| Cervical cancer is the fourth most common cancer in women and the main cause of cancer death in women.Although uterine corpus invasion(UCI)is not considered in the FIGO staging of cervical cancer,it can still be regarded as the spread of tumor to adjacent organs and tissues.Previous studies reported that the incidence of UCI is 4.926.2%,with great difference.The results of the study on the prognosis of UCI and cervical cancer patients are different.In addition,these previous studies also has deficiencie,including less sample size and research variables,no clear diagnostic criteria for UCI and pathological review,and it’s not clear whether the depth of UCI is related to the prognosis of cervical cancer patients.[Objective]To investigate the UCI of cervical cancer patients based on the clinical diagnosis and treatment database of cervical cancer in China(1538 project database)and the method of postoperative pathological review of uterine tissue,and to identify the risk for UCI and the relationship between UCI and the prognosis of cervical cancer patients,and to determine the risk coefficient of UCI and the effect of postoperative adjuvantradiotherapy on the oncological outcomes of cervical cancer patients with UCI.It would provide a theoretical basis for the staging of cervical cancer[Materials and methods]This study is based on the 1538 project database,which continuously collected 46313 inpatient cases of cervical cancer from January 1,2004 to December 31,2016,including 33758 cases of surgical treatment and 12555 cases of non-surgical treatment.The participants included 37 medical institutions in some parts of mainland China.From the pathology departments of 11 hospitals approved to review the pathological results,the cases of original pathological report with UCI were selected as the positive experimental group,and the cases of original pathological report with non-UCI were selected as the negative control group in the proportion of 1:3.These selected cases were received a review of uterine pathology.Finally,2212 cases were obtained.The independent risk factors for UCI and the depth were identified by binary and ordinal logistic regression model.Kaplan-Meier curve was used to describe the change of survival,and log-rank test was used to compare the difference of survival curve.The independent risk factors for the oncological prognosis of patients were determined by Cox regression model.In addition,1:n propensity score matching(PSM)and analysis in different levels of subgroups were used to verify the stability of the main results.[Results]Chapterl:The investigation of UCI of cervical cancer in parts of mainland China1.There were 33758 cases with operation in the 1538 project database.After 151 cases with simple cervical resection and 2922 cases of unknown data were excluded,there were 30685 cases wih hysterectomy.The original pathological report indicated 1598 cases had UCI,Hence he original detection rate of UCI was 5.2%(1598/30685)..2.The database of 2212 cases with a review of uterine pathology include 497 cases with original pathological report of UCI(461 cases with UCI and 36 cases with nonUCI confirmed after pathological-review)and 1715 cases with original pathological report of non-UCI(1661 cases with non-UCI and 54 cases with UCI confirmed after pathological review).Then,after pathological review,515 cases were confirmed as UCI,1697 cases with non-UCI.The overall coincidence rate of a review of uterine pathology was 95.9%(2122/2212);the positive coincidence rate was 89.5%(461/515);the false negative rate(missed diagnosis rate)was 10.5%(54/515);the negative coincidence rate was 97.9%(1661/1697);and the false positive rate(misdiagnosis rate)was 2.1%(36/1697).3.In different cities,the coincidence rate of pathological review of the third-tier and above cities was slightly higher than that of the second-tier cities,but no significant difference was found between the two groups.The coincidence rate of pathological review in different areas was different,and the highest in South China,followed by Central China,North China,East China and Southwest China.Significant difference was found between the groups.Among the three types of hospitals,the coincidence rate of a review of uterine pathology was the highest in women and children’s hospital and the lowest in general hospital,but no significant difference was found between the groups.4.The coincidence rate of a review of uterine pathology fluctuated slightly from 2004 to 2016.The highest coincidence rate was 100.0%in 2004 and 91.4%in 2005.As the year wen on,the tendency of coincidence rate was stable.5.Among 2212 cases with a review of uterine pathology,2122 cases were in accordance with the original pathological report and 90 cases were not in accordance with the original pathological report.According to this,they were divided into two groups.No significant difference was found between the two groups in age,FIGO stage,tumor diameter,histological type,histological grade,preoperative adjuvant treatment,surgical approach,stromal invasion,lymph vascular space invasion(LVSI),parametrial involvement,resection margin involvement and lymph node metastasis.Chapter 2:Analysis of risk factors for cervical cancer patients with UCIFrom the database of 2212 cases with a review of uterine pathology,multivariate logistic regression analysis showed that the older,latter FIGO stage,tumor diameter,adenocarcinoma,paremetrial involvement and lymph node metastasis were all independent risk factors for UCI,and ordinal logistic regression analysis showed that the above risk factors were related to the depth of UCI.The analysis results of the subgroup without preoperative adjuvant therapy were consistent with the main results.Chapter 3:The relationship between the UCI and the oncological prognosis in cervical cancer patients1.First level:Prognosis analysis of the initially enrolled cases:From the database of 2212 cases with a review of uterine pathology,1673 cervical cancer patients withⅠA2-ⅡB stage were enrolled for inclusion and exclusion criteria,with a median followup time of 42 months.There were 338 cases in the UCI group(104 cases of endometrial invasion,1 17 cases of myometrial invasion<1/2 and 117 cases of myometrial invasion≥1/2),and 1335 cases in the non-UCI group.Before matching,patients in the UCI group had a significantly lower 5-year OS and DFS than those in the non-UCI group(OS:74.1%vs.84.7%P<0.001;DFS:67.3%vs.80.7%,P<0.001).On multivariate Cox regression analysis,uterine myometrial invasion≥1/2 was an independent risk factor for 5-year OS and DFS(OS:HR=2.385,95%CI 1.656-3.433,P<0.001;DFS:HR=2.000,95%CI 1.426-2.803,P<0.001).The uterine myometrial invasion<1/2 was an independent risk factor for 5-year DFS(DFS:HR=1.636,95%CI 1.104-2.424,P=0.014),but not for 5-year OS(P=0.176).Uterine endometrial invasion was not an independent risk factor for 5-year OS and DFS(OS:P=0.505;DFS:P=0.396).927 cases were included after 1:2 PSM,including 323 cases in the UCI group and 604 cases in the non-UCI group.Patients in the UCI group had a significantly lower 5-year OS and DFS than those in the non-UCI group(OS:74.2%vs.83.2%,P=0.006;DFS:67.9%vs.77.2%,P=0.002).On multivariate Cox regression analysis,uterine myometrial invasion≥1/2 was an independent prognosis factor for 5-year OS and DFS(OS:HR=2.436,95%CI 1.639-3.620,P<0.001;DFS:HR=2.192,95%CI 1.537-3.126,P<0.001).However,the uterine myometrial invasion<1/2 was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.245;DFS:P=0.068),and uterine endometrial invasion was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.391;DFS:P=0.384).2.Second level:Prognosis analysis of the strictly enrolled cases:Based on 1673 cervical cancer patients with initial inclusion criteria,we further selected 1414 cases without preoperative adjuvant treatment,including 284 cases in the UCI group(90 cases of endometrial invasion,105 cases of uterine myometrial invasion<1/2 and 89 cases of uterine myometrial invasion≥1/2),and 1130 cases in the non-UCI group.Before matching,patients in the UCI group had a significantly lower 5-year OS and DFS than those in the non-UCI group(OS:77.1%vs.85.1%,P=0.003;DFS:69.3%vs.81.5%,P<0.001).On multivariate Cox regression analysis,uterine myometrial invasion≥1/2 was an independent risk factor for 5-year OS and DFS(OS:HR=2.736,95%CI 1.813-4.130,P<0.001;DFS:HR=2.307,95%CI 1.588-3.351,P<0.001).However,uterine myometrial invasion<1/2 was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.818;DFS:P=0.209),and uterine endometrial invasion was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.450;DFS:P=0.509).After 1:2 PSM,803 cases were included,including 276 cases in the UCI group and 527 cases in the non-UCI group.No significant difference was found between the two groups for 5-years OS(77.1%vs.82.3%,P=0.155),but patients in the UCI group had a significantly lower 5-year DFS than those in the nonUCI group(69.1%vs.78.0%,P=0.015).On multivariate Cox regression analysis,uterine myometrial invasion≥1/2 was an independent prognosis factor for 5-year OS and DFS(OS:HR=2.331,95%CI 1.502-3.618,P<0.001;DFS:HR=2.387,95%CI 1.609-3.541,P<0.001).However,uterine myometrial invasion<1/2 was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.810;DFS:P=0.335),and uterine endometrial invasion was not an independent risk factor for 5-year OS and DFS(OS:P=0.327;DFS:P=0.541).Prognosis analysis of subgroups:to determine whether the influence of uterine infiltration on prognosis is stable in different subgroups,the following four subgroups were further selected in 1414 patients without preoperative adjuvant treatment:(1)748 cases with stage IA2-IB;(2)639 cases with stage ⅡA;(3)999 cases without lymph node metastasis;(4)415 cases with lymph node metastasis.The prognosis results in the subgroups revealed that uterine myometrial invasion≥1/2 was an independent prognosis factor for 5-year OS and DFS.The uterine myometrial invasion<1/2 was an independent prognosis factor for 5-year DFS only in the subgroup of ⅠA2-ⅠB stage and without lymph node metastasis,but didn’t have a relationship with 5-year OS.The uterine endometrial invasion was not an independent prognosis factor for 5-year OS and DFS.3.Third level:Prognosis analysis of postoperative standard adjuvant treatment:Based on 1414 cervical cancer patients with strictly enrolled cases,we further selected 769 cases with postoperative standard adjuvant treatment,including 152 cases in the UCI group and 617 cases in the non-UCI group.Before matching,patients in the UCI group had a significantly lower 5-year DFS than those in the non-UCI group(66.2%vs.80.4%,P=0.003),but no significant difference was found between the two groups for 5-years OS(77.3%vs.84.6%,P=0.078),On multivariate Cox regression analysis,uterine myometrial invasion≥1/2 was an independent risk factor for 5-year DFS(HR=2.230,95%CI 1.270-3.913,P=0.005),but not an independent risk factor for 5year OS(P=0.058).However,uterine myometrial invasion<1/2 was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.058;DFS:P=0.664),and uterine endometrial invasion was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.058;DFS:P=0.248).Chapter 4:Assessment of risk coefficient for cervical cancer with UCI1.According to multivariate Cox regression model of the initially enrolled cases,the risk coefficients of UCI for 5-year OS and DFS were obtained.Five-year OS:Cox regression equation:h(t,X)=h0(t)exp(0.481 × preoperative adjuvant treatment+0.869 × uterine myometrial invasion>1/2+1.39 1 × stromal invasion depth>1/2+0.871 × lymph node metastasis).Compared with other postoperative pathological risk factors,the risk coefficient of the uterine myometrial invasion≥ 1/2 was 0.869,lower than that of stromal invasion depth>1/2(1.391),slightly lower than that of lymph node metastasis(0.871).Five-year DFS:Cox regression equation:h(t,X)=h0(t)exp(-0.364 × age 50-59 years-0.430 × year at diagnosis 2009-2016+0.596 × stage ⅡB+0.504 × tumor diameter>4cm +0.331 × laparoscopic surgery+0.492 × uterine myometrial invasion<1/2+0.693 × uterine myometrial invasion≥1/2+1.135 × stromal invasion depth>1/2+0.271 × LVSI+0.728 × lymph node metastasis).Compared with other postoperative pathological risk factors,the risk coefficient of the uterine myometrial invasion≥1/2 was 0.693,lower than that of stromal invasion depth>1/2(1.135),slightly lower than lymph node metastasis(0.728),but higher than tumor diameter>4cm(0.504)and LVSI(0.271).While the risk coefficient of uterine myometrial invasion<1/2 was 0.492,lower than that of stromal invasion depth>1/2(1.135),lymph node metastasis(0.728),and tumor diameter>4cm(0.504),but higher than LVSI(0.271).2.According to the multivariate Cox regression model on prognosis analysis of strictly enrolled cases without adjuvant treatment before operation,the risk coefficients of UCI for 5-year OS and DFS were obtained.Five-year OS:Cox regression equation:h(t,X)=h0(t)exp(0.378 × histological grade G3+1.007× uterine myometrial invasion≥ 1/2+1.450 × stromal invasion depth>1/2+1.744×stromal invasion depth unknown+0.874 × lymph node metastasis).Compared with other postoperative pathological risk factors,the risk coefficient of the uterine myometrial invasion≥1/2 was 1.007,lower than that of stromal invasion depth>1/2(1.450),higher than that of lymph node metastasis(0.874).Five-year DFS:Cox regression equation:h(t,X)=h0(t)exp(0.995 × tumor diameter unknown+0.345 × histological grade G3+0.367 × laparoscopic surgery+0.836 × uterine myometrial invasion ≥ 1/2+1.127 × stromal invasion depth>1/2+1.514 × stromal invasion depth unknown+0.610 × parametrial involvement+0.668 ×resection margin involvement+0.840 ×lymph node metastasis).Compared with other postoperative pathological risk factors,the risk coefficient of the uterine myometrial invasion≥1/2 was 0.836,lower than that of stromal invasion depth>1/2(1.127),slightly lower than that of lymph node metastasis(0.840),but higher than that of resection margin involvement(0.668)and parametrial involvement(0.610).Chapter 5:Assessment of postoperative adjuvant radiotherapy for cervical cancer with UCIAccording to the above prognosis analysis results,the uterine myometrial invasion≥1/2 have an effect on the oncological prognosis of cervical cancer patients,and myometrial infiltration<1/2 is only related to the recurrence of the patients in the subgroup of stage ⅠA2-ⅠB and negative lymph node metastasis,but not related to the death,while the uterine endometrial invasion is not an independent prognosis factor for the death and recurrence of the patients.Therefore,further analysis of postoperative adjuvant radiotherapy for the prognosis of patients with uterine myometrial invasion≥1/2,myometrial invasion<1/2 and endometrial invasion were performed.1.A total of 117 patients with uterine myometrial invasion≥1/2 were enrolled,including adjuvant radiotherapy group(67 cases)and non-adjuvant radiotherapy group(50 cases).Before matching,No significant difference was found between adjuvant radiotherapy group and non-adjuvant radiotherapy group for 5-year OS and DFS(OS:64.5%vs.54.9%,P=0.300;DFS:58.0%vs 53.9%,P=0.208).Multivariate Cox regression analysis revealed that adjuvant radiotherapy was not an independent prognosis factor for 5-year OS and DFS(OS:P=0.096;DFS:P=0.106).The results of prognosis analysis after 1:1 PSM was consistent with that before matching.A total of 89 patients without preoperative adjuvant therapy were further selected from 117 patients with uterine myometrial invasion≥1/2,including adjuvant radiotherapy group(49 cases)and non-adjuvant radiotherapy group(40 cases).Before matching,patients in the adjuvant radiotherapy group had a higher 5-year OS and DFS than those of non adjuvant radiotherapy group(OS:69.7%vs.47.9%,P=0.040;DFS:64.4%vs 45.5%,P=0.031).Multivariate Cox regression analysis revealed that adjuvant radiotherapy was an independent prognosis factor of 5-year OS and DFS(OS:HR=0.435,95%CI 0.204-0.925,P=0.031;DFS:HR=0.484,95%CI 0.249-0.940,P=0.032).The results of prognosis analysis after 1:1 PSM was consistent with that before matching.2.A total of 117 patients with uterine myometrial invasion<1/2 were enrolled,including adjuvant radiotherapy group(83 cases)and non-adjuvant radiotherapy group(34 cases).Before matching,no significant difference was found between adjuvant radiotherapy group and non-adjuvant radiotherapy group for OS and DFS(OS:72.9%vs.86.5%,P=0.592;DFS:66.9%vs.73.4%,P=0.943).Multivariate Cox regression analysis revealed that adjuvant radiotherapy was not an independent prognosis factor of 5-year OS and DFS(OS:P=0.453;DFS:P=0.899).The results of prognosis analysis after 1:2 PSM was consistent with that before matching.A total of 105 patients without preoperative adjuvant therapy were further selected from 117 patients with uterine myometrial invasion<1/2,including adjuvant radiotherapy group(74 cases)and non-adjuvant radiotherapy group(31 cases).Before matching,no significant difference was found between adjuvant radiotherapy group and non-adjuvant radiotherapy group for OS and DFS(OS:78.2%vs.93.3%,P=0.452;DFS:71.0%vs.77.7%,P=0.969).Multivariate Cox regression analysis revealed that adjuvant radiotherapy was not an independent prognosis factor of 5-year OS and DFS(OS:P=0.452;DFS:P=0.893).The results of prognosis analysis after 1:2 PSM was consistent with that before matching.3.A total of 104 patients with uterine endometrial invasion were enrolled,including adjuvant radiotherapy group(58 cases)and non-adjuvant radiotherapy group(46 cases).Before matching,no significant difference was found between adjuvant radiotherapy group and non-adjuvant radiotherapy group for OS and DFS(OS:84.9%vs.87.6%,P=0.737;DFS:77.4%vs.77.2%,P=0.689).Multivariate Cox regression analysis revealed that adjuvant radiotherapy was not an independent prognosis factor of 5-year OS and DFS(OS:P=0.939;DFS:P=0.890).Because the baseline of the two groups before matching was balanced,there was no need to match again to verify the stability of the results.A total of 90 patients without preoperative adjuvant therapy were further selected from 104 patients with uterine endometrial invasion,including adjuvant radiotherapy group(49 cases)and non-adjuvant radiotherapy group(41 cases).Before matching,no significant difference was found between adjuvant radiotherapy group and nonadjuvant radiotherapy group for OS and DFS(OS:87.5%vs.89.3%,P=0.910;DFS:77.9%vs.77.2%,P=0.767).Multivariate Cox regression analysis revealed that adjuvant radiotherapy was not an independent prognosis factor of 5-year OS and DFS(OS:P=0.998;DFS:P=0.987).The results of prognosis analysis after 1:1 PSM was consistent with that before matching.[Conclusions]The incidence of cervical cancer with UCI is 5.2%.The miss diagnosis rate of UCI is 10.5%and misdiagnosis rate is 2.1%.The cervical cancer patients with stageⅠA2-ⅡB and with UCI has a worse oncological outcomes.Further analysis revealed that uterine myometrial invasion≥1/2 is related to the worse oncological outcomes of patients,and its risk coefficient is slightly smaller than that of lymph node which is a high risk factor of postoperative pathology of cervical cancer.The uterine myometrial infiltration<1/2 was only related to tumor recurrence in ⅠA2-ⅠB stage and negative lymph node metastasis,and its risk coefficient was significantly lower than that of lymph node metastasis.The uterine myometrial infiltration isn’t related to the oncological prognosis of patients.Postoperative adjuvant radiotherapy can improve the oncological prognosis of cervical cancer patients with uterine myometrial invasion≥1/2,while couldn,t improve the the oncological prognosis of cervical cancer patients with uterine myometrial invasion<1/2 and endometrial invasion.Hence,cervical cancer patients with uterine myometrial invasion≥1/2 should be paid much attention. |
PDF Full Text Request