| Liver cancer is one of the most common malignant tumors in China,accounting for 45.3% and 47.1% of the world’s liver cancer new cases and death,which is characterized by occult onset,late diagnosis,with 12.1% of 5-year survival rate in our country.Therefore,the prevention and control situation of liver cancer in China is more severe than that in other countries.In order to improve the prognosis of patients with liver cancer and reduce the burden of liver cancer in Chinese population,early screening,early diagnosis and early treatment of high-risk population can improve the prognosis and prolong survival of patients with liver cancer on the whole.According to China’s national conditions,experts in liver cancer screening have proposed a "pyramid" screening pattern," stratified enrichment,precise screening and full consideration of health economics".However,many problems remain unsolved,including how to implement risk stratification in population at risk of liver cancer,what screening protocol should be adopted after risk stratification to achieve precise screening,and how to effectively control screening costs during risk stratification and screening implementation.Objectives:As for population at risk of liver cancer,the long-term prediction efficacy of aMAP score on liver cancer incidence and death will be investigated,and a risk stratification threshold suitable for community population will be established,in order to provide a stratified management basis for liver cancer screening.In terms of liver cancer screening protocol,the screening benefit of adding AFP-L3% and DCP cancer biomarker detection on the basis of traditional screening protocols(AFP combined with liver ultrasound)would be examined to further improve the detection ability of liver cancer,especially early liver cancer.Focusing on liver cancer screening strategies,our study is intended to explore suitable liver cancer screening protocol based on risk stratification of aMAP score,so as to explore how to achieve accurate screening under the premise of controlling screening costs effectively.Methods:1.The first part of this study recruited 1,503 patients with hepatitis B and C infection from Fuyu City,Jilin Province between 2011 and 2012 to conduct a ten-year community-based cohort study.At baseline,all participants underwent questionnaire survey,quantitative detection of hepatitis B and hepatitis C virus,blood routine and liver function test,and abdominal ultrasound.Regular active and passive follow-up was conducted to determine whether the subjects had liver cancer,the time of diagnosis,whether they died,the time of death and the cause of death.Baseline aMAP score was calculated using age,sex,total bilirubin,albumin,and platelet at enrollment.Timedependent ROC curve and calibration curve were plotted to evaluate the differentiation and calibration of baseline aMAP score in predicting liver cancer incidence and death,and then to evaluate its applicability to the community population.The relationship between baseline aMAP score and liver cancer incidence and death was investigated using the restricted cubic spline based on the Fine-Gray competing risk model.The cumulative incidence or death curve was drawn and the incidence and mortality of liver cancer were calculated,then the risk stratification threshold of aMAP score was explored according to absolute risk.The multi-state models with competing risks was used to calculate the hazard ratio of liver cancer death to death from other causes in different aMAP score subgroups,and the risk stratification threshold of aMAP score was explored based on the relative risk.After the optimal stratification threshold of aMAP score was obtained by combining absolute and relative risks,the sensitivity,specificity,positive predictive value and negative predictive value were calculated to evaluate the accuracy of aMAP score at this threshold in predicting the incidence and death of liver cancer at 5 and 10 years.2.The second part of this study enrolled 2,610 apparently healthy participants with chronic liver disease in selected areas of Jilin Province between 2019 and 2020 to conduct a community-based prospective screening study that included baseline and 1-year follow-up.Basic information collection,detection of AFP,AFP-L3% and DCP,abdominal ultrasonography and other standard tests were performed at baseline screening.All participants with negative baseline screening received identical followup screening 1 year after baseline screening,and individuals who tested positive at any one screening were referred for confirmatory testing.After the participants were diagnosed with liver cancer,the patients subsequently received treatment based on clinical practice guidelines.Otherwise,the participants were defined as non-liver cancer if they were screened negative at baseline and follow-up,or if they were screened positive but had negative confirmatory tests.To evaluate the benefit of screening,screening positive rate,detection rate,early diagnosis rate and effective treatment rate were calculated.The improvement level of screening model after adding detection of AFP-L3% and DCP was presented by continuous net reclassification improvement index and integrated discrimination improvement index.3.The third part of this study based on the community-based prospective screening study in selected areas of Jilin Province,the risk stratification threshold of aMAP score defined in the first part was attempted to implement risk stratification for population at risk of liver cancer,and the screening efficacy of the innovative screening protocol adding AFP-L3% and DCP detection in the second part was compared with that of the traditional screening protocol in different risk stratification.De Long test was used to compare the difference between the ROC curve of innovative screening protocol and traditional screening protocol,and categorical net reclassification improvement index was applied to show the improvement degree of screening value of innovative screening protocol compared with traditional screening protocol.Finally,the sensitivity and specificity of screening protocol and the referral rate,detection rate and early diagnosis rate related to the screening population were calculated to explore the improvement of screening efficacy caused by performing different intensity screening protocol after stratification by aMAP score.Results:(I)The stratification effect of aMAP score on the risk of liver cancer incidence and death in community population1.Evaluation of the accuracy aMAP score in predicting the incidence and death of liver cancer,and judgement of its applicability to community populationIn terms of differentiation,the area under time-dependent ROC curve of aMAP score for predicting the incidence of liver cancer was relatively high,which was stable at about 0.80 in 10 years.Moreover,the aMAP score also performed well in predicting liver cancer death,with the optimal area under the curve of 0.85 over the 10-year period.In terms of calibration,the probability of liver cancer incidence or death within 5 years was in good agreement with the predicted probability.From the 6th year onwards,the X predicted probability was slightly higher than the actual probability for both liver cancer incidence and death.2.Exploration of the high-risk threshold of the aMAP score by the absolute risk of liver cancer incidence and deathThe "J" shaped curve showed that the risk of liver cancer incidence and death was stable when aMAP score was <55,but increased rapidly when aMAP score was ≥55.Above 55,each 1 score increase was associated with a 16% [SHR=1.16(1.11-1.21)]increased risk of liver cancer incidence and a 17% [SHR=1.17(1.13-1.23)] increased risk of liver cancer death.After the two groups were divided by aMAP score of 55,the incidence of liver cancer(5-year: 8.3%,10-year: 20.9%)and mortality(5-year: 6.7%,10-year: 17.5%)in aMAP≥55 group were significantly higher than incidence(5-year:1.3%,10-year: 3.6%)and mortality(5-year: 1.1%,10-year: 3.1%)in aMAP<55 group(Gray’s test P<0.001).In terms of absolute risk,the annual incidence of liver cancer in population with an aMAP score of ≥55 had exceeded the threshold of 1.5% for a more cost-effective liver cancer screening,while those with an aMAP score of <55 had not.3.Exploration of the high-risk threshold of aMAP score by the relative risk of death from liver cancer and other causesOur study found no statistically significant differences in the risk of death from liver cancer and other causes in total population [HR=1.05(0.82-1.34)].However,after exploring the aMAP score range of 50-60,it was found that the risk of liver cancer death began to exceed that of death from other causes in population with aMAP score≥55 [HR=1.38(1.02-1.87)],while the risk of death from liver cancer was lower than that from other causes in the population with aMAP<55 [HR=0.56(0.35-0.89)].Thus,55 was the risk stratification threshold of aMAP score based on the relative risk of death from liver cancer and other causes.(II)Exploration the improvement of adding the detection of cancer biomarker for liver cancer based on the screening modelWhile the traditional liver cancer screening model constructed by AFP combined with ultrasound was added by AFP-L3%,the liver cancer screening value was improved[NRI:0.217(0.010-0.425),IDI:0.046(0.009-0.083)],the specificity increased by 8.7%,but the sensitivity decreased slightly.After adding DCP,the new screening model was positively improved [NRI:0.413(0.255-0.571),IDI:0.051(0.009-0.093)],sensitivity was increased by 3.4%,while specificity remained flat.With the addition of AFP-L3%and DCP simultaneously,the positive improvement of the screening model was further enhanced [NRI:0.573(0.364-0.782),IDI:0.098(0.045-0.150)],the specificity increased to 95.3% while the sensitivity remained unchanged.Finally,the screening model constructed by GALADUS score,which was calculated by age and sex in addition to 3cancer biomarkers combined with ultrasound,achieved the best degree of positive improvement [NRI:1.105(0.912-1.298),IDI:0.255(0.199-0.311)],with the significant increase of AUC(De Long P<0.001)and sensitivity increased by 10.1% accompanied by specificity increased by 8.8%.(III)The improvement of liver cancer screening efficacy performing different screening protocols after the population risk stratification1.Comparison of liver cancer screening efficacy of various screening protocols in different populations stratified by aMAP score of 55Considering the operability of liver cancer screening,AFP was divided into categorical variables and then combined with US for screening(AFP/US protocol),and GALADUS was classified into two categories with a boundary of-0.7(GALADUS protocol).After stratified exploration,GALADUS protocol in population with aMAP<55 had the same detection rate(0.5% vs.0.5%,P=1.000)and early diagnosis rate(100% vs.100%)as traditional protocol AFP/US,while GALADUS protocol in aMAP≥55 population greatly improved the detection rate(5.6% vs.4.3%,P<0.001)and early diagnosis rate(61.3% vs.54.4%)compared with AFP/US protocol.2.Screening efficacy of different intensity screening protocol were performed by stratification with aMAP score of 55The "aMAP triage" protocol,performing GALADUS protocol for population with aMAP≥55 and AFP/US protocol for population with aMAP<55,obtained a significant promotion in the detection rate(3.1% vs.2.4%,P<0.001)and early diagnosis rate(64.2%vs.58.7%)relative to AFP/US protocol,which was more conducive to the detection of liver cancer,especially early liver cancer.In addition,the “aMAP triage” protocol was comparable to GALADUS protocol in the detection rate of liver cancer(3.1% vs.3.1%,P=1.000)and early diagnosis rate(64.2% vs.64.2%)in the total population,which saved the cost of liver cancer screening and was more scalable than GALADUS protocol.Conclusions:1.aMAP score has high accuracy in predicting the incidence and death of liver cancer in patients with hepatitis B and C,thus it has good applicability to community population.On the premise of long-term follow-up and considering competing risks,the aMAP score risk stratification threshold was set at 55 by combining absolute and relative risks,so as to guide the follow-up practice of liver cancer screening in the community population.2.Adding cancer biomarker AFP-L3% or/and DCP detection can improve the screening value on the basis of the traditional screening model constructed by AFP combined with ultrasound.The GALADUS score calculated by the additional combination of age and gender can further improve the screening efficacy.3.The "aMAP triage" protocol,performing GALADUS protocol for population with aMAP≥55 and AFP/US protocol for population with aMAP<55,obtained a significant promotion in the detection rate and early diagnosis rate relative to the traditional screening protocol,while it effectively saved the cost of liver cancer compared with population-wide GALADUS protocol. |
PDF Full Text Request