Gastric Cancer: Сlinical Implementation of Artificial Intelligence, Synergetics, Complex System Analysis, Statistics and Simulation of Alive Supersystems.

5-year survival of GCP after radical procedures significantly depended on: 1) PT “early-invasive cancer”; 2) PT N0--N12; 3) Cell Ratio Factors; 4) blood cell circuit; 5) biochemical factors; 6) hemostasis system; 7) AT; 8) GC cell dynamics; 9) GC characteristics; 10) tumor localization; 11) anthropometric data; 12) surgery type. Optimal diagnosis and treatment strategies for GC are: 1) screening and early detection of GC; 2) availability of sufficient quantity of experienced abdominal surgeons because of complexity of radical procedures; 3) aggressive en block surgery and adequate lymph node dissection for completeness; 4) precise prediction; 5) adjuvant chemoimmunotherapy for GCP with unfavorable prognosis.

1. Gastric Cancer: Сlinical Implementation of
Artificial Intelligence, Synergetics, Complex
System Analysis, Statistics and Simulation of
Alive Supersystems
Kshivets Oleg, MD, PhD
Surgery Department, BagrationovskHospital, Bagrationovsk, Kaliningrad, Russia

2. ABSTRACT
OBJECTIVE: 5-survival (5YS) and life span after radical surgery for gastric
cancer (GC) patients (GCP)(T1- 4N0-2M0) - alive supersysems was analyzed.
The importance must be stressed of using complex system analysis, artificial
intelligence (neural networks computing), simulation modeling and statistical
methods in combination, because the different approaches yield complementary pieces of prognostic information.
METHODS: We analyzed data of 803 consecutive GCP (age=57.1±9.4 years; tumor size=5.4±3.1 cm) radically operated (R0) and monitored
in 1975- 2024 (m=560, f=243; distal gastrectomies (G)=463, proximal (G)=166, total (G)=174, combined G with resection of pancreas, liver,
diaphragm, duodenum, colon transversum, jejunum, cholecystectomy, splenectomy=341; T1=241, T2=221, T3=184, T4=157; N0=441,
N1=109, N2=253; G1=225, G2=165, G3=413; early GC=167, invasive=636; only surgery=628, adjuvant chemoimmunotherapy-AT=175: 5-
FU+thymalin/taktivin). Variables selected for prognosis study were input levels of 45 blood parameters, sex, age, TNMG, cell typ e, tumor
size. Survival curves were estimated by the Kaplan-Meier method. Differences in curves between groups of GCP were evaluated using a log-
rank test. Multivariate Cox modeling, clustering, SEPATH, Monte Carlo, bootstrap and neural networks computing were used to determine
any significant dependence.
RESULTS: Overall life span (LS) was 2153±2352.6 days and cumulative 5- year survival (5YS) reached 58.7%, 10 years – 52.5%, 20 years –
40.2%, 30 years – 28.1%. 322 GCP lived more than 5 years (LS=4337.4±2377.7 days), 172 GCP – more than 10 years (LS=5966.5±2159.7 days).
290 GCP died because of GC (LS=651±347.2 days). AT significantly improved 5YS (67.9% vs. 56.8%) (P=0.036 by log -rank test). Cox
modeling displayed that 5YS of GCP significantly depended on: phase transition (PT) early—invasive cancer in terms of synergetics, PT
N0—N12, cell ratio factors (ratio between cancer cells- CC and blood cells subpopulations), G, AT, prothrombin index, hemorrhage time,
residual nitrogen, blood cells subpopulations (P=0.000-0.041). Neural networks, genetic algorithm selection and bootstrap simulation
revealed relationships between 5YS and PT early—invasive cancer (rank=1); PT N0--N12 (2); healthy cells/CC (3), erythrocytes/CC (4),
thrombocytes/CC (5), monocytes/CC (6), segmented neutrophils/CC (7), leucocytes/CC (8), lymphocytes/CC (9), stick neutrophils/CC (10),
eosinophils/CC (11). Correct prediction of 5YS was 100% by neural networks computing (area under ROC curve=1.0; error=0.0).
CONCLUSIONS: 5-year survival of GCP after radical procedures significantly depended on: 1) PT “early-invasive cancer”; 2) PT N0--N12;
3) Cell Ratio Factors; 4) blood cell circuit; 5) biochemical factors; 6) hemostasis system; 7) AT; 8) GC cell dynamics; 9) GC characteristics;
10) tumor localization; 11) anthropometric data; 12) surgery type. Optimal diagnosis and treatment strategies for GC are: 1) screening and
early detection of GC; 2) availability of sufficient quantity of experienced abdominal surgeons because of complexity of radical procedures;
3) aggressive en block surgery and adequate lymph node dissection for completeness; 4) precise prediction; 5) adjuvant
chemoimmunotherapy for GCP with unfavorable prognosis.

3. Data:
•Males…………………………………………………...560
•Females………..……………………………...............243
•Age=57 .1±9.4years
•Tumor Size=5.4± 3.1cm
•Only Surgery.………………………………… …........628
•AdjuvantChemoimmunotherapy (5FU+thymalin/taktivin
•5-6 cycles)………………………...............................175

4. :Radical Procedures
•Total Gastrectomies (R0)…………..……….………………174
•Distal Gastrectomies (R0)……..……..………….…...........463
•Proximal Gastrectomies (R0)……….………………..........166
•Combined Gastrectomies with Resection of Pancreas,
Liver, Duodenum, Jejunum, Colon Transversum,
Diaphragm, Splenectomy (R0)…………….......................341
•D2- Lymphadenectomy………………………………………492
•D4- Lymphadenectomy.………………………………………311

5. Staging:
•T1……241 N0..….441 G1…………225
•T2……221 N1…...109 G2…………165
•T3……184 N2…...253 G3………….413
•T4……157 M0…...803
•Adenocarcinoma……………………………..803
•Early Cancer……………………………...…...167
•Invasive Cancer…………………………..…..636

6. Survival Rate:
•Alive……………………………………….....456 (56.8%)
•5-Year Survivors…………..……………….322 (40.1%)
•10-Year Survivors………………………….172 (21.4%)
•Losses……………………………………….290 (36.1%)
•General Life Span=2153± 2352.6 days
•For 5-Year Survivors=4337.4± 2377.7 days
•For 10-Year Survivors=5966.5± 2159.7 days
•For Losses=651± 347.2 days
•Cumulative 5- Year Survival……………………..58.7%
•Cumulative 10- Year Survival…………………....52.5%
•Cumulative 20- Year Survival…………………....40.2%
•Cumulative 30- Year Survival…………………....28.1%

7. General Gastric Cancer Patients Survival after
Complete Gastrectomies (Kaplan-Meier) (n=803):
Survival Function
5YS=58.7%; 10YS=52.5%; 20YS=40.2%; 30YS=28.1
Complete Censored
-10 0 10 20 30 40 50
Years after Gastrectomies
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
Cumulative Proportion Surviving

8. Results of Univariate Analysis of Phase
Transition Early—Invasive Cancer i n Prediction
of GastricCancer Patients Survival (n=803):
Cumulative Proportion Surviving (Kaplan-Meier)
P=0.0000
Complete Censored
0 5 10 15 20 25 30 35 40 45 50
Years after Gastrectomies
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cumulative Proportion Surviving
Invasive GP
Early GP

9. Results of Univariate Analysis of Phase Transition
N0—N1-2 in Prediction of Gastric Cancer
Patients Survival (n=803):
Cumulative Proportion Surviving (Kaplan-Meier)
Complete Censored
P=0.0000
0 5 10 15 20 25 30 35 40 45 50
Years after Gastrecomies
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cumulative Proportion Surviving
N1-2
N0

10. Results of Univariate Analysis of Type of Procedures
(Total Gastrectomies vs. Distal Gastrecomies) in
Prediction of GastricCancer Patients Survival (n=637):
Cumulative Proportion Surviving (Kaplan-Meier)
P=0.001
Complete Censored
0 5 10 15 20 25 30 35 40 45 50
Years after Gastrectomies
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cumulative Proportion Surviving
Distal Gasrecomies
Total Gastrecomies

11. Results of Univariate Analysis of Type of Procedures
(Total Gastrectomies vs. Proximal Gastrecomies) in
Prediction of GastricCancer Patients Survival (n=340):
Cumulative Proportion Surviving (Kaplan-Meier)
P=0.015
Complete Censored
0 5 10 15 20 25 30 35 40
Years after Gastrecomies
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cumulative Proportion Surviving
Proximal Gastrectomies
Total Gasrectomies

12. Results of Univariate Analysis of Gender
(Males vs. Females) in Prediction of GastricCancer
Patients Survival (n=803):
Cumulative Proportion Surviving (Kaplan-Meier)
P=0.027
Complete Censored
0 5 10 15 20 25 30 35 40 45 50
Years after Gastrectomies
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cumulative Proportion Surviving
Males
Females

13. Results of Univariate Analysis of Adjuvant
Treatment (Adjuvant Chemoimmunotherapy vs Surgery
along) in Prediction of GasricCancer Patients
Survival (n=803):
Cumulative Proportion Surviving (Kaplan-Meier)
P=0.036
Complete Censored
0 5 10 15 20 25 30 35 40 45 50
Years after Gastrectomies
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Cumulative Proportion Surviving
Only Surgery
Adjuvant Chemoimmuotherapy

14. Results of Cox Regression Modeling in Prediction of
Gastric Cancer Patients Survival after Complete
Gastrectomies (n=803):

15. Results of Neural Networks and Monte Carlo
Computingin Prediction of Esophageal Cancer
Patients Survival after Complete
Esophagogastrectomies (n=611
):
Corect Classification Rate=100%
Error=0.000
Area under ROC Curve=1.000

16. Results of BootstrapSimulationin Prediction of Gastric
Cancer Patients Survival after Complete Gastrectomies (n=611) :
Bootstrap Simulation n=611
Significant Factors
(Number of Samples=3333)
Rank Kendall’Tau- A P<
Healthy Cells/Cancer Cells 1 0.170 0.000
Lymphocytes/Cancer Cells 2 0.162 0.000
Leucocytes/Cancer Cells 3 0.159 0.000
Erythrocytes/Cancer Cells 4 0.158 0.000
Thrombocytes/Cancer Cells 5 0.158 0.000
Segmented Neutrophils/Cancer Cells 6 0.150 0.000
Coagulation Time 7 -0.142 0.000
Tumor Size 8 -0.120 0.000
Monocytes/Cancer Cells 9 0.118 0.000
PT N0---N12 10 -0.112 0.000
T1-4 11 -0.112 0.000
Residual Nitrogen 12 -0.109 0.000
Chlorides 13 0.105 0.000
PT Early---Invasive Cancer 14 -0.098 0.001
Procedures Type 15 -0.072 0.01
Localization 16 -0.068 0.05
Age 17 -0.067 0.05
Combined Procedures 18 0.065 0.05
Stick Neutrophils 19 -0.064 0.05
Prothrombin Index 20 -0.064 0.05
Stick Neutrophils abs 21 -0.062 0.05
Stick Neutrophils tot 22 -0.061 0.05

17. Results of Kohonen Self-Organizing Neural Networks
Computing in Prediction of Gastric Cancer Patients
Survival after Complete Gastrectomies (n=611):

18. Gastric Cancer Cell Dynamics:

19. Gastric Cancer Cell Dynamics:

20. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

21. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

22. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

23. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

24. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

25. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

26. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

27. Prognostic Equation Models of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

28. SEPATH Modeling in Prediction of Gastric Cancer
Patients Survival after Complete Gastrectomies
(n=611):

29. 5-year survival of GCP after radical
procedures significantly depended on:
1) PT “Early- Invasive Cancer”;
2) PT N0-- N12;
3) Cell Ratio Factors;
4) Blood Cell Circuit;
5) Biochemical Factors;
6) Hemostasis System;
7) Adjuvant Chemoimmunotherapy;
8) GC Characteristics;
9) GC Cell Dynamics;
10) Anthropometric Data;
11) Surgery Type (Total Gasrectomies vs.
Partial Gasrectomies).
Conclusion:

30. Optimal diagnosis and treatment
strategies for GCP are:
1) Screening and Early Detection
of GC;
2) Availability of Sufficient
Quantity of Very Experienced
Abdominal Surgeons because of
Extreme Complexity of Radical
Procedures;
3) Aggressive en block Surgery
and Adequate Lymph Node
Dissection for Completeness;
4) Precise Prediction;
5) Adjuvant Chemoimmunotherapy
for GCP with Unfavorable
Prognosis.
Conclusion:

31. Address:
Oleg Kshivets,
M.D.,Ph.D.
Consultant Thoracic, Abdominal,
General Surgeon & Surgical
Oncologist
• e-mail: [email protected]
• skype: okshivets
• http: //www.ctsnet.org/home/okshivets