Abstract
Background/Aim: The reimbursement criteria of sorafenib for advanced hepatocellular carcinoma (HCC) were expanded in 2016 by Taiwan’s National Health Insurance (NHI) to include patients without macrovascular invasion or extrahepatic spread. This study explored sorafenib treatment outcomes before and after this expansion. Patients and Methods: The NHI database was searched for patients who initiated sorafenib treatment between January 1, 2013, and December 31, 2017. Clinical variables were retrieved from the Taiwan Cancer Registry database. Overall survival (OS) and time to treatment discontinuation (TTD) were calculated as the times from the first sorafenib prescription date until death and the final prescription date, respectively. Results: A total of 13,862 patients were included. The median age was 64 years, 78.1% of patients were male. Approximately a quarter of patients (25.1%) received sorafenib after the criteria expansion and exhibited significantly longer OS (median 7.9 vs. 6.6 months, p<0.001) and TTD (median 3.0 vs. 2.6 months, p=0.003) compared with patients who started before. These results were verified in patients with available data regarding clinical prognostic factors (n=9,378, 67.7% of the entire study population). In the multivariate analysis, sorafenib prescription after criteria expansion remained an independent predictor of longer OS [hazard ratio (HR)=0.87, p<0.001] and TTD (HR=0.93, p=0.004). In the subgroup analysis, these trends were consistently observed across different patient subgroups. Conclusion: Patients with HCC who received sorafenib treatment after the reimbursement criteria expansion exhibited longer OS and TTD.
- Hepatocellular carcinoma
- sorafenib
- Barcelona Clinic Liver Cancer Stage C
- transarterial chemoembolization
- National Health Insurance
- reimbursement criteria
Despite improvement in hepatitis treatment, hepatocellular carcinoma (HCC) remains one of the most common and deadly cancers worldwide (1). Early-stage HCC can be resected or ablated, and intermediate-stage HCC is mainly treated using transarterial chemoembolization (TACE). Advanced HCC, defined as HCC unsuitable for locoregional therapy, is mainly treated with systemic therapy and tends to exhibit poor prognosis (2-5). Sorafenib, a multikinase inhibitor, has been proven to enhance survival of patients with advanced HCC in two phase III studies and became a standard first-line therapy for advanced HCC (6, 7).
Patients with advanced HCC are highly heterogenous in disease extent, hepatitis etiology, and liver function reserve (8). Advanced HCC includes Barcelona Clinic Liver Cancer (BCLC) stage C diseases with extrahepatic spread (EHS) or macrovascular invasion (MVI) and BCLC stage B diseases, which are unsuitable or refractory to locoregional therapies. In the subgroup analyses of the aforementioned two phase III trials comparing sorafenib with a placebo, survival benefits were more prominent in patients without EHS or MVI (9, 10).
BCLC stage C disease is clearly defined, but designating BCLC stage B disease as unsuitable or refractory for locoregional therapy involves a clinical decision that can dramatically vary depending on geographical region. In the two phase III studies that compared sorafenib with a placebo, the Asia-Pacific study enrolled fewer patients with BCLC stage B disease than the SHARP study, which mainly enrolled patients in Europe and America (9, 10). This may reflect the differences in the criteria for feasibilities of locoregional therapy and the timings of conversion to systemic treatment between these regions.
Taiwan is an endemic area for viral hepatitis, and hence, HCC. Sorafenib prescriptions for BCLC stage C HCC have been reimbursed by the National Health Insurance (NHI) of Taiwan since August 1, 2012 (11). Since November 1, 2016, the reimbursement has been expanded to include patients with BCLC stage B disease refractory to locoregional therapies. Such a policy change resulted in two eras of sorafenib treatment for HCC in Taiwan. Thus, this study was designed to compare prescription patterns and treatment outcomes of sorafenib before and after the reimbursement criteria change.
Patients and Methods
Data source. The NHI program in Taiwan is a mandatory single-payer system covering 97% to 98% of the population according to previous literature (12). We retrieved NHI data from the Health and Welfare Data Science Center, Ministry of Health and Welfare, Taiwan. Death events were extracted from the National Death Registry database. Study participant data were collected and analyzed anonymously. This study was approved by the Research Ethics Committee of National Taiwan University Hospital.
Study population and sorafenib treatment. All inpatient or outpatient clinical visits were reviewed to identify patients who were prescribed sorafenib for the first time between January 1, 2013, and December 31, 2017. We cross-linked these patients by their identification numbers to the Taiwan Cancer Registry database for detailed information regarding initial HCC diagnosis. Patients with missing sex data or aged less than 20 years were excluded.
Under the initial sorafenib reimbursement criteria of NHI, patients with Child–Pugh A liver reserve and HCC that had EHS or MVI were covered (11). Physicians were required to apply for the use of sorafenib with imaging and laboratory evidence demonstrating that these requirements were fulfilled. The application required renewal every 2 months with dynamic imaging evidence showing no disease progression under sorafenib treatment. Since November 1, 2016, the criteria have been expanded to include patients whose liver tumors remained viable after three TACE procedures, even if the disease exhibited no EHS or MVI.
Statistical methods. Statistical analyses were performed using SAS (version 9.4; SAS Institute, Cary, NC, USA). For comparisons between the two sorafenib prescription periods (before and after the expansion of reimbursement criteria), the chi-square test and independent t test were used for categorical and continuous variables, respectively. Overall survival (OS) was calculated as the time from the first sorafenib prescription date until death. Time to treatment discontinuation (TTD) was measured from the first sorafenib prescription date to the last sorafenib prescription date or death. The last follow-up date for death events was December 31, 2018. The OS and TTD of patients were estimated using the Kaplan–Meier method and compared using the log-rank test between various groups of interest. The Cox proportional hazards model was utilized in the multivariate analysis to examine the effects of variables on OS and TTD. Two-sided p-values of ≤0.05 were considered statistically significant.
Results
Treatment patterns. A total of 13,862 patients were included in the study (Figure 1). The median age was 64 years, 78.1% were men, and almost 60% received sorafenib treatment in medical centers (Table I). Less than half (48.4%) of patients initially received the standard sorafenib dose; among them, nearly 60% required dose reduction. In patients who started sorafenib treatment with a reduced dose, 27.9% of them could eventually escalate to the standard dose. The median dose intensity of sorafenib was 533.3 mg/day. During the treatment course of sorafenib, 23.9% of patients received locoregional therapies, including TACE (88.6%), radiofrequency ablation (17.3%), and surgery (4.4%).
Patient flow diagram.
Baseline patient demographic and disease characteristics.
Approximately a quarter of patients commenced sorafenib treatment (n=3,484, 25.1%) after the criteria expansion. Compared with the patients who started before (n=10,378, 74.9%), those who started after the expansion were more likely to be younger than 65 years (p<0.001) and to receive TACE during the course of sorafenib treatment (p<0.001; Table I). The percentage of patients who received standard initial sorafenib dose (49.7% and 44.6%, respectively; p<0.001) and the sorafenib dose intensity (median, 534.5 and 517.6 mg/day, respectively; p<0.001) both decreased after the criteria expansion (Table I).
Treatment outcomes. As of December 31, 2018, 12,149 patients (87.6%) had died, and 13,419 patients (96.8%) had discontinued sorafenib treatment. The median follow-up time was 6.8 months. The median OS and median TTD of the entire population were 6.8 (95%CI=6.7-7.0) and 2.8 (95%CI=2.7-2.8) months, respectively. Patients who commenced sorafenib treatment after the reimbursement criteria expansion, compared with patients who started before, exhibited significantly longer OS (median 7.9 vs. 6.6 months, p<0.001; Figure 2A) and TTD (median 3.0 vs. 2.6 months, p=0.003; Figure 2B).
Kaplan–Meier plots showing (A) overall survival (OS) and (B) time to treatment discontinuation (TTD) of patients who received sorafenib treatment before and after the expansion of reimbursement criteria. p-Values were conducted using the log-rank test.
To confirm whether sorafenib prescription after the reimbursement criteria expansion was an independent predictor for OS and TTD, we adjusted other potential prognostic markers such as hepatitis virus infection and time from initial HCC diagnosis. Among the 13,862 patients, 9,378 patients (67.7%) had available data regarding these prognostic markers. In this selected population, the previously found associations between the reimbursement criteria expansion and age, TACE during sorafenib treatment, the percentage of patients who received standard initial sorafenib dose, and the sorafenib dose intensity remained (Table II). Additionally, patients who started sorafenib treatment after the criteria expansion, compared with patients who started before, exhibited longer times from HCC diagnosis to sorafenib prescription and were more likely to have BCLC stage B disease and initial serum alpha-fetoprotein (AFP) level <400 ng/ml (Table II).
Baseline demographic and disease characteristics of selected patients with available clinical prognostic factors.
In this selected population, patients who started sorafenib treatment after the criteria expansion, compared with patients who started before, still exhibited significant longer OS (median 7.5 vs. 6.1 months, p<0.001; Figure 3A) and TTD (median 3.0 vs. 2.5 months, p<0.001; Figure 3B). Subgroup analysis showed that these OS and TTD survival benefits consistently favored sorafenib treatment after criteria expansion (Figure 4A and B).
Kaplan–Meier plots showing (A) overall survival (OS) and (B) time to treatment discontinuation (TTD) of patients who received sorafenib treatment before and after the expansion of reimbursement criteria. Only patients with available prognostic factors data are included. p-Values were conducted using the log-rank test.
Forrest plot showing subgroup analysis of (A) overall survival (OS) and (B) time to treatment discontinuation (TTD) of patients who received sorafenib treatment before and after the expansion of reimbursement criteria.
In the multivariate analysis adjusting for age, sex, time from initial HCC diagnosis to sorafenib treatment, initial BCLC stage, initial serum AFP level, hospital level, viral hepatitis etiology, and initial standard sorafenib dose, receiving sorafenib treatment after the reimbursement criteria expansion remained an independent predictor for longer OS [hazard ratio (HR)=0.87, 95%CI=0.82-0.92, p<0.001] and TTD (HR=0.93, 95%CI=0.89-0.98, p=0.004; Table III).
Factors associated with overall survival and time to treatment discontinuation.
Discussion
This study compared the prescription pattern and treatment outcomes of patients who received sorafenib prescription for advanced HCC under two different sets of reimbursement criteria. Our data provided an unbiased view of the national Taiwanese population receiving sorafenib and revealed that the reimbursement criteria expansion beyond BCLC stage C was associated with longer TTD and OS. This association remained after adjustment of multiple prognostic factors. One of the possible explanations for our findings was that sorafenib provided higher efficacies in patients refractory to TACE but without MVI or EHS than patients with MVI or EHS. The subgroup analyses in the two phase III clinical trials comparing sorafenib with a placebo both revealed that sorafenib offered greater survival benefits in such patients (9, 10). However, findings regarding OS in our study may be confounded by lead-time bias because patients without MVI and EHS might have continued TACE before the reimbursement criteria expansion, and OS would only be calculated after sorafenib was prescribed later. Nevertheless, two retrospective studies suggested that in patients with BCLC stage B disease refractory to TACE, conversion to sorafenib treatment was associated with longer OS compared with continuing TACE (13, 14).
Lead-time bias could not explain the longer TTD after the reimbursement criteria expansion. A retrospective study in patients with BCLC stage B HCC refractory to TACE reported a longer time to progression to advanced disease if sorafenib treatment was initiated earlier (14). Combining the literature data and our results, we still favor that initiating sorafenib treatment at an earlier disease stage is associated with better treatment efficacy to be a reasonable explanation of our findings. Sorafenib treatment had also been tested at second-line setting for patients with treated advanced HCC and was proved to possessed attenuated anti-tumor activity (15).
The other possibility is treatment pattern shift among BCLC C patients. From the OPTIMIS study (16), an observational study of patients who received TACE for HCC, many patients with BCLC C disease still chose to undergo TACE first. Delaying the initiation of systemic therapy by repeated TACE can lead to impaired liver function reserve (16, 17), which reduces the feasibility of further systemic therapy. After the treatment options for advanced HCC expanded (18, 19), possibly more patients with BCLC C disease initiated systemic therapy directly.
In the global registry study GIDEON, approximately 30% of patients initiated sorafenib treatment at a dose lower than the standard (20). We observed that sorafenib was prescribed at a reduced initial dose in more than half of patients, and even more so after the reimbursement criteria expansion. This illustrates that physicians in Taiwan have been relatively conservative in prescribing sorafenib. Although no randomized studies have confirmed the efficacy of starting sorafenib with a lower dose, such an approach has been evaluated for regorafenib, a multikinase inhibitor with a similar profile of molecular targets and adverse reactions, in patients with refractory metastatic colorectal cancer (21). We revealed that among patients who initiated treatment with a low sorafenib dose, 28% of them escalated to the standard dose. This implies that many physicians in Taiwan preferred to titrate the sorafenib dose actively according to the adverse reactions.
As sorafenib was the mainstream treatment for patients with advanced HCC, some experts applied locoregional therapy to these patients as well (22-24). Furthermore, applying locoregional therapy during sorafenib treatment was also a common strategy, which was demonstrated with favorable outcomes compared with sorafenib treatment alone (25-28). We also found that a considerable 21% of patients received TACE during sorafenib treatment. A randomized-controlled phase III study compared combining TACE and sorafenib treatment for patients with advanced HCC against sorafenib alone; the results indicated the group receiving the combined intervention exhibited a higher disease control rate and longer progression-free survival, but without significant OS improvement (29). The combined treatment also increased the risk of adverse events resulting from TACE and warranted careful patient selection (29). Many guidelines in Asia endorse judicious TACE in patients with minor portal vein thrombosis (5, 30-33).
This study had some limitations. Patients in two eras were imbalanced in demographics and initial sorafenib dose. This is not surprising since the criteria were expanded to include patients without MVI or EHS, but also led to inevitable biases. Nevertheless, the survival benefits of expansion of sorafenib reimbursement were observed across different patient subgroups in subgroup analysis. In multivariate analysis, we found that after adjusting all these clinical variables, receiving sorafenib after the criteria expansion remained an independent predictor of favorable OS and TTD. To examine whether sorafenib efficacy is truly better in patients without MVI or EHS, a direct comparison of such patients with patients with MVI or EHS would be more ideal. However, such an approach is not feasible using the NHI database. Second-line therapies for advanced HCC emerged since 2017 (34-38). Patients who received sorafenib treatment after the reimbursement criteria expansion might have been more likely to receive subsequent therapies. However, no second-line therapies were reimbursed by the NHI in Taiwan until 2019, long after the last follow-up day of our study. The influence of clinical trials may also be limited because this is a nationwide study. Besides the aforementioned possible explanations of our findings, physicians might be more familiar with sorafenib administration and the management of side effects in the later period of this study, which may confound the survival outcomes between two eras.
In conclusion, we demonstrated that HCC patients who received sorafenib treatment after the broadening of reimbursement criteria exhibited longer OS and TTD. Whether it resulted directly from sorafenib treatment efficacies in patients without MVI or EHS remained to be explored. Initiating systemic therapy sooner in patients unsuitable or refractory to locoregional therapy for HCC is a reasonable strategy and supported by the results of this study.
Acknowledgements
The Authors would like to acknowledge the service provided by the RCF5 Lab. of the Department of Medical Research at National Taiwan University Hospital, Taiwan, R.O.C.
Footnotes
Authors’ Contributions
All Authors contributed to the study conception and design. Material preparation and data analysis were performed by Ching-Tso Chen and Yu-Yun Shao. The first draft of the manuscript was written by Ching-Tso Chen and all Authors commented on previous versions of the manuscript. All Authors approved the final manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
Funding
This study was supported by grants from the Ministry of Science and Technology, Taiwan (MOST-103-2314-B-002-181-MY2, MOST-105-2314-B-002-194, MOST-106-2314-B-002-213, MOST-108-2314-B-002-072-MY3, and MOST 111-2314-B-002-120), National Taiwan University Hospital (CTC202001), and Ministry of Health and Welfare (MOHW109-TDU-B-211-114002).
- Received July 7, 2022.
- Revision received July 23, 2022.
- Accepted July 25, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.