3553 Background: In patients with mCRC, the liver represents the primary site of metastasis. In patients with primarily unresectable metastatic disease, primary tumor location and molecular subtypes mainly determine the therapeutic approach. In addition, image-based characterization of hepatic metastatic pattern may provide additional information concerning prognosis and treatment efficacy. Therefore, CRLM were characterized according to size-based heterogeneity in the PanaMa trial (maintenance therapy with 5-fluoruracil/leucovorin (FU/FA) plus panitumumab (Pmab) vs. FU/FA alone following induction therapy with six cycles of FU/FA, oxaliplatin and Pmab). Methods: Assessments were performed in patients with at least two lesions, considering target and non-target lesions, within a central radiological review of the trial. Variance in size expressed as ratio between the smallest and largest lesion measured (≤/ > threefold difference in size) was evaluated for its prognostic and predictive impact on progression-free (PFS) and overall (OS) survival of maintenance therapy. Time-to-event endpoints were expressed by Kaplan-Meier method and compared by log-rank tests. Cox regressions were used to indicate Hazard ratios (HR) with 95% CIs. Results: Imaging data of 211/248 (85.1%) patients receiving maintenance therapy were evaluable for central radiological review (FU/FA+ Pmab, n = 106; FU/FA, n = 105). Of those, 165/211 patients (78.2%) had at least two CRLM. Size of CRLM ranged between 5–180mm, with the smallest metastasis in the individual assessments measuring a median of 12mm and the largest a median of 53mm. Large heterogeneity with more than a threefold difference in size between the smallest and largest CRLM was observed in 49.7% of patients, particularly pronounced in patients with polymetastatic disease ( > 5 lesions). Homogeneous metastasis in terms of lesion size (≤ threefold difference) was associated with favorable OS (HR 0.63; 95% CI, 0.42–0.95; P = 0.027) - irrespective of study arm- compared to heterogeneous lesion disease. Maintenance therapy with FU/FA+ Pmab compared with FU/FA alone seemed favourably active in both patients with heterogeneous metastases (HR 0.53; 95% CI; 0.34–0.85; P = 0.008) as well as to a numerically lesser extent in patients with homogeneous metastases (HR 0.68; 95% CI, 0.43–1.10; P= 0.116). Conclusions: Imaging-based characterization of CRLM focusing on size-related heterogeneity was associated with favorable prognosis potentially indicating more aggressive underlying tumor dynamics in case of heterogeneous lesion size. In addition, the efficacy of Pmab during maintenance was evident in both patient groups (homogeneous and heterogeneous disease lesions) with a numerically greater effect in patients with heterogeneous lesions. Clinical trial information: NCT01991873 .