International cost-effectiveness analysis of chemoradiotherapy plus pembrolizumab for locally advanced cervical cancer

Abstract

Objective

Pembrolizumab administration can improve concurrent chemoradiotherapy (CCRT) efficacy in newly diagnosed, high-risk, locally advanced cervical cancer (LACC). Given the importance of balancing the costs of innovative therapeutics against their efficacy, this study was developed to assess the cost-effectiveness from the perspective of payers in Americas, Europe, and Asia.

Methods

The main survival and other relevant parameters of 1,060 LACC patients from the KEYNOTE-A18 trial were collected to establish a lifetime 3-state Markov model to evaluate the cost and effectiveness of pembrolizumab-CCRT and placebo-CCRT. Primary outcome measures included total cost, life-years (LYs), quality-adjusted LYs (QALYs), incremental cost-effectiveness ratio (ICER), incremental net monetary benefit, and incremental net health benefits (INHBs) at countries’ traditional willingness-to-pay (WTP) thresholds. Model stability was also examined through sensitivity analyses.

Results

The USA, Italy, and China are selected as representative countries for each of the 3 continents, assuming that their WTP thresholds were $150,000, $43,749, and $37,766 per QALY. The increased efficacy and costs of pembrolizumab-CCRT versus placebo-CCRT were 2.52 QALYs (3.11 LYs) and $346,479, 2.30 QALYs (2.81 LYs) and $236,776, 1.79 QALYs (2.12 LYs) and $29,027, calculating the ICER for the 3 countries as $137,500/QALY ($111,499/LY), $102,758/QALY ($84,192/LY), and $16,217/QALY ($13,726/LY), respectively. The respective INHBs were 0.21 QALY, −3.11 QALY, and 1.02 QALY, and pembrolizumab-CCRT was exhibited cost-effectiveness opportunities of 62.68%, 12.53%, and 75.23% at the selected WTP threshold, respectively.

Conclusion

At current prices, pembrolizumab-CCRT represents a cost-effective alternative for patients with LACC in the USA and China, but not in Italy.

Synopsis

Pembrolizumab-concurrent chemoradiotherapy (CCRT) significantly improved survival in locally advanced cervical cancer (LACC). The cost-effectiveness of pembrolizumab-CCRT and CCRT were evaluated based on phase 3 KEYNOTE-A18 trial. Pembrolizumab-CCRT is cost-effectiveness for LACC patients in the USA and China, but not in Italy.

INTRODUCTION

Cervical cancer (CC) is the most common form of human papillomavirus-associated malignancy among females and a leading driver of cancer-related death, with over 560,000 diagnoses and 310,000 deaths annually throughout the globe [1,2]. An estimated 90% of these cases arise in patients in middle-/low-income nations, and approximately two-thirds of patients exhibit locally advanced CC (LACC; International Federation of Gynecology and Obstetrics [FIGO] stage IB2–IVA) when first diagnosed and face high disease recurrence rates [2,3]. The respective 5-year overall survival (OS) rates for patients with phase IB2–IIB and III–IVA disease are reportedly 86% and 60% [4].

Concurrent chemoradiotherapy (CCRT) followed by brachytherapy currently remains the standard of care for LACC patients [5]. However, multiple meta-analyses and randomized controlled trials (RCTs) have determined that CCRT versus radiotherapy-alone increases 5-year OS and progression-free survival (PFS) in CC patients by 6%–10% and 13%–16%, respectively [6,7]. Even with these limited benefits, roughly 50% of high-risk CC patients experience disease recurrence within 2 years [7]. There thus remains a pressing need for the advent of new treatment models and therapies that can improve the prognostic outcomes for women diagnosed with high-risk LACC.

Pembrolizumab is a potent humanized IgG4 mAb that specifically targets programmed cell death-1 (PD-1) and interferes with its ability to interact with its cognate ligands programmed cell death ligand 1 (PD-L1) and PD-L2. In both preclinical and clinical studies, cisplatin-chemotherapy and radiotherapy have exhibited promising synergy with immune checkpoint inhibitors when deployed for cancer patient treatment [8]. To explore the benefits of these combination regimens in gynecological cancer patients, the phase III KEYNOTE-A18 (ENGOT-cx11/GOG-3047, NCT04221945) trial evaluated the benefits of the pembrolizumab-CCRT regimen, revealing that it was associated with significant improvements in newly diagnosed high-risk LACC patient PFS (hazard ratio [HR]=0.70; 95% confidence interval [CI]=0.55–0.89; p=0.0020) with a trend toward the prolongation of OS (HR=0.73; 95% CI=0.49–1.07) [9]. These benefits suggest that this novel combination regimen has been approved by the National Comprehensive Cancer Network (NCCN) Clinical and US Food and Drug Administration guidelines for entry into general clinical practice, providing hope for better outcomes for patients with this devastating disease.

The development of innovative, expensive, and more efficacious treatments has continued to increase the economic burden associated with the management of LACC as in the case of the pembrolizumab-CCRT regimen. Health economic analyses are thus increasingly being employed as a means of guiding treatment-related decision-making. These pharmacoeconomic studies provide an approach to assessing, quantifying, and evaluating the costs and outcomes associated with particular healthcare interventions [10]. Given the clear importance of clarifying the economic viability of novel treatment strategies for both patients and policymakers, the present descriptive and analytical study was developed to assess the cost-effectiveness of pembrolizumab-CCRT when administered to newly diagnosed, high-risk LACC patients.

MATERIALS AND METHODS

All patient, treatment, survival, and safety data in this study were derived from the KEYNOTE-A18 (ENGOT-cx11/GOG-3047) trial, which was conducted from June 9, 2020 through December 15, 2022 [9]. The International Society for Pharmacoeconomics and Outcomes Research guidelines and the CHEERS 2022 reporting standards were observed when conducting this study (Table S1) [11].

1. Model structure

KEYNOTE-A18 trial enrolled patients primarily from the USA, Italy, and China, which happened to represent the Americas, Europe, and Asia, respectively. A decision tree-based Markov model was developed to simulate the disease course and treatment of patients with LACC (Fig. S1). In this model, which had a lifetime time horizon, all patients began in the PFS state before having a chance to transition to the progressive disease (PD) or death state with failure to respond to treatment. Each model cycle was 12 weeks in length, in line with the general clinical practice guidelines for RCT [9]. GetData Graphics Digitizer (v 2.26), R Studio (v 4.2.2), and Matlab (v R2020a) were used to conduct all analyses.

2. Model overview

Long-term clinical and cost outcome predictions were compared between LACC patients in the pembrolizumab-CCRT and placebo-CCRT groups from the perspective of payers in the USA, Italy, and China. Primary outcomes included total costs, life-years (LYs), quality-adjusted LYs (QALYs), incremental cost-effectiveness ratio (ICER), incremental net monetary benefit (INMB), and incremental net health benefit (INHB) were conducted as the primary outcomes [12,13]. The willingness-to-pay (WTP) thresholds of $150,000/QALY, $43,749/QALY, and $37,766/QALY were selected for this study, as this is a frequently chosen value for pharmacoeconomic analyses focused on the USA, Italy, and China, respectively [14,15,16]. All costs and health-related outcomes were discounted at an annual 3%, 3.5%, and 5% rate, respectively [14,16,17].

3. Model transitions

Transition probabilities were calculated with Kaplan-Meier curves, and short-term OS and PFS data were extracted for analysis [9]. These survival curves were fitted to the Weibull, Log-logistic, Log-normal, Exponential, and Gompertz distributions, leading to the selection of the most optimal fit based on visual inspection, clinical rationale, and the Akaike and Bayesian information criterion values (Fig. S2, Table S2) [14,18]. As the Weibull distribution exhibited the best fit, it was selected for use [14,18]. The following formula was utilized to calculate transition probabilities over time:

(1-exp{λ(t-u)^γ-λt^γ})

where u represents the Markov period, t denotes the current model period, and λ and γ correspond to the scale and shape parameters, respectively.

4. Model inputs

Patients and treatments

In total, the model in this study enrolled 1,060 patients with newly diagnosed, high-risk (FIGO 2014 stage IB2–IVA) LACC who had not undergone prior antitumor treatment. These patients were assigned to undergo treatment with pembrolizumab (200 mg every 3 weeks for 5 cycles, followed by 400 mg every 6 weeks for 15 cycles) plus CCRT (cisplatin, 40 mg/m² every 3 weeks for 5–6 cycles; external beam radiotherapy [EBRT] plus brachytherapy, ≥70 Gy [equivalent dose in 2 Gy/f, EQD2] not to exceed 56 days) or placebo-CCRT (Table S3) [9]. Drug dose calculations were performed based on the assumption that all patients were 50-year-old women with body weights (body surface areas) of 70 kg (1.86 m²), 70 kg (1.80 m²), and 68 kg (1.69 m²) in the USA, Italy, and China, respectively [14,16,19]. After completing these treatments, 6.3% (pembrolizumab-CCRT) and 9.2% (placebo-CCRT) of patients exhibited PD and underwent second-line chemotherapy with or without bevacizumab according to the NCCN guidelines and relevant RCTs [5,20]. Best supportive care (BSC) was provided to the remaining patients, and end-of-life was administered for one year before death in cases of treatment-related mortality [5].

Utility and cost

Specific utility weights between 1 (perfect health) and 0 (death) were assigned to the health states in this Markov model and utilized to discount LYs to calculate QALYs. According to the data provided by the KEYNOTE-A18 trial, using 2 methods, direct (standard gambling technique) and indirect measurement (QLQ-C30 scores were mapped to EQ-5D utility scores), Longworth et al.’s algorithm was used to estimate mean baseline utility (PFS) of 0.812 and 0.780 for the pembrolizumab-CCRT and placebo-CCRT [9,21,²²], combining for mean disutility of severe adverse events (AEs, grade 3 or higher and incidence of more than 5%) [14,23,24,25,26,27]. Since QLQ-C30 does not focus on progress scores, the PD utility was 0.65 from previously published studies (Table 1) [28].

Table 1. Clinical and health parameters.

Variables			Baseline value (range)	Distribution
Clinical parameters
	Weibull survival model for OS
		Chemoradiotherapy	Scale=0.0007690	NA
			Shape=1.7633274
		Pembrolizumab-chemoradiotherapy	Scale=0.0009504	NA
			Shape=1.5603459
	Weibull survival model for PFS
		Chemoradiotherapy	Scale=0.0234850	NA
			Shape=1.0090160
		Pembrolizumab-chemoradiotherapy	Scale=0.0211820	NA
			Shape=0.9088340
	Rate of receiving second-line therapy
		Chemoradiotherapy	0.092 (0.074–0.110) [9]	Beta
		Pembrolizumab-chemoradiotherapy	0.063 (0.050–0.076) [9]	Beta
	Risk for main AEs in placebo-chemoradiotherapy
		Anemia	0.158 (0.126–0.190) [9]	Beta
		Leukopenia	0.127 (0.102–0.152) [9]	Beta
		Neutropenia	0.096 (0.077–0.115) [9]	Beta
		Neutrophil count decreased	0.147 (0.118–0.176) [9]	Beta
		WBC count decreased	0.209 (0.167–0.251) [9]	Beta
	Risk for main AEs in pembrolizumab-chemoradiotherapy
		Anemia	0.188 (0.150–0.226) [9]	Beta
		Leukopenia	0.108 (0.086–0.194) [9]	Beta
		Neutropenia	0.106 (0.085–0.127) [9]	Beta
		Neutrophil count decreased	0.146 (0.117–0.175) [9]	Beta
		WBC count decreased	0.193 (0.154–0.232) [9]	Beta
Health parameters
	Utility and disutility
		Utility of PFS in pembrolizumab-chemoradiotherapy	0.812 (0.650–0.974) [9,21,22]	Beta
		Utility of PFS in placebo-chemoradiotherapy	0.780 (0.642–0.936) [9,21,22]	Beta
		Utility of PD	0.650 (0.520–0.780) [28]	Beta
		Disutility of grade 3 or higher AEs	0.087 (0.070–0.104) [14,23–27]	Beta
	Discount rate (USA)		0.030 (0–0.036) [14]	Uniform
	Discount rate (Italy)		0.035 (0–0.042) [16]	Uniform
	Discount rate (China)		0.050 (0–0.060) [17]	Uniform

AE, adverse event; NA, not available; OS, overall survival; PD, progressive disease; PFS, progression-free survival; WBC, white blood cell.

The primary considerations for the developed model were direct medical costs, such as those associated with treatment, drug administration, laboratory tests, tumor imaging, AEs management, PD-L1 testing, BSC, and end-of-life (Table 2) [14,23,24,25,29,30,31,32,33,34]. These costs were based upon data from the related professional website [29,30,31] and prior publications [14,16,23,24,25,32,33,34,35], and were adjusted for inflation to 2024 dollars using the US Consumer Price Index [36].

Table 2. Cost estimates.

Cost parameters ($)		Baseline value (range)			Distribution
		USA	Italy	China
Key treatment per cycle
	Pembrolizumab (concurrent)	55,730 (44,584–66,876) [29]	37,493 (29,994–44,992) [31]	3,154 (2,523–3,785)*	Gamma
	Pembrolizumab (maintenance)	44,584 (35,667–53,501) [29]	29,994 (23,995–35,993) [31]	2,523 (2,018–3,028)*	Gamma
	Cisplatin (concurrent)	180 (144–216) [29]	142 (114–170) [31]	46 (37–55)*	Gamma
	IMRT/VMAT	33,517 (26,814–40,220) [30]	73,796 (59,037–88,555) [32]	9,375 (7,500–11,250)*	Gamma
	Non-IMRT/-VMAT	3,335 (2,668–4,002) [30]	2,282 (1,826–2,738) [33]	6,375 (5,100–7,650)*	Gamma
	Brachytherapy	2,152 (1,722–2,582) [30]	3,768 (3,014–4,522) [32]	2,958 (2,366–3,550)*	Gamma
	Second-line therapy	15,847 (12,678–19,016) [29]	5,253 (4,202–6,304) [32]	2,758 (2,206.4–3,309.6)*	Gamma
AEs
	Placebo-chemoradiotherapy	2,661 (2,129–3,193) [23,24,25]	989 (791–1,187) [34]	220 (176–264) [16,35]	Gamma
	Pembrolizumab-chemoradiotherapy	2,695 (2,156–3,234) [23,24,25]	1,097 (878–1,316) [34]	232 (186–278) [16,35]	Gamma
Drug administration per time		85 (68–102) [30]	34 (27–41) [31]	6 (5–7)*	Gamma
Laboratory tests per time		86 (69–103) [30]	4 (3–5) [31]	25 (20–30) [16]	Gamma
Tumor imaging per cycle		107 (86–128) [30]	350 (280–420) [31]	1,429 (1,143–1,715) [16]	Gamma
Best supportive care per cycle		4,661 (3,729–5,593) [14]	6,004 (4,803–7,205) [32]	243 (194–292) [16]	Gamma
End-of-life per patient		10,889 (8,711–13,067) [14]	17,373 (13,898–20,848) [32]	1,490 (1,192–1,788) [16]	Gamma

AE, adverse event; IMRT/VMAT, intensity modulated radiation therapy/volumetric modulated arc therapy.

^*The costs were estimated based on the price of Xiangya Hospital of Central South University, 2024.

5. Sensitivity and subgroup analyses

Sensitivity analyses were performed to clarify the impact of different variables on model outputs to gauge the robustness of these analyses. In one-way sensitivity analyses, over 30 parameters were individually varied within ±20% of baseline values, and changes in ICERs were examined [14]. In 2-way sensitivity analyses, the effects of simultaneously changing 2 parameters that strongly impacted ICER values were assessed in greater detail [37]. Probabilistic sensitivity analyses were additionally conducted by randomly sampling all parameters from parameter-appropriate distributions based on the results of 10,000 Monte Carlo simulations [14,18]. Tornado diagrams, scatter plots, and acceptability curves were used to present the results of these sensitivity analyses.

6. Ethics approval and consent to participate

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors, it does not require the approval of the independent ethics committee.

RESULTS

1. Base-case analysis

The developed lifetime Markov model revealed that pembrolizumab-CCRT treatment was associated with a life expectancy that was 3.11 LYs (37.92 months), 2.81 LYs (33.72 months), and 2.12 LYs (25.44 months) longer than that for the placebo-CCRT for LACC patients in the USA, Italy, and China, respectively. Following adjustment, net yields for pembrolizumab-CCRT were 2.52 QALYs, 2.30 QALYs, and 1.79 QALYs, resulting in ICERs of $137,500/QALY, $102,758/QALY, and $16,217/QALY, respectively. In addition, INHB (INMB) from pembrolizumab-CCRT was 0.21 QALY ($31,521), −0.31 QALY (−$136,153), and 1.02 QALY ($38,574), relative to placebo-CCRT in the USA, Italy, and China, respectively (Table 3). The data thus suggested that pembrolizumab-CCRT may represent an economically viable therapeutic option for LACC patients in the USA and China.

Table 3. Baseline results.

Treatment		Total cost ($)	Overall LYs	Overall QALYs	ICER ($)		INMB ($)	INHB (QALY)
					Per LY	Per QALY
USA
	Placebo-chemoradiotherapy	216,776	9.99	7.12	Reference	Reference	Reference	Reference
	Pembrolizumab-chemoradiotherapy	563,256	13.10	9.64	111,499	137,500	31,521	0.21
Italy
	Placebo-chemoradiotherapy	207,944	9.64	6.88	Reference	Reference	Reference	Reference
	Pembrolizumab-chemoradiotherapy	444,720	12.45	9.19	84,192	102,758	−136,153	−3.11
China
	Placebo-chemoradiotherapy	57,163	8.71	6.24	Reference	Reference	Reference	Reference
	Pembrolizumab-chemoradiotherapy	86,190	10.82	8.03	13,726	16,217	38,574	1.02

ICER, incremental cost-effectiveness ratio; INHB, incremental net health benefit; INMB, incremental net monetary benefit; LY, life-year; QALY, quality-adjusted life-year.

2. Sensitivity analyses

One-way sensitivity analyses showed that the most sensitive parameters are the utility of PFS in pembrolizumab-CCRT, the utility of PFS in placebo-CCRT, and the cost of maintenance pembrolizumab, which ICER values range from $9,836/QALY to $289,352/QALY, $11,486/QALY to $215,264/QALY, and $14,708/QALY to $156,758/QALY, respectively (Fig. 1A-C). Other variables such as the costs of EBRT, brachytherapy, and concurrent chemotherapy had relatively little impact on ICER values (Fig. 1A-C).

Fig. 1. The 1-way and 2-way sensitivity analyses for the pembrolizumab-chemoradiotherapy compared to the placebo-chemoradiotherapy in the USA (A, D), Italy (B, E), and China (C, F).

AE, adverse event; CCRT, concurrent chemoradiotherapy; ICER, incremental cost-effectiveness ratio; NCD, neutrophil count decreased; PD, progressive disease; PFS, progression-free survival; QALY, quality-adjusted life-year; WBCCD, white blood cell count decreased; WTP, willingness-to-pay.

In 2-way sensitivity analyses of the parameters that most strongly impacted model outcomes (the utility of PFS for each regimen), the ICER remained below the USA and China thresholds for most analyzed utility combinations. Specifically, pembrolizumab-CCRT was predicted to be cost-effective when the utility of PFS for this regimen was greater than 0.893 (USA) or 0.789 (China) and the placebo-CCRT utility value was within the specified range (Fig. 1D and F). In addition, regardless of any combination of the 2 treatments for PFS, pembrolizumab-CCRT was not cost-effective in Italy (Fig. 1E).

In probabilistic sensitivity analyses, the odds of pembrolizumab-CCRT being cost-effective when compared with placebo-CCRT were 62.68%, 12.53%, and 75.23% at the WTP thresholds of $150,000/QALY, $43,749/QALY, and $37,766/QALY in USA, Italy, and China, respectively (Fig. 2, Fig. S3). Further, if 3 times of local per capita GDP are taken as the corresponding WTP threshold, i.e., $228,900/QALY, $104,400/QALY, and $37,766/QALY, the cost-effectiveness probabilities of pembrolizumab-CCRT are 89.89%, 59.83%, and 75.23%, respectively.

Fig. 2. The cost-effectiveness acceptability curves for the pembrolizumab-chemoradiotherapy compared to the placebo-chemoradiotherapy in the USA (A), Italy (B), and China (C).

QALY, quality-adjusted life-year.

DISCUSSION

The total direct medical costs associated with CC are estimated to exceed $441 million annually [38]. Radiotherapy, which is a key component of treatment for affected patients, costs more than surgery [38]. Owing to these high treatment-related costs and medical resource limitations, patients suffering from cancer invariably face high medical costs and associated economic burdens. There is thus a clear need for value-based oncology to ensure that the costs and benefits of treatments for affected patients are effectively balanced. Most pharmacoeconomic analyses of CC patients have focused on individuals with metastatic or recurrent disease based on the results of the KEYNOTE-826 (NCT03635567) trial [39]. Shi et al. [18] and Zheng et al. [27] respectively evaluated the cost-effectiveness of pembrolizumab plus chemotherapy in patients with metastatic or recurrent CC from the perspective of payers in the US (ICER, $64,338.19/QALY) and China (ICER, $247,663/QALY), ultimately determining that this regimen was not cost-effective. As this form of cancer is highly likely to recur after CCRT and there remains a pressing need for new treatment options aimed at significantly improving high-risk LACC patient outcomes, the KEYNOTE-A18 (ENGOT-cx11/GOG-3047) trial was conducted, leading to the recent announcement of encouraging survival outcome data. In light of these promising results, the present study was developed as the first examination of the cost-effectiveness of pembrolizumab-CCRT as a treatment option for newly diagnosed, high-risk LACC patients.

The lifetime Markov model developed herein revealed that the ICERs of the pembrolizumab-CCRT versus placebo-CCRT were $137,500/QALY ($111,499/LY), and $16,217/QALY ($13,726/LY) lower than those of corresponding WTP thresholds of $150,000/QALY and $37,766/QALY for payers in the USA and China, while above the WTP threshold in Italy (ICER, $102,758/QALY and $84,192/LY). The higher cost of the pembrolizumab-CCRT regimen was primarily attributable to the costs of follow-up and BSC, both of which inevitably accompany the prolongation of survival. The unidirectional sensitivity found that pembrolizumab-CCRT was predicted not to be cost-effective if the respective PFS utilities for the pembrolizumab-CCRT in the USA and China were less than 0.791 and 0.680, respectively. However, both groups of utilities did not change the current results in Italy. As such, bidirectional sensitivity analyses were conducted based on the assumption that pembrolizumab-CCRT would be cost-effective for most PFS utility combinations at the $150,000/QALY (USA) and $37,766/QALY (China) WTP thresholds. In the USA, the cost of pembrolizumab additionally contributed to the incremental costs for this regimen, and at pembrolizumab prices exceeding $63/mg, this combination regimen may no longer be cost-effective. As such, healthcare organizations can seek to reduce drug costs or provide Medicare coverage for innovative treatment options to achieve better patient outcomes without compromising the quality of monitoring or services.

Both PD-L1 and PD-L2 are frequently overexpressed in human cancers, with PD-L1 being the PD-1 ligand that is most commonly present in solid tumors. High PD-L1 expression has been reported in CC patients such that PD-1/PD-L1 inhibitors have emerged as promising therapeutic options with the potential to improve patient outcomes [40]. As such, caution is warranted when interpreting data pertaining to patients with a PD-L1 combined positive score <1% until definitive long-term survival data are available in the future. Additional long-term pharmacoeconomic analyses are thus warranted to better explore the utility of PD-L1 and other candidate biomarkers when seeking to identify patients who are most likely to benefit from this combination treatment regimen, thereby facilitating superior individualized treatment planning.

One of the key strengths of this study is the fact that it assessed the relative value of combined immunotherapy and chemoradiotherapy for LACC patients based on the most up-to-date evidence available. Pembrolizumab-CCRT has only recently emerged as a promising option for the treatment of these patients, and pharmacoeconomic analyses are vital to the recommendation of a given regimen under international guidelines. In addition, this study considered the relative value of pembrolizumab-CCRT versus placebo-CCRT from Americas, European, and Asian payers’ perspectives such that these results offer a valuable resource for clinicians, policymakers, and healthcare systems in the USA, Italy, and China, while also offering a robust foundation for multilateral national and international pricing negotiations. The disutility of AEs was also taken into consideration, utilizing QLQ-C30 global health status to estimate mean healthy utility values for CC and correcting these for AE-related disutility to ensure that these findings are more applicable to a clinical setting.

Despite these strengths, the study is subject to certain limitations. For one, the follow-up periods for patients in the phase III KEYNOTE-A18 trial were relatively short such that the Markov model developed herein may have underestimated treatment efficacy when comparing pembrolizumab-CCRT with placebo-CCRT. However, the extrapolation and relatively stable curves that were fitted to the Weibull distribution model helped ensure the improved reliability of these analyses of long-term patient survival. Second, the KEYNOTE-A18 trial did not report survival curves for all patient subgroups, so no cost-effectiveness exploration was conducted for subgroups, and future work is a key task. Lastly, costs associated with the management of AEs affecting >5% of patients were considered without immune-related AEs, but the reanalysis to include immune-related AEs did not show any difference in the economic assessment. In addition, the one-way sensitivity analysis also demonstrated that the costs associated with AEs did not impact the model results.

This is the first pharmacoeconomic analysis to date focused on the long-term survival outcomes and overall healthcare costs for LACC undergoing pembrolizumab-CCRT or placebo-CCRT treatment. Based on these results, pembrolizumab-CCRT is likely to be a cost-effective option for LACC patient care in the USA and China, the opposite in Italy. These results will provide a basis for future pricing negotiations, international guideline formulation, and clinical decision-making for newly diagnosed high-risk LACC patients.

SUPPLEMENTARY MATERIALS

Table S1

The CHEERS 2022 checklist

Table S2

Summary of statistical goodness-of-fit of Kaplan-Meier curve

Table S3

Details of treatment strategy

Fig. S1

Model structure.

Fig. S2

Kaplan-Meier curve fitting and extrapolation.

Fig. S3

Probability sensitivity analysis scatter plot. Each point in the diagram represents a simulation result of 10,000 Monte Carlo simulations of pembrolizumab-chemoradiotherapy versus placebo-chemoradiotherapy in the USA (A), Italy (B), and China (C). The ellipse represents the 95% confidence interval and the dotted line represents WTPs of $150,000/QALY, $43,749/QALY, and $37,766/QALY, respectively. The blue and red dots represent the probability that pembrolizumab-chemoradiotherapy and placebo-chemoradiotherapy are cost effective, respectively.