Article Recommendation: Glioblastoma Subtypes Defined Using Data from TCGA

After reading Verhaak et al. 2010 in Cancer Cell, and I was impressed by this very good study analyzing data from an important resource for genomics research.

The authors were able to define gene signatures to define 4 subtypes of glioblastoma.  The experimental design was pretty straightforward, and the results were quite clear.  Most importantly, their predictive model was trained an a relatively large set of 173 patient samples and validated on an even larger set of 260 patient samples (from 5 independent studies).

The study focused mostly on data provided by The Cancer Genome Atlas (TCGA).  TCGA is a database that contains various types of genomic data (gene/miRNA expression, gene/miRNA copy number, DNA sequence/polymorphism, and DNA methylation), and most or all types of genomic data are available for each patient in the database.  This provides a unique opportunity to integrate many different types of data, usually for a large number of clinical samples.  For anyone not aware of this resource, I would strongly recommend checking out the links provides above as well as the original TCGA paper (also on glioblastoma) published in Nature.

The Personal Benefits of Self-Regulation

Although dishonest individuals do not always experience immediate repercussions for their unethical behavior, there are a number of benefits to having the self-discipline and courage to search for an honest career that is truly helps other people because the development of ethical habits early in one's professional career is likely to pay off later in life.

There can be significant long-term consequences to dishonest behavior, as shown by an increasing number of retractions of papers from scientific journals and a noticeable presence of scientific misconduct in the news.  For example, Anil Potti resigned from Duke University after it was revealed that he published forged results and included inaccurate information in his CV (such as falsely claiming that he was a Rhodes Scholar).  However, there are also a number less drastic consequences that do not involve formal punishment for bad behavior.

Embellishing results early in one's career can make downstream research more difficult.  For example, inaccurate predictions will make it difficult to get positive results from follow-up experiments to validate a preliminary hypothesis.  Also, many scientific disciplines offer rotations for graduate students, and it will be more difficult to recruit top-notch graduate students if other labs can offer more interesting projects with a better experimental design.  It is difficult to maintain a steady stream of publications in a lab with little or no competent personnel.

Although scientists who forge results on a regular basis may not necessarily have to worry about downstream analysis (because all of their results are false anyways), individuals who make false claims on a regular basis are more likely to be caught by others attempting to verify important results.

Networking and social interactions will also be more difficult for individuals with a reputation for behaving dishonestly.   Even if the general public is not aware of a person's reputation, individuals  who behave unethically on a regular basis will probably have difficulties developing a close network of friends.

As my grandmother used to say, the common saying shouldn't be "practice makes perfect" but rather "practice makes permanent" because all habits (good or bad) are difficult to break.  If aspiring scientists start engaging in unethical conduct, then it will become increasingly hard to break those habits at a later stage in their career.

Is it worth the effort to develop personalized cancer treatments?

Robert Langreth has published a series of articles about personalized cancer treatments in Forbes Magazine (see Part I, Part II, Part III, and/or this summary in GenomeWeb).  In a nutshell, the author's main point (in the first article) is that it's very difficult to develop new drugs, and the costs to produce drugs that only help a small proportion of patients may outweigh the benefits for that drug.

I agree with the author in that I don't think it's reasonable to expect to produce an individualized drug for every possible mutation that can cause a disease (such as cancer).  However, I think genetic studies can still help improve treatments for several reasons.

First, there are a wide variety of tools that physicians can use to help patients, and I think it is wrong to view this argument from an "all-or-nothing" point of view.  Pfizer's response in the third article also criticizes the "all-or-nothing" thinking, although they cite the need for "accumulated modest advances" while I am saying it is good to have more options.  For example, the first article mentions the need to develop better surgical methods.  I'd like to see more personalized drug treatments as well as new surgical technologies.  I imagine there will be certain circumstances where a personalized drug therapy is ideal and certian circumstances where surgery will be necessary.  If we reach the point where even 30% of patients can receive personalized drug treatments, then I think that is pretty good.

Second, genetic tools can assist with the diagnosis of existing drugs (or drugs in clinical trails that were not originally designed for individuals with a specific mutation).  For example, a drug company may come very close to bringing a drug to the market, but then realize that the drug has severe side-effects for certain individuals.  If the individuals with the severe side-effects can be identified ahead of time, then the company can prevent total loss of their research costs by targeting individuals without a particular mutation.  Furthermore, scientists can discover new functions for drug candidates (as happened with Viagra), so genetic information may be able to provide researchers with alternative uses for existing drugs (or novel drug candidates).

Finally, it's important to keep in mind that cancer is the 2nd leading cause of death (in the US).  I'm sure there is a going to be a point where a mutation in a particular gene (or related pathway) is too rare to warrant developing a personalized treatment, but drugs that can decrease mortality in even 10-20% of patients may still be able to help a large number of people.