Cancer is not a single disease. It is a broad term that describes a number of related conditions in which cells’ growth has begun to bypass the usual checks and balances. To study the spectrum of cancers, the National Institutes of Health have established The Cancer Genome Atlas (TCGA), a collaborative project aimed at understanding these complex diseases. The project’s goal is to build a map of the genetic changes that occur in different types of tumors, with an ultimate aim of improving physicians’ capacity to prevent, diagnose, and treat cancer in a way that is tuned to be most effective for each individual patient. To date, the project’s members have analyzed >10,000 tumor samples of 33 distinct tumor types. This has resulted in a vast set of data, which TCGA has made publicly available. Having such a large sample size helps researchers have confidence that the conclusions that come out of their analyses are real effects, not the effect of chance.
In The Immune Landscape of Cancer, published April 2018, a team of 600 scientists from all over the world made use of this dizzyingly vast array of data to attempt to separate the 10,000 cancer cases into groups with interesting characteristics. The patient cases accompanying the genetic sequences of tumors were well annotated, which allowed the researchers to look for associations between the different genetic “groups” and various outcomes of interest, with a goal being to allow personalized medicine for future patients.
Thorsson et al. separated the ~10,000 tumors into six immune subtypes, on the basis of a large number of discriminating factors, such as whether the genetic profiles of the cells in the tumor revealed dominance of one type of white blood cell over another, overall white blood cell levels, the balance between inflammatory or anti-inflammatory molecules, and other traits.
Once they had separated the tumors into the six groups, the researchers were interested in comparing relative outcomes. They found that patients with tumors of the most “inflammatory” signature had the best overall odds of survival. This group of tumors was characterized by high expression of genes coding for Th17 and Th1 (which collectively enable your immune system to attack the tumor), low to moderate growth of the tumor cells, and a relatively rare incidence of aneuploidy (cells with the wrong number of chromosomes).
The authors had gathered an immense amount of data and sorted tumors into groups. Now, it was time to begin to see if that data could be used to create new therapeutics, such as a broadly-protective vaccine or drug target. Usually, when immunologists are creating a new vaccine for a bacterium or virus, they can combine some proteins isolated from the pathogen with an adjuvant. The adjuvant stimulates your immune system to recognize the pathogen’s protein as a dangerous substance, and your immune system then learns to attack it upon future encounters, reducing the severity and likelihood of future illness.
This process is really hard to do with cancers, as Sara recently described. Tumor cells are simply too similar to the rest of our healthy cells. In order for a cancer vaccine to work, we have to identify a new protein that is only found in/on the cancer cells.
To identify potential new proteins that were only expressed in tumor cells, the researchers compared the genetic information from all of the tumors to the healthy cells and looked for differences. The results were disappointing. Although new cancer-specific proteins were present in many of the tumors, it was very rare to find the same newly-mutated protein in different patients. In fact, they didn’t find any new individual cancer protein in more than 1% of tumor samples. This means that it is going to be very difficult to produce an “off-the-shelf” cancer vaccine that will be applicable to many patients. Instead, there will either be a library of pre-set vaccines for various cancer subtypes, or cancer vaccines will have to be customized for individual patients. This is still within the realm of possibility.
Although these options will mean that the cost of care will be more expensive, successfully harnessing the power of a patient’s immune system to fight the cancer cells is a strategy that has already shown immense promise. Therapeutic antibodies that modulate the immune system to make it more active have had a strong impact on cancer therapies in recent years. Similarly, CAR-T cell therapies, where a patient’s immune cells are re-engineered to specifically attack leukemia cancer cells, have been highly effective (albeit expensive).
As we aim to improve cancer detection, prevention, and care, new information to allow us to understand individual patients’ tumor types will be a key element. In order to deliver on the promise of personalized medicine, the clinical community needs at least two basic elements: 1) a detailed understanding of an individual patient’s tumor, and 2) specialized tools and protocols to use for various types of patient tumors. With the ever-decreasing cost of genetic sequencing, and large-scale collaborations such as TCGA, analyses like the ones performed in this paper will continue to emerge. The open source nature of this data will allow researchers across the world to develop and test specialized methods to treat individual types of tumors as effectively as possible.
Thorsson, V., Gibbs, D. L., Brown, S. D., Wolf, D., Bortone, D. S., Ou Yang, T. H., . . . Shmulevich, L. (2018). The Immune Landscape of Cancer. Immunity, 48(4), 812-830 e814. doi: 10.1016/j.immuni.2018.03.023
Credit: Tom Deerinck