PWE-002 Home parenteral nutrition in the palliative patient: a single centre experience

Background Few studies have assessed the benefit of nutritional support in advanced cancer patients. We aimed to investigate survival in patients receiving parenteral nutrition (PN) for intestinal failure (IF) in the palliative setting and identify factors associated with improved survival. Methods Patients were identified from a database in an IF tertiary referral centre. Data were collected retrospectively for all cancer patients who commenced PN for a palliative indication between 01/01/2011 and 31/03/2018. These included; age on commencing PN, time from diagnosis (first MDT) to commencing PN (lead time), initial hospital admission length, cancer type, mechanism of IF, and date of death. Co-morbidities were recorded using the validated Charlson index. Mechanism of IF was subdivided into 3 groups; obstruction (n=30), short bowel syndrome/high output fistula (n=22), and other (n=4). Cancer diagnoses were grouped into gynae and peritoneal (n=26), gastrointestinal (GI) (n=18), and other (n=12). Categorical variables were compared using fisher’s exact and continuous variables using Kruskal-Wallis. Survival rates were compared using Kaplan-Meier curves and cox regression. P Results 56 patients (mean age 57; 71.4% female) were identified during the study period. Patients with gynae/peritoneal primary were older when commencing PN (p0.02) compared to other tumour groups. Patients with gynae/peritoneal cancers also had a longer lead time prior to starting PN compared to the other groups (Kruskal-Wallis p0.01; Kaplan Meier p0.04). However once PN was started there was no difference in mortality between groups. 90 day mortality was 38.4% for gynae/peritoneal, 27.7% for GI, and 25.0% for others (p0.633). 150 day mortality, the rate was 61.5% for gynae/peritoneal, 44.4% for GI, and 50.0% for others (p value 0.515). Kaplan-Meier curves appeared to diverge between gynae/peritoneal and GI cancers although this wasn’t significant (p0.56) and cox regression showed near identical survival between these groups (HR 1.02 p0.97). There was also no difference in mortality rates for different mechanisms of IF, 90 day mortality was 36.6% for obstruction, 31.8% for short bowel/fistula, and 0.0% for others (p0.372). 150 day mortality was 60.0% for obstruction, 45.5% for short bowel/fistula, and 50.0% for others (p0.647). Conclusions Patients with slow growing gynae/peritoneal cancers have been diagnosed with cancer for longer than other cancer types prior to commencing palliative PN. However once PN is commenced survival between groups in this cohort was similar. An extended follow up and a larger sample size may yet demonstrate significant differences. A prospective study is required to assess for other factors that may influence survival.