Effects Of Endoscopic Screening On Gastric Cancer And Exploration Of Screening Strategy In High-risk Area

Gastric cancer(GC)remains a major public health concern worldwide.Screening and early detection program on GC is considered an effective way to reduce GC mortality and improve the 5-year survival rate of GC.The endoscopy-based National Upper Gastrointestinal Cancer Early Detection(UGCED)program has been implemented over ten years,but few studies have been conducted in the real-world scenario of population at high-risk of GC and the extra value of repeated endoscopy is unclear.In addition,optimal screening age and screening interval in endoscopic screening program have yet to be studied.The present study was based on the data from three sources in the high-risk area of GC in Linqu County,Shandong Province:cancer registry,a population-based prospective study based on UGCED program and a repeated gastroscopy cohort since 1989.We comprehensively assessed the burden of GC and the effectiveness of endoscopic screening on GC.And then we analyzed the optimal screening age by combining the age tendency turning points by Joinpoint liner regression model with the effects of screening in age subgroup.The screening interval was explored from the perspective of prevention and multistate Markov model.Major findings of the present study were shown as below:1.Trends of gastric cancer incidence and mortality from 2012 to 2019 in LinquA total of 4731 new GC incident cases were documented and 3087 cases died from GC between 2012 and 2019.The average crude incidence and mortality of GC were 65.48/100,000 and 42.13/100,000,respectively.Regression model revealed statistically significant decreasing incidence(APC of crude incidence=-4.5%)and mortality(APC of crude mortality=-3.4%)from 2012 to 2019,in accordance with the trends of standardized rate.Analyses by anatomical location found that the crude incidence and mortality of non-cardia GC decreased by 5.1%and 3.7%,respectively,but the incidence and mortality of cardia GC remained relative stable,even with slightly increase.2.Effects of endoscopic screening on GCEver receiving endoscopic screening significantly decreased by 31%incidence of(HR=0.69,95%confidence interval(CI):0.52-0.92)and 66%mortality from invasive GC(HR=0.34,95%CI:0.20-0.58),particularly for non-cardia invasive GC.Repeated screening strengthened the beneficial effect on invasive GC mortality contributed by one-time screening(one-time HR=0.38,95%CI:0.21-0.69;repeated HR=0.25,95%CI:0.08-0.77).Screening-detected invasive GCs had significantly better overall survival compared with cases in the unscreened group(HR=0.22,95%Cl:0.16-0.30),particularly for those receiving repeated endoscopy(5-year survival rate 90.2%vs.72.5%for one-time screened vs.29.2%for unscreened cases).After stratified by stage,the protective effect of endoscopic screening remained.3.The exploration of screening strategy1)Screening starting ageBased on data of cancer registry,the age-specific incidence for GC showed a significant uptrend as age increased after 40 years old.The age turning points of GC incidence were 40 years for males and 45 years for females while the age turning points of mortality were 50 years for males and 55 years for females.Based on prospective study of UGCED program,the association of invasive GC incidence and mortality with endoscopic screening was statistically significant for subjects aged 50-59 and 60-69 years,but not for those younger than 50 years(P-heterogeneity 0.005 for incidence and<0.001 for mortality).2)Screening intervala.The Association between gastric lesions and risk of developing GC:Compared with subjects with superficial gastritis(SG)or normal mucosa,the HR(95%CI)for progression to GC overall during the follow-up was 3.91(2.21-6.92)for chronic atrophic gastritis(CAG),4.60(2.60-8.14)for intestinal metaplasia(IM),and 17.06(9.28-31.35)for low-grade intraepithelial neoplasia(LGIN).Separate analyses for high-grade intraepithelial neoplasia and invasive GC also reported similar results.b.The screening interval associated with the odds of EGC detection:For endoscopy-detected GC cases that were diagnosed with IM or LGIN at the preceding endoscopy,having the following endoscopy within one to two years(OR=14.92,95%CI:1.83-121.60)or within one year(OR=19.88,95%CI:1.59-248.39)significantly improved the odds of detecting EGC,compared with repeated screening after two years(P-trend=0.02).No such significant findings were found for those with SG or CAG at the preceding endoscopy.c.Multistate Markov model for analyzing the nature history of gastric lesions:Based on repeated gastroscopy cohort in 1989,the mean sojourn time a subject spent in SG/CAG was 9.89 years(95%CI:8.87-11.02)and the average time a subject spent in IM and LGIN were 4.01 years(95%CI:3.43-4.69)and 2.19 years(95%CI:1.78-2.69),respectively estimated by Markov model.Compared with SG/CAG and IM,a subject with LGIN had the highest transition probability to GC and increased with time.The 1-year,3-year and 5-year transition probability were 0.013,0.029 and 0.039,respectively.In conclusion,based on the data from different sources in high-risk area of GC,we found that endoscopic screening prevented GC occurrence and death and improved its prognosis.Repeated endoscopy enhanced the effectiveness,for which the screening interval needs to be defined in conformity with the severity of gastric lesions.For subjects with IM/LGIN,repeated screening within two years is warranted,and within one year would be highly recommended when medical and health resources are enough.Interval for IM could be extended according to mean sojourn time.An interval of more than two years may be acceptable for those with SG/CAG.In addition,the existing screening starting age of 40 years old could be appropriately increased according to the results of age turning points and endoscopic screening effect.Our study based on prospective design may inform policy making on extending the coverage of endoscopy screening and designing suitable screening age and interval in high-risk areas of GC in China.