04:03 PM Abstract No. 109 Transarterial chemoembolization using small diameter drug-eluting beads in patients with neuroendocrine liver metastases

Abstract

and progress-free survival. However, it fails to control the duration of disease prior to presenting for liver directed therapy (LDT). Our study aims to identify potential lead time (L-T) bias that may affect therapy choice and subsequent outcomes. Materials: We reviewed 51 NELM who underwent LDT from 2012-2016 at our institution. Therapy paradigms (Tx) included Y90, cTACE, or mixed. Data collection included L-T from diagnosis to first presentation of LDT, survival duration from first LDT, age, gender, prior hepatectomy, prior octreotide or other chemotherapy, tumor burden, dominate lesion size, lobe involvement, and presence of extrahepatic metastasis. A Cox Regression with a backwards-likelihood ratio (threshold p < 0.10 for inclusion) analysis was performed for L-T and the duration of survivability after first LDT, controlling for the above factors. Results: For L-T, factors meeting threshold criteria included Tx (p ¼ 0.06), prior surgery (p ¼ 0.05), prior octreotide (p ¼ 0.007), dominant lesion size (p ¼ 0.03), multiplicity (p ¼ 0.02), and extrahepatic metastasis (p ¼ 0.09). For Tx, cTACE (24.3 months, p ¼ 0.04) and Mixed (32.6 months, p ¼ 0.06) had earlier presentations when compared to Y90 (52 months). For survivability, factors meeting threshold criteria were Tx (p ¼ 0.08), prior octreotide (p ¼ 0.04), any chemotherapy (p ¼ 0.02), and dominant lesion size (p ¼ 0.008). For Tx, cTACE (30.3 months, p ¼ 0.6) and Mixed (28.2 months, p ¼ 0.02) had longer survivability compared to Y90 (20.4 months). An ad hoc survival analysis accounting for L-T did not alter the above survivability outcomes. Conclusions: Our analysis agrees that cTACE-only may be a superior Tx than Y90-only, but also shines light that a mixed Tx may be even more efficacious in term of survivability than either alone. While Tx was an important factor predicting L-T, L-T does not influence the outcomes in our survivability analysis suggesting that a L-T bias is not present.