Automated prognostic pattern detection shows favourable diffuse pattern of FOXP3(+) Tregs in follicular lymphoma.

BACKGROUND: Histopathological prognostication relies on morphological pattern recognition, but as numbers of biomarkers increase, human prognostic pattern-recognition ability decreases. Follicular lymphoma (FL) has a variable outcome, partly determined by FOXP3 Tregs. We have developed an automated method, hypothesised interaction distribution (HID) analysis, to analyse spatial patterns of multiple biomarkers which we have applied to tumour-infiltrating lymphocytes in FL. METHODS: A tissue microarray of 40 patient samples was used in triplex immunohistochemistry for FOXP3, CD3 and CD69, and multispectral imaging used to determine the numbers and locations of CD3(+), FOXP3/CD3(+) and CD69/CD3(+) T cells. HID analysis was used to identify associations between cellular pattern and outcome. RESULTS: Higher numbers of CD3(+) (P=0.0001), FOXP3/CD3(+) (P=0.0031) and CD69/CD3(+) (P=0.0006) cells were favourable. Cross-validated HID analysis of cell pattern identified patient subgroups with statistically significantly different survival (35.5 vs 142 months, P=0.00255), a more diffuse pattern associated with favourable outcome and an aggregated pattern with unfavourable outcome. CONCLUSIONS: A diffuse pattern of FOXP3 and CD69 positivity was favourable, demonstrating ability of HID analysis to automatically identify prognostic cellular patterns. It is applicable to large numbers of biomarkers, representing an unsupervised, automated method for identification of undiscovered prognostic cellular patterns in cancer tissue samples.

Distinct patterns of infiltrating CD8+ T cells in HPV+ and CD68 macrophages in HPV- oropharyngeal squamous cell carcinomas are associated with better clinical outcome but PD-L1 expression is not prognostic.

Immunotherapies are beginning to revolutionise treatment paradigms in oncology with monoclonal antibodies (mAb) targeting T-cell co-inhibitory (e.g. PD-1/PD-L1) and co-stimulatory pathways (e.g. CTLA-4/CD28) demonstrating clinical utility. Some clinical studies demonstrate that responsiveness to PD-1/PD-L1 mAb therapy is greater in patients with expression of PD-L1 in the tumour microenvironment. However, robust responses have also been observed in patients with low or absent expression of PD-L1. Using multiplex immuno-fluorescent labelling we sought to determine how infiltration of tumours by CD8+ T-cells, their expression of PD-1, and the expression of PD-L1 on both tumours and CD68 cells (macrophages) correlated with HPV status and outcome in a cohort of 124 oropharyngeal squamous cell carcinomas (OPSCC).

Investigation of radiosensitivity gene signatures in cancer cell lines.

Intrinsic radiosensitivity is an important factor underlying radiotherapy response, but there is no method for its routine assessment in human tumours. Gene signatures are currently being derived and some were previously generated by expression profiling the NCI-60 cell line panel. It was hypothesised that focusing on more homogeneous tumour types would be a better approach. Two cell line cohorts were used derived from cervix [n = 16] and head and neck [n = 11] cancers. Radiosensitivity was measured as surviving fraction following irradiation with 2 Gy (SF2) by clonogenic assay. Differential gene expression between radiosensitive and radioresistant cell lines (SF2 median) was investigated using Affymetrix GeneChip Exon 1.0ST (cervix) or U133A Plus2 (head and neck) arrays. There were differences within cell line cohorts relating to tissue of origin reflected by expression of the stratified epithelial marker p63. Of 138 genes identified as being associated with SF2, only 2 (1.4%) were congruent between the cervix and head and neck carcinoma cell lines (MGST1 and TFPI), and these did not partition the published NCI-60 cell lines based on SF2. There was variable success in applying three published radiosensitivity signatures to our cohorts. One gene signature, originally trained on the NCI-60 cell lines, did partially separate sensitive and resistant cell lines in all three cell line datasets. The findings do not confirm our hypothesis but suggest that a common transcriptional signature can reflect the radiosensitivity of tumours of heterogeneous origins.