Tumour development in Lynch syndrome : genes load the gun, lifestyle pulls the trigger?

Lynch syndrome (LS) is caused by a dominantly inherited pathogenic variant in one of the DNA mismatch repair genes. Persons with LS are predisposed to early onset cancer, mainly colorectal and endometrial cancer, and colorectal adenomas which are precursor lesions of colorectal cancer. Cancer risk estimates are variable within and between families with the same mutated gene which suggests that, similar to cancer in the general population, lifestyle-related factors may be involved in cancer development. A limited number of studies on the influence of lifestyle-related factors on cancer risk exist for persons with LS. This thesis aimed at evaluating the association between the inflammatory potential of the diet and the risk of colorectal tumours (i.e. colorectal adenomas and carcinomas), between height and the risk of both colorectal and endometrial cancer, and between body mass index (BMI) at young adulthood and the risk of cancer at all sites and cancer outside the colorectum and/or outside the endometrium for persons with LS. It was also explored whether a colorectal tumour diagnosis was associated with a change in lifestyle habits. For the research aims, data of persons with LS residing in the Netherlands and participating in the GEOLynch study has been used separately or has been  used after harmonization with data of persons with LS residing in Australia, New Zealand, Canada and the USA from the Colon Cancer Family Registry (CCFR).

In chapter 2, dietary intake of 457 participants of the GEOLynch study was determined with a food frequency questionnaire (FFQ) and used to calculate the adapted dietary inflammatory index (ADII). A higher ADII score reflects a higher inflammatory potential of an individual’s diet. After a median follow-up time of 59 months, 200 (43.8%) participants developed a colorectal adenoma or carcinoma (CRT). A higher inflammatory potential of the diet was not associated with the risk of CRTs for persons with LS.

Harmonized data of both the GEOLynch study and CCFR has been used to evaluate the association between height and colorectal cancer risk for men and women separately and between height and endometrial cancer risk for women in chapter 3. Self-reported height of 1155 men and 1553 women was used and cancer diagnoses were obtained from or confirmed, where possible, in medical records and/or pathology reports. After 28 279 and 37 090 person years for men and women respectively, colorectal cancer was diagnosed in 511 (44.2%) men and 436 (28.1%) women. For endometrial cancer, 1544 women were included of whom 171 (11.1%) were diagnosed with endometrial cancer after 39 227 person years. No evidence for an association was observed between height and colorectal cancer for men and women, and between height and endometrial cancer for women with LS.

In chapter 4, the association between body fatness, as reflected by BMI, at young adulthood and cancer risk for persons with LS was evaluated. Harmonized data of 1044 men and 1446 women with LS from the GEOLynch and CCFR studies was used. BMI at young adulthood  was calculated with self-reported height and recalled weight at the age of 18 or 20 years. Where possible, medical records and/or pathology reports were used to  identify  cancer diagnoses. A  5 kg/m2 increment in BMI at young adulthood was associated with an increased risk of cancer  at all sites for women, but not for men. No association was observed between BMI at young adulthood and cancer outside the colon for men and women with LS, and for cancers outside both the colorectum and endometrium for women with LS.

Data of the GEOLynch study was used to explore if a colorectal tumour diagnosis was associated with a change in lifestyle habits for persons with LS in chapter 5. A FFQ and a general questionnaire about lifestyle habits were completed by 324 participants at both baseline and after a median follow-up of 82.0 [interquartile range, 71.4-86.3] months. A CRT was diagnosed in 146 (45.1%) persons between baseline and follow-up. Apart from a potentially higher likelihood of smoking cessation for those with a CRT diagnosis compared to those without a CRT diagnosis, no evidence was observed for a difference in change in intake of energy, alcohol, red meat, processed meat, dairy, fruit, vegetables and dietary fibre, and in adult BMI, physical activity level and non-steroidal anti-inflammatory drugs use for persons with LS.

No previous research has been published in which the association between the inflammatory potential of the diet and colorectal tumour risk has been evaluated for persons with LS. Nor does another publication exist in which the association between a colorectal tumour diagnosis and a change in lifestyle habits has been investigated. For height and BMI at young adulthood, inconsistent results for the association between height and colorectal tumours were observed in previous research for persons with LS whereas a higher BMI at young adulthood was associated with an increased risk of colorectal cancer, but not for endometrial cancer. For the general population, a more pro-inflammatory potential of the diet seems to be associated with an increased risk of colorectal tumours, being taller or having a higher BMI at young adulthood increases the risk of colorectal and endometrial cancer, and inconsistent results are reported for an association between a cancer diagnosis and a change in lifestyle habits. The observed contradiction in associations between lifestyle-related factors and tumour risk for persons with LS compared with the general population may be explained by differences in tumour development for persons with LS versus the general population. Moreover, the discovery of LS-associated colorectal cancer development without adenoma or polyp formation, may  explain why associations between lifestyle-related factors and colorectal adenoma risk do not agree with studies in which LS-associated colorectal cancer was used as endpoint. Methodological issues that resulted from combining data of the GEOLynch study and CCFR, and consequences of the applied data analyses did not introduce biases to such an extent that they can explain the observed results of this thesis.

Overall, results of this thesis suggest that the inflammatory potential of the diet and height are not associated with tumour development for persons with LS. For BMI at young adulthood, a positive association with cancer at all sites is observed for women, but not for men. A colorectal tumour diagnosis does not seem to trigger a change in lifestyle factors. Current cancer prevention recommendations for the general population do not seem to be harmful for the cancer burden in persons with LS and hence, the cancer prevention recommendation to eat a healthy diet and maintain a healthy body weight – also at young adulthood - may be suggested for persons with LS as well. Future studies to lifestyle-related factors and tumour risk for persons with LS may focus on evaluating whether a change in lifestyle-related factors influences subsequent tumour risk. Preferably, such studies should consider molecular characteristics of the developed tumours and/or present results by LS-causing mutated gene. The question mark that ends the subtitle of this thesis, i.e. Genes load the gun, lifestyle pulls the trigger?, will currently remain and cannot be replaced by a period (yet).