Radiological prediction of positive circumferential resection margin in oesophageal cancer. Issue 107 (October 2018)

• Record Type: Journal Article
• Authors: Foley, K.G.; Christian, A.; Patel, N.; Lewis, W.G.; Roberts, S.A.
• Is Part Of: European journal of radiology. Issue 107(2018)
• Page Start: 119
• Page End: 124
• Publication Date: 2018-10
• Subjects: Margins of excision -- Esophageal neoplasms -- Positron-emission tomography -- Endosonography; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2018.08.027
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: Circumferential resection margin status is an important prognostic indicator. EUS T-stage is an independent predictor of a positive CRM. EUS should continue to be used in a multi-modality approach to oesophageal cancer staging. Abstract: Purpose: A positive circumferential resection margin (CRM) is regarded as a poor prognostic indicator in oesophageal cancer (OC) but its prediction can be challenging. MRI is used to predict a threatened CRM in rectal cancer but is not commonly performed in OC unlike PET/CT, which is now routinely used. Therefore, this study assessed the additional predictive value of PET-defined tumour variables compared with EUS and CT T-stage. The prognostic significance of CRM status was also assessed. Materials and Methods: This retrospective study included 117 consecutive patients [median age 64.0 (range 24–78), 102 males, 110 adenocarcinomas, 6 squamous cell carcinoma (SCC), 1 neuro-endocrine] treated between 1st March 2012 and 31st July 2015. A binary logistic regression model tested 5 staging variables; EUS T-stage (≤T2 vs ≥ T3), CT T-stage (≤T2 vs ≥ T3), PET metabolic tumour length (MTL), PET metabolic tumour width (MTW) and the maximum standardised uptake value (SUVmax ). Results: The CRM was positive in 43.6%. Sixty-seven (57.3%) patients received neo-adjuvant chemotherapy (NACT), 31 patients (26.5%) underwent surgery alone and 19 patients (16.2%) had neo-adjuvant chemo-radiotherapy (NACRT). Median overall survival (OS) was 36.0 months (95% confidence interval (CI) 24.1–47.9) and the 2-year OS was 55.4%. A binary logistic regression model showed EUS ≥ T3 tumours were independently and significantly more likely to have a positive CRM than EUS ≤ T2 tumours (HR 5.188, 95% CI 1.265–21.273, p = 0.022). CT T-stage, PET MTL, PET MTW and SUVmax were not significantly associated with CRM status (p = 0.783, 0.852, 0.605 and 0.413, respectively). There was a significant difference in OS between CRM positive and negative groups (X 2 4.920, df 1, p = 0.027). Conclusion: Advanced EUS T-stage is associated with a positive CRM, but PET-defined tumour variables are unlikely to provide additional predictive information. This study demonstrates the continued benefit of EUS as part of a multi-modality OC staging pathway.