There is no question that a “one size fits all” approach to treating cancer is simply not going to be effective. The field of cancer research has shown that the molecular changes driving cancer differ from person to person. As a result, we have seen much more interest in the development of personalized cancer models that can help physicians guide which treatment will work best for their patients.
While the use of personalized cancer models will be important in the battle against cancer, a new study highlights an important limitation in the use of personalized mouse models of cancer.
In some personalized models of cancer, mice with compromised immune systems are implanted with tumors from cancer patients, called patient-derived xenografts (PDX). These animals, sometimes referred to as mouse avatars, are often treated with different anti-cancer drugs to tailor effective and personalized treatments for the cancer patients.
Despite the hype and media coverage that mouse avatars have gotten, limitations of this approach—including the difficulty in creating the models in the first place—have been discussed for many years. There are also concerns that the human tumor may not grow and evolve in the mouse the same way as it would in the human patient, meaning that treatments that showed promise in mice would not be guaranteed to work in people.
In addition, the generation of mouse avatars is a slow and costly approach. There is, therefore, the very real possibility that the cancer patient may pass away before the results from the mouse model are even available.
A new study set out to determine whether the human tumors that have been implanted into mice actually reflect the genomic features of the primary tumors from which they were derived. The study revealed that the tumors transplanted into mice were genomically unstable and differed in “therapeutically meaningful ways” from the original tumor. This is troublesome, given that mouse avatars are often created specifically to test for effective cancer drugs. If the model does not closely represent the primary tumor, then it cannot accurately predict treatments that will work in people.
Fortunately, in vitro approaches to finding personalized cancer treatments are already available and show some advantages over the mouse models of cancer. Unfortunately, however, it will be difficult to change the mindset of some cancer researchers, who will turn a blind eye to data that devalues their animal model.
Ben-David, U. et al. “Patient-derived xenografts undergo mouse-specific tumor evolution,” Nature Genetics. Published online October 9, 2017
Taylor, A. “Study Raises Questions About Patient-Derived Xenografts,” The Scientist. October 9, 2017.