Takeaway: Natera’s Signatera and Guardant360 are well-positioned. How Illumina will leverage Grail/Galleri is less clear.


We spoke with a Ph.D. in genetics with broad consulting, commercial/marketing, and business development experience (10+ years) in the DNA testing and oncology space. We were a bit surprised by his practical, measured views, which were clearly influenced by commercial experience.


1)     Natera’s (NTRA) moat is broad and defensible, but they likely can’t move upstream.

2)     Grail’s Galleri may have a steeper hill to climb regarding market acceptance than the $8B price tag would suggest. Given the sensitivity reported thus far, the commercial use case and overall practical utility are hard to grasp.

3)     Guardant360 is well-positioned – great for treatment selection, partnerships w/ biopharma, etc.

4)     “Is it actionable? Would it trigger any physician behavior change?” – These are the key questions that will determine success for these tests.

Can you briefly explain the tech and use cases for blood-based cancer tests (liquid biopsy)?

  • I try to always focus on the patient’s journey when considering tech and use cases – early detection means what for diagnosis? Ok, screening -> PT scan – do you see something, can you even treat it? After treatment, will the patient (“pt”) have MRD? What does that mean for adjuvant treatment and is there a need for recurrence monitoring? People must factor in treatment response.
  • Guardant360: Standardized large panel for all pts., looks for mutations or possible drivers/risk factors and signals. It’s the same cost for every panel, every time. This is perfect for treatment selection (partnerships w/ biopharma, clinical trial enrollment, etc.). Test is $1k-$1.2k every time.
  • Grail: Methylation – early detection play. Goal = a standard test for every pt. Looking at the patient journey, this is clearly early side – SCREENING. It’s not clear if this can move into MRD – will depend on the panel. With the same test for every pt there should be some economies of scale.
  • Natera’s Signatera: Different, personalized approach. Must have tumor biopsy sample, so pt must have been diagnosed. Need whole exome or genome sequencing. Based on data, don’t sequence everything – just 16 mutations for each patient (they vary!).
    • The biostats suggest no additional benefit beyond the 16, which could be introns, for example. The tests are only performed at Natera’s labs.
    • The 16 are selected based on the algo, and then the custom panel is used to track the patient, every time.
    • Test is $2,500 to start b/c of whole exome or genome, but cheaper thereafter (e.g., $250-300 for each incremental run). The frequency of testing is an important clinical and commercial question (it’s up to the physician).

If Guardant360 can be used as a screening tool for colon cancer, does that alter competitive dynamics up or downstream? Can Oncotype Dx or Galleri be used for MRD? If so, what’s the impact on Signatera?

  • Yes. Guardant or Grail could move down. Signatera can’t move up because of the need for a biopsy. They can/could partner to build a panel based on blood, but thus far it’s been tumor biopsy data.
  • Guardant thinks they can do it by building a larger panel to catch mutations – Natera doesn’t think so because the mutations vary by tumor type/patient (note, he agrees w/ Natera).

OK, is there anything special about the Signatera process that protects its turf?

  • There are a couple of/few things to consider:
    • TRACERx (April 2017) study -> tumor shed varies by stage and type. If that’s true and we think about the post-treatment patient journey, the Signatera test makes a lot of sense. It would take a long time to displace it. The sensitivity and specificity may be better than what Natera says (the criteria – 16 mutations – is more specific). With guardant, it’s hard to ensure specificity.
    • In theory, the clones could change – nobody knows how much that happens. Tracking 16 KNOWN mutations for the patient makes sense. With Guardant360, you may only find 3 mutations. With Signatera, you know.
    • In theory, a bigger panel could produce good results, but costs are an issue (COGS of $1k per panel = not making money on the clinical testing side). On paper, it seems straightforward, but in practice, you can’t just copy it… i.e., the process isn’t easily replicable.

How does someone differentiate? How could competition develop?

  • Purely on cost, maybe, but Signatera is cheaper after 4-5 tests on a per-test basis (vs. Guardant360, for example).
  • I think Natera wants to see if they can help drive decisions. The Oncotype Dx prospective trial – will MDs change behavior? If outcomes improve, there’s strong evidence for MD and payer adoption. Otherwise, it’ll remain peripheral.
  • Code Stacking is an issue – need to get to value-based pricing and have demonstrated value (prospective study data).
  • Think about Guardant – about 50% of the volume looks like it’s lung cancer – pts that can’t be biopsied (Guardant360 is a good option). They got the coverage determination to expand beyond lung, but we haven’t seen an uptick.
  • What pharma will pay (~$2500/test) is well above the level payers will reimburse at (recall, Foundation medicine w/ 50% of revenue from pharma was a disaster).

Clinical Utility – what’s Galleri a tool for? What’s it good at?

  • The clinical utility is clear w/ lung cancer when you can’t get a biopsy (for Guardant). It helps w/ treatment selection (e.g., EGFR mutation = EGFR inhibitor and then prove out better outcomes w/ data).
  • With Signatera, that must be proved out too. Detect residual disease, give patient additional treatment, see what the impact on outcomes is.
    • To the best of my knowledge, Natera hasn’t built out an oncology sales team.
    • With Grail/Galleri, let’s say you screen someone early. There’s early detection… monitoring sounds good, but what do you do if you detect something? It’s not currently actionable so adoption will be tricky. Detect 6 mos. before a CT scan, do you operate or start on chemo? Run more tests? There’s a potential uptick in utilization. The test isn’t sensitive enough.
    • With Oncotype Dx, patients can avoid chemotherapy, so a payer will cover it at $3k - $3,500.
    • Pharma companies don’t want to say, “Use this test to influence treatment duration.” One of the better use cases for Foundation was to deny 3rd line treatments.
    • It’s very hard for me to wrap my head around Grail/Galleri. It’s clear that Illumina was desperate to bring it back, but it could cause a lot of anxiety for patients and challenges for oncologists, especially in the earlier stages due to the low sensitivity. Early-stage detection is a tougher sell, in my opinion; none of the tests are great for early-stage disease. Once a test gets to 75% sensitivity for early-stage cancers, then it’ll change the discussion.

Do you have any thoughts on Exact and Cologuard?

  • I applaud the commercial team. They were able to incorporate Cologuard into treatment guidelines. Also, it’s not invasive – less so than a blood test.
  • This is the hurdle – “Is it actionable? Would it trigger any physician behavior change?” If a Cologuard test is positive or negative, you progress to a colonoscopy, but you don’t cause a spike in utilization.

Please reach out to  with any feedback or inquiries, questions for future field work, or requests for underlying data.

Thomas Tobin
Managing Director


William McMahon


Justin Venneri
Director, Primary Research