Introduction

The first Cancer Predisposition Gene (CPG) testing occurred in Manchester in 1990, with the discovery of the TP53 gene as a high-risk gene for Li Fraumeni syndrome [1], followed by BRCA1, BRCA2 and Lynch (MLH1, MSH2, MSH6 and PMS2) genes in 1991–1995. This enabled the introduction of CPG testing into the cancer genetics clinics and allowed cascade testing to unaffected family members.

The detection of high-risk germline CPG variants in index cases makes available not only cancer prevention and treatment opportunities for that person but also, through family cascade testing, subsequent cancer prevention and early detection possibilities for at-risk family members who test positive for the CpG variant. Family members who test negative can, nearly always, be released from high-risk screening programmes, relieving stress and preventing overdiagnosis and false positive screening tests.

Prior to 2003 practice in cancer genetics in the UK was largely guided by the UK family cancer Study Group (later UK Cancer Genetics Group-CGG) for breast cancer [2] with those for bowel cancer largely coming from collaborations between geneticists and colorectal surgeons [3]. These early guidelines did not set specific thresholds for genetic testing but, emphasised the importance of testing high risk individuals with cancer to open the possibility of testing for BRCA1, BRCA2 and Lynch syndrome genes [2, 3]. Indeed, the main thrust was identifying people at sufficient risk for early detection strategies.

No specific threshold was introduced for testing for HBOC until the 2004 NICE guideline [4] which set a 20% threshold for likelihood of BRCA1/2 and only recommending testing in people affected with a relevant cancer to identify index cases for testing relatives. This later fell to a 10% threshold in 2013 [4] with allowance for testing unaffected relatives where no living affected was available for testing. Since then, NICE recommended testing all triple negative breast cancers aged < 50 years in a short update. More recently (2019-) testing has been governed by algorithms from NHS England’s Genomic Testing directory for rare and inherited disease, with the latest version at time of writing available online as version 5.2 in June 2023 (Table 1) [5]. This reflects testing criteria for all English centres. A Manchester score of ≥ 15 and the CanRisk model reflect a 10% threshold although some of the criteria allow testing at much lower levels. The latter recommendations include testing all non-mucinous epithelial ovarian cancer and male breast cancer cases, with testing of all triple negative breast cancer < 60years and all other breast cancer including high grade DCIS aged < 40years (Table 1). The testing directory reflects the switch in testing of index affected cases from medical genetics departments to mainstream testing in oncology or surgery. Mainstreaming for ovarian cancer in Manchester started in November 2017 [6] and for breast cancer in April 2021 [7]. The initial panel R208 only included BRCA1, BRCA2 and PALB2 with CHEK2 and ATM added in April 2022 and RAD51C and RAD51D in August 2022. Currently missense variants are not reported except the high penetrance c.7271T > G in ATM. Traditionally CHEK2 and ATM were not considered ‘actionable’ genes in the UK and although RAD51C and RAD51D were considered actionable mainly from the ovarian risk point of view [8], this meant that cascade testing has been recommended [8]. There is still no clear guidance on cascade testing for CHEK2 and ATM despite CHEK2 c.1100delC being identified on clinical testing of BRCA1 and BRCA2 as part of Copy Number Variant testing for BRCA2 since around 2001 [9]. However, the recent inclusion of breast cancer risks estimation for both CHEK2 and ATM in the CanRisk model [10] has led to offering testing beyond the initial proband in the last year.

Table 1 NHS England mainstream testing criteria since June 2023 for breast and other cancers for R208 (gene panel of BRCA1, BRCA2, PALB2, CHEK2, ATM, RAD51C., RAD51D)
Full size table

For Lynch syndrome the testing strategy was rather different from the start as pre-screening with either micro-satellite instability (MSI) or mismatch repair immunohistochemistry (IHC) was the usual first step on tumour samples. The original Amsterdam Criteria [11] were developed initially to provide a consistent diagnosis of Lynch Syndrome (previously ‘HNPCC’,-hereditary nonpolyposis colorectal cancer)’. Once germline genetic testing became more available in the late 1990s, the criteria were used to guide testing toward families most likely to harbour a germline variant in one of four mismatch repair genes (MMR) genes. Whilst being fairly specific (about 60% of families had PVs), these criteria lacked sensitivity. Amsterdam criteria were amended in 1999 [12] with extra-colonic cancers added to the diagnostic criteria and looser Bethesda criteria [13] were used to allow inclusion of both tumour testing strategies through MSI/IHC as a pre-screen. This enabled more families to benefit from diagnostic CPG testing [12, 13]. In Manchester we have always tested the germline directly for Lynch syndrome in Amsterdam positive cases as both IHC and MSI miss sufficient numbers to leave a detection rate of 5–10% [14]. In February 2017 NICE recommended pre-screening all colorectal cancers with either MSI or IHC with additional testing for MLH1 methylation and BRAF V600E to rule out somatic involvement of MLH1.[15] This was extended to all endometrial cancers following identification of a similar 3% germline mutation rate for MMR genes in the UK to colorectal cancer [16, 17].

For both Lynch and BRCA1/2 somatic testing of tissue from deceased relatives is a useful mechanism for identification of the family pathogenic variant and is more informative than testing an unaffected at-risk individual [18]. This mechanism is now approved by NHS England [5].

The Manchester Centre for Genomic Medicine (MCGM) serves a population of 5 million in the North-West of England. Outcomes of clinical CPG testing in the clinic, both diagnostic and predictive, have been recorded by the MCGM through the familial registers since diagnostic CPG testing first commenced in 1990 [19, 20]. This record provides a clinical tool to ensure appropriate screening is in place, at risk family members are offered pre-symptomatic CpG testing, as appropriate, and provides a resource for identifying patients when new management approaches or research opportunities become available.

We sought to evaluate, for BRCA1/2 and MMR genes, time trends of variant detection in index cases, and numbers of cascade tests undertaken in relation to guidance changes.

Methods

Genetic family registers were interrogated for numbers of diagnostic index tests and subsequent family cascade tests (positive and negative), and year of testing for BRCA1, BRCA2, PALB2, MLH1, MSH2, MSH6, and PMS2. Only class 4 (likely pathogenic) and class 5 (pathogenic) CPG variants are presented and are referred to, collectively, as ‘variants’ [21, 22]. Testing of moderate risk genes such as CHEK2 and ATM was never comprehensive until April 2022 and was excluded apart from a small sub-study of CHEK2 variants identified through the family history clinic at Wythenshawe hospital. Data was censored at 6th May 2023. Of note there was much reduced capacity for both laboratory testing and clinic appointments for much of 2020 in view of the SARS-CoV2 pandemic. Uptake of risk reducing surgery (prior to a cancer diagnosis) including Bilateral Risk reducing mastectomy (BRRM) and Risk reducing bilateral salpingo-oophorectomy (RRBSO) was also assessed in those who underwent pre-symptomatic testing (had not had a relevant cancer prior to testing). As all first- degree relatives (FDRs) of a PV heterozygote are recorded on our databases [19] we also assessed overall uptake amongst unaffected FDRs of pre-symptomatic tests. Confirmatory tests in those with a relevant cancer were also recorded.

Timeline of testing and strategies

The introduction of diagnostic testing for the relevant CPGs in the MCGM began in the early 1995 mirroring the early days of CPG identification when testing was initially undertaken on a research basis prior to formal incorporation into the clinical diagnostic setting in 1998. Thus, testing began for BRCA1 in 1995 and BRCA2, MLH1 and MSH2 in 1996.

Initial variant detection strategies were by SSCP, dHPLC and Southern Blot analyses, prior to the introduction of Sanger sequencing in 2003, MLPA for copy number detection in 2005/2006 and next-generation sequencing (NGS) in 2014 (2013 for BRCA1/2). A genetic register for BRCA1/2 and Lynch syndrome was introduced in Manchester in 1996 allowing outreach to relatives to encourage cascade testing. Although only previously referred relatives are directly contacted regarding the availability of testing, index cases are provided with information letters to pass the eligible blood relatives.

Results

BRCA1/2 testing

A total of 1600 index BRCA1/2 variant carriers have been identified (BRCA1 = 798; BRCA2 = 802; Table 2). Breast cancer cases predominated but after the 10% NICE threshold meant all non-mucinous epithelial ovarian cancers aged < 60 were eligible in May 2013 the proportion of those tested with ovarian cancer only doubled from 16.5 to 32%. The proportion of pre-symptomatic cases rose from 3.2% before the 2013 NICE guidelines to over 8%. There was also a slight increase in proportion of male index cases rising from 3.5 to > 8% in the last era. Since 2016 30 pathogenic variants have been found from somatic testing of deceased index cases, all but three in ovarian cancer cases.

Table 2 Number of positive index cases and additional family tests generate for BRCA1/2 combined
Full size table

Subsequent non-index cascade tests

We made the main era definition around the NICE guidance of 2013 as this reduced the testing threshold to 10%. In order to allow time for subsequent family testing we separated off those with index positive results after April 2021. The mean number of tests generated was 3.39 for era 1 (< May 2013), 2.09 for era 2 (May 2013–March 2021) and 0.57 for the current era 3. However, when confining tests to within 2 years of family variant report there was a very similar mean of 1.74 for era 1 and 1.73 for era 2 with much of era 3 still not reaching the 2-year follow up. Importantly for era 1 there were 511 additional family tests more than 9 years after index case identification. For number of tests, era 1 had 26% of families generating 5 or more additional tests compared to only 13% in era 2 perhaps reflecting the larger family size of some of the initial families requiring a higher testing threshold.

There were 3568 pre-symptomatic tests and 581 confirmatory tests on people with relevant cancers. Confirmatory tests were far more likely to be carried out within 1 year of the family report 67% versus 32% of all tests in each group (Table 3; P < 0.0001). Male versus female ratios were virtually identical. Around one positive additional family test was generated with around 50% of family tests also being positive (Table 4). A total of 556 RRBSO and 453 BRRM were generated in positive female relatives. Overall, of 6170 FDRs eligible for pre-symptomatic testing 2999 (48.6%) underwent testing. This included 940/2899 (32.4%) males and 2059/3171(64.9%) female FDRs (p < 0.0001).

Table 3 Types of additional tests carried out in families with a positive BRCA1/2 index case
Full size table
Table 4 Number of positive family tests and subsequent and uptake of risk reducing surgery in non-index cases
Full size table

Lynch syndrome

A total of 625 index MMR variant carriers have been identified (MLH11 = 185; MSH2 = 238; MSH6 = 132; PMS2 = 70; Table 5). Colorectal cancer cases predominated but from 2013 a higher proportion of endometrial index cases were identified with close to ¼ since April 2021 when the NICE endometrial guideline for Lynch testing was instituted. As expected with Lynch affecting men almost the same as women nearly 50% of index cases were male. Since 2016, 16 pathogenic variants have been found from somatic testing of deceased index cases mostly in colorectal cancer. In all these 625 index cases have led to the identification of 857 family members positive for the MMR PGV.

Table 5 Number of positive index cases and additional family tests generate for Lynch MMR genes combined
Full size table

Subsequent non-index cascade tests

To compare with HBOC we used the same 2013 cut off. In order to allow time for subsequent family testing we separated off those with index positive results after April 2021. The mean number of tests generated was 3.44 for era 1 (< May 2013), 2.55 for era 2 (May 2013–March 2021) and 0.81 for the current era 3. This was essentially the same as for HBOC in era 1 but with more tests per index being generated after for Lynch. When confining tests to within 2 years of family variant report there was a very similar mean of 1.63 compared to HBOC was seen for era 1 but this increased to 2.13 compared to 1.73 for HBOC in era 2 with much of era 3 still not reaching the 2-year follow up. Importantly for era 1 there were 216 additional family tests more than 9 years after index case identification. For number of tests, era 1 had 25.5% of families generating 5 or more additional tests virtually identical to HBOC compared to only 17.2 a little higher than the 13% in era 2 for HBOC perhaps again reflecting the larger family size of some of the initial families requiring a higher testing threshold.

There were 1514 pre-symptomatic tests and 249 confirmatory tests on people with relevant cancers. Confirmatory tests were again far more likely to be carried out than pre-symptomatic within 1 year of the family report 147/249 (59% Vs 602/1514 (39.8%) (P < 0.0001), though the difference was less marked than for HBOC. The proportion of male pre-symptomatic tests was also much higher than for BRCA1/2 at with 639/1514 (42%) male for Lynch compared to 1058/4158 (25.8%) for BRCA1/2 (P < 0.0001). For FDRs uptake of pre-symptomatic tests was 63.4% in males (568/896) compared to 75% (766/1022) p < 0.0001. The uptake in male FDRs was much higher than in BRCA1/2 (p < 0.0001), but the 75% uptake in females was still significantly higher than the 64.9% for BRCA1/2 (p < 0.0001),

CHEK2 as a moderate risk gene

Although CHEK2 testing outside the index case has not been systematically offered some testing through the Manchester family history clinic has occurred through the FHrisk study. In total 37 index cases have been identified, but only 36 additional family members were tested clinically until very recent testing of stored blood DNA samples from a further 13 women affected with breast cancer. In total of affected non-index cases only 50% 11/22 tested positive for the CHEK2 variant. This compares with 252/290 (87%) for BRCA1 (p < 0.0001) and 328/390 (84%) for BRCA2 (p = 0.0004).

Proportion of index cases found on mainstream testing

Mainstream testing in oncology and surgery started in November 2017 with the focus on driving treatment decisions for the index case rather than cascading. The proportion of index cases found by cascading has increased for BRCA1/2 from 2% in 2017 to 53% in 2019 and 84% so far in 2023 (supplementary Table 1).

Discussion

The provision of CPG testing has evolved enormously revolutionising the practice of clinical cancer genetics over the past 30 years. Where an index case CPG variant is found this enables (i) cancer prevention and early detection strategies for family members at-risk as well as for second cancers in the index case (ii) management strategies in the index case, including surgical management and targeted therapeutic strategies [23]. The focus has very much switched from emphasis on cascade testing for relatives to management of the index case since 2017. This is reflected in the high proportion of index cases found in 2023 from mainstream testing. Although, cascading still occurs from referral of the index case to genetics, this may be slower or may not occur due to the index case being too unwell or refusing referral reducing the rate of family testing [6]. We still do not have enough data to determine whether long term uptake of testing following mainstream testing is lower than the traditional genetics referral model. In era 1 the main motivation for testing was to inform family members of their risk (often generated by unaffected people attending concerned about their risks). Targeted therapies including PARP inhibitors for BRCA1/2 deficient ovarian and more recently breast cancers and PD-1 inhibitors for MMR deficient tumours, were not available until towards the end of era 2 or even era 3 [24, 25]. Changes due to mainstreaming has also changed the cancers from which index cases have suffered, with ovarian cancer becoming more frequent in BRCA1/2 and endometrial cancer in Lynch syndrome.

Decreases in family size from reducing the thresholds for testing from 20% through 10% to effectively only 3% for women with an isolated non-mucinous ovarian cancer over age 70 [26], has meant fewer at-risk relatives per family available for cascading. This is reflected in the larger proportion of families with high numbers of generated cascade pre-symptomatic testing in the earlier era (< 2013) with 14.1% of families generating ≥ 7 tests compared to only 4.8% of families in era 2. Although there were obviously fewer years to generate these tests there were still a higher mean number per family tested to the 9-year point in era 1 compared to era 2 2013–0/4/2021). Nonetheless there were still a similar number of tests generated per family in the first two years at 1.74 and 1.73 respectively. A similar number of pre-symptomatic tests were generated per family for the MMR genes both at 2 years and overall.

The situation in MCGM prior to therapeutic options being available is reflective of the whole UK and is similar internationally. Therefore, consensus strategies were put in place to identify those individuals and families in whom there was the greatest likelihood of detecting a causative germline variant and who would most likely benefit from genetic investigations. This is reflected in NICE guidelines for breast cancer [4] and the use of the original Amsterdam Criteria [11] to drive testing for MMR genes.

CPG variants of the BRCA1, BRCA2 and MMR CPGs are by far the most common of all CPG variants identified in testing with combined population frequencies of 0.3–0.5% for BRCA1/2 and Lynch [27, 28]. The lower number of identified CPG index cases identified for other genes reflects the rarity of these variants for example STK11, CDH1 and PTEN, but also the reduced actionability of the moderate risk genes. Until recently cascading was not offered for either ATM or CHEK2 in the UK. Indeed, although their combined frequency exceeds BRCA1/2 [28] they were only included officially on NHS England breast cancer gene panels in 2021 [5]. The main issue was how to counsel family members on their risk if they test positive or negative. This is reflected in the low proportion of confirmatory tests that were positive (50%) for CHEK2 compared to over 80% for BRCA1, BRCA2 and PALB2 [9]. The availability of breast cancer risk estimates using the CanRisk programme [10] is likely to increase uptake of cascading as cancer genetics professionals now feel more confident providing this service.

As expected, we found higher uptake in males for Lynch syndrome than for BRCA1/2 [27] with a higher proportion of pre-symptomatic tests in males at 42% of all tests in males for Lynch compared to 25.8% for BRCA1/2 (P < 0.0001). Overall uptake was also higher in males FDRs. This higher uptake likely reflects the increased utility of testing in males for personal cancer risks particularly for colonoscopy screening. This is consistent with our previous publications on uptake in both conditions [29, 30]. There were also a number of tests that occurred more than 9 years post index test. This can reflect ineligibility due to being < 18 years at date of index case but also a slower initial uptake between 18 and 25 years of age [29].

Here we have presented the cascading of CPG testing for 33 years at MCGM in the UK. With the rapid increase in mainstreaming of diagnostic testing where immediate treatment decisions are required, pathways need to ensure at-risk family members are offered testing. It is notable that the detection of 2225 index cases for BRCA1/2 and Lynch syndrome combined at MCGM has led to the subsequent identification of over 2989 positive family members, who can then benefit from cancer prevention and early detection strategies resulting in both individual, and societal health economic benefit, Equally important is the reassurance for close to 3000 family members testing negative as well as all their offspring who, for the most part can be reassured regarding risk.