GRAIL’s woes raise questions about the prospects of MCDs and standards for their evaluation

Share on facebook
Share on twitter
Share on linkedin
Share on email
Share on print

In an industry that set its sights on reshaping cancer detection, GRAIL Inc. had the appearance of a pioneer.

If any multi-cancer detection test was heading toward FDA approval and broad reimbursement, GRAIL’s Galleri was on track to get there first. Or so it might have seemed—and so the company said.

A large-scale MCD clinical study was underway in the United Kingdom, with 140,000 people receiving the test through the National Health Service. Another study, the Medicare-Galleri study, called REACH, an acronym for Real-world Evidence to Advance Multi-Cancer Early Detection Health Equity, is enrolling 50,000 individuals, allowing for three annual screens. 

And GRAIL was in conversations with FDA as far back as 2019 about premarket approval of Galleri. The company publicly projected FDA premarket approval by “mid-2024” and “widespread adoption” by 2025.

Investors were bidding up the stock of Illumina Inc., the company that, like a  tango partner, spun off GRAIL, pulled it back in, and ultimately spun it off again. Illumina’s share price peaked at $523 on Aug. 16, 2021. 

Even without FDA approval, Galleri was commercially available as a laboratory developed test, and as of June 30, GRAIL had sold more than 215,000 tests. This, as well as the sale of other similar tests, created an imperative for NCI to push for studying the MCDs, even as the institute’s budget continued to stagnate.

Now, let me just say, once and for all: There are no current data that suggests that these technologies reduce cancer mortality, or even reduce the incidence of advanced stage cancer, which some companies and some people think is an adequate surrogate for the cancer mortality endpoint.

Philip E. Castle

The question before NCI was straightforward: Are people tested with MCDs being helped, or are they swallowing sciencey nonsense and getting harmed?

Today, GRAIL’s prospects have dimmed. The Securities and Exchange Commission is investigating the tangle of corporate moves that include Illumina spinning off the MCD-maker in 2016, then reacquiring it five years later, and ultimately being ordered by the antitrust authorities to divest it.

Both Illumina and newly independent GRAIL are facing class action lawsuits alleging that the insiders had personal financial motives for reacquiring GRAIL in 2021, shelling out $8 billion—more than 90% more than the company was worth, plaintiffs argue. 

The lawsuit states that FDA had told GRAIL directly and unambiguously in 2020 that Galleri’s proposed clinical trials were “wholly insufficient to support FDA approval.” 

In another apparent setback, last May, NHS announced that the preliminary results from the first year of the NHS-Galleri trial didn’t justify proceeding straight away with a large-scale pilot program of the test in NHS clinical practice. The NHS will wait for the final results of the trial, expected in 2026, before making that decision (The Cancer Letter, May 31, 2024). 

Meanwhile, the trial came under fire in an Aug. 7 report in the BMJ, with experts asserting that it is “not suitable to justify a national screening test.”

In August, GRAIL hosted its first earnings call as a public company after Illumina was ordered to divest it. During the call, GRAIL announced the elimination of about 350 employees and three executives. Today, amid questions about Galleri’s potential harms and low sensitivity, the company is operating without a chief medical officer.

The Illumina stock sank to $91 a share, a 10-year low, in November 2023. This amounted to a plunge of more than 82% from the peak in 2021. The stock tanked after Illumina announced that it was forced to take a $821 million write-down on GRAIL—after writing it down nearly $4 billion the year before—which the lawsuit states confirmed that GRAIL “was worth a fraction of what Illumina paid.”

It has hovered between $100 and $150 since.

ILLUMINA STOCK PRICE AND MATERIAL EVENTS INVOLVING SPIN-OFF AND REACQUISITION OF GRAIL

Nasdaq stock price of ILMN from 2016-2024 highlighting key dates
Source: Nasdaq
1/10/2016Illumina forms GRAIL
7/22/2021European Commission announces investigation into Illumina’s proposed acquisition of GRAIL
8/4/2021Illumina board spends $100M to acquire $300M in officers and directors insurance protection covering potential claims related to GRAIL acquisition
8/16/2021Stock peaks at $515.32/share
8/18/2021Illumina announces the closing of the GRAIL acquisition
8/19/2021EC announces investigation into Illumina’s possible breach of standstill obligation
5/6/2022Illumina reports missing revenue targets related to GRAIL
6/10/2022The New York Times questions the clinical benefit and cost of MCDs, focusing on GRAIL
8/12/2022Illumina reports GRAIL missed analyst estimates
6/27/2023STAT reports Illumina to cut 10% of R&D workforce
8/11/2023Illumina announces that the SEC has opened an investigation into the GRAIL acquisition
10/30/2023Investor Carl Icahn files a complaint alleging Illumina and GRAIL’s “personal stake” in GRAIL acquisition
11/10/2023Illumina announces $821M write down
11/13/2023Stock reaches lowest price of $91.11/share
6/24/2024Illumina divests GRAIL
8/13/2024GRAIL lays off ~350 employees and three executives, including CMO Jeffrey Venstrom

None of these woes surprise Donald Berry, professor in the Department of Biostatistics at MD Anderson Cancer Center and co-owner of Berry Consultants LLC, a company that consults with pharmaceutical, biotechnology, and device companies. 

As a former member of the GRAIL scientific advisory board, he found that his counsel was largely ignored. In an interview with The Cancer Letter, Berry described GRAIL executives as religious believers in screening and their MCD technology.

“It was their belief,” Berry said. “It was like a cult, with skeptics banned.

“They arranged to talk to and to listen to people who had drunk the Kool-Aid.”

At an early meeting of the SAB, Berry predicted that GRAIL, should it continue to disregard the fundamentals of screening, was likely to fail. Berry’s explanations of the basics was not acknowledged.

“I would’ve thought, in retrospect, that if somebody on the SAB says what I said, the company executives or other members of the SAB would say, ‘Why do you say that? What are the problems? Can we handle the problems? What would you do?’ Nobody said anything like that,” Berry said. “Nor did they ask me afterwards what I meant. They knew I had not drunk the Kool-Aid.”

Not long after Berry spoke up, GRAIL disbanded and quickly reassembled the SAB—without Berry.

“Certainly, they didn’t want me around,” Berry said. “Maybe the fact that they didn’t ask me further about why I said they were going to fail was writing on the wall.”

Knowledgeable GRAIL staff members did not stick around for long, Berry said.

“There were two people at GRAIL whom I respected,” Berry said. “One was a statistician, one was an MD. They were trying to set up a trial that would do the right thing, namely, address mortality. I had worked with them. They were smart, creative, and knowledgeable. For example, when I would say ‘You know, to show a 20% reduction in cancer mortality across the board is going to be very difficult,’ they understood and suggested alternative routes to FDA approval without sacrificing scientific principles.”

“They were gone a few months after I was gone. I don’t know whether it was of their volition, but it was a major loss for GRAIL.” 

In response to this reporter’s detailed question about the concerns Berry had raised, a GRAIL spokesperson acknowledged only that Berry had indeed served on the company’s SAB.

“Mr. Berry was on GRAIL’s scientific advisory board in the discovery phase of our technology,” the GRAIL spokesperson said in email.

The question of benefit (or harm)

NCI refers to the tests as MCDs while several other entities call them MCEDs, as in multi-cancer early detection tests. Officials say that the word “early” could be misleading because most of the cancers diagnosed by MCDs are detected at an advanced stage.

All MCDs are being marketed as laboratory-developed tests, and there is no mandatory, comprehensive resource for LDTs. FDA’s final rule on LDTs, published April 29, 2024, requires that LDTs be registered and listed as medical devices (The Cancer Letter, May 3, 2024). 

The agency expects laboratories to comply by May 6, 2026. 

At least three MCDs are currently on the market:

  • GRAIL’s Galleri
  • 20/20 GeneSystems’s OneTest 
  • Precision Epigenomics’s (previously Sentinel-10) EPISEEK

So, are MCDs useful?

NCI is hoping to begin to answer this question in a pilot study called Vanguard.

Vanguard, a  project of NCI’s new Cancer Screening Research Network, is designed to address the feasibility of using MCD tests in future randomized controlled trials. It’s intended to inform the design of a much larger randomized controlled trial that would be powered to assess a mortality endpoint.  

“The challenge is: Saying that you can detect cancers—there’s not a single person walking down the street that wouldn’t sign up for it, because the public doesn’t fully understand the difference between detecting cancer and preventing you from dying from cancer,” Philip E. Castle, director of NCI’s Division of Cancer Prevention, said to The Cancer Letter. “So, I think it’s a real challenge in communications about what benefit, if any, these have. And it’s very hard for even the very knowledgeable public to make an informed decision about whether to get these tests. 

“We have to make sure the juice is worth the squeeze.”

The institute has so far selected two MCD technologies for testing in the Vanguard, Castle said. The institute cannot yet publicly disclose which technologies have been chosen because the agreements haven’t been finalized, he said.

Said Castle:

We’re asking healthy people to undergo these new technologies. So, the bar is very high. We have to make sure that we can actually deliver care, and that we don’t end up inadvertently harming people more than the benefits we are hoping we’ll see from some of these technologies. 

And we don’t know which of these technologies are going to work, and for whom, and how. 

So, on one level, we are very excited, of course, because there are cancers for which we do not have an intervention that can reduce the morbidity or mortality, or even the incidence of these cancers. But we have to do our due diligence. That’s what this is about. 

And that’s, I think, the unique role that the NCI can and has played for decades—as a fair evaluator of technologies without prejudice or conflict of interest. I guess that we have a small conflict of interest. 

We want these to work. We want to save lives. We want to prevent cancer and death due to cancer. But aside from that, we have no financial incentives or any other conflict of interest that might sway us. Therefore, we can evaluate these technologies in a way that hopefully brings public trust to the results…

We just don’t have a lot of knowledge of the natural history of cancer as it goes through the stages and how that relates to a mortality benefit. 

And you can get arguments on both sides of this. 

From my perspective, you do the trial to the mortality endpoint, then you can with much greater certainty say, ‘These are surrogates for that mortality.’ And then all the trials subsequently, and even our understanding of the natural history of cancer, are greatly increased. 

And so, the only way to get out of this loop is to have that correlation—whether you look at advanced-stage cancer [as the endpoint] and then at mortality, or you say we have to have mortality [as an endpoint] is, again, part of the debate between some experts and the companies who have developed these technologies. 

While GRAIL and several other MCD developers argue that the clinical benefit of MCDs can be tested using a surrogate endpoint, such as stage shift or absolute reduction in the number of late stage diagnoses, the gold standard of mortality reduction is the only way to know whether a screening test is actually beneficial to the public, NCI officials say (The Cancer Letter, April 12, 2024). 

It was their belief. It was like a cult, with skeptics banned. They arranged to talk to and to listen to people who had drunk the Kool-Aid.

Donald Berry

“It takes a long time to get to a mortality benefit, and they have some bottom-line issues around that,” Castle said. “From our perspective—and again, we have to be in it for the public good and the public trust—we feel fairly strongly the need to show that true benefit [mortality reduction] to the population. As I said, once we have that, then we’ll know what some good surrogates of cancer mortality are.”

Between overscreening, invasive and costly follow-up, and the presence of false positives and false negatives, many screening methods—ovarian and prostate cancer screening among them—haven’t demonstrated reduction of mortality (The Cancer Letter, Jan. 12, 2024).

None of the available clinical studies of Galleri—or any MCD test—provide either cancer-specific or overall mortality data. The NHS-Galleri trial includes a secondary endpoint that examines cancer-specific mortality, but these data will not be available at least until the trial is completed in 2026.

“Now, let me just say, once and for all: There are no current data that suggest that these technologies reduce cancer mortality,  or even reduce the incidence of advanced stage cancer, which some companies and some people think is an adequate surrogate for the cancer mortality endpoint,” Castle said. “There’s been no demonstration of clinical benefit to date or could be extrapolated to a clinical benefit to date.”

GRAIL’s recent challenges may spell trouble for testing MCD technology.

Since Galleri is already commercially available, those who choose to place faith in the test will simply buy it, making further studies of MCDs even more difficult.

“There’s been a lot of publicity about these tests—and, again, there’s been no demonstration of any benefit to the population, no true direct demonstration—there’s a lot of hoopla, a lot of fanfare around this, and that affects people’s willingness to undergo randomization,” Castle said. “Potentially one-third of them are not going to get an MCD test. Therefore, that presents perhaps some challenges; we’ll find out. That’s why we’re doing the feasibility study before we go to the effort of running a very large, very expensive, very long trial. We need to make sure that we can actually implement the trial.”

GRAIL: NHS study continues

GRAIL officials say that the testing of Galleri by NHS England continues. The health service has decided not to move forward with a large-scale pilot program in its clinical practice, but may do so in the future, the company said.

A GRAIL statement reads: 

The NHS-Galleri trial is the first and largest randomized controlled trial of a MCED test which started enrolling participants in 2021 and completed enrollment of more than 140,000 participants in 10.5 months.

The trial was designed to inform implementation of the Galleri MCED test as a national screening programme if ultimately recommended by the UK National Screening Committee on the basis of the final study results, which are expected in 2026. 

The trial, sponsored by GRAIL Inc., was intentionally designed with three consecutive years of screening in order to achieve the primary endpoint, which is the absolute reduction in the number of ‘late stage’ (defined as Stage III and IV) cancer diagnoses. 

GRAIL’s partnership with NHS England recognized that GRAIL had the most advanced MCED product in development, as well as the resources and infrastructure to conduct and fund a clinical trial of this scale.

As one component of NHS England’s ambition to diagnose three-quarters of all cancers at an early stage, it had considered an opportunity to implement an accelerated pilot of the Galleri test in England prior to full completion of the NHS-Galleri trial, if early results were exceptionally compelling.

Following a snapshot of only the first-year results from the trial, NHS England decided to await final results from the three-year trial before determining whether to initiate a pilot of the Galleri test in the NHS. 

NHS England determined that, while the early analysis showed that the assessed clinical performance of Galleri was “very promising”—consistent with or better than Galleri’s clinical performance observed in previous published studies—there was not enough early compelling evidence to accelerate implementation through a pilot programme at that early stage. 

The company argues that Galleri and other MCDs should be evaluated using modeling and intermediate endpoints. 

According to the company: 

Using all-cause mortality endpoints for screening in large trials for new technologies like MCED tests is unrealistic and impractical.

The primary aim of cancer screening is to reduce suffering and premature death from cancer, not all-cause mortality. All-cause mortality is often dominated by unrelated causes of death, such as cardiovascular or infectious diseases, which makes observing the specific benefits (or harm) of cancer screening challenging.

The challenge is: Saying that you can detect cancers—there’s not a single person walking down the street that wouldn’t sign up for it, because the public doesn’t fully understand the difference between detecting cancer and preventing you from dying from cancer.

Philip E. Castle 

Further, cancer screening trials are typically underpowered to detect effects on all-cause mortality. Since no approved cancer screening test has demonstrated (or been required to demonstrate) a reduction in all-cause mortality in randomized controlled trials, insisting on such endpoints for MCED tests is unnecessary. 

Moreover, these trials often take decades to complete, during which time tens of thousands of people will continue to die of cancer every day across the world.

Modern approaches using high-quality cancer screening trials with modeled and intermediary endpoints can establish efficacy more quickly and efficiently, and at a lower cost. 

These approaches have been gaining traction in scientific literature including in papers published with authors from the NHS-Galleri trial, the NCI, and various academic institutions, which highlight endpoints such as late-stage incidence reduction and modeled mortality as informative for implementing new cancer screening technologies. 

Such intermediate endpoints, like the reduction in late-stage cancer diagnoses, provide valuable insights into the effectiveness of screening. For many cancer types, an earlier stage at diagnosis is strongly linked to better survival, with a significant portion of cancers amenable to cure at early stages, but not after metastasis. 

The NHS-Galleri trial also incorporates a secondary endpoint that examines cancer-specific mortality between the intervention and control groups. In a targeted analysis, samples from control group participants who died from cancer will be retrospectively tested using the Galleri test, allowing for an assessment of cancer-specific mortality, while ensuring timely completion of the study.

The future of MCED tests, such as the Galleri test, represents a transformative shift in cancer screening, particularly for cancers that currently lack recommended screening options. With a sensitivity range of 40% to 70% for these unscreened cancers, the Galleri test provides a significant improvement over the current state, where screening detection rates are effectively zero. 

The Galleri test is also intentionally designed for a low false positive rate to improve the efficiency and effectiveness of screening when added to recommended cancer screening tests. It was engineered to balance the detection of a shared cancer signal across many types of invasive and aggressive cancer—which shed measurable amounts of DNA into the blood—while maintaining a single low false positive rate of less than 1%.    

Beyond our clinical studies, real-world use of the Galleri test has detected many of the most aggressive cancers in early stages, including pancreatic, head-and-neck, esophageal, liver, and stomach cancers. For the majority of these cancer types, there are no other screening options available, and most deadly cancers are often found too late. 

GRAIL’s statement notwithstanding, at least one trial of an early detection modality—the National Lung Screening Trial—did find a decrease in all-cause mortality as a result of low-dose spiral computed tomography screening (The Cancer Letter, Nov. 5, 2010).

Shareholder suits allege falsehood, “fraudulent accounting” 

A class action lawsuit filed by shareholders alleges that Illumina and GRAIL officials misled investors about the clinical and commercial promise of Galleri, thus artificially inflating Illumina’s stock price, and causing investors to suffer losses when the stock collapsed.

The complaint focuses on the saga of Illumina’s formation and its subsequent reacquisition of GRAIL. 

At the time, the companies described the reacquisition as a “moral obligation” that would save tens of thousands of lives by combining GRAIL, the manufacturer of the test, with Illumina, the maker of the sequencing equipment.

Antitrust officials in Europe disagreed. 

They viewed the $8 billion merger as the creation of a vertically integrated monopoly and issued a standstill order. U.S. antitrust officials stood back, letting EC take the lead (The Cancer Letter, Sept. 9, 2022). When the merger was closed anyway, EC officials described the move as “an unprecedented and very serious infringement undermining the effective functioning of the EU merger control system.”

According to the lawsuit, before completing the GRAIL acquisition, Illumina board members purchased directors and officers insurance with up to $300 million in coverage, paying an annual premium of up to $100 million.

This was done “specifically to insure the Board’s and senior management’s acquisition-related conduct,” the shareholders’ lawsuit states.

The purchase doubled the company’s then-current coverage for a premium more than 25 times what the company had paid the year before, the suit states. The suit was filed by the law firm of Bernstein Litowitz Berger & Grossmann LLP in the Southern District of California. 

In response to questions emailed by The Cancer Letter, GRAIL officials addressed clinical questions, but didn’t respond to questions about the lawsuit, and Illumina didn’t respond at all. GRAIL is represented by the law firm of Latham and Watkins LLP and Illumina is represented by Cooley LLP. Neither of these firms responded to The Cancer Letter’s request for comment by deadline.

Illumina knowingly oversold the potential value of GRAIL to investors in the midst of the acquisition, the lawsuit alleges.

The defendants are alleged to have misled investors by:

  • Feeding financial projections to investors and the public that were up to quadruple the actual estimates shown secretly to Illumina’s board.
  • Promising certainty of the clinical benefit of Galleri, despite clinical evidence showing that Galleri testing “caused substantial harm to patients, with no discernible clinical benefit, at an extraordinary cost.”
  • Touting a mid-2024 FDA approval timeline for Galleri and claiming to have “been working… with the FDA for a number of years on designing the studies that will be part of the ultimate submission.” In reality, FDA explicitly told GRAIL in private meetings that “existing clinical evidence and proposed clinical trials for Galleri were insufficient to be considered for FDA approval, let alone sufficient to support the baseless statement that acquiring GRAIL would help ‘save lives.’”

The lawsuit attributes the steep drop in Illumina’s stock to a succession of bad news:

  • News of Illumina’s decision to acquire GRAIL against the EC’s standstill order became public,
  • Senior executives involved in the GRAIL transaction, including Illumina’s then-CEO Francis deSouza and Alexander M. Aravanis, GRAIL’s former chief scientific officer who became Illumina’s chief technology officer through the acquisition—“unexpectedly or forcibly resigned,”
  • Experts publicly expressed doubt in Galleri’s clinical effectiveness and potential for FDA approval,
  • Illumina slashed R&D spending and laid off 10% of its R&D team “to offset extraordinary operating losses GRAIL was imposing,”
  • Illumina disclosed that it was being investigated by the SEC into whether Illumina had breached the standstill obligation,
  • Illumina announced that it was forced to take a near-$1 billion write-down on GRAIL. A year earlier, Illumina announced a nearly $4 billion write-down on GRAIL.

Meanwhile, several Illumina and GRAIL top officials were able to invest in GRAIL prior to the $8 billion merger. 

The lawsuit states:

Unbeknownst to investors, Illumina closed the GRAIL transaction because it enabled GRAIL’s private investors, including members of the Illumina and GRAIL leadership, to secretly pocket hundreds of millions of dollars. 

Specifically, as alleged herein, Lead Plaintiffs’ forensic analysis has shown that approximately 70 million shares of GRAIL stock, valued at over $830 million at the time the acquisition was announced, were never publicly accounted for. In concealing this secret insider interest, Illumina violated Generally Accepted Accounting Principles (GAAP) in the way it accounted for its ownership in GRAIL—violations specifically carried out to enable Defendants to avoid disclosing insiders’ hidden interests.

Former NCI Director Richard Klausner, who founded GRAIL, is one of the eight individual defendants named in the suit.

The lawsuit states:

Defendant Klausner, a former Illumina executive and then-current GRAIL Board member, purchased approximately 25 million shares of GRAIL common stock for $5.1080 per share—or an approximately $125 million investment—in the Series D round announced on May 6, 2020, at the exact same time, Illumina was making plans to reacquire GRAIL, and while GRAIL’s Board was expanding its membership and approving substantial personnel changes in anticipation of that acquisition in April and May 2020.

Just four months later, in September 2020, the shares Klausner purchased through Milky Way [an investment vehicle Klausner founded in British Virgin Islands] while he was overseeing and approving these extraordinary measures as a member of GRAIL’s Board would more than double in value when Illumina publicly announced the GRAIL acquisition, delivering Klausner and Milky Way over $100 million in risk free profits.

Tellingly, Milky Way was dissolved on July 4, 2023, just weeks after deSouza’s resignation from Illumina, allowing Defendants to continue concealing what became of Defendant Klausner and Milky Way’s massive stake in GRAIL. 

In April 2023, the Federal Trade Commission issued an opinion and order requiring Illumina to divest GRAIL “to protect competition in life-saving technology market.” In June 2023, Illumina appealed this decision, but the court ruled that there was substantial evidence supporting the FTC’s ruling that the deal was anticompetitive.

On Dec. 17, 2023, Illumina announced that it would spin off GRAIL. Unable to find a buyer, Illumina was forced to divest GRAIL through a distribution of shares to Illumina shareholders. Based on GRAIL’s trading since it became public, the company has a market capitalization of around $500 million, the suit argues.

This “confirmed that GRAIL was, in truth, worth around $500 million—or just 6% of the $8 billion that Illumina paid to acquire it.”

A key piece of Illumina’s justification for breaking the EC’s standstill order was that Illumina would be able to speed the development, regulatory approval, and commercialization of Galleri because Illumina had “the teams that can work on reimbursement and regulatory approval, and so we can dramatically accelerate getting this test into the hands of people whose lives it could save,” the lawsuit states.

The plaintiffs dispute this notion. Illumina had “failed to demonstrate that its claimed ‘regulatory expertise’ was superior to that which GRAIL already possessed,” the plaintiffs allege. 

The lawsuit reads:

GRAIL had already obtained breakthrough device designation for Galleri on its own. Illumina, on the other hand, had only ever obtained premarket approval for one Class III NGS-based diagnostic test, and in that instance, a third party sponsored the clinical study upon which approval was granted.  

In truth, Defendants had no actual ability to accelerate Galleri’s commercialization and never planned or modeled any “acceleration”—rendering that statement and Defendants’ representation that the acquisition would “save lives” materially false and misleading when made. 

As confirmed by documents and testimony in the FTC proceeding, neither Illumina nor GRAIL conducted any analysis to determine whether Illumina could in fact “accelerate” Galleri’s adoption and neither planned for that possibility…

The FTC action also confirmed that Illumina lacked any relevant experience or expertise in FDA approval or reimbursement that would support its ability to accelerate Galleri’s adoption. 

Illumina also misled investors with false financial projections, the lawsuit alleges. In November 2020, Illumina publicly presented two sets of financial forecasts for GRAIL for fiscal years 2021 through 2030: a pessimistic “Case A” and an optimistic “Case B.”

“Case A” predicted that GRAIL would generate $14.4 billion in annual revenues by 2030. In “Case B,” GRAIL’s revenues would approach $24 billion by that time. These financial projections supported GRAIL’s $8 billion valuation.

Driving these numbers was a projected increase in the number of health systems and employers that would purchase and utilize Galleri and that “broad reimbursement would be achieved in calendar 2025.” 

Because of these projections, GRAIL was valued “at $11.15 billion to $43.95 billion, using a discounted cash flow analysis and at up to $19.80 billion using a discounted equity value analysis—figures that made the over $8 billion acquisition look like a steal,” the lawsuit states. 

However, both “Case A” and “Case B” were much more optimistic than the true estimates that were shown only to Illumina’s board, the lawsuit alleges. 

The lawsuit reads:

Unknown to investors, the actual projections considered by Illumina’s Board were far more pessimistic than even the supposedly conservative “Case A” scenario Defendants shared with the public…Illumina’s Board reviewed management prepared slide decks in April 2020 which showed that “[a]n acquisition of GRAIL could generate revenues of >$8B by 2030”—approximately half the revenues projected in the conservative Case A and one-third to one-fourth of the revenues projected in the more optimistic Case B.

Testimony and documents produced in the FTC proceeding, much of which still remains redacted or under seal, likewise confirm that the actual numbers secretly presented to Illumina’s Board were far lower than those disclosed to Illumina’s public investors. 

Specifically, documents produced in the FTC proceeding show that the actual “Deal Model”—the model that Illumina used to value GRAIL—projected test sales of 1.5 million from 2022 to 2024, and 50.5 million from 2025 to 2030, sales numbers that similarly translate into revenues that are roughly half those reported in the S-4’s purportedly “conservative” Case A. 

Specifically, those materials showed valuations for GRAIL as low as $3.3 billion to $11.9 billion—figures far below the valuations Illumina broadcast to investors in the S-4 (or GRAIL’s actual purchase price). 

The most aggressive and optimistic internal valuation in April 2020 barely approached the low end of the range reported in the S-4 under the purportedly “conservative” Case A scenario, and was a small fraction of the more optimistic $28.2 billion to $43.95 billion Case B:  

It was highly misleading for Defendants to present these financial projections to investors to justify the more than $8 billion deal, while holding back the actual financial projections considered by Illumina’s Board in April 2020, which were roughly one-quarter to one-half of those presented to investors.  

The “binary risk is largely retired”

As far back as 2020, Illumina and GRAIL advertised that Galleri’s clinical benefit was already certain. 

Then CEO Francis deSouza even claimed that the “binary risk is largely retired,” the lawsuit alleges.

The suit states:

At a September 24, 2020, Cowen investor conference, Defendant deSouza pointed to the clinical data supporting Galleri’s effectiveness, answering the “binary” question about whether the product would “work or not” in the affirmative. 

He stated that during the “last year, through the data that GRAIL started to publish from its large-scale clinical studies at ASCO, at ESMO, the peer reviewed journal that they published earlier this year that gave us comfort to realize, well, I think they’ve cracked the code. Their technology works, it’s been tested in very large numbers, and the data has been peer-reviewed.” 

According to deSouza, the research on Galleri demonstrated that the “binary risk is largely retired, and that if we get involved now, we can accelerate the development of what’s going to be the largest genomic application we’ve ever seen.”  

The lawsuit highlights real-world experiments aimed to test Galleri’s clinical utility. 

“Galleri’s actual real-world performance provides damning evidence that Defendants’ Class Period statements about Galleri’s “proven technology” were false when made, and underscores the need for the kind of rigorous clinical evaluation required by the FDA that GRAIL did not have,” the lawsuit says.

San Francisco Firefighters Cancer Prevention Foundation funded Galleri screening for 1,786 active and retired firefighters at a cost of over $1 million. In this “real-world” experiment, nearly every concern surrounding MCD testing—high false positives, missed cancers, detection of late-stage disease, invasive workups, and high cost—were on full display.

“The test did not identify the statistically expected number of cancers, provided false positives for half the cancers it did identify, and missed at least three cases of confirmed cancer. In other words, the exercise showed Galleri to be a disaster,” the lawsuit says.

The lawsuit reads:

During that testing round, which was conducted on December 6, 2022—i.e., in the middle of the Class Period—Galleri generated six false positives and identified only five instances of cancer. 

As the clinicians Lead Counsel consulted noted, such a result is alarming in light of the fact that, based on statistical experience, cancer should have been detected in about 29 (not five) of the 1,786 tests.

Moreover, of the five instances of where cancer signals were identified, two individuals already knew they had cancer to begin with.

Further, all of the instances of detected cancer were at a “late stage,” and where the cancer was too advanced to make a difference for the patients. One firefighter was diagnosed with stage 3 pancreatic cancer and is now in hospice, while another firefighter was diagnosed with lung cancer, which has now metastasized to the brain. 

Galleri also missed at least six instances in which the San Francisco firefighters actually had cancer—failing to identify instances of melanoma, prostate cancer, and lymphoma that were identified by private doctors just months after the Galleri test indicated “no cancer signal.” 

Screening increases cancer rates—dramatically. About that we’re certain. What is uncertain is whether it decreases cancer mortality. One needs appropriate clinical trials to address mortality.

Donald Berry 

About six months after the screenings, additional firefighters reported that they had been diagnosed with cancer, even though the Galleri test had shown “no cancer signal.” 

Further, in one instance in the San Francisco firefighters screening where Galleri generated a false positive, the patient underwent a painful bone marrow biopsy procedure to determine whether the patient did have cancer. 

In that case, the patient’s bone marrow biopsy procedure, clinically shown to involve “bearable pain” to the “worst possible pain” a patient can experience, revealed that there was, in fact, no cancer.  

The screenings provided to the San Francisco firefighters were an expensive “disappointment.” 

According to Tony Stefani, the organization’s founder, the screenings demonstrated that the probability of catching an early-stage cancer through the Galleri test was “negligible,” and the screenings only caught “some serious cancers at a later stage,” which likely would have been detected within one or two months without the test. In addition, as Stefani noted, the price of the tests—over $1.1 million, more than 10 times the amount the organization had spent on other cancer prevention studies or tests—was “ridiculous.” 

Given the test’s high costs and useless real-world performance, Stefani stated that Galleri was a total “disappointment” and that the foundation had been sold a “bill of goods.”

According to the lawsuit, this real-world experiment was “prominently featured on GRAIL’s website and in marketing materials as an example of how ‘[m]ulti-cancer early detection screening can be critical for those at elevated cancer risk.’”

“GRAIL knew this kind of thing would happen, or at least they heard it from many people, including from me,” Berry said. “These are standard issues in cancer screening. Regarding, ‘contrary to Defendants’ positive statements about GRAIL’s efficacy’ GRAIL did incorporate these various measures of efficacy and safety in their presentations. But like all screening aficionados, they accentuated the positive. On the other hand, it was wrong to not adequately inform these firefighters about the down sides they knew about. Actually, no one knows all the downsides, and the SF firefighters should have been told that as well.”

One issue this example highlights is money, Berry said. 

Said Berry:

In the U.S., there is a huge divide between drugs and devices, including in-vitro assays. If Galleri were a drug, the company would not be able to charge for it. It’s in development. Regulators have not “approved” or “cleared” it for any purpose. Until they show that the assay is actually worth something, they shouldn’t be allowed to charge for it.

The SF firefighters example could be a clinical trial with IRBs and FDA oversight and all the ethical trappings (such as signed informed consent) associated with clinical experimentation. But no payment from the participants. 

Some might argue that the side effects from drugs are more serious than from devices. Not so in the case of cancer screening. GRAIL and the SFFCPF took healthy young people and messed up many of their lives. And GRAIL profited from doing so. 

It’s possible that some of the participants benefited, but this is uncertain. It’s quite possible that none benefited. Those diagnosed with cancer will think that they benefitted, that seems to be a natural reaction. But they can’t know. The only things they can know is that they have cancer now and without the screening they wouldn’t have cancer. In fact, perhaps they would never have been diagnosed with cancer. A cautionary tale.

A second set of data of 2,000 first responders funded by the City of Mesa, AZ, at the Vincere Cancer Center “produced similarly abysmal results,” the lawsuit stated.

“In that instance, 2,000 first responders were screened—but the test missed at least 28 cancers, leading the program’s director to conclude that Galleri ‘gives a false perception that you have cancer or you don’t,’” the lawsuit says. “The test’s sensitivity in the Mesa program (6.7%) was much lower than that publicly reported by GRAIL (29%), and the test missed a staggering 93 of every 100 cancers in the screened population (93%).”

These real-world experiments are cautionary tales for all MCD developers, Berry said. 

“Screening increases cancer rates—dramatically. About that we’re certain,” Berry said. “What is uncertain is whether it decreases cancer mortality. One needs appropriate clinical trials to address mortality.”

Jacquelyn Cobb
Associate Editor
Table of Contents

YOU MAY BE INTERESTED IN

Arineta, a provider of cardiovascular imaging solutions, announced that its SpotLight Duo cardiac CT scanner has received 510(k) clearance from FDA for low-dose lung cancer screening. This clearance allows providers to use a single ultra-fast CT platform for both cardiac and lung imaging and diagnostics, delivering more complete care to high-risk patients.
According to a recent report, “The Cancer Paradox: Oncologists’ Perspectives on Barriers to Advanced Cancer Care and Recurrence Monitoring,” oncologists believe they are seeing more and more patients with advanced cancers, and that screening barriers are the leading reason why. They worry that current tests may not catch cancer recurrence early enough.
Jacquelyn Cobb
Associate Editor

Never miss an issue!

Get alerts for our award-winning coverage in your inbox.

Login