European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Prediction of endometrial carcinogenesis probability while diagnosed as atypical endometrial hyperplasia: a new risk model based on age, CA199 and CA125 assay Lina Zhou a, Zhenni Meng a, Yueqian Wu a, Huiting Zhu b, Xipeng Wang a,* a b
Department of Gynecology, Shanghai First Maternity and Infant Hospital, Tongji University School of Medicine, Shanghai, China Department of Pathology, Shanghai First Maternity and Infant Hospital, Tongji University School of Medicine, Shanghai, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 21 April 2014 Received in revised form 11 September 2014 Accepted 4 October 2014
Objective: To establish a risk model for predicting the probability of endometrial carcinogenesis while diagnosed as AEH preoperatively. Study design: 149 patients with AEH diagnosed by D&C specimens or hysteroscopy undergoing total hysterectomy were enrolled in this retrospective study. Patients were divided into two groups: 51 cases in AEH to AEH group and 98 cases in AEH to EC group. The clinical pathological characteristics were compared between the two groups. With the use of ROC curves, cutoff values of age, CA199, CA125 and uterine height were determined. The relationships of these evaluators with endometrial carcinogenesis were investigated by multivariate analysis with a logistic regression model. Results: 65.8% (98/149) of preoperative AEH patients become EC after confirmed by paraffin hysterectomy specimens. The cutoff values were 51.5 years of age, 14.06 U/ml of CA199, 14.30 U/ml of CA125 and 45 mm of uterine height, respectively. Logistic regression analysis demonstrated age (OR = 0.58, 95% confidence interval [95%CI]: 0.08–3.94), CA199 (OR = 7.27, 95%CI: 1.43–36.92), CA125 (OR = 13.97, 95%CI: 2.49–78.37), uterine height (OR = 0.49, 95%CI: 0.08–3.08) and abnormal uterine bleeding (OR = 4.27, 95%CI: 0.53–34.52) were associated with endometrial carcinogenesis. However, only CA199 (p = 0.017) and CA125 (p = 0.003) revealed significant differences. Therefore, when entered into the risk model, CA199 14.06 U/ml and CA125 14.30 U/ml were considered as cogent predictors of endometrial carcinogenesis. Conclusions: This risk model may well predict the probability of endometrial carcinogenesis when preoperatively diagnosed as AEH. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Atypical endometrial hyperplasia Endometrial carcinogenesis CA199 CA125 Age
Introduction Atypical endometrial hyperplasia (AEH) is a well-known precursor to endometrial cancer (EC) and multiple reports have indicated that women with a preoperative diagnosis of AEH were frequently found to be EC after hysterectomy [1–4]. According to that statistic, EC represents the most common gynecologic malignancy and the incidence is rising steadily for several decades in the majority developed countries [5].
Abbreviations: AEH, atypical endometrial hyperplasia; D&C, dilation and curettage; EC, endometrial cancer; ROC curves, receiver operating characteristic curves. * Corresponding author at: Department of Gynecology, Shanghai First Maternity and Infant Hospital, Tongji University, School of Medicine, No. 536, Changle Road, Shanghai, China. Tel.: +86 21 854035206; fax: +86 21 854035206. E-mail address: [email protected] (X. Wang). http://dx.doi.org/10.1016/j.ejogrb.2014.10.006 0301-2115/ß 2014 Elsevier Ireland Ltd. All rights reserved.
Nowadays, dilation and curettage (D&C) is still considered as the most common approach in the preoperative evaluation of uterine endometrial disease [6]. However, as a blind operative procedure, D&C has some shortages such as: may curet less than 50% of uterine cavity, obtain few samples, and result in inadequate diagnosis. Several recent studies revealed that hysterectomy specimens obtained from the AEH patients diagnosed by D&C showing AEH are associated with a risk of concurrent EC and the incidence may be as high as 20–50% [7,8]. In addition, the Gynecologic Oncology Group (GOG) published a prospective cohort study involving 289 patients with a preoperative diagnosis of AEH had coexisting EC and reported a similarly rate as 42.6% [9]. As we know, a histological positive endometrial biopsy demands the planning of definitive treatment [10]. Therefore, some researchers indicate that all women with atypical hyperplasia should receive definitive surgical management [11]. In order to
6
L. Zhou et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
ensure appropriate management and patient safety, it is essential to have a better understanding of concurrent EC among women with AEH as diagnosed by D&C or hysteroscopy preoperatively. Multiple clinical factors (such as age, CA125, CA199, abnormal uterine bleeding, et al.) influenced the coexistence incidence of EC in patients diagnosed with AEH. It is important to examine the prevalence of undiagnosed EC among women with a preoperative diagnosis of AEH in correlation to these clinical factors. Nevertheless, there has been no report concerning the risk model which indicates endometrial carcinogenesis probability in AEH patients by now. The management approach for AEH gradually became a broad and controversial topic. When considering management strategies for women with preoperative diagnosis of AEH, clinicians should take into account the considerable rate of concurrent EC and the discrepancy with pathologic diagnosis. We therefore sought to conduct a retrospective observational study to evaluate the clinical factors and to establish a risk model for predicting the probability of endometrial carcinogenesis and ensuring appropriate management of AEH patients. The current study was carried out on 149 patients with AEH diagnosed from D&C specimens or hysteroscopy and underwent total hysterectomy in the short term. Our research may be the first study about comparison of clinicopathologic factors between AEH to AEH and AEH to EC groups in case of establishing an efficient and feasible risk model of predicting endometrial carcinogenesis. We believe our risk model may well predict the probability of endometrial carcinogenesis when preoperatively diagnosed as AEH and could potentially be useful in future clinical trials to determine which patients with AEH should have total hysterectomy as part of their management.
Materials and methods Patients We analyzed retrospectively the medical records of 149 women with primary diagnosis of AEH undergoing treatment in the Department of Gynecology, Shanghai First Maternity and Infant Hospital, Tongji University from Jan 2008 to Mar 2013. They were divided into two groups (AEH to AEH & AEH to EC) according to the postoperative pathological diagnosis. Clinical data were compared between the two groups to establish a risk model for predicting uterine endometrial carcinogenesis probability of AEH patients. This study was approved by the Institutional Review Board Shanghai First Maternity and Infant Hospital. Tissue samples Endometrial tissue sampling for diagnosis was carried out by D&C or hyesteroscopy. All endometrial sampling was completed by resident physicians under the supervision of attending physicians. Patients were all staged surgically by a standard protocol within 6 months. No preoperative radio-therapy or chemotherapy was administered. All histological sections of the endometrial samples were stained with hematoxylin and eosin for histopathological examination. The serum CA199 & CA125 levels were determined using a RIA kit (Fujirebio Diagnostics, Malvern, PA). The pre- and postoperative histopathological findings were evaluated for each case. The histopathological diagnosis of the endometrial sample was compared with the endometrial diagnosis for the hysterectomy specimen. The specimens were reviewed by two gynecologic pathologists at the Department of Pathology in our hospital.
Data analysis Data analyzed between two groups included the following variables: age, abnormal uterine bleeding, CA199 and CA125 assay, uterine height & endometrial thickness (as assessed by ultrasound examination), menarche age, menopausal age, parity, body weight. The results of the measurement were used to obtain ROC curves for endometrial carcinogenesis ability. With use of these curves, the cutoff points of each variable were set. Logistic regression analysis was used to select the risk factors for endometrial carcinogenesis probability. Variables that achieved statistical significance in univariate analysis were included subsequently in a multivariate analysis. The Statistical Package for Social Sciences (Version 20.0; SPSS, Chicago, IL, USA) was used for all statistical analyses. Statistic significant difference was defined as two-tailed p < 0.05.
Results Patient characteristics pre-operation We identified 149 women with AEH on endometrial sampling who underwent hysterectomy within 6 months and did not have evidence of resolution before hysterectomy. The clinicopathologic characteristics of the patients are summarized in Table 1. The mean ages of the two groups were 49.24 9.02 years (range, 47–57 years) in AEH to AEH group and 55.89 8.90 years (range, 54– 58 years) in AEH to EC group, respectively, and statistic significance differences can be found (p = 0.000). Menarche age, menopausal age, menopausal percentages, parity and body weight were also compared between the two groups, however, no statistic significant difference was observed. We used transvaginal ultrasonography to detect endometrial thickness and uterine height in this study. The mean thickness of the two groups were 9.37 4.75 mm (range, 7.98–10.77 mm) in AEH to AEH group and 10.60 6.08 mm (range, 8.87 12.33 mm) in AEH to EC group, respectively, and no statistical significance can be found. The mean uterine height was 54.32 12.08 mm (range, 50.65–57.99 mm) in AEH to AEH group and was 48.22 14.23 mm
Table 1 Characteristics of the two groups. Factors
AEH-AEH
AEH-CA
p value
Mean age (years) (range) Menarche age (years) (range) Mean parity (times) (range) Body weight (kg) (range) Menopausal age (years) (range) Menopausal (%) Endometrial thickness (mm) (range) Uterine height (mm) (range) CA125 (U/ml) (range) CA199 (U/ml) (range) Hypertension (%) Diabetes mellitus (%)
49.24 9.02 (46.70–56.77) 14.53 1.68 (13.91–15.16) 1.27 0.84 (1.03–1.52) 64.13 14.15 (52.30–75.95) 51.33 2.46 (49.77–52.90) 23.53% (12/51) 9.37 4.75 (7.98–10.77) 54.32 12.08 (50.65–57.99) 22.99 35.57 (9.94–36.04) 9.99 7.54 (7.07–12.91) 25.49% (13/51) 3.92% (2/51)
55.89 8.90 (54.10–57.67) 14.66 1.81 (14.09–15.23) 1.39 0.94 (1.14–1.62) 58.23 17.42 (49.83–66.62) 51.03 3.29 (49.86–52.20) 33.67% (33/98) 10.60 6.08 (8.87 12.33) 48.22 14.23 (43.99–52.44) 62.04 170.95 (2.39–121.69) 51.49 109.08 (12.17–90.82) 31.63% (31/98) 12.24% (12/98)
0.000** 0.732 0.598 0.264 0.819 0.260 0.324 0.039* 0.068 0.002** 0.457 0.140
Age, uterine height and CA199 value showed significant differences between the two groups. * p < 0.05. ** p < 0.01.
L. Zhou et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
(range, 43.99–52.44 mm) in AEH to EC group, and statistic significance differences can be observed (p = 0.039). The serum levels of CA199 were also investigated, 9.99 7.54 U/ ml (range, 7.07–12.91 U/ml) in AEH to AEH group and 51.49 109.08 U/ml (range, 12.17–90.82 U/ml) in AEH to EC group, and dramatic significant differences can be obtained (p = 0.002). Additionally, the values of CA125 were 22.99 35.57 U/ml (range, 9.94–36.04 U/ml) in AEH to AEH group and 62.04 170.95 U/ml (range, 2.39–121.69 U/ml) in AEH to EC group, respectively, yet, obvious not significant difference was seen (p = 0.068). In all, 44 patients with hypertension (25.49% in AEH to AEH group and 31.63% in AEH to EC group) and 14 patients with diabetes mellitus (3.92% in AEH to AEH group and 12.24% in AEH to EC group) were observed. Since some cases absent of those medical complications, no statistic significant difference was observed among these factors.
7
Table 2 Final pathological findings post-operation between the two groups. Final pathological findings
Normal Simple hyperplasia Complex hyperplasia Atypical hyperplasia FIGO stage I II III
AEH-AEH
AEH-CA
n (%)
n (%)
8 5 13 25
(15.69) (9.80) (25.49) (49.02) 52 (53.06) 38 (38.78) 8 (8.16)
Final pathological findings indicating that 53.06% of AEH is in low FIGO grade (stage I) and 46.94% of AEH is in high FIGO grade (stages II and III).
Analysis of factors related to endometrial carcinogenesis Final pathological findings post-operation The final pathological findings of AEH to AEH group were 15.69% in normal, 9.80% in simple hyperplasia, 25.49% in complex hyperplasia and 49.02% in atypical hyperplasia. On the other hand, the final pathological findings of AEH to EC group were 53.06% in FIGO stage I, 38.78% in FIGO stage II and 8.16% in FIGO stage III (Table 2). Previous literature reports [12] indicate that AEH is primarily associated with low FIGO grade tumors that mostly infiltrate the myometrium only superficially. However, our findings seem to not that similar to their results, as 53.06% in low FIGO grade (stage I) and 46.94% in high FIGO grade (stages II and III).
We use chi-square test to analyze the factors include age, CA199, CA125, uterine height and abnormal uterine bleeding which are strongly associated with endometrial carcinogenesis (p = 0.000 for age, p = 0.000 for CA199, p = 0.006 for CA125, p = 0.007 for uterine height and p = 0.000 for abnormal uterine bleeding). The results showed when age 51.5 years, CA199 14.03 U/ml, CA125 14.30 U/ml and with post-menopausal abnormal uterine bleeding, patients are more likely to endometrial carcinogenesis when compared to younger and lower CA199 and CA125 expression and no abnormal uterine bleeding patients (Table 3). The smaller uterine height here may be attributed to older age in AEH to EC group and smaller sample in our research.
Receiver operating characteristic curves Logistic regression analysis We obtained ROC curves with the use of age, CA199 and CA125 values (Fig. 1). When determined on the curves obtained from the relationships between age and endometrial carcinogenesis ability, cut-off value was 51.50 years with a sensitivity of 0.755 and a specificity of 0.627. ROC curve also indicated that, at the cutoff value of 14.06 U/ml, the variable CA199 had a sensitivity of 0.594 and a specificity of 0.857 in predicting of endometrial carcinogenesis ability when diagnosed with AEH. Similarly, ROC curve revealed at the cutoff value of 14.30 U/ml, the variable CA125 had a sensitivity of 0.706 and a specificity of 0.645 in predicting of endometrial carcinogenesis ability. Areas under curve were 0.719 of age, 0.668 of CA199 and 0.664 of CA125, respectively.
The multivariate logistic regression analysis in which endometrial carcinogenesis was used as a dependent variable demonstrated that age (OR = 0.58, 95% confidence interval [95%CI]: 0.08–3.94), CA199 (OR = 7.27, 95%CI: 1.43–36.92), CA125 (OR = 13.97, 95%CI: 2.49–78.37), uterine height (OR = 0.49, 95%CI: 0.08–3.08) and abnormal uterine bleeding (OR = 4.27, 95%CI: 0.53–34.52) were associated with endometrial carcinogenesis ability (Table 4). However, the final result identified only CA199 (p = 0.017) and CA125 (p = 0.003) as cogent predictors of endometrial carcinogenesis. Thus, hysterectomy may be necessary when at least one of these predictors is present.
Fig. 1. (a) ROC curve obtained from the relationships between age and endometrial carcinogenesis (area under curve 0.719). (b) ROC curves obtained from the relationships between serum CA199 level and endometrial carcinogenesis (area under curve 0.668). (c) ROC curves obtained from the relationships between serum CA125 level and endometrial carcinogenesis (area under curve 0.664).
L. Zhou et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
8
Table 3 Analysis of factors related to endometrial canceration. Factors
Age (years)
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Prediction of endometrial carcinogenesis probability while diagnosed as atypical endometrial hyperplasia: a new risk model based on age, CA199 and CA125 assay Lina Zhou a, Zhenni Meng a, Yueqian Wu a, Huiting Zhu b, Xipeng Wang a,* a b
Department of Gynecology, Shanghai First Maternity and Infant Hospital, Tongji University School of Medicine, Shanghai, China Department of Pathology, Shanghai First Maternity and Infant Hospital, Tongji University School of Medicine, Shanghai, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 21 April 2014 Received in revised form 11 September 2014 Accepted 4 October 2014
Objective: To establish a risk model for predicting the probability of endometrial carcinogenesis while diagnosed as AEH preoperatively. Study design: 149 patients with AEH diagnosed by D&C specimens or hysteroscopy undergoing total hysterectomy were enrolled in this retrospective study. Patients were divided into two groups: 51 cases in AEH to AEH group and 98 cases in AEH to EC group. The clinical pathological characteristics were compared between the two groups. With the use of ROC curves, cutoff values of age, CA199, CA125 and uterine height were determined. The relationships of these evaluators with endometrial carcinogenesis were investigated by multivariate analysis with a logistic regression model. Results: 65.8% (98/149) of preoperative AEH patients become EC after confirmed by paraffin hysterectomy specimens. The cutoff values were 51.5 years of age, 14.06 U/ml of CA199, 14.30 U/ml of CA125 and 45 mm of uterine height, respectively. Logistic regression analysis demonstrated age (OR = 0.58, 95% confidence interval [95%CI]: 0.08–3.94), CA199 (OR = 7.27, 95%CI: 1.43–36.92), CA125 (OR = 13.97, 95%CI: 2.49–78.37), uterine height (OR = 0.49, 95%CI: 0.08–3.08) and abnormal uterine bleeding (OR = 4.27, 95%CI: 0.53–34.52) were associated with endometrial carcinogenesis. However, only CA199 (p = 0.017) and CA125 (p = 0.003) revealed significant differences. Therefore, when entered into the risk model, CA199 14.06 U/ml and CA125 14.30 U/ml were considered as cogent predictors of endometrial carcinogenesis. Conclusions: This risk model may well predict the probability of endometrial carcinogenesis when preoperatively diagnosed as AEH. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Atypical endometrial hyperplasia Endometrial carcinogenesis CA199 CA125 Age
Introduction Atypical endometrial hyperplasia (AEH) is a well-known precursor to endometrial cancer (EC) and multiple reports have indicated that women with a preoperative diagnosis of AEH were frequently found to be EC after hysterectomy [1–4]. According to that statistic, EC represents the most common gynecologic malignancy and the incidence is rising steadily for several decades in the majority developed countries [5].
Abbreviations: AEH, atypical endometrial hyperplasia; D&C, dilation and curettage; EC, endometrial cancer; ROC curves, receiver operating characteristic curves. * Corresponding author at: Department of Gynecology, Shanghai First Maternity and Infant Hospital, Tongji University, School of Medicine, No. 536, Changle Road, Shanghai, China. Tel.: +86 21 854035206; fax: +86 21 854035206. E-mail address: [email protected] (X. Wang). http://dx.doi.org/10.1016/j.ejogrb.2014.10.006 0301-2115/ß 2014 Elsevier Ireland Ltd. All rights reserved.
Nowadays, dilation and curettage (D&C) is still considered as the most common approach in the preoperative evaluation of uterine endometrial disease [6]. However, as a blind operative procedure, D&C has some shortages such as: may curet less than 50% of uterine cavity, obtain few samples, and result in inadequate diagnosis. Several recent studies revealed that hysterectomy specimens obtained from the AEH patients diagnosed by D&C showing AEH are associated with a risk of concurrent EC and the incidence may be as high as 20–50% [7,8]. In addition, the Gynecologic Oncology Group (GOG) published a prospective cohort study involving 289 patients with a preoperative diagnosis of AEH had coexisting EC and reported a similarly rate as 42.6% [9]. As we know, a histological positive endometrial biopsy demands the planning of definitive treatment [10]. Therefore, some researchers indicate that all women with atypical hyperplasia should receive definitive surgical management [11]. In order to
6
L. Zhou et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
ensure appropriate management and patient safety, it is essential to have a better understanding of concurrent EC among women with AEH as diagnosed by D&C or hysteroscopy preoperatively. Multiple clinical factors (such as age, CA125, CA199, abnormal uterine bleeding, et al.) influenced the coexistence incidence of EC in patients diagnosed with AEH. It is important to examine the prevalence of undiagnosed EC among women with a preoperative diagnosis of AEH in correlation to these clinical factors. Nevertheless, there has been no report concerning the risk model which indicates endometrial carcinogenesis probability in AEH patients by now. The management approach for AEH gradually became a broad and controversial topic. When considering management strategies for women with preoperative diagnosis of AEH, clinicians should take into account the considerable rate of concurrent EC and the discrepancy with pathologic diagnosis. We therefore sought to conduct a retrospective observational study to evaluate the clinical factors and to establish a risk model for predicting the probability of endometrial carcinogenesis and ensuring appropriate management of AEH patients. The current study was carried out on 149 patients with AEH diagnosed from D&C specimens or hysteroscopy and underwent total hysterectomy in the short term. Our research may be the first study about comparison of clinicopathologic factors between AEH to AEH and AEH to EC groups in case of establishing an efficient and feasible risk model of predicting endometrial carcinogenesis. We believe our risk model may well predict the probability of endometrial carcinogenesis when preoperatively diagnosed as AEH and could potentially be useful in future clinical trials to determine which patients with AEH should have total hysterectomy as part of their management.
Materials and methods Patients We analyzed retrospectively the medical records of 149 women with primary diagnosis of AEH undergoing treatment in the Department of Gynecology, Shanghai First Maternity and Infant Hospital, Tongji University from Jan 2008 to Mar 2013. They were divided into two groups (AEH to AEH & AEH to EC) according to the postoperative pathological diagnosis. Clinical data were compared between the two groups to establish a risk model for predicting uterine endometrial carcinogenesis probability of AEH patients. This study was approved by the Institutional Review Board Shanghai First Maternity and Infant Hospital. Tissue samples Endometrial tissue sampling for diagnosis was carried out by D&C or hyesteroscopy. All endometrial sampling was completed by resident physicians under the supervision of attending physicians. Patients were all staged surgically by a standard protocol within 6 months. No preoperative radio-therapy or chemotherapy was administered. All histological sections of the endometrial samples were stained with hematoxylin and eosin for histopathological examination. The serum CA199 & CA125 levels were determined using a RIA kit (Fujirebio Diagnostics, Malvern, PA). The pre- and postoperative histopathological findings were evaluated for each case. The histopathological diagnosis of the endometrial sample was compared with the endometrial diagnosis for the hysterectomy specimen. The specimens were reviewed by two gynecologic pathologists at the Department of Pathology in our hospital.
Data analysis Data analyzed between two groups included the following variables: age, abnormal uterine bleeding, CA199 and CA125 assay, uterine height & endometrial thickness (as assessed by ultrasound examination), menarche age, menopausal age, parity, body weight. The results of the measurement were used to obtain ROC curves for endometrial carcinogenesis ability. With use of these curves, the cutoff points of each variable were set. Logistic regression analysis was used to select the risk factors for endometrial carcinogenesis probability. Variables that achieved statistical significance in univariate analysis were included subsequently in a multivariate analysis. The Statistical Package for Social Sciences (Version 20.0; SPSS, Chicago, IL, USA) was used for all statistical analyses. Statistic significant difference was defined as two-tailed p < 0.05.
Results Patient characteristics pre-operation We identified 149 women with AEH on endometrial sampling who underwent hysterectomy within 6 months and did not have evidence of resolution before hysterectomy. The clinicopathologic characteristics of the patients are summarized in Table 1. The mean ages of the two groups were 49.24 9.02 years (range, 47–57 years) in AEH to AEH group and 55.89 8.90 years (range, 54– 58 years) in AEH to EC group, respectively, and statistic significance differences can be found (p = 0.000). Menarche age, menopausal age, menopausal percentages, parity and body weight were also compared between the two groups, however, no statistic significant difference was observed. We used transvaginal ultrasonography to detect endometrial thickness and uterine height in this study. The mean thickness of the two groups were 9.37 4.75 mm (range, 7.98–10.77 mm) in AEH to AEH group and 10.60 6.08 mm (range, 8.87 12.33 mm) in AEH to EC group, respectively, and no statistical significance can be found. The mean uterine height was 54.32 12.08 mm (range, 50.65–57.99 mm) in AEH to AEH group and was 48.22 14.23 mm
Table 1 Characteristics of the two groups. Factors
AEH-AEH
AEH-CA
p value
Mean age (years) (range) Menarche age (years) (range) Mean parity (times) (range) Body weight (kg) (range) Menopausal age (years) (range) Menopausal (%) Endometrial thickness (mm) (range) Uterine height (mm) (range) CA125 (U/ml) (range) CA199 (U/ml) (range) Hypertension (%) Diabetes mellitus (%)
49.24 9.02 (46.70–56.77) 14.53 1.68 (13.91–15.16) 1.27 0.84 (1.03–1.52) 64.13 14.15 (52.30–75.95) 51.33 2.46 (49.77–52.90) 23.53% (12/51) 9.37 4.75 (7.98–10.77) 54.32 12.08 (50.65–57.99) 22.99 35.57 (9.94–36.04) 9.99 7.54 (7.07–12.91) 25.49% (13/51) 3.92% (2/51)
55.89 8.90 (54.10–57.67) 14.66 1.81 (14.09–15.23) 1.39 0.94 (1.14–1.62) 58.23 17.42 (49.83–66.62) 51.03 3.29 (49.86–52.20) 33.67% (33/98) 10.60 6.08 (8.87 12.33) 48.22 14.23 (43.99–52.44) 62.04 170.95 (2.39–121.69) 51.49 109.08 (12.17–90.82) 31.63% (31/98) 12.24% (12/98)
0.000** 0.732 0.598 0.264 0.819 0.260 0.324 0.039* 0.068 0.002** 0.457 0.140
Age, uterine height and CA199 value showed significant differences between the two groups. * p < 0.05. ** p < 0.01.
L. Zhou et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
(range, 43.99–52.44 mm) in AEH to EC group, and statistic significance differences can be observed (p = 0.039). The serum levels of CA199 were also investigated, 9.99 7.54 U/ ml (range, 7.07–12.91 U/ml) in AEH to AEH group and 51.49 109.08 U/ml (range, 12.17–90.82 U/ml) in AEH to EC group, and dramatic significant differences can be obtained (p = 0.002). Additionally, the values of CA125 were 22.99 35.57 U/ml (range, 9.94–36.04 U/ml) in AEH to AEH group and 62.04 170.95 U/ml (range, 2.39–121.69 U/ml) in AEH to EC group, respectively, yet, obvious not significant difference was seen (p = 0.068). In all, 44 patients with hypertension (25.49% in AEH to AEH group and 31.63% in AEH to EC group) and 14 patients with diabetes mellitus (3.92% in AEH to AEH group and 12.24% in AEH to EC group) were observed. Since some cases absent of those medical complications, no statistic significant difference was observed among these factors.
7
Table 2 Final pathological findings post-operation between the two groups. Final pathological findings
Normal Simple hyperplasia Complex hyperplasia Atypical hyperplasia FIGO stage I II III
AEH-AEH
AEH-CA
n (%)
n (%)
8 5 13 25
(15.69) (9.80) (25.49) (49.02) 52 (53.06) 38 (38.78) 8 (8.16)
Final pathological findings indicating that 53.06% of AEH is in low FIGO grade (stage I) and 46.94% of AEH is in high FIGO grade (stages II and III).
Analysis of factors related to endometrial carcinogenesis Final pathological findings post-operation The final pathological findings of AEH to AEH group were 15.69% in normal, 9.80% in simple hyperplasia, 25.49% in complex hyperplasia and 49.02% in atypical hyperplasia. On the other hand, the final pathological findings of AEH to EC group were 53.06% in FIGO stage I, 38.78% in FIGO stage II and 8.16% in FIGO stage III (Table 2). Previous literature reports [12] indicate that AEH is primarily associated with low FIGO grade tumors that mostly infiltrate the myometrium only superficially. However, our findings seem to not that similar to their results, as 53.06% in low FIGO grade (stage I) and 46.94% in high FIGO grade (stages II and III).
We use chi-square test to analyze the factors include age, CA199, CA125, uterine height and abnormal uterine bleeding which are strongly associated with endometrial carcinogenesis (p = 0.000 for age, p = 0.000 for CA199, p = 0.006 for CA125, p = 0.007 for uterine height and p = 0.000 for abnormal uterine bleeding). The results showed when age 51.5 years, CA199 14.03 U/ml, CA125 14.30 U/ml and with post-menopausal abnormal uterine bleeding, patients are more likely to endometrial carcinogenesis when compared to younger and lower CA199 and CA125 expression and no abnormal uterine bleeding patients (Table 3). The smaller uterine height here may be attributed to older age in AEH to EC group and smaller sample in our research.
Receiver operating characteristic curves Logistic regression analysis We obtained ROC curves with the use of age, CA199 and CA125 values (Fig. 1). When determined on the curves obtained from the relationships between age and endometrial carcinogenesis ability, cut-off value was 51.50 years with a sensitivity of 0.755 and a specificity of 0.627. ROC curve also indicated that, at the cutoff value of 14.06 U/ml, the variable CA199 had a sensitivity of 0.594 and a specificity of 0.857 in predicting of endometrial carcinogenesis ability when diagnosed with AEH. Similarly, ROC curve revealed at the cutoff value of 14.30 U/ml, the variable CA125 had a sensitivity of 0.706 and a specificity of 0.645 in predicting of endometrial carcinogenesis ability. Areas under curve were 0.719 of age, 0.668 of CA199 and 0.664 of CA125, respectively.
The multivariate logistic regression analysis in which endometrial carcinogenesis was used as a dependent variable demonstrated that age (OR = 0.58, 95% confidence interval [95%CI]: 0.08–3.94), CA199 (OR = 7.27, 95%CI: 1.43–36.92), CA125 (OR = 13.97, 95%CI: 2.49–78.37), uterine height (OR = 0.49, 95%CI: 0.08–3.08) and abnormal uterine bleeding (OR = 4.27, 95%CI: 0.53–34.52) were associated with endometrial carcinogenesis ability (Table 4). However, the final result identified only CA199 (p = 0.017) and CA125 (p = 0.003) as cogent predictors of endometrial carcinogenesis. Thus, hysterectomy may be necessary when at least one of these predictors is present.
Fig. 1. (a) ROC curve obtained from the relationships between age and endometrial carcinogenesis (area under curve 0.719). (b) ROC curves obtained from the relationships between serum CA199 level and endometrial carcinogenesis (area under curve 0.668). (c) ROC curves obtained from the relationships between serum CA125 level and endometrial carcinogenesis (area under curve 0.664).
L. Zhou et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 183 (2014) 5–9
8
Table 3 Analysis of factors related to endometrial canceration. Factors
Age (years)
