Journal of Gastroenterology and Hepatology (2000) 15, 1079–1086

Spontaneous regression of HCC is rare but it can occur.The underlying mechanism of this phenomenon isunclear. Here we report a case of HCC at stage IIaccording to Okuda’s classification1 with spontaneousregression as demonstrated by a-fetoprotein (AFP) andimaging studies.

CASE REPORT

A 68-year-old Japanese man with a 7-year history ofchronic hepatitis C was well until January 1994, whenmultiple liver tumours were identified by ultrasonogra-phy and computed tomography (CT) scan. On admis-sion the patient complained of mild fatigue without

See editorial on page 965

INTRODUCTION

Hepatocellular carcinoma (HCC) is a malignanttumour of worldwide distribution with a poor progno-sis. For unresectable HCC, the course of the disease was usually rapid and fatal. Median survival of such apatient who did not receive specific treatment was0.7–8.3 months after the diagnosis and the patients died from hepatic failure, gastrointestinal haemorrhage,intra-abdominal bleeding or cachexia.1

To date, 23 cases of apparently spontaneous regres-sion of HCC have been published in the literature.2–22

CASE REPORT

Spontaneous regression of hepatocellular carcinoma andreview of literature

YOSHIO TAKEDA,* HITOSHI TOGASHI,* HARUHIDE SHINZAWA,* SHINTARO MIYANO,*RIKA ISHII,* TETSURU KARASAWA,* YUMIKO TAKEDA,† TAKAFUMI SAITO,*

KOJI SAITO,* HIROKO HAGA,* TAKU MATSUO,* MASANORI AOKI,*HIDEKI MITSUHASHI,* HISAYOSHI WATANABE* AND TSUNEO TAKAHASHI*

*Second Department of Internal Medicine,Yamagata University School of Medicine,Yamagata and†Department of Internal Medicine, Shinoda General Hospital, Japan

Abstract A 68-year-old man presented with multiple hepatocellular carcinoma, which was consid-ered to be unresectable at the first admission in January 1994. Pathological diagnosis was made bybiopsy of the one lesion among them. From January 1994 to December 1997, 10 transarterial chemoem-bolizations and six percutaneous ethanol injection therapies were performed on the tumours in the cir-rhotic liver. In February 1998 the tumour situated in the right lobe began to increase in size. Themaximum tumour diameter was 6.3 cm measured by computed tomography (CT). In the beginning ofMay 1998 moderate ascites was present and mild hepatic encephalopathy was noticed.The patient wasin the terminal stage of hepatocellular carcinoma and no further treatment was possible at that time.However, serum a-fetoprotein and protein induced by vitamin K absence or antagonist II dramaticallydecreased in June 1998. The CT scan also showed that the tumour had completely regressed withoutspecific treatment. In February 1999 a new biopsy-proven hepatocellular carcinoma, 2 cm in diameter,developed in the lateral segment of the liver. It was well treated by percutaneous ethanol injectiontherapy. The patient was alive in good condition without any symptoms or tumour recurrence in June1999. It was concluded that a rare case of spontaneous regression of hepatocellular carcinoma hadoccurred.© 2000 Blackwell Science Asia Pty Ltd

Key words: hepatitis C, hepatocellular carcinoma, liver cirrhosis, spontaneous regression, tumourthrombosis in the portal vein.

Correspondence: Dr H Togashi, The Second Department of Internal Medicine, Yamagata University School ofMedicine, 2–2-2 Iidanishi,Yamagata 990–9585, Japan. E-mail: htogashi@med.id.yamagata-u.ac.jp

Accepted for publication 30 November 1999.

ascites or hepatic encephalopathy.The patient had beena regular drinker, consuming approximately 540 mL ofthe Japanese alcoholic beverage, Sake, (81 g alcohol)daily for 40 years. He has a past history of left lung lobec-tomy for tuberculosis, and a blood transfusion had been performed in 1955. Laboratory data are shown in Table 1. Hepatis B surface antigen (HBsAg), antihepati-tis B surface antibody and antihepatitis B core antibodywere all negative. Antihepatitis C antibody was positive.

A CT scan of the abdomen and abdominal angio-graphy showed multiple hypervascular tumours in themedial segment (S4), in the inferior anterior segment(S5), in the superior anterior segment (S8), in the infe-rior posterior segment (S6) and in the superior poste-rior segment (S7) of the liver. The largest liver mass,2.8 cm in diameter, in the superior anterior segment(S8) was demonstrated to be a low-density tumour. Anearly phase contrast-enhanced CT showed a high-

1080 Y Takeda et al.

density tumour in S8. A late phase contrast-enhancedCT revealed a low-density tumour in S8 indicatingHCC. Laparoscopic examination revealed liver cirrho-sis and a biopsy was taken from the liver surface of themedial segment (S4). Biopsy specimens from non-tumour area showed liver cirrhosis and the specimensfrom the tumour in the medial segment (S4) revealedmoderately differentiated HCC as shown in Fig. 1a,b.Transarterial chemoembolization (TAE) was per-formed using doxorubicin–lipiodol emulsion andgelform for HCC twice in the left and the right lobesof the liver. Percutaneous ethanol injection therapy(PEIT) was performed in the superior anterior segment(S8) five times in 3 months. Complete tumour necro-sis was elicited in the medial segment (S4), in the infe-rior anterior segment (S5), in the inferior posteriorsegment (S6) and in the superior posterior segment(S7). Partial tumour necrosis was obtained in the

Table 1 Laboratory data

Items Normal range 10 January 1994 8 April 1998

White blood cell count 5000–7000 cells/mL 4200 3300Haemoglobin 12–14.5 g/dL 16.4 10.5Platelet 10–40(¥ 104/mL) 11.1 6.3Total protein 6.3–8.2 g/dL 7.9 6.9Albumin 3.8–5.3 g/dL 3.6 2.8Total bilirubin < 1.3 mg/dL 1.1 1.7AST < 35 IU/L 112 48ALT < 35 IU/L 117 23Gamma GTP < 45 IU/L 59 15Prothrombin time 10–13 s 13.8 16.1Prothrombin time (control) 14.4 12AFP < 10 ng/mL 18.9 8230PIVKA II < 40 mAU/mL not detected 1477

AST, aspartate aminotransferase; ALT, alanine aminotransferase; GTP, glutamyl transpeptidase; AFP, a-fetoprotein; PIVKAII, protein induced by vitamin K absence or antagonist II.

Figure 1 Pathological features of the liver in January, 1994. (a) Pathological features of the non-tumour area. (b) Pathologicalfeatures of hepatocellular carcinoma (HCC) in the medial segment (S4) of the liver at the time of the first admission. Biopsyrevealed microtrabecular HCC with focal pseudoglandular formation. The tumour cells had increased nuclear to cytoplasmicratios with nuclear atypia, anisonucleosis, irregular nuclear outlines and increased numbers of mitotic figures.

tumour in the superior anterior segment (S8) and asmall viable part was seen in the centre of the tumour.

Because the tumour in the superior anterior segment(S8) increased in size, TAE using doxorubicin–lipiodolemulsion and gelform was repeated in September 1994.Mild tumour regression was obtained. During the fol-lowing 4 years, new HCC lesions appeared in the infe-rior posterior segment (S6), the superior posteriorsegment (S7) and the superior anterior segment (S8),and a previously demonstrated HCC in the superioranterior segment (S8) recurred. Conventional treat-ment was performed for those lesions.

In total, the HCC lesions were treated six times byPEIT, and 10 times by TAE using doxorubicin,

Spontaneous regression of HCC 1081

mitomycin c, smancs or epirubicin with lipiodol andgelform.Total cumulative doses used for the procedureswere: doxorubicin, 245 mg; mitomycin C, 4 mg; epiru-bicin, 40 mg; and smancs lipiodol, 10 mL. Complete orpartial regression was noticed following the treatments.Although the tenth TAE was performed against thelesion of HCC in the right lobe of the liver, the lesionsbegan to increase in size in February 1998. Laboratorydata in April 1998 are shown in Table 1. Alpha-fetoprotein and protein induced by vitamin K absenceor antagonist II (PIVKA II) were remarkably elevatedon 6 May 1998, and CT scan showed that the tumouroccupied the anterior and posterior segment (S5, S6,S7 and S8) of the right lobe of the liver (Fig. 2a,b).The

Figure 2 Morphological features of hepatocellular carcinoma (HCC) by delayed phase contrast-enhanced computed tomog-raphy (CT). (a) Heterogeneously enhanced hypoenhanced tumour (black arrows) was seen in the S5 and S6 of the liver, andtumour thrombosis was noticed in the right portal vein as shown (white arrow) on 7 May 1998. PV, portal vein. (b) A hypoen-hanced tumour was observed in S8 of the liver and the inferior vena cava was compressed by the tumour on 7 May 1998. Denselycondensed lipiodol accumulation and a perifocal necrotic low-density area, which were caused by previous transarterial chemoem-bolization (TAE) were also demonstrated. (c) Complete spontaneous tumour regression was seen in S5 and S6, and the portalvein (PV) became patent on 3 September 1998. (d) Complete spontaneous tumour regression was also observed in S8 and theinferior vena cava became patent on 3 September 1998.

maximum tumour diameter was 6.3 cm. Ultrasonogra-phy and CT scan of the liver demonstrated tumourthrombosis in the right branch of the portal vein asshown in Figs 2a,b. Moderate ascites was present andmild hepatic encephalopathy was noticed at the begin-ning of May 1998. At this time, the patient was in theterminal stage of HCC and no further antitumourtherapy was possible.

The general condition of the patient improved andascites gradually disappeared. Serum AFP and PIVKAII levels decreased dramatically in June 1998 (Fig. 4).Tumour regression of the HCC was seen, and CT scansin July 1998 showed that there was more than 90%regression. Complete tumour regression was seen inSeptember 1998 (Fig. 2c,d).The patient later informedus that he had taken two kinds of alternative herbalmedicines by his will from the beginning of May 1998to June 1998. One was crude extract from Agaricusblazei Murill, the other was tahibo tea. A non-selectivebeta-blocker, nadolol, was used for the treatment ofportal hypertension from 16 June 1998. The beta-blocker was stopped on 4 July 1998 because of sicksinus syndrome. Spontaneous regression of HCC wasseen without any conventional medical treatmentsduring this period.

Serum AFP level was normal but serum PIVKA IIlevel gradually increased after the spontaneous regres-sion. Thus, a close follow up was conducted. A newbiopsy-proven moderately differentiated HCC, 2 cm indiameter, developed in the lateral segment (S2) of theliver in February 1999 and it was low-density mass bothon plain and contrast-enhanced CT scan. It was welltreated by PEIT. After PEIT, PIVKA II decreased to88 mAU/mL.The patient is alive without any symptomsand signs in July 1999.

1082 Y Takeda et al.

DISCUSSION

Here we report a rare case of spontaneous regression of HCC with liver cirrhosis due to hepatitis C virus inthe terminal stage.This phenomenon is consistent withspontaneous regression of cancer defined by Cole aspartial or complete disappearance of malignancy in theabsence of specific treatment.23 In May 1998 the patientwas at stage II by Okuda’s classification of HCC; themedian survival of 134 stage II patients was only2.0 months.1 There was a persisting abnormality in theright lobe of the liver on 3 September 1998 (Fig. 2c,d).It seems that this abnormality represents an area ofinfarction following the multiple courses of chemoem-bolization, because a following CT scan of the liver alsodemonstrated unchanged abnormality in 1999. More-over, no evidence of tumour recurrence was seen in theright lobe of the liver.Therefore, this regression seemedto be complete.

Spontaneous regression of cancer is estimated tooccur once in 60 000–100 000 cancer patients, andapproximately 20 cases are published in the world lit-erature each year.23,24 Tables 2 and 3 show 24 cases,including the present case, of HCC reported in theEnglish literature according to MEDLINE from 1966to 1999. Another five cases of partial or completeregression of HCC were also reported by letter formatin the English-speaking literature.25 There is no con-sensus idea for the spontaneous regression of HCCbecause spontaneous regression occurs rarely. Variousmechanisms of spontaneous regression of HCC havebeen suggested. Sepsis,2 bronchopneumonia accompa-

Figure 3 A longitudinal ultrasound scan through the rightlobe of the liver in May 1998. It demonstrated a heteroge-neous tumour (arrow heads) in the right lobe and tumourthrombosis (arrows) in the right branch of the portal vein.

Figure 4 Time course of serum a-fetoprotein (AFP; � inng/mL) and protein induced by vitamin K absence or antag-onist II (PIVKA II; � in mAU/mL) levels. The serum AFPand PIVKA II levels decreased dramatically in July 1998, and AFP was normalized in October 1998. The PIVKA II levelincreased gradually from October 1998 and a new hepatocel-lular carcinoma (HCC) developed in the lateral segment (S2)of the liver. After percutaneous ethanol injection therapy wasperformed on the new HCC ( , March 1999), the PIVKA IIlevel became normal in March 1999. (�) Beta-blocker admin-istration; ( , December 1997) 10th transarterial chemoem-bolization.

nied by fever,4 long-lasting fever15 and liver abscess-likeinfections22 may trigger the regression. Immunologicalreaction to this bacterial infection and fever may playan important role in spontaneous regression.4,23 Bacte-rial infection such as sepsis or abscess, accompanyingproduction of cytokines such as tumour necrosis factor(TNF)26–29 and interleukin (IL)-12,30,31 and fever29,32

may have potential antitumour activities in patients withthese conditions. Other mechanisms such as withdrawalfrom anabolic steroids,2 discontinuation of drinking,3,13

shock and transfusion of 3 L of fresh blood,7 surgeryand transfusion of fresh (1.6 L) and preserved blood(1.4 L),5 herbal medicine,4,10 Silymarian,13 macrobioticdiet,8 infarction due to fast growth,6 infarction due to

Spontaneous regression of HCC 1083

poor vasculature in cirrhotic liver,16 infarction due toarterial thrombosis,11 and abscopal regression afterradiotherapy accompanied by elevation of serum TNFlevels20 have been speculated or suggested as reasons forrequession. No special mechanism was noted in ninecases.3,9,12,14,17–19,21 Iatrogenic subintimal injury duringhepatic angiography, however, can induce necrosis ofHCC.33 Even on closer examination of all publishedcases of spontaneously regressing HCC, precise con-clusions about the mechanisms leading to spontaneousregression are still not clear.

A blood transfusion was not performed, and duringthe 5-year follow-up period there was no subintimalinjury during hepatic angiography, and no gastroin-

Table 2 Characteristics of 24 cases of spontaneous regression of hepatocellular carcinoma

Authors and Year Age Tumour Histologicalreference no. reported (years) Sex Cirrhosis Aetiology size (cm) Metastasis diagnosis

Johnson et al.2 1972 5 F No Unknown 6 No NoGottfried et al.3 1982 65 M Yes Alcoholic Multifocal No YesLam et al.4 1982 50 M No Hepatitis B Unknown + Lung Yes

lung metastasesSato et al.5 1985 78 M No Unknown Two tumours of Bone Yes

6 cm and 10 cm +bone metastases

Suzuki et al.6 1989 65 M Yes Unknown 3.2 No NoTocci et al.7 1990 79 M Yes Non-A, non-B 5 No NoGaffey et al.8 1990 63 M Yes Unknown 10 No YesAyres et al.9 1990 63 F Yes Unknown Multifocal + Lung Yes

lung metastasesChien et al.10 1992 63 M Yes Hepatitis B 12 No YesImaoka et al.11 1994 65 M No Non-A, non-B Unknown No YesMcDermott and 1994 23 F Unknown Unknown 12–15 No Yes

Khettry12 multifocal +Grossmann et al.13 1995 52 M Yes Alcohol 4.5 No YesOzeki et al.14 1996 69 F Yes Unknown 5 No YesMarkovic et al.15 1996 62 M Yes Hepatitis B 13 No Yesvan Halteren et al.16 1997 72 F No No 8 No YesIwasaki et al.17 1997 72 F Yes HCV 3.8 No YesGomez Sanz et al.18 1998 66 M No HCV 8 SCLN Yes

abd LNKaczynski et al.19 1998 73 M Unknown Unknown 7 No YesOhba et al.20 1998 76 M Yes HCV Not measured Bone Yes

> 5 cmMagalotti et al.21 1998 66 M Yes Alcohol 1–3 No Yes

No. 1 multifocal +Magalotti et al.21 1998 75 F Yes Hepatitis C, 1.5 and 4.5 No Yes

No. 2 past historyof hepatitis B

Stoelben et al.22 1998 56 M No Unknown 6 No YesNo. 1

Stoelben et al.22 1998 74 M No Unknown, 6 No YesNo. 2 past history

of hepatitis BPresent case 1999 73 M Yes HCV 6.3 No Yes

multifocal +

HCV, hepatitis virus C; SCLN, Sternoclavicular; abdLN, abdominal lymphnodes.

1084 Y Takeda et al.

Table 3 Characteristics of 24 cases of spontaneous regression of hepatocellular carcinoma

Evidence for regression of HCC Accompanying Proposed Disappearance Survival since

Authors and AFP AFP Radiological Alternative lethal disease or mechanism time diagnosisreference no. (ng/mL) normalized disappearance therapy special remarks of regression (months) (months)

Johnson et al.2 Unknown Yes Partial No Septicaemia Withdrawal from 6 *D, 12anabolic steroid

Gottfried et al.3 Normal Still normal Complete No No Unknown 12 †A, 48Lam et al.4 Unknown Yes Complete Herbal Bronchopneumonia Triggered by 6 A, > 168

medicine and fever infectionSato et al.5 12 600 Yes Complete No Surgery and gastro- Unknown 16 A, > 58

intestinal bleeding;transfusion of 3 L of blood

Suzuki et al.6 7767 Yes Complete No No Infarction due to 15 A, > 70fast growth

Tocci et al.7 625 Yes Complete No Shock; transfusion Infarction due to 30 A, > 48of 3 L of fresh systemic shockblood.

Gaffey et al.8 2690 Yes Partial Macrobiotic Variceal bleeding Unknown 24 D, 24(complete dietat autopsy)

Ayres et al.9 8868 Yes Partial No No Unknown 5 A, > 12Chien et al.10 > 10 000 Yes Complete Herbal No Unknown 5 A, > 37

medicineImaoka et al.11 6609 Yes Surgically No No Infarction due to Not obtained Unknown

resected arterial thrombusMcDermott Unknown Unknown Complete No No Unknown 12 A, > 240

and Khettry12

Grossmann 310 Yes Partial Silymarian Gastric cancer, Unknown 14 D, 14et al.13 abstinence from

alcoholOzeki et al.14 1050 Unknown Surgically No No Unknown Unkown A, > 12

resectedComplete

necrosisMarkovic 11 Unknown Surgically No 3 months of fever Unknown 7 A, > 96

et al.15 resectedComplete

necrosisvan Halteren Unknown Normal Partial No No Infarction due to 14 A, > 28

et al.16 poor vasculature in cirrhotic liver

Iwasaki et al.17 105 340 Yes Partial No No Unknown 4 A, > 9Gomez Sanz 4 Unknown Complete No No Unknown 16 A, > 50

et al.18

Kaczynski Unknown Unknown Complete No No Unknown 15 D, 180et al.19

Ohba et al.20 429 998 Yes Partial No Radiotherapy to Abscopal regression 10 A, > 84bone metastasis. of HCC after elevation of radiation for TNF bone metastasis

Magalotti 2500 Yes Complete Unkown HCC recurred Unknown 12 D, 75et al.21 No.1 5 years later

Magalotti 37 500 Yes Partial Unkown HCC recurred Unknown 5 D, 29et al.21 No.2 5 years later

Stoelben et al.22 3.7 Normal Partial No Liver abscess-like Triggered by Regressed in A, > 24No.1 Complete fever infection 2 months

necrosisSurgically

resectedStoelben et al.22 3850 Decreased Partial No Liver abscess-like Triggered by Regressed in A, > 41

No.2 to 20 Complete fever infection 3 monthsnecrosis

Surgicallyresected

Present case 8230 Yes Complete Two herbal No Unknown 4 A, > 66medicines

HCC, hepatocellular carcinoma; AFP, a-fetoprotein; *D, dead; †A, alive.

testinal bleeding or infection in our patient. Because hehad already stopped drinking alcohol at the first admis-sion, abstinence from alcohol was not related to theregression. After the spontaneous regression of HCCoccurred, the patient informed us that he took twokinds of alternative herbal medicines: Agaricus blazeiMurill and tahibo tea. As shown in Fig. 4, the sponta-neous regression of HCC coincided with the use of theherbal medicines. A possible influence of unconven-tional treatment using herbal medicines has beenreported in the literature.4,10 Lam et al.4 and Chienet al.10 tried the recipe used by their patients on 20 and25 subsequent patients with HCC, respectively.Because tumour regression could not be detected in anyof the patients treated, both authors concluded that itwas unlikely that the herbal preparations had an effecton HCC. Agaricus blazei Murill is a type of mushroomand has antitumour effects.34–38 It is uncertain whethertahibo tea has antitumour effects. Although either orboth of two herbal medicines might be associated withregression of HCC, it is not concluded that the use ofthe herbal medicines affected the regression of HCC inthe present patient.

Another possible factor that affected the spontaneousregression of HCC might be the 19-day period of beta-blocker therapy for portal hypertension.We stopped themedication because sick sinus syndrome, a side-effectattributed to beta-blockers, was noted. It is well knownthat beta-blockers diminish splanchnic blood flow. Wecan not completely rule out the possibility that admin-istration of the beta-blocker facilitates the tumourregression. Because serum levels of AFP and PIVKA IIhad already begun to decline at the time of administra-tion of the beta-blocker (Fig. 4), this possibility may below. In contrast, a fast growing HCC in a poorly vas-cular cirrhotic liver becomes ischaemic and necrotic ona more cirrhotic basis.These situations might affect thespontaneous regression of HCC.

In conclusion, spontaneous regression of HCC is rare but it can occur. Allochronic recurrence of HCCcan occur even if spontaneous regression of HCC is obtained. Although the cause of this spontaneousregression of HCC remains uncertain and differs frompatient to patient, accumulation of these cases will leadto progress of understanding of the phenomenon.

REFERENCES

1 Okuda K, Ohtsuki T, Obata H et al. Natural history ofhepatocellular carcinoma and prognosis in relation totreatment. Study of 850 patients. Cancer 1985; 56:918–28.

2 Johnson FL, Lerner KG, Siegel M et al. Association ofandrogenic-anabolic steroid therapy with development ofhepatocellular carcinoma. Lancet 1972; 2: 1273–6.

3 Gottfried EB, Steller R, Paronetto F, Lieber CS. Sponta-neous regression of hepatocellular carcinoma. Gastroen-terology 1982; 82: 770–4.

4 Lam KC, Ho JC, Yeung RT. Spontaneous regression ofhepatocellular carcinoma: a case study. Cancer 1982; 50:332–6.

Spontaneous regression of HCC 1085

5 Sato Y, Fujiwara K, Nakagawa S et al. A case of sponta-neous regression of hepatocellular carcinoma with bonemetastasis. Cancer 1985; 56: 667–71.

6 Suzuki M, Okazaki N,Yoshino M,Yoshida T. Spontaneousregression of a hepatocellular carcinoma – a case report.Hepatogastroenterology 1989; 36: 160–3.

7 Tocci G, Conte A, Guarascio P, Visco G. Spontaneousremission of hepatocellular carcinoma after massive gas-trointestinal haemorrhage. BMJ 1990; 300: 641–2.

8 Gaffey MJ, Joyce JP, Carlson GS, Esteban JM. Sponta-neous regression of hepatocellular carcinoma. Cancer1990; 65: 2779–83.

9 Ayres RC, Robertson DA, Dewbury KC, Millward-SadlerGH, Smith CL. Spontaneous regression of hepatocellularcarcinoma. Gut 1990; 31: 722–4.

10 Chien RN, Chen TJ, Liaw YF. Spontaneous regression ofhepatocellular carcinoma. Am. J. Gastroenterol. 1992; 87:903–5.

11 Imaoka S, Sasaki Y, Masutani S et al. Necrosis of hepato-cellular carcinoma caused by spontaneously arising arter-ial thrombus. Hepatogastroenterology 1994; 41: 359–62.

12 McDermott WV, Khettry U. Clear cell carcinoma of theliver with spontaneous regression of metastases. J. Surg.Oncol. 1994; 57: 206–9.

13 Grossmann M, Hoermann R,Weiss M et al. Spontaneousregression of hepatocellular carcinoma. Am. J. Gastroen-terol. 1995; 90: 1500–3.

14 Ozeki Y, Matsubara N, Tateyama K, Kokubo M, ShimojiH, Katayama M. Spontaneous complete necrosis of hepatocellular carcinoma. Am. J. Gastroenterol. 1996; 91:391–2.

15 Markovic S, Ferlan-Marolt V, Hlebanja Z. Spontaneousregression of hepatocellular carcinoma. Am. J. Gastroen-terol. 1996; 91: 392–3.

16 van Halteren HK, Salemans JM, Peters H,Vreugdenhil G,Driessen WM. Spontaneous regression of hepatocellularcarcinoma. J. Hepatol. 1997; 27: 211–15.

17 Iwasaki M, Furuse J,Yoshino M, Moriyama N, KanemotoH, Okumura H. Spontaneous regression of hepatocellularcarcinoma: a case report. Jpn. J. Clin. Oncol. 1997; 27:278–81.

18 Gomez Sanz R, Moreno Gonzalez E, Colina Ruiz-Delgado F, Garcia-Munoz H, Ochando Cerdan F,Gonzalez-Pinto I. Spontaneous regression of a recurrenthepatocellular carcinoma. Dig. Dis. Sci. 1998; 43: 323–8.

19 Kaczynski J, Hansson G, Remotti H, Wallerstedt S.Spontaneous regression of hepatocellular carcinoma.Histopathology 1998; 32: 147–50.

20 Ohba K, Omagari K, Nakamura T et al. Abscopal regres-sion of hepatocellular carcinoma after radiotherapy forbone metastasis. Gut 1998; 43: 575–7.

21 Magalotti D, Gueli C, Zoli M. Transient spontaneousregression of hepatocellular carcinoma. Hepatogastroen-terology 1998; 45: 2369–71.

22 Stoelben E, Koch M, Hanke S et al. Spontaneous regres-sion of hepatocellular carcinoma confirmed by surgicalspecimen: report of two cases and review of the literature.Langenbecks Arch. Surg. 1998; 383: 447–52.

23 Cole WH. Efforts to explain spontaneous regression ofcancer. J. Surg. Oncol. 1981; 17: 201–9.

24 Jerry L, Challis E. Oncology. In: Re R, ed. Textbook ofFamily Practice. Philadelphia: W.B. Saunders, 1984;1061–81.

25 Herrera A, Erdozain JC, Molina E, Conde P, Palomo V.Spontaneous regression of hepatocellular carcinoma. Am.J. Gastroenterol. 1996; 91: 1288–9.

26 Pinsky MR, Vincent JL, Deviere J, Alegre M, Kahn RJ,Dupont E. Serum cytokine levels in human septic shock.Relation to multiple-system organ failure and mortality.Chest 1993; 103: 565–75.

27 Hamilton G, Hofbauer S, Hamilton B. Endotoxin. TNF-alpha, interleukin-6 and parameters of the cellularimmune system in patients with intraabdominal sepsis.Scand. J. Infect. Dis. 1992; 24: 361–8.

28 Watanabe N, Yamauchi N, Maeda M et al. Recombinanthuman tumor necrosis factor causes regression in patientswith advanced malignancies. Oncology 1994; 51: 360–5.

29 Fujimoto S, Konno C, Kobayashi K et al. Augmentedantitumour effects of combined treatment with hyper-thermia and tumour necrosis factor on human gastriccancer xenotransplanted into nude mice. Int. J. Hyperther-mia 1991; 7: 511–18.

30 Jansen PM, van der Pouw Kraan TC, de Jong IW et al.Release of interleukin-12 in experimental Escherichia coliseptic shock in baboons: relation to plasma levels of interleukin-10 and interferon-gamma. Blood 1996; 87:5144–51.

31 Nishimura T, Watanabe K, Yahata T et al. Application ofinterleukin 12 to antitumor cytokine and gene therapy.Cancer Chemother. Pharmacol. 1996; 38: S27–34.

32 Downing JF, Taylor MW. The effect of in vivo hyperther-mia on selected lymphokines in man. Lymphokine Res.1987; 6: 103–9.

1086 Y Takeda et al.

33 Takayasu K, Muramatsu Y, Shima Y et al. Necrosis ofhepatocellular carcinoma as a result of subintimal injuryincurred by hepatic angiography: report of two cases. Am.J. Gastroenterol. 1986; 81: 979–83.

34 Fujimiya Y, Suzuki Y, Katakura R, Ebina T. Tumor-specific cytocidal and immunopotentiating effects of rela-tively low molecular weight products derived from thebasidiomycete, Agaricus blazei Murill. Anticancer Res.1999; 19: 113–18.

35 Ebina T, Fujimiya Y. Antitumor effect of a peptide-glucanpreparation extracted from Agaricus blazei in a double-grafted tumor system in mice. Biotherapy 1998; 11:259–65.

36 Fujimiya Y, Suzuki Y, Oshiman K et al. Selective tumori-cidal effect of soluble proteoglucan extracted from thebasidiomycete, Agaricus blazei Murill, mediated via naturalkiller cell activation and apoptosis. Cancer Immunol.Immunother. 1998; 46: 147–59.

37 Ito H, Shimura K, Itoh H, Kawade M. Antitumor effectsof a new polysaccharide-protein complex (ATOM) pre-pared from Agaricus blazei (Iwade strain 101) ‘Himemat-sutake’ and its mechanisms in tumor-bearing mice.Anticancer Res. 1997; 17: 277–84.

38 Kawagishi H, Inagaki R, Kanao T et al. Fractionation and antitumor activity of the water-insoluble residue ofAgaricus blazei fruiting bodies. Carbohy. Res. 1989; 186:267–73.