Mechanical theory: stiffness, pressure and force as causes

The claim. Physical forces are causal inputs to cancer, not by-products. Tissue stiffness, matrix crosslinking, compression and interstitial pressure are sensed through integrins, focal adhesion kinase, the cytoskeleton and the Hippo-YAP/TAZ pathway, and they drive proliferation, invasion, stemness and immune exclusion. Conversely, restoring normal tensional homeostasis can normalise malignant cells. At the tumour scale, solid stress and high fluid pressure collapse vessels, causing hypoxia, poor drug delivery and selection for aggressive clones.

Who and when. Bissell's dynamic reciprocity between cells and matrix (1982) and Weaver's 1997 reversion of malignant breast cells by blocking beta-1 integrin. Paszek, Weaver and colleagues, Tensional homeostasis and the malignant phenotype (2005), showed that matrix stiffness alone drives a malignant phenotype in culture. Levental and colleagues (2009) showed that collagen crosslinking by lysyl oxidase forces tumour progression in mice. Jain's laboratory measured solid stress and interstitial pressure in tumours and proposed vascular normalisation; Nia, Munn and Jain summarised the physical traits of cancer in 2020.

Evidence for. Stiff matrices induce malignant behaviour in three-dimensional culture and soft ones revert it. Mammographic breast density, a measure of stromal content, is a strong risk factor for breast cancer. Fibrotic organs (cirrhotic liver, fibrotic lung) are cancer-prone. Desmoplastic pancreatic tumours have high solid stress, collapsed vessels and hypoxia, and decompressing them improves perfusion in animals. YAP/TAZ mechanotransduction is required for many tumours. Compression from a growing tumour suppresses surrounding cells and selects for invasive ones.

Evidence against and limits. Force is entangled with biochemistry: stiff matrix also concentrates growth factors and changes cell shape, so isolating a purely mechanical cause is difficult. Depleting stroma can make pancreatic cancer worse in mice, so the matrix restrains as well as promotes. Clinical translation is thin: pegvorhyaluronidase, which degrades hyaluronan to lower pressure, failed in a phase 3 pancreatic cancer trial in 2020, and the lysyl oxidase-like 2 antibody simtuzumab failed in phase 2.

Predictions that held or failed. Held: dense stroma predicts poor drug delivery and hypoxia; anti-VEGF therapy transiently normalises vessels and improves delivery; losartan, which reduces collagen and hyaluronan, raised resection rates in a single-arm pancreatic cancer trial (Murphy 2019). Failed: stromal depletion as a strategy; hyaluronidase in phase 3.

Therapies that came from it. Vascular normalisation as a rationale for combining anti-angiogenics with chemotherapy or immunotherapy, angiotensin blockade and focal adhesion kinase inhibitors in trials as mechanotherapeutics, and stiffness-based imaging (elastography) in diagnosis. It supplies the mechanism behind the tissue organisation field theory's experiments and is part of the microenvironment view; Jain's solid stress models belong to mathematical oncology.

Status: partly confirmed. Mechanical signals are established as drivers of malignant behaviour in the laboratory and as prognostic factors in patients; drugs that act through mechanics alone have not yet succeeded in a randomised trial.