Development and validation of a nomogram to predict the risk of postmenopausal uterine leiomyoma degeneration: A retrospective study

doi:10.21203/rs.3.rs-3217875/v1

Download PDF

Research Article

Development and validation of a nomogram to predict the risk of postmenopausal uterine leiomyoma degeneration: A retrospective study

https://doi.org/10.21203/rs.3.rs-3217875/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective: The aim of this study was to investigate the risk factors for postmenopausal uterine leiomyoma degeneration and to develop a nomogram for clinical risk assessment.

Study design and setting: A total of 508 patients were enrolled from 1 August 2012 to 1 August 2022. Univariate and multivariate logistic regression models were used to analyse the association between variables and uterine leiomyoma degeneration in postmenopausal patients. A nomogram for uterine leiomyoma degeneration in postmenopausal patients was developed and validated by bootstrap resampling. The calibration curve was used to assess the accuracy of the model and the ROC, DCA and NRI were compared with the clinical experience model.

Results: DM, breast cancer, number, size5, LDH, TG and Ca²⁺ were independent risk factors for uterine leiomyoma degeneration in postmenopausal patients. We developed a user-friendly nomogram which showed good diagnostic performance (AUC=0.714). The model was consistent and the calibration curve of our cohort was close to the ideal diagonal line. DCA indicated that the model has potential value for clinical application. Furthermore, our model was superior to the previous clinical experience model in terms of ROC (P=0.00039), DCA and NRI (NRI=0.4175722).

Conclusions: We have developed a prediction nomogram for uterine leiomyoma degeneration in postmenopausal patients. This nomogram could serve as an important warning signal and evaluation method for the degeneration and malignant transformation of uterine leiomyoma in postmenopausal patients.

uterine leiomyoma degeneration

postmenopausal patients

prediction nomogram

bootstrap

Uterine leiomyoma (UL) is the most common benign tumor in women, with reports suggesting that 70–80% of women will receive this diagnosis in their lifetime [1, 2]. In the United States, hysterectomy and myomectomy account for a significant percentage of the national surgical volume, with approximately 433,621 and 34,000 cases performed annually, respectively [3, 4]. Fortunately, for postmenopausal women, the size and symptoms of the UL tend to decrease, providing some relief. Studies have shown that the incidence of UL decreases after the menopause [5, 6]. However, menopause does not prevent the occurrence of UL degeneration, and the second highest incidence of UL occurs in the 50–54 age group [7]. In postmenopausal women with UL, the focus is on associated symptoms or uterine leiomyoma degeneration. Degeneration of UL is typically caused by loss of blood supply, often due to rapid growth associated with pregnancy. In cases where UL degeneration is suspected and malignant transformation cannot be ruled out, we recommend that postmenopausal women undergo total hysterectomy. However, despite the low incidence of malignant transformation of UL, surgical treatment is often associated with side effects, especially in postmenopausal women. Therefore, it is important to evaluate UL degeneration, especially malignant transformation in postmenopausal women. There is currently no predictive model available to predict UL degeneration, particularly in postmenopausal women. Among the available models, a nomogram provides a highly accurate, individualised, evidence-based risk estimate. As a simple statistical visual tool, nomograms have been widely used to predict disease incidence, progression, prognosis and survival [8–11]. However, to our knowledge, few nomogram prediction models have been used to estimate the risk of UL degeneration, particularly in postmenopausal patients. To address this knowledge gap, we conducted a case-control study in China. The specific aim of this study was to cover a wide range of biological, behavioural and psychological factors to identify those with significant associations with UL degeneration and to clarify the overall predictive value of UL degeneration risk in postmenopausal patients. The next objective was to develop a nomogram to estimate the likelihood of UL degeneration risk in postmenopausal women with risk factors. Our ultimate goal is to reduce unnecessary surgery in postmenopausal women due to UL, thereby reducing patient discomfort and the economic burden on society.

Our analysis follows the TRIPOD (Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis) guideline [12] and incorporates best practice recommendations for variable selection and model construction when using existing observational datasets. We report model calibration, predictive accuracy, overfitting control and internal bootstrap validation. We compared ROC (receiver operator characteristics curve) [13], DCA (decision curve analysis) [14] and NRI (net reclassification index) [15] between our new model and the clinical experience model. The purpose of these comparisons is to assess the performance of our model and demonstrate its superiority over existing models. By following these guidelines and recommendations, we aim to provide accurate and reliable predictions that will ultimately benefit postmenopausal women with UL.

2.1 Study design and data source

This study was designed as a retrospective review of data from enrolled patients diagnosed with UL by pathology at Women's Hospital School of Medicine Zhejiang University. The study protocol was approved by the Institutional Review Board of Womens' Hospital School of Medicine Zhejiang University under IRB-20230105-R with a waiver of informed consent. This retrospective study included 508 patients who underwent surgical treatment at Women's Hospital School of Medicine Zhejiang University from 1 August 2012 to 1 August 2022 and whose pathological results of the final surgical specimen indicated UL. Inclusion criteria were as follows: (a) the interval between the last menstrual period and enrolment is more than 1 year, (b) there are definite pathological and histological diagnostic results. Exclusion criteria were as follows: (a) patients with malignant ovarian tumours and malignant cervical tumours, (b) patients with multiple UL (including intramural and submucosal myomas) but only submucosal leiomyomas treated by hysteroscopy, and (c) patients without clear imaging information.

2.2 Outcomes

Our primary outcome was UL degeneration [16], which is defined as a series of changes caused by changes in nutrient supply or differentiation of cells in the fibroids during growth. These changes include mucoid change, lipoid change, hyaline change, red-like change, calcification and sarcomatous change [17]. We used the International Classification of Diseases, 11th Revision (ICD-11) [18] clinical codes to identify cases discharged from the Women's Hospital School of Medicine Zhejiang University, with a diagnosis of UL between 1 August 2012 and 1 August 2022. We then selected eligible patients based on our inclusion and exclusion criterias. By following these standards, we aim to ensure the reliability and validity of our study results. The ICD-11 codes used were D25 (uterine leiomyoma), D26 (cervical leiomyoma), D28.2 (broad ligament leiomyoma). Relevant clinical data and laboratory test results of the selected 508 patients were recorded individually through the inpatient medical record system. Their UL pathology results were reviewed. At Women's Hospital School of Medicine Zhejiang University, the pathology results of all patients were reviewed by two pathologists (one pathologist and another senior pathologist).

2.3 Predictor variables

We identified predictor variables based on clinical experience and relevant literature reports [19-23], which are presented in Appendix Table 1. We extracted relevant information such as 'menopausal age', 'menopausal duration', 'sym', 'found time', 'related sur', 'gravida', 'party', 'marriage' and 'education' from the inpatient medical records by asking patients at the time of admission. We collected relevant data on 'HBP', 'DM', 'BRASTCANCER', 'TAMO', 'NUMBER', 'CAPBLOOD', 'size5' and 'BMI' from blood pressure and blood glucose measurements, imaging studies and patient enquiries. For laboratory indices, we obtained relevant preoperative reports from patients, including absolute neutrophil and lymphocyte counts, platelet count, CA125 (carbohydrate antigen 125), lactate dehydrogenase (LDH), total cholesterol, triglyceride, calcium ion, magnesium ion, phosphorus ion and homocysteine. All predictive variables were based on the most recent information recorded in the admission records prior to entry into the cohort. At the time of recording, the attending physician, the laboratory or imaging physician, and the pathologist were all blinded to ensure unbiased data collection. By following these guidelines, we aim to ensure the accuracy and reliability of our study results.

2.4 Sample size calculation

A total of 508 patients were included in the study. We finally selected 7 predictors from 29 candidate predictors to build a multivariate risk prediction model. Of the 508 patients, 206 patients showed UL degeneration according to pathological results, with an incidence of 40.6%. Using the R package 'pmsampsize' to calculate sample size [24] (Appendix Table 2), the maximum sample size we need under current conditions is 369, so our current sample size meets the needs of subsequent calculations.

2.5 Handling of missing values

Before analysis, we assumed that missing values were missing at random and used multiple imputation by chained equations (MICE) using the R programming language to replace missing values for BMI, neutrophil and lymphocyte absolute values, platelet count, CA125, lactate dehydrogenase, total cholesterol, triglyceride, calcium ion, magnesium ion, phosphorus ion and homocysteine. We used random forest imputation to preserve the deterministic relationships between different missing variables in the imputed data. We performed five imputations as this approach has relatively high efficiency and was a pragmatic approach given the size of the datasets and the capacity of the available servers and software [25-29]. Finally, we used Rubin's rules to combine the results across the imputed datasets to ensure the reliability and validity of our study findings.

2.6 Derivation and validation of the models

We used Cox's proportional hazards models to estimate coefficients for each risk predictor in the cohort. We also used univariate and multivariable logistic regression models to determine the association between variables and outcome. In the univariate analysis, we performed crude analyses to identify potential risk factors. All variables with a bivariate association with UL degeneration of P < 0.1 were included in the multivariable model. We developed a stepwise nomogram model of UL degeneration using multivariate logistic regression. The nomogram model was constructed following a stepwise selection process according to the Akaike information criterion (AIC) minimum principle [30]. We assessed the discriminatory ability of the model using ROC analysis and used the calibration curve to determine the concentricity between the model probability curve and the ideal curve. We also performed a Hosmer-Lemeshow test on the final model to assess the goodness of fit of the model. A P value > 0.05 indicates an acceptable model fit. The accuracy of our model was further verified by bootstrap validation using computer resampling for 1000 repetitions of simple random sampling with replacement. The clinical utility of our model was assessed using DCA [31]. In addition, we used NRI [32], ROC and DCA to compare the ability of new predictive models to predict outcomes with clinical empirical models. By following these rigorous methods, we aim to provide accurate and reliable predictions for postmenopausal women with UL.

2.7 Statistical analysis

Continuous variables with a normal distribution are described by the t-test. Otherwise, the U-test is used and described by median (interquartile interval). Classification variables are described by frequency (percentage) using the chi-squared test. A P value < 0.05 was considered significant. All statistical analyses were performed with SPSS version 26.0 (IBM, New York). Other statistical analyses were performed using R version 3.6.3 software with the packages boot, rmda, pROC, calibrate, rms and PredictABEL.

3.1 General characteristics

In our cohort, the clinical information of 508 patients was obtained from Women's Hospital School of Medicine Zhejiang University. Among 508 patients, the number of deletions of serum ions (including calcium and phosphorus ions) and homocysteine concentrations of the predictive variables is the largest, but the number of patients with deletions of their values is still less than 10% of the total number of patients. Multiple imputation was used to obtain complete clinical data for 508 patients (Figure 1, Supplementary Figure 1). Table 1 shows the characteristics of the enrolled patients. Of the 508 postmenopausal patients selected, 206 (40.6%) had UL degeneration. When comparing the two groups of postmenopausal women without UL degeneration and those with UL degeneration, we found that there was no significant difference in menopausal age （51.5 vs. 51.96） and menopausal years (8.6 vs. 8.9）, and years since menopause (8.75 vs 9.17). More interestingly, in clinical practice we often decide whether patients with hysteromyoma have symptoms related to follow-up treatment, but our research found that there was no significant difference between the two groups in terms of symptoms (P=0.91).However, there are significant differences in the number and size of the largest UL between the two groups with P ＜0.05. As we can see, the proportion of multiple ULs and the largest UL with a diameter ≥ 5 cm is higher in patients with UL degeneration than in the group without UL degeneration.

Figure 1. Flow chart of this study

Table 1. Baseline characteristics of the analyzed participants

3.2 Risk factors for the degeneration of UL among Postmenopausal women

Table 2 shows the results of univariate and multivariate logistic regression analyses performed to detect the association between variables and degeneration of UL in postmenopausal women. The risk of fibroid size ≥5 cm in postmenopausal patients was higher than that of fibroid size ≤5 cm (OR = 2.71, 95%CI: 1.87-3.92, P ＜0.001). In addition, the risk was higher for patients with UL number ≥2 than for those with number =1 (OR = 3.2, 95%CI: 2.21-4.62, P ＜0.001). Interestingly, postmenopausal patients with breast cancer developed degeneration of the UL with a significantly reduced risk (OR = 0.33, 95%CI: 0.09-1.17, P = 0.085). For menopausal age, menopausal duration, and symptoms, there were no associations with UL degeneration with P ＞0.05. In addition, there was no significant difference in the incidence of UL degeneration between the two groups with regard to the trend of myoma change after menopause, gravidity, parity and BMI. All variables with a bivariate association with UL degeneration of P < 0.1 were included in the multivariable logistic regression, which also yielded the adjusted ORs shown in Table 2. In the multivariable model, according to the principle of AIC minimum, we finally collected 7 factors to build the model, which included DM, breast cancer, number, size5, LDH, TG, Ca²⁺.

Table 2. Association of degeneration of UL with selected variables in bivariate analysis and in multivariable models

3.3 Nomogram for the degeneration of UL among Postmenopausal women

We analysed 29 clinical variables to determine their association with UL degeneration in postmenopausal women. From the original 29 variables, seven were filtered out, including DM, breast cancer, number, size5, LDH, TG and Ca²⁺. In our study, the stepwise selection model was calculated as follows: -5.821 + 0.727*(DM==1,"have") - 1.12*(BRASTCANCER==1,"have")+1.052*(NUMBER==1,"have") + 0.618 * (size5==1,"have") + 0.006 * (LDH) + 0.113 * (TG) + 1.411 * (Ca²⁺).This model was developed through a rigorous stepwise selection process according to the AIC minimum principle and represents the variables most significantly associated with UL degeneration in postmenopausal women. By using this model, we aim to provide clinicians and patients with accurate and reliable predictions in the management of UL degeneration. The probability of UL degeneration in postmenopausal women can be estimated using the stepwise nomogram described in Figure 2. For each postmenopausal patient, the higher the total score, the higher the risk of myoma degeneration. For example, we selected one of the 508 patients whose pathological results showed that it was a uterine leiomyosarcoma. Supplementary Figure 2 shows that the patient's score was 364, indicating an estimated myoma formation of 69.4% for this case. In addition, the Hosmer-Lemeshow test showed that the model was a good fit (P = 0.7634251). The performance of this nomogram was measured using ROC analysis, and the area under the ROC curve (AUC) of this model was 0.714, indicating good diagnostic performance (Figure 3).

3.4 Model validation

We further validated the stepwise nomogram using an internal bootstrap validation method with 1000 resamples. The ROC was measured by bootstrapping for 1000 repetitions, and the AUC of the bootstrapped stepwise model was 0.7141 (95%CI: 0.6912 - 0.7126), which was statistically similar to the original stepwise model. By using this rigorous validation method, we aim to provide clinicians and patients with reliable and accurate predictions in the management of UL degeneration in postmenopausal women (Supplementary Figure 3). In addition, the probabilities predicted by the nomogram agreed well with clinical outcomes (Figure 4). In general, our model showed good fit and calibration with the ideal curve. Meanwhile, the model had good consistency and the calibration curve of our cohort was also close to the ideal diagonal line. And the decision curve shows that the model has potential value for clinical application (Figure 5).

Red solid line: Prediction model；the gray line: Assume all postmenopausal patients have myoma degeneration. Solid horizontal line: Assume no postmenopausal patients have myoma degeneration. The graph indicates the expected net benefit per patient relative to the nomogram prediction of myoma degeneration.

3.5 Two Model comparison

In clinical practice, we often assess the likelihood of UL degeneration in postmenopausal women based on the presence of associated symptoms (sym), the change trend of postmenopausal uterine leiomyoma (changetrend), and the size of the largest uterine leiomyoma (size5), and use this as the basis for surgery. To evaluate the effectiveness of our 7-factor model compared to the previous clinical empirical model, we conducted a study of 508 patients. Figures 6 and 7 show that our model has advantages in ROC (P=0.00039) and DCA. We also compared the NRI of the two models and the results showed that our model has more advantages than the previous clinical experience models (NRI=0.4175722). By demonstrating the superior performance of our model in predicting UL degeneration in postmenopausal women, we aim to provide clinicians with a reliable and effective tool for decision making in surgical management.

Previous studies have shown that the incidence of Uterine leiomyoma (UL) in postmenopausal women decreases with decreasing hormone levels[33]. Generally, when UL in postmenopausal women becomes large and associated with symptoms such as vitreous degeneration, mucous degeneration, steatosis, red degeneration, cystic degeneration and calcification, surgical treatment is considered. In severe cases, malignant transformation may occur [34,35].

UL degeneration is generally a primary indication for surgery in postmenopausal women. The decision to perform surgery is largely based on a comprehensive assessment by clinicians, taking into account factors such as the patient's symptoms, the size of the UL and the trend of UL change. Although surgery is an effective treatment option, it is important that patients are thoroughly assessed before surgery, taking into account the physical and psychological trauma of surgery, particularly in postmenopausal women, as well as the socio-economic impact of surgery. However, there is a paucity of research on the risk of UL degeneration in postmenopausal women and no relevant studies of the nomogram have been conducted. In this study, we analysed 29 potential variables of UL in 508 postmenopausal women who underwent hysteromyoma-related surgery (myomectomy or hysterectomy). A simple and easy-to-use prediction nomogram for UL in postmenopausal women using multivariate analysis was developed for the first time. Seven variables （DM,BRASTCANCER, NUMBER, size5 ,LDH, TG, Ca²⁺）were filtered for the nomogram using stepwise regression. This nomogram had a good diagnostic performance (AUC = 0.714) and was validated internally using the bootstrap sampling method. Furthermore, this prediction model showed superior performance when used in the clinical setting based on the results of the DCA. More importantly, compared to the clinical experience model (sym+ changetrend+ size5) in AUC, DCA and NRI, the prediction model we built has greater advantages in predicting the outcome of UL degeneration.

This study found that the likelihood of uterine leiomyoma (UL) degeneration in postmenopausal women was 40.6%, which is lower than the 65-66% reported in previous literature [36,37]. After menopause, serum estrogens and progesterone levels decrease significantly, resulting in a reduced likelihood of UL growth. Therefore, only a few cases of UL have been reported in postmenopausal women [38-41] .Interestingly, researchers[42] have found no discernible difference in the expression of sex steroid hormone receptors between postmenopausal UL that required surgery and those that did not. These findings suggest that sex steroid hormones may play a less important role in the growth of postmenopausal UL. In general, we believe that enlargement of postmenopausal UL is one of the indications of UL degeneration or even malignant transformation. In our study, there is no clear relationship between the change trend of postmenopausal UL and UL degeneration. The main reason may be that the concept of health examination has not yet been popularised in China, so some women, especially postmenopausal women, do not have the habit of timely physical examination. Therefore, we cannot get the true trend of postmenopausal fibroid size change. What's more, postmenopausal women who are admitted to hospital for surgical treatment may have recently developed a hysteromyoma.

The presence or absence of UL-related symptoms is usually considered one of the indications for UL and UL surgery.　However, in our study, we found that in postmenopausal women, the presence or absence of UL-related symptoms had no significant correlation with UL degeneration (OR=0.98, P=0.91). This may be related to the decline in hormone levels in postmenopausal women．UL-related symptoms (such as abnormal vaginal bleeding, abdominal pain, and frequent urination) may have a variety of causes, so it is particularly important to assess the causes of these symptoms in this population. We suggest that the presence or absence of symptoms should not be the main reason for UL surgery.

In this study we found a total of 7 predictors, namely DM, BREASTCANCER, NUMBER, size5, LDH, TG and Ca²⁺. Among the seven factors, a history of breast cancer is a protective factor for UL degeneration. Most breast cancers are hormone-dependent, and their occurrence and development are influenced by hormone levels in the body. Previous studies have shown that patients with UL may have an increased risk of breast cancer because UL and breast cancer have some common risk factors, such as obesity and estrogen disruption, Chuang et al[43-45] found that UL was associated with breast cancer (OR=1.20; 95% CI=1.03-1.40). However, Chiaffarino et al [46] investigated the association between the history of UL and the history of breast cancer in black women and found no obvious association (OR=0.99; 95% CI=0.90-1.08).In our study, postmenopausal women with a history of breast cancer have a reduced risk of UL degeneration with OR=0.33(P=0.099).We speculate that this result may be related to hormones, genes, genetics, etc., which needs further research.

To our knowledge, no large epidemiological study has examined the association between diabetes mellitus (DM) and uterine leiomyoma (UL) degeneration, although some studies have reported associations between diabetes and UL. Wei J et al [47] found no association with diagnosed diabetes (OR=0.9; CI=0.4-2.2), but observed an increased relative odds of 2.0 (0.4-12.6) for women taking medication for diabetes based on very small numbers. In contrast, Wise et al [48] found a reduced risk of UL in patients with diabetes in the Black Women's Health Study (OR=0.77; CL=0.60-0.98). In our study, we found diabetes to be a risk factor for postmenopausal UL degeneration with OR=2.07. In this study, we hypothesise that diabetes characterised by hyperinsulinemia may increase the risk of uterine leiomyoma (UL) degeneration, based on previous in vitro experiments [49] demonstrating its proliferative effects on uterine smooth muscle in culture. In addition, vascular dysfunction is part of the pathogenesis of diabetes, and insulin resistance and diabetes may be associated with cytotoxic effects from accompanying hyperglycaemia and other circulating cytotoxic factors [50], which may have a direct promoting effect on UL degeneration.

Lactate dehydrogenase (LDH) is an important coenzyme in glycolysis and is closely associated with malignant tumours. Serum concentrations of LDH are elevated in patients with malignant tumours of the genital tract and particularly in patients with uterine sarcomas [51,52]. In this study, we found that there was some association between LDH and the development of postmenopausal uterine fibroids (OR=0.01), but further research is needed to clarify this. Clinically, the size of fibroids is usually considered an indication for surgery, but whether this indication is appropriate for postmenopausal women needs further discussion. In this study, we found that fibroids larger than 5 cm in diameter were a risk factor for postmenopausal fibroid degeneration (OR=1.85). This may be due to oedema in the tissue during the history of fibroid degeneration, resulting in a larger fibroid volume. Interestingly, our study found that multiple myomas were a risk factor for fibroid degeneration in postmenopausal women (OR=2.86), and we speculate that this may be related to the relative decrease in blood supply associated with multiple myomas. We also found that serum triglycerides were associated with fibroids in postmenopausal women (OR=1.12), which may be related to the role of small amounts of estrogen in the fat of postmenopausal women.

Of all the laboratory indicators, we were most interested in serum calcium ion concentration. In this study, we found for the first time a certain association between serum calcium and postmenopausal uterine leiomyoma (UL) degeneration (OR=4.1, 95% cl=0.73-23). Calcium is the most abundant mineral in the human body and an essential component of bone structure. It is involved in biological processes such as cell proliferation, apoptosis, ageing and signal transduction, acting as a chemical transmitter or second messenger [53, 54]. The dynamic equilibrium formed by serum and cellular calcium provides the basic environment necessary for cells and organs to maintain their structure and function. The intestinal absorption of calcium is influenced by vitamin D, and the metabolism of vitamin D also regulates the normal physiological level of calcium [55]. Appropriate calcium supplementation is usually recommended for postmenopausal women. However, there is increasing evidence that higher calcium intake may be associated with an increased risk of death in the population [56, 57]. Some studies have suggested that abnormally high blood calcium levels caused by high calcium intake may be one of the reasons for the increased risk of death [58]. Currently, research on blood calcium levels and mortality risk is mainly focused on people with disease [59-62], while the number of studies in healthy people is small and the conclusions are inconsistent [58]. There is also epidemiological evidence that abnormal calcium levels and vitamin D deficiency are risk factors for many chronic diseases, including cancer [63]. Prostate cancer is one of the most discussed cancers in this area, but the conclusions are not consistent. Based on the study of the National Health and Nutrition Examination Survey III (NHNES III) in the United States, Gary et al. discussed in detail the effect of total blood calcium and ionic calcium levels on the risk of prostate cancer incidence and death [64, 65]. The study included 49 cases of fatal prostate cancer. The results showed that the RR value of each 0.1 mmol/L increase in total blood calcium level on the risk of fatal colorectal cancer was 1.50 (95% CI: 1.04-2.17), while the RR value of each 0. 1 mmol/L increase in ionic calcium level on the risk of colorectal cancer death was 1.50 (95% CI: 1.04-2.17), while the RR value of each 0.05 mmol/L increase in ionic calcium level was 1.72 (95% CI: 1.11-2.66) during the 96 months of follow-up.

Evidence from the Swedish AMORIS study and meta-analysis shows that serum calcium levels are negatively associated with breast cancer risk [66]. The AMORIS study included 229,674 participants with baseline data on serum calcium and albumin. During 19 years of follow-up, 10,863 cases of breast cancer were identified. The results showed that the HR of the highest group of uncorrected serum calcium levels was 0.94 (95% CI: 0.88-0.99) compared with the lowest group. The results remained almost unchanged after exposure to albumin-corrected calcium. Combined with two other prospective cohort studies [67, 68], the results of the systematic review showed that blood calcium levels had a protective effect on the risk of breast cancer, and the combined HR was 0.80 (95% CI: 0.66-0.97).

Regarding UL, few studies have focused on the relationship between serum calcium concentration, UL and its degeneration. In this study, we found that serum calcium ion concentration may have some positive correlation with UL degeneration in postmenopausal women (OR=4.01). Here we speculate that it may play a role mainly through the vitD-related pathway, but the specific mechanism needs further investigation. Furthermore, this finding may also provide us with new ideas for follow-up treatment and prevention of UL enlargement and even malignant transformation.

The present study has several strengths. First, the nomogram showed good performance in assessing the risk of uterine leiomyoma (UL) degeneration, allowing more accurate personalised clinical decision-making and monitoring. Second, to our knowledge, our study is the first to focus on predicting the risk of UL degeneration, providing a basis for personalised treatment. Finally, the nomogram was confirmed to be robust by internal bootstrap validation and showed a good positive net benefit by decision curve analysis. In addition, our prediction model has more advantages in accuracy and clinical applicability compared to the clinical experience model.

However, our study has several limitations. First, we only used internal verification methods to assess the clinical applicability of the nomograms without external validation. Second, the retrospective nature of the study inevitably introduces internal bias. Finally, the samples from our cohorts can only be considered representative of the population of Southeast China, and we aim to conduct external validation in a multicentre study.

In conclusion, DM, BRASTCANCER, NUMBER, size5, LDH, TG, Ca²⁺ are independent risk factors for UL degeneration in postmenopausal women. This study has established an easy-to-use nomogram model to predict UL degeneration in postmenopausal women. In addition, this model has better advantages than the clinical experience model. The novel nomogram could provide predictive guidance for the management of UL in postmenopausal women. The visual and personalised model of seven predictors provides clinicians with a simple and intuitive tool for early detection and identification of UL degeneration in postmenopausal women, which may be important in the fight to reduce the physical and mental injury and social and economic pressure of postmenopausal women after hysteromyoma surgery.

Ethics approval and consent to participate：This study was designed as a retrospective review of data from enrolled patients diagnosed with UL by pathology at Women's Hospital School of Medicine Zhejiang University. The study protocol was approved by the Institutional Review Board of Womens' Hospital School of Medicine Zhejiang University under IRB-20230105-R with a waiver of informed consent.

Consent for publication：Not applicable.

Availability of data and materials：The data used for analysis is fully available in the manuscript file without restriction.

Competing interests：The authors have no conflicts of interest to declare.

Funding：The authors gratefully acknowledge the financial supports by the National Key Research and Development Program of China under grant number 2022YFC2704100 and the Zhejiang Provincial Public Welfare Fund under Grant numbers LGF19H040019.

Authors' contributions：Concept and design: Juanqing Li, Xiufeng Huang and Yaping Wang. Analysis and interpretation of data: All authors. Methodological supervision: Juanqing Li and Xiufeng Huang. Statistical analysis: Yaping Wang. Drafting the manuscript: Yaping Wang. Revision of the manuscript: All authors. Critical revision of the manuscript for important intellectual content: Juanqing Li and Xiufeng Huang.

Acknowledgments: We thank these researchers who gave their data for this analysis. It is cheerful to acknowledge their contributions.

Walker CL, Stewart EA. Uterine fibroids: the elephant in the room. Science. 2005; 308(5728):1589–1592.
Giuliani E, As-Sanie S, Marsh EE. Epidemiology and management of uterine fibroids. Int J Gynaecol Obstet. 2020;149(1):3–9. doi:10.1002/ijgo.13102
Wright J. D. et al., "Nationwide trends in the performance of inpatient hysterectomy in the United States," (in eng), Obstet. Gynecol., vol. 122, no. 2 Pt 1, pp. 233–41, Aug 2013
Bhave Chittawar P., Franik S., Pouwer A. W., and Farquhar C., "Minimally invasive surgical techniques versus open myomectomy for uterine fibroids," (in eng), Cochrane Database Syst. Rev., no. 10, p.Cd004638, Oct 21 2014,
Ghosh S, Naftalin J, Imrie R, Hoo W-L. Natural History of Uterine Fibroids: A Radiological Perspective. Current Obstetrics and Gynecology Reports. 2018;7(3):117–21.
Baird D. D., Dunson D. B., Hill M. C., Cousins D., and Schectman J. M., "High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence," (in eng), Am. J. Obstet. Gynecol.,vol. 188, no. 1, pp. 100–7, Jan 2003,
Templeman C, Marshall SF, Clarke CA, Henderson KD, Largent J, Neuhausen S, et al. Risk factors for surgically removed fibroids in a large cohort of teachers. Fertil Steril. 2009;92(4):1436–46
Mu X, Li Y, He L, et al. Prognostic nomogram for adenoid cystic carcinoma in different anatomic sites. Head Neck. 2021;43(1):48–59.
Chen H, Liu CT, Hong CQ, et al. Nomogram based on nutritional and inflammatory indicators for survival prediction of small cell carcinoma of the esophagus. Nutrition. 2021;84:111086.
Zhang W, Fang M, Dong D, et al. Development and validation of a CT-based radiomic nomogram for preoperative prediction of early recurrence in advanced gastric cancer. Radiother Oncol.2020;145:13–20.
Hu Q, Shi Y, Hua ZY, et al. A prediction nomogram for acute kidney injury in very-low-birth-weight infants: a retrospective study.Front Pediatr. 2021;8:575097
Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD Statement. BMC Med. 2015 Jan 6;13:1.
Mandrekar JN. Receiver operating characteristic curve in diagnostic test assessment. J Thorac Oncol. 2010;5(9):1315-1316.
Van Calster B, Wynants L, Verbeek JFM, et al. Reporting and Interpreting Decision Curve Analysis: A Guide for Investigators. Eur Urol. 2018;74(6):796-804.
Thomas LE, O'Brien EC, Piccini JP, D'Agostino RB, Pencina MJ. Application of net reclassification index to non-nested and point-based risk prediction models: a review. Eur Heart J. 2019;40(23):1880-1887.
Mathew RP, Francis S, Jayaram V, Anvarsadath S. Uterine leiomyomas revisited with review of literature. Abdom Radiol (NY). 2021;46(10):4908-4926.
Giuliani E, As-Sanie S, Marsh EE. Epidemiology and management of uterine fibroids. Int J Gynaecol Obstet. 2020;149(1):3-9.
Harrison JE, Weber S, Jakob R, Chute CG. ICD-11: an international classification of diseases for the twenty-first century. BMC Med Inform Decis Mak. 2021;21(Suppl 6):206. Published 2021 Nov 9.
Zhang F, Liu Y, Quan Q, Meng Y, Mu X. Diagnostic Value of Preoperative CA125, LDH and HE4 for Leiomyosarcoma of the Female Reproductive System. Cancer Manag Res. 2021;13:4657-4664. Published 2021 Jun 14.
Korkmaz V, Ozkaya E, Özer Kadife S, Kara F, Kucukozkan T. Investigation of cardiovascular disease risk in women with uterine leiomyomas. Ir J Med Sci. 2016;185(3):689-693.
He Y, Zeng Q, Li X, Liu B, Wang P. The association between subclinical atherosclerosis and uterine fibroids. PLoS One. 2013;8(2):e57089.
Duan Y, Peng Y, Shi X, et al. Correlation Between Platelet-Lymphocyte Ratio and Neutrophil-Lymphocyte Ratio in Patients with Uterine Leiomyoma: A Cross-Sectional Study. J Oncol. 2022;2022:3257887. Published 2022 Aug 11.
Cho HY, Kim K, Kim YB, No JH. Differential diagnosis between uterine sarcoma and leiomyoma using preoperative clinical characteristics. J Obstet Gynaecol Res. 2016;42(3):313-318.
Riley R D, Ensor J, Snell K I E, Harrell F E, Martin G P, Reitsma J B et al. Calculating the sample size required for developing a clinical prediction model BMJ 2020; 368 :m441
Schafer JL, Graham JW. Missing data: our view of the state of the art. Psychol Methods2002;7:147-77.
Group TAM. The Academic Medicine Group. Academic medicine: problems and solutions.BMJ 1989;298:573-9.
Steyerberg EW, van Veen M. Imputation is beneficial for handling missing data in predictivemodels. J Clin Epidemiol 2007;60:979.
Moons KGM, Donders RART, Stijnen T, Harrell FE Jr. Using the outcome for imputation of missing predictor values was preferred. J Clin Epidemiol 2006;59:1092-101.
Schafer JL. Multiple imputation: a primer. Stat Methods Med Res 1999;8:3-15.
Kudo S, Fujimoto M, Sato T, Nagano A. Determination of the optimal number of linked rigid-bodies of the trunk during walking and running based on Akaike's information criterion. Gait Posture. 2020;77:264-268.
Van Calster B, Wynants L, Verbeek JFM, et al. Reporting and Interpreting Decision Curve Analysis: A Guide for Investigators. Eur Urol. 2018;74(6):796-804.
Jewell ES, Maile MD, Engoren M, Elliott M. Net Reclassification Improvement. Anesth Analg. 2016;122(3):818-824.
Wise LA, Palmer JR, Harlow BL, Spiegelman D, Stewart EA, Adams-Campbell LL, Rosenberg L. Reproductive factors, hormonal contraception, and risk of uterine leiomyomata in African-American women: a prospective study. Am J Epidemiol. 2004;159(2):113-123.
Dayal S, Kumar A, Verma A. Clinicopathologic Correlation of Leiomyoma With Clinical Findings and Secondary Changes in a Rural Population of North India. Am J Clin Pathol 2014;141(2):275–9；
Simms-Stewart D, Fletcher H. Counselling Patients with Uterine Fibroids: A Review of the Management and Complications. Obstet Gynecol Int 2012;2012:539365.
Persaud V, Arjoon PD. Uterine leiomyoma. Incidence of degenerative change and a correlation of associated symptoms. Obstet Gynecol. 1970;35(3):432–6.
Naz S, Rehman A, Riyaz A, Jehangir F , Naeem S, Iqbal T. Leiomyoma: its variants and secondary changes a five‑year study. J Ayub Med Coll Abbottabad. 2019;31(2):192–5.
Low SC, Chong CL. A case of cystic leiomyoma mimicking an ovarian malignancy. Ann Acad Med Singap. 2004;33(3):371–4.
Yarwood RL, Arroyo E. Cystic degeneration of a uterine leiomyoma masquerading as a postmenopausal ovarian cyst. A case report. J Reprod Med. 1999;44(7):649–52.
Shrestha R, Khanal R, Aryal MR, Pathak R, Karmacharya P , Naqi M, Murukutla S, Bhatt VR, Gottesman A. Fibroid degeneration in a postmenopausal woman presenting as an acute abdomen. J Community Hosp Intern Med Perspect. 2015;5(1):25917.
Okamoto T, Koshiyama M, Yamamoto K. Rapidly growing leiomyoma in a postmenopausal woman. J Obstet Gynaecol Res. 2004;30(4):316–8.
Possible significance of degeneration and decreased expression of progesterone receptor in postmenopausal uterine leiomyoma
Cheng, T. M.. Taiwan's National Health Insurance System[A]. In: Ok-ma KG, Crivelli L, eds. Six countries, six reform models: The health care reform experiences of Israel, the Netherlands, New Zealand, Singapore, Switzerland, and Taiwan[N]. 2010:171-204.
Chen HH, Huang N, Chen YM, et al. Association between a history of periodontitis and the risk of rheumatoid arthritis: a nationwide, population-based, case-control study. Ann Rheum Dis. 2013;72(7):1206-1211.
Wu QJ, Li YY, Tu C, et al. Association between medical conditions and breast cancer risk in Asians: a nationwide population-based study in China. PLoS One. 2013;8(7):e68704.
Chiaffarino F, Parazzini F, La Vecchia C, et al. History of uterine leiomyomata and incidence of breast cancer. Eur J Cancer Prev. 2000;9(2):89-92.
Wei J, Chiriboga L, Mizuguchi M, Yee H, Mittal K. Expression profile of tuberin and some potential tumorigenic factors in 60 patients with uterine leiomyomata. Mod Pathol. 2005;18:179 –188.
Wise LA, Palmer JR, Stewart EA, Rosenberg L. Polycystic ovary syndrome and risk of uterine leiomyomata. Fertil Steril. 2007;87:1108 –1115.
Ciarmela P, Islam MS, Reis FM, Gray PC, Bloise E, Petraglia F, Castellucci M. Growth factors and myometrium: biological effects in uterine fibroid and possible clinical implications. Hum Reprod Update. 2011;17(6):772-790.
Pittenger GL, Malik RA, Burcus N, Boulton AJ, Vinik AI. Specific fiber deficits in sensorimotor diabetic polyneuropathy correspond to cytotoxicity against neuroblastoma cells of sera from patients with diabetes. Diabetes Care. 1999;22:1839 –1844
Kyriazoglou A, Liontos M, Ziogas DC, et al. Management of uterine sarcomas and prognostic indicators: real world data from a single-institution. BMC Cancer. 2018;18(1):1247. doi:10.1186/ s12885-018-5156-1
Mollo A, Raffone A, Travaglino A, et al. Increased LDH5/LDH1 ratio in preoperative diagnosis of uterine sarcoma with inconclusive MRI and LDH total activity but suggestive CT scan: a case report. BMC Womens Health. 2018;18(1):169.
Smaili S, Hirata H, Ureshino R, Monteforte PT, Morales AP, Muler ML, Terashima J, Oseki K, Rosenstock TR, Lopes GS, et al. Calcium and cell death signaling in neurodegeneration and aging. An Acad Bras Cienc, 2009, 81(3): 467-475.
Farber JL. The role of calcium ions in toxic cell injury. Environ Health Perspect, 1990, 84: 107-111.
McCullough ML, Rodriguez C, Diver WR, Feigelson HS, Stevens VL, Thun MJ, Calle EE. Dairy, calcium, and vitamin D intake and postmenopausal breast cancer risk in the Cancer Prevention Study II Nutrition Cohort. Cancer Epidemiol Biomarkers Prev, 2005, 14(12): 2898-2904.
Michaelsson K, Melhus H, Warensjo Lemming E, Wolk A, Byberg L. Long term calcium intake and rates of all cause and cardiovascular mortality: community based prospective longitudinal cohort study. BMJ, 2013, 346: f228.
Xiao Q, Murphy RA, Houston DK, Harris TB, Chow WH, Park Y. Dietary and supplemental calcium intake and cardiovascular disease mortality: the National Institutes of Health-AARP diet and health study. JAMA Intern Med, 2013, 173(8): 639-646.
Reid IR, Gamble GD, Bolland MJ. Circulating calcium concentrations, vascular disease and mortality: a systematic review. J Intern Med, 2016, 279(6): 524-540.
Jensen AC, Polcwiartek C, Sogaard P, Mortensen RN, Davidsen L, Aldahl M, Eriksen MA, Kragholm K, Torp-Pedersen C, Hansen SM. The Association Between Serum Calcium Levels and Short-Term Mortality in Patients with Chronic Heart Failure. Am J Med, 2019, 132(2): 200-208 e201.
Chen Q, Zhang Y, Ding D, Li D, Yang Y, Li Q, Chen X, Hu G, Ling W. Associations between serum calcium, phosphorus and mortality among patients with coronary heart disease. Eur J Nutr, 2018, 57(7): 2457-2467.
Wald DS, Bangash FA, Morris JK, Wald NJ. Mortality from aortic stenosis: prospective study of serum calcium and phosphate. J Intern Med, 2017, 281(4): 407-411.
Rivara MB, Ravel V, Kalantar-Zadeh K, Streja E, Lau WL, Nissenson AR, Kestenbaum B, de Boer IH, Himmelfarb J, Mehrotra R. Uncorrected and Albumin-Corrected Calcium, Phosphorus, and Mortality in Patients Undergoing Maintenance Dialysis. J Am Soc Nephrol, 2015, 26(7): 1671-1681.doi:10.1681/ASN.2014050472
Peterlik M, Cross HS. Vitamin D and calcium deficits predispose for multiple chronic diseases. Eur J Clin Invest, 2005, 35(5): 290-304.
Schwartz GG, Skinner HG. A prospective study of total and ionized serum calcium and time to fatal prostate cancer. Cancer Epidemiol Biomarkers Prev, 2012, 21(10): 1768-1773.
Hemelrijck M, Hermans R, Michaelsson K, Melvin J, Garmo H, Hammar N, Jungner I, Walldius G, Holmberg L. Serum calcium and incident and fatal prostate cancer in the Swedish AMORIS study. Cancer Causes Control, 2012, 23(8): 1349-1358.
Wulaningsih W, Sagoo HK, Hamza M, Melvin J, Holmberg L, Garmo H, Malmstrom H, Lambe M, Hammar N, Walldius G, et al. Serum Calcium and the Risk of Breast Cancer: Findings from the Swedish AMORIS Study and a Meta-Analysis of Prospective Studies. Int J Mol Sci, 2016, 17(9).
Sprague BL, Skinner HG, Trentham-Dietz A, Lee KE, Klein BE, Klein R. Serum calcium and breast cancer risk in a prospective cohort study. Ann Epidemiol, 2010, 20(1): 82-85；
Almquist M, Manjer J, Bondeson L, Bondeson AG. Serum calcium and breast cancer risk: results from a prospective cohort study of 7,847 women. Cancer Causes Control, 2007, 18(6): 595-602.

Tables 1 to 2 are available in the Supplementary Files section

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Development and validation of a nomogram to predict the risk of postmenopausal uterine leiomyoma degeneration: A retrospective study

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods

3. Results

4. Discussion

5. CONCLUSION

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1