Immune checkpoint inhibitors-based therapies for first-line treatment of advanced ovarian cancer with non-BRCAm/HRD-negative: a cost-effectiveness analysis

Abstract

Background

Emerging evidence supports the clinical potential of immune checkpoint inhibitors (ICIs) with or without bevacizumab and poly-(adenosine diphosphate-ribose) polymerase (PARP) inhibitors for refractory advanced ovarian cancer (AOC) with BRCA non-mutated (non-BRCAm) and homologous recombination deficiency-negative (HRD-negative). However, broader adoption hinges on a thorough economic evaluation.

Methods

A lifetime Markov model was developed to assess the cost-effectiveness of six first-line treatment strategies for AOC in non-BRCAm/HRD-negative patients: chemotherapy alone (C); bevacizumab plus chemotherapy (BC); durvalumab plus chemotherapy and bevacizumab with or without olaparib (DBC or DBCO); and pembrolizumab plus chemotherapy with or without olaparib (PC or PCO). Model inputs were informed by patient demographics and clinical outcomes reported in the DUO-O and KEYLYNK-001 trials, as well as aggregated cost data relevant to the American healthcare system. Primary outcomes included incremental cost-effectiveness ratios (ICERs) and incremental net health benefits (INHBs) at the willingness-to-pay (WTP) threshold of $150,000/quality-adjusted life year (QALY).

Results

Compared to C, the ICERs (with INHBs) for BC, DBC, DBCO, PC, and PCO were $2,632,524/QALY (-0.73 QALY), $2,005,771/QALY (-1.73 QALY), $7,351,200/QALY (-2.88 QALY), $484,706/QALY (-1.46 QALY), and $1,382,130/QALY (-2.46 QALY), respectively. The comparison among ICI-based treatment regimens revealed that PC had high efficacy and low cost compared with DBCO and PCO, and the ICER of PC versus DBC was $69,253/QALY.

Conclusions

None of the ICI-based or BC proved cost-effective versus C for newly diagnosed AOC with non-BRCAm/HRD-negative. However, PC offers a cost-effective alternative when compared with other ICI-based strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13048-026-01988-0.

Introduction

Ovarian cancer (OC) ranks as the second leading cause of death among gynecologic malignancies, with an estimated 20,890 new diagnoses and 12,730 deaths anticipated in the United States in 2025 [1]. Due to the deep anatomical location of the ovaries and the nonspecific nature of early symptoms, approximately 70% of cases are diagnosed at an advanced or metastatic stage [2]. Despite initial responsiveness to surgery and platinum-based chemotherapy, over 70% of patients experience recurrence within two to three years, with the overall five-year survival rate remaining at a dismal 30% [2]. Precision medicine for OC now incorporates molecular stratification, particularly regarding BRCA mutation status and homologous recombination deficiency (HRD). While poly (ADP-ribose) polymerase (PARP) inhibitors and standard chemotherapy show efficacy in patients with BRCA mutations or HRD positivity, accounting for only 15–30% of cases [3, 4], more than 70% of OC patients are BRCA non-mutated and HRD-negative (non-BRCAm/HRD-negative), a subgroup characterized by aggressive tumor biology, chemoresistance, and poor outcomes [5, 6]. Thus, developing new therapies with broader efficacy and acceptable toxicity profiles is an urgent unmet need for this population.

The advent of immunotherapy has brought renewed optimism to the treatment landscape of advanced OC (AOC). Immune checkpoint inhibitors (ICIs), such as durvalumab (an anti-PD-L1 IgG1 monoclonal antibody) [7]. and pembrolizumab (an anti-PD-1 IgG4 monoclonal antibody), have emerged as promising therapeutic options. [8]. The phase III DUO-O/ENGOT-ov46/GOG-3025 (NCT03737643) trial demonstrated that the combination of durvalumab, chemotherapy, bevacizumab, and olaparib (DBCO) significantly improved progression-free survival (PFS) over bevacizumab plus chemotherapy (BC) in newly diagnosed non-BRCAm/HRD-negative AOC patients (median PFS: 21.1 vs. 17.5 months; hazard ratio [HR]: 0.68; 95% confidence interval [CI]: 0.54–0.89). The addition of durvalumab without olaparib (DBC) showed a favorable trend toward improved survival, albeit without statistical significance [9]. Similarly, the KEYLYNK-001/ENGOT-OV43/GOG-30 (NCT03740165) trial reported significant PFS improvements with both pembrolizumab plus chemotherapy (PC) (11.9 vs. 10.1 months; HR: 0.69; 95% CI: 0.53–0.90) and pembrolizumab plus chemotherapy with olaparib (PCO) (13.7 vs. 10.1 months; HR: 0.66; 95% CI: 0.51–0.85), when compared to chemotherapy alone. Additionally, both regimens demonstrated encouraging trends in overall survival (OS) (PC: 44.5 vs. 37.3 months, HR: 0.80, 95% CI: 0.60–1.08; PCO: 43.6 vs. 37.3 months, HR: 0.85, 95% CI: 0.63–1.15) [10]. These findings suggest a promising role for ICIs in first-line therapy for this otherwise difficult-to-treat subgroup and support further exploration of ICI-based combination regimens pending longer-term survival data and regulatory approval.

Despite their therapeutic promise, ICI-based regimens impose substantial financial burdens, especially when combined with agents such as bevacizumab and PARP inhibitors. Given the large number of patients who could potentially benefit from these treatments, an economic assessment is vital to determine whether the costs of these regimens are in proportion to the achieved clinical benefit. Cost-effectiveness analysis (CEA) serves as a valuable tool to quantify costs relative to the health gains, guiding healthcare resource allocation and can show for which patient population a certain therapy may be most cost effective. To date, no published CEAs have evaluated ICI-based strategies in non-BRCAm/HRD-negative AOC. The objective of this study is to assess the cost-effectiveness of the first-line treatments evaluated in the DUO-O (BC, DBC, and DBCO) and KEYLYNK-001 (C, PC, and PCO) trials.

Materials and methods

Key clinical and demographic parameters for the model were extracted from the DUO-O and KEYLYNK-001 randomized controlled trials (RCTs) [9, 10]. Although each trial enrolled distinct patient populations, both targeted individuals with newly diagnosed AOC, and the baseline characteristics were sufficiently comparable to support indirect comparisons in the context of cost-effectiveness modeling (Supplementary Table S1). This study followed the Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) guidelines for economic evaluations [11] (Supplementary Table S2).

Model construction

Patients and treatment

The hypothetical cohort comprised treatment-naïve female patients average age of 60 years with advanced (stage III/IV) epithelial ovarian cancer who were both BRCA wild-type (non-BRCAm) and HRD-negative. The model simulated clinical outcomes and costs for six alternative first-line treatment regimens: Chemotherapy alone (C): Paclitaxel (175 mg/m²) plus carboplatin (AUC 5–6), administered for up to six cycles (N = 143); Bevacizumab plus chemotherapy (BC): Includes bevacizumab (15 mg/kg, up to 20 cycles) (N = 216); Durvalumab plus chemotherapy with bevacizumab (DBC): Durvalumab (1,120 mg, up to 35 cycles) plus chemotherapy and bevacizumab (N = 199); Durvalumab plus chemotherapy with bevacizumab and olaparib (DBCO): Olaparib (300 mg twice daily, up to 35 cycles) in addition to DBC (N = 211); Pembrolizumab plus chemotherapy (PC): Pembrolizumab (200 mg, up to 35 cycles) plus chemotherapy (N = 157); and Pembrolizumab plus chemotherapy and olaparib (PCO): As above, with addition of olaparib (N = 149) [9, 10] (Supplementary Table S3). Patients experiencing disease progression or treatment-related adverse events (AEs) were assumed to transition to second-line therapy, consisting of gemcitabine (1,000 mg/m² on days 1 and 8) and carboplatin (AUC 4, up to six cycles), with or without bevacizumab, in accordance with trial protocols and treatment guidelines [9, 10, 12]. Patients who did not receive further active treatment were assumed to enter best supportive care (BSC). End-of-life care costs were included for patients approaching death. For dose calculations, the standard patient profile was assumed: 60-year-old female, body weight 70 kg, body surface area 1.84 m², and serum creatinine level of 1.0 mg/dL [13] (Table 1).

Table 1.

Model parameters

Variable	Baseline value (Range)	Distribution	Reference
Clinical Parameters
Weibull survival model for OS
Chemotherapy group	Scale = 0.0113491, Shape = 1.1354952	NA	[10]
Bevacizumab plus chemotherapy group	Scale = 0.0009981, Shape = 1.7743651	NA	[9]
Durvalumab plus chemotherapy with bevacizumab group	Scale = 0.0027558, Shape = 1.4937163	NA	[9]
Durvalumab plus chemotherapy with bevacizumab and olaparib group	Scale = 0.0008282, Shape = 1.8190000	NA	[9]
Pembrolizumab plus chemotherapy group	Scale = 0.0079970, Shape = 1.1670700	NA	[10]
Pembrolizumab plus chemotherapy with olaparib group	Scale = 0.0092070, Shape = 1.1464741	NA	[10]
Weibull survival model for PFS
Chemotherapy group	Scale = 0.0589300, Shape = 1.0518200	NA	[10]
Bevacizumab plus chemotherapy group	Scale = 0.0124420, Shape = 1.4175400	NA	[9]
Durvalumab plus chemotherapy with bevacizumab group	Scale = 0.0234860, Shape = 0.0234860	NA	[9]
Durvalumab plus chemotherapy with bevacizumab and olaparib group	Scale = 0.0086170, Shape = 1.4095070	NA	[9]
Pembrolizumab plus chemotherapy group	Scale = 0.1026000, Shape = 0.7296300	NA	[10]
Pembrolizumab plus chemotherapy with olaparib group	Scale = 0.0703400, Shape = 0.8321680	NA	[10]
Rate of post-discontinuation therapy
Chemotherapy group	0.105 (0.084–0.126)	Beta	[10]
Bevacizumab plus chemotherapy group	0.215 (0.172–0.258)	Beta	[9]
Durvalumab plus chemotherapy with bevacizumab group	0.263 (0.210–0.316)	Beta	[9]
Durvalumab plus chemotherapy with bevacizumab and olaparib group	0.347 (0.278–0.416)	Beta	[9]
Pembrolizumab plus chemotherapy group	0.106 (0.085–0.127)	Beta	[10]
Pembrolizumab plus chemotherapy with olaparib group	0.204 (0.163–0.245)	Beta	[10]
Risk for main AEs in chemotherapy group
Anemia	0.111 (0.089–0.133)	Beta	[10]
Neutropenia	0.196 (0.157–0.235)	Beta	[10]
Neutrophil count decreased	0.125 (0.100–0.150)	Beta	[10]
Risk for main AEs in BC group
Anemia	0.080 (0.064–0.096)	Beta	[9]
Hypertension	0.109 (0.087–0.131)	Beta	[9]
Neutropenia	0.261 (0.209–0.313)	Beta	[9]
Risk for main AEs in DBC group
Anemia	0.080 (0.064–0.096)	Beta	[9]
Hypertension	0.091 (0.073–0.109)	Beta	[9]
Leukopenia	0.051 (0.041–0.061)	Beta	[9]
Neutropenia	0.279 (0.223–0.335)	Beta	[9]
Risk for main AEs in DBCO group
Anemia	0.251 (0.201–0.301)	Beta	[9]
Hypertension	0.071 (0.057–0.085)	Beta	[9]
Leukopenia	0.079 (0.063–0.095)	Beta	[9]
Neutropenia	0.310 (0.248–0.372)	Beta	[9]
Trombocytopenia	0.061 (0.049–0.073)	Beta	[9]
Risk for main AEs in PC group
Anemia	0.136 (0.109–0.163)	Beta	[10]
Neutropenia	0.160 (0.128–0.192)	Beta	[10]
Neutrophil count decreased	0.130 (0.104–0.156)	Beta	[10]
Risk for main AEs in PCO group
Anemia	0.223 (0.178–0.268)	Beta	[10]
Neutropenia	0.206 (0.165–0.247)	Beta	[10]
Neutrophil count decreased	0.127 (0.102–0.152)	Beta	[10]
Health parameters
Utilities and disutilities
Utility of PFS	0.750 (0.600–0.900)	Beta	[13, 14]
Utility of PD	0.500 (0.400–0.600)	Beta	[13, 14]
Disutility of AEs (≥ 3 grade)	0.061 (0.049–0.073)	Beta	[13–16]
Cost Parameters
Drugs per cycle
Bevacizumab	15,370 (12,296–18,444)	Gamma	[17]
Carboplatin	79 (63–95)	Gamma	[17]
Durvalumab	18,790 (15,032–22,548)	Gamma	[17]
Olaparib	25,973 (20,778–31,168)	Gamma	[18]
Paclitaxel	70 (56–84)	Gamma	[17]
Pembrolizumab	23,041 (18,433–27,649)	Gamma	[17]
Second-line therapy	15,800 (12,640–18,960)	Gamma	[17]
AEs
Chemotherapy group	7,098 (5,678.4–8,517.6)	Gamma	[13]
Bevacizumab plus chemotherapy group	5,742 (4,593.6–6,890.4)	Gamma	[13, 15]
Durvalumab plus chemotherapy with bevacizumab group	7,151 (5,720.8–8,581.2)	Gamma	[13, 15]
Durvalumab plus chemotherapy with bevacizumab and olaparib group	9,395 (7,516–11,274)	Gamma	[13, 15]
Pembrolizumab plus chemotherapy group	6,588 (5,270.4–7,905.6)	Gamma	[13]
Pembrolizumab plus chemotherapy with olaparib group	7,904 (6,323.2–9,484.8)	Gamma	[13]
Imaging/Laboratory per cycle	251 (201–301)	Gamma	[13]
Administration per cycle	174 (139–209)	Gamma	[13]
Germline BRCA testing per patient	3,074 (2,459–3,689)	Gamma	[15]
HRD test per patient	4,961 (3,969–5,953)	Gamma	[15]
Best supportive care cost per cycle	9,795 (7,836–11,754)	Gamma	[13]
End-of-life cost per patients	58,426 (46,741–70,111)	Gamma	[13]
Other parameters
Discount rate	0.030 (0–0.036)	Uniform	[13, 14]
Body surface area, m2	1.840 (1.472–2.208)	Normal	[13]
Body weight, Kg	70 (56–84)	Normal	[13]

Abbreviation OS Overall survival, PFS Progression-free survival, AEs Adverse events, BC Bevacizumab plus chemotherapy, DBC Durvalumab plus chemotherapy with bevacizumab, DBCO Durvalumab plus chemotherapy with bevacizumab and olaparib, PC Pembrolizumab plus chemotherapy, PCO Pembrolizumab plus chemotherapy with olaparib, PD Progressive disease, BRCA Breast cancer susceptibility gene; HRD, homologous recombination deficiency

Model overview and transition probabilities

A Markov model comprising three distinct health states—progression-free survival (PFS), progressive disease (PD), and death—was developed to simulate the long-term clinical and economic outcomes of patients with non-BRCAm/HRD-negative AOC receiving six different first-line treatment strategies (Fig. 1). The model employed a 6-week cycle length to reflect treatment and disease progression intervals. Transition probabilities between health states were derived by extracting data points from Kaplan–Meier curves for PFS and overall survival (OS). These curves were then reconstructed and fitted to five candidate parametric survival distributions: Exponential, Gompertz, Log-logistic, Log-normal, and Weibull [14]. The Weibull model provided the best fit based on Akaike information criterion (AIC), Bayesian information criterion (BIC), and visual inspection (Supplementary Figure S1 and Table S4). The corresponding scale (γ) and shape (λ) parameters for the selected Weibull distributions were estimated and incorporated into the model (Table 1). Software tools used for data extraction and model implementation included TreeAge Pro 2021, GetData Graph Digitizer (v2.26), MATLAB (vR2020a), and R Studio (v4.2.2).

Fig. 1.

Markov Model. Abbreviation: non-BRCAm, BRCA non-mutated; HRD, homologous recombination deficiency

Statistical analyses

Cost and utility estimates

The model adopted a direct medical cost perspective, incorporating expenses associated with medications, drug administration, adverse event (AE) management, BRCA and HRD testing, imaging, laboratory diagnostics, best supportive care (BSC), and end-of-life care (Table 1). Drug costs were sourced from the Centers for Medicare & Medicaid Services and Drugs.com databases [17, 18]. The remaining costs come from previously published literature [13, 15]. Adjusted to 2025 U.S. dollars using the Consumer Price Index (CPI) [19]. The utility was used to reflect patients’ quality-of-life (QoL) weights in the natural history of the disease, on a scale of 0 (death) to 1 (total health). Utilities were used to obtain QALYs by discounting LYs. Health utility values were drawn from prior studies, assigning mean utility scores of 0.750 for the PFS state and 0.500 for the PD state [13, 14] (Table 1). Experiencing AEs (with an incidence rate higher than 5% and at grade 3 or higher AEs) that may have a relatively high negative impact on health utilities (disutility of 0.061) needs to be considered for adjusting the average utilities [13–16] (Table 1). All future costs and health benefits were discounted at an annual rate of 3% in accordance with health economic evaluation standards [13, 14].

Cost-effectiveness analyses

Key outcomes included total healthcare costs, life years (LYs), quality-adjusted life years (QALYs), incremental cost-effectiveness ratios (ICERs), incremental net monetary benefits (INMBs), and incremental net health benefits (INHBs), evaluated at a willingness-to-pay (WTP) threshold of $150,000 per QALY, consistent with USA guidelines [13, 14]. ICER refers to the incremental cost required to achieve an incremental effect, and it makes the difference between different treatments by comparing the ratio of the cost difference to the effective output. ICER = ∆C/∆E (∆C is the incremental cost, ∆E is the incremental effect). The calculated ICER results should also be correlated with the patient’s WTP to determine whether the intervention is cost-effective. INMB and INHB were calculated as follows:

where and respectively represent the efficacy and costs associated with other treatments or C. Per World Health Organization (WHO) cost-effectiveness guidelines, a strategy is deemed not cost-effective if its ICER exceeds the WTP threshold [20]. If the negative ICERs are delivered due to the treatment with high cost and low efficacy, this plan is considered an undominated strategy.

Sensitivity analyses

To account for parameter uncertainty, a one-way deterministic sensitivity analysis was conducted on over 50 variables, including utility values, drug prices, and AE-related costs [14] (Table 1). Each input was varied by ± 20%, and the impact on ICERs was recorded [13]. A probabilistic sensitivity analysis (PSA) was also performed, involving 10,000 Monte Carlo simulations. In the PSA, cost inputs were assigned gamma distributions, while probabilities and utilities followed beta distributions, consistent with standard modeling practices [16].

Results

Base-case results

In terms of survival benefit, patients treated with BC, DBC, DBCO, PC, and PCO experienced life expectancy gains of 0.39 LYs (4.68 months), 0.31 LYs (3.72 months), 0.08 LYs (0.96 months), 1.35 LYs (16.20 months), and 1.12 LYs (13.44 months), respectively, compared to chemotherapy alone. When incorporating quality-of-life adjustments and discounting, the total costs and associated QALYs for each strategy were: $618,269 (3.01 QALYs) for BC, $783,786 (3.11 QALYs) for DBC, $944,050 (3.03 QALYs) for DBCO, $819,105 (3.62 QALYs) for PC, and $917,617 (3.27 QALYs) for PCO. The chemotherapy-only group incurred $502,978 for 2.97 QALYs. Despite modest QALY gains, none of the strategies demonstrated cost-effectiveness under the $150,000/QALY WTP threshold. PC emerged as the most economically favorable option, though its ICER remained high at $484,706/QALY (Table 2; Supplementary Figure S2). Among ICI-based regimens, PC dominated both DBCO and PCO by providing better outcomes at lower costs. The ICER for PC versus DBC was estimated at $69,253/QALY (Supplementary Table S5).

Table 2.

Base-case results

Outcomes	Chemotherapy	Chemotherapy plus bevacizumab	Durvalumab plus chemotherapy with bevacizumab	Durvalumab plus chemotherapy with bevacizumab with olaparib	Pembrolizumab plus chemotherapy	Pembrolizumab plus chemotherapy with olaparib
Total cost, $	502,978	618,269	783,786	944,050	819,105	917,617
Overall LYs	4.93	5.32	5.24	5.01	6.28	6.05
Overall QALYs	2.97	3.01	3.11	3.03	3.62	3.27
INHB, QALY	[Referent]	−0.73	−1.73	−2.88	−1.46	−2.46
INMB, $	[Referent]	−109,291	−259,808	−432,072	−218,627	−369,639
ICER, $/LY	[Referent]	292,951	905,832	5,513,400	234,168	370,213
ICER, $/QALY	[Referent]	2,632,524	2,005,771	7,351,200	484,706	1,382,130

Abbreviation: Lys Life-years, QALYs Quality-adjusted life-years, ICER Incremental cost-effectiveness ratio, INHB Incremental net health benefits, INMB Incremental net monetary benefit

Sensitivity analyses

One-way sensitivity analyses revealed that model outcomes remained robust against input variability across all treatment comparisons with the chemotherapy group, indicating high model stability (Supplementary Figure S3). Key drivers of economic outcomes included utility estimates, drug prices, and AE-related management costs, with ICERs ranging from -$92,911,052 to $54,692,190 per QALY (Fig. 2). PSA results further corroborated the base-case findings. At the $150,000/QALY WTP threshold, chemotherapy had a 95.08% likelihood of being cost-effective, while all other regimens combined accounted for only 4.92%. Notably, PC surpassed a 50% probability of being cost-effective only when the WTP threshold increased to $578,000/QALY (Fig. 3). Pairwise PSA comparisons also favored PC, with probabilities of being cost-effective at 76.95% versus DBC, 99.83% versus DBCO, and 98.17% versus PCO (Supplementary Figure S4).

Fig. 2.

The One-Way Sensitivity Analyses. Abbreviation: PFS, progression-free survival; AEs, adverse events; PD, progressive disease; C, chemotherapy; BC, bevacizumab plus chemotherapy; DBC, durvalumab plus chemotherapy with bevacizumab; DBCO, durvalumab plus chemotherapy with bevacizumab and olaparib; PC, pembrolizumab plus chemotherapy; PCO, pembrolizumab plus chemotherapy with olaparib; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-year

Fig. 3.

The Cost-Effectiveness Acceptability Curves. Abbreviation: BC, bevacizumab plus chemotherapy; DBC, durvalumab plus chemotherapy with bevacizumab; DBCO, durvalumab plus chemotherapy with bevacizumab and olaparib; C, chemotherapy; PC, pembrolizumab plus chemotherapy; PCO, pembrolizumab plus chemotherapy with olaparib; QALY, quality-adjusted life-year

Discussion

ICIs are increasingly being explored as a means to convert immunologically "cold" tumors into "hot" ones, thereby overcoming the therapeutic limitations associated with aggressive and treatment-resistant forms of AOC. Although these regimens have demonstrated encouraging clinical efficacy, their economic viability remains a significant concern. The present analysis indicates that, relative to standard chemotherapy, none of the currently available ICI-based combination therapies for non-BRCAm/HRD-negative AOC are cost-effective under a conventional WTP threshold of $150,000/QALY. This will result in the use being restricted to a small group of patients (such as those with refractory conditions), thereby reducing the impact on the overall budget. However, among the immunotherapy-inclusive regimens, the PC strategy offers a relatively more favorable cost-effectiveness profile. This will be given a higher “replacement priority” and a more favorable modification of the original standard treatment plan. Sensitivity analyses highlighted that the most influential variables in the economic assessment were utility scores, ICI drug costs, and expenses related to AE management. These findings suggest potential avenues to improve cost-effectiveness, such as enhancing patient-reported outcomes, reducing drug prices, and optimizing AE management protocols. Nevertheless, the present results also show that even when assuming zero drug costs, some regimens still failed to meet cost-effectiveness criteria, underscoring the inherent limitations of these strategies and reinforcing the model’s internal validity. These findings offer valuable context for interpreting the clinical outcomes reported in recent pivotal trials such as DUO-O and KEYLYNK-001, particularly with regard to balancing efficacy with economic sustainability. This is also an important bargaining chip in negotiations with the medical insurance and pharmaceutical companies, in order to significantly expand the coverage of reimbursements and reduce prices. The final price after the negotiation alters the most crucial parameters of the budget impact and narrows the calculation scope of the budget impact.

Currently, only two classes of targeted agents are employed in the first-line treatment of AOC: the anti-angiogenic monoclonal antibody bevacizumab and PARP inhibitors. Major clinical trials including GOG-218, ICON-7, SOLO-1, and PAOLA-1 have demonstrated that bevacizumab, olaparib, and their combination regimens can yield favorable survival outcomes [21–24]. Parallel to these clinical advancements, pharmacoeconomic analyses have been conducted to evaluate the cost-effectiveness of such interventions. For example, Mehta et al. reported that the combination of BC is not cost-effective when compared to C alone in the first-line setting for AOC [25]. Further evaluations by Elsea et al. and Zhu et al. indicated that BCO is not cost-effective compared to BC alone, both in the HRD-positive subgroup and the overall AOC population [15, 26]. Notably, these findings were derived from a U.S. healthcare payer perspective. However, cost-effectiveness outcomes appear to vary geographically. Studies by Tan et al., Moya-Alarcon et al., Armeni et al., and Muston et al. have shown that combining chemotherapy with olaparib for BRCA-mutated AOC is cost-effective in Singapore, Spain, Italy, and the United States, respectively [27–30]. This underscores the universal importance of pharmacoeconomic evaluations in guiding the adoption of novel treatment strategies, while also highlighting the limitations of directly extrapolating economic data from one country or region to another. In contrast to existing analyses, this study is the first to assess the cost-effectiveness of ICIs as part of a first-line treatment regimen for AOC. The population modeled in this analysis primarily consisted of patients with non-BRCAm/HRD-negative tumors, a subgroup characterized by higher recurrence rates and more aggressive disease biology.

Emerging evidence from the DUO-O and KEYLYNK-001 trials suggests that incorporating ICIs into frontline therapy may confer survival benefits. Consequently, future international guidelines may endorse these novel regimens. Once such therapies receive formal approval, clinicians and patients may naturally gravitate toward them. However, the present study reveals a significant imbalance between the high costs of these treatments and their incremental clinical benefits. This raises concerns about potential financial toxicity for patients and the broader healthcare system, especially in the absence of long-term follow-up data. Therefore, economic evaluations should be a prerequisite before the widespread adoption of new treatment protocols. Clinically, this means that decisions regarding the use of high-cost regimens in the frontline setting must weigh anticipated benefits against associated risks and financial implications. Importantly, these findings should not be misinterpreted as a blanket rejection of ICI-based therapies for AOC simply due to their low cost-effectiveness. Instead, the clinical value of innovative regimens must be interpreted in conjunction with their economic impact. For instance, patients with non-BRCAm/HRD-negative disease, a subgroup often deemed as harboring "silent tumors" due to their insidious nature, have limited therapeutic options. The model used in this study revealed that chemotherapy alone results in an expected survival of 59.16 months, while adding ICIs and targeted therapies extends life expectancy by 0.96 to 16.20 months, depending on the specific combination. Given the modest but potentially meaningful survival gains, particularly under constraints of limited resources, the ICI-based PC regimen may be a reasonable first-line option.

Immunotherapy represents a promising frontier in ovarian cancer management. While monotherapy with ICIs has yielded suboptimal results, combination approaches have shown more encouraging efficacy. A key challenge is the immunosuppressive tumor microenvironment (TME) in ovarian cancer, which inhibits the infiltration and activity of tumor-reactive immune cells. Overcoming this hurdle requires a deeper understanding of the dynamic interactions among immune cell subsets, tumor cells, and stromal elements within the TME. Looking forward, several important questions remain unanswered. These include the identification of reliable biomarkers such as PD-1 combined positive score (CPS) and tumor mutational burden (TMB), as well as patient-related predictors like histological subtypes, ECOG performance status, and age, which could guide stratified treatment and economic decision-making. Further research is also needed to minimize treatment-related toxicities.

This study offers several notable strengths. First, to the best of our knowledge, it is the first cost-effectiveness analysis comparing ICI-based regimens to standard-of-care options (C or BC) for AOC. These findings provide a valuable complement to existing RCT data and serve as a practical reference for therapeutic decision-making. Second, a Markov model was employed to simulate both clinical and economic outcomes for AOC patients, enabling the extension of survival projections beyond available trial data and thereby enhancing the reliability of long-term estimates. Third, this analysis focused on the non-BRCAm/HRD-negative subgroup, which has not been the primary focus of many prior studies but represents a clinically significant population requiring tailored treatment strategies. Finally, this economic assessment delivers critical insights for patients, clinicians, and healthcare policy-makers alike.

Nonetheless, this study has certain limitations. Foremost among them is the reliance on indirect comparisons between two separate randomized clinical trials, necessitated by the absence of head-to-head studies evaluating all ICI-based regimens. Although the trial populations share broadly similar characteristics, inter-trial variability and inherent biases cannot be excluded. Additionally, the long-term survival estimates in this study are extrapolations derived from existing data, which introduces some degree of uncertainty. Despite this, rigorous sensitivity analyses were employed in concert with robust statistical methods to mitigate these limitations and ensure the reliability of these conclusions.

Conclusion

Based on a WTP threshold of $150,000/QALY, neither the BC nor the ICI-based first-line regimens for AOC patients with non-BRCAm/HRD-negative status are deemed cost-effective in the United States. However, in settings where medical resources are constrained and achieving extended survival is prioritized, the ICI-based PC combination may be a viable frontline treatment option. These findings emphasize the necessity of identifying predictive biomarkers for immunotherapy responsiveness, thereby enabling stratified and personalized treatment strategies in clinical practice for ovarian cancer.

Authors’ contributions

**Hong Zhu:** Conceptualization, methodology, software, resources, data curation, writing-original draft, and writing-review and editing, supervision, project administration, and funding acquisition. **Youwen Zhu:** Conceptualization, methodology, validation, formal analysis, investigation, writing-original draft, and writing-review and editing, and visualization. **Xiaofang Zhou:** Conceptualization, methodology, validation, formal analysis, investigation, writing-original draft, and writing-review and editing. **Kun Liu:** Conceptualization, methodology, validation, formal analysis, investigation, writing-original draft, and writing-review and editing. All authors have read and approved the manuscript.

Funding

This work was partly supported by the Changsha Natural Science Foundation of Hunan Provincial of China (Grant/Award Number: kq2208376 to HZ).

Data availability

All authors had full access to all of the data in this study and take complete responsibility for the integrity of the data and accuracy of the data analysis. The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study is not a clinical trial (Clinical trial number: not applicable). 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.

Competing interests

The authors declare no competing interests.