Accuracy of multidetector-row CT for restaging after neoadjuvant treatment in patients with oesophageal cancer

Abstract

OBJECTIVES: To assess the diagnostic accuracy of 64-multidetector CT (MDCT) for restaging of patients with oesophageal cancer undergoing neoadjuvant therapy. METHODS: Results of pathological staging were correlated with those from 64-MDCT before and after neoadjuvant treatment in 35 patients using the American Joint Committee on Cancer/TNM classification (7th edition). CT response was determined using the Response Evaluation Criteria in Solid Tumours (RECIST) method, modified for one-dimensional tumour diameter measurement. RESULTS: 64-MDCT predicted T stage correctly in 34 % (12/35), overstaged in 49 % (17/35) and understaged in 17 % (6/35). Sensitivity/specificity values were as follows: T0, 20 %/92 %; T1-T2, 31 %/59 %; T3, 60 %/64 %; T4, 100 %/4 %. Negative predictive values for T3/T4 were 80 %/100 %. MDCT accurately predicted complete histopathological response in 20 % (accuracy 74 %) and overstaged in 80 %. Tumour regression grade was predicted correctly in only 8 % (2/25) and underestimated in 68 % (17/25). Accurate N stage was noted in 69 % (24/35). CONCLUSION: Although MDCT tends to be able to exclude advanced tumour stages (T3, T4) with a higher likelihood, the diagnostic accuracy of high resolution MDCT for restaging oesophageal cancer and assessing the response to neoadjuvant therapy has not improved in comparison to older-generation CT. Therefore, the future assessment of oesophageal tumour response should focus on combined morphologic and metabolic imaging. KEY POINTS: • Multidetector CT (MDCT) has been beneficial for the evaluation of many tumours.• However diagnostic accuracy for restaging oesophageal cancer has not improved with MDCT.• MDCT tends to be able to exclude advanced tumour stages (T3/T4).• MDCT has a low accuracy for determining lymph node metastasis.• Oesophageal tumour response should be assessed by combined morphological and metabolic imaging. Conclusion Although MDCT tends to be able to exclude advanced tumour stages (T3, T4) with a higher likelihood, the diagnostic accuracy of high resolution MDCT for restaging oesophageal cancer and assessing the response to neoadjuvant therapy has not improved in comparison to older-generation CT. Therefore, the future assessment of oesophageal tumour response should focus on combined morphologic and metabolic imaging. Key Points • Multidetector CT (MDCT) has been beneficial for the evaluation of many tumours.• However diagnostic accuracy for restaging oesophageal cancer has not improved with MDCT.• MDCT tends to be able to exclude advanced tumour stages (T3/T4).• MDCT has a low accuracy for determining lymph node metastasis.• Oesophageal tumour response should be assessed by combined morphological and metabolic imaging.

Diagnostic performance of MRI for assessment of response to neoadjuvant chemoradiotherapy in oesophageal cancer

“…To implement non‐surgical treatment for advanced oesophageal cancer it is critical to accurately identify pCR. CT and fluorodeoxyglucose (FDG) PET/CT are both inaccurate in discriminating residual disease from pCR owing to the presence of wall thickening and/or radiation oesophagitis. The same is true for endoscopic ultrasonography (EUS), which has an accuracy of only 36 per cent for ypT determination.…”

Performance of different imaging modalities in assessment of response to neoadjuvant therapy in primary esophageal cancer

“…The few published studies evaluating the predictive value of CT response assessment after neoadjuvant therapy in esophageal cancer have been inconclusive. Although changes in CT tumor volume showed greater correlation with pathological response than changes in tumor diameter and were associated with lower interobserver variability, it is only a modest predictor of pathological response at best (Table ) . The cut‐offs used for changes in tumor volume to differentiate responders from non‐responders ranged from 10% to 20% .…”

“…The few studies that evaluated early CT response assessment showed limited utility in the neoadjuvant CRT and chemotherapy setting with poor sensitivity and negative predictive value (NPV) (Table ). The ability of late CT assessment in predicting pathological response and TNM stage is poor (Table ) . There is suggestion that CT may perform better in those treated with neoadjuvant chemotherapy compared with CRT, although this is based on very few studies with no direct comparison of its predictive ability between the two treatment modalities in any of these studies (Table ) …”

“…Although changes in CT tumor volume showed greater correlation with pathological response than changes in tumor diameter and were associated with lower interobserver variability, 23 it is only a modest predictor of pathological response at best ( Table 2). 18,[23][24][25][26][27][28] The cut-offs used for changes in tumor volume to differentiate responders from nonresponders ranged from 10% to 20%. 23,24 Two studies did not find any significant correlation between CT volumetric changes and pathological response, 25,26 with a modest performance as depicted by an area under the curve derived from the receiver operating characteristic curve of 0.63 (95% confidence interval [95% CI] 0.45-0.82).…”

“…The ability of late CT assessment in predicting pathological response and TNM stage is poor ( Table 2). 23,[25][26][27][28] There is suggestion that CT may perform better in those treated with neoadjuvant chemotherapy compared with CRT, although this is based on very few studies with no direct comparison of its predictive ability between the two treatment modalities in any of these studies (Table 2). 23,26,27…”