Objective:The purpose of this study was to compare the predictive value of different prognostic scoring systems on survival and molecular response in patients with newly diagnosed chronic myeloid leukemia treated with tyrosine kinase inhibitor.Method:Clinical data of 321 newly diagnosed chronic myeloid leukemia-chronic phase patients treated with tyrosine kinase inhibitor was analyzed in our hospital from January 2012 to December 2021.The clinical data were stratified according to the Sokal score,Hasford score,EUTOS score,ELTS score,and the sores of predictive major molecular response(MMR)and molecular response4(MR4).To evaluate the relationship between the Sokal score,Hasford score,EUTOS score,ELTS score,and progression-free survival(PFS)、overall survival(OS).To evaluate the relationship between the scoring systems and major molecular response and deep molecular response(DMR);and to evaluate the predictive value of BCR-ABL1 halving time and3-month early molecular response(3M-EMR)in predicting 12-month major molecular response(12M-MMR)and molecular reactions 4.5(MR4.5).Results:1.Risk stratification:According to the Sokal score,161(50.2%),109(34.0%),and 51(15.9%)patients were divided into the low,medium,and high-risk group;according to the Hasford score,155(48.3%),141(43.9%)and 25(7.8%)were divided into the low,medium and high-risk group;according to EUTOS score,257(80.1%)and 64(19.9%)were divided into the low-risk group and high-risk group;according to the ELTS score,221(68.8%),78(24.3%),and 22(6.9%)were divided into the low,medium,and high-risk group,respectively.According to the score of the predicted MMR,99(30.8%),174(54.2%,)and 48(15.0%)were divided into the low,medium,and high-risk group;according to the score of the predicted MR4,39(12.1%),184(57.3%)and 98(30.5%)were divided into the low,medium and high-risk group.2.Survival analysis:PFS: According to the Sokal score,the 5-year PFS rates of patients in the low,medium and high-risk group were 94.4%,93.2% and 75.5% respectively(P=0.004).According to the EUTOS score,the 5-year PFS rates of the low-risk group and the high-risk group were 94.3% and 85.5% respectively(P=0.012).Sokal and EUTOS scores can effectively distinguish the PFS rates of different risk groups.However,Hasford and ELTS scores could not effectively distinguish the PFS rates of different risk groups(P>0.05).OS: According to the Sokal score,the 5-year OS rates of patients in the low,medium and high-risk group were 97.0%,95.2% and 81.7% respectively(P<0.001).According to the Hasford score,the 5-year OS rates of the low,medium and high-risk group were 97.2%,82.7% and 82.4% respectively(P=0.039).According to the EUTOS score,the 5-year OS rates of patients in the low-risk group and the high-risk group were 97.1% and 88.9% respectively(P=0.018).Except for ELTS score,Sokal,Hasford and EUTOS could significantly distinguish the high-risk group with high risk of death(P<0.05).3.Molecular response:12M-MMR: The scoring system of ELTS score,predicting MMR and predicting MR4 can significantly predict the 12M-MMR of different risk groups,in which the scoring system of predicting MR4 is more accurate,and the 12M-MMR rates of low,medium and high-risk group are 77.4%,54.5% and 35.1% respectively(P<0.001).However,Sokal,Hasford and EUTOS scoring system could not predict the12M-MMR of different risk groups(P>0.05).MMR: The six scoring systems of Sokal,Hasford,EUTOS,ELTS,predictive MMR and predictive MR4 can predict the acquisition rate of MMR at any time,and the accuracy of ELTS score,predictive MMR and predictive MR4 scoring system is higher(P<0.001).According to the ELTS score,the MMR acquisition rates of the low,medium and high-risk group were 84.0%,62.1% and 35.7%,respectively;according to the predicted MMR score,the MMR acquisition rates of the low,medium and high-risk group were 91.0%,72.9% and 53.8%,respectively;according to the predicted MR4 score,the MMR acquisition rates of the low,medium and high-risk group were 88.9%,83.7% and 55.7%,respectively.DMR(MR4 and MR4.5): The ELTS score,predictive MMR and predictive MR4 scoring system can all predict the DMR acquisition rate at any time,in which the predictive MR4 scoring system is more accurate in predicting DMR.The MR4 acquisition rates of the low,medium and high-risk group are 85.7%,52.0% and 24.1%respectively(P<0.001),and the MR4.5 acquisition rates are 60.0%,43.4% and 19.0%(P<0.001),respectively.4、Using ROC curve,the correlation between BCR-ABL1 halving time and3-month BCR-ABL1 transcription level in predicting 12M-MMR was the same(AUC=0.804).The best cut-off value for finding the half time of BCR-ABL1 on ROC curve by using Youden index is 17 days,that is,the patients whose BCR-ABL1 half time is less than 17 days have higher 12M-MMR acquisition rate.Of the patients whose BCR-ABL1 halved less than 17 days,80% reached MMR at 12 month(P<0.001)and 56.3% reached MR4.5(P=0.005).Among the patients who achieved the best response in 3 month(3-month BCR-ABL1≤10%,3M-EMR),63.8% of the patients reached MMR in 12 month(P<0.001),and 45.1% of the patients reached MR4.5(P=0.063,no statistical significance).Conclusion:1.ELTS score could not significantly distinguish between PFS and OS rates in different risk groups,while Sokal score could effectively distinguish high-risk group with high disease progression and high risk of death(P<0.05),compared with Hasford,EUTOS and ELTS score,Sokal score still have advantages in evaluating the survival and prognosis of chronic myeloid leukemia-chronic phase patients.2.The ELTS score,predictive MMR and predictive MR4 scoring system can all predict the acquisition rate of 12M-MMR,MMR and DMR at any time,in which the predictive MR4 scoring system can predict the molecular response with higher accuracy,while the Sokal,Hasford and EUTOS scoring system can only predict the acquisition rate of MMR at any time in different risk groups(P<0.05).3.The half time of BCR-ABL1 less than 17 days can predict the acquisition rate of 12M-MMR and MR4.5,while 3M-EMR can only predict the acquisition rate of12M-MMR,but can not predict the rate of MR4.5.The sample size included in this study is relatively small,which still needs to be further verified by follow-up clinical studies with large sample size. |