Spontaneous Regression of Uepa tocellular Carcinoma

A Case Study

K. C. LAM, MBBS, FRACP, J. C. I. HO, MB, MSc, AND R. T. T. YEUNG. MD, FRCP, FRCPE, FRACP

A Chinese patient with documented hepatocellular carcinoma (HCC) satisfied the criteria of Everson and Cole4 for spontaneous regression of malignant tumors. Subsequently he survived a tumor-free period of at least 13 years. During the period of regression, shrinkage of liver coincided with a rise of SCOT to a level comparable to that reported for patients with liver cancer during hepatic arterial ligation and cytotoxic therapy. Postregression liver biopsy from the site of the previous tumor revealed relatively uninflamed HBsAg-positive tissue without dysplasia. The case provided the positive end of the survival spectrum in HCC, evidence that regression of HCC might occur by involution rather than maturation, and histologic data suggesting that regressed HCC might be replaced by surrounding tissue instead of leaving behind dysplasia.

Cancer 50:332-336, 1982.

HE PROGNOSIS of patients with hepatocellular car- T cinoma (HCC) is shrouded in gloom. In our pop- ulation with this disease, the median survival is ap- proximately one month from the date of diagnosis.’ In centers where patients may present earlier, the median survival is still no more than three months.2 However, the survival in this disease varies widely. The authors have noted survivals of up to 4% years without anti- tumor therapy. Reported here is one patient who may form the positive end of the spectrum of survival in this disease. He provided a rare opportunity for studying regression and postregressional state in HCC.

Case Report

A 50-year-old male carpenter from Southern China pre- sented to the University Department of Medicine, Queen Mary Hospital, Hong Kong in mid-November 1965 because of pro- gressive distending discomfort in the epigastrium for one month. An enlarging epigastric mass was noted for two weeks. Food intake decreased; body weight was reduced by 3 Kg during this period. There was no known history of previous hepatitis. but in 1964 he had received three months of med- ication for suspected pulmonary tuberculosis. There was no known history of androgen intake. Physical examination re- vealed the presence of jaundice, finger clubbing, ankle oedema, and hard nodular hepatomegaly. The “liver function tests” are summarized in Table 1 . The blood counts were as follows:

From the Departments of Medicine and Pathology, University of

Address for reprints: Dr. K. C. Lam, University Department of

Accepted for publication May I I , 1981.

Hong Kong, Hong Kong.

Medicine, Queen Mary Hospital, Pokfulam Road, Hong Kong.

hemoglobin 10.8 g/dl, leukocytes 6.1 X 109/1 P57L40E2Ml, platelets 213 X 109/1. Hepatitis B surface antigen (HBsAg) was not tested in the serum at that time. Percutaneous needle biopsy of the liver revealed the presence of a moderately dif- ferentiated hepatocellular carcinoma (Fig. 1 ). Tissue staining later by the aldehyde fuchsin method of Shikata3 was negative for HBsAg. The patient’s chest roentgenogram showed small nodular shadows in the left upper zone; they were compatible with the presence of metastatic tumors. Methyclothiazide was administered for ankle oedema. N o antitumor treatment was given. The patient was discharged from the hospital.

He continued working, although his symptoms progressed. In February 1966 there was an episode of bronchopneumonia. Maximal disturbance was noted in late March 1966 when he had fetor hepatis, jaundice, ascites, and palmar erythema. The biochemical features are summarized in Table 1, column 2. In addition, he had a persistent fever of 38°C. Liver biopsy confirmed the previous diagnosis of H C C (Fig. 2); culture of the tissue yielded a pure growth of paracolon bacillus that was sensitive to tetracycline and chloramphenicol. Tissue staining for HBsAg was again negative. Administration of tetracycline and later chloramphenicol and streptomycin failed to change the remittant pattern of fever: it persisted for ten days, sub- sided for five days, and then repeated in similar cycles for a total period of 1 Y2 months. At the end of this time, the patient felt much improved. The epigastric distension decreased.

Over the next six months, the liver decreased progressively in size. Fluid retention disappeared. H e informed us that he had been taking Chinese herbal medicine (vide infra). There was no change i n his food habit. SCOT was again elevated (Table 1 ). By mid-November 1966, finger clubbing and a spi- der angioma remained as the only abnormal physical findings. There was no clinical evidence of HCC. A chest roentgeno- gram was normal. Minimal biochemical abnormalities were

0008-543X/82/07 15/0332 $0.75 0 American Cancer Society

332

No. 2 SPONTANEOUS REGRESSION OF HCC . Lam et al. 333

still noticable (Table 1 , last column). Repeat liver biopsy from the location of the previous tumor near the site of the first biopsy showed only slight cellular irregularity and some in- crease in binucleated hepatocytes (Fig. 3). There was no dys- plasia. Tissue staining later for HBsAg on this biopsy was positive (Fig. 4).

Since then, the patient has remained fully functioning. There was no clinical or biochemical evidence of liver cirrhosis. A liver scintiscan with ‘I3In performed in 1971 showed the liver to be of normal size, shape and position; there was minor mottling, but the “hot spot” was preserved. (Fig. 5). Hepatic arteriography in 1972 showed no feature of residual HCC. When last assessed in late 1979, he was still working, asyrnp- tomatic, and apparently free from recurrence. HBsAg in serum was positive. His food habits were not noticeably dif- ferent from before.

The recipe of Chinese herbs that he took was subsequently tried by us on about 20 consecutive patients with HCC. No detectable regression of tumor was observed in any patient.

Discussion

Evidence in this patient satisfied all criteria of Ev- erson and Cole4 for complete regression of malignant tumors. It is unlikely, although possible, that the herbal preparation was responsible for the regression of the tumor since subsequent trials in other patients with HCC did not produce detectable benefits. A previously reported case of partial regression of HCC occurred following withdrawal of the androgen that was the ap- parent carcinogen.’ Induction of regression by with- holding environmental factors required for tumor growth is well recognized in such hormone-dependent tumors as mammary carcinoma.6 Better recognized for HCC is attempted total deprivation of oxygen and nutrients by dearterialization. While tumor outstripping its blood

TABLE 1. Serial Features in a Patient with Spontaneous Regression of Hepatocellular Carcinoma

Dates Nov. I965 March 1966 Aug. 1966 Nov. 1966

On presen- At greatest Shrinking After Events tation disturbance liver regression

Albumin, G/L (31- 51)’

Globulin. G/L (23- 40).

Bilirubin. umol/l (7-26).

Alk. phosphalase, umol/min.l. ( 3 5 - 100)’

SCOT. umol/min.l. (7.17).

SGPT. umol/min.l. (8-23).

Prothrombin level (100%)’

Liver size (cm) Below R costal

margin Below xyphi-

sternum Chest rocntgeno-

gram

28

34

20

I54

100

30

85.2%

6

10

Metastases

23

45

78

I42

55

95

71%

10

13 More

metastases

36

38

19

314

100

80

N A

4

6

Metastases

37

33

1 3

I88

40

35

896

I

I

Clear

The normal ranges are given in parentheses. N A = no1 available.

supply is quoted as a mechanism of spontaneous regres- sion in retinoblastoma,’ this mechanism has not been reported to produce complete regression of HCC. Con- sidering the hypervascularity of HCC8 we believe it is unlikely for spontaneous dearterialisation to be the mechanism for tumor regression in this patient.

Spontaneous regression of malignant tumors has been established to occur on average in about three patients in the world per year.’ Many tumors that regress have been subjected to some kind of interference.” This in-

FIG. 1. First liver biopsy showing mod- erately differentiated hepatocellular car- cinoma in trabecular pattern (H & E, X 160).

334 CANCER July 15 1982 Vol. 50

cludes incomplete removal or biopsy, palliative irradia- tion, and intercurrent febrile reactions. The exact trig- gering factor and the mechanism leading to regression is unknown. The first liver biopsy in this patient did not produce detectable regression. Fever is a common oc- currence in our patients with HCC", but in none other than this patient have we observed any significant tumor regression in the absence of cytotoxic therapy.

The processes of regression and their postregressional states vary from tumor to tumor. Histologic studies have shown that experimental epithelial tumors in regression do not convert progressively into normal tissue; instead, they disappear and are replaced by normal tissues.I2 In

FIG. 2. Second liver biopsy confirming the presence of hepatocellular carcinoma ( H & E, X400).

endocrine-dependent tumors, e.g., mammary carci- noma, regression leaves behind residual neoplasia which has the potential to redevelop into carcinoma when suit- able stimulus is applied.' Tumors of embryonic tissues, like neuroblastomas or teratomas, may become differ- entiated and mature out of neoplastic status.' In a more extreme view, WillisI3 was convinced that all tumors had finite life spans which, under special circumstances, might be shorter than that of their hosts. In the context of our patient, the features corresponding to residual neoplasia, premalignancy, carcinogenesis, and matu- ration are cellular dysplasia,I4 persistent hepatitis B virus (HBV) infection,15 exposure to food carc in~gens , '~

FIG. 3. Third liver biopsy taken after spontaneous regression of tumor from the site of the previous hepatocellular carci- noma. Only minor cellular irregularity is seen. There is neither tumor nor dysplasia ( H & E, X120).

No. 2 SPONTANEOUS REGRESSION OF HCC - Lam et al. 335

FIG. 4. HBsAg positive materials (ar- rowed) in the cytoplasm of hepatocytes in the third liver biopsy (Aldehyde fuchsin. X600).

and cellular differentiation. The postregression liver biopsy of this patient showed no evidence of cellular dysplasia. HBV infection was present. The role of food carcinogens was not objectively assessed, but since a change in food habit was denied, no evidence of alter- ation in exposure to food carcinogens was obtained. An association of liver shrinkage with a rise of SGOT was more consistent with tumor involution than maturation into normal tissue. We can find no previous report of spontaneous regression of HCC with which to compare

our findings, but the level of SGOT in our patient during the time of liver shrinkage was comparable to that re- ported for patients with liver cancer during hepatic ar- terial ligation and infusion of cytotoxic drugs.16

Taking all evidence together, it seems reasonable to speculate on the sequence of events in our patient as follows: Interaction of HBV infection with unknown carcinogens produced HCC prior to 1965. The tumor progressively enlarged and metastasized to the lungs. Infection of the tumor with paracolon bacillus or the

FIG. 5. Liver scintigram with "'In five years after spontaneous regression of the hepatocellular carci- noma.

336 CANCER July 15 1982 Vol. 50

HBV in 1966 triggered regression, due perhaps to ac- celerated preprogrammed death of tumor cells as sug- gested by Wil1is.l3 Tumor tissue degenerated in both liver and lung; elevated SGOT occurred simultaneously with shrinkage of liver. This happened over a period of about ten months from March-November 1966. Com- plete involution of carcinomatous cells was followed by replacement with “normal” hepatocytes. Increased to- tal functioning mass of hepatocytes augmented synthe- sis of albumin, thus reversing the hypoalbuminaemia. The residual “normal” cells had little malignant pre- disposition and so regeneration and subsequent survival did not produce another HCC in 13 years. Nevertheless, the remaining hepatocytes were still infected with HBV. Chronic type-B viral hepatitis produced residual bio- chemical and histologic abnormalities. The residual disease followed the typical course of chronic persistent hepatitis.”

Spontaneous regression with a subsequent normal life extends the known survival spectrum of HCC.’*2,’8 Oc- currence of spontaneous regression must be extremely rare. This report neither radically changes the general outlook of the disease nor its accepted therapeutic ap- proach. However, for the individual facing imminent death on diagnosis of unresectable HCC, even the re- motest possibility of cure may provide a thin ray of hope and a reason for pursuing specific or supportive therapy.

REFERENCES

1. Lai CL, Wu PC, Lam KC, et al. Histologic prognostic indicators in hepatocellular carcinoma. Cancer 1979; 44: 1677- 1683.

2. Okuda K. Clinical aspects of hepatocellular carcinoma-analysis of I34 cases. In: Okuda K, Peters RL, eds. Hepatocellular carcinoma. New York, John Wiley & Sons, 1976; 387-436.

3. Shikata, T, Uzawa T, Yoshiwara N, ef al. Staining methods of Australia antigen in paraffin section. Detection of cytoplasmic inclu- sion bodies. Jpn J Exp Med 1974; 44:15-35.

4. Everson TC, Cole WH. Spontaneous regression of cancer. Phil- adelphia; WB Saunders, 1966.

5. Johnson FL, Feagler JR, Lerner KG, ef al. Association of an- drogenic anabolic steroid therapy with development of hepatocellular carcinoma. Lancet 1972; 2:1273-1276.

6 . Smithers DW, Rigby-Jones P, Galton DAG, Payne PM. Cancer of the breast: A review. Br J Radio1 Suppl 1952; 4:l-90.

7. Snell AC, Salazar OM, Orbison L. Tumors of the eye. In: Rubin P, Bakemeier RF, eds. Clinical oncology, ed. 5. Rochester, New York, American Cancer Society, 1978; 191-202.

8. Nakashima T. Vascular changes and hemodynamics in hepa- tocellular carcinoma, In: Okuda K, Peters RL, eds. Hepatocellular carcinoma. New York; Wiley & Sons, 1976; 169-203.

9. Foulds L. Characteristics of neoplasms. In: Foulds L, ed. Neo- plastic development 1. London; Academic Press, 1969; 97-1 36.

10. Boyd WL. The spontaneous regression of cancer. Springfield; Illinois, Charles C. Thomas, 1966.

I I . Lai CL. Lam KC, Wong KP, et 01. Clinical features of he- patocellular carcinoma: Review of 21 I patients in Hong Kong. Cancer 198 1; 47:2746-2755.

12. Rous P, Kidd JG. Conditional neoplasms and subthreshold neoplastic states. A study of the tar tumours of rabbits. J Exp Med 1941; 73: 365-389.

13. Willis RA. The pathology of tumours, ed. 4. London, Butter- worths, 1967.

14. Anthony PP, Vogel CL, Barker LF. Liver cell dysplasia: A premalignant condition. J CIin Pathol 1973; 26: 217-223.

15. Armstrong B. The epidemiology of cancer in the People’s Re- public of China. Internatl J Epiderniol 1980 (in press).

16. Murray-Lyon IM, Dawson JL, Parsons VA, et al. Treatment of secondary hepatic tumours by ligation of hepatic artery and infusion of cytotoxic drugs. Lancet 1970 2: 172- 175.

17. Becker MD, Scheuer PJ, Baptista A. et al. Prognosis of chronic persistent hepatitis. Lancet 1970; 1:53-56.

18. Kubo Y, Shimokawa Y. Arterial injection chemotherapy. In: Okuda K, Peters RL, eds. Hepatocellular carcinoma. New York, John Wiley and Sons, 1976; 477-490.