- Li, Song;
- Sheikh, Farooq;
- Zaghlol, Raja;
- Basyal, Binaya;
- Gallagher, Christopher;
- Molina, Ezequiel;
- Mahr, Claudius;
- Cheng, Richard;
- Barac, Ana;
- Schlam, Ilana;
- Lee, Andy
BACKGROUND: There are limited data to guide oncology and cardiology decision-making in patients with a left ventricular assist device (LVAD) and concurrent active malignancy. OBJECTIVES: The goal of this study was to describe cancer treatment approaches, complications, and survival among patients with active cancer on LVAD support in 2 tertiary heart failure and oncology programs. METHODS: In this retrospective cohort study, LVAD databases were reviewed to identify patients with a cancer diagnosis at the time of or after LVAD implantation. We created a 3:1 matched cohort based on age, sex, etiology of cardiomyopathy, LVAD implant strategy, and INTERMACS profile stratified by site. Kaplan-Meier analysis and Cox proportional hazards models were used to compare survival between patients with cancer and non-cancer comparators. RESULTS: Among 1,123 patients who underwent LVAD implantation between 2005 and 2019, 22 patients with LVADs with active cancer and 66 matched non-cancer comparators were identified. Median age was 62 years (range 41 to 73 years); 50% of patients with cancer were African-American, and 27% were women. Prostate cancer, followed by renal cell cancer and hematologic malignancies were the most common diagnoses. There was no significant difference in unadjusted Kaplan-Meier median survival estimates from the time of LVAD placement between patients with cancer (3.53 years; 95% confidence interval [CI]: 1.41 to 5.33) and non-cancer comparators (3.03 years; 95% CI: 1.83 to 5.26; log-rank P = 0.99). In Cox proportional hazard models, cancer diagnosis as a time-varying variable was associated with a statistically significant increase in death (hazard ratio: 2.05; 95% CI: 1.03 to 4.12; P = 0.04). Patients with cancer had less gastrointestinal bleeding compared with matched non-cancer comparators (P = 0.016). Other complications were not significantly different. CONCLUSIONS: Our study provides initial feasibility and safety data and set a framework for multidisciplinary team management of patients with cancer and LVADs.