Response to induction therapy in oesophageal and cardia carcinoma using Mandard tumour regression grade or size of residual foci

Tumour regression grade (TRG) is used to evaluate responses to induction therapy in cancer of the oesophagus or cardia. This study aimed to determine whether inclusion of node category could improve the prognostic accuracy provided by TRG, and explore the prognostic value of an alternative classification based on size of residual foci and node category.Patients with oesophageal or cardia cancer treated with neoadjuvant chemoradiotherapy followed by resection were studied. Treatment-induced response at the primary site was evaluated by TRG and by a method whereby patients were classified as having no residual cancer, minimal residual disease (MRD) or as non-responders.Between 2000 and 2007, 108 patients underwent resection. Disease-related survival decreased with increasing TRG in node-negative (N0) patients (P < 0.001), whereas in node-positive (N+) patients it was poor irrespective of TRG (P = 0.241). For N0 disease, 3-year survival in patients with MRD (58 (95 per cent confidence interval 26 to 80) per cent) was intermediate between that in patients with no residual cancer (85 (70 to 93) per cent) and non-responders (28 (4 to 59) per cent). Worst prognosis was for N+ disease (21 (9 to 36) per cent).Node category should be considered when evaluating response to induction therapy in oesophageal or cardia cancer. A new classification based on size of residual foci and node category seems promising.