Abstract
Introduction: In the phase 3 RESORCE trial, regorafenib prolonged overall survival (OS) in patients with unresectable hepatocellular carcinoma (uHCC) whose disease progressed on prior sorafenib. The prospective, observational REFINE study aimed to evaluate the safety and effectiveness of regorafenib in a broader population of patients in real-world clinical practice, including patients with Eastern Cooperative Oncology Group performance status (ECOG PS) ≥2, Child-Pugh B liver status, and sorafenib intolerance. Methods: This international, prospective, multicenter study (NCT03289273) enrolled patients with uHCC for whom the decision to treat with regorafenib was made by their physician before enrollment, according to the local health authority-approved label. The primary aim was to evaluate the safety of regorafenib, including the incidence of treatment-emergent adverse events (TEAEs) and dose modifications due to TEAEs. Results: Of the 1,028 patients enrolled, 1,005 initiated regorafenib and were eligible for analysis. Median age was 66 years (range 21–94); most patients were male (83%), Child-Pugh A (61%), and had an ECOG PS of 0 or 1 (82%) at study entry. Overall, 47%, 11%, and 40% of patients initiated regorafenib at 160, 120, and 80 mg/day, respectively. Median treatment duration was 3.7 months (range 1 day to 38.9 months). Dose modifications and permanent discontinuation of regorafenib due to TEAEs occurred in 45% and 31% of patients, respectively. The most common drug-related TEAEs were hand-foot skin reaction (31%), diarrhea (26%), and fatigue (15%). Median OS was 13.2 months (95% confidence interval 11.6, 14.8). Conclusion: The results of the real-world REFINE study confirmed the safety and effectiveness of regorafenib in a broad population of patients with uHCC. Of patients who received standard regorafenib dosing in REFINE, safety and efficacy findings were consistent with those reported in the RESORCE trial.
Introduction
Primary liver cancer, with hepatocellular carcinoma (HCC) accounting for approximately 90% of cases, remains a global health challenge [1]; it was the sixth most commonly diagnosed cancer and the third leading cause of cancer-related mortality worldwide in 2020, with an estimated 906,000 new cases and 830,000 mortalities in the same year [2]. More recent American Cancer Society statistics for liver cancer (2023) reveal an estimated 41,210 new cases and 29,380 mortalities in the US population alone [3]. Following decades of rapidly rising mortality rates, the latest estimates (2017–2020) show that rates have now stabilized in women and declined in men, mirroring patterns in incidence [3], which together highlight the impact of improved treatment. Despite this, the burden of liver cancer varies across the world and is predicted to increase globally by approximately 55% between 2020 and 2040 [4].
The treatment landscape in HCC has changed remarkably over the past 2 decades, following the initial approval of the tyrosine kinase inhibitor (TKI), sorafenib in 2007, as the first-line standard of care for the treatment of advanced HCC [1, 5]. Newer treatments for first- and second-line settings include immunotherapy (e.g., anti-programmed cell death protein-1/death-ligand 1 antibodies) alone or in combination with other agents (e.g., bevacizumab [an anti-vascular endothelial growth factor antibody] or anti-cytotoxic T-lymphocyte-associated antigen 4 antibodies), in addition to newer generation multikinase inhibitors or an anti-vascular endothelial growth factor receptor 2 antibody, as monotherapy treatment options [1, 5].
Regorafenib is an orally administered multikinase inhibitor that was shown to significantly prolong overall survival (OS) versus placebo (10.6 vs. 7.8 months; hazard ratio 0.63, 95% confidence interval [CI] 0.50, 0.79; p < 0.0001) in patients with advanced HCC who tolerated (≥400 mg/day for ≥20 of last 28 days of treatment) and progressed on sorafenib in the randomized, double-blind, placebo-controlled, phase 3 RESORCE trial [6]. Subsequently, regorafenib was the first agent approved in the second-line setting for patients with unresectable HCC (uHCC) [1, 7, 8]. However, like all phase 3 studies, strict inclusion and exclusion criteria may exclude important patient populations seen in everyday practice.
The prospective, observational REFINE study was designed to provide insight into the safety and effectiveness of regorafenib in the real-world setting and to support the results from the phase 3 RESORCE trial [6]. As a result, the study included a more clinically diverse patient population compared with RESORCE [6], including patients with Eastern Cooperative Oncology Group performance status (ECOG PS) ≥2, Child-Pugh B liver status, and intolerance to prior sorafenib (defined per protocol as patients who discontinued sorafenib due to adverse events while on treatment), as well as patients who had received prior immunotherapy, representing real-world patients with uHCC.
Methods
Study Design and Participants
This international, prospective, open-label, multicenter, observational study was conducted in 161 centers in 20 countries in Europe, North America, Asia, Latin America, Levant, Middle East, and North Africa, between September 2017 and January 2022. The full list of principal investigators and participating study sites is shown in online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000542285).
Eligible patients were adults with a confirmed diagnosis of uHCC, and for whom a physician-initiated decision to treat with regorafenib was made before study enrollment. Patients were required to meet the criteria for regorafenib use according to the local health authority-approved product information, including indications and contraindications with respect to the local market authorization and/or summary of product characteristics. The study included patients with ECOG PS 2–4, Child-Pugh B/C liver function, and significant comorbidities and/or medical history, such as moderate or severe ascites, history of or active autoimmune disease or immune deficiency, history of hepatic encephalopathy or encephalopathy, history of transplantation, and varices with an increased risk of bleeding, which were all parameters for exclusion in the phase 3 RESORCE trial [6]. Patients were excluded from this study if they had participated in an investigational program with interventions outside of routine clinical practice, or with unapproved interventions in a clinical trial, and if they had received prior treatment with regorafenib. This trial is registered with ClinicalTrials.gov number NCT03289273.
Assessments
The primary aim was to evaluate the safety of regorafenib in patients with uHCC, including the incidence of treatment-emergent adverse events (TEAEs) and dose modifications due to TEAEs in real-world practice. Dose modifications included dose interruptions and dose reductions.
Secondary aims were to assess the effectiveness of regorafenib based on the following parameters: OS, defined as the time from the start of regorafenib treatment to the date of death, due to any cause; progression-free survival (PFS), defined as the time from the start of regorafenib treatment to the date of first observed disease progression (radiological or clinical, whichever occurred first) or death due to any cause; time to progression (TTP), defined as the time from the start of regorafenib treatment to the first documented disease progression (radiological or clinical, whichever occurred first), assessed by the investigator, according to local standard; and best objective response rate, defined according to investigator-assessed data, with tumor assessments made according to local standard-of-care clinical practice (methods of assessment and/or the frequency of follow-up may therefore have differed across institutions); and regorafenib treatment duration, defined as the time interval from the start of regorafenib treatment to the day of permanent discontinuation of regorafenib (including death). Other secondary aims were to describe the patterns of regorafenib treatment, including actual doses, and other dosing parameters. “Sorafenib intolerance” was defined as any event, leading to sorafenib discontinuation. The time interval between patient assessments was based on the judgment of the treating-physician.
Data Analyses and Statistical Considerations
Safety was evaluated in all patients who received at least one dose of regorafenib, regardless of prior treatment. Patients who received at least one dose of regorafenib and did not violate any eligibility criteria and had at least one follow-up assessment after receiving regorafenib were included in all safety and effectiveness analyses and denoted as the full analysis set. Adverse events were summarized according to the Medical Dictionary for Regulatory Activities v25.0, and severity was categorized by the National Cancer Institute Common Terminology Criteria for Adverse Events v4.03.
OS, PFS, and TTP were evaluated using the Kaplan-Meier method. Patients were observed from the start of regorafenib treatment until the end of observation due to premature discontinuation, withdrawal of consent, death, or until the end of the study. The final analyses were conducted after the final patient who remained on study treatment had been observed for 24 months or was no longer under observation due to withdrawal or death.
The study did not aim to confirm or reject predefined hypotheses. All variables were analyzed descriptively with appropriate statistical methods: categorical variables by frequency tables and continuous variables by sample statistics.
Results
Patient Disposition and Characteristics
In total, 1,028 patients were enrolled, of whom 1,005 initiated regorafenib treatment and were eligible for analyses (online suppl. Fig. S1). The median age was 66 years (range 21–94), most patients had Barcelona Clinic Liver Cancer (BCLC) stage C disease (62%; n = 625), Child-Pugh A disease (61%; n = 618), and an ECOG PS of 0 or 1 (82%; n = 829) at study entry, and the most common causes of HCC were hepatitis B (38%; n = 382), alcohol use (25%; n = 250), and hepatitis C (24%; n = 242) (Table 1). Of the 618 patients with Child-Pugh A disease, 51% (n = 313) were classed as albumin-bilirubin (ALBI) grade 2 compared with 67% (n = 82/123) of patients with Child-Pugh B disease. A numerically higher proportion of Asian versus non-Asian patients were classed as Child-Pugh A at study entry (67% [n = 375/557] vs. 54% [n = 243/448]).
Patient characteristics at study entry
Characteristic . | All patients (N = 1,005) . |
---|---|
Region, n (%) | |
Asia | 557 (55) |
Non-Asia | 448 (45) |
Male sex, n (%) | 835 (83) |
Median age, years (range) | 66 (21–94) |
Race, n (%) | |
Asian | 567 (56) |
White | 254 (25) |
Black or African American | 20 (2) |
Unknown/not reported | 164 (16) |
ECOG PS, n (%) | |
0/1 | 829 (82) |
≥2 | 60 (6) |
Missing | 116 (12) |
Child-Pugh classification, n (%) | |
A | 618 (61) |
B | 123 (12) |
C | 5 (<1) |
Not evaluable | 26 (3) |
Missing | 233 (23) |
ALBI grade, n (%) | |
1 | 318 (32) |
2 | 480 (48) |
3 | 37 (4) |
Missing | 170 (17) |
BCLC stage, n (%) | |
0 | 2 (<1) |
A | 13 (1) |
B | 133 (13) |
C | 625 (62) |
D | 19 (2) |
Missing | 213 (21) |
Extrahepatic spread, n (%) | |
Yes | 591 (59) |
No | 413 (41) |
Missing | 1 (<1) |
Macrovascular invasion (multiple responses), n (%) | |
Yes | 346 (34) |
Portal vein thrombosis | 279 (28) |
Hepatic vein invasion | 79 (8) |
Lack of portal blood flow | 9 (1) |
Missing | 3 (<1) |
No | 655 (65) |
Missing | 4 (<1) |
Etiology of uHCC (multiple responses), n (%) | |
Hepatitis B | 382 (38) |
Alcohol use | 250 (25) |
Hepatitis C | 242 (24) |
NASH | 66 (7) |
Genetic/metabolic | 29 (3) |
Other | 29 (3) |
Unknown | 109 (11) |
Missing | 26 (3) |
Characteristic . | All patients (N = 1,005) . |
---|---|
Region, n (%) | |
Asia | 557 (55) |
Non-Asia | 448 (45) |
Male sex, n (%) | 835 (83) |
Median age, years (range) | 66 (21–94) |
Race, n (%) | |
Asian | 567 (56) |
White | 254 (25) |
Black or African American | 20 (2) |
Unknown/not reported | 164 (16) |
ECOG PS, n (%) | |
0/1 | 829 (82) |
≥2 | 60 (6) |
Missing | 116 (12) |
Child-Pugh classification, n (%) | |
A | 618 (61) |
B | 123 (12) |
C | 5 (<1) |
Not evaluable | 26 (3) |
Missing | 233 (23) |
ALBI grade, n (%) | |
1 | 318 (32) |
2 | 480 (48) |
3 | 37 (4) |
Missing | 170 (17) |
BCLC stage, n (%) | |
0 | 2 (<1) |
A | 13 (1) |
B | 133 (13) |
C | 625 (62) |
D | 19 (2) |
Missing | 213 (21) |
Extrahepatic spread, n (%) | |
Yes | 591 (59) |
No | 413 (41) |
Missing | 1 (<1) |
Macrovascular invasion (multiple responses), n (%) | |
Yes | 346 (34) |
Portal vein thrombosis | 279 (28) |
Hepatic vein invasion | 79 (8) |
Lack of portal blood flow | 9 (1) |
Missing | 3 (<1) |
No | 655 (65) |
Missing | 4 (<1) |
Etiology of uHCC (multiple responses), n (%) | |
Hepatitis B | 382 (38) |
Alcohol use | 250 (25) |
Hepatitis C | 242 (24) |
NASH | 66 (7) |
Genetic/metabolic | 29 (3) |
Other | 29 (3) |
Unknown | 109 (11) |
Missing | 26 (3) |
Child-Pugh classification stratified by ALBI grade, n (%) . | ALBI grade 1 . | ALBI grade 2 . | ALBI grade 3 . | Missing . |
---|---|---|---|---|
A (n = 618) | 248 (40) | 313 (51) | 2 (<1) | 55 (9) |
B (n = 123) | 8 (7) | 82 (67) | 24 (20) | 9 (7) |
Missing (n = 233) | 56 (24) | 74 (32) | 6 (3) | 97 (42) |
Child-Pugh classification stratified by ALBI grade, n (%) . | ALBI grade 1 . | ALBI grade 2 . | ALBI grade 3 . | Missing . |
---|---|---|---|---|
A (n = 618) | 248 (40) | 313 (51) | 2 (<1) | 55 (9) |
B (n = 123) | 8 (7) | 82 (67) | 24 (20) | 9 (7) |
Missing (n = 233) | 56 (24) | 74 (32) | 6 (3) | 97 (42) |
ALBI, albumin-bilirubin; BCLC, Barcelona Clinic Liver Cancer; ECOG PS, Eastern Cooperative Oncology Group performance status; NASH, nonalcoholic steatohepatitis; uHCC, unresectable hepatocellular carcinoma.
Most patients (96%; n = 965) received prior sorafenib treatment, whereas 10% of patients (n = 97) had received a prior immune checkpoint inhibitor and 6% (n = 65) had received a multikinase inhibitor other than sorafenib (online suppl. Table S2). Most patients (84%; n = 848) received regorafenib after one prior systemic treatment, whereas 14% (n = 145) received regorafenib at the third line or later. Following regorafenib, 32% (n = 324) of patients received systemic treatment; the most common systemic treatments after regorafenib were lenvatinib (13%; n = 126), cabozantinib (11%; n = 110), and nivolumab (5%; n = 52).
Of the 965 patients who had received prior sorafenib treatment, 91 (9%) had an adverse event leading to sorafenib discontinuation (defined as sorafenib-intolerant patients). The remaining 874 patients (91%) were considered “sorafenib-tolerant.” The baseline and clinical characteristics of sorafenib-intolerant patients are shown in online supplementary Table S3 and were numerically similar with the overall cohort.
Dosing
Of the total patients on study, 47% (n = 469), 11% (n = 106), and 40% (n = 398) initiated regorafenib treatment at 160, 120, and 80 mg/day, respectively (online suppl. Table S4). A numerically higher proportion of patients starting regorafenib at 160 mg/day had Child-Pugh A liver function and ALBI grade 1 compared with patients initiating regorafenib below 160 mg/day. Both BCLC stage at study entry and underlying causes of HCC were similar across initial dose level 80–160 mg/day groups (online suppl. Table S4). Initial dose in patients who received regorafenib as third- or later-line treatment was similar to that in patients who received regorafenib as second-line treatment; 41% (60/145) initiated regorafenib at 160 mg/day (compared with 48% [405/848] of second-line patients). A numerically higher proportion of patients from Asia initiated regorafenib at 160 mg/day compared with patients from non-Asian countries (51% [n = 282/557] vs. 42% [n = 187/448], respectively).
Dose modifications of regorafenib treatment by initial dose are shown in online supplementary Figure S2 and Table 5. Over the course of the study, dose modifications were used in 54% (n = 253/469), 57% (n = 60/106), and 60% (n = 239/398) of patients who initiated regorafenib at 160, 120, and 80 mg/day, respectively. Of those who initiated regorafenib at 160 mg/day, 49% (n = 229/469) had a dose reduction by ≥1 dose level, compared with 42% (n = 44/106) and 31% (n = 123/398) of patients who initiated regorafenib at 120 and 80 mg/day, respectively. Over the course of the study, a numerically higher proportion of patients who initiated regorafenib at <160 mg/day had a dose escalation by ≥1 dose level (80 mg/day: 40% [n = 160/398]; 120 mg/day: 22% [n = 23/106] vs. 160 mg/day: 7% [n = 32/469]) (online suppl. Fig. S2; online suppl. Table S5). Of those who initiated regorafenib at 80 mg/day, 25% (n = 101/398) had a dose increase as the first modification within the first 4 weeks. Dose escalations included both escalations as a first-dose modification and escalations that followed a dose reduction.
The last daily regorafenib doses were 160 mg (in 30% of patients; n = 299), 120 mg (19%; n = 192), and 80 mg (41%; n = 417) (Table 2). No patient received more than 160 mg/day. The last daily dose of regorafenib was 160 mg in 54% (n = 251/469) of patients who initiated regorafenib at 160 mg/day. The final daily dose of regorafenib was the same as the daily starting dose in 52–65% of patients across all initial dose groups.
Last daily dose of regorafenib by initial dose levela
Last daily dose, mg, n (%) . | Initial regorafenib dose (N = 1,005)b . | |||
---|---|---|---|---|
160 mg (n = 469; 47%) . | 120 mg (n = 106; 11%) . | 80 mg (n = 398; 40%) . | Total (N = 1,005) . | |
160 | 251 (54) | 10 (9) | 35 (9) | 299 (30) |
120 | 80 (17) | 55 (52) | 55 (14) | 192 (19) |
80 | 118 (25) | 35 (33) | 258 (65) | 417 (41) |
40 | 19 (4) | 6 (6) | 43 (11) | 87 (9) |
Other | 1 (<1) | 0 | 7 (2) | 10 (1) |
Last daily dose, mg, n (%) . | Initial regorafenib dose (N = 1,005)b . | |||
---|---|---|---|---|
160 mg (n = 469; 47%) . | 120 mg (n = 106; 11%) . | 80 mg (n = 398; 40%) . | Total (N = 1,005) . | |
160 | 251 (54) | 10 (9) | 35 (9) | 299 (30) |
120 | 80 (17) | 55 (52) | 55 (14) | 192 (19) |
80 | 118 (25) | 35 (33) | 258 (65) | 417 (41) |
40 | 19 (4) | 6 (6) | 43 (11) | 87 (9) |
Other | 1 (<1) | 0 | 7 (2) | 10 (1) |
One patient initiated regorafenib at 140 mg/day.
This patient’s last daily dose was 120 mg.
aDose modifications included dose interruptions and dose reductions.
b31 patients (3%) received a starting dose of 40 mg/day: the last daily doses for these patients were 160 mg/day (10%), 120 mg/day (3%), 80 mg/day (19%), 40 mg/day (61%), and others (6%).
Safety
At the end of the observation period, 936 of 1,005 patients (93%) had discontinued treatment, most commonly due to disease progression (46%; n = 428/936), adverse events (34%; n = 321/936), and physician’s decision (8%; n = 72/936). Most patients (92%; n = 921/1,005) experienced at least one TEAE and 54% (n = 540) had a grade ≥3 TEAE (Table 3), whereas 74% (n = 746) experienced at least one drug-related TEAE and 27% (n = 268) experienced a grade ≥3 drug-related TEAE. Dose modifications and permanent discontinuation of regorafenib due to TEAEs occurred in 45% (n = 450) and 31% (n = 311) of patients, respectively, whereas dose modifications and permanent discontinuation of regorafenib due to drug-related TEAEs occurred in 37% (n = 372) and 16% (n = 161) of patients, respectively. The incidence and severity of TEAEs and drug-related TEAEs in sorafenib-intolerant patients were similar with the overall patient population (Table 3). In this patient subgroup, dose modifications and permanent discontinuation of regorafenib due to TEAEs occurred in 36% (n = 33/91) and 56% (n = 51/91) of patients, respectively. In sorafenib-tolerant patients, the incidence and severity of TEAEs and drug-related TEAEs were generally similar to sorafenib-intolerant patients; however, TEAEs leading to permanent regorafenib discontinuation were more common in sorafenib-intolerant patients. Of patients who received the standard initial dose of regorafenib (160 mg), any grade and grade ≥3 drug-related TEAEs occurred in 70% (n = 330) and 24% (n = 111), respectively (online suppl. Table S6).
Overview of TEAEs
TEAE, n (%) . | Sorafenib-intolerant patients (n = 91) . | Sorafenib-tolerant patients (n = 874) . | All patients (N = 1,005) . |
---|---|---|---|
Any TEAE | 85 (93) | 797 (91) | 921 (92) |
Grade 1/2 | 28 (31) | 334 (38) | 372 (37) |
Grade ≥3a | 56 (62) | 460 (53) | 540 (54) |
Serious | 40 (44) | 319 (36) | 374 (37) |
Leading to dose modificationb | 33 (36) | 401 (46) | 450 (45) |
Leading to permanent discontinuation of study drug | 51 (56) | 244 (28) | 311 (31) |
Any drug-related TEAE | 67 (74) | 641 (73) | 746 (74) |
Grade 1/2 | 39 (43) | 410 (47) | 466 (46) |
Grade ≥3c | 27 (30) | 227 (26) | 268 (27) |
Serious | 9 (10) | 73 (8) | 90 (9) |
Leading to dose modificationb | 27 (30) | 332 (38) | 372 (37) |
Leading to permanent discontinuation of study drug | 27 (30) | 123 (14) | 161 (16) |
TEAE, n (%) . | Sorafenib-intolerant patients (n = 91) . | Sorafenib-tolerant patients (n = 874) . | All patients (N = 1,005) . |
---|---|---|---|
Any TEAE | 85 (93) | 797 (91) | 921 (92) |
Grade 1/2 | 28 (31) | 334 (38) | 372 (37) |
Grade ≥3a | 56 (62) | 460 (53) | 540 (54) |
Serious | 40 (44) | 319 (36) | 374 (37) |
Leading to dose modificationb | 33 (36) | 401 (46) | 450 (45) |
Leading to permanent discontinuation of study drug | 51 (56) | 244 (28) | 311 (31) |
Any drug-related TEAE | 67 (74) | 641 (73) | 746 (74) |
Grade 1/2 | 39 (43) | 410 (47) | 466 (46) |
Grade ≥3c | 27 (30) | 227 (26) | 268 (27) |
Serious | 9 (10) | 73 (8) | 90 (9) |
Leading to dose modificationb | 27 (30) | 332 (38) | 372 (37) |
Leading to permanent discontinuation of study drug | 27 (30) | 123 (14) | 161 (16) |
Adverse event coding was based on MedDRA v25, and event grading was based on NCI CTCAE v4.03.
MedDRA, Medical Dictionary for Regulatory Activities; NCI CTCAE, National Cancer Institute Common Terminology Criteria for Adverse Events; TEAE, treatment-emergent adverse event.
aGrade 5 TEAEs occurred in 14% (n = 139) of all patients.
bDose modifications include dose interruptions and reductions.
cGrade 5 drug-related TEAEs occurred in 1% (n = 7) of all patients.
In the overall population, the most common TEAEs were hand-foot skin reaction (HFSR; 33% [n = 329]; grade ≥3 [6%; n = 58]), diarrhea (29% [n = 296]; grade ≥3 [4%; n = 40]), and fatigue (20% [n = 200]; grade ≥3 [3%; n = 26]) (Table 4). These were considered drug-related in 31% (n = 309), 26% (n = 258), and 15% (n = 151) of cases, respectively (Table 5). In the 91 sorafenib-intolerant patients, the most common TEAEs were diarrhea (27% [n = 25]; grade ≥3 [2%; n = 2]), HFSR (25% [n = 23]; grade ≥3 [11%; n = 10]), and asthenia (19% [n = 17]; grade ≥3 [1%; n = 1]) (Table 4), which were considered drug-related in 25% (n = 23), 25% (n = 23), and 16% (n = 15) of cases, respectively (Table 5). In the overall population of REFINE, drug-related hepatobiliary disorders occurred in 37 patients (3.7%) and hepatic failure occurred in 7 patients (0.7%). A comparison of the safety profile of regorafenib shows generally similar or lower incidences of TEAEs in the sorafenib-intolerant group vs the RESORCE population [6], as per the identified risks displayed in Figure 1.
Most frequent TEAEs (any grade; ≥10% of all patients) and incidence of worst grade 3 or 4
TEAE, n (%) . | Sorafenib-intolerant patients (n = 91) . | All patients (N = 1,005) . | ||
---|---|---|---|---|
any grade . | grade ≥3 . | any grade . | grade ≥3 . | |
HFSR | 23 (25) | 10 (11) | 329 (33) | 58 (6) |
Diarrhea | 25 (27) | 2 (2) | 296 (29) | 40 (4) |
Fatigue | 7 (8) | 1 (1) | 200 (20) | 26 (3) |
Decreased appetite | 16 (18) | 2 (2) | 177 (18) | 15 (1) |
Hypertension | 11 (12) | 5 (5) | 115 (11) | 34 (3) |
Abdominal pain | 13 (14) | 3 (3) | 114 (11) | 28 (3) |
Asthenia | 17 (19) | 1 (1) | 108 (11) | 16 (2) |
TEAE, n (%) . | Sorafenib-intolerant patients (n = 91) . | All patients (N = 1,005) . | ||
---|---|---|---|---|
any grade . | grade ≥3 . | any grade . | grade ≥3 . | |
HFSR | 23 (25) | 10 (11) | 329 (33) | 58 (6) |
Diarrhea | 25 (27) | 2 (2) | 296 (29) | 40 (4) |
Fatigue | 7 (8) | 1 (1) | 200 (20) | 26 (3) |
Decreased appetite | 16 (18) | 2 (2) | 177 (18) | 15 (1) |
Hypertension | 11 (12) | 5 (5) | 115 (11) | 34 (3) |
Abdominal pain | 13 (14) | 3 (3) | 114 (11) | 28 (3) |
Asthenia | 17 (19) | 1 (1) | 108 (11) | 16 (2) |
Adverse event coding was based on MedDRA v25, and event grading was based on NCI CTCAE v4.03.
HFSR, hand-foot skin reaction; MedDRA, Medical Dictionary for Regulatory Activities; NCI CTCAE, National Cancer Institute Common Terminology Criteria for Adverse Events; TEAE, treatment-emergent adverse event.
Most frequent drug-related TEAEs (any grade; ≥10% of patients) and incidence of worst grade 3 or 4
TEAE, n (%) . | Sorafenib-intolerant patients (n = 91) . | All patients (N = 1,005) . | ||
---|---|---|---|---|
any grade . | grade ≥3 . | any grade . | grade ≥3 . | |
HFSR | 23 (25) | 10 (11) | 309 (31) | 57 (6) |
Diarrhea | 23 (25) | 1 (1) | 258 (26) | 33 (3) |
Fatigue | 4 (4) | 1 (1) | 151 (15) | 21 (2) |
Decreased appetite | 13 (14) | 1 (1) | 133 (13) | 10 (1) |
Hypertension | 10 (11) | 5 (5) | 96 (10) | 30 (3) |
TEAE, n (%) . | Sorafenib-intolerant patients (n = 91) . | All patients (N = 1,005) . | ||
---|---|---|---|---|
any grade . | grade ≥3 . | any grade . | grade ≥3 . | |
HFSR | 23 (25) | 10 (11) | 309 (31) | 57 (6) |
Diarrhea | 23 (25) | 1 (1) | 258 (26) | 33 (3) |
Fatigue | 4 (4) | 1 (1) | 151 (15) | 21 (2) |
Decreased appetite | 13 (14) | 1 (1) | 133 (13) | 10 (1) |
Hypertension | 10 (11) | 5 (5) | 96 (10) | 30 (3) |
Incidence of low-grade drug-related asthenia was higher in sorafenib-intolerant patients than in the overall cohort (16% [n = 15]; grade ≥3: 0% [n = 0] vs. 9% [n = 91]; grade ≥3: 1% [n = 13], respectively).
Adverse event coding was based on MedDRA v25, and event grading was based on NCI CTCAE v4.03.
HFSR, hand-foot skin reaction; MedDRA, Medical Dictionary for Regulatory Activities; NCI CTCAE, National Cancer Institute Common Terminology Criteria for Adverse Events; TEAE, treatment-emergent adverse event.
The incidence and severity of TEAEs in sorafenib-intolerant patients in REFINE versus the phase 3 RESORCE population. Incidence (bubble size): Percentage of patients with TEAE in the respective group. Severity (color): Percentage of patients with serious TEAE in the respective group. HFSR, hand-foot skin reaction; TEAE, treatment-emergent adverse event.
The incidence and severity of TEAEs in sorafenib-intolerant patients in REFINE versus the phase 3 RESORCE population. Incidence (bubble size): Percentage of patients with TEAE in the respective group. Severity (color): Percentage of patients with serious TEAE in the respective group. HFSR, hand-foot skin reaction; TEAE, treatment-emergent adverse event.
Effectiveness
The median OS in the overall population was 13.2 months (95% CI: 11.6, 14.8; n = 1,005; Fig. 2). The median OS in patients who received regorafenib as second-line therapy after prior first-line sorafenib was 13.9 months (95% CI: 12.2, 15.3; n = 827; Fig. 3a), compared with 9.7 months (95% CI: 7.9, 13.9; n = 145) in those who received regorafenib as third- or later-line therapy. The median OS was longer in the 557 patients from Asia (15.0 months, 95% CI: 12.6, 16.2) compared with the 448 patients from non-Asian countries (11.4 months, 95% CI: 9.9, 13.8) (Fig. 3b). However, the median OS was similar in Child-Pugh A patients from Asia (16.0 months, 95% CI: 14.6, 17.6; n = 375) and in those from non-Asian countries (14.5 months, 95% CI: 11.4, 16.2; n = 243) (online suppl. Fig. S3). The median OS was longer in patients with Child-Pugh A disease (15.5 months, 95% CI: 13.9, 16.3; n = 618) compared with patients with Child-Pugh B disease (6.3 months, 95% CI: 4.9, 7.8; n = 123) (Fig. 3c). In addition, median OS by initial regorafenib dose in patients with Child-Pugh A was 16.3 months (95% CI: 15.0, 18.8) at 160 mg/day, 13.9 months (95% CI: 9.7, 19.8) at 120 mg/day, 13.2 months (95% CI: 10.7, 16.2) at 80 mg/day, and 16.7 months (95% CI: 6.1, 25.6) at 40 mg/day. Furthermore, the median OS was longer in patients who were classed as ALBI grade 1 (20.2 months, 95% CI: 17.4, 23.2; n = 318) compared with those who were classed as ALBI grade 2 (10.2 months, 95% CI: 8.7, 11.4; n = 480) at baseline (Fig. 3d). For patients who received a prior immune checkpoint inhibitor (n = 97), the median OS was 9.9 months (95% CI: 7.5, 15.0). However, most patients (81%; n = 79) who received a prior immune checkpoint inhibitor also received regorafenib in the third-line setting and beyond. In patients who were unable to tolerate sorafenib, the median OS was shorter (11.1 months [95% CI: 8.1, 18.1]) than in patients who were able to tolerate sorafenib (13.2 months [95% CI: 12.0, 15.1]) (Fig. 3f). The median OS from the start of prior sorafenib in the 822 patients who received regorafenib as second-line therapy was 25.3 months (95% CI: 24.0, 27.5). The median OS in patients who initiated regorafenib at 160 mg/day (n = 469) was 15.8 months (95% CI: 13.6, 17.3) compared with 13.3 months (95% CI: 9.7, 17.9) and 11.1 months (95% CI: 9.9, 13.2) in patients who initiated regorafenib at 120 mg/day (n = 106) and 80 mg/day (n = 398), respectively (online suppl. Fig. S4). The median OS in patients who received prior immunotherapy was 13.5 months (95% CI: 12.0, 15.1).
Overall survival stratified by regorafenib treatment line (a), geographic region (Asia vs. non-Asia; b), geographic region (Asia vs. non-Asia) in patients with Child-Pugh A (c), baseline Child-Pugh classification (d), baseline ALBI grade (e), and sorafenib intolerance (f). ALBI, albumin-bilirubin; CI, confidence interval; OS, overall survival. aData are not shown for regorafenib in the first-line setting due to small sample size. Logrank p value was computed across all treatment lines. bData are not shown for Child-Pugh C disease and “not evaluable” due to small sample size. Logrank p value was computed across all Child-Pugh classification.
Overall survival stratified by regorafenib treatment line (a), geographic region (Asia vs. non-Asia; b), geographic region (Asia vs. non-Asia) in patients with Child-Pugh A (c), baseline Child-Pugh classification (d), baseline ALBI grade (e), and sorafenib intolerance (f). ALBI, albumin-bilirubin; CI, confidence interval; OS, overall survival. aData are not shown for regorafenib in the first-line setting due to small sample size. Logrank p value was computed across all treatment lines. bData are not shown for Child-Pugh C disease and “not evaluable” due to small sample size. Logrank p value was computed across all Child-Pugh classification.
The median PFS in the overall population was 3.9 months (95% CI: 3.6, 4.1; n = 997; online suppl. Fig. S5a). Median PFS was similar in the 554 evaluable patients from Asia (3.7 months, 95% CI: 3.3, 4.0) compared with the 443 evaluable patients from non-Asian countries (4.0 months, 95% CI: 3.6, 4.7; online suppl. Fig. S5b). The median PFS in patients who had received prior immunotherapy was 3.6 months (95% CI: 3.0, 4.3; n = 97). In the overall population, the median TTP was 4.1 months (95% CI: 3.9, 4.6; n = 997). In the 997 patients who were evaluable for antitumor activity, the objective response rate was 14% (2% achieved a complete response and 12% achieved a partial response; online suppl. Table S7).
Duration of Treatment
The median treatment duration for all patients was 3.7 months (range 1 day to 38.9 months) and the median observation duration was 9.4 months (range 1 day to 47.6 months). Patients who prematurely discontinued regorafenib, withdrew consent from the study, or died following one dose of regorafenib had a treatment duration of 1 day. In approximately 25% of the total population, the duration of regorafenib treatment was less than 2 months. The median treatment duration in patients who initiated regorafenib at 160, 120, and 80 mg/day was 3.7 months (range 1 day to 32.7 months), 3.7 months (range 1 day to 32.5 months), and 3.4 months (range 1 day to 38.9 months), respectively. The mean duration of treatment by the initial dose level at 160 mg/day, 120 mg/day, and 80 mg/day was 6.3 months and at 40 mg/day was 5.2 months.
Discussion
The profile of patients from interventional phase 1–3 trials may not reflect their diversity in real-world clinical practice due to the strict eligibility criteria of such trials [9‒11]. For example, in the phase 3 RESORCE trial, only patients with uHCC who tolerated and progressed on prior sorafenib and those with Child-Pugh A liver function were eligible for enrollment [6]. Observational studies are therefore useful since they provide information on subgroups underrepresented in interventional trials, and help the evaluation of the effectiveness of second- and later-line systemic treatments for uHCC in a diverse population of patients in real-world clinical settings to inform treatment decisions [9‒11]. The patient population from REFINE was more diverse than that of the phase 3 RESORCE trial [6]. REFINE included patients who are more often seen in modern clinical practice, including those with ECOG PS ≥2, Child-Pugh B liver function, sorafenib-intolerant patients, and those who had received prior immunotherapy.
Despite the more varied patient population in REFINE, the safety profile of regorafenib was consistent with what is reported in RESORCE [6], with no unexpected safety signals. Treatment discontinuation due to TEAEs occurred in 31% of patients despite 92% experiencing at least one TEAE, which is similar to RESORCE, where 25% of patients discontinued due to adverse events. However, dose modifications due to TEAEs in REFINE were less frequent than in RESORCE (45% vs. 68%). It is important to note that the starting dose of regorafenib was decided by the treating physician, which may have impacted on the frequency of dose modifications compared with RESORCE. Of the patients who received standard initial dosing of regorafenib, 24% experienced a drug-related grade ≥3 TEAE, whereas in RESORCE, in which all patients received the standard dose, 50% experienced a grade ≥3 TEAE. However, the lack of regular, standardized adverse event reporting across study sites may have contributed to the lower incidence of grade ≥3 drug-related TEAEs in REFINE, compared with RESORCE. Across all patients, REFINE had a lower incidence of grade ≥3 TEAEs than RESORCE (54% vs. 66% [grade 3 or 4]). Incidences of grade ≥3 HFSR and fatigue were higher in RESORCE than in REFINE (13% vs. 6% [HFSR] and 9% vs. 3% [fatigue], respectively). Notably, cumulative evidence suggests that the most common regorafenib-related TEAEs, including HFSR, diarrhea, and hypertension (which are associated with TKI use in general), correlate with improved outcomes in advanced HCC [12].
The median OS in REFINE (13.2 months, 95% CI: 11.6, 14.8) was numerically higher than that reported in RESORCE (10.6 months, 95% CI: 9.1, 12.1 [regorafenib arm]) [6], confirming the effectiveness of regorafenib as a treatment option for patients with uHCC. Possible explanations for this difference include the availability of systemic treatments other than sorafenib and regorafenib (i.e., either before or after receiving regorafenib) and increased physician experience with regorafenib at the time of the REFINE study – better management of adverse events invariably translates to improved outcomes with regorafenib, as previously demonstrated for sorafenib [13]. As expected, median OS was nominally longer in patients with Child-Pugh A liver disease and those who were classed as ALBI grade 1 at baseline, reflecting the confounding influence of underlying liver dysfunction on survival rates. Additionally, median OS was longer in patients from Asian countries (15.0 months [95% CI: 12.6, 16.2]) compared with patients from non-Asian countries (11.4 months [95% CI: 9.9, 13.8]). This trend may be explained by multiple factors, including that patients from Asian countries generally had more favorable prognostic factors than non-Asian patients, for example, Child-Pugh A liver function (67% vs 54%) and ALBI grade 1 (37% vs 25%). Furthermore, a numerically greater proportion of Asian patients received concomitant systemic (8% vs 3%) and nonsystemic anticancer therapies (13% vs 4%), as well as a prior locoregional therapy with TACE (71% vs 42%), the latter previously reported to improve survival in patients with advanced HCC when combined with regorafenib in a retrospective, real-world study in Asia [14]. Moreover, as anticipated, the most common viral etiology in patients from Asian countries was hepatitis B (60%), whereas in non-Asian patients, hepatitis C (27%) was the most common. Studies in patients with HCC and hepatitis B have found that viral suppression caused by antiviral usage in these patients may reduce liver-related mortality. However, this benefit may be limited in patients with decompensated HBV cirrhosis [15]. While median OS was longer in Asian patients vs non-Asian patients, it was similar across both groups in patients with Child-Pugh A at study entry (16.0 months, 95% CI: 14.6, 17.6 vs. 14.5 months, 95% CI: 11.4, 16.2, respectively), suggesting that patients with good liver function benefit most from regorafenib. The longer OS observed in patients with Child-Pugh A or ALBI grade 1 validate these important prognostic factors in patients with HCC treated with systemic therapy. The longer median OS in patients who initiated regorafenib at 160 mg/day (15.8 months, 95% CI: 13.6, 17.3) compared with <160 mg/day doses (120 mg/day: 13.3 months [95% CI: 9.7, 17.9]; 80 mg/day: 11.1 months [95% CI: 9.9, 13.2]) may also be related to a higher proportion of patients classed as Child-Pugh A and ALBI grade 1, and different treatment practices in Asian compared with non-Asian countries. With the availability of models such as PROSASH-II [16], which can predict survival across different patient risk groups based on baseline clinical characteristics, treatment of advanced HCC in daily clinical practice can be tailored to the individual patient. Although the validity of such tools was not tested in REFINE, the evaluation of prognostic models for outcome prediction in patients with uHCC treated by regorafenib is a subject for further research. Additionally, these models offer scope for defining preplanned subgroups for future studies.
Patients with poor liver function are often treated with regorafenib in the real-world clinical practice setting. The standard approved daily dose of regorafenib is 160 mg/day in a 3-week-on/1-week-off schedule. Results from this study show that physicians in real-world practice use a lower starting dose of regorafenib and may increase to the standard dose based on patients’ overall health status/liver function and tolerability; in REFINE, 47%, 11%, 40%, and 3% of patients had initial daily doses of 160, 120, 80, and 40 mg, respectively. Of the patients with an initial dose of 160 mg/day, 54% also had 160 mg/day as their last dose.
Median regorafenib treatment duration was similar across initial dose levels. Overall, 57% of patients enrolled in REFINE had dose modifications and 23% had dose escalations, which reaffirms that flexible dosing approaches are commonly used in clinical practice, and that these may differ across regions. The authors acknowledge that real-world data over a longer follow-up are needed to determine the long-term prognosis of patients who receive <160 mg/day starting doses of regorafenib.
An exploratory analysis of RESORCE evaluated outcomes of sequential treatment with sorafenib followed by regorafenib [17]. Regorafenib provided clinical benefit to patients despite the rate of disease progression on prior sorafenib and irrespective of the last sorafenib dose in this very selected clinical trial population requiring well-preserved liver function and performance status [17]. Despite the broader population included in REFINE, median OS from the start of prior sorafenib in patients who received regorafenib as second-line therapy was almost identical to that reported in the RESORCE exploratory analysis (25.3 months, 95% CI: 24.0, 27.5 vs. 26.0 months, 95% CI: 22.6, 28.1 [17], respectively). Notably, patients who received regorafenib as second-line therapy (either after first-line sorafenib or after first-line non-sorafenib) had a nominally longer OS (13.9 months, 95% CI: 12.2, 15.3 and 17.4 months, 95% CI: 6.8, 21.6, respectively) than patients who received regorafenib as third- or later-line therapy (9.7 months, 95% CI: 7.9, 13.9). Taken together, these findings reinforce the clinical significance of regorafenib as a second-line agent to extend survival in patients with uHCC.
In 2020 (during the REFINE study period), the immune checkpoint inhibitor, atezolizumab in combination with bevacizumab, was approved by both the US Food and Drug Administration (FDA) and European Medicines Agency (EMA) as a first-line treatment option for patients with uHCC based on the results of the phase 3 IMbrave150 trial, which revealed the combination to be more effective than sorafenib, based on OS and PFS outcomes [18‒20]. Following its approval, the combination regimen has rapidly replaced sorafenib as the standard initial treatment for patients with advanced HCC with Child-Pugh A liver disease who are without contraindications to immune checkpoint inhibitors or antiangiogenic therapy [21]. Lenvatinib and durvalumab in combination with tremelimumab are also both approved in the first-line setting, whereas nivolumab in combination with ipilimumab (FDA-approved only), pembrolizumab (FDA-approved only), cabozantinib, and ramucirumab (along with regorafenib) are approved as second-line treatments after failure on prior sorafenib [22‒27]. It should be noted that unlike the phase 3 trials of lenvatinib (REFLECT) [23] and tremelimumab plus durvalumab (HIMALAYA) [27] for first-line uHCC, REFINE included high-risk patients with advanced HCC (e.g., those with tumor thrombosis in the main trunk of the portal vein). Optimal sequencing of therapy for patients progressing on atezolizumab-bevacizumab is not yet established, although international treatment guidelines recommend switching to multikinase inhibitor therapy given the current evidence for its efficacy in first- and second-line settings; accordingly, preferred options include sorafenib, lenvatinib, cabozantinib, regorafenib, and ramucirumab (for patients with an alpha-fetoprotein level of ≥400 ng/mL) [28, 29].
In this study, 10% of patients had received prior immunotherapy. The median OS in these patients was 9.9 months (95% CI: 7.5, 15.0). It should be noted, however, that a substantial proportion of patients who had received a prior immune checkpoint inhibitor had received regorafenib as third- or later-line therapy. Notably, a recent retrospective, multicenter, real-world study reported a median OS of 15.8 months (95% CI: 10.5, not evaluable) in patients who received regorafenib following first-line atezolizumab plus bevacizumab. Additionally, the study showed that patients who received regorafenib as second-line therapy, compared with other TKIs, were more likely to receive subsequent third-line therapy (most commonly, cabozantinib) [30].
The median PFS in the overall population from REFINE was 3.9 months (95% CI: 3.6, 4.1), which was similar in patients from Asia (3.7 months [95% CI: 3.3, 4.0]) and in those who had received prior immunotherapy (3.6 months [95% CI: 3.0, 4.3]). This compares with a reported median PFS of 2.6 months (95% CI: 1.9, 4.0) in patients receiving regorafenib following first-line atezolizumab plus bevacizumab, where PFS results were similar versus other TKIs (hazard ratio 1.07) [30].
Based on the phase 3 RESORCE trial eligibility criteria that led to its approval, regorafenib was originally indicated for use in patients with HCC who had tolerated and progressed on prior sorafenib, and thus was not initially recommended in patients who were sorafenib-intolerant due to the unknown safety profile in this patient population [7, 8]. In this study, 9% of patients who received prior sorafenib had an adverse event leading to sorafenib discontinuation and were considered sorafenib-intolerant. A higher proportion of sorafenib-intolerant patients were Child-Pugh B at study entry compared with the total patient population (19% vs. 12%). Overall, 93% of sorafenib-intolerant patients experienced a TEAE of any grade. A similar rate of drug-related TEAEs was reported in sorafenib-intolerant patients versus the overall cohort (74%, both). The most common drug-related TEAEs (any grade, >10%) in sorafenib-intolerant patients were diarrhea (25%), HFSR (25%), asthenia (16%), decreased appetite (14%), and hypertension (11%). In sorafenib-intolerant patients, the incidence and severity of drug-related TEAEs, including diarrhea and HFSR, were similar to the overall cohort, and none were grade 4 or 5. As the occurrence of TEAEs and drug-related TEAEs was similar between sorafenib-intolerant patients and the total population in REFINE, the EMA removed the safety precaution in the prior sorafenib-intolerant population, thereby allowing the use of regorafenib without the need of prior tolerance to sorafenib [8].
Limitations of the global REFINE study are largely those inherent to observational research. The findings generated are subject to biases, including selection bias, limitations to availability of historical medical data, and differences in treatment or reporting due to local guidelines. These limitations will need to be considered along with the benefit of using data in the real-world treatment setting. The results of PFS or TTP must be interpreted carefully as time periods between follow-up visits are more variable than in controlled clinical studies, in which a fixed visit schedule is maintained. Furthermore, the quality of tumor status evaluation will differ from that informed by randomized controlled trials. Comparisons of outcomes after treatment with regorafenib cannot be made with a comparator based on this single-arm study, unless based on historical data, which is subject to bias and confounding. Historical data collected on prior sorafenib use or other treatments may be incomplete and differ according to the standard of care in local practice. Additionally, although the study aimed to include a diverse patient population over a broad geographic reach, local limitations may have undermined the representativeness of patients recruited, including patient access to participating physicians, regorafenib availability and/or reimbursement, and other factors informing the local standard of care. Lastly, this study was not aimed at determining treatment sequencing patterns in patients with advanced HCC. We acknowledge that optimal sequencing following first-line immunotherapy needs to be better defined. Despite the outlined limitations, we would expect our results to be broadly representative of the real-world safety and effectiveness of regorafenib.
Conclusion
In conclusion, this real-world study confirmed the safety and effectiveness of regorafenib as a second-line agent in a clinically diverse uHCC patient population. Moreover, in patients who received standard regorafenib dosing, the findings are broadly consistent with those reported in the phase 3 RESORCE trial, in which sorafenib was the only approved agent for the treatment of patients with advanced HCC. With the emergence of new systemic therapy options for the management of uHCC and the rapidly evolving treatment landscape, the results of this study may help guide future treatment strategies, including patient selection, treatment sequencing, and optimized regorafenib dosing.
Acknowledgments
We thank the patients who participated in the REFINE study and their families. Medical writing support was provided by Robyn Fowler, PhD, of Luna, OPEN Health Communications (London, UK).
Statement of Ethics
The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice Guidelines. The study protocol was reviewed and approved by Ethics Committees at each of the participating sites. The full list of participating sites and central ethics committees and health authority/IRB approvals for each country with participating site(s) is provided in online supplementary Table S1. Written informed consent was obtained from each patient included in the study. This study was a post-authorization safety study; EU PAS Register Number EUPAS20981 (first registered on September 19, 2017).
Conflict of Interest Statement
Yoon Jun Kim has received research grants from BTG, AstraZeneca, Gilead Sciences, Daewoong, Bayer, Samjin, and Yuhan Pharmaceuticals. Philippe Merle has received consulting fees from Bayer, Ipsen, Eli Lilly, Eisai, Roche, AstraZeneca, Bristol Myers Squibb (BMS), and Roche; travel expenses from Bayer, Ipsen, Roche, and BMS; and research funding (to institution) from Ipsen. Richard S. Finn has participated in advisory boards for AstraZeneca, Bayer AG, BMS, CStone Pharmaceuticals, Eisai Co., Ltd., Eli Lilly and Company, Exelixis, Inc., Jiangsu Hengrui Pharmaceuticals Co., Ltd., Merck & Co., Inc., Pfizer Inc., and Roche-Genentech; receives grant funding from Adaptimmune Therapeutics LC, Bayer AG, BMS, Eli Lilly and Company, Eisai Co., Ltd., Merck & Co., Inc., Pfizer Inc., and Roche-Genentech; and is a principal investigator for BMS, Eisai Co., Ltd., Merck & Co., Inc., Pfizer, and Roche-Genentech. Masatoshi Kudo has received speaker fees from AstraZeneca, Bayer AG, Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Eli Lilly and Company, and Takeda Pharmaceutical Company Limited; provides consultancy services for F. Hoffmann-La Roche AG; and currently receives research grants from AbbVie Inc., EA Pharma, Co., Ltd., Eisai Co., Ltd., Gilead Sciences, Inc., Otsuka Pharmaceutical Co., Ltd., Sumitomo Dainippon Pharma, Taiho Pharmaceutical Co., Ltd., Takeda Pharmaceutical Company Limited, and GE HealthCare. Heinz-Josef Klümpen has participated in an advisory board for Janssen, AstraZeneca, and Ipsen; and has received speaker fees to institution for CCO and MEDtalks. Masafumi Ikeda has received honoraria from AstraZeneca K.K., Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Eli Lilly Japan K.K., Incyte Biosciences Japan G.K., Novartis Pharma K.K., and Takeda Pharmaceutical Co., Ltd.; and has research grants from AstraZeneca K.K., Bayer Yakuhin, Ltd., BMS K.K., Chiome Bioscience Inc., Chugai Pharmaceutical Co., Ltd., Delta-Fly Pharma, Inc., Eisai Co., Ltd., Eli Lilly Japan K.K., J Pharma Co., Ltd., Merck BioPharma Co., Ltd., Merus N.V., Merck Sharp & Dohme (MSD) K.K., Nihon Servier Co., Ltd., Novartis Pharma K.K., Ono Pharmaceutical Co., Ltd., Syneos Health Clinical K.K., and Invitees Japan K.K. Gianluca Masi has received fees for consulting/advisory roles from AstraZeneca, Eisai, MSD Oncology, and Roche; receives research funding from Terumo (to institution); patents, royalties, and other IP from Terumo (to institution). Chih-Hung Hsu has research grants from Ministry of Science and Technology and grants from Ministry of Health and Welfare, ROC; and has received personal fees from BMS, Ono Pharmaceutical, MSD, Roche, Merck Serono, BeiGene, and AstraZeneca. Matthias Pinter has received honoraria from Bayer, BMS, Eisai, Lilly, MSD, and Roche; has provided consultancy services for AstraZeneca, Bayer, BMS, Eisai, Ipsen, Lilly, MSD, and Roche; and has received travel support from Bayer, BMS, Ipsen, and Roche. Masayuki Kurosaki has received honoraria from Eisai, AbbVie, Gilead, Chugai, Incyte, and Eli Lilly. Philip Twumasi-Ankrah, Javeed Khan, Maria Awan, and Kirhan Ozgurdal are employees of Bayer. Ho Yeong Lim, Alessandro Granito, René Gerolami, Sung Bum Cho, Yi-Hsiang Huang, Long-Bin Jeng, Do Young Kim, Shi-Ming Lin, Guoliang Shao, Naoya Kato, Kazushi Numata, Kung-Kai Kuo, Yilei Mao, Yih-Jyh Lin, Kangshun Zhu, and Shukui Qin have nothing to disclose.
Funding Sources
This study was funded by Bayer, who also funded medical writing and editorial support in accordance with Good Publication Practice (GPP) guidelines (www.ismpp.org/gpp-2022).
Author Contributions
Philippe Merle, Richard S. Finn, Heinz-Josef Klümpen, Ho Yeong Lim, Masafumi Ikeda, Matthias Pinter, Philip Twumasi-Ankrah, and Kirhan Ozgurdal contributed to study conception. Philippe Merle, MK, Masafumi Ikeda, Alessandro Granito, Gianluca Masi, Ho Yeong Lim, Richard S. Finn, and Kirhan Ozgurdal contributed to the design of the study. Kirhan Ozgurdal and Javeed Khan contributed to the supervision of the study. All authors (Yoon Jun Kim, Philippe Merle, Richard S. Finn, Masatoshi Kudo, Heinz-Josef Klümpen, Ho Yeong Lim, Masafumi Ikeda, Alessandro Granito, Gianluca Masi, René Gerolami, Sung Bum Cho, Chih-Hung Hsu, Yi-Hsiang Huang, Long-Bin Jeng, Do Young Kim, Shi-Ming Lin, Matthias Pinter, Guoliang Shao, Naoya Kato, Masayuki Kurosaki, Kazushi Numata, Kung-Kai Kuo, Yilei Mao, Yih-Jyh Lin, Kangshun Zhu, Philip Twumasi-Ankrah, Javeed Khan, Maria Awan, Kirhan Ozgurdal, and Shukui Qin) contributed to data collection, analysis, and interpretation. Kirhan Ozgurdal and Javeed Khan contributed to the literature search and to the writing of the manuscript. All authors were involved in the reviewing and editing of the manuscript, agree to be accountable for all aspects of the work, accept responsibility for the decision to submit the publication, and had access to all data reported in the study.
Data Availability Statement
Availability of the data underlying this publication will be determined according to Bayer’s commitment to the EFPIA/PhRMA “Principles for responsible clinical trial data sharing.” This pertains to scope, timepoint, and process of data access.
As such, Bayer commits to sharing upon request from qualified scientific and medical researchers patient-level clinical trial data, study-level clinical trial data, and protocols from clinical trials in patients for medicines and indications approved in the USA and European Union (EU) as necessary for conducting legitimate research. This applies to data on new medicines and indications that have been approved by the EU and US regulatory agencies on or after January 1, 2014.
Interested researchers can use www.vivli.org to request access to anonymized patient-level data and supporting documents from clinical studies to conduct further research that can help advance medical science or improve patient care. Information on the Bayer criteria for listing studies and other relevant information is provided in the member section of the portal.
Data access will be granted to anonymized patient-level data, protocols, and clinical study reports after approval by an independent scientific review panel. Bayer is not involved in the decisions made by the independent review panel. Bayer will take all necessary measures to ensure that patient privacy is safeguarded.