S1378
Interdisciplinary - Health economics & health services research
ESTRO 2026
immunotherapy’s heterogeneous timing, late-tail PFS was modestly underestimated, highlighting sensitivity of hazards beyond 36 months. Mean time (months) over 0–60 was: response 16.1, stable disease 5.6, progression 41.2. These translate into concrete planning signals: e.g., prolonged time in progression concentrates resource use (unplanned visits, imaging, palliative RT, systemic switches), while sustained response time supports reduced surveillance intensity after year one. Scenario analyses (±20% piecewise rates) preserved qualitative rankings of state-time outputs, indicating decision robustness to reasonable parameter drift. Conclusion: A piecewise CTMC is a pragmatic bridge between survival curves and decision-ready metrics for HEOR and service design in stage III NSCLC post-cCRT. It preserves interpretability (state occupancy, transition counts), supports value-based health care (time-in- state × unit costs), and generalises to other immunotherapy settings with non-proportional hazards. Current limitations include the absence of an explicit “immune-control” state and dependence on the last calibration window; future work will test hybrid mixture/cure extensions and external validation on broader real-world cohorts. Keywords: Chemoradiotherapy; Continuous-time Markov Chain Scheme-based radiotherapy vs insurance-guided radiotherapy - Analysis of the benefits to the patients from two financial assistance establishments. Vrushab Rao 1 , Pramod Tike 2 , Bhushan Nemade 3 , Sanjay Deshmukh 4 , Bhooshan Zade 1 1 CyberKnife Radiosurgery and Radiation Oncology, Ruby Hall Clinic, Pune, India. 2 Radiation Oncology, Optimus Oncology, Latur, India. 3 Radiation Oncology, Optimus Oncology, Nashik, India. 4 Surgical Oncology, Indrayani Hospital, Pune, India Purpose/Objective: One of the key barriers to timely radiotherapy in the Indian population is the financial constraints that cancer treatment burdens a patient with. The cost of investigations, prior treatments and logistics, compounded occasionally by the inability to work by the sole earning member of the family presents a serious non-pathological toxicity of treatment. Government schemes and insurance, through various agencies and policies, offer assistance to patients who have opted to avail them. This study analyses the role that insurance-guided radiotherapy (IGRT) and scheme-based radiotherapy (SBRT) play in ensuring Digital Poster 362
Digital Poster Highlight 178 Piecewise CTMC of durvalumab after cCRT in stage III NSCLC: turning survival into decision-ready state-time metrics Amadeo Wals Zurita, Maria Ángeles González Ruiz, Carolina Ortiz Bautista, Cristina Nebrera Navarro, Gabriel Campos Rivera, Jeronimo Pachón Ibáñez, David Muñoz Carmona, Carlos Míguez Sánchez Radiation Oncology, Virgen Macarena University Hospital, Seville, Spain Purpose/Objective: Durvalumab after concurrent chemoradiotherapy (cCRT) in stage III NSCLC shows delayed, non- proportional benefits. To support value-based adoption and service planning, we built a transparent, parsimonious continuous-time Markov chain (CTMC) that converts survival endpoints into expected time-in- state metrics directly usable for health-economic and
operational decisions. Material/Methods:
We specified five mutually exclusive clinical states (response, stable disease, progression, post-CR relapse, death) and estimated a piecewise Q-matrix (0– 12, 12–24, 24–60 months). Primary calibration used OS/PFS landmarks from PACIFIC at 12, 24 and 36 months; extrapolation ≥ 36 months was left unconstrained. From Q we derived P(t), cohort distributions at 12/24/36/60 months, and mean time spent in each state over 0–60 months. Face validity included visual fit against PACIFIC (3 and 5 years) and plausibility checks vs real-world summaries from PACIFIC-R. The model also reports transition counts (per 100 patients) to inform care pathways (imaging, salvage, palliative RT/systemic therapy). No cure- fraction was imposed. Results: The CTMC reproduced the early OS/PFS course and reasonably approximated 60-month OS (model ~42– 43%), consistent with PACIFIC durvalumab ( ≈ 43%) and higher than placebo ( ≈ 33%). As expected with
Made with FlippingBook - Share PDF online