Finding. A celecoxib plus nordihydroguaiaretic acid combination kills two genetically different colorectal lines with mitochondrial reactive oxygen species, lipid peroxidation, and labile ferrous iron, and then refuses to be named apoptosis, autophagy, necroptosis, or ferroptosis. Yadala, Palakurthi and Pawar use HCT116 (KRAS) and HT29 (BRAF, different p53). Death does not care about COX-2/5-LOX or apoptotic competence. NAC saves the cells. The usual death-module inhibitors do not. Imaging shows JC-1 green mitochondrial puncta and FerroOrange iron puncta. Mechanistically, p62 is phosphorylated at Ser349, KEAP1 falls, NRF2 stabilizes, and that happens without ULK1 autophagy and without HO-1 coming on.
Why this paper matters
Colorectal heterogeneity is the excuse for every failed targeted drug. A redox death that ignores KRAS versus BRAF and apoptotic competence is worth a look if it is truly not ferroptosis-by-another-name. The mitochondrial pictures are why it is on this desk. The naming fight (non-canonical, autophagy-independent, ferroptosis-inhibitor-resistant) is why you should keep the inhibitor panel in the citation and not flatten it to "they ferroptosed."
What they actually measured
Two lines, combination cytotoxicity, inhibitor panel, NAC, ROS/lipid/iron imaging, p62/KEAP1/NRF2/HO-1. No in vivo arm in the abstract.
How to read the score
Low 80s. Mitochondrial ROS/iron plus a genotype-agnostic claim. Confidence is medium because inhibitor-negative ferroptosis is a fragile category.
Caveats
No mouse. Combination of two old molecules. NRF2 without HO-1 needs the blots. Do not start a celecoxib-NDGA trial from this brief.
What to do with it
If you screen redox deaths in CRC, include this pair and keep ferroptosis inhibitors in the counter-screen. Pull p62-S349. Ask where the iron puncta sit relative to mitochondria.