Ahmed Abbas Suleiman (1), Sebastian Frechen (1), Matthias Scheffler (2), Thomas Zander (2), Deniz Kahraman (3), Carsten Kobe (3), Lucia Nogova (2), Jürgen Wolf (2), Uwe Fuhr (1)
(1) Department of Pharmacology, Clinical Pharmacology Unit, University Hospital of Cologne, Cologne, Germany; (2) Lung Cancer Group Cologne, Department I of Internal Medicine, Center for Integrated Oncology Cologne Bonn, University Hospital of Cologne, Cologne, Germany; (3) Department of Nuclear Medicine, Center for Integrated Oncology Cologne Bonn, University Hospital of Cologne, Cologne, Germany.
Objectives: Positron emission tomography (PET) using [(18)F]-fluorodeoxyglucose (FDG) has been promoted over size-based assessment tools for predicting the response to novel anticancer agents with mechanisms unlikely to result in tumor shrinkage[1-3]. Several criteria to quantify the standardized uptake value (SUV) of FDG have been proposed[4], yet a consensus on which criteria to use has not been reached. Using data from a clinical study conducted in patients with advanced non-small cell lung cancer (NSCLC) first-treated with erlotinib[3], we aimed to develop a survival model describing the times-to-death distribution of the cohort, and to evaluate different factors including the tumor metabolic dynamics represented by the SUV of FDG as overall survival (OS) predictors. Using the model developed, we also aimed to compare the prognostic predictive abilities of different criteria used for SUV quantification.
Methods: Data from patients with stage-IV NSCLC (n=39) first-treated with erlotinib (150mg/day) were used[3]. Three FDG-PET scans were scheduled at baseline, 1 and 6 weeks after starting treatment. A parametric time-to-death model was developed by testing exponential, Weibull, Gompertz, and log-logistic distributions for best description of OS. FDG uptakes (quantified as SUVpeak, SUVmax and SUV50; defined in [4]) measured at baseline, relative changes after 1 and 6 weeks of treatment, demographics, histology, presence of a mutation in the epidermal growth factor receptor domain, smoking and baseline performance statuses were tested as OS predictors. Non-linear mixed effects modeling using NONMEM 7.3 was used for analysis.
Results: OS was described using an exponential distribution. For all SUV definition criteria, the baseline SUV and the relative change in SUV after 1 week of treatment were consistently found as statistically significant predictors of OS (p<0.05). For every unit increase in FDG uptake measured at baseline as SUVpeak, SUVmax or SUV50, the death hazard increased by 25%, 15% and 26%, respectively, while the hazard decreased for every 10% drop in FDG uptake quantified as SUVpeak, SUVmax or SUV50 after 1 week of treatment by 17%, 13% and 13%, respectively. No other tested covariates predicted OS.
Conclusions: Regardless of the criteria used for quantifying FDG uptake, FDG-PET can be used as an early survival predictor for advanced NSCLC patients treated first-line with erlotinib.
References:
[1] Wahl RL et al. J Nucl Med. (2009) 50:122S-50S
[2] Prior JO et al. J Clin Oncol (2009) 27:439-45
[3] Zander T et al. J Clin Oncol (2011) 29:1701-8
[4] Boellaard R et al. Eur J Nucl Med Mol Imaging (2010) 37:181-200
Reference: PAGE 23 (2014) Abstr 3128 [www.page-meeting.org/?abstract=3128]
Poster: Drug/Disease modeling - Oncology