Cost‑effectiveness analysis of tislelizumab plus chemotherapy in Chinese patients with advanced or metastatic oesophageal squamous cell carcinoma

Abstract

The RATIONALE-306 study revealed that patients with advanced or metastatic oesophageal squamous cell carcinoma (OSCC) could benefit from treatment with tislelizumab plus chemotherapy. This study aimed to evaluate the cost-effectiveness of tislelizumab plus chemotherapy for treating OSCC from the perspective of the Chinese healthcare system. Partitioned survival model estimated the cost-effectiveness of tislelizumab plus chemotherapy compared with chemotherapy alone for treating OSCC using RATIONALE-306 data. Costs and utilities were obtained from local databases and published studies. Costs, quality-adjusted life-years (QALYs), life-years, incremental cost-effectiveness ratios (ICER), incremental net health benefits (INHB), and incremental net monetary benefits (INMB) were outcomes. Price simulation were conducted at the willingness-to-pay (WTP) threshold. Sensitivity and subgroup analyses were performed to assess model robustness. Compared with chemotherapy alone, tislelizumab plus chemotherapy yielded an ICER of USD 27,896/QALY, gained an additional 0.414 QALYs and 0.751 life-years, and increased the cost by USD 11,560. Probabilistic sensitivity analysis revealed that tislelizumab plus chemotherapy was cost-effective at the WTP of USD 38,258/QALY with probability of 94.43%. When the price in China was less than USD 3.714 per mg, the price simulation results indicated that tislelizumab plus chemotherapy was cost-effective at a WTP threshold of USD 38,258. Tislelizumab plus chemotherapy yielded an INHB of 0.112 QALYs and an INMB of USD 4,279 compared with chemotherapy alone at a WTP threshold of USD 38,258. Based on the sensitivity analyses, the above results were stable. A general trend was observed for subgroups with better survival benefits related to a higher probability of cost-effectiveness. From the Chinese healthcare perspective, tislelizumab plus chemotherapy is more cost-effective than chemotherapy alone as a first-line therapy for OSCC. These findings can help clinicians make optimal clinical decisions and assist decision-makers in evaluating the cost-effectiveness of tislelizumab in clinical practice.

Introduction

Oesophageal cancer is the tenth most common cancer worldwide and the sixth most common cause of cancer-related death¹. When patients are diagnosed with oesophageal cancer, the majority suffer from advanced or metastatic disease; the 5-year cancer survival rate is 26% in patients with metastatic disease in regional lymph nodes and 5% in patients with distant metastasizing disease². Asian populations (approximately 90%) are most likely to develop oesophageal squamous cell carcinoma (OSCC), with more than half of all cases occurring in China³. In contrast, oesophageal adenocarcinoma is more prevalent in Western European countries and North America^3,4. For decades, chemotherapy has been the standard treatment for patients with metastatic OSCC, and platinum-based chemotherapy has been the primary first-line treatment for OSCC⁵. However, for patients diagnosed with advanced stages of the disease, the overall survival (OS) rate is low, with a median survival time of less than 12 months and a survival rate of only 8% or less^6,7. In recent years, advances have been made in oesophageal cancer treatment, particularly in the development of immunotherapies such as programmed cell death-ligand 1 (PD-L1)/PD-1 inhibitors. PD-1 antibody combination with chemotherapy was related to superior survival benefit compared to chemotherapy alone as a first-line therapy for OSCC, and immune checkpoint inhibitors (ICIs) plus chemotherapy were related to superior antitumor clinical benefits compared to chemotherapy alone^6,8–10. Preclinical studies have reported that immunotherapy combined with chemotherapy can provide a synergistic antitumor effect^10,11, which restrains tumor cell immune escape and enhances the immune response of the host¹². According to the National Comprehensive Cancer Network (NCCN) guidelines¹³, immunotherapy is a recommended method for the treatment of advanced oesophageal cancer.

The emergence of ICIs has had a profound impact on cancer treatment. Tislelizumab is a fully human IgG4 monoclonal antibody with strong affinity and specificity for PD-1^14,15, and is related to superior clinical efficacy in several types of cancers. Several studies have demonstrated the efficacy of ICIs in combination with chemotherapy. KEYNOTE-590 and CheckMate 648, both of which treat global patients, reported superior clinical efficacy with pembrolizumab plus chemotherapy, nivolumab plus chemotherapy, or ipilimumab^10,16. ESCORT-1st, JUPITER-06, ORIENT-15, and ASTRUM-007 revealed that ICIs, including camrelizumab, toripalimab, sintilimab, and serplulimab, plus chemotherapy prolonged OS and progression-free survival (PFS) in Chinese patients^8,17–19. In the era of precision medicine, despite the remarkable rate of development of therapeutic sites with ICIs, the development of biomarkers has been relatively slow²⁰. Unfortunately, microsatellite instability and Epstein-Barr virus positivity are two of the most consistent predictors of response to checkpoint inhibition in oesophagogastric cancer and are almost absent in patients with squamous cell carcinomas²¹. In addition to novel biomarkers reflecting tumor biology, biomarkers that characterize the tumor host, such as the patients' gut microbiome composition, might be relevant to explore. In this context, the occurrence of side effects of reactive capillary endothelial proliferation or hypertransaminasemia of ICIs deserves further attention^21–23. Recently, a randomized, double-blinded, phase III trial, RATIONALE-306, evaluated the clinical efficacy and safety of tislelizumab plus chemotherapy compared to chemotherapy alone as the first-line therapy for advanced or metastatic OSCC⁶. RATIONALE-306 reported that tislelizumab plus chemotherapy significantly prolonged PFS (median, 7.3 vs. 5.6 months) and OS (median, 17.2 vs. 10.6 months), without unexpected treatment-related adverse events (AEs).

Although these clinical results are encouraging, economic evaluation is important, because tislelizumab combined with chemotherapy is relatively more expensive than chemotherapy alone. Hence, this study aimed to compare tislelizumab plus chemotherapy with chemotherapy alone in patients with advanced or metastatic OSCC from the perspective of the Chinese healthcare system. The results of this study will help clinicians and decision-makers allocate resources reasonably.

Methods

Clinical patients and intervention

This study followed the Consolidated Health Economic Evaluation Reporting Standards²⁴. The study was conducted on patients older than 18 years with histologically confirmed unresectable, locally advanced, recurrent, and/or metastatic OSCC without previous systemic therapy. Other key characteristics were common to RATIONALE-306⁶.

Based on the RATIONALE-306 trial, patients received tislelizumab 200 mg plus chemotherapy or chemotherapy alone. Chemotherapy included platinum, (cisplatin 60–80 mg/m² or oxaliplatin 130 mg/m²) plus fluoropyrimidine (fluorouracil [750–800 mg/m²] or capecitabine [1000 mg/m² twice daily]) or paclitaxel (175 mg/m²). Chemotherapy was recommended for a maximum of six cycles. First-line therapy was changed to subsequent treatments if the disease progressed or if intolerable side effects occurred. In the RATIONALE-306 trial, 48% (157/326) of patients in the tislelizumab combined with chemotherapy arm and 55% (177/323) of patients in the chemotherapy arm received subsequent therapies⁶.

Model construction

To compare the direct medical costs and clinical outcomes between tislelizumab plus chemotherapy and chemotherapy alone, we developed a partitioned survival model. The model includes three mutually exclusive health states: PFS, progressed disease (PD), and death. It was assumed that the time horizon of this model was 12 years, with 98% of the patients dead^25,26. The cycle length of the model was 1 week. Considering that the purpose of this analysis was to provide evidence for the direct medical costs associated with the use of tislelizumab plus chemotherapy and the related effectiveness in patients, we chose to perform this study from the perspective of the Chinese healthcare system. The outcomes were life-years, quality-adjusted life-years (QALYs), incremental cost-effectiveness ratio (ICER), incremental net health benefits (INHB), and incremental net monetary benefits (INMB). Annual discounts of 5% were applied to both costs and utilities²⁷. We set the willingness-to-pay (WTP) threshold to USD 38,258 per QALY, which is three times the per capita gross domestic product (GDP) of China. The model was built using R 4.0.2 (https://www.r-project.org/) with “hesim” and “heemod” packages.

Effectiveness

Based on the Kaplan–Meier (K–M) curves of the RATIONALE-306 trial, the probabilities of OS and PFS were calculated using the GetData Graph Digitizer (http://getdata-graph-digitizer.com). Guyot’s method was used to reconstruct the individual patient data (IPD) over the clinical trial²⁸. In the virtual IPD, events and times accounted for the same number of individuals at risk as those at risk at the outset, closely reproducing the digitized K–M curves. A parametric survival function was fitted to these data points: exponential, Weibull, gamma, lognormal, Gompertz, log-logistic, and generalized gamma. A good-fit assessment was conducted via visual inspection using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC)²⁷. Models with lower AIC and BIC values and appropriate visual results are more likely to fit well²⁹. The final survival function parameters of tislelizumab plus chemotherapy and chemotherapy alone are shown in Table 1, and the results of goodness-of-fit are illustrated in Supplementary Fig. 1 and Supplementary Table 1.

Table 1.

Key Model Inputs.

Parameter	Value (95% CI)	Distribution	Source
Log-logistic OS survival model of tislelizumab plus chemotherapya	γ = 1.6547, λ = 0.0130	NA	6
Log-logistic PFS survival model of tislelizumab plus chemotherapya	γ = 1.7730, λ = 0.0302	NA	6
Lognormal OS survival model of chemotherapya	μ = 3.9352, σ = 1.0317	NA	6
Lognormal PFS survival model of chemotherapya	μ = 3.0879, σ = 0.9022	NA	6
Cost input
Drug costs per 1 mg, USD
Tislelizumab	2.16 (1.62 to 2.7)	Gamma	Local database
Cisplatin	0.27 (0.2 to 0.34)	Gamma	Local database
Oxaliplatin	0.48 (0.36 to 0.6)	Gamma	Local database
Fluorouracil	0.065 (0.052 to 0.078)	Gamma	Local database
Capecitabine	0.0015 (0.0012 to 0.0018)	Gamma	Local database
Paclitaxel	0.79 (0.59 to 0.99)	Gamma	Local database
Second-line treatment in tislelizumab plus chemotherapy arm per cycle	426 (383 to 468)	Gamma	6 Local database
Second-line treatment in chemotherapy arm per cycle	373 (336 to 410)	Gamma	6 Local database
Routine follow-up cost per cycle	18.39 (13.79 to 22.99)	Gamma	30
Cost of laboratory tests and radiological examinations	357 (285 to 428)	Gamma	30
Beat supportive care per cycle	41.82 (31.37 to 52.28)	Gamma	30
Cost of terminal care per patientb	1460 (1168 to 1752)	Gamma	34
Cost of managing grade ≥ 3 AEsc
Tislelizumab plus chemotherapy	4886 (3909 to 5863)	Gamma	31–33
Chemotherapy	5113 (4090 to 6135)	Gamma	31–33
Cost of drug administration per unit	20.64 (15.48 to 25.8)	Gamma	37
Health utilities
Disease status utility per year
PFS	0.741 (0.593 to 0.889)	Beta	29,38
PD	0.581 (0.465 to 0.697)	Beta	29,38
Death	0	NA	NA
Disutility due to grade ≥ 3 AEsd
Tislelizumab plus chemotherapy	0.19 (0.152 to 0.228)	Beta	39–41
Chemotherapy	0.207 (0.165 to 0.248)	Beta	39–41
Other parameters
Body surface area, m2	1.80 (1.44 to 2.16)	Normal	35,36

AE, adverse event, NA not available, OS overall survival, PD progressed disease, PD-L1 programmed death-ligand 1, PFS progression-free survival.

^aOnly expected values are presented for these survival model parameters.

^bOverall total cost per patient regardless of treatment duration.

^cCalculated as the average cost of toxic effects using weighted frequencies of grade ≥ 3 treatment related adverse events for each treatment arm in the RATIONALE-306 trial. Costs of individual toxic effects were derived from the literature and include all care required to manage each toxic effect. References for individual toxic effect costs are summarized in Supplementary Table 2.

^dCalculated as the average disutility of toxic effects using weighted frequencies of grade ≥ 3 treatment-related adverse events for each treatment arm in the RATIONALE-306 trial. Disutilities of individual toxic effects were derived from the literature. References for individual toxic effect disutilities are summarized in Supplementary Table 2.

Cost and utility

Only direct medical costs, such as the costs associated with routine follow-up³⁰, laboratory tests and radiological examinations³⁰, best supportive care per cycle³⁰, AE management costs^31–33, and terminal care³⁴ were considered. A local database of the Guangxi Academy of Medical Sciences and the People’s Hospital of Guangxi Zhuang Autonomous Region was used to obtain drug prices. In view of the fact that cisplatin and paclitaxel were available in a variety of dosage forms on the Chinese market, we selected the most reasonable drug specifications that would satisfy both the requirement of improved clinical efficacy and the lowest cost. In our study, we identified severe AEs (≥ grade 3) with rates over 5%. These include decreased neutrophil count, anemia, decreased white blood cell count, and decreased platelet count⁶. The cost of AEs was obtained from published articles^31–33. A mean body surface area of 1.86 m² was used to calculate the drug dosages^35,36. The cost parameters are listed in Table 1.

The utilities of the PFS and PD states were 0.741 and 0.581, respectively, in Chinese patients with OSCC based on two health state utilities studies^29,38. The disutility values resulting from AEs were extracted from other studies and included in this analysis^39–41 (Supplementary Table 2). All the AEs were assumed to occur during the first cycle. PFS utility was reduced by duration-adjusted disutility, Table 1 shows the utility-related parameters.

Base-case analysis

In this study, ICER was expressed as the incremental cost per additional QALY gained. The ICERs of tislelizumab plus chemotherapy below the WTP threshold (USD 38,258/QALY) were assumed to be cost-effective. Furthermore, we calculated INHB and INMB using the following formulae:

$$\boxed{\begin{array}{c}INHB\left(\mathrm{\lambda }\right)=\left(\mathrm{\mu },E_{\text{Tislelizumab plus chemotherapy}},-,\mathrm{\mu },E_{\mathit{Chemotherapy}}\right)-\\ \frac{\left(\mathrm{\mu },C_{\text{Tislelizumab plus chemotherapy}},-,\mathrm{\mu },C_{\mathit{Chemotherapy}}\right)}{\mathrm{\lambda }}=\mathrm{\Delta }E-\frac{\mathrm{\Delta }C}{\lambda }\end{array}}$$

and

$$\boxed{\begin{array}{c}INMB\left(\mathrm{\lambda }\right)=\left(\mathrm{\mu },E_{\text{Tislelizumab plus chemotherapy}},-,\mathrm{\mu },E_{\mathit{Chemotherapy}}\right)\times \lambda -\\ \left(\mathrm{\mu },C_{\text{Tislelizumab plus chemotherapy}},-,\mathrm{\mu },C_{\mathit{Chemotherapy}}\right)=\mathrm{\Delta }E\times \lambda -\mathrm{\Delta }C\end{array}},$$

where μC and μE were the cost and utility of tislelizumab plus chemotherapy and chemotherapy alone, respectively, and λ was the WTP threshold^42,43. We also examined the outputs of the model for only the PFS state.

Price simulation

We also considered the patient assistance program (PAP) as the base-case analysis. To conduct the price simulation, we made the price fluctuate between USD 0.01 and USD 5 per mg without considering PAP, since tislelizumab for OSCC has not been included in the Chinese National Reimbursement Drug List (NRDL), and it was not listed as a first-line treatment for OSCC.

Sensitivity analysis

The model was subjected to a sensitivity analysis to determine its robustness. To conduct the deterministic sensitivity analysis (DSA), all parameters were used the 95% confidence intervals or assumed to be within a reasonable range of the baseline value (± 20%). We also performed probabilistic sensitivity analysis (PSA) using a Monte Carlo simulation conducted for 10,000 iterations. Cost inputs followed a gamma distribution, while probability, proportion, and utility inputs followed a beta distribution^44,45. Cost-effectiveness acceptability curves were constructed to evaluate the cost-effectiveness of the regimens under varying WTP thresholds.

Subgroup analysis

RATIONALE-306 provided subgroup-specific hazard ratios (HRs) for OS, which were used to calculate the ICER for each subgroup. Various subgroup factors were considered, including age, sex, smoking status, Eastern Cooperative Oncology Group (ECOG) performance status score, and PD-L1 tumor proportion score. Due to a lack of sufficient data, all subgroups were assumed to have the same HRs, except for OS, which was based on proportional hazards.

Results

Base-case analysis results

The results of the base-case analysis are presented in Table 2. The cost of tislelizumab plus chemotherapy was higher than that of chemotherapy alone (USD 36,368 vs. USD 24,808). Tislelizumab plus chemotherapy delivered an extra 0.751 life-years (2.377 vs. 1.626 life-years) and 0.414 QALYs (1.364 vs. 0.950 QALYs). Compared to chemotherapy alone, tislelizumab plus chemotherapy yielded an ICER of USD 27,886/QALY, which was less than three times GDP per capita. The INHB and INMB of tislelizumab combined with chemotherapy were 0.112 QALYs and USD 4279, respectively, compared with chemotherapy alone. Additionally, we conducted further studies that focused only on PFS. Tislelizumab plus chemotherapy yielded an ICER of USD 42,019/QALY compared with chemotherapy alone.

Table 2.

Summary of cost and outcome results in the base-case analysis.

Factor	Tislelizumab plus chemotherapy	Chemotherapy	Incremental change
Cost, USD
Druga	22,426	13,346	9079
Nondrugb	13,942	11,462	2480
Overall	36,368	24,808	11,560
Life-years
Progression-free survival	1.062	0.632	0.430
Overall	2.377	1.626	0.751
QALYs
Progression-free survival	0.717	0.442	0.275
Overall	1.364	0.950	0.414
Scenario 1: overall survival
ICERs, USD
Per life-year	NA	NA	15,383
Per QALY	NA	NA	27,886
INHB, QALY, at threshold 38,258a	NA	NA	0.112
INMB, USD, at threshold 38,258a	NA	NA	4279
Scenario 2: only progression-free survival
ICERs, USD
Per life-year	NA	NA	26,861
Per QALY	NA	NA	42,019
INHB, QALY, at threshold 38,258a	NA	NA	− 0.027
INMB, USD, at threshold 38,258a	NA	NA	− 1039

ICER incremental cost-effectiveness ratio, INHB incremental net health benefit, INMB incremental net monetary benefit, NA not applicable, PDL-1 programmed death-ligand 1 QALYs quality-adjusted life-years.

^aCompared with chemotherapy.

^bNondrug cost includes the costs of adverse event management, best supportive care per patient, routine follow-up, laboratory tests and radiological examinations, terminal care, and drug administration.

Price simulation

The results of the price simulation are presented in Fig. 1. The results revealed that tislelizumab plus chemotherapy could be cost-effective if the price of tislelizumab was less than USD 3.714 per mg at the WTP threshold of USD 38,258 and no PAP was considered.

Figure 1.

Results of tislelizumab price simulation. ICER incremental cost-effectiveness ratio, QALY quality-adjusted life-years.

Sensitivity analysis

The DSA results are shown in Fig. 2. The utility for PFS had the greatest influence on ICER, followed by the cost of tislelizumab, the cost of second-line treatment in the tislelizumab plus chemotherapy arm, and the cost of management of AEs in the chemotherapy arm. There were fluctuations in all variables within the upper and lower limit, indicating that the results of the base-case analysis were relatively robust to the model variables.

Figure 2.

Tornado diagram of deterministic sensitivity analysis of tislelizumab plus chemotherapy versus chemotherapy. AE adverse event, PD progressed disease, PFS progression-free survival, QALY quality-adjusted life-years.

The cost-effectiveness acceptability curve (Fig. 3) revealed that when the WTP was set to USD 38,258/QALY, the combination of tislelizumab plus chemotherapy had a 94.43% chance of being considered cost-effective compared to chemotherapy alone.

Figure 3.

Acceptability curves of cost-effectiveness for tislelizumab plus chemotherapy versus chemotherapy. QALY quality-adjusted life-years.

Subgroup analysis

Subgroup analysis was performed using varying HRs for OS based on a threshold WTP of USD 38,258/QALY. Table 3 summarizes the results of the subgroup analysis. Based on the subgroup analysis, all subgroups had positive INHB, and the cost-effectiveness probability of tislelizumab plus chemotherapy was greater than 90%. Tislelizumab plus chemotherapy performed better when the risk of death was lower in the subgroups, and ICERs were influenced by the HRs.

Table 3.

Summary of subgroup analyses obtained by varying the hazard ratios for overall survival.

Subgroup	Unstratified hazard ratio (95% CI)	Change in cost, USDa	Change in QALYsa	ICER, USD/QALY	Cost-effectiveness probability of tislelizumab plus chemotherapy, %b	INHBb
Age
< 65 years	0.73 (0.56 to 0.95)	9,697	0.376	25,774	96.40	0.123
≥ 65 years	0.62 (0.47 to 0.82)	11,951	0.505	23,660	99.29	0.193
Sex
Male	0.72 (0.59 to 0.88)	9,890	0.388	25,488	97.14	0.130
Female	0.46 (0.24 to 0.85)	15,850	0.686	23,089	99.74	0.272
Smoking status
Former or current	0.65 (0.52 to 0.81)	11,307	0.470	24,046	99.03	0.175
Never	0.77 (0.50 to 1.19)	8,945	0.329	27,196	93.30	0.095
Investigator-chosen chemotherapy
Platinum with fluoropyrimidine	0.66 (0.49 to 0.88)	11,139	0.459	24,293	98.89	0.167
Platinum with paclitaxel	0.69 (0.54 to 0.89)	10,418	0.423	24,607	98.41	0.151
ECOG performance status
0	0.72 (0.51 to 1.04)	9,890	0.388	25,488	97.11	0.130
1	0.66 (0.53 to 0.83)	11,097	0.459	24,201	99.00	0.168
Disease status at study entry
Metastatic	0.72 (0.59 to 0.88)	9,890	0.388	25,488	97.45	0.130
Locally advanced	0.44 (0.25 to 0.78)	16,402	0.708	23,151	99.70	0.280
Previous definitive therapy
Yes	0.67 (0.49 to 0.90)	10,890	0.447	24,372	98.82	0.162
No	0.68 (0.53 to 0.87)	10,685	0.435	24,558	98.50	0.156
PD-L1 TAP score
≥ 10	0.62 (0.44 to 0.87)	11,951	0.505	23,660	99.26	0.193
< 10	0.77 (0.59 to 1.01)	8,945	0.329	27,196	93.11	0.095

ECOG Eastern Cooperative Oncology Group, ICER incremental cost-effectiveness ratio, INHB incremental net health benefit, PD-L1 programmed death-ligand 1, QALY quality-adjusted life-year, TAP tumour area positivity.

^aChange in cost and change in QALYs represent the results of tislelizumab plus chemotherapy minus chemotherapy.

^bWillingness-to-pay threshold was set as USD 38,258/QALY.

Discussion

The results of this study revealed that the substantial effects of treatments on OSCC costs, underscoring the value-based oncology concern amid escalating healthcare costs. Despite the lack of prior evidence from, RATIONALE-306 is the first study to examine the potential impact of immunotherapy on survival in a Chinese population with OSCC⁶. Tislelizumab plus chemotherapy has been associated with significant OS and PFS benefits in patients with OSCC, attracting considerable attention. This drug could significantly impact the global immunotherapy market for OSCC treatment. Given the lack of price information, a cost-effectiveness analysis is necessary to aid in determining whether tislelizumab is cost-effective. Thus, this study assessed the cost-effectiveness of tislelizumab plus chemotherapy, providing clinicians with more information for its clinical use.

Based on our base-case analysis, tislelizumab plus chemotherapy was cost-effective at the current price. The WTP threshold was set at USD 38,258/QALY, with an incremental survival of 0.414 QALYs and a cost of USD 11,560, resulting in an ICER of 27,886. Price simulation results indicate that tislelizumab plus chemotherapy is cost-effective if the price of tislelizumab is less than USD 3.714 per mg without considering PAP in China at a WTP threshold of USD 38,258. The DSA and PSA results demonstrated the robustness of this finding. Subgroup analysis indicated that, generally, subgroups with a low risk of mortality were more likely to be cost-effective. Furthermore, the economic outcomes of treatment could improve based on patient characteristics.

The DSA results indicated that the most influential parameters were the cost of tislelizumab, the cost of second-line therapy in the tislelizumab combined with chemotherapy arm, and the cost of managing AEs in the chemotherapy arm. Given the wide range of these variables, the WTP threshold is not cost-effective for combination therapy. The National Healthcare Security Administration (NHSA) in China has recently made concerted efforts to negotiate drug prices with pharmaceutical manufacturers, resulting in a 30–70% reduction in prices for many anticancer drugs⁴⁶. These results provide evidence to assist Chinese policymakers in deciding whether to consider tislelizumab as a first-line treatment for OSCC in the NRDL. Future budget impact analyses are required to determine whether the new treatment is affordable for the patient given that reimbursement policies were not considered in this study. We opted for subsequent treatment strategies derived from the NCCN¹³ as well as the Chinese Society of Clinical Oncology guidelines (version 1.2023, Esophageal Cancer). The cost of subsequent chemotherapy and management of AEs in the chemotherapy arm had a large influence on the total cost of disease progression.

We reviewed of the published economic evaluations of first-line ICIs for the treatment of patients with OSCC in China. In comparison to chemotherapy, the ICERs for camrelizumab plus chemotherapy, toripalimab plus chemotherapy, sintilimab plus chemotherapy, and pembrolizumab plus chemotherapy were USD 31,062/QALY, USD 30,443.629/QALYs, USD 26,773.68/QALY, and USD 41,805.12/QALY, respectively^38,47–49. Our study differs from these studies in several ways: First, some of the key model parameters used in these studies, such as utilities and terminal care cost, were derived from published studies focused on OSCC patients, which could have caused misinterpretation. Second, the studies used different models. The area under the relevant survival curve is calculated to evaluate health state occupancy in a partitioned survival model. Partitioned survival models are different from state transition models such as Markov models, as they do not establish a causal link between intermediate clinical endpoints (e.g., disease progression) and survival. Partitioned survival models directly consider clinical trial outcomes and can be constructed without considering individual patient information. Partitioned survival model and state transition models may produce substantially different survival extrapolations, and extrapolations from the state transition models are influenced by their specification⁵⁰.

This is the first study to evaluate the newest evidence using an economic modeling approach to perform a cost-effectiveness analysis of tislelizumab plus chemotherapy as a first-line therapy for OSCC. However, OSCC remains a major clinical problem in China. Thus, the use of tislelizumab may provide clinical benefits for patients with OSCC. Based on the findings of this study, the process of including tislelizumab in the NRDL and its promotion may be accelerated. Second, the results of this study are robust based on DSA and PSA. The fit and extrapolation of survival data were based on flexible parametric models, which are more accurate than the specified standard survival models. Information on the economic status of the subgroups may facilitate the decision-making process for physicians, patients, and policymakers regarding treatment.

This study had several limitations. First, we did not include other Chinese ICIs, such as camrelizumab, toripalimab, and sintilimab, which have demonstrated superior clinical efficacy, because of a lack of head-to-head results. Second, it remains unclear whether tislelizumab plus chemotherapy is beneficial in the long term for patients with OSCC. In this model, long-term efficacy, subsequent treatment, and long-term follow-up were estimated based on available information regarding the 3-year follow-up. Although the model and parameters have been validated, there may have been some uncertainties. Assumptions of goodness-of-fit were made by matching the reconstructed IPD to the K–M OS and PFS curves for the RATIONALE 306 trial. This could result in uncertainty in the model results and may lead to an underestimation of the survival benefit of tislelizumab plus chemotherapy. Despite this, we are confident that the goodness-of-fit parameters selected in our study are the most suitable models to fit the K–M OS and PFS curves. Third, the utilities in the model were not derived from RATIONALE-306 but from another study in patients with OSCC. Further bias may have been introduced in the cost-effectiveness analysis by assuming the same utility level for the patients in both groups. Fourth, we did not consider grade 1 or 2 AEs, which may have led to an overestimation of the results related to tislelizumab plus chemotherapy. These AEs cannot be ignored in general clinical practice.

Conclusion

In summary, tislelizumab plus chemotherapy is likely a cost-effective treatment option for OSCC from the perspective of the Chinese healthcare system if the WTP threshold is set at USD 38,184. These results may assist clinicians in making appropriate treatment decisions for patients with OSCC. However, further follow-up and real-world data are required owing to the limitations of this study.

Author contributions

Li Zhang and Henghai Su: conceptualization, methodology, data curation and writing original draft preparation. Yan Li: conceptualization, methodology, data curation, funding, reviewing and editing. Xueyan Liang: visualization and editing. Xiaoyu Chen: supervision and validation. Final approval of the manuscript was obtained from all authors.

Funding

This work was supported by the Drug Safety Research Project of Guangxi Zhuang Autonomous Region (Guiyaojiankezhishu[2023]017).

Data availability

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-68399-3.