Multimodality neoadjuvant downstaging therapy before liver transplantation for colorectal liver metastasis: Real-world data from the Cleveland Clinic.

3594 Background: Liver transplantation (LT) has a clear survival benefit in advanced, unresectable, liver-confined colorectal cancer liver metastasis (CRLM). However, pre-transplant management and optimal selection criteria remain understudied and previously reported data indicate recurrence-free survival <40% at 3- and 5-years post-LT. Methods: 23 patients entered the protocol; 22 patients underwent LT (12/2017-12/2025) utilizing the Cleveland Clinic protocol for down-staging prior to LT. The protocol involves intensive locoregional therapy (LRT) with systemic chemotherapy, aiming to reach minimal disease burden based on FDG PET-CT, ctDNA, and CEA. Patients with minimal or no detectable disease or irreversible treatment-induced liver injury undergo transplant. Patients must have disease control for at least 1 year after resection of the primary colorectal tumor. Results: Median age at LT was 45 years (IQR=39-50). Median largest tumor size on pre-LT imaging was 2.2 cm (1.6–7.35), and 18 patients (82%) had bilobar disease. Most patients received intensive systemic therapy (median=18 cycles, IQR=12-37), including 5-FU (n=18), oxaliplatin (n=18), irinotecan (n=13), and bevacizumab (n=17). Tumors were KRAS (n=6) and BRAF non-V600E (n=2) mutated. 20 patients (91%) received pre-LT LRT, including radioembolization (n=11), ablation (n=9), resection (n=10), and HAI pump (HAIP) (n=7). Only one patient dropped out of the pre-LT protocol for liver failure in 2017, suspected from combination of bilobar Y90 and SBRT. Pre-LT PET showed metabolic activity in 8 patients (36%). Median PET-Metabolic Tumor Volume (MTV) at protocol entry was 49.2cm 3 (18-162) and at LT was 0cm 3 (0-6.4) with a median MTV-reduction of 45.2cm 3 (18-156). Median MELD-Na at LT was 9 (IQR 6–13). Pre-LT tumor agnostic ctDNA was detected in 12/17 patients (71%). All patients with ctDNA(-) pre-LT had non-viable tumor on explant (n=5/5 vs. ctDNA(+)=5/12, p=0.028). At median follow-up of 1.6 years, 20 (91%) remain alive, and 18 (82%) are recurrence-free. Median RFS was 5.4 years (95%CI=3.52-7.36) and OS=6.72 years (95%CI=5.34-8.00). Recurrences were lung (n=2), liver (n=1), and liver+para-aortic (n=1) nodes. There was no post-LT recurrence when non-viable on PET scan and on histology (log-rank p<0.001). Pre-LT PET-MTV>70cm 3 correlated with histologically viable tumor (p=0.031). Pre-LT HAIP demonstrated a trend towards non-viable tumor on explant (n=5/7 non-viable, 71%). Conclusions: LT for CRLM shows clear survival benefit in randomized trials. Aggressive pre-LT locoregional and systemic therapies aimed at minimizing disease burden assessed by PET scan and ctDNA may demonstrate improved outcomes, as may HAIP. A clinical trial testing these hypotheses has been designed with expected enrollment in 2026 at three sites.