Textural features based enhanced contrast CT images predicts prognosis to concurrent chemoradiotherapy in stage III esophageal squamous cell cancer

Abstract

INTRODUCTION:

To study the relationship between the tumor heterogeneity based on CT and overall survival (OS) in oesophageal squamous cell carcinoma treated with chemotherapy and radiation therapy (CRT).

METHODS:

Fifty-seventh clinical patients who underwent definitive CRT were analyzed. The results were analyzed in terms of whole-tumor texture, with quantification of entropy, mean gray-level intensity for fine to coarse textures (filters 1.0–2.5, respectively). The association between texture parameters and survival time was assessed by Kaplan-Meier analysis and a Cox proportional hazards model.

RESULTS:

The median, 1 and 3 years OS, were 20.2 months, 75.4%, and 32.1%. In the univariate analysis performed using the log-rank test found global entropies ( $P=$ 0.0119), global mean ( $P=$ 0.088), global Std ( $P=$ 0.0209), and global uniformity ( $P=$ 0.0284) were found to be significant OS prognostic factors for filter value 1.0. Cox proportional hazards models that included a combination of pretreatment GlobalStd and post-treatment volume yields the best performance in predicting OS. Kaplan-Meier curves show that the patients in the high-risk group have significantly worse OS (log-rank test, $P=$ 0.0009) and progression-free survival (PFS) (log-rank test, $P=$ 0.0019) than those in the low-risk group.

CONCLUSION:

Pretreatment texture parameters are associated with survival time, and the combination of post volume performed better in survival models.

1. Introduction

In 2015, an estimated 477.9/100,000 cases of oesophageal cancer (EC) were diagnosed in China, and approximately 375/100,000 people died from this disease [1, 2]. Most patients at diagnosis already have locally advanced oesophageal cancer. Curative chemo-therapy and radiation therapy (CCRT) or preoperative CRT were standard treatment for locally advanced squamous cell carcinoma [3]. The prediction of overall survival before treatment initiation may be of important to the optimization of disease management. Currently, the routine clinical use of imaging in oesophageal cancer risk stratification has been limited to the establishment of the clinical tumor stage. Clinical stage cannot correctly predict patients’ progress with combined modality treatment [3, 4, 5]. Therefore, additional methods of risk-stratification may be useful to stratify clinical management further and enable a more personalized approach to therapy.

Tumor heterogeneity have the relationship with adverse prognosis in oesophageal cancer. To analyze textural qualities in images of tumors is a noninvasive approach to assessing tumor heterogeneity, which would provide some information of tumor microenvironment. Computed tomography (CT) have been used to derive information on tumor texture, which is then analyzed by using structural, model-based, or statistical methods [8, 9, 10]. The textures based on CT have been used to predict the OS in non-small-cell lung cancer, liver cancer, and colorectal cancer [8, 9, 10]. However, a few studies have investigated the importance of texture analysis of CT to predict outcomes in oesophageal squamous cell carcinoma (ESCC) patients with definitive CT. The goal of our study was to assess whether tumor texture in pre- or posttreatment primary tumor contrast CT images can be used as a prognostic marker for overall survival OS and progression free survival PFS in patients with stage III ESCC treated with definitive CRT.

2. Methods

A waiver of institutional review board approval was obtained for this retrospective analysis. From January 2008 to December 2014, 95 patients underwent concurrent radiotherapy and chemotherapy. Twenty-five patients (organ metastasis) and 13 patients (without pre- and post-treatment simulator-enhanced CT scan) were excluded. At last, this datasets included 57 consecutive patients (mean age, 62 years; range, 35–79 years), with pathologically confirmed ESCC having a measurable primary tumor of 3 cm or larger at Sichuan Cancer Hospital and Institution. For inclusion in the study, patients had to have had pre-therapy and post-therapy (RT 40 Gy) CT scans with consistent imaging parameters (see CT Image Acquisition and Texture Analysis below); beam hardening and metal artifacts were not exclusion parameters. Disease stage was determined according to the 7th (2010) edition of the American Joint Committee on Cancer (AJCC) staging manual [11].

2.1 Treatment

Chemotherapy consisting of a Taxol (135 mg/m ${}^{2}$ , D1) plus cisplatin (25 mg/m ${}^{2}$ , D1–3) and Docetaxel (60 mg/m ${}^{2}$ , D1) plus cisplatin (25 mg/m ${}^{2}$ , D1–3) was given concurrently with radiation therapy, for which the median dose was 64 Gy (range: 60–66 Gy), delivered in daily 1.8–2.0 Gy fractions on Monday through Friday. Radiation was delivered by three-dimensional conformal radiation therapy, intensity-modulated radiation therapy. A total of 17 (29.8%) patients also received adjuvant chemotherapy, which included a Taxol plus cisplatin (TP) and Docetaxel plus cisplatin (DP).

2.2 CT image acquisition

All patients underwent standard radiotherapy simulator contrast, material-enhanced CT scans of the thorax and abdomen before treatment. All simulator CT studies in our center were performed with a 16–detector row CT scanner (PHILIPS Healthcare) and the same clinical protocol: 120 kV, 180–280 mA with tube current modulation, a matrix of 512, a field of view of 500 mm, and 3 mm reconstructed section thicknesses after intravenous injection of 60–120 ml of 370 mg/ml (1.5 ml/Kg) iodinated contrast agent at 1.8–2 mL/sec. The thoracic images (from the sternal notch to below the hemidiaphragm) were obtained in the arterial phase 0–5 seconds after contrast agent administration. Abdominal images (from above the hemidiaphragm to the iliac crest) were obtained in the portal venous phase 60 seconds after contrast administration. Contrast CT image sets obtained from all 57 patients before treatment and RT 40 Gy were available in institutional archives for analysis.

2.3 Texture analysis

Whole-tumor texture was analyzed using IBEX (Imaging Biomarker Explorer) software version 1.0 $\beta$ [12]. Two radiation oncologists, Li and Wang with more than 10 years of CT experience and 1 year of texture analysis experience, delineated the whole-tumor volume by using region of interest (ROI) analysis on contiguous 3-mm CT images. Esophageal stents, nasogastric tubes, intraluminal air, and oral contrast material were excluded from the ROIs (Fig. 1). This process was refined by an additional thresholding procedure that excluded any pixels with attenuation of less than $-$ 50 HU.

Figure 1.

Conventional enhanced-contrast CT image (before RT and RT 40 Gy) of an upper-site oesophageal cancer with metastasis node and corresponding CT images selectively displaying fine (1B, 1G), medium 1 (1C, 1H), medium 2 (1D, 1M), and coarse (1E, 1N) texture obtained with filter values of 1.0 (width, 4 pixels), 1.5 (width, 6 pixels), 2.0 (width, 8 pixels), and 2.5 (width, 12 pixels).

Texture analysis was performed in a two-stage process that involved image filtration (image transformation) followed by statistical quantification of texture (Fig. 1). Image filtration involved using a Laplacian of Gaussian spatial band-pass filter to produce a series of derived images highlighting features at different anatomic spatial scales ranging from fine to coarse texture for the ROI. The different filter values used in this study were 1.0 (highlighting fine textures, which may enhance tissue parenchymal features), 1.5–2.0 (highlighting medium textures), and 2.5 (highlighting coarse textures, which may enhance vascular features). The filter widths represented by fine, medium, and coarse textures were 4, 6–8, and 10 pixels, respectively, with four categories of quantitative image features (Table 2): intensity direct and histogram (global entropy, global maximum, global mean gray-level intensity, global standard deviation of the histogram [Global Std], global uniformity, kurtosis, skewness).

2.4 Statistical analysis

The Kaplan-Meier method was used to construct OS and PFS curves, and the log-rank test was used to determine the statistical significance of differences. PFS was computed from the time of treatment to the time of tumor progression, or death without evidence of disease progression, at which point the data were censored. OS was calculated from the time of treatment to the time of death from any cause, or to the time of last follow-up, at which point the data were censored. The proportional changes in each texture parameter after CRT was calculated as the differences between the texture parameter pre-treatment and that post-treatment (radiation dose 40 Gy) with regard to the baseline texture value. Changes in textures ( $C_{t}$ ) after CRT were defined as the following equation:

$\displaystyle C_{t}=\frac{X_{1}-X_{2}}{X_{1}}\times 100\%$

where $X_{1}$ is the pre-texture parameters, $X_{2}$ is the post-texture parameters. Texture parameters at each filter level and their proportional changes after CRT were correlated with survival time. Univariate analyses using the Cox proportional hazards model were performed on each texture parameter as a continuous variable for all four filter values. Median value for each texture parameter was used as threshold levels to divide patients for Kaplan-Meier survival analysis, with comparisons between groups by the log-rank test.

Table 1
Patient, tumor and patient characteristics and postoperative result dataset

Factor	N%	MST (m)	3-y OS(%)	$P$ value
Age				0.812
$\leqslant$ 61	23 (40.4)	21.8	26.1
$ï ¼ ž$ 61	34 (59.6)	21.1	40.0
Gender				0.366
Male	40 (70.2)	24.0	39.6
Female	17 (29.8)	19.0	18.2
ECOG score				0.290
0	12 (21.1)	37.9	58.3
1–2	45 (78.9)	20.1	25.5
Alcohol use				0.800
No	22 (38.6)	21.1	34.0
Yes	35 (61.4)	21.8	33.3
Location				0.877
Upper	24 (42.1)	25.1	32.4
Middle	24 (42.1)	17.0	31.7
Lower	9 (15.8)	24.0	41.7
Squamous cell carcinoma			0.324
No	30 (52.6)	18.2	44.6
Well-moderately	11 (19.3)	20.1	10.6
Poor	16 (28.1)	25.0	25.7
Tumor size (cm)				0.573
$\leqslant$ 5 cm	22 (38.6)	43.4	21.8
$ï ¼ ž$ 5 cm	35 (61.4)	28.4	19.1
Clinical T stage 7th				0.586
T2	12 (21.1)	19.1	31.3
T3	34 (59.6)	25.1	37.7
T4	11 (19.3)	15.7	24.2
Clinical N stage 7th				0.277
N1	27 (47.4)	19.1	21.2
N2	25 (43.9)	27.2	50.0
N3	5 (8.8)	14.0	20.0
Clinical TNM Stage 7th				0.928
Stage IIIa	26 (45.6)	21.8	26.5
Stage IIIb	31 (54.4)	18.2	40.8
Chemo regime				0.020
TP	21 (36.8)	19.1	7.6
DP	36 (63.2)	32.0	48.9
Chemo cycles				0.159
1–2	40 (70.8)	27.0	36.7
3–4	17 (29.8)	19.0	10.0
Pre-tre volume (cc)				0.036
$\leqslant$ 7	28 (49.1)	28.4	43.8
$ï ¼ ž$ 17	29 (50.9)	15.7	25.1
Post volume				0.026
$\leqslant$ 9.5	28 (49.1)	28.4	43.8
$ï ¼ ž$ 9.5	29 (50.9)	15.0	25.1

PDG: Pathological Differentiation GradeOSoverall survivalMST: median Survival time or mOS.

The texture parameters with $p<$ 0.05 in both Cox model and Kaplan-Meier analysis were characterized as significant univariate prognostic factors. The significant texture parameters and morphologic variables in univariate analyses were used in multivariate Cox proportional hazards model. The optimal Cox model (best combination of variables) was selected by a step-wise procedure with exclusion and inclusion of candidate variables, based on the Akaike information criterion (AIC). A risk score of prognosis was predicted by the optimal Cox model. Time-independent receiver operating characteristic (ROC) curve analysis was performed to compare the accuracy of the optimal Cox model to those of the texture parameters alone, or the morphologic assessment alone, in assessing the chance of survival. Areas under the ROC curve (AUCs) of each group were calculated and compared by using the DeLong test. For Kaplan-Meier curves, patients were divided into two subgroups by the median value of risk score or the texture parameters. All the statistical methods were performed by two sides, with $p<$ 0.05 as significance cut-off. These statistical analyses were performed by the R software.

Table 2

Texture of primary tumor and node at univariate analysis performed by using the Cox proportional hazards model and log-rank test

Parameters	$P$ value	HR	Lower.95	Upper.95	$P$ value*
Volume_T_post	0.0141	1.0262	1.0052	1.0477	0.0121
GlobalEntropy_T_fine	0.0036	0.0413	0.0048	0.3526	0.0118
GlobalMean_T_fine	0.0061	0.4828	0.2871	0.8120	0.0880
GlobalStd_T_fine	0.0024	0.6020	0.4336	0.8356	0.0209
GlobalMean_T_medium	0.0266	0.2294	0.0624	0.8430	0.0600
GlobalStd_T_medium	0.0038	0.3111	0.1408	0.6874	0.0036
GlobalStd_T_corse	0.0106	0.1852	0.0508	0.6750	0.0635
Skewness_T_medium_Perc	0.0289	0.8011	0.6566	0.9774	0.2039
Kurtosis_T_Corese_Perc	0.0037	0.8890	0.8211	0.9624	0.5842
GlobalEntropy_N_fine	0.0054	0.1861	0.0570	0.6080	0.2641
GlobalMean_N_fine	0.0068	0.7522	0.6120	0.9245	0.3736
GlobalStd_N_fine	0.0091	0.7903	0.6621	0.9432	0.1357
Skewness_N_fine	0.0197	1.6099	1.0789	2.4021	0.3112
GlobalEntropy_post_N_fine	0.0473	0.3033	0.0933	0.9857	0.0685
GlobalMean_post_N_fine	0.0480	0.8225	0.6777	0.9983	0.0085
GlobalEntropy_Perc_N_fine	0.0010	0.8146	0.7210	0.9204	0.1075

T: primary tumor, N: metastasis lymph node *Log-rank test.

3. Results

3.1 Patients

All patients had stage III ESCC disease. Most patients (48/57, 84.2%) had primary tumors in the upper and middle thirds of the esophagus (Table 1). Most patients (36/57, 63.2%) received concurrent DP chemotherapy at weeks 1 and 5. Other patients received concurrent TP chemotherapy. The median times from the pre-treatment CT study to the start of radiation therapy and from the completion of 40 Gy radiation therapy to the post-treatment CT study were 6 days (range, 3–10 days) and 30 days (range, 28–35 days), respectively. Median follow-up for the entire cohort was 32.0 months (range, 7–60 months). The median 1 and 3 years overall survival (OS) was 20.2 months, 75.4% and 32.1%. For clinical factors, in the univariate analysis performed using the log-rank test (Table 1), chemotherapy regimen ( $P=$ 0.024 HR $=$ 0.463, 95% confidence interval [CI] 0.237–0.905), pre-volume ( $P=$ 0.039 HR $=$ 2.033, 95% CI 1.037–3.988), and post-volume ( $P=$ 0.029 HR $=$ 2.116, 95% CI 1.079–4.152) were found to be significant prognostic factors.

3.2 Primary tumor CT features for survival prediction

For primary tumor, in the univariate analysis performed using the Cox proportional hazards model (Table 2), pre-treatment fine, medium, and coarse texture global entropies, global mean, global Std, and global uniformity as continuous variables were found to be significant prognostic factors. For filter value 1.0, the global Std $P=$ 0.0024, the hazard ratio [HR] was 0.602 (95% [CI]: 0.434–0.836); value 1.5, the HR was 0.311 (95% [CI] 0.141–0.687; $P=$ 0.0039). At value 2.0, the HR was 0.185 (95% [CI] 0.051–0.675; $P=$ 0.0106).

Table 3
Cox regression model in predicting OS of oesophageal cancer patients

Parameters	$P$ value	Coeffient	HR (95% CI)	$P$ value	Coeffient	HR (95% CI)	$P$ value
	Log-rank test	Univariate Cox (continous)			Multivariate Cox (final model)
Volume_pre	0.0361	0.01	1.01 (0.99–1.03)	0.237	$-$	$-$
Volume_post	0.0121	0.026	1.03 (1.01–1.05)	0.014	0.023	1.02 (1.001–1.05)	0.04
GlobalEntropy	0.0118	$-$ 3.186	0.04 (0.005–0.35)	0.004	$-$	$-$	$-$
GlobalStd	0.0209	$-$ 0.507	0.6 (0.43–0.84)	0.002	$-$ 0.493	0.61 (0.44–0.86)	0.004
GlobalUniformity	0.0284	6.8	898.0 (7.67–105111.4)	0.005	$-$	$-$	$-$

Figure 2.

Time-independent ROC analysis revealed that the combined model offers a better predictive value for OS in the checkpoint time.

In the univariate analysis performed using the log-rank test (Table 2), the cutoff value median of pretreatment fine, medium 1–2, and coarse texture global entropies, global mean, global Std, and global uniformity were found to be significant prognostic factors for filter value 1.0, global entropies (cutoff $=$ 0.4942, $P=$ 0.0119), global mean (cutoff $=$ 2.5622, $P=$ 0.088), global Std (cutoff $=$ 4.3551, $P=$ 0.0209), and global uniformity, (cutoff $=$ 0.8133, $P=$ 0.0284). Other filter value is shown in Table 2.

3.3 Metastasis lymph node CT features for survival prediction

For metastasis lymph nodes, in the univariate analysis performed by using the Cox proportional hazards model (Table 2), pre-treatment fine texture global entropy (HR $=$ 0.186, 95% CI: 0.0567–0.608, $P=$ 0.0054), global mean (HR $=$ 0.752, 95% CI: 0.612–0.924, $P=$ 0.0068), global Std (HR $=$ 0.662, 95% CI: 0.662–0.943, $P=$ 0.0091), global uniformity (HR $=$ 34.844, 95% CI: 2.655–457.293, $P=$ 0.0069), and skewness (HR $=$ 1.609, 95% CI: 1.079–2.402, $P=$ 0.0197) as continuous variables were found to be significant prognostic factors. Posttreatment fine texture global entropy and mean can be prognostic factors. In the univariate analysis performed using the log-rank test (Table 2), the cutoff value median of posttreatment fine global mean was found to be a significant prognostic factor ( $P=$ 0.0085).

3.4 Changes in CT textures with CRT and survival prediction

In the univariate analysis performed using the Cox proportional hazards model (Table 2), coarse texture (value 2.5), primary tumor skewness and kurtosis changes as continuous variables were found to be significant prognostic factors: (HR $=$ 0.8011, 95% CI: 0.6566–0.9774, $P=$ 0.0289) and (HR $=$ 0.8889, 95% CI: 0.8211–0.9624, $P=$ 0.0037). Metastasis node-only global entropy change (fine texture value 1.0) was a prognostic factor (HR $=$ 0.8146, 95% CI: 0.7210–0.9204, $P=$ 0.0009).

3.5 Survival models

In the whole group, four parameters (post-volume, entropy, mean, global Std) as continuous variables were found to be significantly correlated with OS in univariate Cox analysis, which was consistent with the Kaplan-Meier result (Table 2). A multivariate Cox prediction model was constructed by the four parameters and was optimized in feature selection by a step-wise exclusion and inclusion of variables. The final optimal prediction model was constructed with a combination of two parameters of post-volume (Fig. 3A) and global Std (Fig. 3B, Table 3). Time-independent ROC analysis revealed that the combined model performs a better predictive value for OS than global Std in the time window of 12 months, and then post-volume in the time window of 24 months. Using 12 months as the endpoint, the combined model got an AUC of 0.79, which was higher than that of global Std (AUC $=$ 0.71, $p=$ 0.107) in assessing the chance of patient death (Fig. 2). Using 24 months as the endpoint, the combined model got an AUC of 0.79, which was higher than that of post-volume (AUC $=$ 0.69, $P=$ 0.107) and global Std (AUC $=$ 0.77, $P=$ 0.607) (Fig. 2). The patients were divided into high-risk and low-risk groups by the median risk score of the combined Cox model. Kaplan-Meier curves show that the patients in the high-risk group had significantly worse OS (log-rank test, $P=$ 0.0009) and PFS (log-rank test, $P=$ 0.0019) than those in the low-risk group (Fig. 3C).

Figure 3.

The risk mode includes post-volume and pre-treatment global Std. High-risk group, from a combined model, has significantly lower OS than the low-risk group.

3.6 Subgroup analysis risk model on different chemotherapy regimes

For patients who underwent DP, low risk increased OS over that of high risk ( $P=$ 0.003). For patients who underwent TP, low risk increased OS over that of high risk ( $P=$ 0.010). Survival models using texture parameters and morphologic response as assessed by volume did perform better than standard morphologic assessment.

4. Discussions

Our findings suggest that textural features extracted from pretreatment primary tumor and metastasis lymph node CT images may serve as an independent predictor of OS in patients with clinical stage III cancer in patients who underwent definitive CRT for primary ESCC. In the current study, we established its ability to predict survival outcomes in ESCC specifically, which not only further validates texture analysis as a prognostic biomarker that warrants prospective evaluation, but also highlights CT texture analysis as a novel method that can be used to risk-stratify and guide management decisions in ESCC. Although textural features have been established to provide predictive data on outcomes in other cancers [8, 9, 10], to date, there have been few studies of texture analysis in oesophageal cancer. There are key differences between this previously published literature and our study. Making a treatment decision mainly on a two-time point for local advanced ESCC, one is pre-treatment, two is RT 40 Gy. The Connie Yip study reported the timing of the pre-treatment at 69 days (range, 19–151 days) and posttreatment at 65 days (range, 18–384 days). CT studies were variable in the investigation and could influence the results [13]. In the current study, we all chose simulator-enhanced CT on two checkpoints, using the same CT equipment to avoid the time interval impact on the results.

In the previously published study, baseline parameters only include primary tumor have not affected the OS, then higher entropy of post treatment has decreased OS [9]. In the current study, we had different results. Higher baseline tumor entropy, mean, and global Std increased OS. One possible reason was that gross tumor volume included normal esophagus tissue. Other results came from non-contrast CT or post-treatment CT. Furthermore, iodine ions can only penetrate tumor tissue or normal tissues that underwent enhanced contrast CT; the concentration of the iodine ion represents the degree of loose tissue. Another important finding was that metastasis lymph node CT textures can be biomarkers for OS. Morphologic CT assessment alone has been shown to correlate poorly with treatment response in oesophageal cancer [14, 15]. Our study found that the survival models using the combination of pretreatment CT textures with tumor volume changes performed better than morphologic response alone. This is in keeping with the findings of a recent retrospective study [16]. Currently, how to choose chemotherapy regimens when patients undergo chemoradiotherapy is a significant challenge. We can use this risk model to guide clinical practice. Other limitations of our study include its retrospective nature. Because our image-derived score was based on a random sample of patients from one institution, the robustness of the cut-point for that score is unclear, and our results may not be generalizable to other centers. Nevertheless, our image-derived score could stratify patients into distinct risk groups and remained an independent predictor of OS even as a continuous variable. Another limitation is the potential for gas to have been present within the delineated tumors, which could influence the values of the extracted quantitative features; in this study, the post-volume is not the real volume after treatment; it is the volume after radiotherapy dose 40 Gy.

5. Conclusion

The use of texture analysis in clinical practice is still exploratory in nature, but it has the potential to add further diagnostic and prognostic information, with the advantage that it is a post-processing step, using existing CT data sets, without subjecting patients to further radiation exposure or imaging. We showed that tumor heterogeneity is associated with survival and adds additional prognostic information to morphologic tumor response. Hence, texture analysis could be a valuable response assessment tool in oesophageal cancer to select poor responders to CRT and potentially improve patient outcomes. This study adds to the growing evidence base that CT texture analysis may be useful in evaluating treatment response, but this still requires further validation in larger prospective studies before widespread implementation in clinical practice.

Footnotes

Acknowledgments

This work was supported by grants from Sichuan science and technology department key research and development project fund [2019YFS0378, 2018JY02 77], the selected excellent projects for foundation from the science and technology projects of Sichuan Province (2016), Wu Jieping Medical Foundation, Grant/Award Number: 320.6750.17237. Cancer Research Foundation of China Anti-cancer Association for Young Scientists (Grant No. CAYC18A33).

Conflict of interest

The authors declare no actual or potential financial conflicts of interest.

References

Chen

Zheng

Baade

P.D.

Zhang

Zeng

Bray

Jemal

X.Q.

and He

, Cancer statistics in China, CA Cancer J Clin 2016 66(2) (2015), 115–132.

Zeng

Zheng

Zhang

Zuo

Xia

Zou

and Chen

, Oesophageal cancer statistics in China, 2011: Estimates based on 177 cancer registries, Thorac Cancer 7(2) (2016), 232–237.

Cooper

J.S.

Guo

M.D.

Herskovic

Macdonald

J.S.

Martenson

J.A.

, Jr. Al-Sarraf

Byhardt

Russell

A.H.

Beitler

J.J.

Spencer

, et al., Chemoradiotherapy of locally advanced oesophageal cancer: long-term follow-up of a prospective randomized trial (RTOG 85-01), Radiation Therapy Oncology Group, JAMA 281(17) (1999), 1623–1627.

Wang

Zhang

Liu

Zhang

Feng

Zhou

Wang

Yin

and Xiao

, Prognosis of oesophageal squamous cell carcinoma patients with preoperative radiotherapy: Comparison of different cancer staging systems, Thorac Cancer 5(3) (2014), 204–210.

Wang

Xiao

Liu

Zhang

Sun

Feng

, et al., Residual lymph node status is an independent prognostic factor in oesophageal squamous cell Carcinoma with pathologic T0 after preoperative radiotherapy, Radiat Oncol 10 (2015), 142.

Borghesi

Hawkins

M.A.

and Tait

, Oesophagectomy after definitive chemoradiation in patients with locally advanced ooesophageal cancer, Clin Oncol (R Coll Radiol) 20(3) (2008), 221–226.

D’Journo

X.B.

Michelet

Dahan

Doddoli

Seitz

J.F.

Giudicelli

Fuentes

P.A.

and Thomas

P.A.

, Indications and outcome of salvage surgery for ooesophageal cancer, Eur J Cardiothorac Surg 33(6) (2008), 1117–1123.

Segal

Sirlin

C.B.

Ooi

Adler

A.S.

Gollub

Chen

Chan

B.K.

Matcuk

G.R.

Barry

C.T.

Chang

H.Y.

et al., Decoding global gene expression programs in liver cancer by noninvasive imaging, Nat Biotechnol 25(6) (2007), 675–680.

Ganeshan

Miles

K.A.

Young

R.C.

and Chatwin

C.R.

, Texture analysis in non-contrast enhanced CT: impact of malignancy on texture in apparently disease-free areas of the liver, Eur J Radiol 70(1) (2009), 101–110.

10.

Aerts

H.J.

Velazquez

E.R.

Leijenaar

R.T.

Parmar

Grossmann

Carvalho

Bussink

Monshouwer

Haibe-Kains

Rietveld

, et al., Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach, Nat Commun 5 (2014), 4006.

11.

Rice

T.W.

Blackstone

E.H.

and Rusch

V.W.

, 7th edition of the AJCC Cancer Staging Manual: esophagus and esophagogastric junction, Ann Surg Oncol 17(7) (2010), 1721–1724.

12.

Zhang

Fried

D.V.

Fave

X.J.

Hunter

L.A.

Yang

and Court

L.E.

, IBEX: an open infrastructure software platform to facilitate collaborative work in radiomics, Med Phys 42(3) (2015), 1341–1353.

13.

Yip

Landau

Kozarski

Ganeshan

Thomas

Michaelidou

and Goh

, Primary oesophageal cancer: heterogeneity as potential prognostic biomarker in patients treated with definitive chemotherapy and radiation therapy, Radiology 270(1) (2014), 141–148.

14.

S.H.

Rau

K.M.

H.I.

Wang

Y.M.

Tien

W.Y.

Liang

J.L.

and Lin

W.C.

, Pre-treatment maximal ooesophageal wall thickness is independently associated with response to chemoradiotherapy in patients with T3-4 ooesophageal squamous cell carcinoma, Eur J Cardiothorac Surg 42(6) (2012), 958–964.

15.

van Heijl

Phoa

S.S.

van Berge Henegouwen

M.I.

Omloo

J.M.

Mearadji

B.M.

Sloof

G.W.

Bossuyt

P.M.

Hulshof

M.C.

Richel

D.J.

Bergman

J.J.

, et al., Accuracy and reproducibility of 3D-CT measurements for early response assessment of chemoradiotherapy in patients with ooesophageal cancer, Eur J Surg Oncol 37(12) (2011), 1064–1071.

16.

Ganeshan

Skogen

Pressney

Coutroubis

and Miles

, Tumour heterogeneity in ooesophageal cancer assessed by CT texture analysis: preliminary evidence of an association with tumour metabolism, stage, and survival, Clin Radiol 67(2) (2012), 157–164.

Textural features based enhanced contrast CT images predicts prognosis to concurrent chemoradiotherapy in stage III esophageal squamous cell cancer

Abstract

INTRODUCTION:

METHODS:

RESULTS:

CONCLUSION:

1. Introduction

2. Methods

2.1 Treatment

2.2 CT image acquisition

2.3 Texture analysis

Table 1 Patient, tumor and patient characteristics and postoperative result dataset

3.1 Patients

3.2 Primary tumor CT features for survival prediction

Table 3 Cox regression model in predicting OS of oesophageal cancer patients

3.4 Changes in CT textures with CRT and survival prediction

3.5 Survival models

4. Discussions

5. Conclusion

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Patient, tumor and patient characteristics and postoperative result dataset

Table 3
Cox regression model in predicting OS of oesophageal cancer patients