Paul and Brett's Alpha
A disquisition upon diagnostics
The diagnostics sub-sector has long been a key overweight for the Trust, averaging a double-digit percentage of our gross investments over the 6½ years of the Trust’s existence. Over that time, exposure peaked at >16% in late 2019 and troughed a low ~6% in mid-2021, the latter being driven mainly by a number of our holdings reaching dizzying valuations resulting in our exit or being acquired in the COVID-driven frenzy of excitement around this sub-sector.
Even at the hubristic heights of mid-2021, this group represented only 2.6% of the MSCI World Healthcare Index and that figure stands at 1.5% as of the end of June 2023, versus our weighting of 10.6% of the portfolio as of the end of June 2023. This serves to illustrate the extent of our overweight position in this area.
Looking back over the past twelve months, diagnostics has been a challenging area for investment, falling into the bottom quartile of the portfolio’s sub-sector NAV contribution for most H2 2022. Our enthusiasm persisted though and Diagnostics has been either mid-pack or right at the top of the portfolio’s sub-sector NAV contribution during H1 2023. As of 30 June, the largest position in the portfolio remains a Diagnostics company.
Why do we feel so compelled to own these companies? The answer lies in their pivotal role at the beginning of the patient journey. In the words of Hippocrates: “as to diseases, make a habit of two things — to help, or at least, to do no harm. Declare the past, diagnose the present, foretell the future”. There is no correct solution that can be foreseen from an incorrect diagnosis.
With so much of the disease burden in modern life coming from chronic conditions or cancer, early detection or even routine screening and prompt interventions are critical to successful disease management. Let us illustrate this compelling value proposition through the core business of our top holding, Exact Sciences.
“Turning muck into gold”
In the United States, colorectal cancer (CRC) is the third leading cause of cancer-related death in both men and women, and moves up to the second most common cause of cancer death when numbers for both sexes are combined. The lifetime risk for CRC is about 4% (i.e. 1 in 23 of us will succumb to it). In the US, there will be around 150,000 new diagnoses of the disease this year and it will kill around 50,000 Americans during that period.
The incidence used to be very low in the under 50 age group (<10%), but diagnosis is becoming more common at earlier ages and now around 1/3 of cases are in the under 55 age group. The risk is significantly higher in the 45-49 age group than in the 40-45 age group. The peak age for diagnosis is 80-85, but around 20,000 of those 150,000 cases will be in the <50 cohort.
The relative survival rate for CRC is highly dependent on the stage of detection. At stage 0-1, 5-year survival is >90%. At stage 2, it is 87%. At stage 3, it is 72%. At stage 4, it is fatal and survival is usually less than two years, even with chemotherapy.
At stage 0, the tumours might be pre-cancerous growths (“polyps”) or a very localised carcinoma-in-situ. These growths could be smaller than 5mm in diameter (so the size of a pea). There is another earlier disease known as an adenoma, which is a growth that may or may not progress to becoming a cancerous lesion.
With these background statistics, it is recommended that people aged 45-75 in the US are routinely screened for CRC and there are various options for what approach is used (described below). In the UK, all registered NHS patients aged 60-74 in the UK have been screened every other year with a FIT test (also described below) that is sent out in the post. In 2021, the UK screening age limit began to be lowered, with the aim of reaching 50 by 2025. Currently, 56 and 58 year olds have been included in the expanded programme. Roughly 70% of these kits are returned for analysis.
There are various types of invasive and non-invasive tests for routine screening and it is surely self-evident that their power to accurately detect early stage disease (sensitivity) is the most ideal characteristic. However, sensitivity to pick up early stage disease must be balanced with specificity, or simply put a low risk of a false positive result. Anyone receiving a call back for further testing after an initial screen is going to be extremely anxious, which is an undesirable health outcome in anyone.
The main approaches to CRC screening are summarised below:
- Faecal immunochemical test (FIT): this test detects blood in a stool sample. Blood in the stool can be a sign of colorectal cancer, but it is also a sign of other GI conditions. As a consequence, the test can have a high false positive rate if it is used at maximum sensitivity (around 25% of positive tests are false positive). The benefits of this approach are its cost (its very cheap) and convenience (it requires nothing more than a small sample of faeces and so can be done by the patient at home).
There is a similar, older test called the Guaiac-based fecal occult blood test (gFOBT), which uses the chemical guaiac to detect blood in the stool. This has a slightly lower sensitivity but higher specificity (so lower false positive result). Both tests are similarly sensitive to small adenomas (~7%). The cost of running a FIT test is ~$25.
- Colonoscopy: this invasive procedure has long been considered to be the gold standard for the detection of colon cancer, especially with respect to the detection of smaller adenomas. The other advantage of a colonoscopy is that any pre-cancerous lesions can be removed as part of the procedure.
However, the process requires anaesthesia and “prep” which cannot be skipped and essentially involves a run in period where you avoid most fibrous foods for several days, and 24 hours of a clear liquid diet including chemicals to help evacuate the bowels with the attendant need to visit the bathroom frequently (and unpleasantly).
The cost of a colonoscopy procedure is around $3,000 and, because it is done in a theatre setting under general anaesthesia. There is finite capacity in any healthcare setting to undertake this procedure, limiting the frequency of screening (usually to 5-10 years depending on background risk factors).
There is a quicker version known as a sigmoidoscopy where a different probe is used to examine the lower third of the bowel and rectum. It does not obviate the need for “prep”, but an enema could be offered instead. However, were any polyps found, you would likely be recommended a colonoscopy in any event to inspect the upper region of the bowel. Local rather than general anaesthesia may also be used, so this is not necessarily a less discomforting experience overall.
- CT Colonography (virtual Colonoscopy or CTC): Computed tomography colonography, also called a virtual colonoscopy, is a less-invasive procedure that creates images of the entire colon that can be analysed by a physician/radiographer. The bowel is inflated with air to aid resolution, so there can be some discomfort during the procedure and the “prep” regimen must still be undertaken. CTC has slightly lower sensitivity across all tumour stages, particularly with respect to very small features like adenomas and carcinoma-in-situ that characterise early stage disease. Like sigmoidoscopy, it does not allow the removal of polyps detected during the examination.
However, it is a much quicker procedure, requires no sedation and is also less costly than a colonoscopy at around $1,500. Because the whole abdominal region is being scanned, it can also detect other conditions outside of the bowel. Re-screening is recommended every 5-10 years.
- FIT-DNA test (aka stool DNA test: Exact Sciences Cologuard): this is a next generation genetic screening test that combines a FIT test with the detection of altered DNA in the stool. For this test, you collect a full stool sample and send it to a lab in pre-supplied box, where it is analysed for altered DNA and for the presence of blood. Patients who have detected anomalies can then undergo a colonoscopy.
Cologuard can be done at home and no “prep” is required. Cologuard costs ~$500 and can be repeated every 1-3 years, depending on background risk. Sensitivity for Stage 1-3 cancer is comparable to colonoscopy, but lower for small adenomas (although better than any of the alternatives). Exact Sciences has reported a test return rate of ~65% over the past few years, which is comparable to that seen by the NHS in the UK with its FIT test programme. Exact has recently shown the clinical results of an improved test (“Cologuard 2.0” that has improved specificity (i.e. lower false positive rate) and higher sensitivity across all stages, but not appreciably different when it comes to adenomas. CG 2.0 has yet to receive regulatory approval but we expect this to occur the near future (patients will not notice any change from the original test as the procedures are unchanged).
The current, approved version of Cologuard requires the patient to collect a complete bowel movement and, as a consequence, the kit supplied to patients is bulky and this has been identified as one of the reasons people do not return the tests. Exact is working on an improved form factor (“CG 3.0”) with a lower sample volume requirement that can thus be supplied in a more discreet package. We think this could enhance return volumes.
“The future, now?”
In the US, around 17 million patients are screened annually by one of the methods described above (note – only one of these approaches advocate annual testing for the majority of patients. This annual figure breaks down to around six million screening colonoscopies, eight million FIT tests and around three million Cologuard tests). There are 33.7m people in the 50-75 year age cohort. If routine screening does move down toward 45, that cohort will increase by a further 20-odd million, although many of these will already have access to some screening via their medical insurance.
Around a third of the current cohort are not routinely screened and this increases to around half for the 45-50 year cohort. There is probably not enough capacity in terms of gastro-intestinal specialists in the US to support a move to widespread screening for 45+ and when one includes this group and the higher risk 75+ cohort, there may be 60m people in the US alone who are not currently compliant with an optimal screening regimen.
Sadly, around 60% of the 150,000 diagnoses mentioned previously (and most of the deaths) are from patients presenting with advanced disease due to overt symptoms, as opposed to early detection from routine screening. Whilst colonoscopy is the gold standard, waiting 10 years in between examinations offers a lot of scope for disease to manifest.
How then do we improve the outcomes for this disease? The answer is beguilingly simple: we need to screen more people more frequently to catch it earlier. This is a problem of cost effectiveness and scale. We cannot realistically use more CTC – there aren’t enough CT scanners and colonoscopy also has structural limitations but even so there are still some 15 million procedures in the US each year. However, only around six million of these are primary screening, the remainder are secondary procedure (e.g. post a positive test with another modality or to remove lesions). Many people reject these procedures due to fears of pain and discomfort, inconvenience of “prep”, cost and the general “ick factor”. The latter point is true for the stool-based tests too, as the sub-100% return rates attest.
The logical solution to this might be a blood test - nobody seems to mind those. Two blood-based genetic tests for CRC have been developed. The first, which is commercially available, is called Shield and was developed by Guardant Health. Whilst this is undoubtedly a step forward, especially for those unwilling to have an invasive procedure or deal with their faeces, it is not comparable in sensitivity to Cologuard or colonoscopy.
The headline specificity may seem okay, suggesting a low-ish false positive rate but it is an annual test like FIT rather than three years or five years like Cologuard and Colonoscopy. When considered against the FIT test alone, the false positive rate is around twice as high and adenoma sensitivity in particular is poor; less than half that of Cologuard. GRAIL’s Galleri multi-cancer early detection (“MCED”) blood test also includes colorectal cancer markers and claims a sensitivity of 82% overall for these tumours (much lower than for the CRC-specific tests described previously).
The lower efficacy of blood-based testing compared to Cologuard is not surprising; the mechanical action of digestion causes bowel cells to slough off and these are present in the faeces in large numbers. Conversely, a carcinoma in situ has not penetrated the inner wall of the intestine, making it unlikely that much cancerous material can appear in the bloodstream. Blood tests may well have a future in CRC screening, but that is not yet.
Exact has itself developed a blood-based test and it evaluated this in parallel to “CG 2.0” in its BLUE-C study. Its own view is that the blood-based approach is not a viable alternative to Cologuard. However, there are some people who will not take up Colonoscopy or Cologuard and Exact expect to launch its blood-based test as a second-line alternative for these people within the next few years – even a less reliable test is better than no screening. Collecting samples in parallel should enable the company to evaluate adding additional markers to the blood-based approach to improve its sensitivity and specificity over time but we do not think it will attain Cologuard levels of performance.
“Diagnostic divination delays deaths”
The example above hopefully explains why we think diagnostics will underpin the ongoing revolution in the healthcare delivery paradigm. Prevention is better than cure but, for silent diseases, early detection is the cornerstone of successful intervention.
Running screening programmes successfully in what would otherwise be thought of as a healthy population requires the successful navigation of the triad of efficacy, cost and acceptance.
The winning formula needs to be accurate, scalable and consistent. It needs to minimise inconvenience for the patient and be cheap enough to justify the economics (lives saved per thousand screened versus costs per thousand screened). For the patient, accuracy means more than just early detection, it also means an acceptably low false positive rate – no-one is coming back for a do-over if they have once been put through the wringer unnecessarily.
Colorectal screening programmes are already in place and are seeing expansion, so this is an area where it is deemed worthwhile to spend. When one considers all of the current options though, it surely seems obvious that increased use of Cologuard is the only practicable way to increase uptake. Everything else is either too expensive, too onerous or not good enough. And that is why Exact Sciences is currently the largest holding in the portfolio.
We always appreciate the opportunity to interact with our investors directly and you can submit questions regarding the Trust at any time via:
As ever, we will endeavour to respond in a timely fashion and we thank you for your continued support during these volatile months.
Paul Major and Brett Darke