1694 Percutaneous Microwave Coagulation Therapy for Patients with Small Hepatocellular Carcinoma Comparison with Percutaneous Ethanol Injection Therapy Toshihito Seki, M.D.1 Masayuki Wakabayashi, Taiichi Nakagawa, M.D.1 Masato Imamura, M.D.1 Toru Tamai, M.D.1 Akira Nishimura, M.D.1 Noriyo Yamashiki, M.D.1 Akiharu Okamura, M.D.2 Kyoichi Inoue, M.D.1 BACKGROUND. The authors compared the efficacy of percutaneous microwave 1 M.D. 1 Third Department of Internal Medicine, Kansai Medical University, Moriguchi, Osaka, Japan. 2 Department of Clinical Pathology, Kansai Medical University, Moriguchi, Osaka, Japan. coagulation therapy (PMCT) and percutaneous ethanol injection therapy (PEIT) in the treatment of patients with cirrhosis and a solitary nodular hepatocellular carcinoma (HCC) # 2 cm in greatest dimension. METHODS. Of 43 patients with well-differentiated HCC, 23 were treated with PMCT and 20 with PEIT. Of the 47 patients with moderately or poorly differentiated HCC, 25 were treated with PMCT and 22 with PEIT. In a retrospective, nonrandomized study, the prognoses of 90 patients during the 12–72 months preceding the study were analyzed according to histologic tumor grade. RESULTS. The overall 5-year survival rates for patients with well-differentiated HCC treated with PMCT (70%) and PEIT (78%) were not significantly different. No difference between the patterns of recurrence was observed. Among the patients with moderately or poorly differentiated HCC, overall survival with PMCT (5-year survival rate: 78%) was significantly better than with PEIT (5-year survival rate: 35%) (P 5 0.03). Nine of 22 patients with moderately or poorly differentiated HCC treated with PEIT experienced recurrence in the original target subsegment. Only 2 of 25 patients treated with PMCT had a recurrence in the same subsegment as the initial tumor. CONCLUSIONS. PMCT may be superior to PEIT for the local control of moderately or poorly differentiated small HCC. Cancer 1999;85:1694 –702. © 1999 American Cancer Society. KEYWORDS: microwave coagulation, hepatocellular carcinoma, tumor cell differentiation, ethanol injection, percutaneous local treatment. S Address for reprints: Toshihito Seki, M.D., Third Department of Internal Medicine, Kansai Medical University, 10-15 Fumizonocho, Moriguchi, Osaka 570-8507, Japan. Received June 25, 1998; revision received November 23, 1998; accepted December 7, 1998. © 1999 American Cancer Society urgical resection, transcatheter arterial embolization (TAE), and ultrasound (US)-guided percutaneous ethanol injection therapy (PEIT) are used alone or in combination as the initial treatment for small hepatocellular carcinoma (HCC). However, surgical resection is not a viable option for all patients due to poor liver function induced by chronic liver disease (cirrhosis). Furthermore, TAE is ineffective sometimes because of inadequate angioneogenesis in small HCCs.1 For these reasons, PEIT is widely performed as a local treatment because of its simplicity and minimal invasiveness. It has been reported by many that PEIT is a satisfactory therapeutic modality.2– 8 However, PEIT occasionally has been found to be ineffective when there is intra- or extracapsular invasion, because the injected ethanol is blocked by the fibrous capsule and fails to reach the tumor cells.2,9 In some cases, the injected ethanol may flow into the vessels surrounding the tumor instead of causing tissue necrosis. Microwave Coagulation for Small HCC/Seki et al. To overcome the disadvantages of PEIT, percutaneous microwave coagulation therapy (PMCT) was developed to induce more complete tumor necrosis for the local treatment of HCC. In a previous study, we suggested that PMCT may be an effective percutaneous local treatment for small HCCs.10 However, we had no available data to compare the long term therapeutic efficacy of PMCT and PEIT. In this study, we retrospectively compared PMCT with PEIT in patients with solitary small HCCs measuring 2.0 cm or less in greatest dimension according to the histologic grade of the tumor, because it has been reported that tumor cell differentiation is associated closely with recurrence and prognosis.11–15 1695 lumped together in a single group because of the small numbers of patients with poorly differentiated HCC. In this study, there were no patients with severe bleeding tendency or ascites. Furthermore, tumors located at sites near the liver surface showing extrahepatic extension were excluded because of the difficulty in maintaining the puncture line. Each patient and one family member chose the initial treatment after they were informed of the advantages and disadvantages of both PMCT and PEIT. The purpose of the study and all procedures were explained thoroughly to the patients, and informed consent was obtained from each patient and one family member. PATIENTS AND METHODS Patients Ninety patients with cirrhosis and a solitary nodular HCC measuring # 2.0 cm in greatest dimension who were admitted to Kansai Medical University and its affiliated hospital between September 1990 and March 1997 were qualified to undergo US-guided percutaneous local treatment. They either were not candidates for surgical resection due to impaired liver function or had requested US-guided percutaneous treatment as the initial therapy despite having operable small HCCs. The tumors were all classified as Stage I according to the International Union Against Cancer.16 Cancer staging was performed using ultrasonography, dynamic computed tomography (CT), dynamic magnetic resonance imaging, and digital subtraction angiography. The histologic diagnosis was confirmed by US-guided fine-needle biopsy. The lesions were biopsied twice, and two specimens were obtained from each tumor. These specimens were assessed blindly by two authors independently. The histologic grade of differentiation was defined as follows: well differentiated, corresponding to Edmondson’s Grade I or I–II; moderately differentiated, corresponding to Edmondson’s Grade II or II–III; or poorly differentiated, corresponding to Edmondson’s Grade III or IV.17 In cases in which there was a discrepancy in the differentiation of the two biopsied specimens, indicating heterogeneity in the differentiation of the tumor, the tumor was assigned to the lower degree of differentiation. Of the 43 patients with well differentiated HCC, 23 were treated with PMCT, and 20 were treated with PEIT. Of the 47 patients with moderately or poorly differentiated HCC, 25 (24 moderately differentiated HCCs, 1 poorly differentiated HCC) were treated with PMCT, and 22 (21 moderately differentiated HCCs, 1 poorly differentiated HCC) were treated with PEIT. The moderately and poorly differentiated HCCs were Methods Microwave coagulation system The frequency of the microwave used for our system is 2450 mHz 6 50 mHz, which is internationally recognized as an Industrial, Scientific, and Medical (ISM) band. The microwave is generated by a magnetron in a microwave generator (Microtaze OT-110M; Nippon Syouji Kaisya, Osaka, Japan). The electric energy of the microwave is supplied to a high frequency coaxial cable with an impedance of 50 Ohm through an isolated sympathizer cavity in the generator and then conducted to an electrode attached to another end of the cable for irradiation from the tip to the tissue. For easy manipulation, we used a coaxial cable 2.5 meters long with a diameter of 5.7 mm. The structure of the electrode (MD-20CDL-10/25; Nippon Syouji Kaisya) from the tip to the other end consists of an antenna 10 mm long made of stainless steel, an insulator 2 mm long made of polytetrafluorethlyene, and a coaxial construction 238 mm long, the inner conductive rod of which is made of silver-plated tool steel and is electrically connected to the antenna. The outer conductor of the coaxial construction is made of nickelplated brass and is an outer electrode. The electrode measures 2.0 mm in diameter and 25 cm in length. It is known from the specific absorption rate (SAR) distribution of this kind of heating system that heating is limited to the area adjacent to the antenna. Our electrode generates an electromagnetic field surrounding the insulator, and the heated area of the tissue is limited to and focused on the area adjacent to the tip of the electrode.18,19 To avoid unintended and unexpected irradiation, the generator is equipped with a fail-safe system so that the microwave cannot be irradiated unless both the inner and outer electrodes completely contact the tissue. 1696 CANCER April 15, 1999 / Volume 85 / Number 8 PMCT The procedure of PMCT was carried out as described previously.10 After local anesthesia under US guidance, a guide needle was inserted in the vicinity of the tumor. After the inner needle of the guide was removed, the microwave electrode was inserted through the outer needle of the guide to place the electrode in the tumor area. The electrode was connected to the microwave generator by a flexible coaxial cable. The tumor area was irradiated with microwaves. The electrode and the outer needle of the guide were removed. The puncture track was irradiated with microwaves to prevent bleeding from the hepatic surface when the electrode was removed. In a previous study, we used of a thin microwave electrode measuring 1.6 mm in diameter and 30 cm in length for PMCT. However, this electrode could not produce an extensive necrotic area with one-time microwave irradiation, because the electrode could not withstand high power microwaves, e.g., 80 W, with microwaves at 60 W for 120 seconds using a thin type electrode, the coagulated area elliptic, with maximal and minimal dimensions of 2.4 cm 6 0.4 cm and 1.6 cm 6 0.3 cm, respectively, in normal rabbit liver tissue in vivo.10 Therefore, repeated electrode insertions are required to obtain a sufficiently large treated margin. In this study, to overcome the disadvantages of the thin type electrode, we used a microwave electrode measuring 2.0 mm in diameter and 25 cm in length that could withstand high power microwave irradiation and a 13-gauge guide needle. The coagulated area induced by high power microwave irradiation is larger than that induced by low power irradiation. With microwaves at 80 W for 60 seconds using a thick electrode, the coagulated area was elliptical, measuring 3.1 cm 6 0.5 cm in maximal dimension and 2.2 cm 6 0.4 cm in minimal dimension on normal rabbit liver tissue in an in vivo test. With regard to the irradiation time, 60 seconds was optimal for obtaining the maximal coagulated area. Even if microwave irradiation was continued for more than 60 seconds, the coagulated area would be much the same as that obtained with 60 seconds of irradiation in an in vivo test using normal rabbit liver tissue. At each session, the multiple electrode insertions were performed on the tumor, including different sites in its proximity, and one microwave irradiation at 80 W for 60 seconds was performed for each electrode insertion until the hyperechogenic change covered the entire tumor to produce extensive necrosis, depending on the size of the tumor and the change seen on echo images. Two to four microwave irradiations (equal to two to four electrode insertions for different sites) could change the echo image of the tumor area to a high echo area that included not only the tumor but also the neighboring noncancerous tissue. When this change in the echo images was achieved, we finished the initial session. PEIT PEIT was performed as described previously.2 After local anesthesia, a 21-gauge, 25-cm needle was introduced into the tumor under US guidance. Approximately 2– 4 mL of ethanol were mixed with 1% lidocaine at a ratio of 9 to 1 (99.5% in 1% lidocaine) and injected into the tumor and into two or three different sites in its proximity (one session). For the purpose of acquiring reliable local control of the tumor, we tried to induce a 5-mm or greater margin of necrosis in the noncancerous tissue surrounding the tumor (the treated margin; Fig. 1). To assess the necrotic area, dynamic CT was performed 2–3 days after each treatment. When the initial session (PMCT or PEIT) failed to obtain the treated margin, the next session was performed under US guidance with reference to postdynamic CT images taken after the initial session. Therefore, the total number of sessions was determined by dynamic CT image findings. We performed both PMCT and PEIT twice a week. To evaluate the efficacy of PMCT and PEIT independently, TAE was not performed as part of the therapy. Follow-up Following discharge from the hospital, all patients were followed closely. Follow-up ultrasonography was performed every 2 months. Dynamic CT and dynamic magnetic resonance imaging were performed every 3–5 months. Serum alpha-fetoprotein concentrations were measured monthly. The clinical observation periods following treatment ranged from 12 months to 72 months. Classification of the pattern of recurrence Intrahepatic recurrences were classified into five types, as follows: Type A, a single recurrent nodule at the margin of treatment or enlargement of the treated tumor; Type B, a single recurrent nodule in the same subsegment as the initial tumor; Type C, recurrent nodules in the same subsegment as the initial tumor or in multiple subsegments, including the same subsegment as the initial tumor; Type D, a single recurrent nodule in a different subsegment than that in which the initial nodule was seen; and Type E, recurrent nodules in multiple subsegments without involvement of the subsegment in which the initial nod- Microwave Coagulation for Small HCC/Seki et al. 1697 FIGURE 1. Dynamic computed tomographies obtained before and 1 week after percutaneous microwave coagulation therapy (PMCT). (a) Before treatment. The tumor site (white arrow; tumor size 5 1.8 cm in greatest dimension) was demonstrated as a enhanced area. (b) One week after treatment (five microwave irradiations). PMCT induced a sufficient treated margin. The tumor and the surrounding area were not enhanced. ule was seen. Subsegments were defined according to Couinaud’s segmental model of the liver.20 When an intrahepatic recurrence was suspected on imagings studies, we performed digital subtraction angiography and US-guided liver biopsy to confirm the diagnosis. Treatment of recurrences In patients with a single recurrent nodule measuring 2.0 cm or less in greatest dimension, we performed PMCT or PEIT. For a recurrence after treatment with PMCT, PMCT was performed again. For patients previously treated with PEIT, PEIT was performed again. In patients with a single recurrent nodule measuring more than 2.0 cm in greatest dimension or with multiple recurrent nodules, we performed TAE in combination with PMCT or PEIT. Statistical analysis Patient age and tumor size are expressed as the mean 6 standard deviation. We used the chi-square test to compare values between groups, applying Yates correction when needed. Continuous variables were compared by means of the Mann–Whitney U test. Cumulative survival rates and cumulative cancer free survival rates were calculated by using the Kaplan– Meier method.21 Differences in survival and cancer free survival were analyzed with the generalized Wilcoxon test.22 A P value of ,0.05 was considered statistically significant. RESULTS There were no differences with regard to age, gender, status of the underlying liver disease, tumor size, or serum alpha-fetoprotein concentration between the two treatment groups (Table 1). At the time of analysis, no difference was evident with regard to the median follow-up times of patients subjected to PMCT (well differentiated HCC: 30 months; range, 12– 63 months; moderately or poorly differentiated HCC: 34 months; range, 14 – 68 months) or PEIT (well differentiated HCC: 32 months; range, 13–71 months; moderately or poorly differentiated HCC: 33 months; range, 12–72 months). None of the patients dropped out during the follow-up period in this study. Treatment Efficacy The PMCT of the tumors was finished within one week in all patients. The number of sessions was one or two (mean, 1.6 sessions). After three to four microwave irradiations (60 seconds comprised one microwave irradiation) were performed on a tumor %1.5 cm, including different sites in its proximity, and four to six irradiations were performed on a tumor .1.5 cm in greatest dimension, including different sites in its proximity, 45 of 48 patients showed complete necrosis of the tumor lesion with a treated margin $ 5 mm on dynamic CT. To obtain a treated margin $ 5 mm, large numbers of microwave irradiations (six electrode insertions) were necessary for some patients. This may have been due to the cooling effect of circulated blood. In the other 3 patients (2 patients with well differentiated HCC and 1 patient with moderately differentiated HCC), despite showing complete necrosis of the tumor lesion, the parenchyma surrounding the tumor 1698 CANCER April 15, 1999 / Volume 85 / Number 8 TABLE 1 Clinical Characteristics of Patients with a Single Hepatocellular Carcinoma Histologic grade Well M or P Characteristic PEIT (n 5 20) PMCT (n 5 23) P valuea PEIT (n 5 22) PMCT (n 5 25) P valuea Gender (M:F) Age (yrs)b Child-Pugh class (A/B/C) HCV/HBV/unknown Maximal tumor size (mm)b AFP (ng/mL) #10 11–100 .100 17:3 61.5 6 8.6 13/5/2 19/1/0 16.3 6 3.0 20:3 60.9 6 7.9 15/8/0 18/4/1 16.0 6 3.4 NS NS NS NS NS 20:2 61.0 6 8.4 11/10/1 21/1/0 18.8 6 2.4 22:3 65.2 6 8.0 12/12/1 23/1/1 18.3 6 2.3 NS NS NS NS NS 11 8 1 8 13 2 NS NS NS 4 15 3 3 20 2 NS NS NS Well: well-differentiated; M or P: moderately or poorly differentiated; PEIT: percutaneous ethanol injection therapy; PMCT: percutaneous microwave coagulation therapy; NS: not significant; HCV: hepatitis C virus; HBV: hepatitis B virus; AFP: alpha-fetoprotein. a Mann–Whitney U test or chi-square test with Yates correction. b Data are expressed as the mean 6 standard deviation. still was viable. This was because PMCT could not be performed extensively due to the presence of the gall bladder or to large vessels near the tumor. There were no clinically serious side effects or complications associated with PMCT in this study. Most patients reported a heat sensation in the upper abdominal region. Half of them experienced some pain during treatment, but it was not severe enough to warrant cessation of treatment. None of the patients in this study refused to continue PMCT. PEIT was performed for three to five sessions per tumor. The total amount of injected ethanol was 12–20 mL for each tumor. Twenty-six of 42 patients had complete necrosis of the tumor lesion with a treated margin $ 5 mm on dynamic CT. In the other 16 patients (7 patients with well differentiated HCC and 9 patients with moderately or poorly differentiated HCC), the dynamic CT showed complete necrosis of the tumor lesion, but the treated margin was not acquired completely, because the injected ethanol flowed into the branch of the portal vein or the hepatic vein around the tumor instead of remaining homogeneously in the marginal area despite multiple injections (maximum, 10 injections). There were no clinically serious side effects or complications associated with PEIT. Almost all patients suffered transient pain during ethanol injection. Survival The 5-year overall survival rate for patients with well differentiated HCC treated with PMCT was 70%, whereas that for patients treated with PEIT was 78%. FIGURE 2. Overall survival of patients with moderately or poorly differentiated hepatocellular carcinoma (HCC) treated with percutaneous microwave coagulation therapy (PMCT) or percutaneous ethanol injection therapy (PEIT). The survival of patients treated with PMCT (n 5 25) was significantly higher than that of patients treated with PEIT (n 5 22) at 5 years (P 5 0.03). Heavy line: PMCT; light line: PEIT; section sign: death by HCC progression (initial therapy, PEIT); cross with dots: death by HCC progression (initial therapy, PMCT). The difference between the two treatment groups was not statistically significant (P 5 0.85). The 5-year overall survival rate for patients with moderately or poorly differentiated HCC treated with PMCT was 78%, whereas that for patients treated with PEIT was 35%. The difference in overall survival between the two treatment groups was statistically significant (P 5 0.03; Fig. 2). Among the patients treated with PEIT, the 5-year overall survival rate associated with well differentiated HCC was considered significantly better than that for Microwave Coagulation for Small HCC/Seki et al. 1699 TABLE 2 Pattern of Recurrence after Initial Therapy Histologic gradea Well FIGURE 3. Cancer free survival of patients with moderately or poorly differentiated hepatocellular carcinoma (HCC) treated with percutaneous microwave coagulation therapy (PMCT) or percutaneous ethanol injection therapy (PEIT). There was no significant difference in cancer free survival between the two groups (P 5 0.17). Heavy line: PMCT: light line: PEIT. patients with moderately or poorly differentiated HCC (P 5 0.04). Conversely, among the patients treated with PMCT, the 5-year overall survival rate was slightly higher for those with moderately or poorly differentiated HCC than for those with well differentiated HCC. However, this difference was not statistically significant (P 5 0.45). Cancer Free Survival Rate The 4-year cancer free survival rate for patients with well differentiated HCC treated with PMCT was 37%, whereas that for patients treated with PEIT was 39%. This difference was not statistically significant (P 5 0.37). The 4-year cancer free survival rate for patients with moderately or poorly differentiated HCC treated with PMCT was 30%, whereas that for patients treated with PEIT was 18%. This difference was not statistically significant (P 5 0.17; Fig. 3). However, 9 of 13 patients who had a recurrence after treatment with PEIT recurred within 2 years. This is in contrast to recurrences that occurred throughout the follow-up period in the patients treated with PMCT. At present, we cannot obtain the 5-year cancer free survival rates for the two treatment groups in this study. Pattern of Recurrence In patients with well differentiated HCC, there were no differences in the pattern of recurrence between the two therapeutic groups. Furthermore, there was no correlation between the recurrence pattern and the response to therapy. Of 7 patients who were without complete necrosis of the surrounding parenchyma after PEIT, 2 patients experienced a recurrence: Type B (1 patient) or Type D (1 patient). Two patients who M or P Recur. p.\initial therapy PEIT PMCT PEIT PMCT Type A Type B Type C Type D Type E 0 3 0 7 1 0 2 0 5 1 2 2 5 2 2 0 2 0 7 1 Well: well differentiated; M or P: moderately or poorly differentiated; Recur. p.: recurrence pattern (the types of recurrences are described in Methods); PEIT: percutaneous ethanol injection therapy; PMCT: percutaneous microwave coagulation therapy. a Statistical significance: M or P, PEIT versus PMCT (Type A 1 Type B 1 Type C) versus (Type D 1 Type E): P 5 0.03; chi-square test. were without complete necrosis of the surrounding parenchyma after PMCT had not experienced a recurrence. Among the patients with moderately or poorly differentiated HCC, there were 2 with Type A recurrence in patients treated with PEIT. Furthermore, Type C recurrence was more frequent with PEIT than with PMCT. Type A and Type C recurrence was not observed in patients treated with PMCT. Of nine patients who were without complete necrosis of the surrounding parenchyma after PEIT, 8 developed recurrence in the original target subsegment: Type A (2 patients), Type B (1 patients), and Type C (5 patients). One patient who was without complete necrosis of the surrounding parenchyma after PMCT experienced a Type B recurrence. Type D recurrence was more frequent in patients treated with PMCT than in patients treated with PEIT. Recurrence after PMCT was observed mainly in a different subsegment than the original tumor (P 5 0.03; Table 2). No extrahepatic recurrence was seen in any patients. Histologic Grade of Recurrent Nodules Recurrent nodules in patients with an initial well differentiated HCC nodule usually were well differentiated, regardless of the therapy. Nine of 11 recurrences (Type B, 2 patients; Type D, 6 patients; Type E, 1 patient) after initial PEIT and 6 of 8 recurrences (Type B, one patient; Type D, 5 patients) after initial PMCT were well-differentiated HCC. In patients with an initial moderately or poorly differentiated HCC treated with PEIT, 9 of 13 recurrences (Type A, 2 patients; Type B, 2 patients; Type C, 5 patients) were moderately or poorly differentiated. 1700 CANCER April 15, 1999 / Volume 85 / Number 8 There were 4 well-differentiated recurrences (Type D, 2 patients; Type E, 2 patients). Conversely in the patients treated with PMCT, 7 of 10 recurrences (Type B, 1 patient; Type D, 6 patients) were well differentiated. The other three patients showed moderately or poorly differentiated recurrence. TABLE 3 Causes of Death in Patients with Hepatocellular Carcinoma Treated with Percutaneous Ethanol Injection Therapy or Percutaneous Microwave Coagulation Therapy Histologic grade Well Treatment of Recurrence In patients with an initial well differentiated HCC, PEIT was performed for 10 recurrences, and PMCT was carried out for 7. TAE was performed for only 2 patients. After treatment for moderately or poorly differentiated HCC, PEIT was performed for 5 recurrences, and PMCT was performed for 9. TAE (in combination with PEIT/PMCT) was performed for 9 recurrences. Of the 9 recurrences treated with TAE, 8 had undergone PEIT as the initial treatment. Of 16 recurrences treated with PMCT, 15 demonstrated complete necrosis of the recurrent lesion with a treated margin $ 5 mm. Of 15 recurrences treated with PEIT, 11 showed complete necrosis of the recurrent lesion with a treated margin $ 5 mm. Definite local recurrences of treated recurrent nodules were observed in 1 patient after retreatment by PEIT. Moreover, in some patients (10 patients retreated with PMCT, 9 patients retreated with PEIT), other types of recurrence were detected during the follow-up period. Of 11 patients treated with TAE, 10 developed the tumor regrowth or multiple recurrent nodules. Of 42 recurrent patients in this study, 28 remain alive at present. Causes of Death No differences in causes of death were found between the two therapeutic groups among the patients with well-differentiated HCC. Among the patients with moderately or poorly differentiated HCC, deaths due to HCC progression were common after PEIT (Fig. 1, Table 3). Seven patients with Type A recurrence (1 patient), Type B recurrence (1 patient), or Type C recurrence (5 patients) after initial PEIT and 1 patient with a Type E recurrence after initial PMCT died of HCC progression despite several retreatments. No patients died of liver failure without HCC progression in either group. DISCUSSION In this study, among the patients with well-differentiated HCC, there were no significant differences in 5-year overall survival, 4-year cancer free survival, or pattern between the patients treated with PMCT and those treated with PEIT. Generally speaking, well-differentiated, small HCC is a good indication for PEIT because of the low incidence of intra- or extracapsular M or P Cause of death\initial therapy PEIT PMCT PEIT PMCT HCC progression Liver failure without HCC progression Variceal bleeding Others 2 0 1 1a 1 0 0 1b 7 0 1 1c 1 0 0 2d Well: well differentiated; M or P: moderately or poorly differentiated; HCC: hepatocellular carcinoma; PEIT: percutaneous ethanol injection therapy; PMCT: percutaneous microwave coagulation therapy. a Myocardial infarction. b Lung carcinoma. c Subarachnoidal haemorrhage. d Lung carcinoma and cerebral haemorrhage. invasion, vascular invasion, and intrahepatic metastases, including microscopic metastatic lesions around the main tumor.11 The therapeutic results with PEIT in this study are in agreement with previous reports.23,24 Therefore, PEIT is indicated in cases of well differentiated, small HCC because of its simplicity and low cost compared with PMCT. Among the patients with moderately or poorly differentiated HCC, the 5-year overall survival for patients treated with PMCT was higher than that for patients treated with PEIT. The better overall survival after PMCT may be related to the fibrous capsules or intratumor septae that often are produced by moderately or poorly differentiated HCC.15,25 The ethanol used in PEIT may not penetrate the capsule or the septa, sparing cancer cells that have invaded beyond the capsule. These few remaining viable cells in the capsule or extracapsular area cannot be detected by modern imaging techniques.9 In fact, although various images indicated that PEIT had induced complete tumor necrosis, we often have observed viable cancer cells persisting in the intracapsular or extracapsular area in surgically resected specimens.2,8 Cancer cells remaining in the tumor capsule or extracapsular area may produce local recurrences as well as intrahepatic metastases and can worsen the patient’s prognosis. It has been reported that the local recurrence rate after PEIT is higher for small HCCs with fibrous septae than for small HCCs without fibrous septae.26 We confirmed recurrences at the margin of treated nodules and a few patients with recurrent nodules in the same subsegment as the initial tumor or in multiple subsegments, including the same Microwave Coagulation for Small HCC/Seki et al. subsegment as the initial tumor, after PEIT in this study. Moderately or poorly differentiated HCC has a high incidence of microscopic metastases adjacent to the primary tumor.11 We therefore try to produce extensive necrosis not only of the tumor lesion but also of the surrounding normal parenchyma. The possibility that occult metastasic lesions persist in the surrounding parenchyma after treatment with PEIT may be high.7 This may be because the ethanol, which is injected into the parenchyma surrounding the tumor, sometimes is distributed inhomogeneously or flows into the vessels around the tumor instead of remaining in the tissue. In this study, some patients treated with PEIT did not obtain adequate margins despite inducing necrosis of the tumor lesion. Moreover, the number of recurrences in the original target subsegment was higher after PEIT than after PMCT. These results indicate that PEIT is less efficacious than PMCT for the local control of moderately or poorly differentiated small HCCs, leading to a poorer prognosis. In fact, in this study, among the patients with moderately or poorly differentiated HCC, deaths due to HCC progression were common after PEIT. However, it is possible that the poorer survival in patients treated with PEIT is related to liver damage due to the nonuniform application of TAE in recurrent cases. The cancer free survival rates for patients treated with PMCT were similar to those for patients treated with PEIT, but the types of recurrences and the histologic grade of recurrences differed. The recurrences after PMCT were observed mainly in a different subsegment than the segment that contained the original tumor. With regard to the histologic grade of recurrent nodules after PMCT, 7 of 10 recurrent nodules were well differentiated. On the other hand, after PEIT, 4 of 13 recurrent nodules were well differentiated. According to the previously reported criteria, well-differentiated, recurrent nodules may be metachronous multicentric developments without a correlation to initial nodules.27–29 Regardless of the presence of a fibrous capsule, intratumor septae, or vessels around the tumor, microwave irradiation can induce tissue coagulation across a designated area if the tip of the microwave electrode is located in the target area.10 The reliable local control capability of PMCT may reduce intrahepatic metastases from a initial tumor or microscopic metastases adjacent to the primary tumor and improve the patients prognosis. However, PMCT cannot inhibit occurrences of metachronous multicentric tumors. Therefore, the improved local control and the ability to treat recurrences more successfully, includ- 1701 ing metachronous multicentric developments, may explain why the overall survival for patients with moderately or poorly differentiated HCC treated with PMCT is better than that for patients treated with PEIT despite the similar cancer free survival rates for the two groups. In fact, it has been reported that the prognoses for patients showing metachronous multicentric development as recurrent nodules were superior to those for patients showing intrahepatic metastasis because percutaneous local treatments are effective for well-differentiated, recurrent nodules.29 Unexpectedly, among the patients treated with PMCT, the 5-year overall survival rate was slightly higher for those with moderately or poorly differentiated HCC than for those with well-differentiated HCC. The number of patients treated with PMCT may have been too small to obtain accurate results. However, this difference was not statistically significant. These data indicate that PMCT demonstrated a good therapeutic effect for small HCCs measuring 2.0 cm or less in greatest dimension despite different grades of tumors. This study was not a prospective, randomized trial and has the potential for selection biases of treatment groups. Therefore, in assessing the therapeutic efficacy of PMCT, there is a great need for a multicenter, randomized, controlled trial compared with PEIT or other treatments with longer follow-up and a larger series. Based on our experience, however, we think that PMCT should have priority over PEIT for moderately or poorly differentiated small HCC because of its better rate of local control. Although we did not perform PMCT for multiple small lesions in this study, we consider that PMCT may be effective for multiple small HCCs and small metastatic liver tumors due to its reliable coagulation capability and mild invasiveness. REFERENCES 1. 2. 3. 4. Kuroda C, Sakurai M, Monden M, Marukawa T, Hosoki T, Tokunaga K, et al. Limitation of transcatheter arterial chemoembolization using iodized oil for small hepatocellular carcinoma. A study in resected cases. Cancer 1991;67:81– 6. Seki T, Nonaka T, Kubota Y, Mizuno T, Samashima Y. Ultrasonically guided percutaneous ethanol injection therapy for hepatocellular carcinoma. Am J Gastorenterol 1989;84: 1400 –7. Ebara M, Ohto M, Sugiura N, Kita K, Yoshikawa M, Okuda K, et al. Percutaneous ethanol injection for the treatment of small hepatocellular carcinoma. Study of 95 patients. J Gastroenterol Hepatol 1990;5:616 –27. Shiina S, Tagawa K, Unuma T, Fujino H, Uta Y, Niwa Y, et al. Percutaneous ethanol injection therapy of hepatocellular carcinoma: analysis of 77 patients. Am J Radiol 1990;155: 1221– 6. 1702 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. CANCER April 15, 1999 / Volume 85 / Number 8 Livraghi T, Bolondi L, Lazzaroni S, Marin G, Morabito A, Papaccini G, et al. Percutaneous ethanol injeciton in the treatment of hepatocellular carcinoma in cirrhosis. Cancer 1992;69:925–9. Livraghi T, Bolondi L, Cottone M, Mazziotte A, Morabito A, Torzilli G, et al. No treatment, resection and ethanol injection in hepatocellular carcinoma: a retrospective analysis of survival in 391 patients with cirrhosis. J Hepatol 1995;22: 522– 6. Kotoh K, Sakai H, Sakamoto S, Nakayama S, Satho M, Morotomi I, et al. The effect of percutaneous ethanol injection therapy on small solitary hepatocellular carcinoma is comparable to that of hepatectomy. Am J Gastorenterol 1994;89: 194 – 8. Shiina S, Tagawa K, Unuma T, Takanashi R, Yoshiura K, Komatsu Y, et al. Percutaneous ethanol injection therapy for hepatocellular carcinoma: a histopathlogic study. Cancer 1991;68:1524 –30. Ishii H, Okada S, Nose H, Okusaka T, Yoshimori M, Takayama T, et al. Local recurrence of hepatocellular carcinoma after percutaneous ethanol injection. Cancer 1996;77: 1792– 6. Seki T, Wakabayashi M, Nakagawa T, Itho T, Shiro T, Kunieda K, et al. Ultrasonically guided percutaneous microwave coagulation therapy for small hepatocellular carcinoma. Cancer 1994;74:814 –25. Kanai T, Hirohashi S, Upton MP, Noguchi M, Kishi K, Makuuchi M, et al. Pathology of small hepatocellular carcinoma: a proposal for a new classification. Cancer 1987;60: 810 –9. The liver cancer study group of Japan: primary liver cancer in Japan. Ann Surg 1989;211:277– 8. Arii S, Tanaka J, Yamazoe Y, Minematus S, Morino T, Fujita K, et al. Predictive factors for intrahepatic recurrence of hepatocellular carcinoma after partial hepatectomy. Cancer 1992;69:913–9. Haratake J, Takeda S, Kasai T, Nakano S, Tokui N. Predictable factors for estimating prognosis of patients after resection of hepatocellular carcinoma. Cancer 1993;72:1178 –3. Kitamoto M, Nakanishi T, Kira S, Kawaguchi M, Nakashio R, Suemori S, et al. The assessment of proliferating cell nuclear antigen immunohistochemical staining in small hepatocellular carcinoma and its relationship to histologic characteristics and prognosis. Cancer 1993;72:1859 – 65. Hermanek P, Sobin LH. TNM classification of malignant tumors. 4th edition Berlin: Springer-Verlag, 1987:53–5. Edmondson HA, Steiner PE. Primary carcinoma of the liver: 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. a study of 100 cases among 48900 necropies. Cancer 1954; 7:462–503. Tabuse K. A new operative procedure of hepatic surgery using a microwave tissue coagulator. Arch Jpn Chir 1979;48: 160 –72. IEEE Standards Coordinating Committee 28 on Non-Ionizing Radiation Hazards. IEEE standard for safety levels with respect to human exposure to radio frequency electromagnetic fields, 3 KHz to 300 GHz. IEEE C95. 1-1991. New York: The Institute of Electrical and Electronics Engineers, Inc., 1992:9 –35. Couinaud C. Les enveloppes vasculo-billiaires du foie ou capsule de Glisson. Lyon Chir 1954;49:489 – 607. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958;63:457– 81. Gehan E. A generalized Wilcoxson test for comparing arbitrarily singly-censored samples. Biometrika 1965;52:203–23. Yamashita Y, Matsukawa T, Arakawa A, Hatanaka Y, Urata J, Takahashi M. US-guided liver biopsy: predicting the effect of interventional treatment of hepatocellular carcinoma. Radiology 1995;196:799 – 804. Toyoda H, Kumada T, Nakano S, Takeda I, Sugiyama K, Kiriyama S, et al. Significance of tumor vascularity as a predictor of long-term prognosis in patients with small hepatocellular carcinoma treated by percutaneous ethanol injection therapy. J Hepatol 1997;26:1055– 62. Kojiro M. Pathomorphology of advanced hepatocellular carcinoma. In: Tobe T, Kameda H, Okudaira M, Ohto M, Endo Y, Mito M, et al., editors. Primary liver cancer in Japan. Tokyo: Springer-Verlag, 1992:31–7. Hirata M, Nishimura N. Hepatocellular carcinoma recurrence after percutaneous ethanol injection therapy: comparisons of tumors with or without a septum. Jpn J Med Ultrasonics 1993;20:548 –54. Okuda K, Tanaka M, Nakayama T, Saitu H, Tanigawa K, Nakashima O, et al. Clinicopathologic comparison between resected hepatocellular carcinoma (HCC) and recurrent tumors: a special reference to multicentric carciogenesis of HCC. Int Hepatol Comm 1993;1:65–71. Takanaka K, Adachi E, Nishizaki T, Hiroshige K, Ikeda, T, Tsuneyoshi M, et al. Possible multicentric occurrence of hepatocellular carcinoma: a clinicopathological study. Hepatology 1994;19:889 –94. Kumada T, Nakano S, Takeda I, Sugiyama K, Osada T, Kiriyama S, et al. Patterns of recurrence after initial treatment in patients with small hepatocellular carcinoma. Hepatology 1997;25:87–92.