Targeting an adhesion molecule to prevent brain colonization of lung cancer

See article by Münsterberg et al in this issue, pp. 955–966.

Brain metastasis is a frequent and feared complication of many solid tumors, compromising quality of life and survival of many patients.¹ About 40% of patients with advanced non-small-cell lung cancer (NSCLC) develop brain metastases, which makes the lung the leading organ of origin. Three questions appear of key importance for brain metastasis research: Which cellular and molecular factors drive early brain seeding? How do tumor cells then thrive in the foreign environment, allowing a clinically relevant macrometastasis to grow? And, last but not least, which treatments are most effective to reduce early seeding on one hand and late growth on the other?

A high number of cancer patients who are brain metastases free at the time of diagnosis have a relevant risk to develop them in the future. Thus, there is a clear clinical need for a better prevention of clinically relevant brain macrometastases, ideally with a non-(neuro)toxic drug.¹ In this context, it is crucial to better understand the biology of early brain seeding. The discovery of a mandatory early angiogenic switch in an NSCLC mouse model and later findings that anti-angiogenics might indeed have brain metastasis preventing effects in NSCLC patients support this concept.^2,3 There is still much to be learned about the early steps of brain colonization, and practical hurdles exist for prospective clinical trials testing brain metastasis preventive drugs⁴—but there is also much to be gained for the course of many oncological diseases if such a trial succeeds.

One of the first steps of brain colonization (and thus a plausible target for preventive concepts) is the adhesion of circulating cancer cells to the brain microvascular endothelium. Here, an earlier study found that multiple adhesion molecules are upregulated in brain endothelial cells during micrometastatic seeding of breast cancer cells—most notably, activated leukocyte cell adhesion molecule (ALCAM), but also many others.⁵ Specifically, the inhibition of ALCAM by short in vitro incubation of breast cancer cells with an ALCAM-blocking antibody reduced the number of brain metastases that later formed in a mouse model, which speaks for a brain metastasis preventive effect.

In this issue of Neuro-Oncology, Münsterberg, Loreth et al extend this earlier finding to NSCLC and provide ample support for a specific role of ALCAM throughout the brain metastatic cascade in this entity.⁶ They first demonstrate that ALCAM is heterogeneously expressed in primary NSCLC tumors, and that high expression predicts worse survival, confirming earlier reports. Interestingly, they find that ALCAM expression on circulating tumor cells matches well the level of expression in the corresponding brain metastatic tissue, which could be investigated for 4 patients only, due to the scarcity of this kind of material. In (brain) metastatic lesions, ALCAM expression was even increased compared with the primary tumor, and patients with brain metastases in which ALCAM protein was not detectable showed the best survival. Together this provides the first indication that ALCAM might play a clinically relevant role in NSCLC brain metastasis.

Using clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated nuclease 9 technology, the authors then inactivate ALCAM expression in a human NSCLC cell line and demonstrate that under these circumstances the adhesion to brain endothelial cells is reduced, under both static and flow conditions. This is unlikely due to an effect on tumor cell autonomous processes: proliferation, migration, and colony formation in vitro where unchanged by genetic ALCAM inactivation. This speaks for a specific role of ALCAM for cancer cell‒endothelial cell attachment. Electron microscopy images support a profound change in this respect, with less direct membrane contacts and a rounded cell shape of ALCAM-inactivated NSCLC cells.

Finally, the authors investigate 2 brain metastasis mouse models using this cell line, one with direct tumor cell implantation into the brain and one with intracardiac systemic injection into the bloodstream. They find that the number of brain metastases is significantly reduced after intracardiac injection when ALCAM expression is compromised, similar to what has been described for breast cancer. Interestingly, no significant effect of the ALCAM knockout on growth after brain implantation was detected. Together this speaks for the relevant role of ALCAM for cancer cell arrest in the brain, or at least the earliest events of brain seeding, and against a relevant role of ALCAM for later, macrometastatic growth in NSCLC. As an interesting side note, the authors find circulating tumor cells in mice even in the intracranial injection model, which provides a first indication that the phenomenon of systemic spread from brain tumors is not limited to gliomas. ALCAM deficiency reduced the number of circulating cancer cells found, in both the intracranial and heart injection models, hinting at a possible role of ALCAM for cancer cell survival in the bloodstream, too.

The data presented here support the concept that the biology of early seeding and later macrometastatic growth in the brain is fundamentally different, which is plausible when considering the diverse steps that need to be sequentially mastered by blood-borne cancer cells to eventually grow to a clinically relevant brain metastasis.² Translationally, the specific implication here is that targeting ALCAM makes sense only in a clinical setting where the prevention of brain metastasis will make a likely change to the patients’ outcome—and not where treatment of existing ones is required. The relevant question is: can targeting of ALCAM, or another adhesion molecule like vascular cell adhesion molecule 1, as recently successfully demonstrated in another preclinical brain metastasis study,⁷ be clinically performed with a reasonable therapeutic window? Several obstacles remain for ALCAM: it is widely expressed in the body, and in the brain ALCAM inhibition might conversely increase blood–brain barrier permeability with the potential risk of aggravating inflammatory and other diseases.⁸ However, preclinically, anti-ALCAM antibodies have been given systemically to mice with apparent tolerability,⁹ and dietary selenium uptake has been shown to reduce both ALCAM expression in brain microvessels as well as brain metastases formation.¹⁰ Thus, it will be interesting to learn whether targeting of those crucial adhesion molecules will emerge as a road to reducing the incidence of brain metastases in patients.