Key Takeaway

IceCure Medical said on August 20 that the institutional review board at West Cancer Center & Research Institute in Germantown, Tennessee, approved the site to run ChoICE, a 400-patient post-marketing study of ProSense cryoablation with adjuvant endocrine therapy in women aged 70 and older. FDA granted the system marketing authorization on October 3, 2025 through the De Novo route, which created a device type where none existed. The company expects its first patients to enroll within weeks.

At a Glance
  • Announced: August 20, 2026 by IceCure Medical (Nasdaq: ICCM) of Caesarea, Israel.
  • Study record: ChoICE carries the identifier NCT07629206 and the sponsor protocol number ICM-BC-003, with 400 patients planned and a July 17, 2026 start date.
  • Primary endpoint: the rate of confirmed ipsilateral breast tumor recurrence out to five years, with primary completion estimated for July 2034.
  • Site lead: Richard Gilmore, a breast surgical oncologist at the Margaret West Comprehensive Breast Center.
  • Regulatory file: De Novo DEN220077, received October 19, 2022 and granted October 3, 2025, now product code QXW under 21 CFR 878.4355.
  • Prior evidence: the ICE3 trial (NCT02200705) enrolled 206 patients and produced a 3.1 percent local recurrence rate.

Thirty Centers That Enroll and Sell at the Same Time

Under the approved design, 30 participating sites will act as ChoICE centers and as active commercial sites at once, taking the 400 study patients while offering ProSense cryoablation to appropriate patients who stay outside the trial. IceCure calls that a dual pathway and attributes it to rising interest from eligible patients in a minimally invasive option. The registry entry lists two locations so far: Adventist Health Glendale in California, and The West Clinic, doing business as West Cancer Center, in Germantown.

Two versions of the recurrence endpoint will be reported, one counting only confirmed ipsilateral breast tumor recurrence and one counting confirmed plus suspected events. Beyond the five-year window the sponsor plans to link participants to claims data, pulling breast cancer surgeries, mammograms and other breast imaging, breast biopsies and all-cause mortality out of administrative records rather than study visits.

What the Enrollment Criteria Ask of the Pathology Report

The registry entry sets the bar at invasive ductal carcinoma diagnosed by core needle biopsy in a woman aged 70 or older, unifocal, no larger than 1.5 cm in greatest diameter, estrogen receptor positive, progesterone receptor positive and HER2 negative. Low-risk biology is defined as Ki67 below 15 percent or confirmed by genomic testing, and the protocol states that where Ki67 reaches 15 percent or more, the genomic result prevails. The lesion has to be sonographically visible at the time of treatment, and the axilla has to be clinically negative.

The exclusion list is where the histologic call decides the matter, beginning with lobular carcinoma. Also barred are microinvasion, invasive carcinoma with an extensive intraductal component defined as a DCIS fraction of 25 percent or more, lymphovascular invasion, multifocal or multicentric disease, multifocal suspicious calcifications, inflammatory features, and any prior radiation to the same breast.

The October 2025 Authorization and Its Written Limits

FDA received the De Novo request on October 19, 2022 and granted it on October 3, 2025, close to three years later. The grant established 21 CFR 878.4355 and product code QXW, a class II device type reviewed by the general and plastic surgery panel. The classification text describes the intended patient as one with "a well-defined solid tumor (e.g., invasive ductal carcinoma)" carrying a low risk of recurrence and metastasis, and states plainly that the device is not intended for breast cancers with a high risk of local recurrence such as lobular carcinoma, for aggressive forms such as those with inflammatory features, or for patients in whom targeted systemic therapy is indicated.

IceCure's announcement uses two different regulatory words for the same file, calling ProSense the first device granted FDA marketing authorization in this setting in one paragraph and describing it as cleared in another. The record shows a De Novo grant, which is a different route from 510(k) clearance and carries no predicate device.

Why This Matters to the APO|APE Reader

ChoICE carries a secondary endpoint asking each site to document complete primary tumor ablation twice, first as adequate iceball coverage of the target lesion during the procedure and again as absence of residual tumor at the six-month visit. The registry entry does not say whether that second confirmation rests on imaging, on a repeat core biopsy, or on both, and the answer decides how much of the work lands in radiology and how much lands in the laboratory at all 30 sites. Cosmetic outcome gets its own instrument, the Sneeuw questionnaire, with items scored from 1 for none to 4 for severe and an overall rating running from poor to excellent. Primary completion is set for July 2034, so the recurrence figure that would settle the question sits eight years out.