Page 203 - EJMO-9-1

Eurasian Journal of Medicine and
Oncology
Cost-effectiveness of nivolumab+chemo for gastric/GEJ cancer

should be analyzed to avoid inaccuracies and uncertainties 2.4. The utility and cost estimates
in results. 15,16 The cycle of the Markov model was set with a During the follow-up, the CheckMate 649 trial used the
cycle length of 3 weeks and a time horizon of 10 years, based gastric cancer subscale to compare the quality of life
on the dosing cycle used in the CheckMate 649 clinical trial. in patients treated with nivolumab plus chemotherapy
Approximately 99% of patients were assumed to be in the versus chemotherapy alone as a first-line treatment
absorption state. A half-cycle correction was applied to for advanced G/GEJ/esophageal adenocarcinoma. The
17
simulate the transition process more accurately. Simultaneous average health utility scores (0.797 for PFS and 0.577
simulation analyses of costs and utility were performed to for PD) of these patients were derived from a previously
estimate cumulative total costs and health outcomes within published study. The top three incidence adverse
27
the cohort’s time frame. 18,19 The research was conducted from events (AEs) with grade three or above were selected
the perspective of American payers, applying a 3% discount for both treatment groups to evaluate the loss of health
on costs and utilities. According to the World Health utility caused by these AEs, simplifying the calculation.
20
Organization, an incremental cost-effectiveness ratio (ICER) In the nivolumab plus chemotherapy group, the most
is acceptable when it is below three times the Gross Domestic common grade three or above AEs included neutropenia
Product (GDP) per capita. This study used 3 times the (14%), decreased platelet count (8%), proteinuria
21
US GDP per capita in 2021 ($69,219.5) as the threshold (5%), increased blood bilirubin (5%), and increased
(https://fred.stlouisfed.org/series/A939RC0Q052SBEA), γ-glutamyltransferase (5%). In the chemotherapy group,
resulting in a willingness-to-pay (WTP) assumption of the top AEs were palmar-plantar erythrodysesthesia
$207,659. The research indicators included costs, life-years, syndrome (12%), hypertension (6%), and increased
quality-adjusted life-years (QALYs), and ICERs.
aspartate aminotransferase (5%).
2.3. The model’s survival and progression risk Costs were reported in 2021 US dollars
estimates (1 US dollar = 6.38 Chinese yuan). The study focused
The original data for constructing the model were obtained solely on direct medical costs, including drug costs,
from the CheckMate 649 clinical trial. When certain subsequent treatment costs, management costs, follow-up
data were unavailable, related published literature was costs, laboratory examination costs, and costs associated
consulted. The GetData Graph Digitizer (version 2.26; with the top three grade three or above AEs, according to
http://getdata-graph-digitizer.com/download.php) was the CheckMate 649 trial. The drug prices used were either
used to extract data from the Kaplan–Meier curves of PFS local hospital prices or obtained through consultations
and OS for both the nivolumab plus chemotherapy group with drug suppliers. The estimated cost of each drug during
and the chemotherapy group. The algorithm developed the specified period is listed in Table 2. The probabilities of
by Guyot et al., which reconstructs pseudo-individual patients in different treatment groups receiving follow-up
22
patient data using R software (version 4.1.0; https://www.r- treatments and specific subsequent treatment modalities
project.org/), was also utilized. This approach was (including systemic therapy other than PD-L1 inhibitors,
22
combined with the Akaike Information Criterion and local regional therapy, radiation therapy, and surgery) were
Bayesian Information Criterion to select the log-logistic derived from the CheckMate 649 trial.
distribution that best fit the reconstructed survival curves The calculated drug doses are based on actual clinical
for both treatment groups (Table S2). This distribution trials. In the nivolumab plus chemotherapy group, patients
23
offers greater flexibility and improved estimation of either received 360 mg of nivolumab every 2 weeks
correlations. 24,25 The transition probabilities for health or 240 mg every 2 weeks and chemotherapy (XELOX
states were estimated using specific parameters of each [capecitabine 1,000 mg/m twice daily on days 1 – 14 and
2
subgroup’s model (Table 1). 26 oxaliplatin 130 mg/m on day 1 every 3 weeks] or FOLFOX
2
[leucovorin 400 mg/m on day 1, fluorouracil 400 mg/m
2
2
Table 1. Base‑case analysis on day 1 and 1,200 mg/m on days 1 – 2, and oxaliplatin
2
85 mg/m on day 1 for every 2 weeks]). In the chemotherapy
2
Parameters Nivolumab plus chemotherapy Chemotherapy group, the patients received chemotherapy alone. We
QALYs (years) 0.75 0.27 assumed an average body surface area of 1.68 m . It
2 35
Total cost ($) 13,941.06 17,565.97 was also assumed that all patients who experienced
ICER/QALY ($) 276,799.67 disease progression would receive follow-up treatment.
WTP/QALY ($) 190,239 Based on the National Comprehensive Cancer Network
Abbreviations: ICER: Incremental cost-effectiveness ratio; 2022.1 guidelines, we selected oxaliplatin combined with
2
QALY: Quality-adjusted life-years; WTP: Willingness-to-pay leucovorin and fluorouracil therapy (oxaliplatin 85 mg/m
Volume 9 Issue 1 (2025) 195 doi: 10.36922/ejmo.7075