Atezolizumab and Bevacizumab Combination Therapy and Sequential Conversion Hepatectomy for Advanced Fibrolamellar Hepatocellular Carcinoma Presenting Pseudoprogression

Dear Editor,

Fibrolamellar hepatocellular carcinoma (FLHCC) is a subtype of hepatocellular carcinoma (HCC) accounting for 0.6–8.6% of all HCCs [1], typically developing in young adults without underlying hepatitis or cirrhosis. The incidence of lymph node (LN) metastasis is 1.6% in conventional HCC, while it is 40% in FLHCC, and the treatment strategy for FLHCC with LN metastasis remains unclear [2]. Systemic chemotherapy might be an option in patients with advanced FLHCC; however, FLHCC cases were excluded from the recent clinical trial (IMbrave150) showing a better survival of atezolizumab and bevacizumab combination therapy (ABT) than that of sorafenib [3]. We report the first case of ABT for advanced FLHCC followed by hepatectomy presenting pseudoprogression of LN metastases which were pathologically proven after surgery.

The patient was a thirty-year-old man without a history of liver disease. Dynamic computed tomography (CT) showed a hepatic tumor in segment 7 and segment 8 measuring 9 cm in diameter, showing strong enhancement in the arterial phase and washed out in the portal phase with central calcification (Fig. 1a), and multiple LN metastases behind the pancreatic head sized maximally 3 cm in diameter (Fig. 1b). Laboratory examination showed a slight elevation of the serum α-fetoprotein level, 9.8 ng/mL. Percutaneous biopsy from the liver lesion revealed polygonal tumor cells with an eosinophilic cytoplasm surrounded by abundant fibrous stromatolites with a lamellar distribution, and reverse transcription polymerase chain reaction showed positive for DNAJB1-PRKACA, which is a fusion gene. At the diagnosis of advanced FLHCC (Barcelona Clinic Liver Cancer, Stage C), ABT once every 3 weeks was administered after obtaining informed consent on the possible benefits and risks of ABT for FLHCC because FLHCC was excluded from IMbrave150. After four courses of ABT, dynamic CT showed a faint decrease in the intratumor vascularity of the liver tumor, and laboratory examination showed normalization of the serum α-fetoprotein level. However, the size of metastatic LNs increased within stable disease in RECIST 1.1 (Fig. 1c).

Fig. 1.

a Dynamic CT of the hepatic tumor before chemotherapy. A hepatic tumor sized 9 cm in diameter in segment 7 and segment 8, showed strong enhancement in the arterial phase with central calcification. b Dynamic CT of the peripancreatic LN before chemotherapy. Peripancreatic LN sized 3 cm in diameter showed enhancement in the arterial phase, suggesting metastasis (arrow head). c Dynamic CT of the peripancreatic LN after 4 courses of the ACT. The size of the metastatic LN increased to 3.7 cm in diameter. d, e Dynamic CT 2 weeks after percutaneous transhepatic portal vein embolization. The intratumor vascularity of the liver tumor was reduced remarkably. The size of the peripancreatic LNs increased to 4.4 cm; however, intranodal vascularity markedly decreased.

At the moment, the patient admitted bloody stool and colonoscopy revealed immune-related colitis caused by atezolizumab; therefore, the fifth course was canceled. We decided to undergo conversion surgery after recovery from the colitis to rescue the patient. The scheduled surgery was right hemihepatectomy plus nodal dissection after percutaneous transhepatic portal vein embolization (PTPE). Two weeks after PTPE (20 weeks after ABT initiation and 8 weeks after ABT discontinuation), dynamic CT showed an objective response to ABT (Fig. 1d, e) and stable disease in RECIST 1.1 (7% increase in the LN size and no change of liver tumor) and partial response in modified RECIST (47% decrease of intratumor vascularity of the liver tumor and LNs). After adequate hypertrophy of the left hemiliver 3 weeks after PTPE, right hemihepatectomy plus nodal dissection was successfully performed. The operation time was 7 h, and the estimated blood loss was 700 mL. Pathological findings showed approximately 60–70% of necrosis of liver tumor and 70–80% of metastatic LNs, showing a good response to ABT (Fig. 2a-d). The patient was discharged on day 15 and is doing well without any sign of recurrence 5 months after conversion surgery.

Fig. 2.

a, b Macroscopic findings of the liver tumor (a) and metastatic LN (b). a The liver tumor was the confluent multinodular type, and the center of the tumor was associated with scarring and calcification. Scattered areas of necrosis were seen within the tumor. b Most areas in the metastatic LNs were necrotic. c Microscopic findings of the liver tumor. In microscopic findings, the liver tumor was composed of polygonal tumor cells with an eosinophilic cytoplasm, surrounded by abundant fibrous stroma with a lamellar distribution. Hyalinization and necrosis were observed in part of the tumors, suggesting the therapeutic effect of the ACT. d Microscopic findings of the metastatic LN. LN metastasis of fibrolamellar HCC was seen. The majority of the tumor was necrotic.

To the best of our knowledge, the present case is the first that the objective response of FLHCC to ABT for FLHCC and conversion surgery was successfully performed to eradicate the primary tumor and the peripancreatic LNs which showed pseudoprogression induced by ABT. The first response to metastatic LNs was observed 20 weeks after initiation of ABT, and the metastatic LNs showed both increase in nodal volume and a decrease in vascularity. Apparent tumor growth associated with edema and infiltration of immune cells treated with an immune checkpoint inhibitor is known as pseudoprogression [4]. Pseudoprogression can be found in the patients with solid malignancies undergoing immune checkpoint inhibitor treatment; however, it rarely occurs in HCC. In the updated data of the IMbrave150 trial, 19% of first responses occurred after week 24 [5]. Thus, as in the present case, it is sometimes difficult to differentiate pseudoprogression from general tumor progression in patients undergoing ABT. Physicians should consider that ABT may also be effective in FLHCC and may cause pseudoprogression before determining a treatment strategy.

Conflict of Interest Statement

Naohiro Okano has received honoraria from Taiho Pharmaceutical, Eli Lilly Japan, Chugai Pharma, and Ono Pharmaceutical. Other authors declare no conflicts of interest associated with this manuscript.

Funding Sources

There were no funding sources for this manuscript.

Author Contributions

Clinical diagnosis and liver biopsy: Hideaki Mori. ACT: Naohiro Okano and Fumio Nagashima. Operation and perioperative management: Ryota Matsuki, Nobuhiro Hasui, Shohei Kawaguchi, Hirokazu Momose, Masaharu Kogure, Yutaka Suzuki, and Yoshihiro Sakamoto. Pathological diagnosis: Keiichiro Kitahama, Kiyotaka Nagahama, and Junji Shibahara. Manuscript writing: Ryota Matsuki. All the authors have seen and approved the final version of the manuscript being submitted.

Funding Statement

There were no funding sources for this manuscript.