PhD defence M.J. (Maartje) Valkema

Below is a brief summary about the dissertation:

In patients with locally advanced esophageal cancer, standard treatment comprises neoadjuvant chemoradiotherapy (nCRT) and esophagectomy. One-third of these patients has no residual tumor cells in the surgical resection specimen. An alternative strategy after nCRT might be active surveillance, which is currently investigated in the SANO trial. During active surveillance, surgery is performed only in case of proven residual tumor after nCRT, without distant dissemination. In that context, the goal of this thesis was to find tumor characteristics ("features") to distinguish patients with different treatment responses to nCRT and surgery.

The main focus was on different image analysis techniques. Visual and complex analysis methods (using “radiomics”) of FDG PET/CT scans were shown inaccurate for tumor detection in the esophagus. The FDG PET/CT up to 12 weeks after nCRT cannot reliably distinguish residual tumor from radiation-induced esophagitis. The performance of another scan, the FDG PET/MRI, was found comparable to the FDG PET/CT. In contrast, serial FDG PET/CT scans up to one year after nCRT during active surveillance were found valuable to detect primary tumor regrowth.

In the next part of the thesis, histopathological features were studied on pre-treatment biopsies and post-treatment resection specimens. A “fragmented pattern” in the resection specimen after nCRT, as opposed to a “shrinkage” pattern, was associated with higher risk of death and a higher risk of disease recurrence. Finally, acellular mucin pools, signet-ring cells and poorly cohesive cells in biopsies before nCRT were not correlated with histopathological response.