Lynch syndrome in the Asian populations

Abstract

Lynch syndrome is an autosomal dominant genetic disorder. Mutation carriers are at significant risk of developing colorectal and a variety of extra-colonic cancers, often at a younger age compared to sporadic cancers. The current guidelines on detection and management of Lynch syndrome are based upon studies in Caucasian populations with Lynch syndrome. The applicability of these guidelines to Asian populations is not known, as the prevalence and risks associated with mutations in mismatch repair genes may be different compared to Caucasian populations. The aims of this thesis were:

1. To assess the applicability of the current guidelines for the screening of Lynch syndrome and cancer surveillance strategies in the Asian populations (Chapter 2). 2. To study the phenotype of Lynch syndrome in Asian populations (Chapter 3). 3. To describe the mutation profile of novel pathogenic Asian variants and the associated phenotype of Lynch syndrome (Chapter 4). 4. To examine the applicability of next generation sequencing technology in microsatellite instability testing (Chapter 5).

Chapter 2 describes the performance of seven predictive models (PREMM1,2,6, MMRpro, MMRpredict, Wijnen, Myriad, Amsterdam Plus, and Amsterdam Alternative), which were derived from European/Caucasian populations, on the screening of Lynch syndrome in the Asian populations with colorectal cancer. There was no evidence that these models will perform poorly in Asian families with performance similar to that reported for Caucasian families. However, many mutation carriers were not detected when these models were tested on Asian families simulated under the one-child policy (China). In addition, similar inferior performance was observed when these models were applied to families with a low penetrant gene associated with Lynch syndrome (e.g.: PMS2). These findings provide strong evidence that these models are applicable to Asian families where the size is not restricted. Their superiority over existing clinical guidelines indicates that case detection of Lynch syndrome can be maximised while minimizing false positive results when these models are applied instead of simply the Bethesda Guidelines or Amsterdam Criteria. However, alternative methods for assessing who should be tested for mutations need to be developed apart from family history in settings where small families are common. All models in this study may be deficient for the families with a PMS2 mutation, and alternative methods should be explored.

Chapter 3 examines the phenotype of Lynch syndrome in Asian populations. To date, many conflicting results are observed about phenotype of Lynch syndrome in the Asian populations amongst the published studies. While the majority of Asian studies indicated that Lynch syndrome phenotype in Asian and Caucasian populations share many similarities, other studies have disagreed. In Chapter 3, no differences in the risk of cancers related to Lynch syndrome were observed when the phenotype of Asian and Caucasian mutation carriers were compared. These findings suggest that Lynch syndrome in the Asian and Caucasian populations are indeed quite similar, and raise the possibility of a similar carcinogenesis pathway that is not affected by ethnicity. Consequently, Asian mutation carriers with Lynch syndrome should be managed according to the current cancer surveillance programs, which are based on data of Lynch syndrome in Caucasian populations, until an ethnicity-specific program becomes available. The findings of this chapter also explain why the predictive models, which were developed using European/Caucasian populations, performed well in Asian populations as demonstrated in Chapter 2. The discrepancy of Lynch syndrome phenotypes reported amongst the Asian studies can be explained by their recruitment strategies. Many of these studies derived their phenotypic data from families that only fulfilled clinical guidelines without germline testing, and many sporadic cancers may have been included in their analyses. In contrast, all the families with Lynch syndrome in the present study have been confirmed to carry a pathogenic mismatch repair (MMR) gene mutation.

Chapter 4 describes the phenotype of Lynch syndrome in Asian families with a novel pathogenic mutation. These mutations are yet to be described in the literature pertaining to Lynch syndrome in Caucasian populations. The reported phenotypic manifestation of Lynch syndrome in the Asian populations varies widely amongst the available studies, with some studies concluding that the cancer risk of Asian and Caucasian populations with Lynch syndrome is similar while other studies disagree. It remains uncertain if these novel variants confer a different cancer risk compared with Caucasian populations with Lynch syndrome. In Chapter 4, Asian families with a novel pathogenic MMR gene mutation were demonstrated to share many phenotypic manifestations as described in the Lynch syndrome literature based on Caucasian populations. This strengthens the hypothesis that Lynch syndrome mutation carriers of different ethnicities progress through similar carcinogenesis pathway despite the presence of novel mutations. These findings also strengthen the conclusions derived from the study of Lynch syndrome phenotype in the Asian populations in Chapter 3, and explain why the MMR predictive models, that were developed and validated in the Caucasian populations, performed well in Asian populations with colorectal cancer as described in Chapter 2.

Chapter 5 describes the testing of microsatellite instability (MSI) using next generation sequencing (NGS) technology. The current family history-based screening strategies are inadequate in the detection of Lynch syndrome in small Asian families and families harbouring a gene with reduced penetrance (e.g.: PMS2) as described in Chapter 2. An alternative screening strategy would be to implement universal MSI testing, regardless of age of diagnosis or family history. The current MSI testing platform is based on multiplex polymerase chain reaction (PCR) technology, which may not meet the demand if MSI testing of cancers related to Lynch syndrome becomes universal. In chapter 5, NGS technology has been demonstrated to be a feasible platform for MSI testing. The main advantage of NGS over the traditional multiplex PCR-based method is that it allows simultaneous testing of large batches of cancer samples, thus improving the efficiency of testing. In addition, compared to other NGS-based MSI testing approaches, the method described in the present study obviates the need for genome-wide alignment in data analysis. Therefore, complex data processing pipelines are not required and data analysis can be performed using standard computing resources.

In conclusion, this thesis describes the characteristics of various cancers related to Lynch syndrome and the screening of Lynch syndrome using the prediction models in the Asian populations. The data suggests that Asian and Caucasian populations with Lynch syndrome share many clinical characteristics, and the current Caucasian-based clinical guidelines and screening models are applicable to the Asian populations. NGS technology is applicable to MSI testing, and this would be a promising platform if universal colorectal cancer screening for Lynch syndrome becomes the standard of practice.