significant prediction in rise in cancer cases over the next 20 years and that as clinicians and professionals working in this field is a big challenge for all of us and as organizations responsible for delivery of cancer care for women in the future, you know, we can't ignore this. Oops, let's see how I can move this forward. The traditional, so why are we doing testing for cancer diagnosis? Because the traditional approach which was used was used to use family history as an approach to identify people at increased risk, identify people with gene mutations as multiple limitations and it misses about half the people at risk and we know that, for example, for ovary cancer, the BRCA genes are associated with an increased risk which is 17 to 44 percent and they're also associated with an increased risk of breast cancer. Despite the screening or the testing probability threshold having fallen, you know, there is a limitation by using a family history approach and if we do unselected gene testing, you hit the lower probability threshold. The other thing to bear in mind, I think, particularly as we're looking at it from outside the Western population context is that a lot of these models or rules or criteria using family history which correlates the probability have been designed from largely Western datasets and when you try and use them in non-Western populations which might be, for example, non-Caucasian populations or the African population or the age from the Asian subcontinent, they don't perform as well. So, you have higher probability of carrying gene mutations than if you use traditional criteria and traditionally, if you look at complete unselected gene testing, whether you look at cancer case series or you look at biobanking data, unless you do unselected testing, you're going to miss anything between 50 percent to 80 percent of the carriers. So, there's been a move clearly towards unselected gene testing which will lead to better identification of high-risk groups, better prevention and risk management in those who are unaffected, largely through cascade testing, particularly for ovary cancer, but better prognostication opportunities for novel treatments. I'm going to talk about that and we are all aware of PARP inhibitors and Asima does a huge amount of work in that area and, of course, access to clinical trials which can lead to improved survival. So, who should we test for a cancer diagnosis? I think ovary, endometrium and bowel are now recommended in all clinical guidelines. Whether we should do it for all women with breast cancer is slightly gray area and there are proponents and people who don't agree with that. For example, it's not part of your current NCCN guidelines. My personal view is that I think it should be done. I think this will come. In the UK, we have just completed a trial around this and we are about to start an implementation pilot program across some of the London NHS hospitals this year and I suspect this will then move towards solid tumors. So, looking at the model of implementing gene testing at cancer diagnosis is going to be critically important for the future. From the benefit for women who have ovary cancer, these are the more recent data which clearly show an improvement in overall survival with the hazard ratio published last year of 0.55 using PARP inhibitors in first-line maintenance treatment. So, there's a clear benefit. There's also benefit in the recurrence setting demonstrated for PARP inhibitors which again is linked to testing the tumor tissue or the germ line to establish who may be eligible for treatment. When we are doing gene testing and looking at screening and prevention and particularly in unaffected individuals and some of these things will apply for affected individuals, the options are there for screening and prevention depending on the cancer type here. Particularly for BRCA, we're looking at breast and ovarian cancer and there's the options of surgical prevention for breast and ovarian cancer, medical prevention for breast cancer, potentially for ovarian cancer. Screening for breast cancer in particular, screening for bowel cancer if you're looking at women with Lynch syndrome and lifestyle reproductive choices, pre-implantation genetic diagnosis, pre-implantation genetic diagnosis. Again, surgical prevention if you go to the top of the slide for Lynch syndrome women with hysterectomy and BSO. So, these are the benefits that you can get through cascade testing for the unaffected population and of course these need to be weighed against some of the cons of gene testing which relate to family dynamics, potential emotional impact, anxiety, distress, issues around confidentiality, insurance, manageability, stigmatization, cost effectiveness of these whole strategies and of course cost for the patient depending on the health system you're working in. A little bit around terminology. When we say germ line variant, we're really talking about a variant in all the cells of the body. When we use tumor variant, it means that there's a variant present in the tumor tissue that may or may not be in the germ line. If you have a variant of the tumor tissue that's absent in the germ line, we tend to call it a somatic variant. So, when I'll be using these terminologies, this is what I mean by all these different types of variants. The other thing to mention is that variants are of different classes and the ones we are really interested in, in particular, are class four and class five. Those are disease-causing or we call pathogenic or likely pathogenic variants. There is another class of variant called a VUS or a variant of uncertain significance. That's a class three variant where we don't believe that the variant is currently classified as disease-causing, but a small proportion of these may in the future get reclassified as either class four or class five. So, they need to be monitored. The large majority, of course, will remain as VUSs or get downgraded. All international guidelines today recommend gene testing at cancer diagnosis for ovarian cancer and RBGCS guidelines are no different. So, which genes should we test for? For ovarian cancer, in particular, we now move towards panel testing. So, we would test for BRCA1, BRCA2, RAD51c, 51d, BRP1, PALB2, the Lynch syndrome genes. These are now listed in our NHS test directory. I think, I'm not sure if all the guidelines have moved towards this, but definitely some, I suspect some of the guidelines still say BRCA1 and BRCA2, but we know now that the large number of genes associated with ovarian cancer, which have implications both for treatment and also for cascade testing and screening and prevention, and really those should be included in a panel. From the tumor perspective, we do at least BRCA1 and BRCA2, but also there's the opportunity to do HRD testing, which could impact on influence patient management. Some of these genes, of course, are associated with breast cancer, you know, BRCA1, 2, and PALB2, in particular, high-risk breast cancer genes, but 51c, 51d, and CHECK2ATM are more moderate risk genes. Things to think about when implementing testing at cancer diagnosis include the logistics. I don't think there can be one solution or one model that fits all. I think even in different health systems, you'll have different models in different regions or even different hospitals. I don't think there's a single model, for example, in all the London hospitals. It's important that we develop pathways and models which are context-specific and are going to work for teams or cancer centers or those delivering the care and these pathways, and also for the population. And we need to look at logistics, turnaround time, psychosocial factors, considering, you know, when you're talking to the patient, also you need to consider impact on the family. I've spoken about some of the limitations of testing, so patient factors and issues around pre-test counseling and how to deliver pre-test counseling. Ideally, for ovary cancer, I think we should be testing as near possible as cancer diagnosis. And so this provides various opportunities, whether it's at the time of surgery or at the time of the diagnostic biopsy or if patients are in the actual chemotherapy and at the time of starting chemotherapy or for that matter, subsequently after the patient comes to surgery. It depends on when you see the patient, our tendency is to try and do this as soon as we can near the time of diagnosis, but at least sensitize the patient to that. I'm just going to highlight some data from a multi-center cohort study, which we've collected and analyzed. And this is from about 2400 ovary cancer patients. And you find that the mutation prevalence is about 17% for BRCA1 and 2. And if you look at the other genes, this doesn't have the lynch genes in it, but if you look at the other genes, it contributes another two and a half percent, so two to 3% prevalence. Somatic prevalence is 7% and about 4% to 5% vast prevalence rate. And 50% of people do not give you a family history or just significant for breast or ovary cancer. The other thing to flag up is when you're doing tumor testing is that you get large genomic rearrangements. And if you just did tumor testing, you'd miss this. So you need to do MLP and germline to be able to pick these up. In our data set, it's about 11% of the population. So it has implications for what strategy you used to do gene testing. And if you did only somatic When we use about 23 percent in our dataset at biopsy and about three to five percent at surgery, whether it's primary surgery or interval site-reductive surgery. We know that HRD testing is also important because 50 percent of all high-grade ovarian cancers, especially serious ovarian cancers, are likely to be HRD deficient, and this can be picked up by either the two assays that are widely used at the moment on myriad and the foundation medicine assays, but lots of colleagues, including Asimov, are working at novel ways and new assays, which will hopefully make a good contribution to the field. So our philosophy is to do parallel testing. I think if you do germline only, you miss somatic mutations. If you do tumor only, you miss large genomic rearrangements, and parallel testing is ideal. I'm just going to skip through a bit. It's just a cost-effectiveness analysis model we've run on panel testing, and I'll come to the last slide, which is the crux and also the interest of time. We show that unselected panel gene testing is cost-effective compared to just bracket testing alone, and really we should move towards panel testing, but if you include PARP inhibitor from a PARP inhibitor treatment perspective, unless your list price or cost is discounted by about 30%, 35%, at the current listed cost price, the PARP inhibitors are not necessarily cost-effective strategy for both the UK and US health systems. Coming down to endometrial cancer, the endometrial cancer, the four Lynch syndrome genes, MLH1, MSH2, and MSH6. This is Dr. Warden, and it was initially discovered in women with endometrial cancer, but then lost for many years and then picked up by Henry Lynch who popularized Lynch syndrome. So the ovary cancer risk can vary from 11% to 17%. PMS2 is not associated with a hugely increased risk of ovary cancer, so please bear that in mind. The bowel and endometrial cancer risks are fairly similar, between 13% to 50%. So we've spoken about limitations of using family history criteria. If you use just the Amsterdam 2 family history based criteria for Lynch syndrome, you only pick up about 30% of patients. So NICE and lots of other guidelines recommend unselected gene testing at endometrial cancer diagnosis. So this is built into our care pathway now, and anybody who has endometrial cancer does get immunohistochemistry, which reflects immunohistochemistry. And depending on the immunohistochemistry result, you may or may not go for gene testing. So it's 3% of endometrial cancer patients have Lynch syndrome, 4% of bowel cancer patients have Lynch syndrome. And what you're really looking for is the stain deficiencies. You can see these two slides, MLH1 and PMS2 are negative stained, while MSH2 and MSH6 are positively stained. So this means that maybe the protein's not being produced and the gene's not working here. So if you have a loss of expression of MSH2 or MSH6 or isolated PMS2, then you go for germline testing. However, like I showed in the slides, if it's MLH1 and PMS2 that's lost, then you need to make sure there's not a false positive due to MLH1 promoter hypermethylation. And if the promoter hypermethylation is present, then you don't have gene testing. But if it's absent, the patient should be offered testing for Lynch syndrome genes. So the advantages clearly may influence treatment from the point of view of use of PD-L1 inhibitors. So it's got implications for prognostication and treatment, especially at recurrence for the cancers, but also cascade testing for family members and secondary cancer prevention for colorectal cancer prevention for the patient. For unaffected women, we do tend to do endometrial cancer screening. The data around surveillance for this is limited, but we have picked up both pre-invasive and invasive disease. There's the option of surgical prevention we spoke about earlier, targeted treatments and PGD. Aspirin is recommended for reducing the risk of bowel cancer. The CAP trials have shown a clear benefit. What the actual dose should be is not yet quite decided, and there's an ongoing trial on that. Our guidelines at the moment in the BSGE suggest 150 milligrams per day with a higher dose in obese women. The other benefit, of course, and especially if you include P53 and poly, so molecular classification, which we know is going to influence treatment and clinical trial improvement going forward. So at our center, we've implemented this over the last six, seven months in a systematic way, and we have a separate MDT actually set up for endometrial cancer. It is just the way that it was feasible to implement in our setup. This is the pathologist, the molecular lab person, the cancer clinician fellow. We have CNSs and relevant members from our high-risk clinic and the gynecology team. I can today tell you what every single patient who goes through my MDT or our GMO board across our Northeast London Cancer Network, and I can tell you where that patient is in the pathway because we track everybody, whether they've had immunohistochemistry, whether they've had hypomethylation, whether the samples in the lab or the patient's got the result or she's gone for gene testing or not and what the result is. So we also have risk management special interest clinic, which we run once a week, which is a multidisciplinary clinic, includes nurses and gynecologists and menopause specialists and psychologists, and it's linked into genetics colleagues, and we have a monthly MDT around this. We are building some endometrial cancer teaching modules for training mainstreaming at endometrial cancer diagnosis, and I hope we should be, I'm optimistic we'll be able to make this available free for everybody, both nationally and I hope internationally going forward. This will be on our NHS training system called eLearning for Health, and we're in the process of just finalizing these modules and I hope in the next few months they'll be up. So this is just to show the different types of methods of doing, so think about the pathways of implementing mainstreaming at cancer diagnosis, moving away from the traditional methodology of counselling, there is good data showing you can do this on the telephone today, you can do telemedicine, video counselling, you can do group or DVD-based approaches, and you can use a web-based information system and helpline. So there's published literature around all these models, and I'm sure Donna will talk about and you'll hear about the Nugena program, which is the nurse-led work that she's been delivering with Asima and their team and doing a huge amount of work in low-resource settings. One of the things we are trying to set up is also a direct patient approach using a digital end-to-end process to build web apps to deliver information and deliver the counselling testing and moving away from blood- to saliva-based DNA testing, and this is just an RCT we are going to start maybe by the middle of this year in the UK, where we will randomise to a direct-to-patient versus traditional mainstreaming approach for gene testing, and hopefully in the next three years we can produce results. So just to summarise, yeah, I think we need to move towards mainstream implementation of mainstream gene testing at both ovary and endometrial cancer diagnosis for ovary, we need to consider both tumour testing and germline testing, you should do it as soon as you can in the treatment pathway, you should need to think about cascade testing and secondary cancer prevention, we need to build context-specific pathways and models to implement it, I don't think there's a one-size-fits-all solution here, and unselected gene testing and endometrial cancer diagnosis also being implemented today needs to be implemented, it's also got potential benefits from molecular classification, I think this will happen for more and more solid tumours going forward, and we will move into breast and pancreas and prostate. Thanks, so just a whistle-stop tour from my side.

cancer cases

gene testing

context-specific pathways

high-risk groups

clinical trials

molecular classification