Roger McLendon on Solving the Jigsaw Puzzle of Glioblastoma
Special Topic of Glioblastoma Interview, October 2011
Subsequent experiments have identified various growth factors that are not only produced by the stem cells, but also induced in other cells. What keeps us up at night is wondering how this all fits together. Recently we published a theoretical report on the application of Chaos theory and self-organizing systems into understanding these tumors at the histologic as well as at the macroscopic levels (McLendon RE, Rich JN, "Glioblastoma Stem Cells: A Neuropathologist's View," Journal of Oncology 2011: 397195-397195).
You've also participated in several genome studies for gliomas. Could you give us an overview of these studies, and tell us what sort of discoveries or advances in gliomas have come about from them over the years?
Prior to 2000, the focus of brain tumor research was in developing models of gliomas, either through cell culture, xenografting, or developing model systems. Cell cultures, in particular, were interesting to the molecular biologists for their genetic amplicons, called double minutes. Because of the availability of these models at Duke, Bert Vogelstein collaborated with Darell Bigner to characterize the genes that were contained in these double minutes. From those studies, EGFR variant 3 and GLI were discovered.
By 2000, molecular techniques were advancing at a rapid pace and it was becoming possible to examine significant quantities of DNA or RNA at a time, which, of course, has expanded to whole genome studies now. As a neuropathologist, my role was to refine the tissue collection and banking methods to preserve DNA and RNA as well as to better characterize the histologic contents of each banked tissue block. It was also important to educate my clinical colleagues about the importance of obtaining significant quantities of tumor tissue for banking, tissue that would otherwise be discarded as "in excess of that needed for pathologic examination."
By not only banking the tissues, but also maintaining a database that contains descriptive data on the blocks as well as clinical information, we have become a very valuable resource not only within our institution, but also attracting the attention of top research institutions.
"When I was a child, I was given a wood block jigsaw puzzle. Cancer is very much like that puzzle, however, not only do we have to put the pieces together, we have to determine the size and shape of each intricately small piece from experiments."
By having such granular clinical and pathologic data available on these tumors, they have been instrumental in the discovery of PIK3CA mutations in anaplastic oligodendrogliomas and the amplification of OTX2 in anaplastic medulloblastomas by Dr. Hai Yan, the discovery of Isocitrate Dehydrogenase 1 and 2 mutations in progressive gliomas in collaboration with Vogelstein, the discovery of ERRFI1 as a potential tumor suppressor gene and TACC3 as a potential oncogene in GBM by Duncan and Yan, and the recent reports of the significance of ATRX and DAXX mutations in sustaining telomere lengths by Heaphy and Meeker.
Hai Yan is another young researcher who joined the Brain Tumor Center after his post-doc with Vogelstein; it has been very gratifying to watch him become one of the top young brain tumor researchers.
You and Edward Halperin published a very interesting study in Cancer in 2003 using the Duke University Medical Center Tumor Registry, "Is the long-term survival of patients with intracranial glioblastoma multiforme overstated?" (McLendon RE, Halperin EC, 98[8]: 1745-8, 15 October 2003). How was your call for caution received? Has the situation changed (for better or for worse) since publication of this paper?
I really do not know how it was received; however, it was just something a neuropathologist does: collect cases, review their clinical histories, and look for any common issues. Unfortunately, for these cases, at least at the time, the common issue was that survival for about half of these glioblastoma patients had been calculated from the appearance of their original lower-grade tumors. The result was an initial false impression of a larger-than-expected number of long-term survivors.
However, the end results were not entirely negative. In the IDH1/2 work described above with Hai Yan, we demonstrated that patients whose glioblastomas exhibit either the Isocitrate Dehydrogenase 1 or 2 gene mutation were tumors that had progressed from lower-grade tumors. Of particular interest was that the survivals of the patients whose tumors exhibited the mutation were longer, and correcting the diagnosis dates forward to the date of glioblastoma diagnosis in the fashion described in the Cancer paper anticipated that issue. Thus, that project proved to be a blessing in disguise.
What are the particular issues and challenges in using brain tissue to learn about glioblastoma and similar malignancies?
Although we have been banking tissues for at least 30 years, each new molecular technique puts new demands on tissue preservation. As a result of recent advances in molecular expression analysis, we find that we are still in the early years of learning about the tolerances of tissue components to various handling techniques. We are just beginning to learn the influences of warm ischemic time, or the time the tissue remains in the brain while the surgeon is removing it, and cold ischemic time, or the time the tissue is out of the brain but before it is cut into blocks and preserved.
"Although we have been banking tissues for at least 30 years, each new molecular technique puts new demands on tissue preservation."
We have found that phosphorylation status of the second messenger proteins are exquisitely sensitive to ischemic time, with dephosphorylation occurring very rapidly after surgical devitalization. Furthermore, while we understand that dephosphorylation is occurring, we are not sure of either the rate of dephosphorylation or the relative differences in rates among proteins. If the process were uniform, different assumptions could be made versus if one molecule was dephosphorylating faster than others.
A significant advance will come when we determine those biological markers that correlate with quality of preservation relative to the technique to be used. However, molecular preservation is not the only issue. Hopefully one day we will develop cryopreservation techniques that allow the block to be sectioned for outstanding histologic appearance as well as for molecular extraction. Another advance will be the preparation of tissue microarrays with frozen tissues. These are all issues where an understanding of neuropathology is critical and neuropathologists should be supported for their knowledge and efforts.
How has the field as a whole changed over the past 10 years? Would you say we are in a better position today in terms of our knowledge of glioblastoma (and other brain tumors) than we were in the 1990s? Why or why not?
The biggest change that I have seen is the increased demand for high-quality tissue. With high-throughput analysis, studies that once took years now take weeks. Exciting genetic discoveries in glioblastoma are being published almost monthly now. As such, the advances are attracting highly talented and qualified scientists and clinicians into the field. This is great news for everyone concerned, both now and in the future.
New avenues of therapy are being explored, such as the vaccine trials being done by Drs. John Sampson and Duane Mitchell here at Duke. The excitement in the field, combined with the absolutely devastating results of the tumor itself, has also attracted the attention of a number of philanthropies and concerned citizens.