Volumetric histogram analysis of apparent diffusion coefficient for predicting pathological complete response and survival in esophageal cancer patients treated with chemoradiotherapy

https://doi.org/10.1016/j.amjsurg.2019.07.040Get rights and content

Highlights

  • Esophageal cancer patients achieving pathological complete response (pCR) showed better survival.

  • Apparent diffusion coefficient (ADC) is a biomarker derived from diffusion-weighted MRI.

  • Histogram analysis of ADC was useful for predicting pCR with accuracy of 86.2%.

  • Histogram analysis of ADC can predict survival of esophageal cancer treated with chemoradiotherapy.

Abstract

Background

The purpose of the study was to evaluate whether histogram analysis of apparent diffusion coefficient (ADC) can predict pathological complete response (pCR) and survival in patients with esophageal squamous cell carcinoma (ESCC) after chemoradiotherapy (CRT).

Methods

We retrospectively identified 58 patients with ESCC who underwent surgery after CRT between 2007 and 2016. Associations of pretreatment histogram derived ADC parameters with pathological response and survival were analyzed.

Results

Tumors achieved pCR (10 patients, 17.2%) showed significant lower ADC, higher kurtosis, and higher skewness than those of non-pCR (p = 0.005, 0.007, <0.001, respectively). Receiver operating characteristics analysis demonstrated skewness was the best predictor for pCR (AUC = 0.86), with a cut off value of 0.50 (accuracy, 86.2%). In Kaplan-Meier analysis, patients with higher skewness tumors (≥0.50) showed a significantly better recurrence free survival (p = 0.032, log-rank).

Conclusions

Histogram analysis of ADC can enable prediction of pCR and survival in ESCC patients treated with preoperative CRT.

A short summary

ADC histogram analysis can be an imaging biomarker for esophageal cancer patients treated with CRT.

Introduction

Esophageal squamous cell carcinoma (ESCC) have extremely poor prognosis due to high metastatic potential and high tendency of tumor invasion into the surrounding organs. Esophagectomy with radical lymph node dissection is the gold standard treatment, but it has been reported that an early recurrence (within two years after surgery) is often observed in patients with locally advanced ESCC.1 In a couple of previous meta-analyses, preoperative chemoradiation therapy (CRT) for patients with resectable ESCC has shown a significant increase of survival, compared to surgery alone.2,3 However, it was reported that preoperative CRT significantly prolongs survival in only those who could achieve pathological complete response (pCR), defined as the absence of disease in both the esophagus and lymph nodes (T0N0) in the resected specimen.4,5 On the other hand, CRT induces several side effects, such as bone marrow suppression, esophagitis, pericarditis, pneumonia, that sometimes result in death.6 Furthermore, if a patient is insensitive to preoperative CRT, this patient may miss the opportunity for curative surgery due to progression of disease beyond resectability and the distant spread. Therefore, it is important to predict the response to preoperative CRT in order to avoid unnecessary adverse events and choose the best treatment for each individual.

On the other hand, the abnormal structure of tumor is a well-recognized feature of malignancies. Malignant tumors essentially has heterogenous biological structures, which may influence drug delivery and therapeutic outcome.7 Therefore, analysis of “tumor heterogeneity” can be a biomarker for cancer treatment, and various techniques are proposed for the quantification of intratumor heterogeneity.8,9 Histogram analysis of medical images is one of the methods to quantify structural abnormality of the tumor, and has a potential to be a biomarker for cancer treatment and progression.10,11 In magnetic resonance imaging (MRI), it has been reported that application of histogram analysis to diffusion-weighted MR imaging (DWI) could help to further increase prediction of histological features of tumors and response to treatment in various tumor types, such as orbital tumor,12 thyroid cancer,13 head and neck cancer.14 DWI is a functional imaging technique based on measuring the random Brownian motion of water molecules, which can be quantified by the apparent diffusion coefficient (ADC) value. ADC value has been reported as a tool that can reflect tumor structures such as stroma, fibrosis, cellularity, and angiogenesis.15,16

However, no study has reported that abnormality in the tumor structure measured by histogram analysis of ADC can predict pCR in patients with ESCC. Therefore, we applied histogram analysis to ADC values in the tumor, and retrospectively assessed whether histogram parameters before CRT could be predictive biomarkers for pCR and survival in patients with ESCC.

Section snippets

Patient population

Fifty-eight patients with ESCC admitted to our institution between February 2007 and December 2016 were included into this retrospective study. All these 58 patients underwent curative surgery after CRT, and all of them got MRI before CRT. They had an upper endoscopic examination with tumor biopsy, barium esophagography, chest and abdominal computed tomography (CT) scans, and positron emission tomography (PET) with 18F-fluorodeoxy glucose (18F-FDG) to determine the clinical stage according to

Patient background and pathological response to CRT

The study population included 58 patients (51 male and 7 female), with their ages ranging from 45 to 78 (median 61) years old. The characteristics of the 58 patients are given in Table 1. The median follow-up period was 17.9 months (range 1.9–114.3 months). Pathological response to CRT of the main tumor was as follows: 12 patients were Grade 3 (20.7%), 24 were Grade 2 (41.4%), 22 were Grade 1 (37.9%). However, 2 of the 12 Grade 3 patients had pathologically proven metastatic lymph nodes in

Discussion

ADC mean, which is the average value within the ROIs in tumor tissues, is commonly used for getting qualitive and quantitative information on tissue characterization (differentiating benign from malignant lesions) and monitoring treatment response,20 but pretreatment ADC value may not be valuable enough for predicting CRT response in a meta-analysis.21 Heterogeneity in the tumor structure is a well-recognized feature of malignancy, which is associated with adverse tumor biology. It can be

Conclusions

To our knowledge, no studies have been published on the efficacy of ADC histogram analysis for predicting pCR in ESCC. Our results indicate that ADC skewness is the most useful ADC derived parameter for predicting pCR and survival in ESCC patients treated with preoperative CRT. We believe that prediction of pCR prior to treatment could contribute to choosing the optimal therapeutic strategy for each individual.

Conflicts of interest

There are no financial or other relations that could lead to a conflict of interest.

Acknowledgements

This work was supported by JSPS KAKENHI Grant Number JP 18K07666.

References (35)

  • N. Just

    Improving tumour heterogeneity MRI assessment with histograms

    Br J Canc

    (2014)
  • L. Alic et al.

    Quantification of heterogeneity as a biomarker in tumor imaging: a systematic review

    PLoS One

    (2014)
  • K. Hayano et al.

    Imaging biomarkers for the treatment of esophageal cancer

    World J Gastroenterol

    (2019)
  • M. Chen et al.

    Systematic review and meta-analysis of tumor biomarkers in predicting prognosis in esophageal cancer

    BMC Canc

    (2013)
  • X. Dong et al.

    The impact of intratumoral metabolic heterogeneity on postoperative recurrence and survival in resectable esophageal squamous cell carcinoma

    Oncotarget

    (2017)
  • X.Q. Xu et al.

    Utility of histogram analysis of ADC maps for differentiating orbital tumors

    Diagnostic Interv Radiol

    (2016)
  • S. Schob et al.

    Histogram analysis of diffusion weighted imaging at 3T is useful for prediction of lymphatic metastatic spread, proliferative activity, and cellularity in thyroid cancer

    Int J Mol Sci

    (2017)
  • View full text