Retrospective analysis of perioperative and long-term outcomes after pancreas-sparing duodenectomy for duodenal cancer

Abstract

Aims: Pancreaticoduodenectomy (PD) is the procedure of choice for malignant periampullary tumours, but is associated with significant morbidity. Pancreas-sparing duodenectomy (PSD) is technically feasible for distal duodenal cancers, and has been considered potentially safer than pancreaticoduodenectomy. However, there is little published data on either short or long-term outcomes after PSD. Our aim was to compare the outcomes of PSD and PD for duodenal cancer in a single UK centre. Methods: Patients who underwent either PD or PSD for duodenal cancer over a 21-year period (1993–2014) were identified from a prospectively maintained database. Patient demographics, tumour location, postoperative complications and histological data were compared between the two groups. Long-term overall and disease-free survivals were analysed. Results: 81 consecutive patients (median age 66 years) underwent surgical resection of duodenal cancer during the study period (PSD 21, PD 60). Patient demographics were similar in both groups. Tumours were more commonly located in the distal duodenum in patients undergoing PSD (85% vs. 9%; p < 0.001). There was no significant difference in perioperative morbidity (PSD 38% vs. PD 55%; p = 0.311), length of hospital stay (median: 10 vs. 11 days; p = 0.135) or 90-day mortality (5% vs. 12%; p = 0.672) between groups. R0 resection rates were not significantly different (PSD 75% vs. PD 92%; p = 0.112). However, lymph node yield (median 6 vs. 17; p < 0.001) and the proportion of specimens with more than six lymph nodes (35% vs. 91%; p < 0.001) were significantly lower after PSD. There was no significant difference in five-year overall (p = 0.932), disease-specific (p = 0.646) or recurrence-free (p = 0.731) survival rates between PSD (45%, 70%, 38%) and PD (45%, 61%, 36%). Conclusions: Pancreas-sparing duodenectomy is a feasible option in patients with distal duodenal cancer, but does carry a significant morbidity rate, comparable to pancreaticoduodenectomy. Although lymph node yield is lower after a pancreas-sparing resection, this does not appear to negatively impact on long-term survival.