Kent C. Sasse, MD, MPH, FACS, FASMBS
The stomach has a pacemaker. Most clinicians know this in the abstract, the way we know the heart does — but where the cardiac conduction system has been mapped in exhaustive human detail for decades, almost everything we believed about the human gastric pacemaker was extrapolated from mice, guinea pigs, and dogs.
A study published this year in The Journal of Physiology set out to fix that, and its findings are worth knowing for anyone who sees patients with gastroparesis.
The gap in what we knew
Gastric contractions are driven by electrical events called slow waves, generated by specialized pacemaker cells in the stomach wall known as interstitial cells of Cajal, or ICC. Loss or dysfunction of ICC is the central pathologic finding in gastroparesis.
The problem has been measurement. Human studies have relied almost entirely on extracellular recordings — electrodes pressed against the stomach during surgery, or electrogastrography from the abdominal surface. These techniques tell you roughly how often a wave occurs. They cannot resolve the shape of the wave, they can’t distinguish real electrical signals from movement artifact, and they don’t allow you to apply drugs to work out which ion channels are responsible.
The alternative — impaling individual cells with glass microelectrodes to record transmembrane potentials directly — requires fresh, viable human tissue. That’s the bottleneck this study got past.
What we did
Gastric antral muscle was obtained as surgical waste from 90 patients undergoing sleeve gastrectomy, with IRB approval and consent. Tissue moved to the laboratory within an hour. Investigators made intracellular microelectrode recordings, measured contractile force simultaneously, and used thin vibratome sections to sample at different depths through the stomach wall. A panel of pharmacologic antagonists probed which channels mattered, with multiple compounds tested for each question to guard against off-target effects.
Four findings worth knowing
The human stomach has two pacemaker regions, not one. Near the serosal surface, in longitudinal muscle, slow waves were large, long, and slow — about 3 cycles per minute. Deeper, near the myenteric plexus and through the circular muscle, waves were smaller, shorter, and faster — about 5 cycles per minute. In intact muscle the faster pacemaker dominated, which is the same principle by which the sinoatrial node dominates in the heart. This dual-pacemaker arrangement had not previously been described in humans.
The upstroke is largely T-type calcium. Blocking T-type channels sharply reduced slow-wave amplitude, rate of rise, and frequency. L-type channels contributed too — and notably, both CaV1.2 and CaV1.3 appeared to participate. A role for CaV1.3 in gastric pacemaking hadn’t previously been established.
The plateau is calcium-induced calcium release driving ANO1. Calcium entering during the upstroke triggers further calcium release from endoplasmic reticulum stores through both IP3 and ryanodine receptors. That released calcium opens ANO1, a calcium-activated chloride channel found exclusively in ICC within the GI tract, which sustains the plateau phase. Blocking ANO1 with either of two selective inhibitors abolished slow waves.
Sustaining the plateau requires ongoing calcium entry. T-type channels inactivate within milliseconds, yet the plateau lasts seconds. The data implicate the L-type “window current” and, interestingly, sodium-calcium exchange running in reverse mode. Store-operated calcium entry through ORAI channels refills the depleted stores.
Slow waves coupled 1:1 with phasic contractions throughout — confirming that these electrical events are what actually drive gastric peristalsis.
Why a clinician should care
Three reasons.
It gives us drug targets in human tissue. ANO1 expression is known to be altered in human diabetic gastroparesis. This study demonstrates that ANO1, T-type calcium channels, and intracellular calcium handling are each load-bearing in the human stomach specifically — not just in mouse. That is the difference between a plausible target and a validated one.
Animal models were partly misleading. In canine stomach, L-type blockade with nifedipine substantially shortens slow waves. In human tissue, nifedipine had only modest effects. If we’d designed a human drug program on the canine data, we’d have aimed at the wrong channel.
You can’t classify abnormal without knowing normal. This is the argument the authors make directly, and it explains a good deal of why gastroparesis therapeutics have disappointed. Gastric electrical stimulation, for instance, remains a humanitarian-use device with increasingly promising published results — developed largely in the absence of the kind of mechanistic human data this study provides.
Caveats
The tissue came from patients undergoing sleeve gastrectomy — non-diabetic, relatively young, with obesity — so these are healthy stomachs, not gastroparetic ones. Recordings were made in vitro after a period of ischemia during tissue transport, and in vitro slow-wave frequencies run higher than those recorded in living patients, for reasons that remain unexplained. Pharmacologic agents have off-target effects, which the authors mitigated but could not eliminate. And there was substantial patient-to-patient variability in slow-wave frequency that no one can yet account for.
The bottom line
Nothing here changes what you do in clinic tomorrow. But the reason gastroparesis has so few effective therapies is that the disease was being targeted without a detailed map of the normal human system it disrupts. This is a meaningful piece of that map — and mapping the normal is how the treatable abnormal eventually gets defined.
Reference: Hwang SJ, McErlain T, Seaton-Kelly R, Blair PJ, Hegarty RC, Kim M, Baker S, Sasse KC, Sanders KM, Ward SM. Mechanisms responsible for pacemaker activity in human gastric muscles. J Physiol. 2026. doi:10.1113/JP289300
Dr. Sasse is a board-certified general surgeon in Reno, Nevada and a co-author of the study described above.