An invention to prevent strokes, first sketched out in hospital basements and hotel restaurants in the Twin Cities a decade ago, hit a milestone of 100,000 human implants last month in a product niche racing toward $1 billion in yearly sales.

The Watchman left-atrial appendage closure device, invented at Plymouth startup Atritech and partly manufactured by 150 people at Boston Scientific in Maple Grove today, is designed to prevent strokes in people with atrial fibrillation. The pluglike system permanently seals off an obscure corridor of the heart where blood clots form when the heart is quivering, allowing patients to stop taking blood-thinning drugs.

Boston Scientific executives said the device is available at 600 U.S. hospitals, but they don't disclose pricing or revenue totals. Analysts with Decision Resources Group said U.S. hospitals pay an average $15,800 for the device, not including rebates, and Boston Scientific is on track to generate more than $500 million in U.S. sales this year.

The idea of blocking clots before they can migrate to the brain may sound straightforward today, but back in the early 2000s, it was anything but, the Watchman's inventors said.

The safety of "obliterating" the source of stroke-causing clots had to be proved, and a design using materials and techniques borrowed from stents was sketched out. Regulators applied unprecedented scrutiny. Hundreds of doctors were taught to safely prescribe and implant, and insurers were persuaded the implant could be as cheap and safe as blood thinners. Ads began flickering across TVs.

Back in late October 2003, Mayo Clinic cardiologist Dr. David Holmes, a co-inventor and co-founder of Atritech, wiped a little sweat from his brow after implanting the first Watchman in the U.S., at the Rochester hospital. "The striking thing is that now, in 2019, ... my colleague Dr. [Mohamad] Alkhouli at Mayo put in the 100,000th device that's ever been used," Holmes said in an interview.

Getting a Watchman does carry upfront health risks and costs — most complications are procedure-related, and hospitals were getting about $16,100 from Medicare to implant the nitinol-wire device under 2016 Medicare rates published last year in a study in Stroke.

But the health risks and systemic costs of taking anti-coagulation drugs mount steadily over time, unlike the Watchman, where costs and risks stabilize after a few months. The study in Stroke found that a Watchman implant becomes more cost-effective than drugs five or six years after implant, making it "more effective and less costly" in the long run.

A typical Medicare patient will pay $2,200 to get a Watchman, Boston Scientific said. Average out-of-pocket spending is lower for the Watchman two years after implant than the costs for anti-coagulation drugs, when the costs of health complications are factored in.

The Watchman is a major player in a lucrative market, growing at 19% a year. Energias Market Research estimates the marketwide revenue for left-atrial appendage closure devices will hit $958 million by 2025, though high device costs might hinder that growth, the firm's analysis last May found.

But as the market grows, challengers are rising. A second device used to plug the appendage, Abbott Laboratories' Amplatzer Amulet for LAA closure, is in clinical trials but unlikely to enter the U.S. market before 2021. California's SentreHeart is working on a clinical trial of its Lariat device, to see whether suturing plus ablation can reduce incidents of atrial fibrillation. AtriCure, which acquired SentreHeart in August, also sells the AtriClip device for clamping off the appendage from the outside.

Aggressive growth projections for the left-atrial appendage device market are based largely on demographics.

Atrial fibrillation, in which the heart quivers or beats erratically, is the most common significant heartbeat problem in the world. About 30 million people worldwide have it, and it becomes more common as a person ages, according to a large meta-analysis in Circulation.

Among other health problems, having "Afib" increases a person's risk of stroke fivefold because it disrupts the normal circulatory patterns that prevent blood from clotting, especially inside the natural cul-de-sac connected to the heart's left atrium. The first-line treatment for Afib patients is typically to take anti-coagulation drugs like warfarin or dabigatran, which reduce clotting but also cause bruising and bleeding.

Some patients, like Coon Rapids' William Bonin, have so many bleeding-related problems that blood-thinning drugs aren't a viable long-term solution.

Bonin, 76, had a Watchman implanted at Allina Health's Minneapolis Heart Institute in 2016 after five episodes of serious bleeding in his gastrointestinal tract. "I had one implanted so I could get off of the Coumadin," Bonin said. "It's worked very well." He was put under general anesthesia for the one-hour procedure and stayed overnight in the hospital.

Dr. Mario Goessl, one of the doctors who implanted Bonin's device, said that under existing medical guidelines patients need to have a medical reason to have a Watchman implanted: "You can't just say, I don't like to be on anticoagulants, therefore I want a Watchman." But a prescriber also has to consider that when patients stop taking their drugs, they are left at much greater risk of stroke, he said.

Implanting a Watchman is considered "minimally invasive," because it's implanted in the heart using thin, flexible tools inserted through a small incision in the leg, avoiding the need for a chest incision. The device is folded in the delivery system, and unfurled inside the heart's left atrium.

But minimally invasive doesn't mean risk-free. Federal data show that Watchman implantation is an elective medical procedure that carries a mortality risk. At least 46 death reports for Watchman patients have been reported to the FDA's MAUDE database in 2019.

Boston Scientific executives said such tallies can be misleading, because they don't account for how sick the patients are or how many successful cases have happened. The company cites peer-reviewed studies that put the rate of major complications at 1.5% in real-world patients. A report on its marketing website, www.watchman.com, said procedural complications have declined as physicians gain more experience with the device, while bleeding complications from anti-coagulation drugs have been constant or increasing.

Compared with warfarin (often sold as Coumadin), Watchman is associated with a statistically significant 55% reduction in the rate of fatal or disabling strokes, and 72% relative risk reduction in bleeding events six months after the procedure, peer-reviewed studies have found.

"What happens is, with a lot of these patients, they'll take [anti-coagulation drugs] for four or five years and then they have a major bleeding event. And ... then they live with the risk of the procedure," said Nick Spadea-Anello, general manager of the Watchman franchise for Boston Scientific. "So in some ways, if you are a patient and you understand that you have a chance of bleeding by being on these blood thinners, your risk ... could be significantly more by living with the risk of taking it first, and then living with the risk of the procedure."

Dr. Howard Schumaker, 59, a retired emergency-room doctor who worked at Mayo Clinic Health System hospitals in western Wisconsin, said his atrial fibrillation made him a candidate for anti-coagulation drugs. But his active lifestyle, including regular hiking trips through remote wilderness areas, combined with the effects of a previous surgery that make gastrointestinal bleeding "basically fatal" for him, persuaded him to seek a Watchman.

There was a problem. Left-atrial appendages can vary widely, and medical imaging showed that a Watchman wasn't going to fit perfectly inside Schumaker's LAA. So his doctor, device co-inventor Holmes, got Schumaker into a program a year ago that let him get the next iteration of the device, the Watchman FLX, which is legal in Europe and likely to launch domestically next year.

Like Bonin, Schumaker spent a night in the hospital, and went off blood-thinning drugs after tissue healed over the device. (Guidelines call for patients to take an anti-platelet medication like clopidogrel temporarily after implant.) Although the left-atrial appendage is a natural part of the heart, Schumaker said having it permanently sealed was like having an appendix removed.

"The left-atrial appendage is really kind of a remnant organ, like the appendix," he said. "In my situation, it was the source of significant risk. The risk of continuing to have a left-atrial appendage was too high, from a stroke standpoint."