The Value Of A Novel Systemic Inflammation Index In The Prognosis Of Cervical Cancer

Objective: The purpose of this study was to investigate and compare the value of SII,NLR,PLR,LMR,and other hematological inflammation markers(neutrophils,lymphocytes,platelets,etc.)in the evaluation of prognosis of cervical cancer,and to establish the predictive model of cervical cancer death risk relying on inflammatory markers based on hematology.Methods: This retrospective study included a total of 211 cervical cancer patients undergoing radical chemoradiotherapy who were treated at Sichuan Cancer Hospital from January 2009 to April 2016.Using the receiver operating curve(ROC)to determine the optimal cut-off value for index based on inflammation,the patients were divided into two groups based on the optimal cut-off values for each index,the differences in clinical characteristics and other hematological parameters were compared between groups.Kaplan-Meier univariate analysis was used to compare overall survival(OS),local recurrence-free survival(LRFS),distant metastasis-free survival(DMFS),and progression-free survival(PFS)in different groups.Using the multivariate analysis to determinethe independent prognostic factors of cervical cancer.The area under the curve(AUC)was used to compare the ability of these independent prognostic factors to differentiate between OS,LRFS,DMFS,and PFS.The propensity score matching(PSM)of the case data was completed through the empowerstats software(www.empowerstats.com,X&Y solutions,Inc.Boston MA).Finally,basing on the results of multivariate analysis,Using the independent prognostic factors of cervical cancer,a predictive model for the risk of cervical cancer death from inflammation markers based on hematology was established.Results: The best cut-off value for SII was 427.822,NLR was 2.449,PLR was 95.306,LMR was 0.208,WBC was 7.05×109/L.The levels of NLR,PLR,and LMR in the high SII group were significantly higher than those in the low SII(NLR: 3.16±1.21 vs.1.89±0.61,p<0.001;PLR: 177.16±65.65 vs.96.57±28.76,p<0.001;LMR: 0.25 ± 0.10 vs.0.18±0.06,p<0.001).In addition,WBC,GR#,and PLT levels were obviously higher in the high SII group than in the low SII group(WBC: 6.14±1.44 vs.5.42±1.12,p<0.001;GR#: 4.17±1.22 vs.3.16±0.77,p<0.001;PLT : 233.38±62.01 vs.163.85±44.34,p<0.001),the levels of LYMPH,RBC,HGB,MCV,MCH,and MPV in the high SII group were significantly lower than those in the low SII group(LYMPH: 1.40±0.38 vs.1.77±0.52,p<0.001;RBC: 3.98±0.57 vs.4.17±0.45,p=0.007;HGB: 113.99±20.88 vs.125.21±12.88,p<0.001;MCV: 89.97±7.80 vs.93.35±5.37,p<0.001;MCH: 28.63±3.19 vs.30.08±1.95,p<0.001;MPV: 10.73±1.66 vs.11.97±1.41,p<0.001).The survival analysis shows that the survival rate of the high SII group was significantly lower than that of the low SII group.The 1-,3-,and 5-year OS were(94% vs.100%;79% vs.96%;65% vs 90%,respectively,p<0.001).The 1-,3-,and 5-year OS of the high NLR and low NLR groups were(93% vs.100%;82% vs.93%;63% vs.88%,respectively,p<0.001),the 1-,3-,and 5-year OS in the high and low PLR groups were(96% vs.100%;85% vs.96%;71% vs.91%,p=0.004).The 1-,3-,and 5-year OS of the high-and low-LMR groups were(99% vs.95%;93% vs.83%;86% vs.68%,p=0.004).Univariate analysis showed that age,stage,SII,NLR,PLR,LMR,WBC,GR#,LYMPH,MONO,HGB,MCV,MCH were risk factors for OS.Multivariate analysis showed that SII,PLR,WBC,and stage were independent prognostic factors of OS.The AUC of SII was 0.728(95%CI: 0.639-0.816,p<0.001),and the AUC of PLR was 0.621(95% CI: 0.526-0.717,p=0.031).The AUC of WBC was 0.684(95%CI: 0.569-0.798,p=0.001).The AUC of stage was 0.591(95%CI: 0.486-0.697,p=0.104),The AUC of each indicator was compared on a pairwise basis and it was found that SII was better than PLR and stage(AUC differences were: 0.104 and 0.133,z=2.812 and 2.515,p=0.0049 and 0.0119,respectively).Then we establish a predictive model for the risk of cervical cancer death based on relevant inflammatory markers,prediction scores=-5.01798+0.00146*SII-0.00066*PLR+0.36564*WBC+0.47067*stage(stage: ≥3=2,<3=1).The predicted value of the nomogram was evaluated by the ROC curve.The AUC was 0.715,95% CI: 0.610-0.819,and the best cut-off score was-1.718.The sensitivity was 61.3%,the specificity was 72.8%,and the accuracy was 71.1.%,the diagnostic odds ratio was 4.233.Since this study is a retrospective one without randomized controls,there may be differences in patient baseline.Therefore,we use the PSM method to make the baseline data of the cohort similar.A total of 75 pairs of baseline data matching after matched by the empowerstats software.Again,we compared the inflammatory index between the 150 patients and the results were almost the same as before.The levels of NLR,PLR and LMR in the high SII group were significantly higher than those in the low SII group(p<0.001;p<0.001;p<0.001).In addition,WBC,GR#,and PLT levels in the high-SII group were obviously higher than those in the low-SII group(p<0.001;p<0.001;p<0.001).The levels of LYMPH,HGB,MCV,MCH,and MPV were significantly lower in the high-SII group than those in the low-SII group(p<0.001;p=0.006;p=0.032;p=0.016;p<0.001).The survival rates between groups were compared after matching,and the results were consistent with those before matching.The 1-,3-,and 5-year OS of the high SII and low SII groups were(95% vs.100%;78% vs.98%;58% vs.91%,respectively,p<0.001).The 1-,3-,and 5-year OS of the high NLR and low NLR groups were(94% vs.100%;81% vs.92%;58% vs.86%,respectively,p=0.001),the 1-,3-,and 5-year OS of the high and low PLR groups were(96% vs.100%;84% vs.97%;66% vs.97%,p=0.005).And the 1-,3-,and 5-year OS of the high-and low-LMR groups were(100% vs.94%;93% vs.82%;82% vs.66%,p=0.013).Post-matching univariate analysis showed that stage,SII,NLR,PLR,LMR,WBC,GR#,LYMPH,MONO,HCT,MCV,MCH,and PLT were risk factors for OS.Multivariate analysis showed that SII,WBC,LYMPH,and stage were independent prognostic factors for OS in cervical cancer patients.Compared with WBC,LYMPH,and stage,SII had a higher area under the curve.After comparing each AUC,it was found that SII was superior to stage(AUC difference: 0.104,z=1.658,p=0.0373).After matching,predictive models were again established based on independent prognostic factors of cervical cancer OS.Prediction scores=-4.53428 +0.00071*SII +0.52112*WBC-0.71099*LYMPH+0.50025*(staging: ≥3 = 2,<3 = 1),Comparing the pre-and post-matching prediction models,the results showed that there was no significant difference between the two models in predicting the risk of cervical cancer death,p=0.2665.Conclusion:Systemic inflammation indicators can predict the prognosis of cervical cancer.SII,PLR,and stage were independent prognostic factors of OS in cervical cancer.For the predictive ability of cervical cancer OS,SII is superior to PLR and stage.The nomogram can be used to predict the risk of death from cervical cancer patients.However,more large samples,multicenter,prospective,randomized controlled clinical trials are needed to verify this result in the future.