Are Migrating Cancer Cells Like Rafting Monkeys?

Early in the morning, before getting sucked into the e-world, sometimes even before having coffee, I’ve been reading a book called Cancer: The Evolutionary Legacy by a British medical researcher named Mel Greaves. This probably sounds like an astoundingly bad way to start the day, but Greaves’ tone is pretty light, and he doesn’t get into the individual stories of cancer sufferers. It’s a concept book, and an incredibly fascinating one. I’m only making it through a few pages a day, because I’m constantly stopping to digest some point, and turning the margins into snarls of exclamation points, arrows, and scribbled thoughts.

Greaves obviously loves a good analogy, and when he describes metastasis—the establishment of cancer cells in new locations within the body—what comes out, strangely enough, is pure biogeography, and, in particular, long-distance colonization. The malignant cells in their original location (colon, breast, or wherever) are hemmed in by physical boundaries and barriers, like “an overpopulating and localized species of animal” confined to an island, with the surrounding sea cutting off escape. When a few cells manage to break out into the bloodstream or lymphatic circulation, “it is rather like a desperate migrant crossing the China Sea in a fragile vessel.” Most of these cells will not survive the journey, but “given the number that try and the time available, inevitably a few may succeed.”

In other words, the movement of cancer cells is a lot like monkeys crossing the Atlantic or other instances of chance, overseas dispersal, even extending to the idea that a small probability of success for any single attempt can translate to a much higher one given enough time. That quote about “the number that try and the time available” could have come from the paleontologist George Gaylord Simpson writing about overwater dispersal by Cenozoic mammals. Or, for that matter, from Darwin arguing for the natural colonization of oceanic islands.

Figure drawing: Edward Burne-Jones. Map: Ktrinko—own work, CC0, https://commons.wikimedia.org/w/index.php?curid=17169364

The analogy doesn’t stop there, though; in fact, the detailed parallels pile up almost to the point of absurdity. Thus, once a cancer clone makes it to a new location, it is more likely to thrive “in a similar environment or ecosystem to that from which it came—one part of the bladder to the other, skin to skin.” So the successfully colonizing cancer cell is like, say, a shrub from a tropical rainforest that, after dispersing, gains a foothold because it happens to find itself in another tropical rainforest.

Continuing on, some types of cancer cells are better migrants than others, just as ducks and rats are more likely overseas colonists than salamanders and kangaroos. Numbers at the source location are important in both instances—of cancer cells in the one case and of individual organisms in the other. Access to thoroughfares can be critical—cancer cells multiplying next to a large blood vessel are like monkeys living by a river that might wash them down to the sea. And lymph nodes can act as way stations for migrating cancer cells, the equivalent of stepping-stone islands for migrating species.

Greaves’ analogy makes it clear that metastasis is, at its core, a random process, but with the probabilities heavily influenced by various factors. The randomness of it—chance cell movement, like chance dispersal—gives some insight into the fact that two people could develop the same kind of cancer simultaneously, yet, in one, the disease might spread almost immediately whereas, in the other, it might take years to metastasize. Meanwhile, the influencing factors, such as access to many blood vessels and lymph nodes, make sense of why some kinds of cancer often persist for years without spreading, while others can be expected to metastasize very quickly.

Following up on these thoughts, I did a quick search in the Web of Science for mathematical models of  metastasis. None of the studies mentioned overseas dispersal or rafts of vegetation, but they did give me some reason to tweak the biogeographic analogy to more closely match the disease situation. It appears that, for at least some kinds of cancer, the initial and subsequent tumors might interact with each other beyond the original one-way colonizations. In particular, the tumors within a person seem to receive new “seeds”—colonizing cancer cells—from one another, and the re-seeding influences their growth. This makes me think that monkeys rafting across the Atlantic, which happened only once as far as anyone knows, is not the best analogy for these cases. The migrating cancer cells are more like Darwin’s finches colonizing and re-colonizing islands in the Galápagos.

Darwin’s evolutionary ideas have been fruitfully translated to all sorts of areas outside of their original domain, including such things as constructing genealogical trees of languages, describing the fates of memes, and creating computer algorithms that solve problems by mimicking natural selection. Close to the topic at hand, many people have recognized that the growth and spread of cancer within the body is a Darwinian process—selection occurring at the level of competition among cells—and that this knowledge might prove useful in developing new treatments.

To end on a vaguely practical note, I wonder if models of chance dispersal—the mathematical embodiment of ideas set forth by Darwin, Simpson, and others—might be applied to the process of metastasis. Could these biogeographic models give us some novel insight into that process, either by introducing new variables or a different kind of analysis? Could they provide something more medically useful than a metaphor?

Hawaiian ramblings: koa trees, broad-nosed weevils, and a Darwinian experiment

On bristletail collecting trips to Hawaii, we’ve often run across koa (Acacia koa), a native tree that’s common in so-called “mesic forest” (that is, not too dry and not too wet, by Hawaiian standards). Koa have sickle-shaped leaves that bring to mind eucalyptus, but it turns out that these leaves are actually modified leaf stalks, not true leaves. The real leaves can be seen on very young koa and are compound like those of typical Acacia species.

Compound leaves of a young koa tree

Modified leaf stalks (phyllodes) of an adult koa

The strange leaf development story of koa, and the fact that bristletails have been found on them, made me especially aware of these trees, and I’ve tended to notice them among the welter of mostly unknown Hawaiian plants (unknown to me, that is). That familiarity made it especially satisfying to hear about a new study by Le Roux et al. (also described here) indicating that koa dispersed naturally from Hawaii to Réunion, one of the Mascarene Islands, in the Indian Ocean, giving rise to a similar species on Réunion, Acacia heterophylla.

Yes, the Mascarenes are a long way from Hawaii, more than 10,000 miles, in fact. While writing The Monkey’s Voyage, I looked at a lot of maps, of course, but to get a good sense of this koa dispersal, I felt like I needed a more planetary perspective. So I pulled our kids’ globe off the shelf and, with my right index finger fixed on Hawaii, I spun the globe and put my left index finger on Réunion. My two fingers were not quite directly opposite each other on the globe, but they weren’t too far off. The implication was obvious: Hawaii to Réunion is a very, very long trip, even assuming the most direct route.

In addition to the sheer distance, several things make this case especially striking. One is that the area of origin is an island, which means that the source population would have been small compared to many continental ones, decreasing the already slim odds of a successful dispersal. Another is that koa grow in the mountains, not on the coast, so just getting to the ocean is a chance event for them. And, finally, the fruits and seeds of koa aren’t adapted for dispersal by wind, water or birds. For these reasons, it’s not a species you’d pick out as especially likely to make a tremendously long overwater journey.

Basically, the colonization of Réunion by Hawaiian koa must have been a major fluke, an extremely improbable event that ranks with monkeys rafting across the Atlantic or carnivorous sundew plants somehow dispersing from Australia to a tepui in northern South America.

********************

On koa trees at night, dark, broad-nosed weevils can be found munching on those sickle-shaped leaf stalks, something this group of insects has probably been doing for several million years. The weevils are large—some of the species can be 3 cm long, which is big for a weevil—and, probably for that reason, they’re coveted by beetle collectors. And, as it turns out, when it comes to unexpected long-distance dispersal, the koa trees have nothing on these weevils.

The koa-eating weevils are in the genus Rhyncogonus, a group that, like the koa itself, doesn’t seem like a great candidate for oceanic dispersal. In addition to being bulky for insects, Rhyncogonus are flightless, and their larvae live, not on plants, but in the soil. They seem like just the sort of insect that would not be good at dispersing long distances.

Rhyncogonus howarthi in the Waianae Range, Oahu. Photo by Karl Magnacca.

In fact, on one level, there are indications that Rhyncogonus are not good dispersers. In the Hawaiian chain and other islands where they’re found, species of Rhyncogonus are almost always restricted to single islands and, often, to small parts of those islands. For instance, of nearly 50 Hawaiian species, only one or two are known to occur on more than one island. That kind of geographic pattern suggests the weevils do not move freely between islands, even if the islands are quite close together. Once a lineage of Rhyncogonus establishes itself on an island, it essentially evolves on its own, rarely mixing its gene pool with populations on other islands.

However, counterintuitively, when it comes to extreme long-distance colonization, Rhyncogonus weevils have been ridiculously successful. Their ancestors probably came from Asia or western Pacific islands, and from there they spread all over the warm parts of the central and eastern Pacific; endemic Rhyncogonus are now found on Wake Island, Hawaii, the Line Islands, the Kermadecs, and the French Polynesian islands of the Marquesas, Societies, Australs, and Mangarevas, among other places. Defying intuition, these large, flightless beetles have somehow managed dozens of overwater colonizations, some of them involving voyages thousands of miles long.

********************

On a 2012 trip to Kauai, the bristletail collecting crew met up with Robin Rice, who runs the Kipu Ranch, one of the places where we wanted to look for the insects. One of Robin’s forebears, William Hyde Rice, who was the last governor of Kauai under the Hawaiian monarchy, apparently bought the ranch in the 1870s from Princess Ruth Ke‘elikolani, and the property has been in the family ever since.

I learned that history from Wikipedia, not from Robin; most of my conversations and email exchanges with him have been intensely biological and, especially, entomological. Studying insects is Robin’s hobby, but it’s more like a passion, one that’s resulted in significant discoveries in Hawaii and many other Pacific islands. I’m not really the person to judge, being an entomological neophyte, but my impression is that he knows the insects of Pacific islands like very few people do.

These days, Robin’s focus is on tree roaches, but at one point, back in the 1970s, he was especially interested in Rhyncogonus. And, like others who have studied the genus, he wondered how these bulky, flightless beetles had colonized islands scattered over a vast expanse of the Pacific.

Some people had suggested that Rhyncogonus got around attached to birds, but Robin doubted that. He had another notion about their dispersal, and thought he could shed some light on the subject by doing a quick-and-dirty experiment using Rhyncogonus simplex that he found feeding on native cotton plants on Oahu.

In an email last month, here’s how he described what he did:

I would take some simplex and drop them into vials of sea water, where they would immediately sink to the bottom and become immobile. Later I would take them out and put them on filter paper in a petri dish to see if they would revive. After one or two days completely submerged nearly all would revive. If I recall the most extreme experiment was 4 days on the bottom of a jar of seawater, whereupon about 2-3 of 15 weevils were walking around in the petri dish a few hours later.

Basically, Robin had performed an experiment along the lines of the famous seeds-in-seawater trials that Darwin had used to establish the plausibility of plants colonizing remote islands. Robin’s experiment (and Darwin’s) was what the philosopher of history William Dray has called a “how possibly” experiment. How possibly could weevils have survived a long rafting voyage in the ocean? Well, it would help if they could survive being inundated for days at a time. And, strangely enough, they can!

Of course, the Lazarus-like recovery of the weevils doesn’t prove or even strongly suggest that the main way that Rhyncogonus reached all those islands was by rafting. But it does push one toward thinking that answer might be the right one. The experiment helps to suspend one’s disbelief.

Darwin, a dispersalist, master of the “how possibly” argument, and, in his youth, an avid beetle collector, would have loved it.