| Objective:Ovarian endometrioma(OMA)is an important type of endometriosis(EMT),which is difficult to diagnose at an early stage,complicated to operate at a late stage,and prone to recurrence of residual lesions,Therefore,their preoperative diagnosis and evaluation has become an urgent and complex problem of medical research.In recent years,magnetic resonance imaging technology has been rapidly developed and the increasing number of serum markers are gradually applied to gynecological diseases,but the combined use of both for preoperative assessment of endometriosis severity,combined deep-infiltrating endometriosis(DIE)and OMA staging is still in the exploratory stage.Therefore,we retrospectively analyzed MRI features and serologic data in OMA patients to clarify the clinical value of combining these two factors in the differential diagnosis and evaluation of ovarian endometriosis and to enable clinicians to develop individualized treatment plans.Methods:A retrospective case-control study based on inpatients was used,and 179 patients with inpatient surgery and postoperative pathology of OMA at the Affiliated Hospital of Qingdao University from October 2019 to October 2022 were collected as the study group(OMA group),and 136 patients with postoperative pathology confirmed as benign ovarian tumors who underwent surgery at our hospital during the same period were collected as the control group(non-OMA group)according to the inclusion and exclusion criteria.General clinical data [age,age at menarche,body mass index(BMI),number of abortions,previous surgical history],magnetic resonance signs(T1WI,T2 WI and DWI cyst signal,ovarian Kiss sign,intracapsular compartment,fluid-fluid plane,cyst side and diameter)and serological indices(CA125,CA199,CA724,IL-6,IL 8,IL-10,AMH,anti-endometrial antibodies),and the diagnostic efficacy of single and combined indicators for moderate to severe OMA was analyzed by single-factor,multi-factor and subject workup(ROC)curves;a preoperative prediction model for ovarian endometriosis patients with combined DIE and ovarian endometriosis cysts typing was developed using logistic regression analysis to assess the condition of OMA patients in various aspects.Results:1.Comparison of general data,MRI and serological indexes in the OMA and non-OMA groupsThe mean age and number of abortions in the OMA group were greater than those in the non-OMA group(P<0.001),and their age at menarche and BMI were less than those in the non-OMA group(P<0.05);the incidence of MRI signs of T1 WI,T2WI high signal,ovarian Kiss sign,intracapsular segregation,and fluid-fluid plane were higher in the OMA group than in the non-OMA group(P<0.001).The serum levels of CA125,CA724 and IL-6 in the OMA group were higher than those in the non-OMA group(P<0.05).2.Comparison of MRI and serological indexes and the value of combined diagnosis of moderate-severe endometriosis in the moderate-severe OMA group and the mild groupThe incidence of ovarian Kiss sign and cyst diameter in the moderate-severe group of OMA were greater than those in the mild group(P<0.001),and the serum levels of CA125,CA199 and IL-6 in the moderate-severe group were higher than those in the mild group(P<0.05).Ovarian Kiss sign,cyst diameter,CA125,CA199 and IL-6 were all positively correlated with r-AFS score(r>0.1,P<0.001),among which the stronger correlation was between ovarian Kiss sign and patients’ preoperative serum IL-6 level(r>0.5,P<0.001).The results of multifactorial logistic analysis suggested that ovarian Kiss sign,cyst diameter and CA125 were independent risk factors for moderate to severe endometriosis,and the combination of the three was more accurate than each index alone in diagnosing moderate to severe endometriosis(AUC=0.899),with a sensitivity of94.9% and specificity of 70.5%;the optimal cut-off value for cyst diameter was 5.65 cm and for CA125 was 5.65 cm.The optimal diagnostic threshold for CA125 was 29.75U/ml.3.Logistic regression modeling and predictive efficacy assessment of DIE in patients with ovarian endometriosisThe risk prediction model for OMA combined with DIE was: logistic(P)=-2.419+0.271×(IL-6 level)+1.085×(Kiss sign: yes=1,no=0);the H-L test model fitted well with the observed values(c2=10.464,P=0.234),and the AUC of ROC was 0.730,suggesting that the model has good predictive good predictive performance.4.Logistic regression modeling and assessment of predictive efficacy for ovarian endometriosis cysts typingThe risk prediction model for type II OMA was: Logistic(P)=-1.378+0.464×(cyst diameter)+2.117×(Kiss sign: yes=1,no=0);the H-L test model fitted well with the observed values(c2=11.354,P=0.182),and the AUC of ROC was 0.864,suggesting that the model has good predictive efficacy.Conclusion:1.Patient’s age,age at menarche,BMI and number of abortions,characteristic MRI signs such as ovarian Kiss sign and serum CA125,CA724,IL-6 have differential diagnostic value in ovarian type endoheterosis and other benign ovarian tumors.2.Ovarian phase anastomosis sign,cyst diameter,serum CA125,CA199,and IL-6were associated with the stage of endoheterosis.Magnetic resonance ovarian phase anastomosis sign combined with serum CA125 and IL-6 can improve the diagnosis of stage III and IV ovarian type endoheterosis and is recommended for clinical evaluation.3.The logistic risk prediction model for ovarian endometriosis patients with combined DIE and ovarian endometriosis cysts typing has superior predictive efficacy and can be promoted in the clinic. |
PDF Full Text Request