Accuracy of preoperative staging for a priori resectable esophageal cancer

Deep learning prediction of esophageal squamous cell carcinoma invasion depth from arterial phase enhanced CT images: a binary classification approach

BACKGROUND

Precise prediction of esophageal squamous cell carcinoma (ESCC) invasion depth is crucial not only for optimizing treatment plans but also for reducing the need for invasive procedures, consequently lowering complications and costs. Despite this, current techniques, which can be invasive and costly, struggle with achieving the necessary precision, highlighting a pressing need for more effective, non-invasive alternatives.

METHOD

We developed ResoLSTM-Depth, a deep learning model to distinguish ESCC stages T1-T2 from T3-T4. It integrates ResNet-18 and Long Short-Term Memory (LSTM) networks, leveraging their strengths in spatial and sequential data processing. This method uses arterial phase CT scans from ESCC patients. The dataset was meticulously segmented by an experienced radiologist for effective training and validation.

RESULTS

Upon performing five-fold cross-validation, the ResoLSTM-Depth model exhibited commendable performance with an accuracy of 0.857, an AUC of 0.901, a sensitivity of 0.884, and a specificity of 0.828. These results were superior to the ResNet-18 model alone, where the average accuracy is 0.824 and the AUC is 0.879. Attention maps further highlighted influential features for depth prediction, enhancing model interpretability.

CONCLUSION

ResoLSTM-Depth is a promising tool for ESCC invasion depth prediction. It offers potential for improvement in the staging and therapeutic planning of ESCC.

Computed tomography radiomics identification of T1-2 and T3-4 stages of esophageal squamous cell carcinoma: two-dimensional or three-dimensional?

BACKGROUND

To evaluate two-dimensional (2D) and three-dimensional (3D) computed tomography (CT) radiomics analysis for the T stage of esophageal squamous cell carcinoma (ESCC).

METHODS

398 patients with pathologically confirmed ESCC were divided into training and testing sets. All patients underwent chest CT scans preoperatively. For each tumor, based on CT images, a 2D region of interest (ROI) was outlined on the largest cross-sectional area, and a 3D ROI was outlined layer by layer on each section of the tumor. The radiomics platform was used for feature extraction. For feature selection, stepwise logistic regression was used. The receiver operating characteristic (ROC) curve was used to assess the diagnostic performance of the 2D radiomics model versus the 3D radiomics model. The differences were compared using the DeLong test. The value of the clinical utility of the two radiomics models was evaluated.

RESULTS

1595 radiomics features were extracted. After screening, two radiomics models were constructed. In the training set, the difference between the area under the curve (AUC) of the 2D radiomics model (AUC = 0.831) and the 3D radiomics model (AUC = 0.830) was not statistically significant (p = 0.973). In the testing set, the difference between the AUC of the 2D radiomics model (AUC = 0.807) and the 3D radiomics model (AUC = 0.797) was also not statistically significant (p = 0.748). A 2D model was equally useful as a 3D model in clinical situations.

CONCLUSION

The performance of 2D radiomics model is comparable to that of 3D radiomics model in distinguishing between the T1-2 and T3-4 stages of ESCC. In addition, 2D radiomics model may be a more feasible option due to the shorter time required for segmenting the ROI.

Non-curative resection for surgical T4b esophageal cancer: esophagectomy or non-esophagectomy?

BACKGROUND

Recently, with the development of multidisciplinary treatment, the treatment outcomes of esophageal cancer (EC) have improved. However, despite advances in diagnostic imaging modalities, preoperative diagnosis of T4 EC is still difficult, and the prognosis of T4 EC remains very poor. In addition, the prognosis of surgical T4b EC (sT4b EC) after surgery remains unclear. In this study, we retrospectively reviewed sT4b EC.

METHODS

We evaluated the clinical course of sT4b EC and compared palliative esophagectomy with R2 resection (PE group) with other procedures without esophagectomy (NE group) (e.g., only esophagostomy) for sT4b EC.

RESULTS

Forty-seven patients with thoracic EC underwent R2 resection at our institution between January 2009 and December 2020. Thirty-four patients were in the PE group, and 13 patients were in the NE group. The 2-year overall survival rate was 0% in the PE group and 20.2% in the NE group (p = 0.882). There was one case of long-term survival in the NE group that underwent surgery followed by definitive chemoradiation. Postoperative complications (Clavien-Dindo grade ≥ 3) were observed in 25 patients (73.5%) in the PE group and in three patients (23.1%) in the NE group (p = 0.031). The median time to the initiation of postoperative treatment was 68.1 days in the PE group and 18.6 days in the NE group (p = 0191).

CONCLUSIONS

If EC is diagnosed as sT4b, palliative esophagectomy should be avoided because of the high complication rate and the lack of long-term survival.

Histological prognosticators in neoadjuvant naive oesophageal cancer patients

PURPOSE

Prognosis of oesophageal cancer is primarily based upon the TNM stage of the disease. However, even in those with similar TNM staging, survival can be varied. Additional histopathological factors including venous invasion (VI), lymphatic invasion (LI) and perineural invasion (PNI) have been identified as prognostic markers yet are not part of TNM classification. The aim of this study is to determine the prognostic importance of these factors and overall survival in patients with oesophageal or junctional cancer who underwent transthoracic oesophagectomy as the unimodality treatment.

METHODS

Data from patients who underwent transthoracic oesophagectomy for adenocarcinoma without neoadjuvant treatment were reviewed. Patients were treated with radical resection, with a curative intent using a transthoracic Ivor Lewis or three staged McKeown approach.

RESULTS

A total of 172 patients were included. Survival was poorer when VI, LI and PNI were present (p<0.001), with the estimated survival being significantly worse (p<0.001) when patients were stratified according to the number of factors present. Univariable analysis of factors revealed VI, LI and PNI were all associated with survival. Presence of LI was independently predictive of incorrect staging/upstaging in multivariable logistic regression analysis (OR 12.9 95% CI 3.6-46.6, p<0.001).

CONCLUSION

Histological factors of VI, LI and PNI are markers of aggressive disease and may have a role in prognostication and decision-making prior to treatment. The presence of LI as an independent marker of upstaging could be a potential indication for the use of neoadjuvant treatment in patients with early clinical disease.

Feasibility study of adjuvant chemotherapy with S-1 after curative esophagectomy following neoadjuvant chemotherapy for esophageal cancer

BACKGROUND

Despite advances in surgical techniques, long-term survival after esophagectomy for esophageal cancer remains unacceptably low, and more effective perioperative chemotherapy is expected. However, an important concern regarding the application of postoperative adjuvant chemotherapy is treatment toxicity. We aimed to evaluate the feasibility of adjuvant chemotherapy with S-1 in patients after esophagectomy.

METHODS

We investigated the tolerability of a 2-week administration followed by 1-week rest regimen of S1 as postoperative adjuvant therapy in 20 patients with esophageal squamous cell carcinoma who received neoadjuvant chemotherapy (NAC) and 22 patients who did not receive NAC during 2011-2020.

RESULTS

In the non-NAC group, the mean and median relative dose intensity (RDI) were 78.7% and 99.4%, respectively, and 11 patients (50%) had altered treatment schedules. The corresponding rates in the NAC group were 77.9% and 100%, respectively, and nine patients (45%) had altered treatment schedules, with no significant difference among the groups. Moreover, 17 patients (77.2%) in the non-NAC group and 16 patients (80.0%) in the NAC group continued S-1 treatment as planned for one year postoperatively, with no significant difference in the S-1 continuation rate (p = 0.500). Seventeen of 22 patients (77.3%) and 15 of 20 patients (75.0%) experienced several adverse events in the non-NAC and NAC groups, respectively. The frequency, severity, and type of adverse events were consistent among patients with and without NAC.

CONCLUSIONS

S-1 could be safely and continuously administered as adjuvant chemotherapy for patients with esophageal cancer regardless of NAC. Long-term prognosis should be evaluated for S-1 to become the standard treatment after esophagectomy.

Impact of endoscopic ultrasonography on the accuracy of T staging in esophageal cancer and factors associated with its accuracy: A retrospective study

The sensitivity and specificity of endoscopic ultrasound (EUS) for esophageal cancer are variable. The aim of the present study was to determine the accuracy of EUS for the T staging of esophageal cancer and to explore the factors that affect the accuracy.This was a retrospective study of patients with esophageal cancer who underwent EUS between January 2018 and September 2019 at the author's hospital. All patients underwent EUS, surgery, and pathological examination. The diagnostic value of ultrasound-based T (uT) staging was evaluated using the pathological T (pT) staging as the gold standard.Finally, 169 patients were included. Among the 169 patients, 37 were overstaged by EUS, 33 were understaged, and 99 were correctly staged. The overall accuracy of EUS was 58.6%. Sensitivity was low, at 0% to 70.8% depending upon the pT stage, but specificity was higher, at 71.0% to 100.0%, also depending upon the pT stage. The multivariable analysis revealed that highly differentiated tumors (odds ratio = 9.167, P = .041) and pT stage ≥T2 (odds ratio = 2.932, P = .004) were independent factors of accurate uT stage.The staging of esophageal cancer using EUS has low sensitivity but high specificity. Highly differentiated tumors and pT stage ≥2 tumors were associated with the accuracy of uT staging.

Clinical Predictors of Locally Advanced Pathology in Esophageal Adenocarcinoma

Background

In patients with resectable esophageal adenocarcinoma (EAC), the decision for neoadjuvant treatment depends on clinical staging with endoscopic ultrasound (EUS) and positron-emission tomography (PET) scan. Patients with locally advanced EAC pathology misclassified as early EAC by clinical staging are missing the opportunity to receive neoadjuvant therapy. We aim to identify predictors of locally advanced pathology in EAC to determine more accurately those who benefit from neoadjuvant therapy. 

Methods

Retrospective study of patients who underwent upfront endoscopic or surgical resection for EAC without neoadjuvant therapy from January 2011 to December 2017 was performed. Clinical characteristics, EUS, PET scan and histologic findings were analyzed. Multivariable analysis of predictors of locally advanced stage was performed and a risk prediction score was developed.

Results

A total of 97 patients were included; 68 patients were staged as early EAC (pT1 or pT2 and pN0) and 29 patients were staged as locally advanced EAC (pT1 or pT2 with pN1 and pT3 or pT4 irrespective of N status). In a predictive model of EAC, patients presenting with dysphagia, tumor size >2 cm, exophytic mass appearance on endoscopy and absence of hiatal hernia were more likely to be have locally advanced pathology with a probability of 70% (C-statistic 0.766).

Conclusions

A risk prediction model based on the presence of dysphagia, tumor size >2 cm, exophytic mass appearance and absence of hiatal hernia can be used to identify locally advanced pathology in EAC patients.

Computed Tomography-Based Radiomics in Predicting T Stage and Length of Esophageal Squamous Cell Carcinoma

Background

Because of the superficial and infiltrative spreading patterns of esophageal squamous cell carcinoma (ESCC), an accurate assessment of tumor extent is challenging using imaging-based clinical staging. Radiomics features extracted from pretreatment computed tomography (CT) or magnetic resonance imaging have shown promise in identifying tumor characteristics. Accurate staging is essential for planning cancer treatment, especially for deciding whether to offer surgery or radiotherapy (chemotherapy) in patients with locally advanced ESCC. Thus, this study aimed to evaluate the predictive potential of contrast-enhanced CT-based radiomics as a non-invasive approach for estimating pathological tumor extent in ESCC patients.

Methods

Patients who underwent esophagectomy between October 2011 and September 2017 were retrospectively studied and included 116 patients with pathologically confirmed ESCC. Contrast-enhanced CT from the neck to the abdomen was performed in all patients during the 2 weeks before the operation. Radiomics features were extracted from segmentations, which were contoured by radiologists. Cluster analysis was performed to obtain clusters with similar radiomics characteristics, and chi-squared tests were used to assess differences in clinicopathological features and survival among clusters. Furthermore, a least absolute shrinkage and selection operator was performed to select radiomics features and construct a radiomics model. Receiver operating characteristic analysis was used to evaluate the predictive ability of the radiomics signatures.

Results

All 116 ESCC patients were divided into two groups according to the cluster analysis. The chi-squared test showed that cluster-based radiomics features were significantly correlated with T stage (p = 0.0254) and tumor length (p = 0.0002). Furthermore, CT radiomics signatures exhibited favorable predictive performance for T stage (area under the curve [AUC] = 0.86, sensitivity = 0.77, and specificity = 0.87) and tumor length (AUC = 0.95, sensitivity = 0.92, and specificity = 0.91).

Conclusions

CT contrast radiomics is a simple and non-invasive method that shows promise for predicting pathological T stage and tumor length preoperatively in ESCC patients and may aid in the accurate assessments of patients in combination with the existing examinations.

Accuracy of clinical staging for T2N0 oesophageal cancer: systematic review and meta-analysis

Oesophageal cancer is the sixth commonest cause of overall cancer mortality. Clinical staging utilizes multiple imaging modalities to guide treatment and prognostication. T2N0 oesophageal cancer is a treatment threshold for neoadjuvant therapy. Data on accuracy of current clinical staging tests for this disease subgroup are conflicting. We performed a meta-analysis of all primary studies comparing clinical staging accuracy using multiple imaging modalities (index test) to histopathological staging following oesophagectomy (reference standard) in T2N0 oesophageal cancer. Patients that underwent neoadjuvant therapy were excluded. Electronic databases (MEDLINE, Embase, Cochrane Library) were searched up to September 2019. The primary outcome was diagnostic accuracy of combined T&N clinical staging. Publication date, first recruitment date, number of centers, sample size and geographical location main histological subtype were evaluated as potential sources of heterogeneity. The search strategy identified 1,199 studies. Twenty studies containing 5,213 patients met the inclusion criteria. Combined T&N staging accuracy was 19% (95% CI, 15-24); T staging accuracy was 29% (95% CI, 24-35); percentage of patients with T downstaging was 41% (95% CI, 33-50); percentage of patients with T upstaging was 28% (95% CI, 24-32) and percentage of patients with N upstaging was 34% (95% CI, 30-39). Significant sources of heterogeneity included the number of centers, sample size and study region. T2N0 oesophageal cancer staging remains inaccurate. A significant proportion of patients were downstaged (could have received endotherapy) or upstaged (should have received neoadjuvant chemotherapy). These findings were largely unchanged over the past two decades highlighting an urgent need for more accurate staging tests for this subgroup of patients.

Value of preoperative computed tomography in predicting the local staging of cancer of the tongue at primary surgery

Preoperative staging is essential for the planning of treatment of cancer. This study was designed to evaluate the accuracy of computed tomography (CT) in predicting the local stage of tongue cancer by comparing it with the gold standard of histopathology. A total of 233 patients with newly-diagnosed tongue cancer was retrospectively reviewed, and the size of the tumour and the status of the cervical lymph node were compared between CT images and histopathological results. Patients with stage II cancer were followed up to assess the influence of inaccurate preoperative staging on prognosis. The accuracy of local staging by CT was 47.6% (111/233), with 59.7% (139/233) for tumour stage, and 70.4% (164/233) for nodal stage. The greatest dimension of the tumour on the CT image was about 2mm less than that measured by histopathology. The estimated volume of tumour was a quarter smaller. The accuracy of predicting malignant lymph nodes by CT was 68.9% (n=161). Among patients with stage II disease, simultaneous neck dissection was less likely in the understaged group than in the accurately staged one. The reoperation rate was a little higher but not significantly so. We conclude that the accuracy of CT in predicting local staging for tongue cancer was only moderate, because it underestimated the size of the tumour and needed to improve the criteria for detecting malignant lymph nodes. Understaging on CT images may influence the prognosis of patients with early stage tongue cancer.

Combined CT texture analysis and nodal axial ratio for detection of nodal metastasis in esophageal cancer

OBJECTIVE

To assess the accuracy of a combination of CT texture analysis (CTTA) and nodal axial ratio to detect metastatic lymph nodes (LNs) in esophageal squamous cell carcinoma (ESCC).

METHODS

The contrast-enhanced chest CT images of 78 LNs (40 metastasis, 38 benign) from 38 patients with ESCC were retrospectively analyzed. Nodal axial ratios (short-axis/long-axis diameter) were calculated. CCTA parameters (kurtosis, entropy, skewness) were extracted using commercial software (TexRAD) with fine, medium, and coarse spatial filters. Combinations of significant texture features and nodal axial ratios were entered as predictors in logistic regression models to differentiate metastatic from benign LNs, and the performance of the logistic regression models was analyzed using the area under the receiver operating characteristic curve (AUROC).

RESULTS

The mean axial ratio of metastatic LNs was significantly higher than that of benign LNs (0.81 ± 0.2 vs 0.71 ± 0.1, p = 0.005; sensitivity 82.5%, specificity 47.4%); namely, significantly more round than benign. The mean values of the entropy (all filters) and kurtosis (fine and medium) of metastatic LNs were significantly higher than those of benign LNs (all, p < 0.05). Medium entropy showed the best performance in the AUROC analysis with 0.802 (p < 0.001; sensitivity 85.0%, specificity 63.2%). A binary logistic regression analysis combining the nodal axial ratio, fine entropy, and fine kurtosis identified metastatic LNs with 87.5% sensitivity and 65.8% specificity (AUROC = 0.855, p < 0.001).

CONCLUSION

The combination of CTTA features and the axial ratio of LNs has the potential to differentiate metastatic from benign LNs and improves the sensitivity for detection of LN metastases in ESCC.

ADVANCES IN KNOWLEDGE

The combination of CTTA and nodal axial ratio has improved CT sensitivity (up to 87.5%) for the diagnosis of metastatic LNs in esophageal cancer.

Clinical Relevance of Circulating Tumor Cells in Esophageal Cancer Detected by a Combined MACS Enrichment Method

INTRODUCTION

Current modalities to predict tumor recurrence and survival in esophageal cancer are insufficient. Even in lymph node-negative patients, a locoregional and distant relapse is common. Hence, more precise staging methods are needed. So far, only the CellSearch system was used to detect circulating tumor cells (CTC) with clinical relevance in esophageal cancer patients. Studies analyzing different CTC detection assays using advanced enrichment techniques to potentially increase the sensitivity are missing.

METHODS

In this single-center, prospective study, peripheral blood samples from 90 esophageal cancer patients were obtained preoperatively and analyzed for the presence of CTCs by Magnetic Cell Separation (MACS) enrichment (combined anti-cytokeratin and anti-epithelial cell adhesion molecules (EpCAM)), with subsequent immunocytochemical staining. Data were correlated with clinicopathological parameters and patient outcomes.

RESULTS

CTCs were detected in 25.6% (23/90) of the patients by combined cytokeratin/EpCAM enrichment (0-150 CTCs/7.5 mL). No significant correlation between histopathological parameters and CTC detection was found. Survival analysis revealed that the presence of more than two CTCs correlated with significantly shorter overall survival (OS) and progression-free survival (PFS).

CONCLUSION

With the use of cytokeratin as an additional enrichment target, the CTC detection rate in esophageal cancer patients can be elevated and displays the heterogeneity of cytokeratin (CK) and EpCAM expression. The presence of >2CTCs correlated with a shorter relapse-free and overall survival in a univariate analysis, but not in a multivariate setting. Moreover, our results suggest that the CK7/8+/EpCAM+ or CK7/8+/EpCAM- CTC subtype does not lead to an advanced tumor staging tool in non-metastatic esophageal cancer (EC) patients.

18F- FDG PET/CT-derived parameters predict clinical stage and prognosis of esophageal cancer

BACKGROUND

Although ¹⁸F- FDG PET/CT is validated in baseline workup of esophageal cancer to detect distant metastases, it remains underused in assessing local staging and biology of the primary tumor. This study aimed to evaluate the association between ¹⁸F- FDG PET/CT-derived parameters of esophageal cancer, and its clinico-pathological features and prognosis.

METHODS

All patients (n = 86) with esophageal adenocarcinoma or squamous cell cancer operated between 2005 and 2014 were analyzed. Linear regression was used to identify clinico-pathologic features of esophageal cancer associated with the tumor's maximal Standardized Uptake Value (SUV_max), Total Lesion Glycolysis (TLG) and Metabolic Tumor Volume (MTV). ROC curve analysis was performed to precise the optimal cutoff of each variable associated with a locally advanced (cT3/4) status, long-term survival and recurrence. Kaplan Meier curves and Cox regression were used for survival analyses.

RESULTS

High baseline SUV_max was associated with cT3/4 status and middle-third tumor location, TLG with a cT3/4 and cN+ status, whereas MTV only with active smoking. A cT3/4 status was significantly predicted by a SUV_max > 8.25 g/mL (p < 0.001), TLG > 41.7 (p < 0.001) and MTV > 10.70 cm³ (p < 0.01) whereas a SUV_max > 12.7 g/mL was associated with an early tumor recurrence and a poor disease-free survival (median 13 versus 56 months, p = 0.030), particularly in squamous cell cancer.

CONCLUSIONS

Baseline ¹⁸F- FDG PET/CT has a high predictive value of preoperative cT stage, as its parameters SUV_max, TLG and MTV can predict a locally advanced tumor with high accuracy. A SUV_max > 12.7 g/mL may herald early tumor recurrence and poor disease-free survival.

Olfactomedin 4 (OLFM4) expression is associated with nodal metastases in esophageal adenocarcinoma

To date no informative biomarkers exist to accurately predict presence of lymph node metastases (LNM) in esophageal adenocarcinoma (EAC). We studied the discriminative value of Olfactomedin 4 (OLFM4), an intestinal stem cell marker, in EAC. Patients who had undergone esophagectomy as single treatment modality for both advanced (pT2-4) and early (pT1b) adenocarcinoma of the esophagus or gastro-esophageal junction were selected for this study from an institutional database (Erasmus MC University Medical Center, Rotterdam, The Netherlands). Surgical resection specimens of 196 advanced and 44 early EAC were examined. OLFM4 expression was studied by immunohistochemistry and categorized as low (<30%) or high (> = 30%) expression. Low OLFM4 was associated with poor differentiation grade in both advanced (60% vs. 34.8%, p = 0.001) and early EAC (39.1% vs. 9.5%, p = 0.023). LNM were present in 161 (82.1%) of advanced and 9 (20.5%) of early EAC respectively. Low OLFM4 was independently associated with the presence of LNM in advanced EAC in multivariable analysis (OR 2.7; 95% CI, 1.16-6.41; p = 0.022), but not in early EAC (OR 2.1; 95% CI, 0.46-9.84; p = 0.338). However, the difference in association with LNM between advanced (OR 2.7; 95% CI, 1.18-6.34; p = 0.019) and early (OR 2.3; 95% CI, 0.47-11.13; p = 0.302) EAC was non-significant (p = 0.844), suggesting that the lack of significance in early EAC is due to the small number of patients in this group. OLFM4 was not of significance for the disease free and overall survival. Overall, low expression of intestinal stem cell marker OLFM4 was associated with the presence of LNM. Our study suggests that OLFM4 could be an informative marker with the potential to improve preoperative assessment in patients with EAC. Further studies are needed to confirm the value of OLFM4 as a biomarker for LNM.

Prognostic significance of preoperative lymph node assessment for patients with stage pN0 esophageal squamous cell carcinoma after esophagectomy

Background

Accurate preoperative lymph node (LN) staging is important for surgical treatments of esophageal squamous cell carcinoma (ESCC). The aim of the study was to investigate the role of preoperative lymph nodes assessment by computed tomography (CT) scans in prognostic estimates and lymph nodes dissection strategy.

Methods

A total of 233 stage pN0 ESCC patients who underwent radical esophagectomy from 2009 to 2016 were included, with the last follow-up time in 2018. Survival analysis was conducted to assess the relationship between preoperative clinical LN metastasis and the prognosis of patients with pN0 ESCC.

Results

Ninety-nine patients were classified as clinical positive LN metastasis by CT scans, but were confirmed as stage N0 by postoperative pathological examination, and survival analysis suggested that these patients had relatively poorer prognosis (P=0.027). Cox regression analysis indicated that the clinical LN metastasis on CT scans was an independent negative prognostic factor for patients with pN0 ESCC (P=0.031). The number of LNs dissected affected the prognosis of pN0 patients. Patients with positive LN metastasis on CT would have better prognosis when the number of dissected LNs was equal to or more than 15 LNs (P=0.036). Especially for patients with higher T stage, they were would obtain prognostic benefit with at least 17 LNs dissected (P=0.037). On the other hand, for those with negative LN metastasis on CT, at least 12 LNs dissected indicated better prognosis (P=0.019).

Conclusions

Preoperative LN assessment for ESCC patients is critically important, the optimal LN dissection number should refer to the preoperative CT performance.