Doesn't it suck when a press release hits the wires, leads to a big drop in one of your positions, and you honestly have no idea why the market is reacting negatively?
The upcoming data readout for the BLUE-C study from Exact Sciences has all the makings of one of those events. Shares have had a good run recently, but they're near fair value based on expected 2023 operations. That creates a loftier perch to fall from if results disappoint.
How should investors define "disappointing" in this context. Conversely, what results would be needed to claim success?
Focus On The Big Three Metrics
I provided a deep dive into the nuances of colon cancer screening diagnostics when Guardant Health unveiled disappointing results for its liquid biopsy candidate in the SHIELD portfolio. To recap briefly, the U.S. Food and Drug Administration (FDA) has clearly communicated the key metrics it looks for in an experimental tool:
- Sensitivity of colon cancer: The ability to correctly identify individuals with colon cancer. Overall sensitivity is most often communicated, which is the rate of detecting cancer at any stage. Tests are differentiated by their ability to detect early-stage cancers (Stage I and Stage II). It's easier to detect advanced cancers, which makes early-stage detection the biggest driver of overall sensitivity.
- Specificity of colon cancer: The ability to correctly identify healthy individuals. It's generally not great to scare the shit out of people by telling them they have cancer when they don't.
- Sensitivity of precancerous lesions: The ability to detect benign masses that can become cancer. Precancerous lesions are more important for some cancers (cervical or colon) than others (lung).
Although scoring above a minimum level across these three metrics has been adequate to earn regulatory approval, a crowded landscape makes it much more difficult to earn reimbursement coverage and achieve commercial success. For example, the Epi proColon liquid biopsy tool from Epigenomics earned FDA approval, but was rejected by the Centers for Medicare & Medicaid Services (CMS) for Medicare reimbursement. Many private insurers followed suit, essentially making the product noncompetitive.
The next-generation Cologuard test from Exact Sciences won't have that problem, but it must clear a high bar nonetheless. Investors need the diagnostic candidate to outperform the existing Cologuard product, specifically:
- Cologuard 2.0 must deliver minor improvements on colon cancer sensitivity and specificity, with the most valuable improvement being sensitivity in early-stage cancer.
- (Most Important) Cologuard 2.0 must deliver significant improvement on precancerous lesion sensitivity, with 60% representing the closely-watched benchmark. That's roughly the precancerous lesion detection rate of colonoscopies alone (they're >95% when paired with a second diagnostic tool). Achieving at least 55% would be considered a success.
Exact Sciences announced results from a case control study in early 2023 that delivered exactly what investors wanted to see. Cologuard 2.0 delivered colon cancer sensitivity of 95% (vs. 92% for the current test), colon cancer specificity of 92% (vs. 90%), and precancerous lesion sensitivity of 57% (vs. 42%). However, real-world studies typically underperform case control studies – sometimes by a surprising margin.
A case control study uses a diagnostic candidate with tissue samples stored in a laboratory. Scientists and doctors already know the health outcome of these tissue samples ("Is this fecal sample from an individual who had Stage I colon cancer? A healthy individual? An individual with precancerous lesions?"), which allows them to gauge the performance of a diagnostic candidate and informs the design of a real-world study.
A real-world study uses a diagnostic candidate in real humans with unknown colon health. Real humans also take grandkids to soccer practice, sleep past alarms, and fail to follow exacting directions for sample preparation, shipping, and so on. These real-world variables tend to introduce noise and error into clinical studies, which explains why real-world study results tend to underperform case control study results.
If Cologuard 2.0 delivers on the metrics outlined above, then it could reset the market. There's a correlation between the multiplied value of the Big Three metrics and Medicare reimbursement. If precancerous lesion detection is significantly improved relative to the current product, then it would be difficult for other tests to match the economic and diagnostic value.
Note: Open in new tab to see larger, higher quality image.
What are some specific examples of "resetting the market"?
- Successful results could significantly reduce Medicare reimbursement levels for emerging liquid biopsy tools. That includes the SHIELD test from Guardant Health (which could be rejected) and the unnamed liquid biopsy candidate from Exact Sciences. The latter is also being studied in the BLUE-C study, but will have a separate data readout later this year. Liquid biopsy screening tools in colon cancer might struggle to earn $200 per test -- less than half of what Guardant Health expects.
- Successful results could allow favorable inclusion in colon cancer screening guidelines in 2026. The existing Cologuard tool has been creeping higher up the list, but significantly improved precancerous lesion detection rates could put Cologuard 2.0 on equal footing with colonoscopies for the general population. Prices wouldn't necessarily increase, but ordering rates would meaningfully increase in the back half of the decade.
Forecast & Modeling Insights
Solt DB Invest will need to update (increase) our forecast and modeling for Exact Sciences when full-year 2022 results and full-year 2023 guidance are provided in the coming weeks. That's what happens when a business pulls another $100 million in revenue out of thin air to demolish its fourth-quarter revenue guidance.
The current (soon to be outdated) modeling can be viewed in the most recent article.
As a reminder, the margin of safety range for Exact Sciences is based on 2025 modeling. The midpoint for my 2023 model is a valuation of roughly $12.1 billion ($67.35 per share). The midpoint for my 2024 model is a valuation of roughly $14.8 billion ($80.74 per share). The margin of safety range for all three years modeled will increase in the coming weeks, but that's the ballpark.
- To keep it simple, stick to the margin of safety range. A lot of work goes into my research, forecasts, modeling, pleasant emails to researchers, and angry emails to investor relations teams. Exact Sciences will be a great long-term investment.
- For those who want to acknowledge the nuances, shares are bumping up against fair value based on expected 2023 operations. There are more attractive opportunities for putting capital to work in the short term.
Margin of Safety & Allocation
(No change.)
Exact Sciences is considered an Anchor position. The current margin of safety range for the company is below:
- Current Price (market close January 24): $65.30 per share
- Likely Undervalued: <$67.62 per share
- Midpoint: $84.53 per share
- Likely Overvalued: >$101.43 per share
- Allocation Range: Up to 15%
Exact Sciences reported 177.684 million shares outstanding as of November 2, 2022. The margin of safety range above assumes no change in the number of shares outstanding.
Further Reading
- January 2023 company update discussing preliminary 2022 revenue
- December 2022 company update discussing the colon cancer screening market, in the context of Guardant Health's disappointing liquid biopsy results