Spontaneous regression of hepatocellular carcinoma: myth or reality?

Abstract

We present a case of a 68-year-old man with chronic hepatitis C infection, with no evidence of chronic liver disease during the first years of follow-up, diagnosed with a hepatocellular carcinoma (HCC) with 40 mm (α-fetoprotein (AFP) 205 ng/mL). He underwent segmental liver resection and pathology analysis was consistent with HCC and cirrhosis in the adjacent liver. Four months after surgery, AFP raised up to 126 661 ng/mL and abdominal MRI revealed a multinodular HCC. Patient rejected treatment with sorafenib and started megestrol and an herbal medicine, soursop (Annona muricata). Six months later, AFP markedly decreased (28 ng/mL) and abdominal MRI showed decreasing size and number of lesions. At 5 years of follow-up, he has no evidence of HCC. Spontaneous regression of HCC is a rare condition and the underlying mechanism is unclear. In this case there is a temporal relation between the start of megestrol and Annona muricata and HCC regression.

Background

Hepatocellular carcinoma (HCC) is the most frequent primary liver cancer and one of the leading causes of cancer-related death worldwide.^{1 2} Despite recent advances, there are still limited options for treatment advanced HCC.^{2 3} Spontaneous regression of HCC is a rare condition defined as partial or complete disappearance of the tumour in the absence of any specific treatment.³ In this report, we describe a case of regression of postoperative HCC recurrence without any systemic therapy; patient was only medicated with megestrol for anorexia and soursop (Annona muricata) on his own.

Case presentation

A 68-year-old Caucasian man presented with chronic hepatitis C virus (HCV) infection, genotype 1b, in the context of blood transfusion in 1977. He had a recent history of prostatic adenocarcinoma and underwent surgery and radiotherapy. Patient was not under any medication, had no history of smoking, alcohol consumption or intravenous drug use. During the first 8 years of follow-up, between 2001 and 2009, he was asymptomatic and had no evidence of advanced fibrosis (liver stiffness of 4.5 kPa by transient elastography in 2009). He was not eligible for hepatitis C treatment at that time. Patient was then lost to follow-up for 3 years and returned in 2013. At that time, he had a rise in alpha-fetoprotein (AFP 208 ng/mL) and abdominal ultrasonography revealed a suspected lesion in the liver. Abdominal CT showed a solid lesion with 40×40 mm in segment IV with arterial enhancement and ‘washout’ in portal venous phase and hypertrophy of the caudate lobe suggestive of early-stage chronic liver disease (figure 1). Patient underwent segmental liver resection and pathology analysis was consistent with trabecular HCC with vascular invasion and cirrhosis in the adjacent liver. In the postoperative period AFP decreased to 1.3 ng/mL.

Figure 1.

Abdominal CT after detection of suspected lesion in the liver in abdominal ultrasonography showed a solid lesion with 40×40 mm in segment IV with arterial enhancement and ‘washout’ in portal venous phase.

Outcome and follow-up

Four months after surgery, patient was asymptomatic, had an Eastern Cooperative Oncology Group performance score of 0, was a Child-Pugh A5 and had a MELD score of 8; however, there was an increase in AFP (3614 ng/mL) and abdominal MRI revealed multiple focal solid lesions in the right lobe of the liver, the largest with 30×30 mm, with typical features of HCC (figure 2). Hence, an advanced stage HCC was diagnosed. Patient and his family refused treatment with sorafenib; he started on megestrol for anorexia and self-medicated with a herbal medicine, soursop (A. muricata). Seven months after surgery, AFP peaked at 126 661 ng/mL. Three months later, AFP markedly decreased (28 ng/mL) and abdominal MRI showed decreasing size, number and vascularisation of lesions previously described (figure 3). One year after HCC recurrence, AFP was 2 ng/mL and abdominal CT showed residual lesions in the liver. At this time, patient started sofosbuvir/ledipasvir for 24 weeks, with sustained virologic response 24 weeks after the end of treatment. Five years later, patient remains asymptomatic, with well-compensated cirrhosis, an AFP of 1.3 ng/mL and no has evidence of HCC in abdominal ultrasonography.

Figure 2.

First abdominal MRI performed 4 months after surgery to clarify α-fetoprotein increase showed multiple focal solid lesions in the right lobe of the liver, the largest with 30×30 mm, with typical features of hepatocellular carcinoma.

Figure 3.

Second abdominal MRI performed after α-fetoprotein reduction, 6 months after the first MRI, revealed decreasing size, number and lesions vascularisation compared with figure 2.

Discussion

Herein, we describe a case of spontaneous regression of HCC. Although there is no histological confirmation of postoperative HCC recurrence, AFP raised up to 126 661 ng/mL and MRI showed typical features of HCC in a cirrhotic liver. Besides, there is a previous surgical specimen with evidence of cirrhosis and HCC.

Spontaneous regression of HCC was first described by Eversion and Cole in 1956 and is defined as partial or complete disappearance of the tumour in the absence of any specific treatment.³ It is a rare condition with an estimated incidence of 0.4%.⁴ The mechanism underlying it is unclear but may include tumour hypoxia and systemic inflammatory response.⁵ Several factors have been proposed for HCC regression, including alcohol abstinence, use of herbal medicine and withdrawal of anabolic steroids.⁵

In this case report, patient started megestrol for anorexia and self-medicated with soursop (A. muricata) 7 months before HCC regression. We hypothesise that tumour regression in this patient was due to enhanced antitumoral immunogenicity which could eventually be related with the effects of hormonal and/or herbal therapy.

There are increasing evidence of immunogenicity in HCC such as lower recurrence rates after liver transplantation in patients with high CD4:CD8 ratio, higher overall survival in patients with HCC with low intratumoral T-regulatory lymphocytes level and high activated CD8+ cytotoxic T cells (CTLs), and high prevalence of tumor-associated antigens recognised by CTLs in patients with HCC. However, this antitumor immunity is usually suppressed by tolerance mechanisms, namely upregulation of inhibitory immune checkpoints (eg, CTLA-2 and programmed cell death protein 1 (PD-1)) that prevent the activation of T cells.⁶

Previous cases of HCC spontaneous regression suggested the potential role of inflammatory response in this phenomenon. Abiru et al reported three cases of HCC regression in HCV-infected patients with increased serum interleukin (IL)-18 levels which may induce cytotoxic activity of T cells and natural killer (NK) cells against the tumour.⁷ Jozuka et al also detected high levels of NK cell activity and cytokines (IL-2, IL-6, IL-12 and interferon-γ) in an HCV-patient under antidepressant drugs.⁸

Spontaneous regressions of other tumours potentially related to tumour immunogenicity have also been described, namely neuroblastoma,⁹ melanoma,¹⁰ pancreatic¹¹ and rectal¹² cancer. The former has a unique propensity for spontaneous regression which can partially be explained by immunological mechanisms such as upregulation of human leucocyte antigen class I expression mediated by interferon-γ which increases the recognition of tumour cells by cytotoxic T cells.⁹

Some studies suggest that A. muricata compounds have cytotoxic activity against liver cancer cells.¹³ Annonaceous acetogenins, the main bioactive component extracted from A. muricata, induced apoptosis in vitro¹⁴ and inhibited the growth of human hepatoma cells in animal models.¹⁵ More recent studies revealed that ethanol extract of A. muricata leaves causes apoptosis of liver cancer cells through endoplasmatic reticulum stress pathway.¹⁶ Megestrol, a synthetic derivative of progesterone, is usually used as palliation therapy in breast and endometrial cancer and for anorexia and cachexia in advanced cancer.¹⁷ Several studies suggest an association between sex hormones and HCC.^{18 19} However, currently published guidelines do not recommend the use of hormonal therapy on HCC, namely tamoxifen, since there is no evidence of survival benefits.¹ Tamoxifen is an antioestrogen hormone that acts at the receptor level, competing for the oestrogen-binding site.²⁰ The occurrence of oestrogen receptor mutations may justify its lack of efficacy in HCC treatment.^{20 21} In contrast, megestetrol acts at a postreceptor level which might explain its favourable outcomes in some studies.^{20 22} In a randomised controlled trial (RCT) including 45 patients with HCC characterised by variant liver oestrogen receptors, megestrol slowed down tumour growth (median time to progression of 22 vs 9 months, p=0.0212) and improved survival (18 vs 9 months, p=0.0090) as compared with placebo.²⁰ However, a more recent RCT with 204 patients with advanced HCC found no survival benefits with megestrol.²³ These contradictory results might be explained by different inclusion criteria and preclude any definitive conclusion on the effectiveness of megestrol in patients with HCC.

In this case report, there is a clear temporal relation between the start of megestrol and A. muricata and HCC regression. Although we cannot assume that any of these drugs induced tumour regression, these findings are encouraging to pursue investigations on these drugs for treatment of advanced HCC. Furthermore, better understanding of the immunology of spontaneous regression may help the evolution of targeted immunotherapeutic approaches.

Learning points.

• Spontaneous regression of hepatocellular carcinoma (HCC) is a rare condition; still the estimated incidence is 0.4%.

• Understanding the mechanisms of spontaneous regression of HCC may provide insights for new treatments for advanced HCC.

• Megestrol and Annona muricata might have beneficial effects in patients with advanced HCC.