Cisplatin plus high dose infusional 5-fluorouracil and leucovorin for breast cancer (BC) patients (pts) with hyperbilirubinemia secondary to liver metastases (mets).

e13111 Background: Treatment options for BC with severe hepatic impairment caused by liver mets, i.e., liver visceral crisis, are limited. We evaluated the efficacy and safety of a non-hepatic metabolized regimen: weekly cisplatin plus a 24-hour infusion of high-dose 5-fluorouracil and leucovorin (P-HDFL), in this challenging scenario. Methods: We retrospectively searched the National Taiwan University Hospital IntegrativeMedical Database to identify BC pts with liver mets and serum total bilirubinlevels (Bil-T) greater than 3 mg/dL (3-fold upper limit of normal) who received P-HDFL from January2008 to December 2018. Pts with obstructive jaundice were excluded. Liver function reserve was graded using the albumin-bilirubin (ALBI) score. Biochemical response was defined as the normalization of Bil-T.The 6-month treatment failure-free survival rate (treatmentfailure defined as treatment discontinuation for any reason,including progression, toxicity, or death) (6M-FFSR) and median overallsurvival (mOS) were calculated. Results: A total of 26 pts with liver mets and Bil-T above 3 mg/dL (range 3.2-21.6 mg/dL) without biliary obstruction were treated with P-HDFL (cisplatin 25-35 mg/m 2 on day 1, 8 plus a 24-hour infusion of 5-FU 2000-2600 mg/m 2 and leucovorin 200-300 mg/m 2 on day 1, 8 in a 3-week cycle). Five pts (19%) received prophylactic treatment of granulocyte colony-stimulating factors. Thirteen pts (50%) had hormone receptor-positive disease, and 11 pts (42%) had HER2-positive disease. Thirteen pts (50%) received P-HDFL as first or second-line chemotherapy in the metastatic setting, while the remainder of pts were more heavily pretreated. The majority (82%) of HER2-positive pts received trastuzumab concurrently. Among the 15 HER2-negative pts, 5 (33%) received bevacizumab concurrently. All pts had an ALBI grade of 2 or 3, indicating poor liver reserve. A total of 13 pts (50%) had a biochemical response. The 6M-FFSR was 15.4%. The mOS was 5.4 months. mOS was significantly longer in pts with biochemical response compared to pts without biochemical response (16.8 vs 0.7 months, hazard ratio [HR] 0.09; 95% confidence interval [CI], 0.02-0.30, p = 0.0001). Biochemical response was achieved in 8 (44%) out of the 18 pts with Bil-T greater than 6 mg/dL; for those pts, the mOS was 16.7 months. ALBI score, line of chemotherapy, and concurrent use of targeted therapies were not significantly associated with biochemical response or overall survival. 15 pts (58%) had grade 3/4 hematological adverse events, including 6 (23%) with grade 3/4 neutropenia. Only 4 pts (15%) had a grade 2 creatinine increase. Conclusions: P-HDFL is safe and effective for BC pts with non-obstructive hyperbilirubinemia secondary to liver mets, and should be considered as a favored treatment option in pts with liver visceral crisis.