Cost-effectiveness analysis of anlotinib plus penpulimab versus sorafenib in the treatment of unresectable hepatocellular carcinoma

ABSTRACT

The APOLLO trial demonstrated that anlotinib plus penpulimab significantly prolonged progression-free survival (PFS) and overall survival compared to sorafenib in patients with unresectable hepatocellular carcinoma (HCC). However, its cost-effectiveness remains uncertain. This study aimed to evaluate the cost-effectiveness of this regimen from the perspective of the Chinese healthcare system. We constructed a partitioned survival model comprising three health states to evaluate the cost-effectiveness of anlotinib plus penpulimab versus sorafenib for unresectable HCC. Over a 10-y time horizon, we compared the total costs, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) between two groups. The robustness of the results was validated through one-way sensitivity analysis and probabilistic sensitivity analysis (PSA). Compared to sorafenib, anlotinib plus penpulimab provided an additional 0.21 QALYs at an incremental cost of $18,194.86. This resulted in an ICER of $86,546.80/QALY. One-way sensitivity analysis revealed that the utility value for PFS exerted the greatest influence on the model results, followed by penpulimab prices and the disutility due to adverse events (grade ≥3) in the anlotinib plus penpulimab group. PSA indicated a 0% probability of the anlotinib plus penpulimab regimen being cost-effective at a willingness-to-pay threshold of $40,335/QALY. Scenario analysis results showed that the Patient Assistance Program of penpulimab could help the regimen achieve favorable cost-effectiveness. Compared with sorafenib, anlotinib plus penpulimab for unresectable HCC patients was unlikely cost-effective under the perspective of the Chinese healthcare system.

Introduction

Hepatocellular carcinoma (HCC) accounted for 865,269 newly diagnosed cases and 757,948 deaths globally, positioning it as the sixth most prevalent malignancy and the third highest cancer-related mortality rate worldwide.¹ The global number of HCC cases is projected to increase by 50% between 2020 and 2040.² China carries a disproportionate global burden of HCC, accounting for 45.3% of worldwide HCC cases and 47.1% of HCC-related deaths. These epidemiological data leading HCC as the fourth most prevalent incident malignancy and the second leading cause of cancer mortality in China.^2,3 Only 30–40% of HCC cases are diagnosed at an early stage, while most patients already have advanced-stage disease at diagnosis.⁴ Advanced HCC carries a dismal prognosis,⁵ with a 5-y survival rate of merely 22%.⁶

The Food and Drug Administration (FDA) initially approved sorafenib in 2007 for advanced HCC.⁷ Subsequent years witnessed regulatory clearance of multi-target oral tyrosine kinase inhibitors (TKIs) – regorafenib⁸ (2017), lenvatinib⁹ (2018) and cabozantinib¹⁰ (2019) – for advanced HCC management. Nevertheless, molecular-targeted therapies demonstrate constrained clinical efficacy, narrow patient benefit cohorts, and inevitable acquired resistance.^11,12 The advent of immune checkpoint inhibitors (ICIs) has fundamentally transformed the therapeutic landscape.¹² Despite this, phase III trials of nivolumab¹³ and pembrolizumab¹⁴ versus sorafenib failed to demonstrate overall survival (OS) benefits. In contrast, the groundbreaking IMbrave150 trial established atezolizumab plus bevacizumab combination therapy achieved unprecedented survival advantages, which marked a pivotal shift toward ICI-based combinatorial approaches as the new standard of care in unresectable HCC treatment.¹⁵

The APOLLO trial investigated the efficacy and safety of anlotinib plus penpulimab versus sorafenib in patients with unresectable HCC. Anlotinib plus penpulimab demonstrated remarkable improvements in both progression-free survival (PFS) (median PFS: 6.9 vs. 2.8 months; hazard ratio [HR] = 0.52) and OS (median OS: 16.5 vs. 13.2 months; HR = 0.69) compared to sorafenib. Furthermore, the incidence of grade ≥3 adverse events (AEs) was comparable between the two groups (59% vs. 55%).¹⁶

These findings position anlotinib plus penpulimab as an attractive therapeutic option for unresectable HCC. However, over 70% of Chinese HCC patients face unaffordable financial burdens due to treatment costs,¹⁷ and the high cost of penpulimab and anlotinib further exacerbates this challenge. Therefore, this study aimed to evaluate the cost-effectiveness of the anlotinib plus penpulimab versus sorafenib for unresectable HCC from the perspective of the Chinese healthcare system.

Materials and methods

This study was conducted in accordance with the Consolidated Health Economic Evaluation Reporting Standards (CHEERS 2022) guidelines¹⁸ (Supplementary Material Appendix I). Ethical approval from an institutional review board (IRB) or patient consent was not required because the data used in this study came from previously published studies and did not involve identifiable patient data.

Patient population and intervention

Consistent with the APOLLO trial,¹⁶ we modeled a cohort of patients aged 18–75 y with histologically or cytologically confirmed locally advanced metastatic or unresectable HCC. Additional inclusion criteria were Child-Pugh class A or B7 liver function; no prior systemic therapy for advanced disease; a baseline Eastern Cooperative Oncology Group performance status of 0 or 1; an expected survival of at least 3 months; and adequate hematological and organ function.

Patients randomized to the anlotinib plus penpulimab group received oral anlotinib (10 mg orally once daily on days 1–14 of a 21-d cycle), combined with intravenous penpulimab (200 mg on day 1, for up to 2 y). Those assigned to the sorafenib group received sorafenib (400 mg orally twice daily). Treatment was discontinued upon unacceptable toxicity or confirmed progressed disease (PD). Subsequent second-line therapy was administered to patients in both groups according to the proportions reported in the APOLLO trial (37% in the anlotinib plus penpulimab group versus 44% in the sorafenib group), while those not receiving second-line therapy transitioned to best supportive care (BSC). It was assumed that all deceased patients in both groups received equivalent end-of-life care.

To align with clinical practice guidelines in China, we focused on the major therapeutic categories with clear report in the APOLLO trial: PD-(L)1 inhibitors, targeted therapy (monotherapy), traditional Chinese medicine, and interventional therapy. Minor categories such as surgical resection, combined targeted therapy plus PD-(L)1 inhibitors, ablation therapy, and others were excluded due to their low incidence (<1%–2%) or lack of a standardized treatment definition. The proportion for each second-line regimen was calculated as its reported incidence percentage divided by the sum of all selected incidence percentages within that treatment group. For example, in the anlotinib plus penpulimab group, the proportion for PD-(L)1 inhibitors is 13% divided by the total of 59%, yielding 22.03%. The choice of specific agents (e.g., regorafenib for targeted therapy,⁸ pembrolizumab for PD-1 inhibition,¹⁹ Icaritin soft capsules for traditional Chinese medicine,²⁰ and transarterial chemoembolization (TACE) for interventional therapy)²¹ was based on the National Comprehensive Cancer Network (NCCN) guidelines,²² the Chinese Society of Clinical Oncology (CSCO) guidelines²³ and the Chinese Standard for Diagnosis and Treatment of Primary Liver Cancer²⁴ (Table 1).

Table 1.

Second-line treatment regimen and duration.

Group	Drug/Treatment	Dosage	Frequency of administration	Duration	Proportion	References
Anlotinib plus penpulimab group	Pembrolizumab	200 mg	Every three weeks	3.48 months	22.03%	19
	Regorafenib	160 mg	21–d oral daily, 7–daybreak	3.6 months	38.98%	8
	Icaritin	600 mg	Twice a day orally	2.79 months	18.64%	20
	TACE	/	Every six weeks	3 months	20.34%	21
Sorafenib group	Pembrolizumab	200 mg	Every three weeks	3.48 months	28.38%	19
	Regorafenib	160 mg	21–d oral daily, 7–daybreak	3.6 months	35.14%	8
	Icaritin	600 mg	Twice a day orally	2.79 months	17.57%	20
	TACE	/	Every six weeks	3 months	18.92%	21

Model structure

In line with recommendations from the National Institute for Health and Care Excellence (NICE)²⁵ and the China Guidelines for Pharmacoeconomic Evaluations (2020),²⁶ partitioned survival models are preferred for cost-effectiveness analyses of interventions with a limited health states. Accordingly, we constructed a partitioned survival model using TreeAge Pro 2019, including three mutually exclusive health states: PFS, PD, and death, to compare the cost-effectiveness of different interventions for unresectable HCC (Figure 1). Patient survival proportions were estimated using the area under the curve (AUC) of OS data from the APOLLO trial, while PFS proportions were derived from the PFS AUC. The proportion of patients with PD proportions were calculated as the difference between OS and PFS. Following the WHO-CHOICE recommendations²⁷ and the China Guidelines for Pharmacoeconomic Evaluations (2020),²⁶ the willingness-to-pay (WTP) threshold in China was set at $40,335/QALY, equivalent to three times the 2024 per capita GDP in China.²⁸

Figure 1.

Partitioned survival model consisting of three health states.

Abbreviations: HCC, hepatocellular carcinoma.

The model cycle length was set at 21 d to align with the treatment cycles in the APOLLO trial. We set a 10-y time horizon for the model because most of patients died within this period. The primary model output included total costs, quality-adjusted life-years (QALYs), and the incremental cost-effectiveness ratio (ICER).

Clinical data

We used the automated extraction software SurvdigitizeR (https://pechlilab.shinyapps.io/SurvdigitizeR/.), developed by Zhang et al.²⁹ to extract data points from the Kaplan–Meier (KM) curves for OS and PFS reported in the APOLLO trial. These data points were then reconstructed into pseudo-individual patient data using the algorithm developed by Guyot.³⁰ These pseudo-individual patient data were then fitted to seven parametric survival distributions: Exponential, Gamma, Gompertz, Weibull, Log-logistic, Log-normal, and Generalized gamma (Supplementary Material Appendix II). The Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), and visual inspection were used to select the optimal survival function. The fitting results and model extrapolation outcomes can be found in Supplementary Material Appendix III (Figure S1–Figure S6). Lower AIC and BIC values, combined with a visual assessment of the alignment between the reconstructed and reported survival curves, indicated a superior fit for the chosen parametric model.^31–33

Cost

Given the perspective of the Chinese healthcare system adopted in our study, only direct medical costs were considered. Direct medical costs including costs of drugs, TACE, follow-up and monitoring, management of AEs, BSC, and end-of-life care. Table 2 lists all AEs from the APOLLO trial, along with their respective incidence rates and severity. We assumed that AEs occurred exclusively during the first model cycle.³⁴ The management costs for AE were exclusively applied to grade ≥3 events because these AEs often require necessitate significant medical interventions, such as hospitalization or administration of supportive care medications. In contrast, grade 1–2 AEs are often self-managed, and managed with minimal resources,³⁵ resulting in negligible cost implications. This approach ensures that our cost estimation reflects the true economic burden driven by clinically significant events.

Table 2.

All AEs from the APOLLO trial.

	Rate of Anlotinib plus penpulimab group (%)		Rate of Sorafenib group (%)
	Grade 1–2	Grade ≥3	Grade 1–2	Grade ≥3
Hypertension	27.08	17.36	27.49	10.43
Platelet count decreased	29.63	9.03	27.01	6.16
Aspartate aminotransferase concentrations increased	30.56	4.17	27.96	6.16
Blood bilirubin concentrations increased	25.69	5.32	25.12	2.37
White blood cell count decreased	22.69	5.56	16.11	3.32
Alanine aminotransferase concentrations increased	24.77	2.31	18.96	3.32
Hypothyroidism	20.60	0.00	5.21	0.00
Neutrophil count decreased	13.66	5.79	9.00	3.32
Palmar-plantar erythrodysesthesia syndrome	17.13	1.85	24.17	8.06
Hypoalbuminaemia	17.82	0.69	11.37	0.00
Weight loss	17.82	0.46	11.85	1.42
Proteinuria	17.36	0.69	14.22	0.00
Diarrhea	13.43	3.24	16.59	0.95
Thyroid-stimulating hormone concentrations increased	15.74	0.00	6.16	0.00
Anaemia	12.96	2.08	8.06	1.90
Lymphocyte count decreased	10.19	3.94	6.64	1.90
Gamma-glutamyltransferase concentrations increased	12.04	1.62	10.90	2.37
Fatigue	12.96	0.23	11.37	0.95
Occult blood positive	12.50	0.00	11.85	0.00
Anorexia	10.88	0.46	8.53	0.47
Alkaline phosphatase concentrations increased	10.65	0.23	12.32	0.95
Rash	10.19	0.46	4.27	0.95
Hypokalaemia	7.87	2.55	6.64	2.84

Drug costs were derived from the median bid prices across Chinese provinces using the YAOZHI database,³⁵ while other costs were sourced from published studies. A 5% annual discount rate was applied to all costs and outcomes.²⁶ All expenditures were adjusted to 2025 values using the Consumer Price Index (CPI) of China and converted to US dollars (1 USD = 7.2776 CNY).

Utility

Health utility values for PFS and PD were derived from published studies,³¹ as these were not reported in the APOLLO trial. Utility values were assigned as follows: 0.76 for PFS, 0.68 for PD, and 0 for death. Additionally, disutilities associated with AEs were incorporated to account for their impact on quality of life³⁶ (Table 3).

Table 3.

Base-case model parameter.

Parameter	Baseline value	Minimum	Maximum	Distribution	Reference
Survival data distribution parameter input
Log-logistic OS survival model of anlotinib plus penpulimab group	Shape = 1.6031, scale = 16.6560	NA	NA	Fixed	16
Log-logistic OS survival model of sorafenib group	Shape = 1.339, scale = 11.743	NA	NA	Fixed	16
Log-normal PFS survival model of anlotinib plus penpulimab group	Meanlog = 1.8643, sdlog = 1.0894	NA	NA	Fixed	16
Log-logistic PFS survival model of sorafenib group	Shape = 1.758, scale = 3.254	NA	NA	Fixed	16
Costs input
Diagnosisa	4.67	3.74	5.60	Gamma	37
Nursinga	1.16	0.93	1.39	Gamma	37
Beda	1.87	1.50	2.24	Gamma	37
Intravenous injectiona	0.71	0.36	1.56	Gamma	38
Enhanced CTa	21.30	17.04	25.56	Gamma	39
Routine blood testa	1.42	1.14	1.70	Gamma	39
Blood biochemistry a	22.09	17.67	26.51	Gamma	37
Urine routinea	0.57	0.46	0.68	Gamma	39
Thyroid functiona	8.52	6.82	10.22	Gamma	39
12 lead electrocardiographya	3.13	2.50	3.76	Gamma	39
Coagulation functiona	5.68	4.54	6.82	Gamma	39
Alpha fetoproteina	2.27	1.82	2.72	Gamma	39
Myocardial enzyme testa	4.67	3.74	5.60	Gamma	37
BSC	130.38	104.30	156.46	Gamma	40
End-of-life careb	1,873.21	1,498.57	2,247.85	Gamma	40
TACE treatment	1,936.72	1,549.38	2,324.06	Gamma	41
Anlotinib (10 mg)	35.16	28.13	42.19	Gamma	35
Penpulimab (100 mg)	580.21	464.17	696.25	Gamma	35
Sorafenib (200 mg)	12.27	9.82	14.72	Gamma	35
Pembrolizumab (100 mg)	2,462.08	1,969.66	2,954.50	Gamma	35
Regorafenib (40 mg)	23.70	18.96	28.44	Gamma	35
Icaritin (100 mg)	2.34	1.87	2.81	Gamma	35
Hypertension	12.22	9.78	14.66	Gamma	40
Platelet count decreased	160.74	128.59	192.89	Gamma	37
Blood bilirubin concentrations increased	70.36	56.29	84.43	Gamma	37
White blood cell count decreased	458.92	367.14	550.70	Gamma	42
Neutrophil count decreased	387.72	310.18	465.26	Gamma	42
Palmar-plantar erythrodysesthesia syndrome	15.27	12.22	18.32	Gamma	40
Aspartate aminotransferase concentrations increased	70.36	56.29	84.43	Gamma	37
Utility
PFS	0.76	0.61	0.91	Beta	40
PD	0.68	0.54	0.82	Beta	40
Disutility of AEs
Grade 1–2	0.014	0.0112	0.0168	Beta	31
Grade ≥3	0.157	0.1256	0.1884	Beta	36
Subsequent active treatment rate
Anlotinib plus penpulimab group (%)	37	30	44	Beta	16
Sorafenib group (%)	44	35	53	Beta	16

Abbreviations: OS, overall survival; PFS, progression-free survival; CT, computed tomography; AE, adverse events; PD, progressed disease; BSC, best supportive care; TACE, transarterial chemoembolization.

^aIt is assumed that these costs will continue until the health state transitions.

^bOverall total cost per patient regardless of treatment duration

Sensitivity analysis

To assess the robustness of the cost-effectiveness results and identify key parameters of uncertainty, we conducted one-way sensitivity analysis and probabilistic sensitivity analysis (PSA).

In the one-way sensitivity analysis, the impact of each parameter on model outcomes was evaluated by varying it individually while holding the others constant. If available, parameters were adjusted within their reported 95% confidence intervals; otherwise, a range of ±20% from baseline values were applied. The results were visualized using a tornado diagram to rank parameters by their influence on the ICER. We set a range of 0–8% variation for the discount rate.²⁶ For the PSA, we conducted 1,000 Monte Carlo simulations, performed by randomly sampling parameters from predefined distributions: costs were assigned a gamma distribution, while probabilities, proportions, and utility values were modeled using beta distribution. The results of the PSA were presented as ICER scatter plot and a cost-effectiveness acceptability curve (CEAC). CEAC shows the probability that each intervention is cost-effective at different WTP thresholds.

Scenario analysis

Scenario analyses were performed for the model’s time horizon (5 and 15 y) to evaluate the robustness of the model’s outcomes. Furthermore, since penpulimab has not yet been included in China’s National Reimbursement Drug List (NRDL), we simulated a price reduction scenario assuming its post-NRDL approval. Based on data disclosed by China’s National Healthcare Security Administration, the average price reduction achieved through the 2024 national reimbursement negotiations was 63%.⁴³ Additionally, we incorporated penpulimab’s Patient Assistance Program (PAP) into the scenario analysis. Specifically, the initial assistance phase follows a “2 + 1” model, whereby patients receive one treatment cycle free when purchasing two cycles. Subsequent assistance adopts a “2+PD” model, providing continuous drug support after two treatment cycles until PD, with a maximum duration of 2 y.⁴⁴ Anlotinib’s PAP was excluded from this study due to the announced termination of its program effective December 31, 2024.⁴⁵ Finally, to test the robustness of our findings, we conducted a scenario analysis using alternative utility values sourced from the IMbrave150 study⁴⁶: PFS utility of 0.78 for the anlotinib plus penpulimab group, 0.77 for the sorafenib group, and a PD utility of 0.72.

Results

Base-case analysis and scenario analysis

Base-case and scenario analysis results are presented in Table 4. In the base-case analysis, the anlotinib plus penpulimab regimen provided an additional 0.21 QALYs at an incremental cost of $18,194.86, resulting in an ICER of $86,546.80/QALY, which substantially exceeded the WTP threshold of China. Scenario analyses across different time horizons demonstrated persistently unfavorable ICERs: $88,649.51/QALY (5-y horizon) and $89,023.00/QALY (15-y horizon). Additionally, our results demonstrate that despite a potential price reduction for penpulimab following its inclusion in the NDRL, anlotinib plus penpulimab remained not cost-effective ($42,366.86/QALY). It was only when the PAP was incorporated into the analysis that the ICER fell to $34,984.47/QALY, making the regimen a cost-effective option. The scenario analysis using alternative utility values yielded an ICER of $84,527.81/QALY. Although this represents a reduction from the base-case ICER of $86,546.80/QALY, the conclusion of the analysis nevertheless remained unaltered.

Table 4.

The cost and outcome results of the base-case analysis and scenario analysis.

	Total cost ($)		Overall QALYs		ICER ($/QALY)
	Anlotinib plus penpulimab	Sorafenib	Anlotinib plus penpulimab	Sorafenib
Base-case Analysis	28,380.78	10,185.92	1.25	1.04	86,546.80
Scenario Analysis
5 y	27,429.77	9,458.08	1.09	0.89	88,649.51
15 y	28,657.59	10,452.86	1.31	1.10	89,023.00
Penpulimab at price of NRDL	19,092.77	10,185.92	1.25	1.04	42,366.86
Penpulimab at price of PAP	17,540.76	10,185.92	1.25	1.04	34,984.47
Alternative utility values	28,380.78	10,185.92	1.34	1.12	84,527.81

Abbreviations: NRDL, China’s National Reimbursement Drug List; PAP, patient assistance program; QALY, quality-adjusted life-years; ICER, incremental cost-effectiveness ratio.

Sensitivity analysis

One-way sensitivity analysis identified the utility value of PFS as the most influential parameter on model outcomes, followed by penpulimab prices, the disutility due to AEs (grade ≥3) in the anlotinib plus penpulimab group, the utility value of PD, the disutility due to AEs (grade ≥3) in the sorafenib group and anlotinib price. Other parameters had a marginal impact (Figure 2). However, no single parameter reduced the ICER below the WTP threshold, which reinforced the robustness of the base-case analysis findings. In the PSA involving 1,000 Monte Carlo simulations, all simulated ICER estimates on the scatter plot fell above the WTP threshold. This result shows that the anlotinib plus penpulimab was not cost-effective compared to sorafenib in any of the iterations (Figure 3). The CEAC showed a 0% probability of cost-effectiveness for the anlotinib plus penpulimab regimen under China’s WTP threshold (Figure 4). Considering the potential price reduction of penpulimab after its inclusion in the NRDL, the probability of the anlotinib plus penpulimab regimen being cost-effective was 43.0%. When considering the subsidies of PAP, this probability was 75.5%.

Figure 2.

Tornado diagram of one-way sensitivity analysis for anlotinib plus penpulimab versus sorafenib.

Abbreviations: ICER, incremental cost-effectiveness ratio; PAP, patient assistance program; A: anlotinib; P: penpulimab; S: sorafenib; QALY: quality-adjusted life years; AE: adverse events; PD: progressed disease; PFS: progression-free survival.

Figure 3.

The incremental cost-effectiveness scatters plot of anlotinib plus penpulimab versus sorafenib.

Abbreviations: WTP, willingness-to-pay; QALY: quality-adjusted life years.

Figure 4.

Cost-effectiveness acceptability curves for anlotinib plus penpulimab versus sorafenib.

Abbreviations: PAP, patient assistance program; WTP, willingness-to-pay; QALY: quality-adjusted life years.

Discussion

The advent of ICIs has substantially transformed the treatment landscape for unresectable HCC.¹² While the success of atezolizumab plus bevacizumab has established ICI-based combination therapy as the standard of care for unresectable HCC, not all combination regimens have proven effective.^47,48 The dual benefits in OS and PFS demonstrated by anlotinib plus penpulimab offer a new treatment option for unresectable HCC patients. Although the anlotinib plus penpulimab regimen has been approved by the NMPA for unresectable HCC, there is insufficient evidence to enable healthcare policymakers, healthcare providers and patients to determine whether this new treatment is cost-effective.

Our findings indicated that compared to sorafenib, anlotinib plus penpulimab provides an additional 0.21 QALYs for patients with unresectable HCC. However, this comes with an incremental cost of $18,194.86, resulting in an ICER of $86,546.80/QALY. This significantly exceeds the commonly accepted WTP threshold in China. Consequently, from the perspective of the Chinese healthcare system, anlotinib plus penpulimab is unlikely to be a cost-effective option. This conclusion was robust, as supported by both one-way sensitivity analysis and PSA.

Our findings align with the majority of published pharmacoeconomic studies. Wu’s study demonstrated that atezolizumab plus bevacizumab provided an additional 0.811 QALYs compared to sorafenib. However, the incremental cost of $49,994 resulted in an ICER of $61,613/QALY, exceeding China’s 2019 WTP threshold ($30,828/QALY).⁴⁰ Xu’s results indicated that sintilimab plus bevacizumab biosimilar generated an additional 0.33 QALYs at an incremental cost of $17,552.17 in China, corresponding to an ICER of $52,817.89/QALY. The probability of this regimen being cost-effective at China’s WTP threshold ($38,334/QALY) was only 1.28%.⁴⁹ Furthermore, Sun’s cost-effectiveness analysis, based on a network meta-analysis of five first-line unresectable HCC treatments, revealed that despite ICI-based combinations providing greater QALY gains, lenvatinib remained the preferred option for Chinese unresectable HCC patients.⁵⁰ This conclusion was corroborated by similar findings in Zhao’s research.⁵¹

In contrast, several cost-effectiveness analyses have demonstrated that the combination of camrelizumab and apatinib is a cost-effective regime compare with sorafenib.^52–54 It is noteworthy that both camrelizumab and apatinib are included in the NRDL, and the substantial price reduction following their inclusion likely explains this finding. However, in our analysis, even under the assumption of a 63% price reduction for penpulimab upon its inclusion in the NRDL, the anlotinib plus penpulimab regimen remained not cost-effective compared to sorafenib. Only the scenario incorporating the PAP resulted in an ICER below the WTP threshold. Therefore, to enhance patient access to novel therapies, we recommend that policymakers consider formally integrating such financial assistance mechanisms into the healthcare financing framework. Simultaneously, pharmaceutical companies should develop aggressive pricing and market access strategies to ensure affordability upon NRDL inclusion.

The scenario analysis employing alternative utility values did not alter the conclusion of the base-case analysis, thereby reinforcing the credibility of our primary findings. However, the inherent and unavoidable differences in cultural backgrounds and preferences across study populations underscore the necessity for future research to derive utility values specific to their respective cohorts. Furthermore, our scenario analyses across different time horizons (5, 10, and 15 y) revealed minimal fluctuation in the ICERs ($88,649.51/QALY, $86,546.80/QALY and $89,023.00/QALY, respectively). This stability stands in contrast to the findings of Xiong, who evaluated the cost-effectiveness of tremelimumab plus durvalumab for unresectable HCC in the United States and reported substantial ICER variations across 1-y, 2-y, 4-y, and lifetime horizons ($375,584.97/QALY, $53,106.25/QALY, $97,995.51/QALY and $110,652.15/QALY, respectively).⁵⁵ This discrepancy could be attributed to the fundamental difference in the therapeutic mechanisms between the two regimens. The combination of anlotinib and penpulimab may demonstrate a distinct long-term survival profile and associated cost accumulation pattern compared to the dual-immunotherapy approach studied by Xiong, potentially leading to different ICER trajectories over time. Given the current lack of head-to-head trials comparing these strategies, our findings highlight a critical evidence gap. We therefore call for future studies to perform systematic indirect comparisons or network meta-analyses to rigorously evaluate their relative effectiveness, especially in terms of long-term survival.

In the one-way sensitivity analysis, the utility value of PFS had the greatest impact on our model results, followed by the prices of penpulimab and the disutility due to AEs (grade ≥3) in the anlotinib plus penpulimab group. While these parameters significantly influenced the model outcomes, no variation in a single parameter within plausible ranges was sufficient to reduce the ICER below the WTP threshold. While both the price of anlotinib and penpulimab are key parameters influencing the model results, the potential for further price reductions of anlotinib is limited. Since its initial inclusion in the NRDL in 2018, anlotinib has undergone two price reductions, dropping from $10.14 to $3.52 per mg. Furthermore, the PAP for anlotinib was due to end in 2024, so further significant price reductions are unlikely. Conversely, penpulimab offers significant cost-saving potential. It was first approved by the NMPA in August 2021 for relapsed/refractory classical Hodgkin lymphoma and approved in 2025 in combination with anlotinib for the first-line treatment of unresectable HCC. However, it has yet to be included in the NRDL. Although the inclusion of the hypothetical penpulimab in the NRDL did not reduce the ICER below the WTP threshold, the substantial price reduction led to a dramatic decrease in the ICER (from $86,546.80/QALY in the base case to $42,366.86/QALY). This marked improvement significantly alleviates the economic burden for Chinese patients with unresectable HCC, moving the anlotinib plus penpulimab regimen closer to being a cost-effective option.

Our study has several key strengths. First, to our knowledge, this is the first cost-effectiveness analysis evaluating anlotinib plus penpulimab versus sorafenib for the treatment of unresectable HCC, providing economic evidence for the treatment options of unresectable HCC. Second, we conducted scenario analyses on utility values, and the findings were consistent with the base-case results. This suggests that while differences in baseline population characteristics and cultural preferences may influence the specific ICER value, they do not change the overall study conclusions, thereby reinforcing the credibility of our findings. Third, we simulated penpulimab pricing under different policy scenarios, and the results offer valuable insights for policymakers. Fourth, unlike most cost-effectiveness analyses that only consider grade 3 or higher AEs exceeding a preset incidence threshold,^34,56 our model incorporated AEs of all grades and any reported incidence. This approach provides a more comprehensive reflection of the treatment’s safety profile and enhances the real-world relevance of our results.

Nevertheless, our study has several limitations. First, our study has methodological limitations. The use of digitization software to extract individual patient data from published KM curves may introduce bias due to variations between individual extractors. To minimize this bias, we used the SurvdigitizeR program developed by Zhang et al.²⁹ an automated extraction tool that ensures the reliability and reproducibility of the extraction results. Second, there are limitations related to quality of life. As health-related quality of life data were not published in the APOLLO trial, the utility values for PFS and PD were obtained from other published studies. Third, there are cost limitations. We excluded the management costs of grade 1 and 2 AEs. This may have led to an underestimation of the total cost of the anlotinib plus penpulimab regimen, consequently overestimating its cost-effectiveness.⁵⁶ Finally, our safety analysis was comprehensive in cataloging the incidence and severity of AEs but did not explore potential correlations between treatment-emergent AEs and treatment efficacy. This is a recognized limitation, as evidence from the literature suggests that certain treatment-related AEs may have prognostic significance. For instance, in patients treated with sorafenib, the development of specific AEs such as hypertension or hand-foot skin reaction has been associated with improved survival outcomes, potentially serving as a clinical biomarker of treatment response.⁵⁷ While similar analyses for the anlotinib plus penpulimab regimen were beyond the scope of our current economic model and the available trial data, acknowledging this relationship is crucial. It underscores a complex dimension of oncology care where treatment toxicity and efficacy can be intertwined. Future clinical studies of this combination should investigate such correlations, as they could provide valuable insights for patient management and prognostication. Despite these limitations, the ICER for the anlotinib plus penpulimab regimen was double the WTP threshold. Furthermore, our findings were validated through sensitivity analyses, indicating that these limitations are unlikely to alter our study’s primary conclusion. Therefore, our results still provide important economic insights for policymakers, pharmaceutical companies, and healthcare providers.

Conclusion

From the perspective of the Chinese healthcare system, the anlotinib plus penpulimab regimen was unlikely to be a cost-effective option compared to sorafenib for patients with unresectable HCC. Nevertheless, the PAP for penpulimab is expected to enable this regimen to achieve cost-effectiveness.

Biography

Jiefeng Luo is a pharmacist at Fudan University Shanghai Cancer Center. His research interests include evidence synthesis, pharmacoeconomics and hospital pharmacy. He is actively engaged in multidisciplinary research aiming to optimize cancer treatment strategies and support evidence-based decision-making in healthcare. His ORCID is https://orcid.org/0000-0001-9658-5100.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data generated or analyzed during this study are included in this published article and its supplementary materials. The model code and input parameters are available from the corresponding author upon reasonable request.

Supplementary materials

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2025.2582294

Ethics approval and consent to participate

Ethical approval and informed consent were not required for this study, as all data was obtained from previously published sources. The study did not involve any human participants, human tissue, or identifiable personal data.