Abstract 470P
Background
After radical treatment of locally advanced gastric cancer (LAGC), there is a high risk of metachronous peritoneal dissemination (MPD), even after using various variants of intraperitoneal chemotherapy (IPC) and modern imaging techniques. The difficulty of early diagnosis of metachronous peritoneal dissemination (MPD) determines the necessity to estimate the probability of its development.
Methods
The analysis included long-term results of treating 1,311 patients with LAGC (pT1-4bN0-3M0, R.Borrmann scale III-IV) who underwent adjuvant therapy, normothermic intraperitoneal chemotherapy (NICT), hyperthermic intraperitoneal chemotherapy (HIPEC), adjuvant polychemotherapy (APCT) and their combinations. MPD risks were assessed using the Fine-Gray model.
Results
The intensity of MPD development was found to change over the 5-year follow-up period. To examine time-varying risk factors for MPD development, we used the Fine-Gray competing risks model with stratification by follow-up time intervals (1, 2, 3-5 years). The influence of age, pT, pN, degree of differentiation, tumor growth form, volume of surgical and adjuvant treatment was assessed. Risk ratios (95% Confidence Interval) were determined from the model and found to be:1) decreasing risk at the age >65 years with each year of follow-up - year 2 OR=0.53 (0.32-0.87), year 3-5 OR=0.59 (0.37-0.92); 2) MPD risks increased in patients with pN+, high grade adenocarcinoma and infiltrative LAGC; 3) risk reduction after APCT during the 1st year - OR=0.21 (0.08-0.52); 4) risk reduction during the 1st-5th year after IHT: NIHT OR=0.20 (0.09-0.45); IITHT OR=0.30 (0.18-0.50). The multidirectional influence of risk factors on MPD leads to the necessity estimating the period of maximum risk for MPD development based on individual patient factors. So the prognostic model was created and has a high prognostic value (C-index is 0.798, the AUC=0.835).
Conclusions
The non-linearity of risks should be taken into account when predicting MPD. Individualization of observation and risk assessment according to the proposed model will allow to make valid decisions on the timely application of ICT, which will increase the effectiveness of LAGC treatment.
Legal entity responsible for the study
Reutovich Michail.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.