New online dynamic nomograms: recurrence-free and overall survival after resection of endometrial cancer

doi:10.21203/rs.3.rs-3145923/v1

Download PDF

Research Article

New online dynamic nomograms: recurrence-free and overall survival after resection of endometrial cancer

https://doi.org/10.21203/rs.3.rs-3145923/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

This study aimed to propose nomograms based on clinicopathological variables to predict recurrence-free survival (RFS) and overall survival (OS) in patients after surgical resection therapy for endometrial cancer (EC).

Methods

Patients who underwent surgical therapy for EC were identified in this single-institution retrospective cohort study from January 1, 2012, to August 31, 2022. Web-based nomograms were developed to predict RFS and OS following EC resection. C-statistics, calibration plots, and Kaplan-Meier curves were used to assess the discriminative and calibration abilities of the nomograms.

Results

This study included 289 patients who underwent surgery for EC. At a median follow-up of 51.1 (range, 4.1-128.3) months, 13.5% (39 of 289) of the patients showed a relapse or died from their disease. The median age of the participants was 56 years. Approximately one-tenth of the patients (10.7% [31/289]) had non-endometrioid tumor with a median size of 2.8 cm. RFS was predicted using the Akaike data criteria and a backward stepwise selection of clinically relevant variables: positive peritoneal cytology (PC; hazard ratio [HR], 35.58; 95% confidence interval [CI], 3.43-369.39; P = 0.0028), age-adjusted Charlson comorbidity index (AACCI) (HR, 25.51; 95% CI, 8.74–74.49; P < 0.001), FIGO stage III (HR, 3.01; 95% CI, 1.25–7.22; P = 0.0136), FIGO stage IV (HR, 82.92; 95% CI, 22.27-308.76; P < 0.001). Similarly, OS was predicted using depth of myometrial invasion ≥ 1/2 (HR, 1.18; 95% CI, 0.57–2.46; P = 0.6551), AACCI (HR, 20.09; 95% CI, 6.82–59.16; P < 0.001), FIGO stage III (HR, 3.23; 95% CI, 1.35–7.72; P = 0.0083), FIGO stage IV (HR, 81.56; 95% CI, 20.49-324.57; P < 0.001). Using C-statistics, the nomograms showed good predictive capability and positive discriminative ability and calibration (RFS: 0.895 and OS: 0.891).

Conclusions

Independent factors were chosen to predict survival and recurrence risk following surgical therapy for EC. The nomograms performed well in the internal validation when patients were stratified into prognostic groups.

Oncology

Dynamic nomogram

Recurrence-free survival

Overall survival

Endometrial cancer

FIGO staging

lymphatic vessel space invasion

Endometrial cancer (EC), which accounted for an estimated 417,000 new cases and 97,000 fatalities among women in 2020, is the sixth most frequently reported cancer in women. Compared with countries with low human development index (HDI) countries, the incidence rates vary by a factor of ten in high HDI countries. Globally, an estimated 47% increase in death rates due to population growth and aging, which is projected to occur by 2040, may exacerbate the situation (1). Fortunately, most patients with EC survive the initial diagnosis, making them the largest group of survivors of gynecological malignancies. However, Relapse continues to be the primary cause of death in patients with EC (2). Currently, most models of EC recurrence rely on conventional clinicopathological variables (3). The nomogram model developed by Ouldamer predicts poor prognosis based on histological type and grade, age, lymphatic vessel space invasion (LVSI), FIGO stage, as well as surgical nodal staging in patients with stages I–III EC (2). Takahashi et al. recently developed a scoring system for stages I-II EC recurrence according to various factors such as age, pathology, cervical stromal invasion, and peritoneal cytology (4). Nevertheless, additional predictive indicators can be applied to optimize such models (5).

A reliable prognostic marker is essential for physicians to decide on the adjuvant therapy required and follow-up for patients after cancer resection. For general survival predictions, FIGO or NCCN classification may be helpful (6, 7), and individual patient prognosis may not be as accurate as using risk stratification systems. With the advent of individualized tumor treatment, nomograms have the advantage of being safe and easy to understand as a statistical tool that can combine multiple factors to help physicians make individualized patient assessment and treatment and follow-up choices. To our knowledge, there is no such dynamic prediction model for endometrial cancer. Therefore, this study aimed to develop an online dynamic nomogram to help physicians provide personalized treatment and follow-up guidance.

Data collection and patient population

Through a single-institution retrospective cohort study, patients who underwent surgical therapy for EC at Dongyang People’s Hospital of Wenzhou Medical University, Zhejiang, China were identified from January 1, 2012, to August 31, 2022. The follow-up termination date was December 31, 2022. Data were collected using the medical record mining software of Shanghai Le9 Healthcare Technology Co., Ltd (Le9 Healthcare Technology, Shanghai, China). This study included only patients who underwent surgery as a treatment. Patients aged < 18 years and those with missing data were excluded from the study. To prevent the incorporation of postoperative complications, patients who did not survive for one month postoperatively were excluded. Finally, 289 patients were included in the cohort analysis based on the inclusion and exclusion criteria. This study was approved by the Ethics Committee of Dongyang People’s Hospital of Wenzhou Medical University (DRY-2023-YX-078), and written informed consent was obtained from individual participants. Since this was a retrospective study, no additional consensus was required.

A survey of demographics and clinicopathology was conducted, including age at surgery, presence of adenomyosis, menopause, age-adjusted Charlson comorbidity index (AACCI), FIGO stage, histologic type, primary tumor diameter, LVSI, histologic grade, peritoneal cytology, depth of myometrial invasion, presence or absence of lymph node metastases, and cervical stromal invasion. The resection specimens were measured based on their maximum diameters. The tumors were segmented as endometrioid or non-endometrioid. The depth of myometrial invasion, lymph node metastasis, and cervical stromal invasion status were determined according to the final pathological assessment. LVSI was classified as negative, focal, or substantial. Outcome status was defined as cancer recurrence or cancer-related death, and the primary outcomes were long-term OS and RFS.

Statistical analysis

Medians and interquartile ranges (IQRs) were provided for continuous variables, and complete values were provided for categorical variables, except as otherwise specified. RFS and OS were determined using Kaplan-Meier curves, and comparisons were made using the log-rank test. Clinical and pathological factors related to survival risk and recurrence were evaluated a priori (8–11). Using a correlation matrix, variable collinearity and interactions were examined, including interactions between age, adenomyosis, menopause, AACCI, FIGO stage, histological type, primary tumor diameter, LVSI, histological grade, peritoneal cytology, depth of myometrial invasion, lymph node metastases, and cervical stromal invasion. Multivariate analysis did not include an interaction term since there was no significant interaction. Linear correlation between continuous predictors (i.e., age and primary tumor diameter) and risk of recurrence or death was evaluated using restricted cubic splines to categorize the continuous predictors (12). Age (56 years) and primary tumor diameter (2.8 cm) were associated with increased risk of recurrence and death. Nomograms were constructed with tumor size and age as categorical variables for consistency according to prior research and clinical practice guidelines (< 2 vs. ≥2 cm and < 56 vs. ≥56 years, respectively) (6, 7, 11).The association of clinicopathological factors with RFS and OS was identified using a Cox proportional risk regression model. Backward stepwise selection using the Akaike information criterion (AIC) was employed to develop multivariable Cox proportional hazards regression models. Hazard ratios (HRs) with 95% confidence intervals (CIs) were reported (13). Using statistical software, specific factors were included in the nomograms to predict the 3-, 5-, and 10-year RFS and OS values following EC surgical resection with curative intent (RMS in R, version 4.1.1; http://www.r-project.org). Linear predictors were defined by applying regression coefficients to each individual observation (14).

The performance of the model was evaluated based on its ability to predict correctly individual outcomes (discriminative ability) and point estimates of the survival function (calibration). Harrell et al. evaluated nomogram performance using C-statistics (15), which is equivalent to the area under the curve (AUC) of the receiver operating characteristic (ROC) curve and measures the agreement between the anticipated and actual results (15). To determine the effectiveness of calibration, a calibration plot was created, which represents the correlation between observed outcomes and predicted probabilities using a bootstrapped sample. Ideally, the line of prediction should be at a 45° angle. Clinical utility of the model was measured using decision curve analysis (DCA) (16). Kaplan-Meier curves were drawn across the threshold for patients classified by the nomogram scores in the dataset to assess calibration further. Overfitting was quantified by bootstrapped resampling of the model, and model improvement was judged using the integrated discrimination improvement (IDI) index (17). Finally, a website was established for clinicians to access and utilize the proposed paradigm. Two-tailed tests were employed, and the findings were considered significant at P < 0.05.

Demographic and clinicopathological features

The flow chart of the study is shown in Fig. 1. The median age was 56 years (IQR, 51–61 years), and median AACCI score was 1 (1–3). In total, 60.9% (176/289) of patients were menopausal, 13.8% (40/289) had adenomyosis. Only 0.7% (2/289) of the patients were positive for peritoneal cytology, and 5.5% (16/289) were positive for lymph node metastases. Regarding histologic type, 10.7% (31/289) of the participants had non-endometrioid tumor. The overall pathology report indicated that the majority of the participants had early-stage tumors were early, with 83.4% (241/289) at stage I, while 6.2% (18/289) had stage II, 37.8% (56/148) had stage III, and 10.8% (16/148) had stage IV cancer. Cervical stromal invasion was detected in 10.0% (29/289) of the patients. The median primary tumor diameter was 2.8 cm (IQR, 1.5–4.2 cm), and 19.0% (55/289) had deep myometrial invasion. High histologic grade was observed in 16.6% (48/289) of the patients. Among all patients, 11.1% (32/288) were LVSI-positive: 9.7% (28/289) had focal LVSI, while only 1.4% (4/289) had substantial LVSI (Additional file 1).

At a median follow-up of 51.1 (range, 4.1-128.3) months, 13.5% (39 of 289) of the patients showed a relapse or died from their disease. The unadjusted median RFS was 47.1 months (95% CI, 7.6-108.3), and the unadjusted median OS was 51.1 months (95% CI, 8.6-108.9). The 3- and 5-year RFS rates were 89.3% (95% CI, 85.7%-93.3%) and 82.6% (95% CI, 77.5%-88.0%), respectively, whereas the OS rates were 93.7% (95% CI, 90.7%-96.8%) and 83.9% (95% CI, 78.7%-89.5%), respectively (Additional file 2).

Model specifications and predictors of RFS and OS

Candidate variables for the predictive models were selected based on tumor characteristics of clinical importance and demographics. The following three variables with the strongest correlation with OS were identified in the Cox proportional risk regression model through AIC stepwise backward selection: AACCI score > 2, FIGO stage, and a depth of myometrial invasion ≥ 1/2 (Table 1). Similarly, the following three variables were associated with recurrence risk: AACCI score > 2, FIGO stage, and positive peritoneal cytology (Table 2). On multivariable analysis, AACCI score > 2 (HR, 52.08; 95% CI, 12.35-219.61; P < 0.001), positive peritoneal cytology (HR, 35.06; 95% CI, 1.12-1095.64; P = 0.0428), and FIGO stage IV (HR, 138.33; 95% CI, 17.38-1101.05; P < 0.001) were each separately related to RFS.

Table 1

Cox Proportional Hazards Regression Model Showing the Association of Variables with Overall Survival
	Univariable		Multivariable
Variable	HR (95% CI)	P Value	HR (95% CI)	P Value
Factors Selected
AACCI
≤ 2	1 [Reference]	NA	1 [Reference]	NA
> 2	23.07 (8.17–65.13)	< 0.001	93.63 ( 14.87–589.44 )	< 0.001
FIGO
I	1 [Reference]	NA	1 [Reference]	NA
II	2.79 (1.05–7.46)	0.04	0.21 ( 0.03–1.69 )	0.1437
III	4.71 (2.07–10.71)	< 0.001	1.23 ( 0.35–4.36 )	0.7441
IV	92 (28.97–292.2)	< 0.001	608.26 ( 73.41–5039.66 )	< 0.001
Depth of myometrial invasion
< 1/2	1 [Reference]	NA	1 [Reference]	NA
≥ 1/2	4.32 (2.28–8.17)	< 0.001	1 ( 0.46–2.19 )	0.995
Factors Not Selected
Histologic grade
1–2	1 [Reference]	NA	1 [Reference]	NA
3	4.98 (2.63–9.42)	< 0.001	2.19 ( 0.88–5.42 )	0.0912
Primary tumor diameter,cm
< 2	1 [Reference]	NA	1 [Reference]
≥ 2	5.29 (1.63–17.22)	0.006	6.28 ( 1.33–29.72 )	0.0205
Age at surgery,y
< 56	1 [Reference]	NA	1 [Reference]	NA
≥ 56	2.42 (1.2–4.88)	0.014	0.42 ( 0.17–1.04 )	0.0608
Cervical stromal invasion
No	1 [Reference]	NA	1 [Reference]	NA
Yes	4.33 (2.18–8.59)	< 0.001	6.15 ( 1.02–36.96 )	0.0472
Histologic type
Endometrioid	1 [Reference]	NA	1 [Reference]	NA
Non-endometrioid	4.52 (2.27–8.99)	< 0.001	1.08 ( 0.4–2.95 )	0.8815
Lymph node metastases
No	1 [Reference]	NA	1 [Reference]	NA
Yes	5.34 (2.33–12.25)	< 0.001	2.08 ( 0.52–8.32 )	0.3021
LVSI
Negative	1 [Reference]	NA	1 [Reference]	NA
Focal	5.13 (2.49–10.54)	< 0.001	2.25 ( 0.93–5.45 )	0.0724
Substantial	17.42 (5.96–50.92)	< 0.001	0.05 ( 0.01–0.38 )	0.0043
Peritoneal cytology
Negative	1 [Reference]	NA	1 [Reference]	NA
Positive	21.79 (2.67-178.07)	0.004	1.68 ( 0.09–30.91 )	0.7261
Menopause
No	1 [Reference]	NA	1 [Reference]	NA
Yes	2.16 (0.99–4.71)	0.054	NA	NA
Adenomyosis
No	1 [Reference]	NA	1 [Reference]	NA
Yes	0.19 (0.03–1.37)	0.099	NA	NA
AACCI: Age-Adjusted Charlson Comorbidity Index; FIGO: Federation of Gynecology and Obstetrics; HR: Hazard ratio; LVSI: lymphatic vessel space invasion; NA, not applicable.

Table 2

Cox Proportional Hazards Regression Model Showing the Association of Variables with Recurrence-Free Survival
	Univariable		Multivariable
Variable	HR (95% CI)	P Value	HR (95% CI)	P Value
Factors Selected
AACCI
≤ 2	1 [Reference]	NA	1 [Reference]	NA
> 2	25.65 (9.1–72.3)	< 0.001	52.08 ( 12.35–219.61 )	< 0.001
FIGO
I	1 [Reference]	NA	1 [Reference]	NA
II	3.11 (1.17–8.25)	0.023	1.22 ( 0.18–8.25 )	0.837
III	4.6 (2.03–10.42)	< 0.001	0.95 ( 0.23–3.96 )	0.9434
IV	102.59 (29.66-354.82)	< 0.001	138.33 ( 17.38–1101.05 )	< 0.001
Peritoneal cytology
Negative	1 [Reference]	NA	1 [Reference]	NA
Positive	14.11 (1.87-106.36)	0.01	35.06 ( 1.12–1095.64 )	0.0428
Factors Not Selected
Age at surgery,y
< 56	1 [Reference]	NA	1 [Reference]	NA
≥ 56	2.65 (1.32–5.33)	0.006	0.36 ( 0.15–0.88 )	0.025
Histologic grade
1–2	1 [Reference]	NA	1 [Reference]	NA
3	5.8 (3.09–10.87)	< 0.001	2.95 ( 1.1–7.92 )	0.0314
Cervical stromal invasion
No	1 [Reference]	NA	1 [Reference]	NA
Yes	4.73 (2.39–9.35)	< 0.001	1.68 ( 0.35–8.11 )	0.5194
Depth of myometrial invasion
< 1/2	1 [Reference]	NA	1 [Reference]	NA
≥ 1/2	5.36 (2.85–10.06)	< 0.001	1.35 ( 0.61–3.02 )	0.4608
Histologic type
Endometrioid	1 [Reference]	NA	1 [Reference]	NA
Non-endometrioid	4.34 (2.19–8.58)	< 0.001	0.89 ( 0.28–2.89 )	0.8523
Lymph node metastases
No	1 [Reference]	NA	1 [Reference]	NA
Yes	4.83 (2.12-11)	< 0.001	2.96 ( 0.66–13.21 )	0.155
LVSI
Negative	1 [Reference]	NA	1 [Reference]	NA
Focal	6.58 (3.26–13.27)	< 0.001	3.03 ( 1.34–6.85 )	0.0077
Substantial	35.8 (11.58–110.7)	< 0.001	0.24 ( 0.03–2.17 )	0.2046
Primary tumor diameter,cm
< 2	1 [Reference]	NA	1 [Reference]	NA
≥ 2	4.06 (1.44–11.43)	0.008	2.37 ( 0.71–7.87 )	0.1604
Menopause
No	1 [Reference]	NA	1 [Reference]	NA
Yes	2.08 (0.98–4.37)	0.055	NA	NA
Adenomyosis
No	1 [Reference]	NA	1 [Reference]	NA
Yes	0.16 (0.02–1.18)	0.073	NA	NA
AACCI: Age-Adjusted Charlson Comorbidity Index; FIGO: Federation of Gynecology and Obstetrics; HR: Hazard ratio; LVSI: lymphatic vessel space invasion; NA, not applicable.

Nomograms and model performance

The nomograms for predicting RFS and OS after surgical resection in patients with EC are shown in Fig. 2. The predictive nomogram for RFS was established using the following three independent predictive variables: AACCI score (≤ 2 or > 2), peritoneal cytology (negative or positive), and FIGO stage (I-IV). The predictive nomogram for OS was established based to the following three prognostic factors: AACCI score (≤ 2 or > 2), FIGO stage (I, II, III or IV), depth of myometrial invasion (< 1/2 or ≥ 1/2). A worse prognosis was related to greater total scores, which were quantified by summing all the points for each element in the nomogram. For instance, a patient with high AACCI score (> 2), early FIGO stage (II), and negative peritoneal cytology could score 83 points (73, 10, and 0 points for each item, respectively) for projected 3-,5- and 10-year RFS rates of 69.0%, 45.0%, and 38.0%, respectively. Similarly, a patient with high AACCI score (> 2), early FIGO stage (II), and depth of myometrial invasion < 1/2 could score 74 points (68, 6, and 0 points for each item, respectively). The 3-, 5-, and 10-year OS rates in this patient were 88.0%, 64.0%, and 38.0%, respectively. To evaluate calibration further, Kaplan-Meier curves were plotted over the cutoff of patients segmented by the nomogram scores in the dataset (Fig. 3). Patients with the least projected 5-year RFS rate (high) showed a significantly poorer prognosis (41.4%, 5-year RFS), compared with patients with low points (99.2%, 5-year RFS) (P < .00001). Likewise, patients with reduced projected 5-year OS rate (high) showed a significantly poorer prognosis (41.6%, 5-year OS) than those with low scores (92.4%, 5-year OS) (P < .00001). C-statistics was used to identify the discriminant ability of the RFS and OS models (0.895 for RFS and 0.891 for OS). Model accuracy and potential model overfitting were examined by bootstrap validation with 1000 resamplings. A 39-sample bootstrap calibration plot was used to predict the 5-year RFS and OS (Additional file 3). Owing to the general prognostic preference for the collected data, our correction curve deviated significantly from the diagonal when the predicted probability of actual occurrence was low. Therefore, our model may not be suitable for identifying patients with a low probability of actual occurrence. DCA is a straightforward technique that can be used to evaluate the feasibility and clinical value of predictive technologies. DCA showed that our constructed nomograms could better predict RFS and OS in patients with EC (Fig. 4).

Furthermore, we established a website for clinicians to use the proposed models for individualized risk prediction. The online dynamic nomogram for RFS is accessible at https://dyyyfkhz.shinyapps.io/DynNomRFSapp/ and that for OS at https://dyyyfkhz.shinyapps.io/DynNomOSapp/ (Fig. 5A and B).

While total surgical resection remains the preferred course of action for EC, 65–85% of patients experience recurrence after surgery (18). In recent years, with advances in detection and treatment technology, the recurrence rate has decreased (19). Approximately 70% of patients with EC have been reported to have stage I disease, with a 5-year OS rate ranging from 74–91%. However, the 5-year OS rate has been reported as 57–66% for stage III and only 20–26% for stage IV, suggesting that metastasis is associated with a poorer outcome (20, 21). The prognostic heterogeneity following EC surgery varies greatly by tumor stage (22). Accurate prognosis following surgery for EC is essential to inform patients accurately of their long-term prognosis and to select patients for adjuvant therapy. Although the FIGO staging system is currently the most commonly used staging system, the reported prognostic factors vary, and the best way to stratify the risk of EC patients is unknown (23). Herein, two nomograms were created to predict RFS and OS in individuals following EC resection, based on patient-related factors. This information makes it possible to decide on what individualized treatments and monitoring methods to pursue and to determine a patient's prognosis. It is important to note that the nomograms created herein were constructed using information obtained from nearly a decade's worth of record of patients with EC who underwent surgery. Moreover, unlike the previously suggested nomograms based on the Surveillance, Epidemiology, and End Results program database (11), the effectiveness of the current nomograms has been thoroughly evaluated and internally validated. In addition, the present study included several variables mentioned previously that affect prognoses (24–26). Various factors have been identified to be associated with outcomes, with little agreement on which factors determine prognosis (10, 26). For example, some studies (10, 27, 28) have reported age, tumor stage, grade, LVSI, myometrial invasion, cervical stromal invasion, and tumor size. However, other researchers found no link between long-term survival and LVSI, histological type, menopause, and adenomyosis (29–32). Similarly, we found no association between menopause, adenomyosis, and RFS or OS in the current study (Tables 1 and 2). Comparatively, several studies on EC have revealed that LVSI, advanced FIGO stage (III/IV), histologic grade, myometrial invasion depth, cervical stromal invasion, and tumor size at diagnosis are associated with poorer outcomes (8, 27, 28, 33, 34). In the present study, patients with an AACCI score > 2 and primary tumor diameter > 2 cm had 93-fold and 6-fold higher mortality risks, respectively, and patients with an AACCI score > 2, stage IV disease, and high histologic grade had 52-fold, 138-fold, and 3-fold higher risks of recurrence, respectively. AACCI score was an independent risk variable for OS and RFS, consistent with previous conclusions (35).

The prognosis for patients with tumors such as EC can be heterogeneous, making accurate risk stratification of these patients important. The nomograms may provide more individualized prognostic information to patients rather than the FIGO staging system, which is determined using data from the population. Li et al. (11)suggested a nomogram for patients with EC that considered age, FIGO stage, histologic grade, histologic type, distant metastasis, and tumor stage in the model, and we consider this model to lack convenience and practicality. Hence, we limited our model factors to patients undergoing surgical treatment for EC. Nomograms stratified by cutoffs helped identify distinct groups of patients with different relapse and mortality risks (Fig. 2). Moreover, both nomograms showed good discriminatory capability, with C-statistics of 0.895 for predicting RFS and 0.891 for predicting OS (Fig. 1). The nomograms also made median 5-year survival predictions, similar to those of the Kaplan-Meier curve survival. Taken together, these data suggest that the proposed nomograms can offer patients with postoperative EC information on the risk of recurrence and survival. Compared with previous models (36), the IDI evaluated the overall improvement of the model: 0.189 for 5-year and 0.173 for 10-year RFS prediction.

This study has several limitations. First, this was a single-institution retrospective study. Patient differences will affect the universality and prediction accuracy of the model. Second, the prognostic significance of OS was not entirely accurate because data that may affect patient survival, such as chemotherapy, radiotherapy, and surgery, were excluded from the nomogram. Moreover, next generation sequencing molecular classification has demonstrated superiority over conventional pathological assessment of grade and histotype (37). Future modeling should consider these factors. Finally, although the proposed nomograms were internally validated using bootstrapping, additional studies are required for external validation.

In conclusion, this study is the first to develop online nomograms based on three variables to predict RFS and OS dynamically in women with EC. The proposed nomograms can be used to stratify patients into different predictive groups based on relapse and long-term outcomes. Clinicians can use our model through the web calculator we created. Additionally, the nomograms performed well when internally validated. To demonstrate their usefulness in assessing long-term prognosis following therapeutic EC excision, external validation of the proposed nomograms is required.

AACCI age-adjusted Charlson comorbidity index

DCA decision curve analysis

EC endometrial cancer

HDI human development index

LVSI lymphatic vessel space invasion

OS overall survival

Ethics approval and consent to participate

Institutional review board approval was obtained prior to data abstraction by the Ethics Committee of Dongyang People’s Hospital, and written informed consent was obtained from individual participants.

Consent for publication

Not applicable.

Availability of data and materials

The authors will make the raw data used to support the conclusions of this article available with no excessive delay.

Competing interests

The authors declare no competing interests.

Funding

None.

Authors’ contributions

ZH and JL designed, drafted and revised the paper, and analyzed the data. JD supervised and submitted the paper for publication. Each author contributed to the article and approved the submitted version.

Acknowledgments

We would like to thank Editage (www.editage.cn) for English language editing.

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209-49.
Ouldamer L, Bendifallah S, Body G, Touboul C, Graesslin O, Raimond E, et al. Predicting poor prognosis recurrence in women with endometrial cancer: a nomogram developed by the FRANCOGYN study group. Br J Cancer. 2016;115:1296-303.
Awtrey CS. Nomograms for predicting endometrial cancer recurrence. Gynecol Oncol. 2012;125:513-4.
Takahashi K, Yunokawa M, Sasada S, Takehara Y, Miyasaka N, Kato T, et al. A novel prediction score for predicting the baseline risk of recurrence of stage I-II endometrial carcinoma. J Gynecol Oncol. 2019;30:e8.
van der Putten LJM, Visser NCM, van de Vijver K, Santacana M, Bronsert P, Bulten J, et al. Added value of estrogen receptor, progesterone receptor, and l1 cell adhesion molecule expression to histology-based endometrial carcinoma recurrence prediction models: an ENITEC collaboration study. Int J Gynecol Cancer. 2018;28:514-23.
Koskas M, Amant F, Mirza MR, Creutzberg CL. Cancer of the corpus uteri: 2021 update. Int J Gynaecol Obstet. 2021;155:45-60.
Abu-Rustum N, Yashar C, Arend R, Barber E, Bradley K, Brooks R, et al. Uterine Neoplasms, Version 1.2023, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw. 2021;21:181-209.
Tortorella L, Restaino S, Zannoni GF, Vizzielli G, Chiantera V, Cappuccio S, et al. Substantial lymph-vascular space invasion (LVSI) as predictor of distant relapse and poor prognosis in low-risk early-stage endometrial cancer. J Gynecol Oncol. 2021;32:9.
Arslan SA, Avci GG, Akkas EA, Guney Y. Improved disease-free survival with adjuvant radiotherapy in early-stage endometrial cancer: 10-year outcome analysis. J Contemp Brachytherapy. 2020;12:572-8.
Feldman A, Chaugle S, Burmeister C, Munkarah A, Elshaikh MA. A matched analysis on the prognostic impact of race on survival endpoints of women with early-stage endometrial cancer. Gynecol Obstet Invest. 2019;84:283-9.
Li R, Yue Q. A nomogram for predicting overall survival in patients with endometrial carcinoma: a SEER-based study. Int J Gynaecol Obstet. 2022;161:744-50.
Harrell FE, Jr., Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15:361-87.
Simon R. Confidence intervals for reporting results of clinical trials. Ann Intern Med. 1986;105:429-35.
Shariat SF, Karakiewicz PI, Suardi N, Kattan MW. Comparison of nomograms with other methods for predicting outcomes in prostate cancer: a critical analysis of the literature. Clin Cancer Res. 2008;14:4400-7.
Harrell FE, Jr., Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the yield of medical tests. JAMA. 1982;247:2543-6.
Zhang Z, Rousson V, Lee WC, Ferdynus C, Chen M, Qian X, et al. Decision curve analysis: a technical note. Ann Transl Med. 2018;6:308.
Steyerberg EW, Vickers AJ, Cook NR, Gerds T, Gonen M, Obuchowski N, et al. Assessing the performance of prediction models: a framework for traditional and novel measures. Epidemiology. 2010;21:128-38.
Ben Arie A, Lavie O, Gdalevich M, Voldarsky M, Barak F, Schneider D, et al. Temporal pattern of recurrence of stage I endometrial cancer in relation to histological risk factors. Eur J Surg Oncol. 2012;38:166-9.
Ouldamer L, Bendifallah S, Body G, Touboul C, Graesslin O, Raimond E, et al. Incidence, patterns and prognosis of first distant recurrence after surgically treated early stage endometrial cancer: results from the multicentre FRANCOGYN study group. Eur J Surg Oncol. 2019;45: 672-8.
Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021. CA Cancer J Clin. 2021;71:7-33.
Creasman WT, Odicino F, Maisonneuve P, Quinn MA, Beller U, Benedet JL, et al. Carcinoma of the corpus uteri. FIGO 26th annual report on the results of treatment in gynecological cancer. Int J Gynaecol Obstet. 206;95:S105-43.
Bendifallah S, Genin AS, Naoura I, Buffet NC, Chapelon FC, Haddad B, et al. A nomogram for predicting lymph node metastasis of presumed stage I and II endometrial cancer. Am J Obstet Gynecol. 2012;207:197.e1-197.e8.
Oaknin A, Bosse TJ, Creutzberg CL, Giornelli G, Harter P, Joly F, et al. Endometrial cancer: ESMO clinical practice guideline for diagnosis, treatment and follow-up. Ann Oncol. 2022;33:860-77.
Raffone A, Travaglino A, Raimondo D, Neola D, Maletta M, Santoro A, et al. Lymphovascular space invasion in endometrial carcinoma: a prognostic factor independent from molecular signature. Gynecol Oncol. 2022;165:192-7.
Capozzi VA, Sozzi G, Uccella S, Ceni V, Cianciolo A, Gambino G, et al. Novel preoperative predictive score to evaluate lymphovascular space involvement in endometrial cancer: an aid to the sentinel lymph node algorithm. Int J Gynecol Cancer. 2020;30:806-12.
Sahin H, Meydanli MM, Sari ME, Kocaman E, Cuylan ZF, Yalcin I, et al. Recurrence patterns and prognostic factors in lymphovascular space invasion-positive endometrioid endometrial cancer surgically confined to the uterus. Taiwan J Obstet Gynecol. 2019;58:82-9.
Euscher ED, Bassett R, Duose DY, Lan C, Wistuba I, Ramondetta L, et al. Mesonephric-like carcinoma of the endometrium: a subset of endometrial carcinoma with an aggressive behavior. Am J Surg Pathol. 2020;44:429-43.
Jin X, Shen C, Yang X, Yu Y, Wang J, Che X. Association of tumor size with myometrial invasion, lymphovascular space invasion, lymph node metastasis, and recurrence in endometrial cancer: a meta-analysis of 40 studies with 53,276 patients. Front Oncol. 2022;12:881850.
Okugawa K, Yahata H, Hachisuga K, Tomonobe H, Yasutake N, Kodama K, et al. Evaluation of clinical significance of lymphovascular space invasion in stage IA endometrial cancer. Oncology. 2022;100:195-202.
Iida Y, Komatsu H, Okawa M, Osaku D, Nosaka K, Sato S, et al. Is adjuvant chemotherapy necessary in patients with early endometrial cancer? Yonago Acta Med. 2022;65:82-7.
Yu Z, Wei S, Zhang J, Shi R, An L, Feng D, et al. Development and validation of a novel prognostic model for endometrial cancer based on clinical characteristics. Cancer Manag Res. 2021;13:8879-86.
Casadio P, Raffone A, Maletta M, Travaglino A, Raimondo D, Raimondo I, et al. Clinical characteristics of patients with endometrial cancer and adenomyosis. Cancers (Basel). 2021;13:4918.
Ghanem AI, Modh A, Burmeister C, Munkarah AR, Elshaikh MA. Does age-adjusted Charlson comorbidity score impact survival endpoints in women with federation of gynecology and obstetrics-stage III endometrial cancer? Gynecol Obstet Invest. 2018;83:290-8.
Xie G, Qi C, Yang W, Wang R, Yang L, Shang L, et al. Competing risk nomogram predicting cancer-specific mortality for endometrial cancer patients treated with hysterectomy. Cancer Med. 2021;10:3205-13.
Di Donato V, D'Oria O, Giannini A, Bogani G, Fischetti M, Santangelo G, et al. Age-adjusted Charlson comorbidity index predicts survival in endometrial cancer patients. Gynecol Obstet Invest. 2022;87:191-9.
Cheng Y, Dong Y, Tian W, Zhang H, Li X, Wang Z, et al. Nomogram for predicting recurrence-free survival in Chinese women with endometrial cancer after initial therapy: external validation. J Oncol. 2020. doi:10.1155/2020/2363545.
Rao Q, Liao J, Li Y, Zhang X, Xu G, Zhu C, et al. Application of NGS molecular classification in the diagnosis of endometrial carcinoma: a supplement to traditional pathological diagnosis. Cancer Med 2022. doi:10.1002/cam4.5363.

Additionalfiles.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

New online dynamic nomograms: recurrence-free and overall survival after resection of endometrial cancer

Status:

Version 1

Abstract

Figures

Background

Methods

Data collection and patient population

Statistical analysis

Results

Demographic and clinicopathological features

Model specifications and predictors of RFS and OS

Nomograms and model performance

Discussion

List of abbreviations

Declarations

References

Supplementary Files

Status:

Version 1