Lynch syndrome (LS) is a hereditary cancer syndrome underlying ~5% of all endometrial cancer diagnoses. It is caused by a dominant germline mutation in one of the mismatch repair (MMR) genes, namely MLH1, MSH2, MSH6 or PMS2. Relative to the general population, MMR mutation-positive women have a substantial increased lifetime risk of endometrial cancer (30-70%), and also increased risk of colorectal (48%), kidney, renal pelvis or ureter (11%); urinary bladder (9%); and breast (11%) cancer after endometrial cancer diagnosis. Cancer prevention strategies are proven to reduce morbidity and mortality of mutation carriers. It is thus important to facilitate identification and timely referral of endometrial cancer patients who are possible mutation carriers, and subsequently their at-risk relatives. However, studies have shown that LS patients are often under-diagnosed in routine clinical practice.
Clinical criteria incorporating detailed family history and/or molecular tumour testing such as the Amsterdam II, revised Bethesda and the Society of Gynaecologic Oncology (SGO) have been developed to aid identification of individuals at risk for LS. Though these criteria are commonly used to triage patients suspected for having LS, they have been criticised for being too complicated to apply in routine clinical practice, and for poor sensitivity and/or specificity. Consequently, molecular tumour screening tests such as immunohistochemistry (IHC) and microsatellite instability (MSI) have been performed to provide additional means for identifying patients at risk for LS. However, to date, there are no standard criteria for referral of endometrial cancer patients at risk of LS. There has been no previous study comparing the performance of a variety of approaches to inform population-level triage of endometrial cancer patients for germline MMR gene mutation testing. Further, barriers to referral for genetic services have been reported, but it is unknown whether clinicians in Australia encounter the same challenges as their peers abroad in providing genetic services for patients in daily patient care.
In view of these problems, the studies conducted within this thesis aimed to improve identification and referral of endometrial cancer patients at increased risk of LS for genetic services. We have conducted the study in three parts: Part 1 – a retrospective chart review to evaluate referral of endometrial cancer patients with possible LS, based on family history information, to genetic services; Part 2 – mixed-method study: (1) a systematic literature review regarding genetic referral of patients with suspected hereditary cancer syndromes, focussing particularly on clinician-related factors such as documentation of family history of cancer, knowledge of hereditary cancer syndromes and awareness of genetic services; (2) semi-structured interviews to explore clinicians views on barriers and motivators of genetic referral; and (3) a web-based survey to describe prevalent practices and attitudes regarding referral and testing for LS in Australia; and Part 3 – study of a large population-based case-control endometrial cancer cohort to evaluate and compare the performance of clinical criteria, including BMI, age, IHC and DNA MLH1 methylation testing for screening of possible mismatch mutation carriers in the clinical practice. A simplified cost-effective analysis was also performed for selected screening strategies.
Findings from our systematic review demonstrated clinicians lack knowledge of LS, and poorly discriminated between low and high-risk patients. Family history documentation was inadequate and referral of patients was more likely when clinicians were aware of genetic services. Correspondingly, our research results revealed poor documentation of family history and under-referral of endometrial cancer patients with possible LS for genetic services. Numerous clinician-related barriers referral included: clinicians’ unfamiliarity with LS and uncertainty of who or when to refer; patient disinterest; lack of family history; lack of awareness of genetic services; lack of access and referral guidelines for services. General practitioners were less likely than gynaecologists, oncologists, gastroenterologists or surgeons to recognize the risk of colorectal and endometrial cancer. Male general practitioners who had practiced for ≥10 years were less likely to make genetic referral than other provider groups, and many specialists did not clearly recognize that IHC testing is not a germline test. Our results also show that stepwise testing for tumour MMR IHC loss in patients aged ≤60 years, tumour MLH1 methylation in individuals with MLH1 IHC loss, and germline mutations in patients exhibiting loss of MSH6, MSH2, PMS2 or loss of MLH1/PMS2 with absence of MLH1 methylation is an optimal and cost-effective strategy to identify MMR mutation-positive endometrial cancer patients at the population level.
Using various approaches in this study, we have demonstrated that identification and referral of endometrial cancer patients who are at elevated risk of LS should be improved. Implementation of stepwise IHC and MLH1 methylation testing in clinical setting could further improve triage of endometrial cancer patients who are at increased risk of LS. A research-based family history questionnaire can be used in specialist clinics to improve family history documentation, and a targeted educational intervention for general practitioners and specialists may increase knowledge of LS.