Population kinetics assessment of limited small cell lung cancer (L-SCLC).

Abstract

e20602 Background: For most advanced malignancies, progression free survival (PFS) and overall survival (OS) curves approximate first order kinetics.1 PFS and OS half-lives can be calculated by curve exponential decay nonlinear regression analysis (EDNLRA). Curve characteristics vary significantly with treatment and malignancy type.1 On log-linear plots, PFS curves are significantly more likely to be convex for extensive SCLC than for other malignancies, with late acceleration of tumor growth after completion of 1st line therapy.1,2 Here we assess L-SCLC by EDNLRA. Methods: We used https://apps.automeris.io/wpd/ to digitize PFS curves from L-SCLC studies published from 1990 to 2021. We analyzed curves for study arms having > 50 patients. We used GraphPad Prism 7 for 1 and 2 phase decay EDNLRA as previously described.1 We also constructed log-linear plots. Results: We analyzed 55 L-SCLC PFS curves from 34 published trials. Median PFS half-life was 15 months for studies published 1990 to 2010 vs 19.5 months for 2011 to 2021 (P = 0.003). This improvement over time might be explained in part by stage migration (eg, with PET scanning). Of 55 log-linear plots, 53 (96%) had an inflection to the right. The curve inflection indicates there are 2 distinct subpopulations: namely progressing patients vs cured patients. Of the 55 curves, 46 fit 2 phase decay EDNLRA models. Across these 46 2-phase decay curves, EDNLRA calculations indicated that the progressing subgroup comprised 86% of the total population and had a PFS half-life of 10.5 months. The cured subgroup accounted for the other 14% of the population, with a PFS half-life greater than 15 years. The proportion cured increased from a median of 12% for 1990-2010 studies to 22% for 2011-2021 studies (p = 0.01). Stage migration might again explain this difference. For patients who are progression-free at any time point, one can also use these EDNLRA data to calculate the proportion who would eventually progress in the future. Of those still progression free at 12, 24, 36, 48, 60 and 120 months, respectively, the proportion that would eventually progress would be 73%, 56%, 36%, 20%, 10%, and 0.2%. As with extensive SCLC, all 55 L-SCLC PFS log-linear curves displayed early convexity, with a downward inflection after a period of relative stability. This downward inflection occurred at a median of 4.4 months, in keeping with acceleration of progression after completion of planned systemic therapy. Conclusions: Population kinetic assessments permit calculation of probability of eventual progression for L-SCLC patients who remain progression-free at different time points after initiation of therapy. Of L-SCLC patients who remain progression-free at 60 months, approximately 10% are nevertheless destined to eventually relapse.