Synchron Has 72 Patient-Months of Zero Serious Adverse Events. Neuralink Has a Thread Retraction Problem.

Seventy-two. That is the number of patient-implant-months Synchron has accumulated in the COMMAND trial (NCT05035823) without a single device-related serious adverse event leading to death or permanent disability. Six patients. Twelve months each. Zero brain hemorrhages. Zero vascular injuries. Zero infections requiring explant. For a device threaded through a blood vessel into the motor cortex, that safety record is the single most important data point in the entire brain-computer interface industry right now.

Neuralink, by comparison, has implanted 12 patients as of September 2025, but only two were enrolled in the formal PRIME trial under FDA oversight. The first patient, Noland Arbaugh, experienced 85% thread retraction within weeks of implantation. Brain pulsation pushed the ultra-thin polymer threads away from cortical tissue, degrading signal quality until a software workaround partially restored function. Neuralink changed its surgical technique for later patients, but the thread retraction risk had been flagged during animal testing before the first human ever received one.

Two Philosophies of Getting Inside the Brain

Neuralink drills a quarter-sized hole in the skull. A surgical robot called R1 inserts 1,024 electrodes on 64 flexible threads directly into cortical tissue. The approach maximizes electrode count and, theoretically, signal bandwidth. It also requires a neurosurgeon, an operating room, general anesthesia, and the acceptance of a craniotomy's inherent risks: infection, bleeding, brain swelling.

Synchron avoids the skull entirely. Its Stentrode device is a stent fitted with electrodes, delivered through the jugular vein and navigated into a blood vessel sitting on top of the motor cortex. The procedure looks closer to a cardiac catheterization than brain surgery. In the COMMAND trial, median deployment time across all six patients was 20 minutes. Every device landed on target with full motor cortex coverage.

The tradeoff is bandwidth. Synchron's electrodes sit outside the brain, separated from neurons by a vessel wall. Signal resolution is lower. The first-generation Stentrode supported just two command outputs: a foot-lift intent and a hand-movement intent. That is enough to click, scroll, and navigate a computer interface, but it is orders of magnitude fewer channels than Neuralink's architecture promises.

A Metric Nobody Is Tracking: Patient-Implant-Months

The FDA does not clear brain-computer interfaces based on YouTube demos or Twitter posts. It clears them based on accumulated safety and efficacy data, measured in patient-implant-months: the number of patients multiplied by the number of months each has had the device.

Here is where the two companies stand:

Company	Trial	Patients in Formal FDA Trial	Months per Patient	Patient-Implant-Months	Device-Related SAEs
Synchron	COMMAND (US)	6	12	72	0
Synchron	SWITCH (Australia)	4+	Up to 72	~150+	0 reported
Neuralink	PRIME (US)	2	~18	~36	Thread retraction (patient 1)
Neuralink	Broader implants	10 (non-PRIME)	Varies	Unknown (unpublished)	Unknown

Synchron's combined patient-implant-months across COMMAND and the earlier Australian SWITCH trial exceeds 220. One SWITCH patient, 65-year-old Rodney Gorham, has been living with a first-generation Stentrode since 2020. That is over six years of continuous implantation with zero device-related complications. He uses two thought commands to text, browse the web, and check football scores. When asked if he would do it again, he nodded several times.

For FDA reviewers assembling a risk-benefit calculus, patient-implant-months with zero SAEs carry enormous weight. Synchron's 72 months in COMMAND alone represent a clean feasibility dataset. Neuralink's 36 months in PRIME include a significant device-related event: the 85% thread retraction in patient one required a software workaround and prompted a change in surgical technique for subsequent implantations. Whether FDA classified this as a formal SAE is not public, but the functional degradation and procedural change are the kind of signals reviewers weigh heavily.

The Regulatory Chessboard

Synchron received FDA Breakthrough Device designation in August 2020, placing it in a program that provides dedicated FDA staff interaction, priority review, and a collaborative pathway through regulatory clearance. Synchron has completed its IDE feasibility study (COMMAND) and is positioning for a pivotal trial.

Neuralink has not publicly confirmed a Breakthrough Device designation. It is still in the first-in-human phase. A pivotal trial is likely years away.

Both companies will probably seek De Novo classification, since permanently implanted motor BCIs have no predicate device. But Synchron enters that process with a completed feasibility trial and a clean safety record. Neuralink enters it with a smaller formal dataset and a known hardware failure mode.

The Next-Gen Leap: FOCUS-AU

On March 25, 2026, Synchron announced FOCUS-AU, the first clinical trial of its next-generation Stentrode. Ten patients with motor neurone disease will be enrolled across Royal Melbourne Hospital, Alfred Hospital in Melbourne, and Liverpool Hospital in Sydney. Surgeries begin in April 2026.

The upgrade is substantial. The new device supports 16 command outputs, up from two, and eliminates the external transmitter module that first-generation patients wore on their chest. Led by Prof. Peter Mitchell, the same interventional neuroradiologist who performed the first Stentrode implant in 2020, FOCUS-AU will test whether the endovascular approach can close the bandwidth gap with cortical implants.

Sixteen commands opens a new design space. Instead of binary click-scroll, patients could potentially access multi-directional cursor control, app switching, text selection, and device-specific gestures. It is still far fewer channels than Neuralink's 1,024 electrodes, but the question is whether 16 well-placed, reliably captured commands outperform 1,024 electrodes that occasionally detach from the tissue they are reading.

Apple Changed the Game Quietly

In May 2025, Apple added a Brain-Computer Interface Human Interface Device (BCI HID) protocol to its accessibility stack. In August 2025, Synchron demonstrated thought-controlled navigation of an iPad through the protocol, making Synchron the first BCI to be recognized as a native input device alongside touch, keyboard, and voice.

That integration extends to iPhone, iPad, and Apple Vision Pro. For a patient with motor neurone disease, native OS-level input recognition means the BCI works everywhere, in every app, without custom software. It transforms the Stentrode from a research device into a consumer accessibility product.

Synchron has raised $345 million in total funding, including a $200 million Series D in November 2025. Investors include Jeff Bezos and Bill Gates. Australia's National Reconstruction Fund contributed AUD $54 million in equity. Total BCI market projections reach $4.66 billion by 2030 at a 14.4% CAGR.

Strongest Counterargument

Bandwidth matters, and Neuralink has more of it. The 1,024-electrode Utah-array-inspired architecture can resolve individual neuron firing patterns. Synchron's endovascular electrodes record aggregate field potentials through a vessel wall, analogous to holding a microphone outside a concert hall versus sitting in the front row. For applications beyond simple computer control (restoring hand dexterity, enabling speech synthesis, or achieving high-speed typing), cortical resolution may prove essential. Neuralink's first patient typed at rates exceeding those of any endovascular BCI patient. If the thread retraction problem is solved through better materials or surgical technique, Neuralink's ceiling is significantly higher.

Limitations

This analysis has important blind spots. Patient-implant-months is a crude proxy that does not weight for condition severity, procedure complexity, or patient demographics. COMMAND enrolled six patients, a sample too small for population-level safety conclusions. Zero SAEs in six people is encouraging but not statistically robust. Neuralink's broader 12-patient cohort includes patients outside formal FDA trials, meaning their outcomes are not publicly documented in peer-reviewed literature. Direct signal-quality comparisons between endovascular and cortical BCI approaches do not exist in controlled settings. Synchron's patient-implant-month figures for the SWITCH trial are estimated from public reports, not official disclosures, and may undercount or overcount actual follow-up time. The 85% thread retraction figure for Neuralink's first patient comes from media reports citing company disclosures, not a peer-reviewed publication. Regulatory timelines are inherently uncertain, and FDA can accelerate reviews, delay them, or impose additional requirements at any stage.

The Bottom Line

The BCI company closest to FDA clearance is not the one with the most Twitter followers. Synchron has 72 patient-implant-months of clean safety data in a completed feasibility trial, a Breakthrough Device designation, Apple integration, $345 million in funding, and a next-generation device entering clinical trials this month. Neuralink has higher bandwidth, more funding ($650 million in June 2025), and far more public attention. But regulatory clearance is not a popularity contest. It is a data contest. Right now, Synchron has more of the data that matters.