P-098 - Local advanced gastric cancer: optimization of management

Date 04 July 2015
Event WorldGI 2015
Session Posters
Topics Gastric Cancer
Presenter O. Kshivets
Citation Annals of Oncology (2015) 26 (suppl_4): 1-100. 10.1093/annonc/mdv233
Authors O. Kshivets
  • Kaluga Cancer Center, Kaluga/RU

Abstract

Introduction

Search of best treatment plan for local advanced gastric cancer (GC) patients (GCP) (T4bN0-2M0) was realized.

Methods

We analyzed data of 144 consecutive GCP (age = 55.7 ± 9.5 years; tumor size = 8 ± 3 cm) radically operated (R0) and monitored in 1975-2015 (m = 95, f = 49; total gastrectomy = 41, distal gastrectomy = 70, proximal gastrectomy = 33, combined gastrectomy with resection of 1-6 adjacent organs (pancreas, liver, diaphragm, colon transversum, splenectomy, small intestine, kidney, adrenal gland, etc.) = 144; T4b = 144; M1 = 0; N0 = 47, N1 = 10, N2 = 87; G1 = 37, G2 = 25, G3 = 82; only surgery-S = 97, adjuvant treatment-AT = 47 (chemoimmunotherapy: 5-FU + thymalin/taktivin). Survival curves were estimated by the Kaplan-Meier method. Differences in curves between groups of GCP were evaluated using a log-rank test. Cox modeling, clustering, SEPATH, Monte Carlo, bootstrap simulation and neural networks computing were used to determine any significant dependence.

Results

For total of 144 GCP overall life span (LS) was 1685.3 ± 2100.3 days, (median = 728 days) and cumulative 5-year survival (5YS) reached 43.9%, 10 years – 36.5%, 20 years – 30.4%. 40 GCP lived more than 5 years without GC progressing. 72 GCP died because of GC during the first 5 years after surgery. 5YS was superior significantly after AT (69.3%) compared with S (35.1%) (P = 0.001 by log-rank test). Cox modeling displayed that 5YS significantly depended on: phase transition (PT) N0-N12 in term of synergetics, tumor growth, histology, localization, age, AT, ESS, color index, residual nitrogen, blood chlorides, hemorrhage time (P = 0.000-0.049). Neural networks computing, genetic algorithm selection and bootstrap simulation revealed relationships between 5YS and PT N0-N12 (rank = 1), color index (rank = 2), eosinophils (3), ESS (4), age (5), thrombocytes/cancer cells – CC (6), eosinophils/CC (7), healthy cells/CC (8), AT (9). Correct prediction of 5YS was 100% by neural networks computing.

Conclusion

Optimal management strategies for local advanced GCP are: 1) availability of experienced surgeons because of complexity of radical procedures; 2) aggressive en block surgery and adequate lymph node dissection for completeness; 3) high-precision prediction; 4) adjuvant treatment for GCP with unfavorable prognosis.