Intratumoural immune signature to identify patients with primary colorectal cancer who do not require follow-up after resection: an observational study

Background: Following surgical and adjuvant treatment of primary colorectal cancer, many patients are routinely followed up with axial imaging (most commonly computerised tomography imaging) and blood carcinoembryonic antigen (a tumour marker) testing. Because fewer than one-fifth of patients will relapse, a large number of patients are followed up unnecessarily. Objectives: To determine whether or not the intratumoural immune signature could identify a cohort of patients with a relapse rate so low that follow-up is unnecessary. Design: An observational study based on a secondary tissue collection of the tumours from participants in the FACS (Follow-up After Colorectal Cancer Surgery) trial. Setting and participants: Formalin-fixed paraffin-embedded tumour tissue was obtained from 550 out of 1202 participants in the FACS trial. Tissue microarrays were constructed and stained for cluster of differentiation (CD)3+ and CD45RO+ T lymphocytes as well as standard haematoxylin and eosin staining, with a view to manual and, subsequently, automated cell counting. Results: The tissue microarrays were satisfactorily stained for the two immune markers. Manual cell counting proved possible on the arrays, but manually counting the number of cores for the entire study was found to not be feasible; therefore, an attempt was made to use automatic cell counting. Although it is clear that this approach is workable, there were both hardware and software problems; therefore, reliable data could not be obtained within the time frame of the study. Limitations: The main limitations were the inability to use machine counting because of problems with both hardware and software, and the loss of critical scientific staff. Findings from this research indicate that this approach will be able to count intratumoural immune cells in the long term, but whether or not the original aim of the project proved possible is not known. Conclusions: The project was not successful in its aim because of the failure to achieve a reliable counting system. Future work: Further work is needed to perfect immune cell machine counting and then complete the objectives of this study that are still relevant.