RECONSIDERING LYMPHADENECTOMY FOR LOCALIZED PANCREATIC NEUROENDOCRINE TUMORS
Susanna W. de Geus*2, Heidi Overton3, Jin He3, Sing Chau Ng2, Tara S. Kent1, David McAneny2, Christopher L. Wolfgang3, Jennifer F. Tseng2, Teviah Sachs2
1Surgical Oncology, Beth Israel Deaconess Medical Center, Boston, MA; 2Surgery, Boston Medical Center, Boston, MA; 3Johns Hopkins Hospital, Baltimore, MD
Background: Management of nonfunctional pancreatic neuroendocrine tumors (PNETs) remains controversial, due to their rarity and ill-defined natural history. The value of surgical resection has been questioned for small PNETs, but the current guidelines recommend complete oncologic resection, including lymphadenectomy, for PNETs > 2 cm in size. Although the prognostic value of lymphadenectomy remains undisputed, the effect of extensive lymph node dissection on prognosis remains unclear. The purpose of this study is to assess the survival impact of lymphadenectomy in PNETs > 2 cm.
Methods: Patients undergoing pancreatic resection between 2004 and 2014 for non-metastatic PNETs measuring > 2 cm in size, were identified from the National Cancer Database. Multivariable logistic regression analysis was performed predicting resection of > 10 lymph nodes. In addition, a propensity score predicting the odds of undergoing resection > 10 lymph nodes was created. Patients were matched 1:1 based on their propensity score. Survival analysis was performed using the Kaplan-Meier method and log-rank test.
Results: In total, 1,620 patients were identified. The median number of lymph nodes examined was 10 nodes (IQR, 4 - 16 nodes), with 9.7% (n=157) and 43.3% (n=702) having positive lymph nodes on preoperative imaging and pathological examination, respectively. On multivariable analysis, tumor size > 40 mm (vs. £ 40 mm: Adjusted Odds Ratio [AOR], 1.261; p=0.0293), academic center (vs. non-academic: AOR, 1.511; p=0.0006), pancreaticoduodenectomy (vs. distal pancreatectomy: AOR, 2.103; p<0.0001), treatment at a high-volume center (vs. low-volume: AOR, 2.188; p<0.0001), and having positive lymph nodes on preoperative imaging (vs. negative: AOR, 2.654; p<0.001) were predictive for resection of > 10 lymph nodes. After matching, median survival was similar for patients with £ 10 nodes compared to > 10 nodes resected, resulting in a 3-year survival rate of 89.5% vs. 86.2% (log-rank, p=0.2801). On subset analysis in patients with positive lymph nodes, no significant survival difference was observed between patients with £ 13 and > 13 lymph nodes excised (3-year survival rate: 81.6% vs. 81.8%; log-rank p=0.3597) after matching. On sensitivity analyses using resection of 4 (3-year survival after matching: 90.4% vs. 89.9%; log-rank p=0.7517) and 6 (3-year survival after matching: 86.4% vs. 87.0%; log-rank p=0.5673) lymph nodes as cutoffs the results also remained robust.
Conclusions: Although positive lymph nodes remain associated with less favorable survival outcomes, the results of this study suggest that lymph node analysis is not associated with improved survival. The rationale for lymphadenectomy for PNET > 2 cm warrants further investigation, including its impact on optimal surveillance and adjuvant therapy
Back to 2019 Posters
