Fibrosis

Fibrosis is what wound healing becomes when it does not stop. The programme that is adaptive over days turns pathological when injury persists or resolution fails: myofibroblasts survive rather than apoptose, matrix accumulates and becomes cross-linked, and cross-linked collagen resists the metalloproteinases that would clear it, so the process ratchets. TGF-beta1 is the central mediator, amplified by mechanical feedback, since a stiffer matrix activates latent TGF-beta and drives further myofibroblast differentiation.

The drug record sets realistic expectations. Pirfenidone and nintedanib, both approved in 2014 for idiopathic pulmonary fibrosis, slow the annual decline in forced vital capacity by roughly half. Neither reverses established fibrosis, and the pivotal trials were powered on lung function rather than survival. In the liver, fibrosis stage is the strongest histological predictor of clinical outcomes in metabolic dysfunction-associated steatohepatitis, and resmetirom received accelerated approval in 2024 on fibrosis improvement seen on biopsy.

This changes how to read any anti-fibrotic claim. Slowing progression and reversing fibrosis are separate assertions requiring different evidence, and only paired biopsies or a validated imaging equivalent can tell them apart.

The reproducibility trap is the bleomycin mouse lung model. A very large number of compounds look anti-fibrotic in it and essentially none of that has translated. Much of the problem is timing: dosing in the first week suppresses the inflammatory response to bleomycin and pre-empts fibrosis, a different claim from treating fibrosis already established at two weeks. Preclinical anti-fibrotic claims for peptides almost invariably use the prophylactic schedule, and none has been tested against a fibrosis endpoint in a controlled human trial.