PT - JOURNAL ARTICLE AU - Woff, Erwin AU - Hendlisz, Alain AU - Salvatore, Lisa AU - Genovesi, Dario AU - Critchi, Gabriela AU - Ameye, Lieveke AU - Guiot, Thomas AU - Paesmans, Marianne AU - Flamen, Patrick TI - Lowering threshold to 15% for early metabolic response assessment using <sup>18</sup>F-FDG PET/CT in metastatic colorectal cancer: results from prospective studies with external validation cohort DP - 2019 May 01 TA - Journal of Nuclear Medicine PG - 215--215 VI - 60 IP - supplement 1 4099 - http://jnm.snmjournals.org/content/60/supplement_1/215.short 4100 - http://jnm.snmjournals.org/content/60/supplement_1/215.full SO - J Nucl Med2019 May 01; 60 AB - 215Objectives: Metastatic colorectal cancer (mCRC) has a poor prognosis, however strong variation of survival exists among patients, at least partially related to the rapid apparition of treatment resistance. 18F-FDG PET/CT is a sensitive imaging tool for early detection of treatment resistance. Early non-response detection should have a high negative predictive value through the use of a low response threshold to avoid unjustified discontinuation of an effective treatment. This prospective study aimed to assess and validate the 15% response threshold for early metabolic response assessment in large exploratory and validation cohorts of mCRC patients treated respectively with multikinase inhibitors and chemotherapy. Methods: The exploratory cohort was composed of 224 patients with unresectable chemorefractory mCRC enrolled in two prospective Belgian multicenter non-randomized clinical trials investigating new targeted agents: combined sorafenib/capecitabine (EudraCT number: 2010-023695-91) and regorafenib (EudraCT number: 2012-005655-16). The validation cohort included 122 mCRC patients in an Italian prospective monocentric study treated with chemotherapy and bevacizumab. Standardized baseline and early 18F-FDG PET/CT scans were performed within 2 weeks before treatment and 2-3 weeks after the beginning of the therapy, respectively. After independent quality control, 216 and 102 patients for the exploratory and the validation cohorts, respectively, were considered suitable for analysis. The European Organization for Research and Treatment of Cancer (EORTC) criteria with threshold of 15%, PET Response Criteria in Solid Tumors (PERCIST) 1.0 with thresholds of 15 and 30%, and CONSIST method (non-responder patient identified when at least one target lesion showed no significant decrease of SUVmax (&lt;15%)) were tested. Univariate analyses for overall survival (OS) were performed to assess the predictive values of these different response criteria and multivariate analyses to assess their prognostic independency along with well-known clinical factors (i.e. age, gender, BMI, ECOG performance status, time since diagnosis, previous use of bevacizumab, KRAS mutation, and use of sorafenib vs. regorafenib). Results: In the exploratory cohort, univariate analyses showed that early 18F-FDG PET/CT response was strongly related to outcome whatever the criteria used (hazard ratio (HR): 1.65, P &lt; 0.001 for CONSIST 15%; HR: 1.48, P = 0.005 for PERCIST with 15% threshold and HR: 1.48, P = 0.006 for PERCIST 30%). Early 18F-FDG PET/CT response was identified by multivariate analyses as independent predictor of OS along with clinical factors for the different response criteria (HR: 1.62, P = 0.001 for CONSIST 15%; HR: 1.45, P = 0.01 for PERCIST 15%; and, HR: 1.44, P = 0.02 for PERCIST 30%). In the validation cohort, only early 18F-FDG PET/CT response using the 15% threshold was confirmed to be strongly related to outcome (HR: 2.05, P = 0.002 for CONSIST 15%; HR: 2.08, P = 0.003 for EORTC 15%; and, HR: 2.13, P = 0.002 for PERCIST 15%). PERCIST with 30% threshold was not significantly related to outcome (HR: 1.55, P = 0.055). Conclusions: This prospective study based on two exploratory and one validation cohort validates the strong predictive value of early 18F-FDG PET/CT response for OS in mCRC patients treated with different treatment modalities (targeted therapy/chemotherapy). A lowered metabolic response threshold of 15% was validated as the most appropriate method for OS prediction in a setting of early metabolic response assessment.