Hereditary predisposition and prognostic prediction in cancer

Cancer is a genetic disease. The cancer cells evolve clonally through changes of their genome, gaining malignant characteristics. In hereditary cancer syndromes, a genetic aberration which puts the cells one step closer to being malignant is inherited and present throughout whole body. Not all genes associated with hereditary cancer syndromes have been discovered, and the spectrum of cancers connected to each cancer-associated gene is not yet known. When it is feasible, the genetic changes within the tumour can be analysed by performing a tissue biopsy. Alternatively, analysis of cell-free DNA from a blood sample might give the same information. Cell-free DNA is released into the blood stream from both malignant and nonmalignant cells, and it carries the same genetic aberrations as the cells it comes from. The aims of this thesis were to improve oncogenetic counselling, by increasing the knowledge about who should be offered genetic tests and what genes to test, and to improve treatment decision-making, by developing a clinical pipeline for a new cancer biomarker, cell-free tumour DNA.

In Study I, we explored the genetic cause for a suspected highly penetrant, autosomal dominant rectal cancer syndrome in one family. We analysed the blood samples from six family members by massive parallel sequencing, and found six variants that could contribute to the increased risk for rectal cancer in the family. The variants were in the genes CENPB, ZBTB20, CLINK, LRRC26, TRPM1, and NPEPL1. None of the variants have a known connection to hereditary cancer syndromes and further studies are needed to investigate their potential involvement.

In Study II, we evaluated the efficacy of performing massive parallel sequencing in patients with three or more primary tumours. We found two likely pathogenic variants in two individuals. In a woman with a clinical diagnosis of MEN1, we found a synonymous variant in the MEN1 gene. By RNA analysis, we could show that the variant activates a cryptic splice site, which leads to a deletion of 14 nucleotides, and a frameshift in the RNA transcript. The variant segregated with disease in the family, and it was classified as likely disease-causing. In a woman with multiple Lynch-associated tumours and microsatellite instability, we found a synonymous variant in the MLH1 gene. This variant has recently been described as an epimutation, leading to methylation of the MLH1 promoter, and Lynch syndrome. Both these genes had previously been analysed in blood samples from the two women, but the variants had not been recognized as pathogenic. We therefore conclude that genetic re-analysis may be beneficial, especially for patients with a clear clinical diagnosis of a specific hereditary cancer syndrome.

In Study III, we developed a clinical method for detecting cancer-associated CNAs (copy number alterations) in cell-free tumour DNA from patients with gastro-oesophageal cancer. The same clinical laboratory protocol as for non-invasive prenatal testing at Karolinska University Hospital, Stockholm, was used, and for the last bioinformatic steps WISECONDOR and ichorCNA were used. This generated a list of potentially cancerassociated CNAs, which we manually annotated. We compared the variants to the tissue sample from the same participant, and also with a control set of samples from pregnant women. We found cancer-associated CNAs in 14/26 (54%) of the individuals with detectable cancer-associated CNAs in the tissue samples. Potentially clinically actionable amplifications were detected in the genes VEGFA, EGFR, and FGFR2. This study showed that the clinical pipeline we have set up for the non-invasive prenatal testing workflow can be used in the oncogenetic field, and CNA analysis in patients with gastro-oesophageal cancer might be used as a biomarker in the future.

In conclusion, we report six new genes with a potential involvement in a hereditary rectal cancer syndrome, we suggest re-analysis might be beneficial for patients without a molecular diagnosis but who fulfil testing criteria for specific cancer syndromes, and we show that the non-invasive prenatal testing clinical pipeline can also be used in the oncogenetic setting.

List of scientific papers

I. Wallander K, Thutkawkorapin J, Sahlin E, Lindblom A, Lagerstedt-Robinson K. Massive parallel sequencing in a family with rectal cancer. Hered Cancer Clin Pract. 2021 Apr 7;19(1):23.
https://doi.org/10.1186/s13053-021-00181-2

II. Wallander K, Thonberg H, Nilsson D, Tham E. Massive parallel sequencing in individuals with multiple primary tumours reveals the benefit of reanalysis. Hered Cancer Clin Pract. 2021 Oct 28; 19(1):46.
https://doi.org/10.1186/s13053-021-00203-z

III. Wallander K, Eisfeldt J, Lindblad M, Nilsson D, Billiau K, Foroughi H, Nordenskjöld M, Liedén A, Tham E. Cell-free tumour DNA analysis detects copy number alterations in gastro-oesophageal cancer patients. PLoS One. 2021 Feb 4;16(2):e0245488.
https://doi.org/10.1371/journal.pone.0245488