The world’s richest doctor had a very bold plan.

He’d assemble an unprecedented collaboration of companies, researchers, and doctors. Their mission: to vanquish cancer. By the year 2020, they’d build a working vaccine and test therapies in 20,000 patients. They would forever transform medicine.

Or so Dr. Patrick Soon-Shiong vowed when he launched his audacious “Cancer MoonShot 2020’’ a year ago.

The supremely self-confident billionaire behind that vision­ has drawn attention at the highest levels: He has talked cancer research with Joe Biden, Bill Clinton, even the pope. He has met with President Trump at least twice since the election. Riding high on his pledge to win the war on cancer, Soon-Shiong is said to have pitched a role for himself as national health care czar in talks with Trump’s team.

But a STAT investigation of Soon-Shiong’s cancer moonshot has found very little scientific progress.

At its core, the initiative appears to be an elaborate marketing tool for Soon-Shiong — a way to promote his pricey new cancer diagnostic tool at a time when he badly needs a business success, as his publicly traded companies are losing tens of millions of dollars per quarter. STAT also found several inflated claims, with the moonshot team taking credit for progress that doesn’t appear to be real.

Soon-Shiong’s use of the moonshot to advance his business interests may be good for his investors. But it increasingly looks destined to disappoint patients — the latest in a long trail of failed quests to win the war on cancer.

Those conclusions are based on a review of hours of video and dozens of scientific presentations, legal filings, clinical trial summaries, and press releases, as well as exchanges with more than a dozen physicians and institutions named as moonshot collaborators.

Soon-Shiong declined repeated requests for interviews in recent months, though spokeswoman Jen Hodson did provide two lengthy statements she said to attribute to him. They touted nine areas of “remarkable progress’’ in the moonshot initiative’s first year.

STAT asked several independent scientists to review Soon-Shiong’s claims. Their conclusion: The data don’t back up the hype.

“The clinical breakthroughs touted by Patrick Soon-Shiong are less than modest — they are the most minuscule and vague findings,’’ said Dr. Vinay Prasad, a hematologist-oncologist at Oregon Health and Science University.

He called them “overblown beyond what is reasonable or fair.’’

The moonshot focuses on the hot field of immunotherapy, which involves revving up a patient’s immune system to fight cancer. Soon-Shiong, like others, plans to focus on testing drugs in various combinations. But he’s bent on doing it faster than everyone else. His plans center on two types of immunotherapies: a treatment vaccine and genetically ­engineered “natural killer’’ cell therapies ­being developed by one of his businesses.

STAT found that the moonshot has aggressively promoted an expensive diagnostic tool, called GPS Cancer, which analyzes patients’ tumors and recommends a course of treatment. It’s sold by one of Soon-Shiong’s companies and drives business to labs owned by another of his companies.

Under the banner of the moonshot, Soon-Shiong has paid for researchers around the world to use GPS Cancer in their work, even though there have been no published studies validating it as an effective diagnostic test. Indeed, several of the “milestones’’ the moonshot initiative has cited as exciting progress in the war on cancer simply involve various doctors or hospitals agreeing to use GPS Cancer.

Among STAT’s most striking findings were cases of Soon-Shiong’s team taking credit for progress that doesn’t appear to have been made.

Soon-Shiong’s statements, for instance, said that the global pharma companies Pfizer and Merck KGaA had joined the moonshot — which, if true, would be a big step in building a worldwide coalition. But representatives of those companies said they’re not aware of any involvement. The moonshot’s website also lists Johns Hopkins as a collaborator; neither the university nor its affiliated hospital was aware of any such involvement.

Other companies and institutions are characterized as moonshot collaborators mostly, it appears, because they have committed to using or paying for doctors to use GPS Cancer.

In another example of inflated claims, Soon-Shiong cited launching “over 20’’ clinical trials as an example of the “remarkable progress’’ made in the first year of the moonshot. He pointed STAT to a federal registry listing 23 trials. But 13 of them were launched years before the moonshot started — in some cases, almost a decade before. Ten were completed before the moonshot was launched.

Asked why he was counting these old trials as current accomplishments, Soon-Shiong said he and his team have selected experimental therapies from old trials as “worthy candidates’’ to be tested in new combinations.

Soon-Shiong has also spoken extensively about promoting collaboration between big players in the world of cancer research. But so far, STAT has found, it appears that all of the corporate collaboration involves research conducted by a handful of early-stage biotech companies — all of which Soon-Shiong owns or holds stakes in.

Some academics have credited the moonshot with advancing their work, in part by launching “working groups’’ focused on specific types of cancer. Among the topics discussed by at least some of the working groups: how to use Soon-Shiong’s GPS Cancer tool in clinical trials.

Visionary, or shameless self-promoter?

Soon-Shiong, 64, has long had a reputation as a daring visionary — or a “shameless self-promoter,’’ depending on who’s describing him. He launched his moonshot the very same week that then-Vice President Biden introduced the entirely separate federal moonshot to fight cancer.

In the year since, Soon-Shiong’s enormous ambition has only inflated. What was originally called “the nation’s’’ most comprehensive collaboration against cancer is now billed as “the world’s’’ most comprehensive.

On social media, Soon-Shiong’s team boasts that it’s planning to develop not just an effective vaccine, but one that could treat “all cancer types at every stage’’ — within about three years.

His team has also inflated the initiative’s name: It’s now called Cancer Breakthroughs 2020, because calling it a moonshot “hardly does us justice,’’ the website explains.

Researchers who have worked closely with him say the former transplant surgeon, whose net worth is estimated at $8.8 billion, genuinely cares about patients and is passionate about pushing science forward.

Other scientists, however, are deeply skeptical of Soon-Shiong. In the early 1990s, he seemed to be suggesting he had cured diabetes after he implanted cells that secrete insulin in a patient. But the results didn’t hold up.

Bold claims with little evidence to back them

It’s hard to overstate what a presumptuous goal it is to aspire to develop a cancer vaccine that could treat any patient by 2020.

To be sure, Soon-Shiong’s general idea for a vaccine — to customize it for each patient based on the unique genetic markers of their tumor — is widely seen as a promising ­approach within the field of immunotherapy.

But the science is incredibly hard. There are just one or two — depending on your definition — approved cancer treatment vaccines on the market. They’re meant for narrow subsets of patients and have shown mixed results.

Drug companies have poured years, and many tens of millions, into immunotherapy; while they still think it has promise, they point to an array of serious challenges, including patient deaths in clinical trials.

And some of Soon-Shiong’s ideas about the field are “far-fetched,’’ said Dr. Eric Topol, a geneticist who directs the Scripps Translational Science Institute.

Consider Soon-Shiong’s oft-stated conviction that the engineered natural killer cell therapies he’s developing through his company NantKwest could be a universal treatment, effective for any patient. There’s no evidence to back up that claim.

“Some of his ideas may ultimately click someday, but he talks about them as if there’s more support than there is. And if there is more, it sure isn’t published,’’ Topol said.

In his statements to STAT, Soon-Shiong boasted that his initiative has achieved “clinical breakthroughs’’ in its first year.

As evidence, though, he pointed to just one research poster, not yet peer-reviewed: It found that tumors shrank in one of three patients treated with an experimental therapy involving NantKwest’s natural killer cells for a rare cancer called Merkel cell carcinoma.

“This is the type of poster that you pass quickly on your way to credible data in an oncology conference,’’ said Prasad, the Oregon oncologist.

As for the pledge to treat 20,000 patients in the next few years, the moonshot appears to be far off the pace needed to reach that goal.

STAT looked at all the studies affiliated with the moonshot’s primary clinical trial program, known as QUILT, since the initiative was launched. There are slots for just 252 patients, and not all have been filled.

The moonshot is breaking new ground this month with the first US human test of an experimental cancer treatment that combines a “fusion protein’’ with natural killer cells. Outside scientists say it is an intriguing concept. But this is just a trial to test safety, most likely in a patient with Merkel cell carcinoma. Even in a best-case scenario, it is many years removed from being widely used.

The moonshot as marketing tool

When you look closely, Soon-Shiong’s moonshot initiative looks less like a diverse coalition than a roll call of his tangled web of business interests.

The five biotech companies that are participating in the moonshot are the only ones sponsoring registered QUILT trials. And they are all closely tied to Soon-Shiong: He is either the CEO, a board member, or the controlling owner in each of them.

Though Soon-Shiong has talked for a year about bringing major drug companies into the coalition, so far, just two have joined: Amgen­ and Celgene. He is a shareholder in both. And both are investors in Soon-Shiong’s companies. Neither is involved in the moonshot’s clinical trial program; they wouldn’t provide any information about what, if anything, they are doing for the initiative.

The moonshot website also touts a “historic alliance’’ with companies like Bank of America, which insures its own employees, and Independence Blue Cross in Philadelphia. The role of both appears to be simply that they cover doctors’ use of the GPS Cancer diagnostic for patients on their insurance plans.

From the start, the GPS Cancer test has been a central element of the moonshot. Press releases call it a “key enabler’’ and a “cornerstone’’ of the whole initiative. And a promotional video for the moonshot project doubles as an advertisement for GPS Cancer.

Soon-Shiong has high hopes for GPS Cancer, which involves two of his businesses: NantHealth sells the diagnostic tool and Nant­Omics labs interpret the results.

He debuted the tool at a big scientific meeting for cancer specialists last summer. In an unusual move for a CEO, Soon-Shiong sat at the Nant booth in the sprawling exhibition hall, explaining the company’s products and taking questions from physicians — a role usually left to sales representatives.

Soon-Shiong’s team claims that GPS Cancer is the most comprehensive test of its kind. It doesn’t just analyze a tumor’s genetics, but also reports on the precise amount of different proteins expressed by genes in the tumor — and then advises oncologists about which drugs may work best.

The test is expensive, about $11,500, according­ to a financial analyst’s note from last summer. And physicians have not quickly warmed to it. Only 524 tests were ordered in the third quarter of last year.

NantHealth needs GPS Cancer to do well. The company made its public debut in June, but its stock price has slipped 53 percent since then.

And it’s bleeding money: The company had lost $324 million as of the first quarter of last year; it tacked on a net loss of $54 million in the second quarter of last year, and $37 million more in the third quarter of last year.

The moonshot project is helping Soon-Shiong get the word about GPS Cancer out far and wide and establish it as a cutting-edge tool — even before studies assessing its efficacy have been published. (They’re expected sometime this year.)

Last fall, in what was billed as a moonshot initiative “milestone,’’ Soon-Shiong’s research institute awarded a $20 million grant to several hospitals to use GPS Cancer to sequence brain tumors in 1,600 children.

The press release quoted Soon-Shiong as calling GPS Cancer “an incredibly comprehensive test that provides oncologists with the peace of mind that they are making the most informed decisions before developing a treatment plan.’’

Over the past year, Soon-Shiong’s team has also set up at least a half a dozen working groups to take on different types of cancer. Several participating scientists told STAT they hadn’t done much. But others were optimistic about their progress.

The experience of the German researcher Dr. Peter Fasching offers a window into how the moonshot is spurring academic research — even while it advances Soon-Shiong’s ­business.

Fasching, a women’s health specialist, is ­involved in the working groups for breast cancer and for inflammatory breast cancer, an aggressive and deadly form of the disease. Their in-person and virtual meetings so far, he said, have centered on discussing their experiences with GPS Cancer and planning clinical trials.

Fasching is the lead investigator for an ­early-stage moonshot QUILT trial that will test one of the natural killer cell compounds developed by Soon-Shiong’s NantKwest ­company. He is also collaborating with Soon-Shiong to test GPS Cancer in another clinical trial, sponsored by a German research network.

Fasching said he thinks it’s too early to evaluate the success of the moonshot project, but he likes GPS Cancer and said he’s pleased with how things are going. Among clinical ­development programs, he said, “it’s moving faster than anything I’ve been part of yet.’’

Soon-Shiong is plowing confidently ahead.

He’s due to unveil his next big thing next week at a health technology conference. A press release describes it, in typically breathless Soon-Shiong fashion, as “the world’s first whole genomics transcriptomics supercomputing medical reasoning engine’’ and boasts that it’s “covered by over 100 seminal patents.’’

Its name: the NantDaVinci Engine.

Rebecca Robbins can be reached at rebecca.robbins@statnews.com. Follow her on Twitter @rebeccadrobbins.