Pancreatic cancer provides testbed for first mechanotherapeutics

Pancreatic cancer treatments are forging the path for a new line of agents that target the physical properties of a tumor and its surrounding microenvironment.
Dublin, Ireland

Search for this author in:

In pancreatic cancer, here shown, the connective tissue around the tumor thickens and scars. Cancerous cells are red; nuclei are blue and the stroma is cyan.Credit: Neelima Shah and Edna Cukierman, Fox Chase Cancer Center, National Cancer Institute, National Institutes Of Health / Science Photo Library

In recent weeks, FibroGen initiated a phase 3 trial of pamrevlumab in patients with locally advanced, unresectable pancreatic cancer. Pamrevlumab is a monoclonal antibody that targets connective tissue growth factor, thereby reducing the fibrotic tissue and making unresectable tumors amenable to surgical excision. It is one of several novel drugs in clinical development in the emerging area of mechanotherapies for neoplasms. Others are looking to repurpose existing agents originally developed for other conditions. A recently reported open-label phase 2 trial based on preclinical research led by Rakesh Jain of Massachusetts General Hospital (MGH) and Harvard Medical School suggests that the generic angiotensin system inhibitor losartan, in combination with chemotherapy, boosts survival in unresectable, locally advanced pancreatic ductal adenocarcinoma.

As the role of physical forces in cancer is becoming clearer, trials such as these are spurring increased interest. “Cells are mechanical entities; they have physical properties. Just like a table or chair, they can be soft or stiff,” says Kandice Tanner, who studies how mechanobiology influences cancer at the National Institutes of Health’s National Cancer Institute.

But the approach remains largely uncharted territory for drug makers. “The idea of targeting mechanics is so new, compared with looking for genes or molecular biomarkers,” says Cynthia Reinhart-King, professor of biomedical engineering at Vanderbilt University. Scientists are still relying on the traditional molecular biology approaches. “It’s hard to target a mechanical property. You want to target the molecules that alter or interpret the mechanical properties,” she says.

The density of breast tissue, for example, is recognized as one of the strongest risk factors for developing breast cancer, even if the underlying biology is incompletely understood. Pioneering studies in the laboratory of Mina Bissell at Lawrence Berkeley National Laboratory established the central role of the extracellular matrix (ECM) and microenvironment in the initiation, promotion and progression of breast cancer. More widely, Bissell has long championed the idea that “tumorigenicity is context-dependent”—and that the genetic changes that accompany the development of a tumor arise from alterations or disruptions to a transformed cell’s normal physiological context.

A tumor’s surroundings have thus emerged as a key research focus in oncology. The ECM, a three-dimensional complex of collagen, fibronectin and other macromolecules, is important not only for the structural support it lends to tissues but also for its many roles in providing environmental cues to cells. “Drugs that target the extracellular matrix are exciting because you don’t have to get them across the cell membrane,” says Donald Ingber, founding director of the Wyss Institute for Biologically Inspired Engineering at Harvard University, who is another pioneer of mechanobiology research. In cancer, ECM stiffening results from a complex interplay of cellular and molecular processes, including remodeling of stroma by cancer-associated fibroblasts, disruptions to ECM homeostasis and increased crosslinking between the protein components of the ECM mesh. Enzymes such as lysyl oxidases and transglutaminases, which catalyze ECM crosslinking, also contribute to ECM stiffening. So too do advanced glycation end products, a heterogeneous category of macromolecules that arise from uncontrolled glycation of proteins, lipids and nucleic acids. “They tend to build up with age, ironically,” Reinhart-King notes.

Conditions such as hypoxia can promote overproduction of ECM components, as can transforming growth factor-β (TGF-β), which has long been a target of intense interest in cancer (Table 1). The attention arises primarily from its immunosuppressive properties within the tumor microenvironment, but this multifaceted cytokine also promotes tumor progression by stromal modification, angiogenesis and induction of epithelial–mesenchymal transition, as described in a publication last year by scientists at the EMD Serono Research and Development Institute arm of Merck KGaA, of Darmstadt, Germany. Merck’s bifunctional fusion protein bintrafusp alfa, which comprises the extracellular domain of TGF-β receptor type 2 covalently linked to the C terminus of the Fc domain of a monoclonal antibody targeting programmed cell death ligand-1 (PD-L1), may therefore operate on multiple levels.

Table 1 | Selected novel cancer drugs in development with possible mechanobiology effects








Connective tissue growth factor

Pancreatic cancer, idiopathic pulmonary fibrosis, Duchenne muscular dystrophy

Phase 3


SFJ Pharmaceuticals, Merck KGaA (Darmstadt, Germany)


Colorectal cancer

Phase 2


Verastem Oncology


Pancreatic cancer

Phase 2




Pancreatic cancer

Phase 2


Eli Lilly


Solid tumors

Phase 2




Solid tumors

Phase 1

Bintrafusp alfa (M7824)

Merck KGaA, GlaxoSmithKline


Solid tumors

Phase 1

Integrins represent another important set of players. This 24-member family of heterodimeric cell adhesion receptors physically connect the ECM to the cellular cytoskeleton, as well as providing a bidirectional signaling link between a cell’s interior and its environment. Although overexpression of integrin αvβ6 has long been identified as a negative prognostic biomarker in certain cancers, several integrin inhibitors failed in clinical trials. Most notably, cilengitide, a peptide inhibitor of αvβ3 and αvβ5 integrins, in development at Merck KGaA, had no effect on survival in a phase 3 trial in glioblastoma. The same company has since allied with SFJ Pharmaceuticals to test abituzumab, which targets all αv integrins.

Targeting integrin adhesion and signaling in cancer may be complicated by redundancy among the different integrin receptors, however, and indirect approaches acting on targets downstream of integrin-mediated signaling—including focal adhesion kinase (FAK), Rho, Src and extracellular signal-related kinase (ERK)—may prove more successful. Statins may affect this system too. Statins lower cholesterol by inhibiting 3-hydroxy-3-methyl-glutaryl-CoA reductase, which catalyzes the rate-limiting step in the mevalonate pathway. This disrupts attachment of a prenyl lipid group to Rho, a post-translational modification that leads to its translocation to the cell membrane and its activation. It may be a stretch at this point to suggest that statins may have a future as a cancer drug, but epidemiological studies suggest that statin use is associated with improved survival in ovarian cancer and with lowering the risk of developing pancreatic cancer.

In contrast, Rakesh Jain’s longstanding research into the pathophysiology of desmoplastic cancers—which are characterized by a dense tumor stroma comprising fibrous and connective tissue—is now close to generating definitive clinical evidence that ‘normalizing’ the extracellular matrix by angiotensin system inhibition (ASI) reduces compressive stress on the blood vessels within the tumor. It’s more than twenty years since Jain and colleagues reported that solid stress—generated by the growth of cancer cells within the confines of a tumor—inhibits in vitro growth of tumor spheroids. They later showed that this pressure causes the collapse of blood vessels—and that the collapse can be reversed in vivo by the selective elimination of human xenograft cells in a murine model. A subsequent study showed that depleting cancer cells, fibroblasts, collagen or hyaluronan each increases tumor perfusion. Shortly afterwards, the group reported similar effects with losartan, an inexpensive drug with a known safety profile that is widely used to treat hypertension.

At a molecular level, losartan and other ASIs work by blocking signaling downstream of the angiotensin I receptor. “By blocking the receptor, many things happens,” Jain says. Receptor blockade improves tumor perfusion, reduces hypoxia and improves the delivery of chemotherapy. Animal studies also suggest a likely effect on the immune response. “In transcriptome analysis, we found increased expression of genes related to cytotoxic T cells and also dendritic cells,” says Jain’s colleague Yves Boucher, an associate professor at MGH and Harvard Medical School. The group will conduct similar studies on patient material taken during the recent trial.

The recently reported single-arm trial builds on an earlier retrospective study, which found that pancreatic cancer patients with inoperable tumors who were also on chronic ASI therapy were twice as likely to convert to having resectable tumors after undergoing chemotherapy than those who were not taking ASIs. Solidifying the evidence from both studies is now an urgent priority. “We’re doing a randomized controlled phase 2 trial right now,” Jain says. This academic study is sponsored by MGH; Bristol-Myers Squibb and the Stand Up To Cancer initiative are collaborating. It is comparing the effect of adding losartan and/or the PD-1 inhibitor Opdivo (pembrolizumab; Bristol-Myers Squibb) to standard chemotherapy, radiation and surgery.

A prospective study of losartan in ovarian cancer is also due to get underway at MGH and several other centers, following a recent study led by Jain and Lei Xu of MGH and Harvard Medical School, who found that losartan dramatically reduced abdominal fluid build-up in a mouse model of ovarian cancer and improved uptake of—and responses to—the chemotherapy drug paclitaxel. “Losartan reduces extracellular matrix in the diaphragm, and that opens up the lymphatics,” says Jain. A retrospective analysis of patient records supported the presumed effect: patients receiving standard therapy for ovarian cancer who were also on ASI therapy lived thirty months longer than those who were not.

Intriguingly, mechanical forces may not only influence cancer progression at cellular or molecular levels but also at the macro level (Box 1). Ingber found that, when growing lung cancer tumors on a three-dimensional chip that mimics both lung tissue structure and the accompanying breathing motions, stopping the breathing movement increased tumor growth. It surprised even him: “Breathing motions regulate tumor growth, that’s totally wild,” he says.

Box 1 | Force drives direction of pancreatic cancer growth

In London, scientists at the Francis Crick Institute have recently reported on an aspect of cancer mechanobiology that could define the course pancreatic cancer may take in different patients. Cancer biologist Alex Behrens and colleagues identified why some ductal carcinomas grow inwards (endophytically), into the lumen of the duct, and why others grow outwards (exophytically), into the surrounding tissue. They did so in a mouse model of pancreatic cancer using a three-dimensional, whole-organ imaging technique that enabled them to conduct tissue analyses of epithelial tubes at single-cell resolution. They found that the critical factor that determines the direction of growth is the diameter of the duct itself: if it’s below 17 micrometers, the tumor grows exophytically, whereas if it’s above that threshold, it grows endophytically.

What causes a tumor to grow in one or another direction arises from mechanical forces acting directly on the cells. In a small duct, the tight curve on the inner surface creates high tension between the cells, which requires more energy to break, so the tumor is forced outward. In larger ducts, the tension is less, so the growing tumor can exert enough force to grow inward. “The types of pancreatic cancer that form are fundamentally different,” Behrens says. “They are histologically different, and different in how they interact with non-tumor cells.” Exophytic tumors grow into the parenchyma and interact with the cells around them, inducing the recruitment of cancer-associated fibroblasts and the process of epithelial-to-mesenchymal transformation—both markers for aggressiveness. Endophytic tumors may be less aggressive but ultimately more dangerous because they will eventually occlude the whole ductal system.

These findings may help clinicians to understand why pancreatic cancers present so differently and opens up a new path of inquiry, to determine whether such variations may influence patients’ responses to therapy.

Translating the myriad findings on tumor mechanics into useful therapies or biomarkers is a vast challenge given the complexity of the biology. But the field is now an established and ‘respectable’ area of biological research. The skepticism that initially greeted the ideas of Bissel, Jain and Ingber has evaporated. “When I first suggested that mechanical forces could drive cancer, everyone thought I was absolutely nuts,” says Ingber. “It’s great to see all the interest and all the work being done with it now.”

doi: 10.1038/d41587-019-00019-2


Additional reporting by Brian Owens, New Brunswick, Canada.

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.