Dramatic response to fourth-line ramucirumab after atezolizumab plus bevacizumab failure in advanced hepatocellular carcinoma: A case report

Abstract

BACKGROUND

Systemic treatment options for hepatocellular carcinoma (HCC) following atezolizumab plus bevacizumab (Atez/Bev) failure are limited, particularly for patients with deteriorated liver function and poor tolerance to tyrosine kinase inhibitors (TKIs). This case highlights the potential role of ramucirumab (RAM) as later-line therapy in challenging clinical settings.

CASE SUMMARY

An 83-year-old man with a history of hepatitis C-related HCC developed disseminated recurrence involving the splenic hilum 12 years after curative resection. The lesion was surgically resected following four courses of Atez/Bev. After 17 months, intrahepatic and liver surface recurrences were detected. Atez/Bev was reintroduced as a first-line therapy but resulted in disease progression after 17 courses. Second-line lenvatinib and third-line sorafenib were discontinued because of severe fatigue, appetite loss, marked tumor progression, and worsening liver function (Child-Pugh B). Given the patient’s elevated alpha-fetoprotein (AFP) level and intolerance to TKIs, RAM was initiated as fourth-line therapy. Tumor marker levels rapidly normalized, and imaging demonstrated a partial response. Liver function also improved, and the response was maintained for over 1 year with manageable toxicity.

CONCLUSION

Following Atez/Bev failure, RAM may provide mechanistically rational and effective therapy in AFP-high HCC for patients intolerant to kinase inhibitor.

Key Words: Alpha-fetoprotein; Atezolizumab plus bevacizumab; Child-Pugh B; Hepatocellular carcinoma; Later-line therapy; Ramucirumab; Tyrosine kinase inhibitor intolerance; Case report

Core Tip: Hepatocellular carcinoma with elevated alpha-fetoprotein levels that progresses after treatment with atezolizumab plus bevacizumab may involve enhanced vascular endothelial growth factor receptor-2 signaling, making ramucirumab a potentially effective treatment option.

INTRODUCTION

In Japan, there are currently six available regimens as first-line drug therapies for hepatocellular carcinoma (HCC), including immune checkpoint inhibitor regimens such as nivolumab plus ipilimumab, tremelimumab plus durvalumab, atezolizumab plus bevacizumab (Atez/Bev), and durvalumab monotherapy, as well as the multi-kinase inhibitors sorafenib and lenvatinib (LEN)[1]. Regorafenib, ramucirumab (RAM), and cabozantinib can also be used as second-line and subsequent drug therapies, with nine regimens available. Therefore, it is necessary to select the appropriate medication for each patient. RAM can extend overall survival compared with placebo in patients with HCC and alpha-fetoprotein (AFP) levels ≥ 400 ng/mL who had a history of sorafenib treatment. In the REACH-2 trial, RAM showed an overall response rate of 4.6% (combining complete and partial responses) and a disease control rate of 59.9% when patients with stable diseases were included[2]. Therefore, rather than expecting significant tumor shrinkage, RAM is often used in second- or later-line treatment, with the goal of disease control. Herein, we report a case in which RAM administered as a fourth-line treatment produced remarkable therapeutic effects.

CASE PRESENTATION

Chief complaints

An 83-year-old man experienced HCC-related recurrence in both the intrahepatic and liver surface areas and was referred to our department.

History of present illness

Upon referral to our hospital at 68 years old for HCC and chronic hepatitis C, a 10 mm HCC was identified in liver segment 3, and partial resection was performed. Pathologically, the tumor was a well-differentiated HCC. After HCC treatment, a sustained response to hepatitis C was achieved with pegylated interferon plus ribavirin therapy. Twelve years later, at age 80 years old, a single 20 mm mass was identified at the splenic hilum. The tumor markers for HCC, namely, serum AFP and des-gamma carboxyprothrombin (DCP), were elevated, and recurrence due to HCC-related dissemination was suspected. Because disseminated recurrence was considered, a total of four courses of combination therapy with Atez (1200 mg)/Bev (15 mg/kg) were administered every 3 weeks. After confirming that there were no recurrences at other sites, the tumor and spleen were surgically resected. Seventeen months later, the patient experienced recurrence in both the intrahepatic and liver surface areas and was referred to our department.

History of past illness

The patient achieved a sustained virologic response for hepatitis C with pegylated interferon plus ribavirin therapy at 68 years of age.

Personal and family history

The patient had no significant family history of malignancy or known genetic predisposition, nor did he have any significant history of environmental exposure.

Physical examination

No notable findings were observed on physical examination.

Laboratory examinations

Blood test results at the time of initial referral were as follows: Platelet count, 17.6 × 10⁴/μL; total bilirubin, 1.0 mg/dL (normal range: 0.4-1.5 mg/dL); albumin, 3.8 g/dL (normal range: 4.1-5.1 g/dL); prothrombin time (PT), 76% (PT-international normalized ratio: 1.14); Child-Pugh score, 5 (Class A), and albumin-bilirubin (ALBI) score, -2.42. His serum AFP and DCP levels were 99.5 ng/mL and 1897 mAU/mL, respectively.

Imaging examinations

Contrast-enhanced computed tomography (CT) revealed two intrahepatic tumors measuring 10 mm and 18 mm in maximum diameter, both showing early arterial enhancement and late-phase washout. In addition, an extrahepatic lesion with a maximum diameter of 28 mm was identified on the liver surface. Given the elevated AFP and DCP serum levels and the history of treatment for splenic hilum area seeding in HCC, HCC recurrence was suspected.

FINAL DIAGNOSIS

Recurrence of HCC disseminated nodules on the liver surface.

TREATMENT

With a Child-Pugh score of 5 and a performance status of 1, Atez/Bev was initiated as first-line therapy. Tumor markers gradually increased during treatment, and after 17 courses, contrast-enhanced CT demonstrated growth of the intrahepatic target lesions from a total of 28 mm to 43 mm, along with growth of the extrahepatic lesion from 28 mm to 35 mm in maximum diameter, leading to a diagnosis of progressive disease according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. At this time, the Child-Pugh score was 7 points (Class B) and the ALBI score was -1.88 (mALBI grade 2b), indicating decreased hepatic reserve. LEN (8 mg/day) was started as second-line therapy but discontinued 3 weeks later because of adverse events (grade 3 fatigue and grade 2 anorexia, according to CTCAE version 5.0). After the symptoms subsided, based on the second-line treatment course, it was determined that the patient had low tolerance to tyrosine kinase inhibitors (TKIs). Therefore, third-line treatment was initiated with a low dose of sorafenib (200 mg/day). However, this was discontinued after 4 months because of fatigue and anorexia. During this time, serum AFP and DCP levels increased to 714 ng/mL and > 75000 mAU/mL, respectively, with a marked increase in tumor volume (Figure 1). The patient's overall condition had deteriorated, with a Child-Pugh score of 7 (Class B), an ALBI score of -1.32 (mALBI grade 3), and a performance status of 1. However, after careful discussion with the patient and his family confirming his willingness to receive treatment, we considered a fourth-line treatment. Given his tolerance to LEN and sorafenib after a poor response to Atez/Bev, we selected RAM as a fourth-line treatment as it is well-tolerated. RAM (8 mg/kg) was administered every 2 weeks. Tumor markers began to decrease after eight courses (AFP, 2.9 ng/mL; DCP, 63 mAU/mL). By the twelfth course, contrast-enhanced magnetic resonance imaging (MRI) demonstrated that the sum of the longest diameters of the intrahepatic target lesions had decreased from 95 mm (35 mm and 60 mm) to 40 mm (17 mm and 23 mm), representing a 61% reduction. The longest diameter of the extrahepatic lesion also markedly decreased from 108 mm to 28 mm. These findings fulfilled the criteria for a partial response according to the RECIST version 1.1 (Figure 2).

Figure 1 Progression chart based on tumor markers and magnetic resonance imaging findings. Atezolizumab plus bevacizumab therapy (every 3 weeks) was administered for 12 months, and progressive disease (PD) was determined. Lenvatinib was discontinued after 3 weeks due to adverse events, and sorafenib was administered for 4 months but was discontinued because of adverse events and PD. The image shows the late phase of contrast-enhanced magnetic resonance imaging. The tumor gradually increased in size. Tumor response was evaluated according to the Response Evaluation Criteria in Solid Tumors version 1.1. The tumors are indicated by white arrowheads. The maximum tumor diameters are indicated by white lines, with the corresponding measurements (mm) shown. Atez/Bev: Atezolizumab plus bevacizumab; LEN: Lenvatinib; SOR: Sorafenib; PD: Progressive disease; q3w: Every 3 weeks; DCP: Des-gamma carboxyprothrombin; AFP: Alpha-fetoprotein; ALBI: Albumin-bilirubin.

Figure 2 Progression chart during ramucirumab administration. Ramucirumab (RAM) was initiated as a fourth-line treatment 20 months after the initiation of systemic therapy. Magnetic resonance imaging performed 6 months after RAM administration revealed marked tumor shrinkage. The sum of the longest diameters of the two target hepatic lesions decreased from 95 mm to 40 mm, while the longest diameter of the extrahepatic lesion decreased from 108 mm to 28 mm, corresponding to a partial response according to the Response Evaluation Criteria in Solid Tumors version 1.1. The tumors are indicated by white arrowheads. The maximum tumor diameters are indicated by white lines, with the corresponding measurements (mm) shown. PR: Partial response; q2w: Every 2 weeks; q3w: Every 3 weeks; RAM: Ramucirumab; SD: Stable disease; DCP: Des-gamma carboxyprothrombin; AFP: Alpha-fetoprotein; ALBI: Albumin-bilirubin; SOR: Sorafenib.

OUTCOME AND FOLLOW-UP

After 14 courses of RAM (7 months after initiation), the patient developed grade 2 fatigue, resulting in treatment interruption for approximately 2 months. RAM was subsequently resumed with a modified schedule of administration every 3 weeks. To date, treatment has been continued for more than 14 months, and a partial response has been maintained. Concomitantly, liver function improved to a Child-Pugh score of 6 (Class A), an ALBI score of -2.09 (mALBI grade 2b), and a performance status of 1 (Table 1).

Table 1 Laboratory data at each treatment initiation and 14 months after ramucirumab.

Parameter	Reference range	Atez/Bev	LEN	SOR	RAM	RAM (14 months)
WBC (μL)	3300-8600	4900	3500	6100	5300	6000
Hb (g/dL)	13.7-16.8	14.5	13.7	13.9	13.4	11.8
Plt (× 104/μL)	15.8-34.8	19.2	22.7	22.3	29.2	14.4
PT-INR	1.00	1.14	1.23	1.07	1.19	0.98
TBIL (mg/dL)	0.4-1.5	1.0	1.1	0.9	1.3	0.7
Alb (g/dL)	4.1-5.1	3.8	3.2	3.4	2.6	3.3
AST (U/L)	13-30	33	37	47	65	42
ALT (U/L)	10-42	25	18	47	26	15
ALP (U/L)	38-113	145	96	153	194	137
GGT (U/L)	13-64	77	116	195	213	108
Child-Pugh score		5	7	6	7	6
ALBI score		-2.42	-1.88	-2.11	-1.32	-2.09
AFP (ng/mL)	< 7	99.5	260	692	714	2.3
DCP (mAU/mL)	< 28	1897	34797	19206	> 75000	161

WBC: White blood cell; Hb: Hemoglobin; Plt: Platelet; PT-INR: Prothrombin time-international normalized ratio; TBIL: Total bilirubin; Alb: Albumin; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; GGT: Gamma-glutamyl transferase; ALBI: Albumin-bilirubin; AFP: Alpha-fetoprotein; DCP: Des-gamma carboxy prothrombin; Atez/Bev: Atezolizumab plus bevacizumab; LEN: Lenvatinib; SOR: Sorafenib; RAM: Ramucirumab.

DISCUSSION

We encountered a patient for whom RAM, administered as fourth-line therapy, yielded a remarkable effect. The patient showed slow progressive disease with Atez/Bev, intolerance to LEN, and progressive disease even with a low dose of sorafenib. However, fourth-line RAM was well tolerated, demonstrating efficacy and extended survival. This case involved an older adult patient with a Child-Pugh score of 7 (Class B), indicating decreased liver functional reserve. However, RAM was used safely.

The patient developed disseminated recurrence in the splenic hilum, leading to recurrence in the liver and on the liver surface. Recurrence in the splenic hilum was surgically treated after four courses of Atez/Bev; however, pathology findings showed that 90% of the HCC was viable, indicating that Atez/Bev was largely ineffective. Therefore, it was deemed highly likely that the subsequent recurrent lesions shared similar characteristics with HCC. The patient did not respond further to Atez/Bev. Therefore, we did not combine durvalumab and tremelimumab or ipilimumab and nivolumab as second-line therapy. Moreover, out of concern for immune-related adverse events associated with these agents, we chose to administer a TKI. Treatment with LEN markedly decreased tumor marker levels, suggesting that treatment may have been effective if tolerability had not been an issue. However, sorafenib was ineffective, possibly because of the low dose used.

Vascular endothelial growth factor (VEGF)-A, released by tumor cells and vascular endothelial cells at the tumor site, activates and recruits numerous types of inhibitory immune cells, including tumor-associated macrophages, regulatory T cells, and myeloid-derived suppressor cells, creating an immune-inhibitory microenvironment that favors tumor immune escape[3]. Bev inhibits VEGF-A, thereby improving the immunosuppressive microenvironment and enhancing the therapeutic effect when used in combination with immune checkpoint inhibitors. VEGF-A is also a central regulator of tumor angiogenesis, endothelial proliferation, permeability, and survival. VEGF-A binds with high affinity to two structurally similar tyrosine kinase receptors, VEGFR-1 and VEGFR-2, both of which are expressed in the tumor vasculature[4]. Thus, the inhibition of angiogenesis by Bev also has antitumor effects. Additionally, LEN and sorafenib exert antitumor effects by inhibiting multiple tyrosine kinases involved in angiogenesis in addition to VEGFR-1 and VEGFR-2, including PDGFRβ. LEN has a stronger affinity for receptors than sorafenib[5,6]. Yang et al[7] reported that Bev inhibition of VEGF-A leads to a compensatory increase in VEGF-C and VEGF-D, which act on VEGFR-2, resulting in reactivation of VEGFR-2-mediated signaling and potentially promoting tumor progression. RAM is a monoclonal antibody that acts specifically on VEGFR-2[4]; in this case, RAM was considered because enhanced VEGFR-2 signaling is the main driver of tumor growth. This was further supported by the observed decrease in tumor markers with LEN treatment, which strongly inhibited VEGFR-1 and VEGFR-2.

Approximately 10 months passed between the final administration of Atez/Bev and RAM administration. Atez’s half-life is reported to be 27 days[8]; therefore, we deemed it unlikely that an enhanced therapeutic effect owing to a pseudo-combination effect with Atez at the time of RAM administration applied. Regarding the therapeutic effect of RAM after Bev, Shimose et al[9] reported that RAM after Atez/Bev treatment had better progression-free survival than other combinations. These results support the model in which RAM is effective after Bev because of increased VEGFR-2 signaling[7].

AFP-high tumors exhibit a distinct phenotype characterized by poor differentiation, enrichment of progenitor features, and enhanced proliferation. Montal et al[10] reported that, VEGF-B is overexpressed and preferentially occupies VEGFR-1 in AFP-high HCC, resulting in a decreased decoy function of VEGFR-1 and a relative increase in VEGF-A binding and signaling through VEGFR-2.

Therefore, in addition to the compensatory increase in VEGF-C/D associated with resistance to Atez/Bev therapy, a phenotypic shift in the tumor to an AFP-high phenotype may occur. Against the background of increased VEGF-B, it is possible that VEGFR-2 signaling, re-induced by VEGF-A, is further enhanced. While VEGFR-1 has a high affinity for VEGF-A, its kinase activity is relatively weak and it primarily functions as a regulatory receptor that modulates VEGFR-2–mediated angiogenic signaling. Under such circumstances, where VEGFR-2–dependent angiogenesis becomes predominant, RAM acts effectively through direct inhibition of VEGFR-2.

CONCLUSION

In conclusion, in patients with high AFP levels following Atez/Bev treatment, the VEGFR-2 pathway may be activated. For cases in which tolerability to TKIs such as LEN, which also has strong VEGFR-2 inhibitory effects, cannot be ensured, there is value in proactively considering RAM as a therapeutic option.