermia or surgery [16, 17], and

5 AJA:164, May 1995 PERCUTANEOUS MICROWAVE COAGULATION FOR CARCINOMAS 1163 microwave therapy has been applied to several surgical fields to prevent bleeding [18-22]. Recently, microwaves were applied to the surgical treatment of neoplasms; a tumor can be treated in situ with minimal loss of surrounding normal parenchyma by inserting the electrode directly into the lesion [23]. In 1986, Tabuse et al. [24] reported a percutaneous method to prevent malignant seeding, bleeding, and bile leakage in the needle track after liver biopsy. It was suggested that the tumor tissue could be completely necrotized by several approaches with the percutaneous procedure [14, 24]. We have introduced a new percutaneous interstitial technique of microwave coagulation for hepatocellular carcinomas. Hepatocellular carcinomas less than 3 cm in diameter are considered good candidates for percutaneous injection of ethanol [6, 7]. However, complete necrosis often is difficult to accomplish with ethanol injection, especially for fibrotic tumors. The ethanol is not dispersed evenly throughout the tumor but flows along interstices in the tissue. However, the induced heat of microwaves produces tissue coagulation independent of tissue texture, septa, or capsules. If the needie electrode is correctly placed within the lesion, coagulation necrosis invariably occurs. Even metastatic fibrotic tumors may be good candidates for PMCT. Several punctures with the guiding needle and several emissions per puncture are necessary for larger tumors. In experimental work with egg white and with the same parameters as those used in this clinical study, microwaves increased the tissue temperature to more than 43#{176}Cwithin regions 2.5 cm from the tip of the needle electrode [14]. Tumor cells cannot survive this temperature. In experiments with animals bearing a liver tumor and with the same parameters, microwaves produced coagulation necrosis of regions about 2 cm in diameter [14]. The histopathologic findings of coagulation necrosis were nonspecific and indistinguishable from those produced by other interstitial techniques. Immediately after microwave emission, congestion developed around the area of coagulation necrosis. About 3 weeks later, the congestion disappeared, and a fibrotic layer formed around the area of coagulation necrosis [24]. The coagulated tissue finally became a fibrotic scar. A higher energy level and a longer duration of emission may produce a larger area of coagulation necrosis. In this study, microwaves were produced with an emission of 60 W because the energy limit of the needle electrode was 60 W. Once coagulation necrosis occurs, water molecules around the needle electrode escape in steam, and induction heat is not effectively generated. Furthermore, patients may not tolerate emission for a longer duration under local anesthesia. For these theoretical, technical, and clinical reasons, we have selected 60 W for energy and 60 sec for duration. With these parameters, all patients tolerated the treatment well, and all lesions were effectively treated. PMCT is safer and more effective than laser therapy, because microwaves for shorter durations produce larger areas of coagulation necrosis. Moreover, the deeper tissue is not damaged because microwave coagulation is limited to the level of the monopolar electrode (Fig. 1). Evaluation of the effects of PMCT requires special attention. Real-time sonography revealed expansion of the highly echogenic area with strong echogenic foci around the tip of the needle electrode; evaluation of the deep portion of a tumor would be difficult with real-time sonography (Fig. 4). Therefore, all patients were treated with two sittings regardless of tumor size or sonographic appearance, and the second sitting was done on another day. Animal experiments and clinical studies of resected specimens have shown that the hyperechogenic areas that appear during microwave emissions can be interpreted as coagulation necrosis [14, 16, 24]. The strong echogenic foci often accompanied by acoustic shadows probably represent steam and gas produced by microwaves, as with lasers [ii]. Although the lesions appeared as markedly echogenic regions immediately after the procedure, the appearances of necrotic tissue and viable tumor were indistinguishable on the following day. Consequently, the evaluation oftreated tumors requires additional imaging techniques. Dynamic CT clearly showed PMCT as creating well-defined, unenhanced areas. The unenhanced areas represented regions of complete coagulation necrosis, a result that has been proved in clinical studies [14, 23]. Biopsy of the treated area in our study also revealed complete tumor necrosis. For routine clinical practice, we believe that dynamic CT is sufficient for the evaluation of treated areas and that biopsy usually is unnecessary. PMCT was effective for all nine lesions. Although the responses might have been partly attributable to TAE, we believe that tumor shrinkage was caused mainly by PMCT, not only because the accumulation of iodized oil within the tumors was poor on CT scans obtained after TAE [5] but also because needle biopsy after TAE proved the presence of viable tumor cells. All lesions, including recurrent and nonrecurrent lesions, appeared smaller without enhancement and were presumed to have undergone complete necrosis 1 month after treatment. During the follow-up period, however, tumor recurrence was seen in four of nine lesions. The tumors were poorly differentiated and!or were larger hepatocellular carcinomas. The factors used to evaluate recurrences might be histology and size of the tumor rather than the appearance of the tumor on dynamic CT scans. Serious complications were not encountered with this technique in our small series because we used a 14-gauge guiding needle, although potential complications include bleeding, bile leakage, and malignant seeding. The effectiveness of this technique may be attributable to the additional microwave emissions applied to the needle track during withdrawal of the needle electrode [24]. In a patient whose lesion was adjacent to the diaphragm, however, ascites and pleural effusion developed. If the microwave energy stimulated the diaphragm and the pleura, complications such as massive ascites, massive pleural effusion, and/or pneumothorax might have occurred. Other potential problems include vascular occlusion by a clot, bile duct injury by heat, and infection. Although our experience is preliminary, this technique has proved feasible, effective, and safe. In the present study, we used PMCT for relatively large hepatocellular carcinomas. We could extend its use to hepatic tumors, such as small hepatocellular carcinomas and metastatic hepatic tumors. PMCT might be considered a promising local thermotherapy technique for hepatic tumors and could be a safer and more effective alternative to percutaneous injection of ethanol. Further studies will be required to establish the true efficacy, especially long-term results, and safety of this technique.

6 1164 MURAKAMI ET AL. AJR:164, May 1995 REFERENCES i. tin -r Lee C5, Chen KM, Chen CC. Role of surgery in the treatment of pnmarycarcinoma ofthe liver a 31-yearexpenence. BrJSutgl98l;74: Kanematsu T, Matsumata T, Shirabe K, et al. A comparative study of hepatic resection and transcatheter arterial embolization for the treatment of primary hepatocellular carcinoma. Cancer i993;7i:2i8i Takayasu K, Shima Y, Muramatsu Y, et al. Hepatocellular carcinoma: treatment with intraarterial iodized oil with and without chemotherapeutic agents. Radiology 1987;162:345-35i 4. Beers By, Roche A, Cauquil P, Jamart J, Pariente D, Ajavon Y. Transcatheter arterial chemotherapy using doxorubicin, iodized oil and gelfoam embolization in hepatocellular carcinoma. Acta Radioli989;30: Murakami A, Yoshimatsu 5, Yamashita Y, 5agara K, ArakawaA, Takahashi M. Transcatheter hepatic subsegmental arterial chemoembolization thorapy using iodized oil for small hepatocellular carcinomas: correlation between Lipiodol accumulation pattern and local recurrence. Acta Radiol 1994;35: Livraghi T, vettori C. Percutaneous ethanol injection therapy of hepatoma. Cardiovasc Intervent Radiol I 990; 13: Shiina 5, Tagawa K, Niwa Y, et al. Percutaneous ethanol injection therapy for hepatocellular carcinoma: results in 146 patients. AiR 1993;160:i023-i Masters A, 5teger AC, Bown 5G. Role of interstitial therapy in the treatmont of liver cancer. Br J Surg : Nolsee CP, Torp-Pedersen 5, Burcharth F, et al. Interstitial hypertherrnia of colorectal liver metastases with a US-guided Nd-YAG laser with a diffuser tip: a pilot clinical study. Radiology 1993;187: Amin Z, Donald JJ, Masters A, et al. Hepatic rnetastases: interstitial laser photocoagulation with real-time US monitoring and dynamic CT evaluation of treatment. Radiology 1993;i87: ii. Malone DE, Lesiuk L, Brady AP, Wyman DA, Wilson BC. Hepatic interstitial laser photocoagulation: demonstration and possible clinical importance of intravascular gas. Radiology 1994;i93: Onik G, Kane A, Steele G, et al. Monitoring hepatic cryosurgery with sonography. AJR 1986;i47: Dritschilo A, Grant EG, Harter KW, Holt AW, Austgi SN, Rodgers JE. Interstitial radiation therapy for hepatic metastases: sonographic guidance for applicator placement. AJR i986;147: i4. Seki T, Kunieda K, Sato M, et al. Local treatment for large hepatocellular carcinoma: combination therapy with percutaneous microwave coagulation therapy and percutaneous ethanol injection therapy. (In Japanese.) Acta Hepatol Jpn i992;33: i5. Hand JW, Ter-Haar G. Heating techniques in hyperthermia. Br J Radiol 1981;54: Tabuse K. A new operative procedure of hepatic surgery using a microwave tissue coagulator. Arch Jpn Chir i979;48:160-i Homback NB, Shupe A, Shidnia H, et al. Radiation and microwave therapy in the treatment of advanced cancer. Radiology 1 979;i 30: Tabuse K, Katsumi M, Kobayashi Y, et al. Microwave surgery: hepatectomy using a microwave tissue coagulator. World J Surg 1985;9:136-i Zhou XD, Tang ZY, Yu YQ, et al. Microwave surgery in the treatment of hepatocellular carcinoma. Semin Surg Oncoli993;9:i Toy FK, Reed WP, Taykr LS. Expenmentd spleruc preservathn employng microwave surgicaltechniques: a preliminary report. Surgeiyl984;96:i Muraki J, Schwalb DM, Cord J, et al. Application of microwave tissue coagulation in partial nephrectomy. Urology : Tabuse K, Katsumi M, Nagai Y, et al. Microwave tissue coagulation applied clinically in endoscopic surgery. Endoscopy i985;i7:i Saitsu H, Mada Y, Taniwaki 5, et al. Investigation of microwave coagulonecrotic therapy for 21 patients with small hepatocellular carcinoma less than 5 cm in diameter. (In Japanese.) J Jpn Surg Soc i993;94: Tabuse Y, Tabuse K, Mon K, et al. Percutaneous microwave tissue coagulation in liver biopsy: experimental and clinical studies. Arch Jpn Chir i986;55: The reader s attention is directed to the commentary on this article, which appears on the following pages.