Convergence Watch: The Bubble Cloud Quietly Doing Immunotherapy’s Job

In May 2019, a 67-year-old man with stage IV colon cancer that had spread throughout his liver lay on a table at the University of Michigan while a transducer outside his body fired ultrasound pulses into a single 1.2-centimeter metastasis. The pulses arrived focused, intense, and so short that they passed below the threshold of heating. At the focal point, dissolved gas in the tissue nucleated into a cloud of bubbles, the bubbles collapsed against the tumor cells, and the targeted lesion liquefied. He went home.

When his physicians scanned him a week later, the treated lesion had shrunk, as expected. So had the lesions they had not treated. Eight weeks on, the untreated metastases were still shrinking. His blood marker for colon cancer was falling. One year later, he was alive with slow disease. The same patient had received thermal high-intensity focused ultrasound — heat-based ablation — on other tumors months earlier. That treatment killed what it cooked. Nothing else moved. The bubble cloud apparently did something the heat could not. The case is documented by the Focused Ultrasound Foundation; it is the cleanest single-patient evidence that histotripsy, the official name for the bubble-cloud technique, can trigger an immune response against tumors a transducer never touched.

That observation is now the most interesting thing happening at the intersection of medical devices and oncology drugs. Two industries that share almost no engineering DNA — focused ultrasound transducers built by a University of Michigan spin-out, and checkpoint-inhibitor antibodies built by Merck and Bristol-Myers Squibb — are being yoked together by a single physics result.

The patent record makes the shift visible

A search of US utility grants for the word “histotripsy” returns 45 patents through May 5, 2026, with the curve sharply asymmetric: 22 of those 45 grants — nearly half the corpus — were issued in the last 16 months. Issuances ran twelve patents in 2025 and six more by mid-May 2026, after sitting at one to five per year through 2020. The institutional concentration is similarly lopsided. The University of Michigan, where the technique was invented in the early 2000s by Charles Cain and Zhen Xu, holds 17 grants on histotripsy methods and systems. HistoSonics, the Michigan-founded company that licenses the work, holds nine. The University of Washington holds four. Philips and Siemens Healthineers each have a few, as does a single Veterans Affairs grant on thrombolysis. It is, at the patent level, a one-university story with one corporate vehicle and a small cohort of imaging-system partners.

Read the recent grants and the platform expansion becomes concrete. US patent 12,220,602, issued to Michigan in February 2025, describes histotripsy for brain tumors — a problem focused ultrasound has spent two decades trying to solve, because the skull is a chaotic acoustic lens that smears any beam aimed through it. Michigan’s solution is hybrid: a transducer array sits outside the skull, but a thin drainage catheter, threaded into the target region, carries a piezoelectric sensor at its tip. The sensor measures the pulses arriving on the inside, an aberration-correction algorithm calculates the skull’s distortion, and the next pulses fire with that distortion subtracted out. The catheter then drains the liquefied tumor. The destructive event is noninvasive; only the listening device and the drain are inside. US 12,303,152, issued May 2025, applies the same cavitation cloud to thrombolysis — mechanically pulverizing blood clots without thrombolytic drugs, the same trick Boston Scientific’s Farapulse system is using electrically for atrial fibrillation. US 12,551,222, from February 2026, describes a headset that mounts to the cranium and runs the procedure with no operator hand on a probe.

The business has stopped being quiet

On August 8, 2025, a consortium led by K5 Global with Bezos Expeditions and Wellington Management took majority control of HistoSonics at a $2.25 billion valuation, the company’s BME-school announcement and the Focused Ultrasound Foundation both confirmed. Two months later, on October 16, the company closed an oversubscribed $250 million growth round at a $3 billion valuation, MedTech Dive and Endovascular Today reported. CEO Mike Blue told Bloomberg the company would book over $100 million in revenue in 2025 and projects $200 million for 2026. For a Michigan engineering spin-out that received FDA clearance for its first indication — primary and metastatic liver tumors — only in October 2023, that is a notably steep curve. The original HOPE4LIVER pivotal trial, published in Radiology in 2024, ran as parallel single-arm studies across 14 US and European sites. A real-world international safety analysis published in the Journal of Gastrointestinal Surgery in 2025 covered 230 cases and reported a complication profile that compared favorably with surgical and thermal liver-directed therapies.

The platform is multiplying faster than the FDA file

ClinicalTrials.gov lists 14 histotripsy studies; 12 of them are recruiting, not-yet-recruiting, or active. The list reads like a tour of high-margin oncology and minimally invasive surgery:

• NCT06282809 — pancreatic adenocarcinoma, HistoSonics’s GANNON trial, opened December 2024, enrolling 50 patients. Pancreatic cancer is the FDA Breakthrough Device indication HistoSonics secured in 2024; it is the cancer with the worst five-year survival in the developed world.
• NCT07214675 — benign prostatic hyperplasia, opened March 2026, enrolling 80 patients. BPH is a $5 billion device market today divided among laser, electrocautery, and water-jet incumbents.
• NCT05820087 — primary solid renal tumors (HOPE4KIDNEY), active, 67 patients.
• NCT06524570 — Phase I, histotripsy combined with checkpoint-inhibitor immunotherapy. This is the trial that turns the abscopal-effect mouse data into a human signal. Twenty-four patients.
• NCT07044362 — colorectal liver metastases, histotripsy plus chemotherapy versus chemotherapy alone, 100 patients.
• NCT06697834 — plantar fasciitis, completed. (Yes, foot pain. The same physics works on inflamed tendons.)

The most consequential of these is the checkpoint-inhibitor combination. Thermal ablation has been around for decades — radiofrequency, microwave, conventional HIFU — and despite repeated attempts to show that it primes systemic immunity, the answer has mostly been no. Heat denatures proteins. The tumor antigens that an immune system would need to recognize are cooked into unrecognizability before the immune system arrives. Histotripsy does not heat the tissue; cavitation tears cell membranes mechanically. The intact tumor antigens come out into circulation, antigen-presenting cells find them, and downstream T-cell priming follows. The 2020 paper in the Journal for ImmunoTherapy of Cancer (200 citations and still climbing) showed this in mice; subsequent work has reproduced the finding in breast and pancreatic models. The Michigan group’s 2024 Annual Reviews of Biomedical Engineering primer states the immunostimulation property as a defining feature of the technique, on par with the ability to image and target in real time.

Why an R&D director should care

For an oncology pharma scout, this collapses a calculation that has been driving combination-trial spend for five years. Checkpoint inhibitors work spectacularly in tumors that are inflamed and antigen-rich, and barely at all in tumors that are “cold.” Pharma’s working answer has been to make the cold ones hotter with intratumoral injections, oncolytic viruses, radiation, and a long list of mechanisms whose manufacturing is a nightmare. A focused ultrasound device that converts any visible tumor into an in-situ antigen depot, in under an hour, with no incision, would be the cheapest combination partner ever offered to a checkpoint program. The trial is small. The mechanism is plausible. The exit math, for a company already at $3 billion on liver alone, is straightforward.

For a device strategist, the more uncomfortable observation is that the patent backbone for the entire field sits at one Big Ten university and one Wisconsin-headquartered company. Philips and Siemens are imaging partners, not therapy owners. Medtronic, Boston Scientific, Johnson & Johnson — the ablation incumbents whose pulsed-field, microwave, and radiofrequency portfolios dominate the global ablation market — appear in the histotripsy patent record only as cited prior art. If histotripsy proves out as an immune-priming platform, the largest ablation businesses in medicine are looking at a noninvasive competitor that they do not own and cannot easily acquire after a 2025 take-private at $2.25 billion and a follow-on at $3 billion.

The bubble cloud was invented to ablate. It is being used to liquefy liver tumors today. The interesting question is whether, by the time the next checkpoint-inhibitor patent cliff arrives, it will also be the thing that finally taught the immune system to see them.

Method note. Patent counts were drawn from 9.3M US utility grants in the USPTO public dataset (full-text indexed locally) by filtering on the word “histotripsy” across title, abstract, and claims, with publication dates through May 5, 2026. Assignee tallies combine variant spellings (“HistoSonics, Inc.” and “HISTOSONICS, INC.”; the two casings of “The Regents of the University of Michigan”) and reflect only US-issued grants — international filings and continuations not yet granted are excluded. Trial counts come from a ClinicalTrials.gov mirror queried on May 12, 2026 for studies whose brief or official title contains “histotripsy.” Literature counts come from a local OpenAlex mirror indexed through early 2026; the abscopal-effect case report and the HOPE4LIVER pivotal trial were verified against the original publications. Funding figures come from contemporaneous coverage by MedTech Dive, Endovascular Today, Crain’s Detroit Business, and the University of Michigan College of Engineering.