Finding. Isoflurane, at clinically relevant concentrations, depolarizes mitochondria in helper T cells, raises ROS, damages DNA, and activates caspase-3/7 through p38 MAPK. The same cells under propofol keep mitochondrial integrity, run less oxidative stress, and proliferate better. Caspase-3/7 blockade, N-acetyl cysteine, and p38 inhibition each rescue the isoflurane death program. The injury is transient: it recedes after surgery in patients and within 24–48 hours of washout in Jurkat cells.
Why this paper matters
Onco-anesthesia arguments usually stall at retrospective survival curves and vibe-level immune claims. Saha, Chakrabarti, Das and Sinha go after a mechanism in the cell type that actually organizes adaptive immunity, using both perioperative CD4+ Th cells from women with breast cancer (n=15 per anesthetic) and a Jurkat mechanistic bed.
The mitochondrial claim is the entry point, not a side stain. Isoflurane opens a depolarization–ROS–DNA-damage–caspase sequence. Propofol does not. If you care about how a volatile anesthetic talks to lymphocyte mitochondria, this is a usable axis: ROS–p38–caspase-3/7, closed with three independent inhibitors.
What closes the axis
Patient-derived CD4 cells and Jurkat cells agree on the isoflurane phenotype: mitochondrial depolarization, ROS, DNA damage, caspase-3/7. Z-DEVD-FMK and NAC rescue apoptosis, which puts caspase execution and oxidant load on the required path. Isoflurane, more than propofol, activates p38 MAPK; SB203580 suppresses the apoptosis. That is a real signaling loop, not a biomarker stack.
Propofol's side of the comparison is conservation: mitochondrial integrity held, oxidative stress lower, proliferative capacity higher. The abstract does not prove propofol is actively mitoprotective — only that it is not isoflurane.
Reversibility is the result most likely to be dropped in tweets and the one that should not be. Patients recover postoperatively. Jurkat cells recover 24–48 hours after the drug is removed. This is a hit-and-recover mitochondrial apoptotic pulse, not a permanent deletion of helper T cells.
How to read the score
Clear mitochondrial apoptosis biology in a clinically used drug pair, with human cells and pharmacologic epistasis. Confidence is medium: modest n, one cancer type, one sex, no oncologic endpoint, Jurkat as the mechanistic workhorse, and no other volatile compared. Do not read a recurrence-risk paper into a signaling paper.
What to do with it
If you model perioperative immunity, give isoflurane a reversible mitochondrial-ROS–p38–caspase edge on CD4 T cells and give propofol a preserve-ΔΨm edge. If you replicate, keep the three rescues and the washout time course; drop either and the axis opens again. If you want clinical importance, this preprint does not supply it — you still need infection, vaccine, or recurrence endpoints. The honest use is mechanistic: the mitochondrion is where these two anesthetics first disagree in helper T cells.