Abstract

Objectives

Endometrial cancer is the second most common gynaecologic cancer in women worldwide. Due to the biology of endometrial cancer, most patients are overweight. Standard uptake value (SUV) measurements are known to vary depending on the patient’s body weight. We aimed to evaluate whether lean body mass-adjusted ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) parameters [lean body mass-corrected standardized uptake value (SUL)-based metrics] are superior to conventional SUV-based parameters for predicting adverse histopathologic features in overweight patients with endometrial cancer, and to explore their association with progression-free survival (PFS).

Methods

In this retrospective single-center study, 73 overweight patients with endometrial cancer who underwent preoperative ¹⁸F-FDG PET/CT followed by primary surgery were included. SUV- and SUL-based metabolic parameters were compared with histopathologic risk factors. ROC analyses were performed to determine discriminatory performance and optimal cut-off values. PFS was assessed using Kaplan-Meier analysis.

Results

SUVmean (SUL_mean) demonstrated significant associations with deep myometrial invasion, lymphovascular space invasion, lymph node involvement, and higher tumor grade. In ROC analysis, SUL_mean showed moderate discriminatory ability for lymph node involvement (area under the curve: 0.78). However, PFS did not differ significantly between groups stratified by the ROC-derived SUL_mean cut-off (log-rank p=0.46).

Conclusion

Lean body mass-adjusted PET parameters, particularly SUL_mean, were more strongly associated with adverse histopathologic features than conventional SUV metrics in overweight patients with endometrial cancer. Although SUL-based parameters may contribute to preoperative risk assessment, their prognostic value for survival remains uncertain.

Introduction

Endometrial cancer is the second most common gynecological malignancy among women worldwide (1, 2). However, in terms of mortality, it ranks lower, which can largely be attributed to the fact that endometrial cancer is typically diagnosed at an early stage (3). Obesity is a well-established risk factor for endometrial cancer (1, 4). With increasing obesity rates worldwide, the incidence of endometrial cancer has been rising steadily, posing a growing public health challenge. Importantly, obesity not only increases the risk of developing endometrial cancer but may also influence tumor biology, imaging characteristics, and perioperative risk assessment, thereby complicating preoperative evaluation (5).

Histopathological features are essential for prognostication and clinical management in endometrial cancer. Histologic grade and myometrial invasion depth are key determinants of tumor aggressiveness and recurrence risk. Lymphovascular space invasion (LVSI), and lymph node involvement are strong predictors of extrauterine spread and poor survival. Cervical involvement contributes to FIGO staging and surgical planning. p53 mutations reflect tumor proliferation and molecular subtype, particularly in high-grade lesions. Additionally, estrogen receptor/progesterone receptor (ER/PR) receptor expression is often associated with hormone responsiveness and favorable prognosis. Evaluating these parameters provides critical insight into disease behavior and supports individualized treatment strategies. Accurate preoperative identification of these adverse pathological features remains clinically relevant, as it may influence the extent of surgical staging, lymph node assessment, and adjuvant treatment decisions (6).

In the diagnostic and staging process of endometrial cancer, ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) imaging is widely used to assess tumor metabolism and disease burden. Traditional parameters such as the standardized uptake value (SUV) have long played a critical role in evaluating tumor metabolic activity. The metabolic data obtained from ¹⁸F-FDG PET/CT can help guide more accurate and personalized treatment decisions, ultimately supporting cancer management and improving patient outcomes (7). Although SUV-based metrics provide a convenient and reproducible assessment of tumor metabolism, their accuracy can be compromised in populations with altered body composition, such as obese patients.

SUV measurements are known to vary depending on the patient’s body weight. As body weight and body mass index (BMI) increase, SUV values in tumor tissues and normal tissues also tend to rise (8), potentially leading to inconsistent results across patients, and predicting therapy response within a patient (9). To overcome these limitations, lean body mass-corrected standardized uptake value (SUL) have emerged as a promising alternative, offering superior reproducibility across individuals with varying body compositions (10). Recent studies have shown that parameters like SUL maximum (SUL_max) provide more stable measurements than SUV, especially in patients with high BMI (11, 12, 13). However, data specifically addressing the clinical utility of lean body mass-adjusted (LBM) PET parameters in overweight or obese patients with endometrial cancer remain limited.

In this study, we aimed to evaluate whether metabolic and volumetric PET/CT parameters corrected for lean body mass are more effective in predicting pathological features in overweight patients diagnosed with endometrial cancer. In addition, we explored the potential prognostic implications of LBM PET parameters by performing progression-free survival (PFS) analyses based on ROC derived cut-off values. Our goal was to determine whether SUL-based imaging metrics can enhance the accuracy of preoperative risk stratification and support more personalized treatment planning.

Materials and Methods

Study Population

This single-center retrospective study included consecutive patients diagnosed with endometrial cancer between January 2021 and December 2024. A total of 412 patients were initially screened. After applying predefined inclusion and exclusion criteria, 73 patients were included in the final analysis. The patient selection process is summarized in Figure 1. Surgical treatment, consisting of total abdominal hysterectomy with bilateral salpingo-oophorectomy and pelvic lymph node dissection, was performed within two weeks following PET/CT imaging in all cases.

Demographic characteristics, clinicopathological variables, and postoperative histopathological findings were retrieved from the institutional electronic medical records system. Histopathological evaluation included tumor grade, depth of myometrial invasion, LVSI, cervical stromal involvement, and lymph node status, when available. The approval has been granted by the Ethics Committee of Ege University Medical Research with (decision no: 25-3.1T/87, date: 20.03.2025).

PET/CT Acquisition and Image Analysis

All PET/CT scans were performed using the same PET/CT scanner (Biograph True Point 16; Siemens Healthcare, Henkester, Germany) in accordance with the Uniform Protocols for Imaging in Clinical Trials (UPICT) guidelines (14). PET/CT imaging was performed from the vertex to the mid-thigh following at least 6 hours of fasting, and blood glucose levels were confirmed to be below 150 mg/dL prior to tracer injection. The PET/CT images were independently analyzed by two experienced nuclear medicine physicians who were blinded to the patients’ clinical information. Any discrepancies were resolved by consensus. PET/CT data were evaluated both visually and semi-quantitatively. Semi-quantitative parameters, including maximum SUV (SUV_max), SUV mean (SUV_mean), and peak (SUV_peak), were calculated for each lesion. Additionally, [maximum SUL (SUL_max), SUL mean (SUL_mean) and peak (SUL_peak] were derived using the James formula, implemented within the PET/CT workstation software. Volumetric parameters such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG) were also obtained, along with TLG-LBM to assess overall metabolic burden.

Follow-up and Outcome Assessment

Patients were followed through review of electronic medical records. Recurrence status was assessed based on radiologic and/or histopathologic confirmation during follow-up. Recurrent disease was defined morphologically as radiologic evidence of locoregional or distant tumor reappearance after primary surgical treatment. PFS was calculated from the date of surgery to the date of documented recurrence or last clinical follow-up. Patients without recurrence were censored at the time of last follow-up.

Statistical Analysis

For statistical analysis, we compared PET/CT-derived metabolic and volumetric parameters—including SUV_max, SUV_mean, SUV_peak, SUL_max, SUL_mean, SUL_peak, MTV, TLG, and SUL-TLG—with key clinicopathological features. These included histological grade, depth of myometrial invasion, LVSI, lymph node involvement, cervical stromal invasion, hormone receptor status (ER/PR), p53 mutation status, age at diagnosis, and tumor size. Continuous variables were assessed for correlations using Spearman’s rank correlation coefficient. Group comparisons were performed using non-parametric tests (Mann-Whitney U or Kruskal-Wallis) as appropriate. ROC curve analyses were conducted to evaluate the discriminative ability of PET/CT parameters for relevant binary outcomes, and optimal cut-off values were determined using the Youden J index. All tests were two-tailed, and a p-value of <0.05 was considered statistically significant. All statistical analyses were conducted using SPSS version 25 (IBM Corp., Armonk, NY, USA).

Results

A total of 73 patients with endometrial cancer who underwent preoperative ¹⁸F-FDG PET/CT followed by primary surgical treatment were included. Baseline clinical characteristics, histopathologic features, and follow-up data are summarized in Table 1. The study population predominantly consisted of patients with early-stage disease, while individual histopathologic risk factors—including deep myometrial invasion, LVSI, cervical stromal involvement, and lymph node metastasis—were variably present.

The distribution of PET-derived metabolic and volumetric parameters, including SUV-based, SUL-based, and volumetric measurements, is presented in Table 2. All PET parameters demonstrated non-normal distributions and were therefore summarized using median and interquartile range values.

Comparisons of PET-derived metabolic parameters according to adverse histopathologic features are detailed in Table 3. Among evaluated parameters, SUL-based measurements demonstrated significant differences across several adverse pathologic characteristics, including LVSI and lymph node involvement. Representative PET/CT images illustrating differences in primary tumor FDG uptake according to LVSI status are shown in Figure 2. Both SUV_mean and SUL_mean differed significantly according to the depth of myometrial invasion.

PET-derived metabolic parameters also differed significantly across histologic tumor grades, as shown in Table 4. Global differences were observed for multiple SUV- and SUL-based parameters, including SUL_mean. Pairwise post-hoc analyses revealed significant differences primarily between low-grade and higher-grade tumors, whereas no significant differences were observed between intermediate- and high-grade tumors.

Spearman correlation analyses demonstrated moderate to strong positive correlations between tumor size and volumetric PET parameters, particularly MTV and TLG (r values ranging from approximately 0.68 to 0.73, p<0.001). SUV_mean and SUL_mean also showed significant positive correlations with tumor size, although with lower correlation coefficients (r=0.38-0.50, p<0.001).

Preoperative cancer antigen 125 levels demonstrated weak to moderate correlations with tumor size and volumetric PET-derived parameters (r=0.30-0.40, p<0.05).

The diagnostic performance of PET-derived parameters for predicting individual histopathologic risk factors was evaluated using receiver operating characteristic analysis, with results summarized in Table 5. SUL_mean demonstrated higher discriminative ability compared with SUV-based parameters for the prediction of LVSI and lymph node involvement.

ROC analysis was performed to evaluate the diagnostic performance of PET-derived parameters for predicting individual histopathologic risk factors (Table 5). Among the evaluated parameters, SUL_mean demonstrated the highest discriminative ability for lymph node involvement. An optimal SUL_mean cut-off value of 6.04 was identified based on the Youden index (Youden J =0.55), yielding a sensitivity of 80%, specificity of 75%, positive predictive value of 76%, and negative predictive value of 78% for the prediction of lymph node involvement.

Because lymph node involvement represents a clinically relevant adverse prognostic factor and SUL_mean demonstrated the strongest diagnostic performance for this endpoint, the ROC-derived cut-off value of 6.04 was selected for subsequent survival analyses. Patients were dichotomized into low (≤6.04) and high (>6.04) SUL_mean groups according to this threshold. Kaplan-Meier analysis was performed to assess PFS. During follow-up, recurrence events occurred in a limited number of patients. Survival distributions did not differ significantly between the two groups (log-rank p=0.46; Figure 3). Median PFS was not reached in either group due to the low event rate. Given the limited number of events, further survival modeling was not conducted.

Discussion

In this retrospective cohort of overweight patients with endometrial cancer, lean body mass-adjusted PET/CT parameters—particularly SUL_mean—demonstrated stronger and more consistent associations with adverse histopathologic features than conventional SUV-based metrics. Specifically, SUL_mean was significantly higher in tumors with deep myometrial invasion, LVSI, lymph node involvement, and higher histologic grade. In ROC analyses, SUL_mean showed moderate discriminative ability for LVSI and the highest performance for lymph node involvement, whereas SUV-based parameters were less consistent across these endpoints.

From a biological and methodological perspective, these findings suggest that lean body mass correction may reduce variability introduced by excess adipose tissue and provide a metabolically more stable estimate of tumor glycolytic activity in overweight individuals. SUV measurements are known to be influenced by total body weight and adiposity, potentially leading to overestimation of uptake in obese patients. In contrast, lean body SUL reduce this dependency and improve inter-patient comparability (12, 13, 15).

While both SUV_mean and SUL_mean were associated with depth of myometrial invasion, SUL_mean demonstrated additional discriminatory value for LVSI and nodal metastasis, two features closely linked to extrauterine spread and the need for comprehensive surgical staging in accordance with current guidelines (6, 16). Notably, SUL_mean values increased stepwise across tumor grades, with the most pronounced differences observed between low-grade and higher-grade tumors, supporting its potential role as a marker of tumor aggressiveness.

Importantly, although a ROC-derived SUL_mean cut-off of 6.04 demonstrated reasonable sensitivity and specificity for lymph node involvement, Kaplan-Meier analysis did not reveal a statistically significant difference in PFS between low and high SUL_mean groups (log-rank p=0.46). This negative survival result must be explicitly acknowledged. Several factors likely contributed to the absence of a significant PFS difference: the predominance of early-stage disease, the very low number of recurrence events (n=5), and a follow-up duration that may be insufficient to capture late recurrences in this generally favorable-risk population (1, 2, 3). In addition, although SUL_peak is emphasized in PERCIST criteria for treatment response assessment (7), the prognostic role of baseline SUL-based parameters in endometrial cancer—particularly in relation to PFS—remains insufficiently studied. Therefore, our findings do not support the use of SUL_mean as an independent survival predictor and should be interpreted cautiously.

The present results are partially consistent with prior studies reporting associations between FDG uptake and adverse pathologic features in endometrial cancer. According to Yao et al. (17), no significant correlation was observed between estrogen receptor (ER) expression and FDG uptake on PET/CT. Takagi et al. (18) demonstrated a significant association between SUV_max and histological grade, while Vural Topuz et al. (19) reported correlations between SUV_max, MTV, TLG, and several adverse pathologic features, including nodal involvement. However, these analyses were based on weight-normalized SUV metrics and did not incorporate lean body mass correction. Studies focusing on SUL in gynecologic malignancies have suggested potential advantages in reflecting tumor differentiation (11), yet direct comparisons between SUV- and SUL-based parameters in overweight endometrial cancer populations remain limited. Sürer Budak et al. (20) reported that SUV_max and apparent diffusion coefficient minimum were independently associated only with deep myometrial invasion, with moderate predictive performance. In contrast, our findings suggest that lean body mass-adjusted PET parameters were associated with a broader spectrum of adverse histopathologic features, including LVSI and nodal involvement. Although magnetic resonance imaging was not evaluated in our study, the incorporation of SUL-based metrics may provide complementary metabolic information beyond myometrial invasion assessment.

The theoretical advantage of SUL is particularly relevant in endometrial cancer, a malignancy strongly associated with obesity (4, 5). Because SUV is normalized to total body weight, excess adipose tissue—which exhibits relatively low FDG uptake—may artificially inflate calculated tumor uptake values (8). Lean body mass correction addresses this limitation and has been shown to reduce variability related to body composition (12, 13, 15). Our findings support this methodological rationale by demonstrating stronger associations between SUL_mean and adverse histopathologic characteristics in an overweight cohort.

Nevertheless, not all evaluated parameters performed equally. Volumetric metrics such as MTV and TLG correlated strongly with tumor size, which is expected given their volumetric nature, but they did not consistently outperform SUL-based indices for specific adverse features. Furthermore, metabolic parameters did not show significant associations with certain molecular markers, including p53 status, despite its recognized biological relevance in endometrial carcinoma (21). This lack of association may reflect limited statistical power, early-stage predominance, or the fact that FDG uptake captures glucose metabolism rather than the full spectrum of molecular alterations.

This study has several strengths. The exclusive inclusion of overweight patients addresses a clinically relevant population in which SUV variability is most problematic. PET/CT imaging was performed within a short preoperative interval using a standardized protocol consistent with UPICT recommendations (14). Image interpretation was conducted by experienced nuclear medicine physicians blinded to clinical data, minimizing observer bias. Comprehensive histopathologic correlation and ROC-based cut-off analyses further enhance the interpretability of the findings, although these thresholds require external validation.

Study Limitations

However, important limitations must be acknowledged. The retrospective, single-center design introduces potential selection bias. The sample size was modest, and the number of lymph nodes–positive cases and recurrences was low, limiting statistical power. Multivariable survival analysis could not be performed due to the small number of PFS events. External validation in an independent cohort was not undertaken. Moreover, the predominance of early-stage disease restricts extrapolation to advanced-stage or high-risk populations.

Conclusion

Lean body mass–adjusted PET/CT parameters, particularly SUL_mean, demonstrated stronger associations with adverse histopathologic features than conventional SUV-based metrics in overweight patients with endometrial cancer. SUL_mean showed moderate discriminatory performance for LVSI and lymph node involvement, suggesting that lean body mass correction may enhance metabolic characterization in populations with elevated BMI. However, SUL_mean did not predict PFS in this predominantly early-stage cohort. These findings indicate that SUL-based parameters may contribute to preoperative risk assessment in overweight patients, but their prognostic significance remains uncertain. Prospective, multicenter studies with larger event numbers are required to validate these observations and to clarify the role of SUL metrics in clinical decision-making.